Full Oral

The 12
th
ASEAN FOOD CONFERENCE 2011
16 -18 June, 2011
BITEC Bangna, Bangkok, Thailand

78

OA : Functional Food Innovation
OB : Food Productivity Improvement
OC : Malaysia Palm Oil Symposium
OD : Food Safety and Quality Management
OE : Driving Trends in Aquatic Food Market
OG : Innovative Fermented Foods and Functional Ingredients
OH : Novel Food Processing & Packaging
OJ : Trends in Food Research: ASEAN Perspective
(Graduate Papers)

The 12
th
16 -18 June, 2011

79

OA : Functional Food Innovation
The 12
th
16 -18 June, 2011

80

OA-61

Fermentation of Tropical Fruit Juices by Lactic Acid Bacteria

Lin Kiat Saw
1,*
, Suijing Chen
2
, Siew Hwa Wong
2
, Soon Ann Tan
1
, Kelvin K.T Goh
2

1
School of Chemical & Life Sciences, Singapore Polytechnic, 500 Dover Road, Singapore 139651
2
Massey University, Institute of Food, Nutrition and Human Health, Private Bag 11 222, Palmerston North,
New Zealand
*
Corresponding author: LKSaw@sp.edu.sg

Abstract

The majority of products containing probiotics are dairy-based, which include yogurt and
fermented milk beverage. In the last decade, there is an increasing interest in using non-
dairy ingredients as substrates for certain strains of lactic acid bacteria to deliver the
physiological benefits of probiotics to wider group of consumers. This research aimed to
explore the use of tropical fruit juices as substrates for lactic acid bacteria fermentation.
The fruit juices studied were watermelon, honeydew melon, rockmelon, China Pear and
dragon fruits (red and white flesh). Three lactic acid bacteria strains, Lactobacillus
acidophilus, Lactobacillus casei and Lactobacillus delbrueckii subsp bulgaricus were used
in this study. The greatest decrease in pH was observed in fermentation using L. casei,
followed by L. acidophilus and L. delbrueckii subsp bulgaricus. Rockmelon, watermelon
and honeydew melon were identified to be suitable substrates, especially for L. casei. The
growth profile of L. casei was further evaluated in honeydew melon juice fermentation at
37
o
C over 48-hour. The maximum cell count achieved was approximately 10
9
CFU/mL.
Glucose was utilized and corresponded with the lactic acid production. However, fructose
was not utilized during the fermentation. This study demonstrated the potential of
producing value-added non-dairy probiotic beverages from certain tropical fruits.

Keywords: Lactic acid bacteria, probiotics, prebiotic, non-dairy beverages, tropical fruits

Introduction
Probiotics has been a subject of interest in the last few decades in food research,
particularly in the area of functional food. Probiotic foods and beverages are manufactured
by either method: (a) by adding the probiotic strains simultaneously with the standard
cultures in the fermentation tank; (b) by adding the probiotic culture directly into non-
fermented final products (Saxelin, 2008). Generally, species of Lactobacillus and
Bifidobacterium are used in most of the probiotic applications (Parvez and others 2006).
Presently, the majority of products containing probiotics are dairy-based, which include
yogurt and fermented milk beverage. However, due to some drawbacks related to dairy
products, there are emerging interests in using non-dairy ingredients as substrates for
delivering the physiological benefits of probiotics to wider group of consumers (Prado and
others 2008; Rivera-Espinoza and Gallardo-Navarro 2010).

Non-dairy substrates that have been used for lactic acid bacteria (LAB) fermentation
include soy protein and cereals. In recent years, several studies have reported the use of
fruits and vegetables juices as base medium for LAB fermentation. Juices from these
sources are deemed to be advantageous because of their low allergenicity, perceived health
benefits and appeal to a wide segment of the population. Sheehan and others (2007)
showed that different probiotic cultures, added to orange and pineapple juices, varied in
The 12
th
16 -18 June, 2011

81

their ability to tolerate low pH (~ 3.5) and their survival during storage at low temperatures
(~ 4
o
C). However, majority of the probiotic bacteria were killed if the fermented juices
were subjected to thermal or high pressure pasteurisation treatments. Amongst the cultures
studied, L. casei, L. rhamnosus and L. paracasei displayed good survival in orange and
pineapple juice, as compared to cranberry juice. The cultures studied survived at levels of
above 10
6
CFU/ml for at least 12 weeks when the bacteria are added to shelf-stable orange
and pineapple juices without subjecting to further pasteurisation treatments. Yoon and
others (2005, 2006) reported that L. acidophilus and L. plantarum were able to grow well
in non-supplemented beetroot and cabbage juice to nearly 10
8
CFU/mL at 30 C after 48
hours of fermentation. Several Bifidobacterium strains (B. lactis and B. bifidum) have also
been used to ferment carrot juice successfully (Kun and others 2008).

The wide varieties of fruits and vegetables and the huge number of LAB strains provide
new challenges and opportunities for the development and commercialisation of value-
added non-dairy fermented probiotic beverages. The survival of probiotic strains depend
on factors, such as, nutrients, pH, temperature and the presence of inhibitors.
This study utilised several tropical fruit juices widely found in the South East Asia as
substrates for the fermentation of different strains of LAB. The growth profiles of these
cultures and the utilisation of sugars were also reported in this study.

Materials and Methods
Preparation of bacterial culture
The lactic acid bacteria used in this study were L. casei (4114, NCIMB), L. acidophillius
(FD DBS LA5, Chris Hansen) and L. delbrueckii subsp bulgarius (11778, NCIMB). The
lyophilized cultures were reactivated by rehydrating the cultures in Lactobacilli MRS broth
(Acumedia, Langsing, MI), followed by streaking on MRS agar (Acumedia, Langsing,
MI). The cultures were incubated at 37
o
C in a 5% CO
2
atmosphere (3110 Series Direct
Heat CO
2
incubator, ThermoForma) for 48 to 72 hours. The bacteria colonies were
subsequently transferred from the agar plate to cryobeads and stored at -70
o
C in 10%
glycerol solution. Gram staining was carried out to ensure no contamination had occurred.

Preparation of fermentation substrates
The fruits used in this study were watermelon (Citrullus lanatus), honeydew melon
(Cucumis melo var. inodorus), rockmelon (Cucumis melo var. cantalupensis), China Pear
(Pyrus pyrifolia), and white flesh dragon fruits (Hylocereus undatus) and red flesh dragon
fruit (Hylocereus costaricensis). These fruits were purchased from the local supermarkets
and stored at 4
o
C prior to use. The fruit juices were obtained using a juice extractor (MJ-
W171P, Panasonic). The juices were pre-filtered using a sieve and then filtered through a
cheese cloth before they were centrifuged at 3200 g for 15 minutes at 10
o
C. The
supernatants were collected and the pH of the juices was adjusted to 7 using 1M NaOH.
The pH-adjusted clear juices were frozen at -20
o
C prior to use. The pH and total soluble
solids (brix
o
) of the juices were measured using a pH meter (Metrohm 827 pH lab) and
digital refractometer (CDX-1, Vee Gee).

The 12
th
16 -18 June, 2011

82

Screening test
The screening involved the fermentation of each of the three bacteria cultures (L. casei,
L. acidophilius and L. delbrueckii subsp bulgarius) with each of the six juices mentioned
earlier. In total, 18 bacteria-juices fermentation mixtures were screened. The frozen juice
substrates described earlier were thawed at room temperature and pasteurized at 105
o
C for
15 minutes in an autoclave (HiClave HV-50, Hirayama). The pasteurized juices were left
to cool and stored at 4
o
C before inoculation. Prior to bacteria inoculation, the pasteurized
juices were allowed to equilibrate to the room temperature, holding at about 25
o
C.

To prepare the inoculums, a cryobead of each culture was transferred aseptically into 40 ml
of Lactobacilli MRS broth (Acumedia, Langsing, MI). The bacteria culture was incubated
at 37
o
C in the 5% CO
2
incubator for 18 hours before being inoculated into the pasteurized
juices.

All 18 bacteria-juices fermentations were carried out in duplicates. The fermentations were
conducted using 50ml centrifuge tubes, each containing 45g of pH-adjusted or non-pH
adjusted pasteurized juice. All samples were inoculated with 2 % (v/w) mother culture and
fermented at 37
o
C in a 5% CO
2
incubator for 48 hours. The pH measurements were
recorded at 0, 24 and 48 hours of the fermentation process. The samples which gave the
highest decrease in pH were considered for subsequent trials.

Growth profile of L. casei in Honeydew melon fermentation
Honeydew melon juice was selected to determine the growth profile of the L. casei during
fermentation. The fruit juice and inoculum were prepared according to the methods
described earlier. The pH adjusted and pasteurized honeydew melon juice (250g) was
fermented in 250ml Duran bottles. All samples were inoculated with 2 % (v/w) mother
culture that had been incubated for 22 hours. The capped bottles were fermented at 37
o
C in
the incubator. All fermentations were conducted in duplicates and samples were obtained
at 0, 4, 8, 18, 24, 32 and 48 hours of fermentation.

Analysis of fermented products
Bacteria enumeration was carried out by decimally diluting the samples with MRS broth
(Neogen, acumedia) and 1ml was plated using MRS agar (Neogen, acumedia) via pour-
plate method. The plates were incubated for 48 hours at 37
o
C in an anaerobic incubator
with 5% CO
2
(ThermoForma, 3110 series). Glucose, fructose and lactic acid were analyzed
using High Performance Liquid Chromatography (HPLC). Samples collected were
centrifuged at 4 000 rpm for 10 minutes to remove the bacteria cells. The supernatant
were diluted 1:1 with ultra pure water and vortexed. This was followed by filtering the
diluted sample through a 0.22 m filter (MS Simplepure, membrane solutions) into a vial.
The HPLC system (LC-20AD, Shidmadzu) was fitted with Aminex HPX-87H (Bio-Rad)
column of 300 x 7.8 mm. The column temperature was maintained at 60
o
C throughout the
separation. The solvent used was 0.005M H
2
SO
4
with 10% Acetonitrile. The sample was
separated at a flow rate of 0.6 ml/min.

A refractive index detector (RID-10A, Shimadzu) was used for the identification of
glucose and fructose. A diode array detector (SPD-M20A, Shimadzu) was used to detect
lactic acid. Calibration standards were prepared using glucose and lactic acid (Sigma-
Aldrich), and fructose (Scharlau, Chemie) to obtain calibration curves of R
2
=0.999.
The 12
th
16 -18 June, 2011

83

Results and Discussions
Screening test

Figure 1 shows the changes in pH during the fermentation of the fruit juices by the three
LAB strains, L. casei, L. acidophilius and L. delbrueckii subsp bulgarius.

The figure shows that samples with pH adjustment had a relatively large decrease in pH
during the first 24 hours of fermentation. The pH values only decreased slightly when the
fermentation was continued for another 24 hours. The initial pH of all juices was adjusted
to more than 6 to avoid the inhibition of bacteria growth caused by the differences in
natural acidities of different fruits. Bacteria growth is known to be retarded in an acidic
environment, typically at pH lower than 4.5.

The amount of soluble solids (degree brix) in the fruit juices were measured using a
refractometer. The results showed degree brix of the juices in descending order were
China Pear (13.5%) > red flesh dragon fruit (12.6%) > white flesh dragon fruit (12.5%)>
watermelon (8.8%) > rockmelon (6.2%) > honeydew melon (5.8%). We noted that the pH
decrease was more drastic for juices with lower soluble solids (6-9%) compared to those
with higher level of soluble solids (12.5-13.5%). It is possible that juices containing higher
degree brix and the pH conditions could retard the growth of the LAB cultures. However,
further work is required to ascertain this observation.

Table 1 shows, in descending order, the change in pH values of the fruit juice samples
(with pH adjustment) in combination with the LAB strains. The change in pH is based on
the difference between the final pH value (at 48 hours fermentation) and the initial pH
value (at 0 hour fermentation).

Watermelon, rockmelon and honeydew melon juices showed relatively higher decrease in
pH (>2.7) than dragon fruits and China Pear juices (<2.2). Watermelon, rockmelon and
honeydew melon showed pH values between of 3.5-3.9 at 48 hour of incubation. The
values are close to the pH of fermented milk suggesting that the three juices (watermelon,
rockmelon and honeydew melon) had provided the necessary nutrients for growth and acid
production of the LAB strains. At this low pH range, further growth and acid production
of most LAB cultures would be retarded.

On the other hand, the fermentation of white dragonfruit, red dragonfruit or China Pear
juices resulted in pH decrease of lower than 2.2 at the end of 48 hour fermentation. The
final pH values of these samples were above pH 4, suggesting that these juices was
relatively less conducive for the growth and acid production of the LAB cultures.
Informal sensory evaluation of the odour of the fermented juices was carried out. We
noted that the odours of different fruits were unique and somewhat pungent except for
honeydew melon which was rather sweet smelling. Based on the above results, the growth
profile of L. casei in honeydew melon was studied.

Growth profile of L. casei in honeydew melon juice fermentation
The growth profile of L. casei was studied based on the fermentation of honeydew melon.
Figure 2 shows the total plate count and pH development over 48 hours. From the results,
the total plate count increased exponentially from 10
7
to around 10
9
CFU/ml during the
The 12
th
16 -18 June, 2011

84

first 8 hours of fermentation before entering the stationary phase. Death phase was not
observed for the L. casei throughout the 48 hour fermentation of honeydew melon juice.
Rapid decrease in pH was also observed during the first 8 hours of fermentation, where pH
value drops from 6.3 to about 4.9. This was followed by an intermediate decrease in pH
until approximately 24 hours of fermentation, when pH reaches 3.4. There was very little
decrease in pH from 24 to 48 hours of fermentation, with the pH reaches about 3.2 at the
end of the experiment.

The increase in viable cell count corresponds to the decrease in pH during the first 8 hours
of fermentation. The pH value at the end of the fermentation period was 3.2. The L. casei
appeared to be relatively acid tolerant since the total count was maintained throughout,
indicating that this strain may be suitable for use in fruit juice fermentation.

(a) L. casei

(b) L. acidophilius

(c) L. delbrueckii subsp bulgarius

Figure 1 pH values of fruit juices at 0, 24 and 48

hour of fermentation by different lactic acid bacteria.

The 12
th
16 -18 June, 2011

85

Table 1 Changes in pH for all 18 samples screened. The samples are ranked in descending order based
on the absolute change in pH.

Rank Bacteria Fruits
Initial pH
(at 0 hour)
Final pH
(at 48 hour)
pH change
(0-48 HR)
1 L. acidophillius Watermelon 6.74 0.11 3.67 0.04 3.08 0.15
2 L. delbrueckii Watermelon 6.80 0.08 3.81 0.08 2.99 0.07
3 L. acidophillius Honeydew melon 6.80 0.02 3.81 0.04 2.99 0.06
4 L. casei Watermelon 6.64 0.10 3.70 0.01 2.95 0.09
5 L. casei Rockmelon 6.36 0.03 3.46 0.01 2.91 0.04
6 L. casei Honeydew melon 6.55 0.04 3.68 0.01 2.88 0.04
7 L. delbrueckii Honeydew melon 6.73 0.02 3.92 0.08 2.81 0.06
8 L. delbrueckii Rockmelon 6.58 0.01 3.80 0.08 2.78 0.01
9 L. acidophillius Rockmelon 6.56 0.02 3.85 0.09 2.71 0.06
10 L. acidophillius Red dragonfruit 6.33 0.05 4.20 0.01 2.13 0.04
11 L. casei Red dragonfruit 6.17 0.01 4.10 0.09 2.07 0.08
12 L. casei White dragonfruit 6.14 0.01 4.19 0.01 1.96 0.02
13 L. delbrueckii White dragonfruit 6.20 0.00 4.96 0.60 1.25 0.60
14 L. casei China Pear 6.25 0.03 4.71 0.17 1.54 0.20
15 L. delbrueckii Red dragonfruit 6.23 0.01 4.92 0.04 1.31 0.15
16 L. acidophillius White dragonfruit 6.01 0.01 4.56 0.01 1.45 0.00
17 L. delbrueckii China Pear 6.33 0.04 5.34 0.15 1.00 0.19
18 L. acidophillius China Pear 6.40 0.03 5.85 0.04 0.56 0.01

Figure 2 Changes in total plate count for L. casei and pH development over 48 hours fermentation of
pH adjusted honeydew melon juice.

Figure 3 shows the plot of lactic acid production and pH reduction over 48 hours of
fermentation at aerobic and anaerobic condition. As the lactic acid production increased,
the pH of the medium decreased. 14.3 g/l and 13.7g/l of lactic acid were produced
respectively for aerobic and anaerobic condition of fermentation.

The 12
th
16 -18 June, 2011

86

Figure 3 Concentration of Lactic acid, glucose and fructose of L. casei-Honeydew melon fermentation.

In theory, for homo-fermentation, lactic acid is the major product of the fermentation
whereby one mole of glucose consumed will results in two moles of lactic acid. As for
hetero-fermentation, one mole of glucose will results in one mole of lactic acid together
with other end products such as acetic acid and carbon dioxide (Axelsson, 2004).
According to Alxelsson (2004), L. casei was classified as facultatively heterofermenters
while Walstra and his co-workers (2006) had classified L. casei as homofermenters, with
exceptions of some strains that are facultatively heterofermentative. The amount glucose
consumed is shown in figure 3.

Figure 3 shows the plot of the utilization of glucose and fructose with the total plate count
over 48 hours for L. casei. The results showed that only glucose was utilized but the
amount of fructose remained relatively constant. The lactic acid produced correlate closely
(R
2
value = 0.9986) with the amount of glucose utilized. Approximately 2.6 mole/L of
lactic acid were produced with 1 mole/L of glucose consumed. This value is close to the
theoretical value, in which 1 mole of glucose yields 2 moles of lactic acid, suggesting that
the strain of L. casei used in this study belongs to the homofermenters. It has to be noted
that the above preliminary results only showed the main acid and sugars and did not take
into considerations other compounds, such as, proteins, oligosaccharides, minerals,
vitamins, that are present in the fruit juice.

It is interesting to note that since fructose was not utilized, the sweetness of the juice could
be retained. As expected, the rate of glucose utilization was rapid when L. casei was at its
exponential growth phase. The rate of utilization slowed down when the bacteria entered
its stationary phase. However, there was still relatively high amount of glucose that
remained in the juice (~24g/L) after 48 hour fermentation. This could indicate that the
carbon source was not the limiting factor that retarded the growth of the bacteria. Instead,
pH appeared to be the limiting factor that retarded L. casei fermentation. The stationary
phase began when the pH of the juice was around pH 3.9. In order to increase the viable
The 12
th
16 -18 June, 2011

87

cell count, pH adjustment throughout the fermentation may be necessary to optimize the
cell count in a semi-batch fermentation.

Further work is underway to evaluate the flavour profile of the fermented fruit juice since
both heat treatment and the volatiles produced by the LAB culture during fermentation
would have altered the flavour profile of the fruit juice. This study shows the potential of
producing value-added non-dairy probiotic beverages from certain tropical fruits.

Acknowledgements
The author would like to acknowledge the support of Toteboard, Singapore under the
Toteboard Model project grant M121.

References
Axelsson, L. 2004. Lactac acid bacteria: Classification and physiology. In: Salminen, S,
Wright, AV, Ouwehand, A, editors. Lactic acid bacteria: Microbiological and
functional aspects. 3
rd
ed. New York: Marcel Dekker Inc.
Kun, S, Rezessy-Szabo, JM, Nguyen, QD, Hoschke, A. 2008. Changes of microbial
population and some components in carrot juice during fermentation with selected
Bifidobacterium strains. Process Biochemistry 43(8):816-821.
Parvez, S, Malik, KA, Kang SA, Kim H-Y. 2006. Probiotics and their fermented food
products are beneficial for health. J Appl Microbiology 100(6):1171-1185.
Prado, FC, Parada, JL, Pandey, A, Soccol, CR. 2008. Trends in non-dairy probiotic
beverages. Food Research International 41(2):111-123.
Rivera-Espinoza, Y, Gallardo-Navarro, Y. 2010. Non-dairy probiotic products. Food
Microbiology 27(1):1-11.
Saxelin, M. 2008. Probiotic formulations and applications, the current probiotics market,
and changes in the marketplace: A European perspective. Clinical Infectious
Diseases 46 (SUPPL. 2).
Sheehan,VM, Ross P, Fitzgerald GF. 2007. Assessing the acid tolerance and the
technological robustness of probiotic ultures for fortification in fruit juices.
Innovative Food science and Emerging Technologies 8(2):279-284.
Walstra, P, Wouters, J, Genurts, T. 2006. Dairy science and technology. 2
nd
ed. Boca
Raton: CRC Press.
Yoon, KY, Woodams, EE, Hang YD. 2005. Fermentation of beet juice by beneficial lactic
acid bacteria. Lebensmittel-Wissenschaft und Technologie 38(1):73-75.
Yoon, KY, Woodams, EE, Hang, YD. 2006. Production of probiotic cabbage juice by
lactic acid bacteria. Bioresource Technology 97(12):1427-1430.

The 12
th
16 -18 June, 2011

88

OA-91

Potential Probiotic Evaluation of Bacteriocin Producing Lactobacillus
Plantarum ST16PA Isolated from Papaya (Carica Papaya)

Manuela Vaz-Velho
1
, and Svetoslav Dimitrov Todorov
2,*

1
Escola Superior de Tecnologia e Gesto, Instituto Politcnico de Viana do Castelo, Portugal
2
Universidade de So Paulo, Faculdade de Cincias Farmacuticas, Departamento de Alimentos e Nutrio
Experimental, Laboratrio de Microbiologia de Alimentos, So Paulo, SP, Brasil
*
Corresponding author: slavi310570@abv.bg

Abstract

The World Health Organization defines probiotics as live microorganisms which, when
administrated in adequate amounts, confer health benefits to the host. The aim of this work
was to evaluate the potential probiotic features of the bacteriocin producing Lactobacillus
plantarum isolated from papaya. L. plantarum ST16Pa survived conditions simulating the
GIT and produced bacteriocins active against a number of L. monocytogenes. Good growth
of L. plantarum ST16Pa was recorded in MRS broth with initial pH values between 5.0
and 9.0 and this strain was able to survive in pH 4.0, 11.0 and 13.0. L. plantarum ST16Pa
grew well in the presence of oxbile at concentrations ranging from 0.2% to 3.0%. L.
plantarum ST16Pa was determined to have an auto-aggregation level of 37.05%. Various
degrees of co-aggregation were observed with different strains of L. plantarum,
Enterococcus, L. sakei and Listeria spp... Low levels of co-aggregation of L. plantarum
with the Listeria strains may be important in preventing both the formation of Listeria
biofilms and its persistence in the GIT. Adherence of the bacteria to Caco-2 cells was
within the range reported for other probiotic strains. Hydrophobicity values for L.
plantarum ST16Pa were also determined and the presence of adhesion genes (mub, mapA
and EF-Tu) was detected by PCR analysis. Although these properties are all characteristic
of a good probiotic, further in vivo studies will have to be performed to demonstrate that
the strain is active in the GIT of animals. To the best of our knowledge, this is the first
report on potential probiotics properties of L. plantarum isolated from papaya.

Keywords: Lactobacillus plantarum, papaya, probiotic, encapsulation, bacteriocin

Introduction
In addition to important technological properties in food manufacturing, such as production
of lactic acid, reduction of lactose content and improvement of organoleptic and physical
characteristics, several lactic acid bacteria (LAB) are used in foods as probiotics.
Probiotics are defined as live microorganisms which when administered in adequate
amounts (10
6
-10
7
CFU per gram of food) confer health benefits to the host (FAO/WHO,
2001; Bertazzoni-Minelli and others 2004; von Mollendorff, 2008). Probiotic benefits
include suppression of growth of pathogens, control of serum cholesterol level, modulation
of the immune system, improvement of lactose digestion, synthesis of vitamins, increase in
bio-availability of minerals and possible anti-carcinogenic activity (Chan and Zhang,
2005). Beneficial effects depend on the ability probiotic strains have to survive the natural
defenses of the host and to multiply in the gastrointestinal tract (GIT) (Todorov and others
2008). Besides these effects, the production of antimicrobial compounds, such as
The 12
th
16 -18 June, 2011

89

bacteriocins, by probiotic LAB, may contribute to colonization of the GIT, as they give
them competitive advantage over other bacteria (Garriga and others 1993; Todorov, 2009).

This study evaluated the probiotic potential of a bacteriocinogenic Lactobacillus plantarum
strain (Lb. plantarum ST16Pa) isolated from papaya fruit (Todorov and others 2011) and
studied the effect of its encapsulation in alginate on survival in conditions simulating the
human GIT. These features are important when selecting a starter culture for food
biopreservation as potential beneficial properties to the consumer health should always be
taken into account.

Materials and methods
Strains: The study was conducted with Lactobacillus plantarum ST16Pa, a
bacteriocinogenic strain isolated from papaya fruit (Todorov and others 2011). Other
strains used in the study were L. rhamnosus GG, L. plantarum ST202Ch, L. plantarum
ST69BZ, Enterococcus faecium ST62BZ, Enterococcus faecalis ATCC 19443, L. sakei
subsp. sakei 2a, Listeria ivanovii subsp. ivanovii ATCC 19119 and Listeria innocua
2030C. Pure cultures were stored at -20
o
C with 20% glycerol.

Effect of pH and oxbile on growth of L. plantarum ST16Pa: This experiment was
performed according to Todorov and others (2008). L. plantarum ST16Pa was grown in
MRS broth (Difco) adjusted to pH 3.0, 4.0, 5.0, 6.0, 7.0, 9.0, 11.0 and 13.0. Resistance to
oxbile salts was tested by growing the strain in MRS broth (Difco) containing 0.2, 0.4, 0.6,
0.8, 1.0, 2.0 and 3.0 % (w/v) oxbile (Sigma). All tests were conducted in sterile flat-bottom
96-well micro titer plates (TPP, Switzerland). Each well was filled with 180 L of tested
MRS broth and 20 L of the culture presenting an OD
650nm
adjusted to 0.2. Optical density
readings at 600 nm were recorded every hour for 12 h.

Adhesion of L. plantarum ST16Pa to Caco-2 cells: The experiment was performed
according to Todorov and others (2008). Experiments were done in triplicate. L.
rhamnosus GG was used as control of adhesion to Caco-2 cells.

Identification of genes encoding MapA and Mub adhesion proteins and EF-Tu
elongation factor in L. plantarum ST16Pa: DNA from L. plantarum ST16Pa was isolated
according to Dellaglio and others (1973). Primers Mub423F (5-GTA GTT ACT CAG
TGA CGA TCA ATG-3), Mub423R (5-TAA TTG TAA AGG TAT AAT CGG AGG-
3), Map423F (5-TGG ATT CTG CTT GAG GTA AG -3), Map423R (5-GAC TAG
TAA TAA CGC GAC CG-3), EFTu423F (5-TTC TGG TCG TAT CGA TCG TG-3)
and EFTu423R (5-CCA CGT AAT AAC GCA CCA AC-3) were used for amplification
of genes MapA, Mub and EF-Tu by PCR, as described by Todorov and Dicks (2008).

Determination of cell surface hydrophobicity in L. plantarum ST16Pa: Cell surface
hydrophobicity was measured as described by Doyle and Rosenberg (1995), and modified
by Todorov and others (2008). In summary, L. plantarum ST16Pa was grown in MRS
broth at 37C for 18 h. Cells were harvested (6 700 g, 4C, 6 min), washed twice with
sterile saline solution (pH 6.5), re-suspended in the same solution and the optical density
(OD
580nm
)

determined. The cell suspension (1.5 mL) was mixed with equal volume of n-
hexadecane (Sigma) and vortexed for 2 min. After 30 min at room temperature for
separation of the two phases, one ml of the aqueous phase was removed and the optical
The 12
th
16 -18 June, 2011

90

density (OD
580nm
) determined. The experiment was repeated and the average optical
density value determined. The percentage hydrophobicity was calculated as follows: %
hydrophobicity = [(OD
0
OD
30
)/OD
0
] x 100, where OD
0
and OD
30
refer to the initial OD
and OD measured after 30 min, respectively. Experiments were conducted in triplicate.
Aggregation properties of L. plantarum ST16Pa: The experiment was performed
according to Basson and others (2007) with modifications (Todorov and others 2008). L.
plantarum ST16Pa was grown in MRS broth (Difco) for 24 h at 37
o
C, harvested (7 000 x
g, 10 min, 20
o
C), washed, re-suspended in sterile saline solution (pH 6.5) and diluted to
OD
660nm
= 0.3. One mL of the cell suspension was transferred to a 2 mL sterile plastic
cuvette and the OD
660nm
recorded over 60 min using a spectrophotometer. After 60 min,
the cell suspension was centrifuged (300 x g, 2 min, 20
o
C) and the OD
60
of the supernatant
determined. Auto-aggregation was determined using the following equation: % Auto-
aggregation = [(OD
0
OD
60
)/OD
0
] x 100, where OD
0
and OD
60
refer to the initial OD
and to the OD measured after 60 min, respectively.

For evaluation of co-aggregation of L. plantarum ST16Pa with L. plantarum ST202Ch, L.
plantarum ST69BZ, E. faecium ST62BZ, E. faecalis ATCC 19443, L. sakei subsp. sakei
2a, L. ivanovii subsp. ivanovii ATCC 19119 and L. innocua 2030C, the lactobacilli were
grown in MRS broth for 24 h at 37
o
C and the other cultures were prepared in BHI for 24 h
at 37
o
C. The cells were harvested (7 000 x g, 10 min, 20
o
C), washed, re-suspended in
sterile saline solution (pH 6.5) and diluted to OD
660nm
= 0.3. One mL of the L. plantarum
ST16Pa culture and one mL of each test cell suspension were transferred to a 2 mL sterile
plastic cuvette and the OD
660nm
recorded over 60 min using a spectrophotometer. After 60
min the mixture was centrifuged (300 x g, 2 min, 20
o
C) and the OD
60
of the supernatant
determined. Co-aggregation of L. plantarum ST16Pa with the test cultures was calculated
using the same equation used for calculation of auto-aggregation.

Results and Discussion
A previous study indicated that L. plantarum ST16Pa produces a large spectrum
bacteriocin capable of inhibiting many food spoilage bacteria and foodborne pathogens,
including Listeria monocytogenes, a psychrotrophic foodborne pathogen of increasing
importance (Todorov and others 2011). Further research evaluating potential probiotic
characteristics of this strain is important as this additional beneficial property can lead to a
broader application of the strain in foods, for improving quality and safety of food products
as well as for conferring beneficial health effects to the consumer.
Capability to grow and survive to low pH and high concentrations of oxbile salts, as those
encountered in the human GIT, is a major characteristic of a strain with potential probiotic
activity (Havenaar and others 1992; Carvalho and others 2009). Results reported in this
study indicate that L. plantarum ST16Pa grew well in MRS broth (Difco) with initial pH
values of 5.0, 6.0, 7.0 and 9.0 (Fig. 1), and that at pH 4.0 the strains still survived well but
the growth rate was lower. No growth was detected at initial pH of 3.0. At pH 11.0 and
13.0, lower growth rates than at pH 6.0 were recorded. Previous studies indicate that pH
affects the growth of bacteriocinogenic Lactobacillus spp. at different levels. Growth of
several strains of L. plantarum, L. rhamnosus, L. pentosus and L. paracasei was
suppressed at pH 3.0 and 4.0, and variable results were recorded for pH of 11.0 and 13.0,
but poor growth recorded for strains L. paracasei ST242BZ and ST284BZ, L. rhamnosus
ST462BZ, L. plantarum ST664BZ and L. pentosus ST712BZ at pH 13.0 (Todorov and
others 2008).
The 12
th
16 -18 June, 2011

91

Lactobacillus plantarum ST16Pa grew well in the presence of oxbile at concentrations
ranging from 0.2% to 3.0% (Fig. 1), suggesting that this strain would be able to survive in
the human GIT. Bile salts have different effects in different Lactobacillus strains. Growth
of strains L. plantarum ST194BZ, ST414BZ and ST664BZ, L. rhamnosus ST462BZ, and
L. pentosus ST712BZ were less affected by the presence of 0.3% bile, compared to strains
L. paracasei ST242BZ and ST284BZ and L. rhamnosus ST461BZ (Todorov and others
2008). Bile at 0.6% (w/v) suppressed the growth of L. plantarum ST194BZ and ST441BZ,
L. paracasei ST242BZ and ST284BZ and L. rhamnosus ST462BZ, but not L. rhamosus
ST461BZ, L. plantarum ST664BZ and L. pentosus ST712BZ. Bile levels above 0.6%
(w/v) suppressed the growth of all tested strains (L. plantarum ST194BZ, ST441BZ and
ST664BZ, L. paracasei ST242BZ and ST284BZ, L. rhamnosus ST461BZ and ST462BZ
L. pentosus ST712BZ) (Todorov and others 2008).

Fig 1 Growth of Lactobacillus plantarum ST16Pa in MRS broth (Difco) supplemented with ox-bile and
at different pH levels. Each result represents an average of three independent experiments

Adhesion of probiotic strains to intestinal cells such as Caco-2 cells is believed to be a
critical factor for increasing the possibility of colonizing the GIT and of surviving in this
hostile environment (Tuomola and Salminen, 1998). L. plantarum ST16Pa adhered to
Caco-2 cells at a rate similar to that recorded for the world known probiotic reference
strain L. rhamnosus GG (9.5% and 11.3%, respectively) (data not shown). These values are
similar to those reported by Todorov and others (2008) for several bacteriocins producing
LAB isolated from boza (0.26% to 9.0%) and by Tuomola and Salminen (1998) for some
probiotic and dairy Lactobacillus strains (3.2% to 14.4%). Bertazzoni-Minelli and others
(2004) reported much lower adhesion rates for probiotic Lactobacillus casei strains (0.08%
to 0.74%).

The expression of mucus adhesion proteins, such as those encoded by the mub and mapA
genes, and of GTP-binding EF-Tu protein has shown to be critical in adhesion of probiotic
strains to human intestine cells (Ramiah and others 2008). Results of PCR amplification
using specific primers and sequence analysis of the PCR amplicons indicated that L.
plantarum ST16Pa contains the mub, mapA and EF-Tu genes (Fig. 2). The presence of
these adhesion genes in this L. plantarum strain is not surprising due to the high adhesive
properties of this microbial species. Whole genome sequence data of L. plantarum
WCFS1, a human isolate, revealed the presence of 223 extracellular proteins, most of them
involved in the adhesion of the cell to its environment (Kleerebezem and others 2003). The
domain composition of L. plantarum WCFS1 proteins, predicted to be associated with
The 12
th
16 -18 June, 2011

92

adhesion, indicated the presence of seven proteins to be involved with mucus. Some of
these possible mucus binding proteins contain the Mub domain, a trait unique to lactic acid
bacteria (Boekhorst and others 2006). The presence of these adhesive proteins in lactic
acid bacteria is imperative for probiotic strains to colonize the GIT.

Bacterial cells with a high hydrophobicity usually form strong interactions with mucosal
cells. The interactions between microbial and host cells are non-specific, but in probiotic
strains, there is a good correlation between surface hydrophobicity and the ability to adhere
to the intestinal mucosa (Wadstrm and others 1987). The initial interaction may be weak;
it is often reversible and precedes subsequent adhesion processes mediated by more
specific mechanisms involving cell-surface proteins and lipoteichoic acids (Roos and
Jonsson, 2002). Hydrophobicity varies among even genetically closely related species and
among strains of the same species (Schar-Zammaretti and Ubink, 2003). Although
hydrophobicity may assist in adhesion, it is not a prerequisite for strong adhesion to human
intestinal cells. The hydrophobicity recorded for L. plantarum ST16Pa was higher than that
recorded for the reference probiotic L. rhamnosus GG strain (68.7% and 53.3%,
respectively). Similar results were observed by Todorov and others 2008, for strains L.
rhamnosus ST461BZ, L. rhamnosus ST462BZ and L. plantarum ST664BZ (75% to 80%)
and L. rhamnosus GG (55%).

Fig. 2 Agarose gels showing DNA fragments characteristic for genes MapA (lanes 1, 2 and
3), Mub (lanes 4, 5 and 6) and EF-Tu (lanes 7, 8 and 9) amplified by PCR. Lanes 2, 5
and 8 correspond to Lactobacillus plantarum ST16Pa and lanes 1, 4 and 7 to
Enterococcus faecium ST88Ch. Lanes 3, 6 and 9 contain no DNA and lane M
corresponds to 100bp molecular weight marker.

Aggregation is an important feature for biofilm formation. L. plantarum has a number of
genes encoding for surface proteins responsible for recognition or binding of components
present in the environment. Several of these genes are homologous to proteins with
predicted functions, such as mucus-binding, aggregation-promotion and intracellular
adhesion (Kleerebezem and others 2003). Auto-aggregation is a strain-specific trait
(Todorov and Dicks, 2008).

L. plantarum ST16Pa was determined to have an auto-aggregation level of 37.05% (Fig.
3), which is lower than the levels reported by Todorov and others (2008) for other
lactobacilli: L. pentosus ST712BZ (67%) and L. paracasei ST284BZ (99%). The co-
aggregation of L. plantarum ST16Pa with L. plantarum ST202Ch, L. plantarum ST69BZ,
M 1 2 3 4 5 6 7 8 9 M 1 2 3 4 5 6 7 8 9
The 12
th
16 -18 June, 2011

93

E. faecium ST62BZ, E. faecalis ATCC 19443, L. sakei subsp. sakei 2a, L. ivanovii subsp.
ivanovii

Fig.3 Auto-aggregation and co-aggregation of Lactobacillus plantarum ST16Pa with Lactobacillus
plantarum ST202Ch, Lactobacillus plantarum ST69BZ, Enterococcus faecium ST62BZ,
Enterococcus faecalis ATCC 19443, Lactobacillus sakei subsp. sakei 2a, Listeria ivanovii subsp.
ivanovii ATCC 19119 and Listeria innocua 2030c. Each result represents an average of three
experiments.

ATCC 19119 and L. innocua 2030C also varied according to the strain (Fig. 3), but the
lowest levels were observed for Listeria and Enterococcus strains. Low levels of co-
aggregation may play an important role in preventing the formation of biofilms, and thus
preventing the persistence of pathogenic species in the GIT.
To our knowledge, this is the first report of a bacteriocinogenic LAB isolated from fruit
that presents relevant application to food biopreservation and may be beneficial to
consumer health due to its potential probiotic characteristics.

Acknowledgments
The authors would like to thank Coordenao de Aperfeioamento de Pessoal de Nvel
Superior (CAPES, Brasil), Conselho Nacional de Desenvolvimento Cientfico e
Tecnolgico (CNPq, Brasil).

References
Basson, A., Flemming, L.A., & Chenia, H.Y. (2008). Evaluation of adherence,
hydrophobicity, aggregation characteristics and biofilm development of
Flavobacterium johnsoniae-like isolates. Microbial Ecology, 55, 1-14.
Bertazzoni-Minelli, E., Benini, A., Marzotto, M., Sbarbati, A., Ruzzenente, O., Ferrario,
R., Hendriks, H., & Dellaglio, F. (2004). Assessment of novel probiotic
Lactobacillus casei strains for the production of functional foods. International
Dairy Journal, 14, 723-736.
Boekhorst, J., Wells, M., Kleerebezem, M., & Siezen, R.J. (2006). The predicted
secretome of Lactobacillus plantarum WCFS1 sheds light on interactions with its
environment. Microbiology, 152, 3175-3183.
Carvalho, K.G., Kruger, M.F., Furtado, D.N., Todorov, S.D., & Franco, B.D.G.M. (2009).
Evaluation of the role of environmental factors in the human gastrointestinal tract
on the behaviour of probiotic cultures Lactobacillus casei Shirota and Lactobacillus
casei LC01 by the use of a semi-dynamic in vitro model. Annals of Microbiology,
59, 439-445.
0
10
20
30
40
50
60
S
T
1
6
P
a
/
S
T
1
6
P
a
S
T
1
6
P
a
/ S
T
2
0
2
C
h
S
T
1
6
P
a
/
S
T
6
2
B
Z
S
T
1
6
P
a
/
S
T
6
9
B
Z
S
T
1
6
P
a
/
A
T
C
C
1
9
4
4
3
S
T
1
6
P
a

/ 2
a
S
T
1
6
P
a

/ A
T
C
C
1
9
1
1
9
S
T
1
6
P
a

/ 2
0
3
0
C
S
T
2
0
2
C
h
/
S
T
2
0
2
C
h
S
T
6
2
B
Z
/
S
T
6
2
B
Z
S
T
6
9
B
Z
/
S
T
6
9
B
Z
A
T
C
C
1
9
4
4
3

/ A
T
C
C
1
9
4
4
3
2
a
/
2
a
A
T
C
C
1
9
1
1
9

/ A
T
C
C
1
9
1
1
9
2
0
3
0
C
/
2
0
3
0
C
%

a
g
g
r
e
g
a
t
i
o
n
0
10
20
30
40
50
60
S
T
1
6
P
a
/
S
T
1
6
P
a
S
T
1
6
P
a
/ S
T
2
0
2
C
h
S
T
1
6
P
a
/
S
T
6
2
B
Z
S
T
1
6
P
a
/
S
T
6
9
B
Z
S
T
1
6
P
a
/
A
T
C
C
1
9
4
4
3
S
T
1
6
P
a

/ 2
a
S
T
1
6
P
a

/ A
T
C
C
1
9
1
1
9
S
T
1
6
P
a

/ 2
0
3
0
C
S
T
2
0
2
C
h
/
S
T
2
0
2
C
h
S
T
6
2
B
Z
/
S
T
6
2
B
Z
S
T
6
9
B
Z
/
S
T
6
9
B
Z
A
T
C
C
1
9
4
4
3

/ A
T
C
C
1
9
4
4
3
2
a
/
2
a
A
T
C
C
1
9
1
1
9

/ A
T
C
C
1
9
1
1
9
2
0
3
0
C
/
2
0
3
0
C
%

a
g
g
r
e
g
a
t
i
o
n
The 12
th
16 -18 June, 2011

94

Chan, E.S., & Zhang, Z. (2005). Bioencapsulation by compression coating of probiotic
bacteria for their protection in an acidic medium. Process Biochemistry, 40, 3346-
3351.
Dellaglio, F., Bottazzi, V., & Troatelli, L.D. (1973). Deoxyribonucleic acid homology
and vase composition in some thermophylic lactobacilli. Journal of Geneneral
Microbiology, 74, 289-297.
Doyle, R.J., & Rosenberg, M. (1995). Measurement of microbial adhesion to hydrophobic
substrates. Methods in Enzymology, 253, 542-550.
FAO/WHO. (2001). Evaluation of health and nutritional properties of powder milk and
live lactic acid bacteria. In: Food and Agriculture Organization of the United
Nations and World Health Organization expert consultation report. FAO, Rome,
Italy.
Garriga, M., Hugas, M., Aymerich, T., & Monfort, J. M. (1993). Bacteriocinogenic activity
of Lactobacillli from fermented susages. Journal of Applied Bacteriology, 75, 142-
148.
Havenaar, R., Ten Brink, B., & Huis in't Veld, J.H.C. (1992). Selection of strains for
probiotic use. In Probiotics: The Scientific Basis ed. Fuller, R. London: Chapman
and Hall.
Kleerebezem, M., Boekhorst, J., van Kranenburg, R., Molenaar, D., Kuipers, O.P., Leer,
R., Tarchini, R., Peters, S.A., Sandbrink, H.M., Fiers, M.W.E.J., Stiekema, W.,
Lankhorst, R.M.K., Bron, P.A., Hoffer, S.M., Groot, M.N.N., Kerkhoven, R., de
Vries, M., Ursing, B., de Vos, W.M., & Siezen, R.J. (2003). Competitive genome
sequence of Lactobacillus plantarum WCFS1. Proceedings of National Academy of
Science USA, 100, 1990-1995.
Ramiah, K., Van Reenen, C.A., & Dicks, L.M.T. (2008). Surface-bound proteins of
Lactobacillus plantarum 423 that contribute to adhesion of Caco-2 cells and their
role in competitive exclusion and displacement of Clostridium sporogenes and
Enterococcus faecalis EF-Tu protein binds to GTP. Research in Microbiology, 159,
470-475.
Roos, S., & Jonsson, H. (2002). A high-molecular-mass cell-surface protein from
Lactobacillus reuteri 1063 adheres to mucus components. Microbiology, 148,
2481-2489.
Schar-Zammaretti, P., & Ubbink, J. (2003). The cell wall of lactic acid bacteria: surface
constituents and macromolecular conformations. Biophysics Journal, 85, 4076-
4092.
Todorov, S.D., & Dicks, L.M.T. (2008). Evaluation of lactic acid bacteria from kefir,
molasses and olive brine as possible probiotics based on physiological properties.
Annals of Microbiology, 58, 661-670.
Todorov, S.D., Prvost, H., Lebois, M., Dousset, J., LeBlanc, J. G., & Franco, B. D. G. M.
(2011) Bacteriocinogenic Lactobacillus plantarum ST16Pa isolated from papaya
(Carica papaya) from isolation to application: Characterization of a bacteriocin.
Food Research International in press
Todorov, S.D., Botes, M., Guigas, C., Schillinger U., Wiid I., Wachsman M.B., Holzapfel
W.H., & Dicks L.M.T. (2008). Boza, a natural source of probiotic lactic acid
bacteria. Journal of Applied Microbiology, 104, 465-477.
Todorov, S. D. (2009). Bacteriocins from Lactobacillus plantarum production, genetic
organization and mode of action. A review. Brazilian Journal of Microbiology, 40,
209-221.
The 12
th
16 -18 June, 2011

95

Tuomola, E.M., & Salminen, S. (1998). Adhesion of some probiotic and dairy
Lactobacillus strains to Caco-2 cell cultures. International Journal of Food
Microbiology, 41, 45- 51.
Von Mollendorff, J.W. (2008). Characterization of bacteriocins produced by lactic acid
bacteria from fermented beverages and optimization of starter cultures. McS
Thesis, Stellenbosch University, Stellenbopsch, South Africa.
Wadstrm, T., Andersson, K., Sydow, M., Axelsson, L., Lindgren, S., & Gullmar, B.
(1987). Surface properties of lactobacilli isolated from the small intestine of pigs.
Journal of Applied Bacteriology, 62, pp. 513520

The 12
th
16 -18 June, 2011

96

OA-97

Polyphenolic Antioxidants and Phytoalexins Changes in Germinating
Legume Seeds with Food Grade Fungal Rhizopus oligoporus Stress

Ziyun Wu, Lixia Song, Dejian Huang
*

Food Science and Technology Programme, Department of Chemistry, National University of Singapore, 3
Science Drive 3, Singapore 117543, Singapore
*
Corresponding author: chmhdj@nus.edu.sg

Abstract

This study aimed to investigate the phytochemical changes in food legume seeds under
ungermination, germination, and germination with fungus Rhizopus oligoporus stress.
Thirteen selected legumes are used in this study, including sword bean, big red bean,
black-eyed bean, brown bean, small red bean, white bean, green bean, broad bean, yellow
bean, black bean, big peanut, small peanut and chick pea. Five sets of experiments are
prepared: non-germinated seeds (UG), germinated seeds without fungal stress (G),
germinated seeds with fungal stress (GS), deactivated seeds without fungal stress (D) and
deactivated seeds with fungal stress (DS). The seeds or sprouts are extracted with
methanol/acetone/water (2:2:1), and compounds analyzed by reversed-phase high-
performance liquid chromatography (RP-HPLC) with photodiode array (PDA) detection,
liquid chromatographyelectrospray ionization mass spectrometry (LC-ESI-MS). Two
antioxidant-related polyphenolic compositions, total phenolic content (TPC) and total
flavonoids content (TFC), as well as antioxidant activity for oxygen radical absorbance
capacity (ORAC) are investigated. The results showed that germination could enhance the
generation of phytochemicals in most legumes and fungus-stressed germination can release
some novel phytoalexins in several legume sprouts, such as sword bean, soybeans and
peanuts. Remarkably, we found that the small peanuts synthesized the most number of
phytoalexins when the sprouts are stressed by the fungus. This study suggested that food
grade fungal stressed germination of legume seeds might be used as a processing method
to induce phytoalexins and enhance production of polyphenolic antioxidants, which are
proven to have many health enhancing benefits (e.g. antioxidant, anti-inflammation, anti-
cancer, anti-obesity, and anti-aging).

Keywords: Phytoalexin, antioxidants, legume seeds, germination, fungal stress

Introduction
Well-known grain legumes include beans, lentils, lupins, peas, and peanuts and are
cultivated for their seeds, and are also known as pulses. The seeds are used for human and
animal consumption or for production of oils for industrial uses. Legumes have high
nutritional value and play an important role in traditional diets throughout the world.
Recent studies have suggested that legumes especially soybean and peanuts could be as a
good functional food for health promotion (Francisco & Resurreccion 2008). Legume seed
sprouts are popular foods in globally. During germination, some seed components are
degraded and used for respiration and synthesis of new cell constituents for the
development, thus causing significant changes in the biochemical characteristics (Dueas
and others 2009). Numerous investigations have shown that germination has been
identified as an inexpensive and effective way to improve the nutritional quality of
legumes by increasing amino acids content, total dietary fibers, total soluble sugars, while
The 12
th
16 -18 June, 2011

97

reducing antinutrients levels such as -galactosides (Khattak and others 2008; Martn-
Cabrejas and others 2008; Arora and others 2010; Paucar-Menacho and others 2010).
Germination is also a convenient process to enhance polyphenolic contents and related
antioxidant activity (Cevallos-Casals & Cisneros-Zevallos 2009; Dueas and others 2009;
Shi and others 2010).

Plants possess both constitutive and inducible mechanisms to resist stress from wounds,
freezing, ultraviolet light, and microorganisms (e.g. oomycetes, fungi, bacteria, viruses,
and insects). In the past few years, a large number of reports have been published on
identification of phytoalexins, and effects of stress on gene expressions, transcription
factors, signaling pathways, metabolic pathways against both compatible and incompatible
plant-pathogen (van Loon and others 2006). Nevertheless, there is a paucity of literatures
on the changes of such treatments in nutritional values, antioxidant capacity as well as
phenolic composition. Phytoalexin concentration can be produced at much higher level
when plants responses to the stress (Lpez-Amors and others 2006). Their functions in
plant mainly include antimicrobe and antioxidant activities which are some of the
beneficial activities to human health and disease prevention (Hammerschmidt 1999; Boue
and others 2009). Resveratrol is a well-studied polyphenolic phytoalexin and has received
tremendous attention because of its broad range of health benefits in a variety of human
disease models (Pervaiz & Holme 2009). Logically, it is an emerging field of functional
food research by introducing phytoalexins through bioprocesses (Boue and others 2009).

Probiotics are food grade microorganisms, some of them used in starters such as Rhizopus
oligosporus for tempeh fermentation and Bacillus subtilis in natto fermentation. They are
attractive stresser to induce phytoalexins from legume seeds. Our previous work has shown
that food grade microbial-stressed germination of living soybeans leads to generation of a
group of oxooctadecadienoic acids and their glyceryl esters in addition to glyceollins, a
known phytoalexins present in wild and stressed soybeans (Feng and others 2007).
Furthermore the nutritional values of the soybean foods made from the bean seeds may be
particularly beneficial with higher content of total isoflavones (Feng and others 2008). To
expand the research into other legumes seeds, the present study is carried out with the aim
to evaluate the influence of R. oligosporus stressed germination and non-fungal stressed
germination process on phytochemicals and phytoalexins changes, phenolic and flavonoids
yields as well as the antioxidant capacity in 13 well known and used legume seeds. This
study will also provide comparative information for further identification of phytoalexins
in legumes.

Materials
Food grade fungus, Rhizopus oligosporous, was bought from PT. Aneka Fermentasi
Industri (Bandung, Indonesia). Sword bean (Canavalia gladiata (Jacq.) DC., Indonesia),
kidney bean (Phaseolus vulgaris L., China), Black-eyed pea (Vigna unguiculata subsp.
unguiculata, Myanmar), yardlong bean/cowpea, Vigna unguiculata subsp. sesquipedalis,
China), azuki bean (Vigna angularis (Willd.) Ohwi & H. Ohashi, China), hyacinth bean
(Lablab purpureus (L.) Sweet, India), mung bean (Vigna radiata (L.) R. Wilczek,
Thailand), broad bean (Vicia faba L., China), yellow soybean (Glycine max (L.) Merr.,
Canada), black soybean (Glycine max (L.) Merr., China), big peanut (Arachis hypogaea L.,
China), small peanut (Arachis hypogaea L., India) and chickpea (Cicer arietinum L.,
The 12
th
16 -18 June, 2011

98

Turkey) from supermarket in and the origins of production were obtained from Singapore
Trade Statistics.

Peanut Germination and Fungal Inoculations
The seeds germination and fungal inoculations were carried out according to the method
described previously (Feng and others 2007). In brief, legume seeds were allowed to
imbibe distilled water for 24 h at room temperature (Figure 2). The skins of the legume
seeds were peeled off afterwards without destroying radicals. The legumes seeds were
divided equally into five kinds, namely non-germinated (UG), germinated with/without
stress (GS/G), and deactivated seed with/without fungal stress (DS/D). The seeds that were
prepared for germination were put into petri dishes. The petri dishes were covered with
filter papers that were sprayed with distilled water or the fungal suspension. Those dishes
were placed at room temperature under dark condition and germinated for four days. The
deactivated seeds were prepared by putting them into oven at 150C for twenty minutes
and then germinated in the petri dishes for four days. Approximately two to four seeds of
each legume sample were collected and accurately weighed in a 15 mL screw-cap tube,
then extracted with 5 mL acetone/ethanol/water (2: 2: 1; v/v) mixture containing 0.1%
acetic acid on a shaking incubator at 200 rpm and room temperature for 12 h. The
supernatant was collected and stored at 20C for analysis.

Quantification of Antioxidant Capacity and Total Phenolics
Three antioxidant capacity related values including total phenolics content (TPC), total
flavonoid content (TFC) and oxygen radical absorbance capacity (ORAC) values are
measured. TPC was measured based on Folin-Ciocalteau method according to Wu et al
(Wu and others 2004). Gallic acid (50, 25, 12.5, 6.25, 3.125, 1.5625 mg/L, correlation
coefficient, r = 0.999) was used to establish the standard curve. These results were
expressed as gallic acid equivalents (mg GAE/100 g fresh weight sample). TFC was
determined using a colorimetric method described previously (Heimler and others 2005).
The results were calculated and expressed as catechin equivalents (mg CAE/100 g fresh
weight sample). Hydrophilic ORAC procedure based on previous report (Huang and others
2002). The results were expressed as Trolox equivalents (mol TE/100 g fresh weight
sample).

Phytochemicals and Phytoalexins Identification
HPLC analysis was carried out on a Waters HPLC system (Milford, MA) with a Alliance
2659 separation module, a 2996 photodiode array detector (PDA), and a Waters C18
column (5 m, 4.6250 mm, Atlantis, Ireland). The detection wavelength was set from 210
to 800 nm. The separation was accomplished with water (A), acetonitrile (B) and 2% acetic
acid in water (C) as mobile phase. The column temperature was 30C. The injection
volume was 20 L. Solvent C composition was maintained at an isocratic 5% for 60 min.
Solvent A and B gradient was as follows: 0 1 min, A 95%; 1 8 min, A from 95% to
85%; 8 24 min, A from 85% to 70%; 24 34 min, A from 70% to 40%; 34 50 min, A
from 40% to 20%; 50 55 min, A from 20 % to 5%; 55 58 min, A from 5% to 95 %; 58
60 min, A 95%. The flow rate was 1.0 mL/min. LC-MS spectra were acquired using a
Finnigan/MAT LCQ ion trap mass spectrometer (San Jose, CA) equipped with a TSP 4000
HPLC system and an electrospray ionization (ESI) source, which consisted of a P4000
quaternary pump, UV6000LP PDA detector, and AS3000 autosampler. The LC conditions
used solvent A (water with 0.05% acetic acid) and B (acetonitrile with 0.05% acetic acid)
The 12
th
16 -18 June, 2011

99

as mobile phase. The gradient was identical to those used for HPLC analysis above. The
injection volume of each sample was 20 L.

Comprehensive Evaluation of Antioxidant Contents of Fungal-Stressed Sprouts
TPC of ungermination (UG, 0 day), germination (G, 1-4 day) and germination with fungal
stress (GS, 1-4 day) of 13 selected legume seeds are presented in Table 1. Germination
improves TPC in most cases except yellow soybean that has less changes during 4-day
germination, and this result is similar with the investigations on phenolic contents of
germinated edible seeds (Cevallos-Casals & Cisneros-Zevallos 2009) and 9 selected
legumes (Lin & Lai 2006). Meanwhile, Legume seeds with food grade fungi R.
oligosporous stress have much higher (P < 0.01) TPC than that of without fungal-stressed
germination. Overall, our data suggest that fungal-stressed germination could greatly
increase the TPC in legume seeds.

TFC ranges from the minimum 3.19 (chickpea) to the maximum 19.68 (yardlong bean) mg
CAE/100g FW in UG seeds (Table 1). Germination significantly improves TFC in broad
bean by 9.69 times, 2.57 times for yellow soybean and 2.42 times for chickpea. In sharp
contrast, TFC reduce in germinated yardlong, azuki, hyacinth and mung beans. Majority of
germinated legume seeds with fungal-stress have higher (P < 0.05) TFC than that of
without fungal-stress.

ORAC is a method of measuring antioxidant capacities in biological samples in vitro. A
wide variety of foods has been tested using this assay. Different legume seeds have very
large differences (P < 0.05) of ORAC values ranging from 456 (mung bean) to 2805
(yellow soybean) (Table 1). Germination significantly increases ORAC in all legumes.
Fungal stress results in higher ORAC value than that of without fungal-stressed
germination. The sprouts from mung bean and yellow soybean are the most popular
traditional food. ORAC of mung bean and yellow soybean sprout from supermarket in
Singapore are 1606 and 3126 mol TE/100g FW respectively (Isabelle, and others 2010),
while soybean sprout is 962 mol TE/100g FW from USDA Database for the ORAC of
Selected Foods (Release 2, 2010). In this study, the ORAC of mung bean sprouts are 1759
mol TE/100g FW, and yellow soybean sprouts are 4003 mol TE/100g FW.

The 12
th
16 -18 June, 2011

100

Table 1 Comparative evaluation of total phenolic content (TPC), total flavonoid content (TFC) and
oxygen radical absorbing capacity (ORAC) of 13 legume seeds.
TPC TFC ORAC
UG G GS UG G GS UG G GS
Sword bean 42.0g 42.3 63.3 10.11d 10.82 12.99 614g 1298 1903
Kidney bean 33.2h 48.5 77.6 6.20g 8.89 12.07 888e 1855 2145
Black-eyed pea 28.1i 59.6 68.2 9.17e 10.70 12.63 753f 1708 2195
Yardlong bean 54.8f 60.0 67.3 19.68a 12.38 12.88 1139d 1863 2601
Azuki bean 43.1g 54.2 57.2 11.68b 10.33 9.56 858e 1495 1682
Hyacinth bean 15.0j 34.2 45.0 4.26h 3.61 5.63 682g 1186 2021
Mung bean 42.0g 70.6 85.7 8.78e 7.70 9.79 456h 1759 2705
Broad bean 60.6e 129.6 147.0 7.80f 56.73 59.13 679g 2816 2679
Yellow soybean 107.2b 106.5 110.0 6.10g 12.95 13.85 2805a 4003 5038
Black soybean 69.2d 76.9 86.4 4.78h 7.65 14.48 1667c 2434 3632
Big peanut 97.8c 125.4 151.6 8.15f 9.21 9.77 762f 2428 3157
Small peanut 116.5a 153.9 152.4 10.80c 13.19 30.92 1768b 2249 3484
Chickpea 54.6f 91.7 98.6 3.19i 6.03 6.54 874e 1997 2352
Minimum 15.0 34.2 45.0 3.19 3.61 5.63 456 1186 1682
Maximum 116.5 153.9 152.4 19.68 56.73 59.13 2805 4003 5038
Mean 58.8C 81.0B 93.1A 8.52C 13.09B 16.17A 1073C 2084B 2738A
TPC expressed as mg GAE/100g FW; data present as mean, n = 4, RSD < 5%. TPC of 13 legume seeds are
determined at three different treatments: UG = ungermination at 0 day, G = germination for 1 to 4 day
without fungal stress, GS = germination with stress of food grade fungus R. oligosporous for 1 to 4 day.
Means of UG (lowercase letters in the UG column) or among UG, G and GS (uppercase letters in the last
row) were compared with Duncans multiple range test (P < 0.05), different letters showed significant
differences. TFC expressed as mg CAE/100g FW; data present as mean, n = 4, RSD < 7%. ORAC value
expressed as mol TE/100 g FW, data present as mean, n = 4, RSD < 8%.

In order to rank antioxidant capacity of 13 GS samples based on TPC, TFC and ORAC, the
three criteria are given same priority and ranking of antioxidant capacity based on means
of membership function values f(x) of TPC, TFC and ORAC, and the order from largest to
smallest is listed in Table 2. Interestingly, we find that broad bean has the highest
antioxidant capacity. This might be germination of broad bean can significantly increase
the production of catechin derivatives, since USDA Flavonoid Database (2003) shows the
concentration of (-)-epicatechin, (-)-epigallocatechin and (+)-catechin in immature raw
broad bean seeds are 22.51, 14.03, 12.83 mg/100 g FW respectively. From Table 1, we
also note the change of broad bean on TFC is much higher than TPC and ORAC.
Additionally, our results indicate that the antioxidant capacity of soybean, peanut and
mung bean are ranked top.

The 12
th
16 -18 June, 2011

101

Table 2 Ranking of antioxidant capacity based on three criteria of TPC, TFC and ORAC under GS
Mean
a
membership function value f(x)
b

Samples
TPC TFC ORAC TPC TFC ORAC Mean Ranking
Sword bean 63.3 12.99 1903 0.17 0.14 0.07 0.12 11
Kidney bean 77.6 12.07 2145 0.30 0.12 0.14 0.19 9
Black-eyed pea 68.2 12.63 2195 0.22 0.13 0.15 0.17 10
Yardlong bean 67.3 12.88 2601 0.21 0.14 0.27 0.21 8
Azuki bean 57.2 9.56 1682 0.11 0.07 0.00 0.06 12
Hyacinth bean 45.0 5.63 2021 0.00 0.00 0.10 0.03 13
Mung bean 85.7 9.79 2705 0.38 0.08 0.30 0.25 6
Broad bean 147.0 59.13 2679 0.95 1.00 0.30 0.75 1
Yellow soybean 110.0 13.85 5038 0.61 0.15 1.00 0.59 3
Black soybean 86.4 14.48 3632 0.39 0.17 0.58 0.38 5
Big peanut 151.6 9.77 3157 0.99 0.08 0.44 0.50 4
Small peanut 152.4 30.92 3484 1.00 0.47 0.54 0.67 2
Chickpea 98.6 6.54 2352 0.50 0.02 0.20 0.24 7
a
TPC (mg GAE/100g FW), TFC (mg CAE/100g FW) and ORAC (mol TE/100g FW) values are expressed
as mean of GS in four days.
b
f(x) = (x x
min
)/(x
max
x
min
), x is the mean of TPC, TFC and ORAC
respectively, ranking based on means of f(x) values in TPC, TFC and ORAC, the number of ranking is
smaller indicates the antioxidant capacity of sample is stronger.

Phytochemical and Phytoalexin Changes in Germinating Legume Seeds
HPLC chromatograms of the thirteen legumes under different conditions are then collected
and analyzed. The numbers of peaks at three wavelengths 260, 300 and 340 nm are
counted respectively. The AU of peaks that are counted is above 0.002 AU. After the data
of the numbers of peaks are collected, the thirteen legume seeds are then divided into four
different classes (Table 3).

The difference in the numbers of peaks between the chromatograms under UG and those
under G indicates the change in the phytochemical contents in the legumes during
germination. Whereas, the difference in the numbers of peaks between G and GS shows
the increase or decrease in the phytoalexins that are released after fungal stress
germination. The legume seeds are divided into the four classes so that it will be very
helpful in determining which legume seeds will produce more or less phytochemicals when
they are germinated and which will synthesize more phytoalexins when they are
germinated under fungal stress.

The 12
th
16 -18 June, 2011

102

Table 3 Comparison of the number of peaks in HPLC chromatograms for the thirteen legume seeds
under non-germination (UG), germination (G) and germination and fungal stress (GS).
Number of Peaks
260 nm 300 nm 340 nm Categories Legumes
UG G GS UG G GS UG G GS
Small peanut 30 52 83 28 58 69 19 49 55
Big peanut 32 69 75 26 64 75 14 54 65
Sword bean 43 55 65 39 55 72 27 42 52
Black soybean 35 70 75 30 59 71 16 36 51
Rich phytoalexins &
enhanced phytochemicals
Azuki bean 18 32 37 13 18 36 1 15 33
Broad bean 38 52 59 18 43 58 9 30 45
Kidney bean 23 73 75 21 73 77 14 53 55
Black-eyed bean 17 58 63 20 60 62 12 35 40
Hyacinth bean 17 50 55 13 37 39 5 22 25
Low phytoalexins &
enhanced phytochemicals
Chickpea 18 64 67 9 55 60 3 35 40
Mung bean 9 50 46 5 36 30 3 15 12 Less phytochemicals
under GS Yardlong bean 26 50 26 37 46 19 20 30 11
Less phytochemicals under G Yellow soybean 49 48 60 47 42 47 34 26 32

Rich phytoalexins and enhanced phytochemicals are the legume seeds that show high
difference in the numbers of peaks between the chromatograms under UG and G, as well
as between the chromatograms under G and GS. High difference means that the difference
in the numbers of peaks between UG and G or G and GS are greater than 10 in average of
the three wavelengths. For example, peanut is one of the most widely used legumes
because of their high nutritional value and good taste. After GS, there are 69 compounds
observed at 300 nm, more than 11 of which may be phytoalexins. As can be seen from
Figure 1, there are more peaks that observed from HPLC chromatogram under GS
compared to that of G. This shows that fungal stress can induce the production of
phytoalexins in peanuts during germination. To confirm whether these new released
compounds are produced by fungi themselves or by the fungal action on peanuts, the D3d
peanut is analyzed and found clearly no those compounds (Figure 1). In order to identify
these phytoalexins, we re-extract the GS small peanut sprouts at higher amount of sample,
and remove the phenolic acids (before resveratrol) and oils compounds after a traditional
silica column chromatography elution with hexane and ethyl acetate. From the LC-MS

spectral data, 45 compounds were identified in the peanut sprouts, including 14 coumaric
acids, 3 ferulic acids, 4 sinapinic acids, 2 hydroxybenzoic acids, 1 caffeic acid, 2
flavonoids and 19 stilbenoids. Small peanut sprouts produced the highest amount of
phytoalexins after GS with 55 compounds detected. Forty five of these compounds were
stilbenoid phytoalexins, 3 flavonoids, 4 oxooctadecadienic acids, 1 pterocarpanoid
phytoalexin aracarpene and 2 unknown compouds (data not show). So far, only several
simple stilbenoid phytoalexins from peanuts have been reported. More complex stilbenoid
derivatives have not previously been found in peanuts, but have been reported from other
sources and considered important factors in plant defense. The potential therapeutic value
of stilbenoid oligomers has promoted research activity on the occurrence of this class of
compounds in various plants around the world.
The 12
th
16 -18 June, 2011

103

10.00 12.00 14.00 16.00 18.00 20.00 22.00 24.00 26.00 28.00 30.00 32.00 34.00 36.00 38.00 40.00 42.00 44.00 46.00 48.00 50.00 52.00 54.00
10.00 12.00 14.00 16.00 18.00 20.00 22.00 24.00 26.00 28.00 30.00 32.00 34.00 36.00 38.00 40.00 42.00 44.00 46.00 48.00 50.00 52.00 54.00
10.00 12.00 14.00 16.00 18.00 20.00 22.00 24.00 26.00 28.00 30.00 32.00 34.00 36.00 38.00 40.00 42.00 44.00 46.00 48.00 50.00 52.00 54.00
10.00 12.00 14.00 16.00 18.00 20.00 22.00 24.00 26.00 28.00 30.00 32.00 34.00 36.00 38.00 40.00 42.00 44.00 46.00 48.00 50.00 52.00 54.00
A
U
0.00
0.01
0.02
0.03
0.04
0.05
10.0
A
U
0.00
0.01
0.02
0.03
0.04
0.05
10.0
A
U
0.00
0.01
0.02
0.03
0.04
0.05
10.0
A
U
0.00
0.01
0.02
0.03
0.04
0.05
10.0
0.00
Minutes
10.0 15.0 20.0 25.0 30.0 35.0 40.0 45.0 50.0 55.0
small peanut
UG
small peanut
G3d
small peanut
GS3d
small peanut
D3d
Phenolic acids
Phytoalexins
Oils
resveratrol

Figure 1 HPLC chromatogram (300 nm) of small peanut (UG) after 3 days germination without R.
oligosporous stress (G3d), 3 days with the fungal stress (GS3d), and thermal-deactivated
seeds inoculated without/with fungal stress for 3 days (D3d).
The 12
th
16 -18 June, 2011

104

Minutes
10.00 12.00 14.00 16.00 18.00 20.00 22.00 24.00 26.00 28.00 30.00 32.00 34.00 36.00 38.00 40.00 42.00 44.00 46.00 48.00 50.00 52.00 54.00 56.00 58.00 60.00
Kidney bean
UG
G 4 day
0.05
0.04
0.03
0.02
0.01
0.00
0.05
0.04
0.03
0.02
0.01
0.00
10 20 30 40 50 60 Minutes

Figure 2 HPLC chromatogram (300 nm) of kidney bean (UG) after 4 days germination without R.
oligosporous stress (G 4 day)

Low phytoalexins but enhanced phytochemicals are the legume seeds that show low
difference in the numbers of peaks between G and GS, but high difference in the numbers
of peaks between UG and G. Low difference means that the difference between the
numbers of peaks concerned are below 10 in average. From Table 3 and Figure 2, after
germination, the phytochemicals in kidney beans increase quite significantly by 50 at 260
nm, by 52 at 300 nm and by 39 at 340 nm. The large difference in the numbers of peaks
between UG and G indicates that kidney beans synthesize many new compounds during
germination. While slight difference in the numbers of peaks between chromatograms
under G and GS indicates that germinating kidney beans with Rhizopus oligoporus stress
may not be an effective way to induce phytoalexins.
Less phytochemicals under GS means that the legume seeds show lesser peaks in GS
chromatograms when they are compared with G chromatograms. For example, yardlong
bean has been reported that contains many anthocyanin derivatives (Ha and others 2010).
The phytochemicals in yardlong bean decrease significantly by 24 at 260 nm, by 27 at 300
nm and by 19 at 340 nm (Figure 3). However, mung bean shows less decrease in the
numbers of peaks between GS and G compared to yardlong bean.
The 12
th
16 -18 June, 2011

105

Minutes
10.00 12.00 14.00 16.00 18.00 20.00 22.00 24.00 26.00 28.00 30.00 32.00 34.00 36.00 38.00 40.00 42.00 44.00 46.00 48.00 50.00 52.00 54.00 56.00 58.00 60.00
Yardlong bean
G 4 day
GS 4 day
10 20 30 40 50 60 Minutes
0.05
0.04
0.03
0.02
0.01
0.00
0.05
0.04
0.03
0.02
0.01
0.00

Figure 3 HPLC chromatogram (300 nm) of yardlong bean after 4 days germination without R.
oligosporous stress (G 4 day) and 4 days with the fungal stress (GS 4 day)

Less phytochemicals under G means that the legume seeds show lesser peaks in G
chromatograms when they are compared with UG. Yellow soybean is rich in isoflavones.
The phytochemicals slightly reduce by one at 260 nm, by 5 at 300 nm and by 8 at 340 nm
after germination (Figure 4), which represents that germination might not be an effective
way to enhance the production of phytochemicals in yellow soybeans.
Overall, different legume seeds have different responses towards germination and fungal-
stress germination. Although some legume seeds belong to the same species, they still have
different responses towards germination and fungal stress germination. In this experiment,
the two peanuts also show a slight difference in their responses towards germination and
fungal stress germination. Both peanuts have produced many phytochemicals and are rich
in phytoalexins after fungal stress germination. However, it can be observed that big
peanuts produced more phytochemicals upon germination compared to small peanuts.
Although black soybeans and yellow soybeans are the same species, the phytochemicals
contents in both soybeans are quite different as well as the response towards fungal stress.
Black soybeans produce more phytochemicals as well as phytoalexins under fungal-
stressed germination, while yellow soybeans do not.
The 12
th
16 -18 June, 2011

106

Minutes
10.00 12.00 14.00 16.00 18.00 20.00 22.00 24.00 26.00 28.00 30.00 32.00 34.00 36.00 38.00 40.00 42.00 44.00 46.00 48.00 50.00 52.00 54.00 56.00 58.00 60.00
Yellow soybean
UG
G 4 day
10 20 30 40 50 60 Minutes
0.05
0.04
0.03
0.02
0.01
0.00
0.05
0.04
0.03
0.02
0.01
0.00

Figure 4 HPLC chromatogram (300 nm) of yellow soybean bean (UG) after 4 days germination
without R. oligosporous stress (G 4 days)

In conclusion, these results show the food grade fungus R. oligosporus infected sprout can
greatly improve antioxidant capacity, enhance production of phytochemicals and induce
phytoalexins of 13 selected food legumes. This study proposes that we may develop a new
kind of functional food using a novel process of germinated legume seed with stress of
food grade microorganisms such as probiotics. It is possible that these functional foods
contain significantly higher concentration of phytochemicals (such as flavonoids, phenolic
acids, saponins and also phytoalexins) that may lead to many health enhancing benefits
(e.g. antioxidant, anti-inflammation, anti-cancer, anti-obesity, cholesterol-lowering and
anti-aging). However, we would firstly establish a relatively comprehensive metabolite
profiling of the stressed sprouts and analyze the safety of these food in next way to develop
this novel food and validate our hypothesis.

Acknowledgements
The authors are grateful for the financial support of National University of Singapore
Virtual Institute for the Study of Aging (VISA) (grant number: R-143-000-437-290).

References
Arora S, Jood S, Khetarpaul N. 2010. Effect of germination and probiotic fermentation on
nutrient composition of barley based food mixtures. Food Chem 119, 779-784.
Boue S, Cleveland T, Carter-Wientjes C, Shih B, Bhatnagar D, McLachlan J, Burow M.
2009. Phytoalexin-enriched functional foods. J Agric Food Chem 57, 2614.
Cevallos-Casals B, Cisneros-Zevallos L. 2009. Impact of germination on phenolic content
and antioxidant activity of 13 edible seed species. Food Chem 119, 1485-1490
Dueas M, Hernndez T, Estrella I, Fernndez D. 2009. Germination as a process to
increase the polyphenol content and antioxidant activity of lupin seeds (Lupinus
angustifolius L.). Food Chem 117, 599-607.
The 12
th
16 -18 June, 2011

107

Feng S, Saw C, Lee Y , Huang D. 2007. Fungal-stressed germination of black soybeans
leads to generation of oxooctadecadienoic acids in addition to glyceollins. J Agric
Food Chem 55, 8589-8595.
Feng S, Saw C, Lee Y, Huang D. 2008. Novel process of fermenting black soybean
[Glycine max (L.) Merrill] yogurt with dramatically reduced flatulence-causing
oligosaccharides but enriched soy phytoalexins. J Agric Food Chem 56, 10078-
10084.
Francisco M, Resurreccion A. 2008. Functional components in peanuts. Crit Rev Food Sci
48, 715-746.
Ha TJ, Lee MH, Park CH, Pae SB, Shim KB, Ko JM, Shin SO, Baek IY, Park KY. 2010.
Identification and Characterization of Anthocyanins in Yard-Long Beans (Vigna
unguiculata ssp. sesquipedalis L.) by High-Performance Liquid Chromatography
with Diode Array Detection and Electrospray Ionization/Mass Spectrometry
(HPLC- DAD- ESI/MS) Analysis. J Agric Food Chem 58, 2571-2576.
Hammerschmidt R. 1999. Phytoalexins: what have we learned after 60 years? Annu Rev
Phytopathol. 37, 285-306.
Heimler D, Vignolini P, Dini M, Romani A. 2005. Rapid tests to assess the antioxidant
activity of Phaseolus vulgaris L. dry beans. J Agric Food Chem 53, 3053-3056.
Huang D, Ou B, Hampsch-Woodill M, Flanagan J, Prior R. 2002. High-throughput assay
of oxygen radical absorbance capacity (ORAC) using a multichannel liquid
handling system coupled with a microplate fluorescence reader in 96-well format. J
Agric Food Chem 50, 4437-4444.
Khattak A, Zeb A, Bibi N. 2008. Impact of germination time and type of illumination on
carotenoidcontent, protein solubility and in vitro protein digestibility of chickpea
(Cicer arietinum L.) sprouts. Food Chem 109, 797-801.
Lpez-Amors M, Hernndez T, Estrella I. 2006. Effect of germination on legume
phenolic compounds and their antioxidant activity. J Food Composit Anal 19, 277-
283.
Lin P, Lai H. 2006. Bioactive compounds in legumes and their germinated products. J
Agric Food Chem 54, 3807-3814.
Martn-Cabrejas M, Daz M, Aguilera Y, Bentez V, Moll E, Esteban R. 2008. Influence
of germination on the soluble carbohydrates and dietary fibre fractions in non-
conventional legumes. Food Chem 107, 1045-1052.
Paucar-Menacho L, Berhow M, Mandarino J, de Mejia E, Chang Y. 2010. Optimisation of
germination time and temperature on the concentration of bioactive compounds in
Brazilian soybean cultivar BRS 133 using response surface methodology. Food
Chem 119, 636-642.
Pervaiz S, Holme A. 2009. Resveratrol: its biologic targets and functional activity.
Antioxid Redox Sign 11, 2851-2897.
Shi H, Nam P, Ma Y. 2010. Comprehensive profiling of isoflavones, phytosterols,
pocopherols,minerals, crude protein, lipid, and sugar during soybean (Glycine max)
germination. J Agric Food Chem 58, 4970-4976.
van Loon L, Rep M, Pieterse C. 2006. Significance of inducible defense-related proteins
in infected plants. Annu Rev Phytopathol 44, 135-162.
Wu X, Beecher G, Holden J, Haytowitz D, Gebhardt S, Prior R. 2004. Lipophilic and
hydrophilic antioxidant capacities of common foods in the United States. J Agric
Food Chem 52, 4026-4037.

The 12
th
16 -18 June, 2011

108

OA-106

Antihyperuricemic Effect of Ethanol Extract of Snake Fruit (Salacca
Edulis Reinw.) Var. Bongkok on Wistar Male Rat

Leni Herliani Afrianti Priyatno
1,*
, Elin Yulinah Sukandar
2
, Slamet Ibrahim
3
,
I Ketut Adnyana
2

1
Food Technology Department of Pasundan University, Bandung, Indonesia;
2
Pharmacology-Clinical Pharmacy, School of Pharmacy-Institute Technology Bandung, Indonesia;
3
Pharmaceutical chemistry Research Group, School of Pharmacy-Institute Technology Bandung, Indonesia
*
Corresponding author: leni_priyatno@yahoo.com

Abstract

The aim of the study was to investigate antihyperuricemic effect of snake fruit (Salacca
edulis Reinw.) var. Bongkok Wistar male rates. Experimental animal model of
hyperuricemia induced by uricase inhibitor potassium oxonate has been used to study.
Wistar rat were injected intraperitoneally with potassium oxonate 1 hr before the final drug
administration to increase the serum urate level. Whole blood samples were collected from
rat by tail vein bleeding. Serum uric acid and urine 1 + 10 with dist.water were determined
by the phosphotungstic acid method. To evaluate the efficacy, statistical test of
significance with the analysis of variations in student t test was performed.
Antihyperuricemic investigation on Wistar male rats showed that administration of
ethanol extract at doses of 200 mg/kg bw decreased serum uric acid level significantly
compared to control group at hour 6 and 7 (p < 0.05) after inducing with potassium
oxonate intraperitoneally simultaneously with uric acid orally. Whereas, administration of
ethanol extract at doses of 100 mg/kg bw did not decrease serum uric acid level
significantly compared to control group at hour 6 and 7 (p < 0.05). Determination of uric
acid level in urine, administration of ethanol extract at a dose of 200 mg/kg bw, or
probenecid as a standard drug, at a dose of 45 mg/kg bw increased excretion of urine uric
acid level significantly different compared to control group in day of 7 (p < 0.05) after
inducing with potassium oxonate intraperitoneally simultaneously with uric acid orally.
However, increase of uric acid excretion by ethanol extract was lower compared to that of
probenecid at a dose of 45 mg/kg bw. Mechanism of action of the ethanol extract as an
antihyperuricemia has been proposed by inhibition of xanthine oxidase and finally
decreased the synthesis of uric acid and increased the excretion of urine uric acid level.
This study revealed the existence of antihyperuricemia of snake fruit var. Bongkok

Keywords: Snake fruit var. Bongkok, ethanol extract, antihyperuricemic, probenecid,
Wistar male rat

Introduction
Snake fruit (Salacca edulis Reinw) belongs to the class of Salacca originated from
southeast Asia. The fruit was named snake fruit because skin of the fruit is brown and
looks like a snake skin. Form of fruit is egglike in shape; it contains three pieces of seeds
covered with white flesh. In Indonesia there are many snake fruit cultivars; it is known in
Java, Sumatera and other island. There are some variety of snake fruit such as Manonjaya,
Bongkok, Banjarnegara, Condet, Pondoh, Bali, Enrengkang and Sidempuan (Yustina and
Farry, 1993). Most of snake fruit have an astringent taste and are not sweet. Snake fruit
The 12
th
16 -18 June, 2011

109

var. Bongkok from Conggeang a sub district of Sumedang West Java. The fruit has more
sour, bitter, an astringent taste and it is not sweet than the other snake fruit.
Snake fruit (Salacca edulis Reinw.) var. Bongkok, which grows in Sumedang Regency,
West Java, it contained flavonoid, alkaloid, terpenoid, tannin, and quinone compound
groups, whereas saponin was not found (Afrianti and others, 2006). Isolated were extract
of snake fruit var. Bongkok the present of these two compounds i.e. 3-hydroxystigmastan-
5(6)-en (-sitosterol) and pyrolle-2.4-dicarboxylic acid-methylester. The pyrolle-2.4-
dicarboxylic acid-methylester isolated is from the snake fruit var. Bongkok is a new
compound (Afrianti and others, 2007). Snake fruit var. Pondoh contains sucrose, glucose,
fructose and volatile compounds as methyl esters of butanoic acids, 2-methylbutanoic acid,
hexanoic acid, pentanoic acid and carboxylic acids (Supriyadi and others, 2002).

HO
5
6
7
4
3
2
1
25
10
19
26
27
20
21
29
23
9
11
12
13
18
14
8
17
16
15
28
24
22

Figure 1 3-Hydroxystigmastan-5(6)-En (-Sitosterol)

H
N
OCH
3
O
O
OH
2
3 4
5
1

Figure 2 Pyrolle-2.4-Dicarboxylic Acid-Methylester

Snake fruit var. Bongkok becomes an unfavorable fruit and wasted product. In 2003, the
harvested snake fruit var. Bongkok decreased by 24 %, leading to the extinction. In order
to overcome this problem, it is needed to gain the additional economic values of the snake
fruit var. Bongkok by studying pharmacological effects in vivo test using experimental
animal to become medicine or functional food.

Xanthine oxidase is a flavin enzyme that catalyzes the oxidation of both hypoxanthine and
xanthine to uric acid. During the process of purin oxidation by xanthine oxidase, reactive
oxygen species such as peroxides generated (Fields and others, 1995). Therefore the aims
of the study were to determine antihyperuricemia, as well as to predict its mechanism of
action of snake fruit (Salacca edulis Reinw.) var. Bongkok.
The 12
th
16 -18 June, 2011

110

The composition of the species is very attracted to study, its compound has not been
known up to now. High Uric acid level in blood known as gout can enhance cardiovascular
disorder. Allopurinol is used commercially as anti gout with mechanism of action of
xanthine oxidase inhibitors. Allopurinol is indicated when uricosuric drugs fail to reduce
serum urate lower than 7.0 mg per dL. Xanthine oxidase catalyzes the oxidation of
hypoxanthine and xanthine to uric acid. Xanthine oxidase is a complex
metalloflavoprotein (Vogel and others 1996).

The ethanol extract at concentrations of 0.01, 0.02, 0.2, 2, and 2000 g/mL showed
xanthine oxidase inhibition by 20.89, 32.78, 44.96, 50.30, and 50.25 % respectively, with
IC
50
of 44.95 g/mL. At the same concentrations the pyrolle-2.4-dicarboxylic acid-
methylester, showed xanthine oxidase inhibition by 27.7, 30.5, 37.3, 50.27 and 50.55 %
respectively, with IC
50
of 48.86 g/mL. Allopurinol as a standard drug showed IC
50
of 0.92
g/mL (Afrianti and others 2007). Moreover, according to Fields and others (1996), during
the oxidation, free radicals are also generated. Allopurinol (4-hydrxipirazolo [3,4-d)
pyrimidin), an analog hypoxanthine, is a specific potent inhibitor for xanthine oxidase,
hence decreases blood uric acid level. In this research was found that the ethanol extract of
snake fruits var. Bongkok showed antioxidant activity decreased serum uric acid level with
its mechanism of action similar to that showed xanthine oxidase inhibition in vitro
(Afrianti, and others, 2007).

This paper describes mechanism of action of the ethanol extract of snake fruit var.
Bongkok as an antihyperuricemia has been proposed by inhibition of xanthine oxidase and
finally decreased the synthesis of uric acid and increased the excretion of urine uric acid
level. This study revealed the existence of antihyperuricemia of snake fruit var. Bongkok

Plant material
The snake fruit (Salacca edulis Reinw) var. Bongkok was collected from Conggeang a sub
district of Sumedang West Java, Indonesia, and it was weighed, peeled, fractionated into
little pieces, and dried at 40
o
C in tunnel drier to constant weight. The dried samples were
ground to fine powder by using a grinder. The dried powdered snake fruit (Salacca edulis
Reinw) var. Bongkok (10 g) was macerated in ethanol (100 ml) and kept in container
overnight then filtered (Whatman No.1 filter paper). The filtrates were evaporated at 40
o
C

Animal model of hyperuricemia on male rat (in vivo)
Experimental animal model of hyperuricemia induced by uricase inhibitor potassium
oxonate has been used to study (Al-Qirim and others 2002). Briefly, male rat were injected
intraperitoneally with potassium oxonate (200 mg/kg) and natrium urate (15 mg/kg) orally
1 h before the final drug administration to increase the serum urate level. Whole blood
samples were collected from rat by tail vein bleeding. The blood was allowed to clot for
approximately 1 hr at room temperature and then centrifuge at 12000 rpm min to obtain the
serum. The serum was stored at -20
o
C until assayed. Serum uric acid was determined by
the phosphotungstic acid method (Zhu and others 2004).

The 12
th
16 -18 June, 2011

111

Drug administration
Snake fruit var. Bongkok extract and probenecid at various concentrations were dissolved
in CMC-Na 0.5%. The volume of the suspension administered was based on body weight
measured immediately prior to each dose, respectively.

The extract and positive control (drug) orally for 5 consecutive days performed on twenty-
four male Wistar rats randomly divided into four groups, as the following:
(i) Control group (CMC 0.5%)
(ii) Ethanol group (dose 100 mg/kg bw)
(iii) Ethanol group (dose 200 mg/kg bw)
(iv) probenecid group (45 mg/kg bw).

Whole blood samples were collected from rat by tail vein bleeding. The blood was allowed
to clot for approximately 1 hr at room temperature and then centrifuged at 12000 rpm for 5
min to obtain the serum. The serum was stored at 20
o
C until assayed. Dilute urine 1 + 10
with dist.water. Determination of uric acid was done by the phosphotungstic acid method.

Result and Discussion
Figure 3, antihyperuricemic investigation to Wistar male rats showed by administration of
ethanol extract at doses 200 mg/kg bw, decreased serum uric acid level significantly
different compared to control group at hour 6 and 7 (p < 0.05) after induced with
potassium oxonate intraperitoneally simultaneously with uric acid orally. Whereas,
administration of ethanol extract at dose of 100 mg/kg bw did not decreased serum uric
acid level significantly different compared to control group at hour 6 and 7 (p < 0.05). But,
probenecid did not decrease serum uric acid level significantly; it is to accelerate the
excretion of urine uric acid.

Figure 3 Antihyperuricemic investigation to wistar male rats showed by administration of ethanol
extract of snake fruit var. Bongkok at doses 100, 200 mg/kg bw

Nguyen and the others (2004), have reported inhibitory effect on xanthine oxidase some of
medicinal plants in the South of Vietnam such as Artemisia vulgaris (leaf), caesalpinia
sappan (wood), Blumea blasamifera (aerial parts) and T. scandens (stem).
The 12
th
16 -18 June, 2011

112

The efficacy of Biota orientalis extract, and its main flavonoids constituents quercetin and
rutin in reducing serum urate levels in mouse model of hyperuricemia induced by the
uricase inhibitor potassium oxonate and in vivo inhibiting xanthine oxidase activities in
mouse liver (Zhu and the others, 2004)

Figure 4 Percentage of Decreasing Serum Uric Acid In Male Rats

Figure 4: It shows determination of uric acid level in urine, administration of ethanol
extract at a dose of 200 mg/kg bw, or probenecid as a standard drug, at a dose of 45 mg/kg
bw increased excretion of urine uric acid level significantly different compared to control
group in day of 7 (p < 0.05) after induced with potassium oxonate intraperitoneally
simultaneously with uric acid orally. However, increase of uric acid excretion by ethanol
extract was lower compared to that of probenecid at a dose of 45 mg/kg bw. Probenecid is
used as anti gout with mechanism of action to accelerate the excretion of urine uric acid (as
uricosuric); it has no effect to decrease uric acid serum.
Figure 5: Determination of uric acid level in urine, administration of ethanol extract at a
doses of 100 and 200 mg/kg bw, show increase of uric acid excretion in urine was lower
compared to that of probenecid as a standard drug, at a dose of 45 mg/kg bw increased
excretion of urine uric acid level significantly different compared to control group in day
of 7 (p < 0.05) after induced with potassium oxonate intraperitoneally simultaneously with
uric acid orally.

Mechanism of drug action antihyperuricemia consisted of two types that were inhibit the
activity of xanthine oxidase and to accelerate the excretion of urine uric acid (Zhu and
others, 2004), Moreover, according to Katzung (2002), probenecid increases urine uric
acid excretion by inhibition of proximal tubulus reabsorption, hence reduces serum uric
acid level. In this research it was found that ethanol extract of snake fruit var. Bongkok
increased the excretion of urine uric acid level.

This research has discovered the existence of antihyperuricemia of ethanol extract of snake
fruit var. Bongkok. Mechanism of action of the ethanol extract as an antihyperuricemia has
been proposed decreased the synthesis of uric acid. The mechanism of antihyperuricemia
of the ethanol extract to be an uricosuric because being increasing the excretion of urine
uric acid.
The 12
th
16 -18 June, 2011

113

Figure 5 The Excretion of Urine Uric Acid Level

The existence of pharmacological active compound from extract of snake fruit (Salacca
edulis Reinw.) var. Bongkok can be used as a potential lead compound to develop the
extract of snake fruit var. Bongkok as medicine and or functional food.
Further researches that can be conducted to development of antihyperuricemic medicine as
well as functional food from extract of snake fruit var. Bongkok are safety tests and
formulation development of the extracts and its active compound.

Conclusion
Mechanism of the ethanol extract of snake fruit var. Bongkok as an antihyperuricemia has
been proposed. It decreases the synthesis of uric acid and acts as a uricosuric because of
an increase in the excretion of urine uric acid.

Acknowledgments
I would like to thank the Hibah Kompetitif Batch 2, 2011 for the financial support.

References
Afrianti LHA, Yulinah EY, Ibrahim S, Adnyana IK. 2006. Inhibisi Xanthin Oksidase.
Ekstrak Daging Buah Salak Varietas Bongkok (Salacca Edulis Reinw.). J.
Infomatek, 8(1), 1-5.
Afrianti LHA, Yulinah EY, Ibrahim S, Adnyana IK. 2007. Xanthine Oxidase inhibitor
activity of terpenoid and pyrrole compounds isolated from snake fruit (Salacca
edulis Reinw) cv. Bongkok. J. Appl. Sci, 7(20), 3127-3130.
Al-Qirim TM, Shahwan M, Zaidi KR, Uddin Q, Banu N. 2002. Effect of khat, its
constituent and restraint stress on free radical metabolism of rats. J. Ethnopharm,
83:245-250
Fields M, Charles GL, and Mark DL. 1996. Allopurinol, an inhibitor of xanthine oxidase,
reduces uric acid levels and modifies the signs associated with copper deficiency in
rats fed fructose. J. Free Radical Biol & Med, 20(4), 595-600.
Katzung BG. 2002. Farmakologi Dasar dan Klinik. Salemba Medika. Jakarta, 487-493p.
The 12
th
16 -18 June, 2011

114

Nguyen, MTT, Awale S, Tezuka Y, Tran QL, Watanabe H, and Kadota S. 2004. Xathine
oxidase inhibitory activity of Vietnamese medicinal plants. J. Biol. Pharm. Bull.
27,1414-1421.
Supriyadi S, Suzuki M, Yoshida K, Muto T, Fuujita A, and Watanabe. 2002. Changes in
the volatile compounds and in the chemical and physical properties of snake fruit
(Salacca edulis Reinw.) cv. Pondoh during maturation. J. Agric.chem. ,(50),7627-7633.
Vogel HG, and Wolfgang HV. 1996. Drug discovery and evaluation pharmacological
assay. Springer, Philadelphia. 178-179p.
Zhu ZX, Wang Y, Kong LD, Yang C, and Zang X. 2004. Effects of Biota orientalis extract
and its flavonoid constituents, quercetin and rutin on serum uric acid levels in
oxonate-induced mice and xanthine dehydrogenase and xanthine oxidase activities
in mouse liver. J. Ethnopharmacology, 93:133-140.

The 12
th
16 -18 June, 2011

115

OA-233
A Review of Research Developments of Malaysians
Biodiversified Resources
Abdul Salam Babji
*
, Salma Mohamad Yusop

and Masomeh Ghassem

Food Science Programme, School of Chemical Sciences and Food Technology,
Universiti Kebangsaan Malaysia, 43600, Bangi, Selangor, Malaysia
*
Corresponding author: daging@ukm.my

Abstract
Malaysia is blessed as one of the twelve countries in the world with mega-diverse bio-
resources. This allows the utilization of biodiversified materials of agro plant, animal and
marine to turbocharge in Malaysia, which is very keen to play a significant part to
contribute to various fields of human knowledge and its application. Malaysia flora is
estimated to contain about 12,500 species of flowering plants and more than 1,100 species
of ferns. Marine ecosystem is also rich in a variety of life forms while the coral community
is considered to be the most diverse in the world. Agro plant and animal materials and their
by-products seemed to be a major cause of environmental pollution. Specifically addressed
is the inefficiency of farming and practices of harvesting yield from the agriculture sector.
Plant materials of herbal/spice nature currently are poorly exploited for their functional
properties which can be beneficial to mankind. Antioxidants extracted from agro materials
currently used crude methods to yield many products available in the local market.
However, none of these anti-aging products meet the requirement for export market, due to
the low quality and insufficient scientific information and tests to evaluate the safety and
efficacy of herbal supplements. More important is the long-term damage these crude
preparations can caused to the local population consuming a variety of well-advertised
concoctions associated with youth vitality and strength for man.
Keywords: Biodiversity, Bio-resource, Functional properties, Antioxidant, Anti-aging.

Introduction
The past eight years resulted in 5 research projects by University Kebangsaan Malaysia,
exploring the Malaysian local resources resulting in basic scientific and technological
information that contributed to obtaining functional components deemed using to mankind
(Babji and others 2005; Chan and Babji 2006a,b; Huda-Faujan and others 2007; Noriham
and others 2004a,b). Phytochemicals, herbs and spices were identified for antioxidant
potentials, collagen/gelatin was obtained from local freshwater fish and bioactive peptides
from fish or waste yielded antihypertensive oligopeptide (Ghassem and others 2009).
Collagen derived from animal tissues like skin, bone and entrails now comes in many
forms available to consumers, however most of them are imported and of questionable
origin as to the Halalness. Studies carried out showed the potentials of derivatives as
antioxidants, antimicrobial, anti-hypertensive and anti-thrombotic peptides. Collagen, in
particular, is of interest to enhance its usage from fish skin and chicken bone waste to yield
high quality of nano-particles for efficient absorption to target organs related to anti-aging
The 12
th
16 -18 June, 2011

116

and extension of facial expression related to the beauty and health industry (See and others
2010).

Birdnest and Carrageenan in freeze-dried nano form along with hydrolysed collagen and
selected phytochemicals will be upscale using modern processing machineries to optimise
parameters to yield functional ingredients for the nutraceuticals and food supplements
industries. Based on previous results on R & D of animal and plant sources, it is prudent to
look into aspects of applied R & D that will bring these crude nutraceuticals and functional
ingredients to higher level in technological processes and nutritional acceptance.
Developing prototypes of diverse range such as collagen hydrolysate, nutraceuticals,
bioactive oligopeptides, and glyco/mucoproteins by natural products from agro-resources,
are expensive, requiring product development protocols to obtain prototypes that are
similar to market products. Nanotechnology of the freeze-dried functional components
could be more efficiently absorbed and/or utilised by the consumer for the benefit of
mankind. Therefore, the objectives of this study were to review the application of derived
natural bioactive peptides such as anti-aging, anti-oxidative and ACE inhibitory peptides
from biodiversified resources available in Malaysia.

Application of Extracted Collagen from Freshwater Fish
Annually, more than 100 million tons of fish are being harvested worldwide. 29.5% of the
total catch is used for fishmeal due to its poor functional properties (Kristinsson and Rasco
2000). Processing discards from fisheries accounted to as much as 7085% of the total
weight of the catch and 30% of the waste is in the form of bones and skins with high
collagen content. These wastes are excellent raw materials for the preparation of high
protein food especially gelatin. Conversion of these wastes into value-added products to
yield additional income has both economic and waste management benefits for the fish
industry (Choi and Regenstein 2000).

The primary biochemical and structural features of collagen include the presence of
hydroxyprolyl and hydroxylsyl residues and a triple helical structure. Type I collagen are
characterized by triple helix of two
1
chains and
2
chain. Each collagen chain adopts a
left-handed helical conformation, and the three stands intertwine with a right-handed
superhelical twist exhibiting an apparent molecule weight of 300 kDa.

The most important source of collagen for gelatin and collagen hydrolysate production is
bovine hide, bone and pigskin. According to Gelatin Manufacturers of Europe (GME)
(2008), the demand for gelatin has been increasing over the years. The annual world output
of gelatin in year 2008 was 326,000 tons, with pig skin gelatin accounting for the highest
(46%) output, followed by bovine hides (29.4%), bones (23.1%), and other sources (1.5%).
The fish gelatin production is minor, yielding only about 1% of the annual world gelatin
production. However, recently, the utilization of fish by-products including fish skin and
bone to produce gelatin and collagen hydrolysate increases (See and others 2010). It is
likely due to factors such as the outbreak Bovine Spongiform Encephalopathy (BSE) and
increasing demand for non-mammalian gelatin and collagen hydrolysate for Halal and
Kosher food markets. Furthermore, fish gelatin and collagen hydrolysate production is not
only an effective waste disposal management but also contributes to additional income to
the fish processing industry by converting the waste into value-added products. These
The 12
th
16 -18 June, 2011

117

products are generally recognized as safe food ingredients by regulatory agencies (Choi
and Regenstein 2000).

Functional foods claiming health benefits also have increased greatly in recent years.
Researchers studied peptides derived from protein hydrolysates as potential nutraceuticals
and an ingredient is functional food. Gelatin and gelatin hydrolysate have been reported to
possess beneficial biological functions for some time, which justifies its use in food
supplements and pharmaceuticals preparations. Some fish gelatin hydrolysates have been
found to have noticeable antioxidant, anti-hypertensive, anticancer and antimicrobial as
well as immuno-modulatory and cholesterol-lowering effects (See and others 2010).
Recently, a number of research works have shown that peptides derived from various fish
gelatin hydrolysates act as potential antioxidants. It has been reported that the skin gelatin
hydrolysates of Alaska Pollock (Kim and others 2001) and hoki fish (Mendis and others
2005) contain antioxidant peptides. Catfish is common farm-raised, warm-water fish in
Malaysia, supplying a large amount of fish skins year-round. According to the Department
of Fisheries Malaysia (2007), the total amount of catfish production in year 2007 was
5,784.44 tones metric tons. Comprising about 5% of the whole fish, Catfish skin has
become an interesting raw material for gelatin production.

Bird Nest as a High Protein Source
There are more than 24 species of swiflest distributed around the world, but only a few
produce nests that are deemed edible. The majority of edible birds nest (EBN) traded
worldwide come from two heavily exploited species, the White-nest swiflet (Aerodramus
fuciphogus) and the Black-nest swiflet (Aerodramus maximus). Their habitats range from
the Nicobar Islands in the Indian Ocean to sea caves in the coastal regions of Thailand,
Vietnam, Indonesia, Borneo and the Palawan Islands in the Philippines. Malaysia is
situated right at the heart of the golden triangle of swiftlet birdnest production, making
our country as a strong and potential competitor in this industry.

Edible birds nest is the nest of the swift that is made from its saliva. The therapeutic
effects of EBN, including replenishing deficiency and expelling phlegm, were recorded in
a Chinese ancient literature first publish in 1695. In the past two decades, hormone-like
substances such as mitogen and avian epithelial growth factor have been reported found in
EBN. As these substances stimulate cell division and growth, they can enhance tissue
growth and regeneration and, therefore, believed to be the source of EBNs reputed
rejuvenating properties. For these therapeutic and beauty-rejuvenating applications, EBN is
much in demand in the international market.

In the past, people could only buy dried edible bird's nests. For the advancement in food
technology, a large variety of EBN related products emerge in to the healthy food market.
They are ready to serve products - no cooking process is required. Most of them are still in
the traditional form as bird's nest soup, such as instant bird's nest in different
concentrations. Some instant bird's nest may be also supplemented with other traditional
Chinese medicines. Apart from the traditional form, there is a trend of using edible bird's
nest extract/essence as one of the chief ingredients of the products, such as vitamin and
mineral supplements. These products focus mainly on the medicinal use of edible bird's
nest. On the other hand, cosmetics industry has been incorporating birds nest extract in
The 12
th
16 -18 June, 2011

118

their products to promote skin cell renewal. Hence, it has been claimed to maintain a
youthful, radiant complexion and prevents the effects of ageing (Soo 2009).

Preliminary analysis revealed that more than half of EBNs weight is consisted of protein,
making EBN a good source of protein. According to Marcone (2005), the order of
composition (form lowest to highest) in birds nest is: lipid (0.14 - 1.28%), ash (2.1%),
carbohydrate (25.62 - 27.76%) and protein (62 - 63%). The nests consist largely of a
salivary mucin-like substance, which contains glycoprotein with abundant sialic acid-
containing sugar chains. Glycoprotein is of great importance in all higher organisms,
vertebrate and invertebrate, that the body cavities of the respiratory, digestive and
urogenital tracts are lined by a cell layer secreting viscous glycoproteins. The glycoprotein
in birds nest contains about 9% sialic acid, 4.19 - 7.2% galactosamine, about 5.3%
glucosamine, 5.03 - 16.9% galactose, and about 0.7% fucose (Marcone 2005). The most
abundant amino acids are serine, threonine, aspartic acid, glutamic acid, proline, and
valine. Since the protein content in EBN is high (62-63%), and, glycopeptides are possible
to be extracted from EBN using protease, this makes birds nest a potential raw material
for enzyme-mediated production of ACE inhibitory peptides.

Carrageenan Gel from Seaweeds
Carrageenan and alginates are two of the most promising fat substitutes which are
polysaccharides derived from red seaweeds (Rhodophyceae) and brown seaweed
(Phaeophyceae) respectively. Both are able to form hydrocolloids which act as gelling
agent in meat-base products. Utilization of these ingredients in low-fat meat products is not
popular in Malaysia. It is hydrocolloid consisting of potassium, sodium, magnesium and
calcium sulphate esters of galactose and 3,6-anhydrogalactose copolymers. It has three
main fractions, kappa- and (I and II), iota-, and lambda-carrageenans, with varying
numbers and positions of the sulphate groups on the galactose dimmer (van de Velde
2008). Iota- and kappa-carrageenans have the ability to form thermo-reversible gels
whereas lambda-Carrageenan is a thickener, not a gelling agent. Due to their different
structural characteristics, these three fractions are commercially available as mixture
prepared considering specific food applications. As a result of the interaction with water
through both ionic and hydrogen bonding, carrageenans are capable of structuring water,
thus, have high water binding ability.

In the meat industry, carrageenan is used as a gelling agent in canned meats and pet foods
and allows an important reduction in fat content in comminuted meat products like
frankfurters. In cooked sliced meats carrageenan is used to improve moisture retention,
cooking yields, consistency, slice ability, cohesiveness mouthfeel and juiciness, while
simultaneously decreasing purge. In these products, the application of carrageenan is based
on its low viscosity when dispersed in the brine to be injected in the meat, its hydration
during the cooking of the ham and its gelation upon cooling (Imeson 2000). Kappa-
carrageenan was found to increase the hardness of meat batters when replacing fat by
watergum solution, whereas iota-carrageenan importantly improves the water-holding
ability. Both kappa-and iota-carrageenan were found to increase the cooking yield,
hardness and bind strength of low-fat sausages, although these effects were less
pronounced at higher salt concentration. Hsu and Chung (2001) observed an increase in
cooking yield, hardness, and other textural profile analysis parameters when adding up to
2% kappa--carrageenan to low-fat emulsified meatballs.
The 12
th
16 -18 June, 2011

119

Although, there are some successful reports regarding to replacement of animal fat by
carrageenan and pectin in frankfurters (Jimenez-Colmenero and others 2010) the research
on other meat products are scarce. No studies have been reported on frequently-consumed
meat products such as burger and meatball, which are of considerable economic
importance and enjoy wide consumer acceptance in certain sectors of the population.
Burger and meatball are popular foods, however, they are becoming of health concern for
the consumers because of their high animal fat contents. The design of healthier meat
products in which fat and salt contents are reduced and the presence of edible seaweeds is
added is a promising avenue of research.

ACE Inhibitory Peptides Derived from Freshwater Fish
Most people have a concern for their diet from a health aspect. Dietary food guides, such
as the USDA food guide pyramid, are tools to inform the public about diet, nutrition and
health (Lachance and Fisher 2005). Due to increasing concerns about health in recent
years, much attention has been paid to the physiological functions of foods, including
anticarcinogenicity, antimutagenicity, antioxidative activity and anti aging activity
(Arihara 2004). Efforts have been made by the food industries in many countries to
develop new physiological functions. Foods having these physiological functions are
known as functional food.

Several attractive meat-based bioactive substance have been studied (Arihara 2004),
including carnosine, anserine, L-carnitine, conjugated linoleic acid, glutathione, taurine
and creatine. In addition to these bioactive compounds, protein-derived peptides are
another group of promising functional compounds of meat. Although the activities of these
peptides in the sequence of proteins are latent, they are released by proteolytic enzymes
(i.e., muscle, microbial and digestive proteinases). Therefore, meat proteins have possible
bioactivities beyond a nutritional source of amino acids alone. Examples of such bioactive
peptides are anti-hypertensive, opioid, immunostimulating, antimicrobial, antithrombotic,
hypocholesterolemic and antioxidative peptides (Arihara 2004). Although most proteins
would contain bioactive sequences, those sequences are inactive or incomplete within the
parent proteins. There are several ways to produce bioactive peptides from proteins:
gastrointestinal proteolysis, aged meats, fermentation of meat and commercial proteases.

The most extensively studied bioactive peptides generated from food proteins are
angiotensin I-converting enzyme (ACE) inhibitory peptides (Meisel and others 2005;
Vermeirssen and others 2004). ACE inhibitory peptides have attracted much attention
because of their ability to prevent hypertension. These peptides could be used as potent
functional food additives and would constitute a natural and healthier alternative to ACE
inhibitory drugs (Ghassem and others 2010). ACE is a dipeptidylcarboxypeptidase that
converts an inactive form of decapeptide, angiotensin I, to a potent vasoconstrictor,
octapeptide angiotensin II, and inactivates bradykinin, which has a depressor action (Li and
others 2004). Therefore, it is capable of suppressing the elevation of blood pressure by
inhibiting the catalytic action of ACE. Many ACE inhibitors have been isolated in
enzymatic digestive products from food proteins such as zein, milk, wheat, blood, soy
bean, chicken muscle dried bonito, and sardine muscle (Fujita and others 2000).

The 12
th
16 -18 June, 2011

120

Phytonutrients from Local Malaysian Herbs
Natural antioxidants from local Malaysian herbs and vegetables may have more consumer
appeal as those plants are accepted as healthy and safe for consumption. The recent revival
of consumer interest in the therapeutic properties of herbs has led to an increase in
consumer demand for herbal-based food products. Malaysia being a tropical country
enjoys the privilege of abundant rain forests, rich in phytochemicals (Babji and others
2005; Chan and Babji 2006a,b; Huda-Faujan and others 2007; Noriham and others 2004a,b).
Over a short period of ten years, scientists in Asia have explored the active components in
plants that exhibited antioxidative, antimicrobial, anti-inflammatory, and anti-cancer
properties. It has long been known that herbs and spices possessed anti-oxidative property
and since ancient time those plant chemicals have been used to prolong shelf-life and
improve taste of meat products (Mielnik and others 2003).

Several studies had been conducted to evaluate the correlation between phenolic
compounds and antioxidant activity. The antioxidant activity of Du-Zhong (Eucomnia
ulmoides), Mung Bean Hulls, ear mushrooms and anise seed (Pimpenella anisum L.)
(Glcin and others 2003) were found to correlate with the phenolic compounds. Studies on
local plant such as turmeric (Gurcuma domestica), betel leaf (Piper betel), pandan leaf
(Panadanus odorus), asam gelugur (Garnicia atroviridis), mengkudu (Morinda citrifolia),
pegaga (Centella asiatica), ginger (Zingiber officinale), and cassava shoot (Manihot
asculenta) also exhibited good antioxidant activity (Noriham and others 2004). Wang and
others (1999) reported that the antioxidative properties of some vegetables and fruits are
partly due to the low molecular weight phenolic compounds, which are known to be potent
as an antioxidant.

Preliminary screening of potential herbs and spices showed higher anti-oxidation activities
than BHT and vitamin C (Huda-Faujan and others 2007): Garcinla atroviridis (asam
gelugor), Allium sadiumium (garlic), Akerrhoe belimbi (belimbing buluh), Murraya
koenigii (curry leaf), Citrus micocaspa (Musk lime), Curcinma domestica (tumeric leaf),
Indian pennyword (pegaga), Cosmos candatus (ulam raja), Poygonum adoratum (kesum),
Centella asiatica L. (Indian pennywort), Melicope lunu-ankenda (tenggek burung),
Morinda citrifolia (mengkudu), Manihot asculenta (Cassava shoot), Mentha arvensis
(mint), Cymbopogon citrates (lemon grass), Momordica charantia L. (bitter gourd), and
Psohorcarpus tetragonolobus (four angled bean).

Water extracts from clove and cinnamon that are commonly found in Malaysian curry
formulation exhibited powerful antioxidant activity. These water extracts were similar to
ascorbic acid and combination of BHA/BHT (50% / 50%) in chicken meat ball emulsion.
Ethanol and methanol extracts exhibited strong antioxidative properties but not used due to
food safety aspects and toxicity consideration. Addition of herbal extracts was effective in
overall quality improvement of beef and chicken meat products. Most herbal extracts were
effective and able to compete with synthetic additives.

Conclusion
The past 8 years of R & D at University Kebangsaan Malaysia, has resulted in research
projects exploring the local resources resulting in basic scientific and technological
information that contributed to obtaining functional components deemed useful to
mankind. Phytochemicals, herbs and spices were identified for antioxidant potentials;
The 12
th
16 -18 June, 2011

121

collagen/gelatin was obtained from local freshwater fish with cooperation from the
government and food industry. Bioactive peptides from fish or waste yielded anti-
hypertensive oligopeptide and collagen drinks. Basic research and development on birdnest
and seaweed is on-going. A total of RM 1,226,450 (USD 402,000) was spent with
outcomes of publications, prototype products and scientific and technological know-how to
elucidate the potentials of collagen, anti-hypertensive, birdnest bioactives and seaweeds.
From crude extractions of biodiversified components, application of nanotechnology and
freeze-drying will bring forward the supplement/nutraceutical industry to the ultimate level
of global competitive for many of our diversified resources. Innovative enzymatic/physical
operation process of purification, biologically active freeze-dry particle size optimization
will enhance the ultimate usage of those active ingredient. Current laboratory methods of
extraction will be extended to optimize innovative nano-particle freeze-dried to produce
competitive nutraceuticals/supplements currently not available in the market.

References
Arihara K. 2004. Functional foods. In: Jensen WK, Devine C, Dikeman M, editors.
Encyclopedia of meat sciences. Oxford: Elsevier. p. 492499.
Babji AS, Noriham A, Chan KW, Teoh SW. 2005. Screening of various plant extracts as
potential natural dietary antioxidants. Proceeding of 20
th
Scientific Conference &
Annual General Meeting on Nutrition Society of Malaysia. 107 p.
Chan KW, Babji AS. 2006a. Effects of adding clove buds and cinnamon bark extracts on
the oxidative stability of chicken meatballs. Proceeding of 52
nd
International
Congress of Meat Science and Technology; 2006 August 1318; Dublin, Ireland.
439 -440 p.
Chan KW, Babji AS. 2006b. Effects of cinnamon barks and clove buds extracts on the
oxidative stability, physico-chemical properties and sensory acceptability of
chicken meatballs. Malay J Anim Sci 11: 8389.
Choi SS, Regenstein JM. 2000. Physico-chemical and sensory characteristics of fish
gelatin. J Food Sci 65:194-9.
Fujita H, Yokohama K, Yoshikawa M. 2000. Classification and anti-hypertensive activity
of angiotensin I-converting enzyme inhibitory peptides derived from food proteins.
J Food Sci 65(4):5649.
Ghassem M, Khoo TC, Feni HS, Babji AS, Tengku Rozaina TM. 2009. Proximate
Composition, Fatty Acid and Amino Acid Profiles of selected Malaysian
freshwater fish. Malay Fish J 8(1): 7-16.
Ghassem M, Said M, Babji AS. 2010. Proteolytic activity of Haruan sarcoplasmic and
myofibrillar proteins. Proceeding of International Seminar. PATPI 2010- Emerging
issues and technology developments in food and ingredients. 2010 September 29-
30, Jakarta International Expo, Arena PRJ Kemayoran Jakarta-Indonesia.
GME. 2008. Gelatin Manufacturers of Europe. Available from: www.gelatine.org/en/
gelatine/overview/127.htm. Accessed March 15 2008.
Glin I, Oktay M, K1reci E, Kfreviolu I. 2003. Screening of antioxidant and
antimicrobial activities of anise (Pimpella anisum L.) seed extracts. Food Chem
83:371-82.
Hsu SY, Chung HY. 2001. Effects of j-carrageenan, salt, phosphates and fat on qualities of
low fat emulsified meatballs. J Food Eng 47(2):11521.
Huda-Faujan N, Noriham A, Norrakiah AS, Babji AS. 2007. Antioxidative activities of
some Malaysian herbal water extracts. ASEAN Food J. 14(1):61-68.
The 12
th
16 -18 June, 2011

122

Imeson AP. 2000. Carrageenan. In: Phillips GO, Williams PA, editors. Handbook of
hydrocolloids. Cambridge: Woodhead. p 87102.
Jimnez-Colmenero F, Cofrades S, Lpez-Lpez I, Ruiz-Capillas C, Pintado T. Solas MT.
2010. Technological and sensory characteristics of reduced/low-fat, low-salt
frankfurters as affected by the addition of konjac and seaweed. J Meat Sci 84:356
63.
Kim SK, Kim YT, Byun HG, Nam KS, Joo DS, Shahidi F. 2001. Isolation and
characterization of antioxidative peptides from gelatin hydrolysate of Alaska
Pollack skin. J Agri Food Chem 49:1984-9.
Kristinsson HG, Rasco BA. 2000. Fish protein hydrolysates: production, biochemical, and
functional properties. Crit Rev Food Sci Nutr 40:4381.
Lachance PA, Fisher MC. 2005. Reinvention of the food guide pyramid to promote health.
Adv Food Nutr Res 49:139.
Li GH, Le GW, Shi YH, Shrestha S. 2004. Angiotensin Iconverting enzyme inhibitory
peptides derived from food proteins and their physiological and pharmacological
effects. Nutr Res 24:46986.
Marcone Massimo F. 2005 Characterization of the edible bird's nest the "Caviar of the
East". Food Res Int 38(10):1125-34.
Meisel H, Walsh DJ, Murry B, FitzGerald RJ. 2005. ACE inhibitory peptides. In: Mine Y,
Shahidi F, editors. Nutraceutical proteins and peptides in health and disease. Boca
Raton, FL: CRC Press. p. 269315.
Mendis E, Rajapakse N, Byun HG, Kim SK. 2005. Investigation of jumbo squid
(Dosidicus gigas) skin gelatin peptides for their in vitro antioxidant effects. Life Sci
77: 2166-78.
Mielnik MB, Aaby K, Skrede G. 2003. Commercial antioxidants control lipid oxidation in
mechanically deboned turkey meat. Meat Sci 65: 1147-55.
Noriham A, Babji AS, Aminah A. 2004a. Determination of antioxidative activities of
selected Malaysian plant extracts. ASEAN Food J 13(4): 193-99.
Noriham A, Babji AS, Aminah A. 2004b. Antioxidant and antimicrobial effects of plant
extracts in chicken sausage. ASEAN Food J 13(4): 211-218.
Okamoto A, Hanagata H, Kawamura Y, Yanagida F. 1995. Anti-hypertensive substances
in fermented soybean, Natto. Plant Foods Hum Nutr 47:3947.
See SF, Hong PK, Ng KL, Wan Aida WM, Babji AS. 2010. Physicochemical properties of
gelatins extracted from skins of different freshwater fish species. Int Food Res J 17:
809-816.
Soo WJ. 2009. Youthful beauty. The Sun 28 Jul 2009.
Van de Velde F. 2008. Structure and function of hybrid carrageenans. Food Hydroco
22(5):72734. Vermeirssen V, Camp JV, Verstraete W. 2004. Bioavailability of
angiotensin I converting enzyme inhibitory peptide. Br J Nutr 92:357366.
Wang H, Nair MG, Straburg GM, Booren AM Gray JI. 1999. Antioxidant polyphenols
from tart cherries (Prunus cerasus). J Agric Food Chem 47: 840-44.
The 12
th
16 -18 June, 2011

123

OA-254

Adherence Properties of Lactic Acid Bacteria Isolated from Breast Milk
Lilis Nuraida
1,2*
, Dhenok Anggraeni
3
and Ratih Dewanti-Hariyadi
1,2
1
Southeast Asia Food and Agricultural Science and Technology (SEAFAST) Center, IPB, Indonesia
2
Department of Food Science and Technology, IPB, Indonesia
3
Alumni of Graduate School majoring in Food Science, IPB, Indonesia
*
Corresponding author: lilis@seafast.org

Abstract
The ability of bacteria to adhere to intestinal cells has been considered as one of selection
criteria for probiotic strain. The aim of this study was to evaluate the adherence properties
of nine lactobacilli isolated from breast milk. Those isolates are able to resist low pH and
bile acid. The present study aimed to evaluate adherence properties including
hydrophobicity, autoaggregation, and adhesion to the surface of rat intestine; and to
evaluate competition of adhesion between lactobacilli and Enteropathogenic E. coli
(EPEC) including exclusion and displacement. All of nine lactobacilli tested show
hydrophilic properties and have low autoaggregation ability. Exposure of rat intestine to
suspension of lactobacilli raised the total number of lactic acid bacteria (LAB) indicating
the attachment of lactobacilli isolates on the surface of rat intestine while reducing the total
number of indigenous E. coli indicating the LAB isolates were able to displace indigenous
E. coli on the surface of rat intestine. The most adhesive lactobacilli were A27 followed by
B16, R14, and R23. The ability of lactobacilli to compete for adhesion on the surface of rat
intestine was strain dependent. The lactobacilli were able to compete with EPEC for
adhesion at the dose of LAB higher than EPEC, i.e. 10
8
cfu/ml vs 10
6
cfu/ml. The higher
the number of lactobacilli exposed, the higher the number of adhering cells. The exclusion
and displacement test using R23 and B16 showed that the lactobacilli tested were likely
able to attach well on the surface of rat intestine. Only small amount of B16 (0.2 log
cfu/m
2
) was excluded by EPEC, however, also only small amount of attached EPEC could
be displaced by B16. R23 was also not excluded by EPEC, but was unable to displaced
attached EPEC. Different binding site or mechanism may involve in the attachement of
LAB isolates and EPEC. This finding also confirm that cell surface characteristic and
adherence ability are strain-dependent.

Keywords: Probiotic, hydrophobicity, autoaggregation, adherence, competition

Introduction
Probiotic is defined as life microorganism when ingested in sufficient amount will confer
beneficial effect to the host (FAO/WHO, 2002). Several criteria have to be met for
selecting probiotic strains. Those include acid and bile tolerance, survival through the
gastrointestinal track, ability to adhere to intestinal surfaces, exhibiting antimicrobial
activity against potential pathogenic bacteria and good technological properties
(Ouwehand and others, 2001).

The 12
th
16 -18 June, 2011

124

Adherence bacteria to mucosal surfaces is prerequisite for transient colonization and one of
probiotic criteria to be able to control the balance of intestinal microbiota (Juntunen and
others 2001). Adhesion ability in lactobacilli is species-specific and there is positive
correlation between adhesion and competitive exclusion ability of lactobacilli (Li and
others 2007). Lee and others (2003) characterised mechanism of adhesion showed that
lactobacilli were able to compete with, exclude and displace pathogenic bacteria when
there were incubated together with the degree of adhesion was strain dependent.

Servin and Coconnier (2003) stated that the microbial adhesion process of lactic acid
bacteria includes passive forces, electrostatic interactions, hydrophobic, steric forces,
lipoteichoic acids and specific structures. The ability to adhere to ephitelial cell has been
shown to correlate with hydrophobicity (Zavaglia and others 1998). In addition to those,
mechanism of adhesion also involve binding on specific receptor (Wardstrom and Ljungh
2006), and their ability to survive and presist in the gastrointestinal track correlates with
their ability to aggregate (Jankovic and others 2003, Morelli dan Callegari 2006). The cell
surface physicochemical properties of lactobacilli is particular to the bacterial species
(Pelletier and others 1997), hence adherence properties is strain dependent.

Human breast milk has been reported as source of lactic acid bacteria potential as
probiotics. B. longum was the most widely found in breast milk, followed by B. animalis,
B. bifidum, B. catenulatum (Gueimonde and others 2007). Occurrences of lactobacilli in
human breast milk were also reported such as L. gasseri, L. fermentum, L. salivarius
(Martin and others 2005). They showed that the lactobacilli isolates had potency as
probiotic at least, similar to that of the strains commonly used in commercial probiotic
products. Study done in our laboratory showed that most of lactobacilli and bifidobacteria
isolated from 28 lactating mothers in Bogor area showed good survival in pH 2 and bile
salt, and showed antimicrobial activities against pathogenic bacteria (Nuraida and others
2007). To further explore the potency of those lactic acid bacteria, the adherence properties
including hydrophobicity, autoagregation, adhesion and ability to compete for adhesion
with pathogenic bacteria, exclusion and displacement were evaluated using intestinal
surface of rats as a model.

Bacterial culture
Nine lacticobacilli i.e. L. rhamnosus A15, Lactobacillus A27, L. rhamnosus A29, L.
rhamnosus R14, L. rhamnosus R21, L. rhamnosus R23, L. rhamnosus R26, L. rhamnosus
B10, L. rhamnosus B13, dan L. rhamnosus B16, previously isolated from breast milk were
studied. Enteropathogenic Escherichia coli (EPEC) K1.1 used for competition study for
adhesion was obtained from Biotechnology Inter University Center, Bogor Agricultural
University. All cultures were maintained in stock culture kept in 20% glycerol at -20
o
C.
When required the lactobacilli were grown in MRS broth while EPEC in NB at 37
o
C.

Cell surface hydrophobicity (modification of Misra and Prasad, 2005)
Hydrophobicity of cell surface was assesed based on MATS (Microbial Adhesion to
Solvent). Lactic acid bacteria were harvested after growth for 16-18 h at 37
o
C by
centrifugation for 15 min at 5000 rpm, then washed twice in PUM buffer (composition
(g/l): K
2
HPO
4
.3H2O: 22.2, KH
2
PO
4
: 7.26, urea: 1.8, MgSO
4
.7H
2
O: 0.2, pH 7.1) and
finally suspended in the same buffer at the level of 10
8
cfu/ml. The absorbance of the
The 12
th
16 -18 June, 2011

125

suspension was measured at 600 nm (A). Five ml of cell suspension in PUM buffer was
taken into clean and dry round bottom test tubes. Then, 1 ml of different hydrocarbon
(xylene, ethyl acetate and chloroform) was added and mixed by vortexing at high speed
for 1 min. The tubes were left undisturbed for 1 h at 37
o
C to allow the phase separation.
The lower aqueous phase was carefully removed with a sterile pasteur pipette and final
absorbance (A
0
) was recorded at 600 nm. The decreased absorbance in aqueous phase was
taken as measure of cell surface hydrophobicity (H%), calculated using following
equation: H% = (A-A
0
)/A x 100.

Autoaggregation properties (Kos and others 2003)
The culture was grown in MRSB for 18 h at 37
o
C. The pellets were washed twice in
phosphate-buffered saline (PBS) and re-suspended in similar solution to reach number of
cells of 10
8
cfu/ml. Autoaggregation was determined by measuring their absorbancy at 0 h
(A
0
) and after 5 h (A
t
) incubation at room temperature. Measurement of absorbancy was
done by taking 0.1 ml of upper aqueous phase and diluted with 3.9 ml PBS. The
absorbance was measured at 600 nm. Percentage of autoaggregation was calcuates using
the following formula: Autoagregasi (%) = 1- (A
t
/A
0
) x 100.

Adhesion of lactobacilli isolates to the surface of rats intestine
Rat intestine was taken from 8 weeks old rat (Sprague Dawley). The intestine of were cut
for about 10 cm long, opened and washed three times with PBS. To evaluate the adhesion,
the LAB isolates were grown for 24 h in MRSB, then centrifuged at 5000 rpm for 15 min.
The cell pellet was resuspended in PBS to reach the number of LAB of 10
6
cfu/ml. A
piece of rat intestine was placed in petri dish and 10 ml of LAB cell suspension was added.
The suspension was then incubated at room temperature for 60 min. At the end of
incubation period, the rat intestine was washed twice with PBS. To calculate the number of
lactic acid bacteria (LAB) and E. coli on the surface of rat intestine, 1 cm
2
of rat intestine
surface was swabbed using sterile cotton bud and then the cells were released in 0.85%
NaCl and subsequent dillution was done. The LAB count was performed in MRSA while
E. coli in EMBA. The evaluation was done for 3 replicates. A set experiment using PBS
as exposure media was used as control. The number of total LAB and E. coli after
incubation were enumerated using swab methods in similar way as above. The increase in
total LAB or the decrease of E. coli indicating the number of attached lactobacilli and
displaced indigenous E. coli was calculated as the difference between LAB or E. coli count
of control and the count after exposure of rat intestine to the lactobacilli cell suspension.

Competition for adhesion between lactobacilli isolates and EPEC K1.1
The experiment was done in three set of intestine. The first set was expose to LAB cells
suspension as control LAB, the second set was exposed to EPEC as control EPEC and the
third set was exposed to a mixture of LAB and EPEC. Each set of experiment used three
pieces of rat intestine. The number cell of bacteria in exposure media was 10
6
cfu/ml.
Incubation was done for 60 min at room temperature. Following incubation, the number of
LAB and/or E. coli were enumerated as described above. To evaluate the effect of dose on
the ability to adhere, the rat intestine was exposed to the higher number of lactobacilli i.e.
10
8
cfu/ml and compared with 10
6
cfu/ml.

Exclusion and displacement
The 12
th
16 -18 June, 2011

126

For evaluation of exclusion, the lactobacilli was allowed to adhere to the surface intestine
first and then EPEC was added subsequently, meanwhile for displacement study, EPEC
was allowed to adhere first and then lactobacilli was added. Four set of rat intestine were
used, i.e. the first set was exposed to lactobacilli as control LAB, the second set was
exposed to EPEC as control EPEC, the third set was exposed to lactobacilli and
subsequently to EPEC, the fourth set was exposed to EPEC and subsequently to
lactobacilli. The exposure dose of lactobacilli was 10
8
cfu/ml, while that of EPEC was 10
6

cfu/ml. Incubation for each exposure was 60 min. For exclusion and displacement study,
prior to exposure to the second bacteria, the intestine was washed twice with PBS to wash
un-attached the first bacterial cell from the surface. Enumeration of LAB and/or E. coli
was done similarly with the adhesion study describe above.

Hydrophobicity of lactobacilli isolated from breast milk
Three different solvent were used to evaluate hydrophobic/hydrophobic cell surface
properties and acidic-basic character. The results (Figure 1) revealed that most isolates
showed negative affinity to xylene. Low affinity to xylene indicate hydrophylic properties
of cell surface. While Lactobacillus A15 and R23 indicated being slightly hydrophobic as
they have positive affinity to xylene i.e. 15.24% and 9.43% respectively. The present
finding is in accordance with previous study of Pelletier and others (1997) showed that of 8
Lactobacillus (L. casei, L. paracasei, dan L. rhamnosus), all are hydrophylic as shown by
low affinity (2.7 26.5%) on non-polar solvent.

Figure 1 Affinity of LAB ioslates to chloroform, ethylacetate and xylene

Affinity to chloroform shows the character of cell surface as electron donor and basic
character, while to ethyl acetate shows as electron acceptor and acid character (Pelletier
and others 1997). Lactobacillus A15 and R23 shows higher affinity to chloroform as
compared to xylene and ethyl acetate indicating that the cell surface of A15 and R23 act as
electron donor and has basic character. Meanwhile, lactobacilli A27, A29, B10, B13, B16,
R14, and R26 shows higher affinity to ethyl acetate indicating the cell surface character of
most lactobacilli isolates being electron acceptor and acidic. Only isolate A15 and R23
were in accordance with finding of Pelletier and others (1997) and Hamadi and others
(2004) showing that at neutral pH, the microbial surface has character as electron donor.
The character of electron acceptor was observed on low pH due to deprotonation.

Autoaggregation properties of lactobacilli isolated from breast milk
The 12
th
16 -18 June, 2011

127

The results revealed that autoaggregation ability was relatively low, i.e. ranged between
4.13% - 39.10%. The highest autoaggreation ability shows by alctobacilli R23 (39.10%),
followed by B13 (31.8%), B16 (29.42%), B10(28.04%), R14 (26%), and A15 (14.96%).
The lowest autoaggregation ability was shown by A29 (4.13%) and R26 (4.93%). The
results is in accordance with hydrophobic properties of all lactobacilli tested. Morelli dan
Callegari (2006) stated that strains with high autoaggreagtion properties generally show
hydrophobic properties of cell surface.

Adhesion of lactobacilli isolates to the surface of rats intestine
Figure 2 shows that the highest adhesion was shown by Lactobacillus A27 followed by
Lactobacillus B16, R14, dan R23. Exposure of rat intestine to the lactobacilli cell
suspension decreased the number of indigenous E. coli, except for isolate B10. This
indicate the LAB isolated from breast milk were able to displace indigenous E. coli.
All lactobacilli were able to adhere to the surface of rat intestine regardless their
hydrophilic properties. The higher adhesion was observed for Lactobacillus A27 and B16
which had hydrophilic properties. In contrast, adhesion ability of Lactobacillus A15 and
R23 with higher affinity to xylene was lower than Lactobacillus A27 and B16. As all
isolates shows more hydrophilic properties, this study revealed that adhesion ability of
lactobacilli isolates was not hydrophobic-dependent. This study was in aggreement with
study by Ouwehand and others (1999) showing no correlation between hydrophobicity of
cell surface and adhesion ability. Research done by Collado and others (2007) also showed
that there was no correlation between autoaggreagtion ability of 4 strains of commercial
probiotics (L. rhamnosus GG, L. rhamnosus LC705, B. breve 99, and P. freudenreichii ssp.
shermanii JS) and their attachments on human intestinal mucosa. L. rhamnosus LC705 had
the highest autoaggreagtion ability but its attachment on mucosa was low i.e. 1.2%.

Change in LAB count Change in E. coli count

Figure 2 Change of LAB and E. coli count on the surface of rat intestine after exposure to lactobacilli
suspension for 60 min.

Competition for adhesion between lactobacilli isolates and EPEC K1.1
This study used four lactobacilli isolates i.e. A27, B16, R14, and R23, while EPEC K1.1.
used as competitor. The results (Figure 3) shows that when the surface of intestine was
exposed to the mixture of lactobacilli and EPEC, both lactobacilli isolates and EPEC were
able to adhere to the surface of rat intestine. When the surface of rat intestine was exposed
The 12
th
16 -18 June, 2011

128

to lactobacilli alone, the number of E. coli decreased, indicating that the lactobacilli
isolates were able to exclude indigenous E. coli. However, when the surface of intestine
was exposed to the mixture of lactobacilli and EPEC, only Lactobacillus B16 could
slightly reduced the number of attachement of EPEC as compared to the exposure on
EPEC alone.

The lactobacilli isolates were able to compete with indigenous E. coli as shown with the
decrease of the number of E. coli when the surface of rat intestine was exposed to
lactobacilli alone. However, when the surface of rat intestine was exposed to the mixture
of lactobacilli and EPEC, Lactobacillus A27, B16 and R14 were unable to decrease the
number of EPEC attaching to the surface of rat intestine. In contrast, the number of LAB
attached to the surface decreased as compared to control (exposed to lactobacilli alone),
except for Lactobacillus A27. Only isolate B16 was able to slighlly decreased the number
of EPEC as compared to control (exposed to EPEC alone). Lee and Puong (2002) stated
that the degree of competition was strain-dependent and determined by the affinity of
adhesin on respective bacterial surface for stero-specific receptors. Study by Lee and
others (2003) revealed that each lactobacilli could only compete with limited range of
gatrointestinal bacteria for adhesion site.


Figure 3 Change in LAB and E. coli count on the surface of rat intestine after exposure to EPEC,
LAB, and mixture of LAB and EPEC

To evaluate the effect of dose on the ability of lactobacilli to attached and to compete with
EPEC, two isolates i.e. Lactobacillus B16 and R23 were used as models. The results
(Figure 4 ) show that increasing number of lactobacilli cells exposed to the surface of rat
intestine, i.e. from 10
6
cfu/ml to 10
8
cfu/ml increased the reduction of indigenous E. coli
while increased the number of attached LAB cells. The observation was in aggrement with
the results of Tuomola dan Salminen (1998) evaluating adhesion of 12 strains of
Lactobacillus to Caco-2 cells that highly depent on lactobacilli cells concentration. The
higher the number of lactobacilli, the higher the number of cells attached.
On the study of competition for adhesion between Lactobacillus B16 and EPEC, the results
(Figure 4) shows that increasing number of lactobacilli exposed to the surface of rat
intestine (10
8
cfu/ml) improved the competition ability as shown by the decrease in
attached E. coli as compare to that of 10
6
cfu/ml. Increasing the dose of lactobacilli also
increase the number of Lactobacillus B16 attached on the surface of rat intestine for both
The 12
th
16 -18 June, 2011

129

when the lactobacilli exposed alone or as mixture with EPEC as competitor (Figure 4).
Increasing the dose of Lactobacillus R23 cells exposed to the surface of rat intestine only
increased reduction of indigenous E. coli and increased the number of LAB attached, but
did not decreased the number of EPEC when exposure was done to the mixture of cells of
R23 and EPEC (data not shown). This results revealed that increasing the dose of isolate
R23 only increased attachement, but did not improve competition ability toward EPEC.
This finding also confirmed that competition ability for adhesion is strain-dependent.

Figure 4 Effect of exposure dose of LAB isolate B16 on the change of number of LAB and E. coli on
the surface of rat intestine

Exclusion and displacement
Exclusion study to evaluate if the attached LAB could be replaced by EPEC, while
displacemnet study was to evaluate the ability of LAB to displace attached EPEC on the
surface of rat intestine. Study using isolate B16 (Figure 5) shows that only small amount of
the attached isolate B16 on the surface of rat intestine was excluded by EPEC, i.e. only 0.2
log cfu/cm
2
as compared to exposure to LAB alone as control. Similarly, only small
amount EPEC that has attached to the surface of rat intestine could be displaced by B16,
i.e. only 0.2 log cfu/cm
2
as compared to esposure to EPEC alone as control. Observation
on R23 shows that the number of R23 attached on the surface of rat intestine did not
change following exposure to EPEC cell suspension indicating that no exclusion of B16 by
EPEC. The results indicate that both lactobacilli and EPEC were able to attached to the
surface of rat intestine with attached EPEC or attached LAB on it, but they did not
displaced each other. The present finding suggest that the adhesin on respective bacterial
surface may differ, hence they did not compete for similar receptor. Lee and others (2003)
stated that LGG was unable to displaced bacterial cells attached to Caco-2 cells unless the
bacterial cells detaches from the receptor and the binding of LGG hinder the re-attachment
of bacterial cell to the receptor. The indigenous bacteria attached to the surface of rat
intestine may also affect the ability of lactobacilli to displace EPEC. The number of LAB
present on the surface of rat intetsine was 10
4
-10
5
cfu/cm
2
. In contrast, the number of
indigenous E. coli present on the surface of rat intetstine was quite low, i.e. about 10
2

cfu/cm
2
, hence the space for adherence of E. coli still available while for LAB may be
saturated.
The 12
th
16 -18 June, 2011

130

Lactobacillus B16 Lactobacillus R23

Figure 5 Change in LAB and E. coli count on the surface of rat intestine after exposure to EPEC,
Lactobacillus B16 or R23, and Lactobacillus B16 or R23 and subsequently to EPEC, EPEC
and subsequently to Lactobacillus B16 or R23

Correlation between adherence characteristic and the functional properties of LAB as
probiotics still need further study. Several strains with low adherence properties both in
vitro and in vivo showed positive effect on the host (Saarela and others 2000). Study done
by Nuraida and others (2010) employed three Lactobacillus isolates used in the present
study i.e. R14, R23 and B16 showed that they were able to prevent diarrhea on rat due to
infection of EPEC K1.1. when the lactic acid bacteria was regularly introduced prior to
infection. Isolate R23 showed the best capabilities of preventing diarrhea in rats compared
to two other isolates. L. acidophilus LA1, L. casei Shirota, and Lactobacillus GG have
been proved to deliver beneficial health effect although their ability to attached to Caco-2
cell were low, i.e. LA1 (12.6%), Shirota (3.2%), and LGG (9.7%) (Tumuola and Salminen
1998). The present finding suggest dose and method of delivery should be considered
when developing probiotic product using these lactobacilli isolates.

Acknowledgements
The research team acknowledge SEAFAST Center, Bogor Agricultural University for
research funding and Dr. Sri Budiarti of Inter University Center of Biotechnology, Bogor
Agricultural University for EPEC K1.1.

References
Collado MC, Meriluoto J, Salminen S. 2007. Development of new probiotics by strain
combinations: is it possible to improve the adhesion to intestinal mucus? J Dairy
Sci. 90:27102716.
FAO/WHO. 2002. Guidelines for The Evaluation of Probiotics in Food. London, Ontario,
Canada.
The 12
th
16 -18 June, 2011

131

Gueimonde M, Laitinen K, Salminen S, Isolauri E. 2007. Breast milk: A source of
Bifidobacteria for infant gut development and maturation? Neonat 92(1):64-66.
Hamadi F, Latrache H, El Ghmari A, El Louali M, Mabrrouki M, Kouider N. 2004. Effect
of PH and ionic strength on hydrophobicity and electron donor and acceptor
characteristics of Escherichia coli and Staphylococcus aureus. Annals of Microbiol
54 (2): 213-225.
Jankovic I, Venura M, Meylan V, Rouvet M, Elli M, Zink R. 2003. Contribution of
aggregation-promoting factor to maintenance of cell shape in Lactobacillus gasseri
4B2. J Bacteriol 185: 32883296.
Juntunen M, Kirjavainen PV, Ouwehand AC, Salminen SJ, Isolauri E. 2001. Adherence of
probiotic bacteria to human intestinal mucus in healthy infants and during rotavirus
infection. Clinical And Diagnostic Laboratory Immuno 8: 293296.
Kos B, Suskovic J, Vukovic S, Simpraga M, Frece J, Matosic S. 2003. Adhesion and
aggregation ability of probiotic strain Lactobacillus acidophilus M92. J Appl
Microbiol 94: 981-987.
Lee YK, Puong KY. 2002. Competition for adhesion between probiotics and human
gastrointestinal pathogens in the presence of carbohydrate. Br J Nutr 88: S101-
S108.
Lee YK, Puong KY, Ouwehand AC, Salminen S. 2003. Displacement of bacterial
pathogens from mucus and Caco-2 cell surface by lactobacilli. J Medical Microbiol
52: 925930.
Li XJ, Yue LY, Guan XF, Qiao SY. 2007. The adhesion of putative probiotic lactobacilli
to cultured epithelial cells and porcine intestinal mucus. Appl Microbiol 104:
10821091.
Martin R, Olivares M, Marn ML, Fernndez L, Xaus J, Rodrguez JM. 2005. Probiotic
potential of 3 Lactobacilli Strains Isolated from Breast Milk. J Human Lactation
21(1):8-17. available from http://jhl.sagepub.com. Assesed October 29, 2009
Mishra V, Prasad DN. 2005. Application of in vitro methods for selection of Lactobacillus
casei strains as potential probiotics. Int J Food Microbiol 103: 109-115.
Morelli L, Callegari ML. 2006. Taxononomy dan biology of probiotics. In Goktepe I,
Juneja VK, Ahmedna M, editors. Probiotics in Food Safety and Human Health.
Boca Raton: Taylor and Francis.
Nuraida L, Soetikno S, Hartanti AW. 2007. Lactic acid bacteria and bifidobacteria profile
of breast milk and their potency as probiotics. Programme and Abstracts: 10
th

ASEAN Food Conference 2007 Food for Minkind-Contribution of Science and
Technology.
Nuraida L, Hartanti Aw, Hana, Prangdimurti E. 2010. Potency of lactic acid bacteria
isolated from breast milk to prevent diarrhea caused by infection of Epec K.1.
Program. The 3rd International Conference of Indonesia Society for Lactic Acid
Bacteria. 21 January 21-22, 2010.
Ouwehand AC, Kirjavainen PV, Grondlund MM, Isolauri E, Salminen SJ. 1999. Adhesion
of probiotic micro-organisms to intestinal mucus. Int Dairy J 9: 623-630.
Ouwehand AC, Tuomola EM, Ikko ST, Salminen S. 2001. Assessment of adhesion
properties of novel probiotic strains to human intestinal mucus. Int J Food
Microbiol 64: 119-126.
Pelletier C, Bouley C, Cayuela C, Bouttier S, Bourlioux P, Bellon-Fontaine M. 1997. Cell
surface characteristics of Lactobacillus casei subsp. casei, Lactobacillus paracasei
The 12
th
16 -18 June, 2011

132

subsp. paracasei, and Lactobacillus rhamnosus Strains. Appl and Environment
Microbiol 63: 1725-1731.
Saarela M, Mogensen G, Fonde R, Matto J, Mattila-Sandholm T. 2000. Probiotic bacteria:
safety, functional, and technological properties. J Biotechnol 84: 197215.
Servin AL and Coconnier MH. 2003. Adhesion of probiotic strains to the intestinal mucosa
and interaction with pathogens. Best Practices & Research Clinical
Gastroenterology 17(5):741-54.
Tuomola EM, Salminen SJ. 1998. Adhesion of some probiotic and dairy Lactobacillus
strains to Caco-2 cell cultures. Intl J Food Microbiol 41:4551.
Wadstrom T , Ljungh A. 2006. Lactic acid bacteria as probiotic. Current Issues in
Intestinal Microbiol 7: 73-90.
Zavaglia AG, Kociubinzki G, Perez P, De Antoni G. 1998. Isolation and characterization
of Bifidobacterium strains for probiotic formulation, J Food Protect 61: 865-873.

The 12
th
16 -18 June, 2011

133

OB : Food Productivity Improvement
The 12
th
16 -18 June, 2011

134

OB-9

Development of a Small-Scale Ohmic Pasteurization System for Flavor
Retention in Fruit Juice Production

P. Somboonsilp
1
, S. Tia
2
, T. Yoovidhya
1,
*

1
Department of Food Engineering;
2
Department of Chemical Engineering
King Mongkuts University of Technology Thonburi
126 Pracha U-tid Road, Bangmod, Tungkru, Bangkok 10140, Thailand
* Corresponding author: tipaporn.yoo@kmutt.ac.th

Abstract
Conventional pasteurization of fresh fruit juices generally results in flavor deterioration.
Ohmic heating, on the other hand, despite its comparative capability to inactivate
microorganisms, results in much less flavor deterioration. In the present study, an ohmic
pasteurization system with a static ohmic heater was applied to four fruit juices, namely,
orange, pineapple, guava and coconut juices. The system was operated in a batch mode
with a capacity of 400 mL; electric field strength of about 23 V/cm was used. A proper
design and fine tuning of the system was based on the juice electrical conductivity and the
changes in the appearance of the juices. The concentration of the tested fresh fruit juices
was varied within the range normally found in the market by adding sugar and/or water.
The quality of the treated fruit juices, in terms of color and flavor as well as shelf-life, was
also evaluated. The results were compared with those of fresh juices and juices treated by a
conventional heating process. The results indicated that the temperature dependency of the
electrical conductivity of all the juice samples was linear. Pure fruit juices exhibited the
highest values of the electrical conductivity when compared with that of the juices
containing sugar and/or water. In terms of quality of fruit juice, ohmic pasteurized fruit
juice could be stored at 5C for 7 days like the conventional pasteurized fruit juice but
color and flavor deterioration was less.

Keywords: Electrical conductivity, flavor deterioration, fruit juices, pasteurization, static
ohmic heater

Introduction
Fruit juices play an important role in a healthy diet because they provide a variety of
nutrients found naturally in fruits. Nowadays, there are two main types of fruit juice
product in the market; freshly squeezed and pasteurized fruit juices. The former gives
natural flavor but has short shelf-life in refrigeration. The later has longer shelf-life but
flavor deterioration occurs due to heat treatment employed. Heat treatment by conventional
method destroys pathogens and spoilage microorganisms but causes nutrient loss and
flavor deterioration to the juices due to high temperature and long time of processing.
Ohmic heating can inactivate microorganism like the conventional pasteurization process
(Sastry 1994, Khunvirojpanich 2005) but flavor deterioration was less as reported by
Leizerson and Shimoni (2005).

Ohmic heating is an efcient thermal processing method where heat is generated by
passing electricity through the food material. Electrical energy is dissipated into heat,
which results in rapid and uniform heating. Therefore, the quality of food especially
The 12
th
16 -18 June, 2011

135

volatile components can be better preserved compared with using the conventional method
(Sastry 1994).. The aim of this study is to design and develop a small-scale ohmic
pasteurization system for flavor retention in commercial production of fruit juices.

Fruit juices studied were orange, pineapple, guava and coconut juices since they are the
most favorite for Thai consumers. The electrical conductivity and appearance changes of
each juice were observed when electricity was passed through it, using static ohmic heater
developed by Khunvirojpanich (2005).

For design and development of ohmic pasteurization system, the type of ohmic heater
desired should be easy and safe to operate and control by the operator. . The process used
for fruit juice pasteurization in this study was 70C for 2 min. After heat treatment, the
juice was cooled in ice slush (water mixed with crush ice).

The quality of the three fruit juices; freshly squeezed juice, juices pasteurized by ohmic
heating and by conventional method, was evaluated in terms of color and flavor. Color
measurement was conducted using a colorimetric spectrophotometer (Hunter Lab, model
Color Quest XE, Virginia, USA) based on three-color co-ordinates, namely L, a and b,
which were measured in transmittance mode and light source used was D65. Flavor of fruit
juice was described in terms of volatile compounds using solid-phase micro-extraction Gas
Chromatograph with Mass spectrometer (SPME-GC/MS). 10 mL of fruit juice was
transferred into 40 mL vial containing magnetic stirring bar. The sample vial was air
tightly sealed by septum and an aluminum cap. The SPME fiber coated with 50/30 m
divinylbenzene-carboxen-polydimethylsiloxane (DVB/CAR/PDMS) (Supelco Inc.,
Bellefonte, PA, U.S.A.) was manually inserted into the headspace of sample vial. The
SPME fiber coating which isolated headspace flavor compounds by desorption was
injected into the GC injection port. The analysis condition of volatile components was
varied for each type of fruit juice. For orange and pineapple juices, the volatile components
were absorbed for 20 min at 60C and desorbed in the injector port at 220C for 2 min. The
desorbed flavor compounds were separated by capillary column (30 m 0.53 mm i.d.)
coated with a 2.65 m film of 5% phenyl substituted methylpolysiloxane (Supelco). The
temperature was programmed from 60 to 120C at 10 C/min, then increased to 200 C at
4 /min, and held for 10 min at the final temperature, detector temperature was 250C (Jia
and others 1998). For guava juice, the volatile component was absorbed for 20 min at 60C
and desorbed in the injector port at 300C for 5 min. The desorbed flavor compounds were
separated by CP-Sil 8 fused silica capillary column (30 m, 0.25 mm i.d., 0.25 m phase
thickness) (Supelco). The temperature was programmed at 40C and held for 2 min and
then increased to 200 C at 5 C/min; this temperature was held for 2 min and then
increased to 260 C at 30 C/min, detector temperature was 250C (Carasek and
Pawliszyn, 2006). For coconut juice, the volatile component was absorbed for 45 min at
42-45C and desorbed in the injector port at 220C for 5 min. The desorbed flavor
compounds were separated by DB-5 column (Zebron, 30 m 0.32 mm ID 0.50 m FT).
The temperature was programmed from 40 to 265C at 7 C/min, and held for 5 min at the
final temperature, detector temperature was 270C (Damar 2006).

Final evaluation was the shelf-life at storage temperature of 5C; all juice types were
microbiologically evaluated every day using spread plate method by AOAC (Cunniff
The 12
th
16 -18 June, 2011

136

1995) and compared with the standard values (Thai Community Product Standard). Total
plate count (TPC) and yeast and mold were determined. The agar used for TPC
determination was plate count agar or PCA (Difco Laboratories, Detroit, Mitch., U.S.A.)
and dextrose agar or PDA (Difco Laboratories) was used for yeast and mold determination.

Results and discussion
1. The electrical conductivity of fruit juice
The electrical conductivity and appearance change of 4 fruit juices were observed. Figure 1
shows that the temperature dependence of the electrical conductivity of four fruit juice
samples was linear. And they were suitable for pasteurization by ohmic heating because
they did not leave any deposits on the electrode and no obvious appearance changes of the
juices were observed.

Moreover, the electrical conductivity of each fruit juice with added sugar and/ or water was
studied. The results indicated that the pure fruit juices showed the highest electrical
conductivity values when compared with those containing added sugar or water. This is
because sugar did not ionize into free ion but obstructed the movement of free ions in the
juice. So, the electrical conductivity of the juice was decreased. Regarding water, which is
a pure substance and does not contain any free ions, it could therefore dilute the free ions
in the juice and reduced the electrical conductivity.

Figure 1 Electrical conductivity of 100% juice with different TSS (a) orange juice, (b) pineapple
juice, (c) guava juice, (d) coconut juice

The 12
th
16 -18 June, 2011

137

From the experimental data, electrical conductivity of fruit juices could be predicted using
dimensionless form of electrical conductivity and temperature:
* *
aT = (1)
where
*
20
/ 1 = (1.1)
and
*
/ 20 1 T T = (1.2)
and the regression equation between electrical conductivity at 20 C and the amount of
sugar and/ or water added:
The electrical conductivity of orange juice could be written as:
* = 0.544T* and
20
= 0.3251.16 S0.115 W (2)
The electrical conductivity of pineapple juice could be written as:
* = 0.512T* and
20
= 0.3260.87S0.142 W (3)
The electrical conductivity of guava juice could be written as:
Fresh guava juice: = 0.010T + 0.235 (4)
Guava juice with added sugar or water:
* = 0.523T* and
20
= 0.3210.302S0.0801W (5)
Guava juice with added sugar and water:
* = 0.523T* and
20
= 0.3670.453S0.116W (6)
And the electrical conductivity of coconut juice could be written as :
Fresh coconut juice: = 0.019T + 0.419 (7)
Coconut juice with added sugar or water:
* = 0.504T* and
20
= 0.6790.719S0.221W (8)
Coconut juice with added sugar and water:
* = 0.504T* and
20
= 0.5240.556S0.144W (9)
where is the electrical conductivity of fruit juice (S/m)
S is the amount of sugar added (kg/ liter of juice)
W is the amount of water added (kg/ liter of juice)

2. Development of ohmic pasteurization system
Concerning development of ohmic pasteurization system; the selected type of ohmic
pasteurizer was the static one because it is easy and safe to operate by the operator. Figure
2 shows a diagram of ohmic heater. It has capacity of about 400 mL per batch. The voltage
used to provide the electric field strength of about 23 V/cm is 300 V.

Figure 2 Diagram of ohmic pasteurizer

The 12
th
16 -18 June, 2011

138

To operate the system, the prepared fresh fruit juice is transferred into the feeding tank
before being moved to the ohmic heater through the pipe and ball valve. The pasteurization
starts by pushing the operating switch at the controller, heat occurs when passing electric
current through the fruit juice. Temperature of the juice increases linearly and electrical
heating stops when fruit juice reaches the desired temperature. The hot juice is then filled
into the cleaned bottle and held for a few minutes to destroy spoilage microorganism. The
pasteurized fruit juice is cooled in the cooling tank containing ice slush. Electrical source
for this system is alternating current of 220 volts which is transformed to alternating
current of 300 volts by the use of transformer. This is to generate electric field strength of
about 23 V/cm. Ohmic pasteurization system is shown in Figure 3.

Figure 3 Ohmic pasteurization system

3. Heating rate comparison between ohmic and conventional heating
Heating rate of ohmic and conventional pasteurization of fruit juice is evaluated. Come-up
time at the slowest heating point of orange juice to reach 70 C by conventional method is
about 2 times longer than ohmic heating, 4 times longer for pineapple juice, 2 times longer
for guava juice and 3 times longer for coconut juice. Due to shorter come-up time of
ohmic heating, it can be concluded that volatile components in fruit juice are better
preserved by ohmic heating than by conventional method due to shorter contact time
between fruit juice and heat.

4. Quality evaluation
Color
For color evaluation, all juices pasteurized by ohmic heating showed the color similar to
the fresh juice more than the juices processed by conventional method with the exception
of coconut juice. It was found that coconut juices processed by ohmic heating turned to
pink color after 2-3 day-storage at 5
o
C. This may be because the shorter heating time by
ohmic was not enough to inactivate the enzyme in coconut juice.

The 12
th
16 -18 June, 2011

139

Flavor
Quality evaluation in terms of volatile components was observed by SPME-GC/MS and
expressed as relative percentage. Volatile components of orange juices are shown in
Figure 4.

Figure 4 % Relative of volatile components in fresh orange juice, ohmic pasteurized and conventional
pasteurized orange juices.

Limonene is hydrocarbon found naturally in citrus fruit (Izquiredo 1993). The minimum
quantity was found in fresh orange juice and it increased due to heat treatment. It increased
to 114% in ohmic pasteurized orange juice and 135% in conventional pasteurized orange
juice. The reason for the increase of limonene in pasteurized orange juice is because
limonene has higher boiling point than other volatile components when heat is applied
during GC-MS analysis, it is then changed into volatile again (Tuner and Harris 2004).
Nonylphenol content is unchanged but acetic acid is decreased after heat treatment.

Figure 5 % Relative of volatile components in fresh pineapple juice, ohmic and conventional
pasteurized pineapple juices.

Change in volatile components of pineapple juice is shown in Figure 5. Most of volatile
components in pineapple juice are decreased after pasteurization. Methyl-3-
methylthiopropionate gave flavor in sweet taste fruit such as pineapple and banana
The 12
th
16 -18 June, 2011

140

(Tokimoto and others 2005). Methyl 3-acetoxyhexanoate and 2,5-Dimethy-4-hydroxy-
3(2H) furanone gave the specific flavor in fruit (Tokimoto and others 2005).

Figure 6 % Relative of volatile components in fresh guava juice , ohmic pasteurized guava juice and
conventional pasteurized guava juice

Most volatiles components in guava juice are decreased after heat treatment as shown in
Figure 6. -copaene is an essential oil found naturally in orange, guava and mango
(Nishida and others 2000). Caryophyllene and -humulene are the essential oils from the
stems and flowers of hemps and rosemary (Jaoui and Kamens 2003). -bisabolene is
present in the essential oils of a wide variety of plants including cubeb, lemon and oregano
(Theophilus and others 1997). In case of nerolidol which is the essential oil of many types
of plants and flowers including ginger, jasmine, lavender, tea tree and lemon grass (Moser
and others 2001), it decreases after ohmic pasteurization but increased after conventional
heating method.

The last fruit juice studied is coconut juice as shown in Figure 7. The quantity of
nonylphenol and benzaldehyde after ohmic pasteurization are decreased but increased after
conventional method. For lauric acid, palmitic acid and stearic acid, these volatiles
increased after ohmic pasteurization but decreased after conventional method. Lauric acid
is the main acid in coconut oil and in palm kernel oil (Dawson and others 2002).
Benzaldehyde is the colorless liquid has a characteristic pleasant almond-like odor.
Palmitic acid is the most common saturated fatty acids found in animals and plants. It is a
major component of the oil from palm trees and coconut (Marmesat and others 2005).
Stearic acid occurs in many animal and vegetable fats and oils, but it is more common in
animal fat than vegetable oil (Emken and others 1994).
The 12
th
16 -18 June, 2011

141

Figure 7 % Relative of volatile components in fresh coconut juice , ohmic and conventional
pasteurized coconut juices

5. Shelf-life evaluation
Shelf-life of 4 fruit juices was determined as microbiological evaluation using spread plate
method and compared with the standard values of Thai Community Product Standard. The
results indicated that ohmic pasteurized fruit juices as well as conventional pasteurized
fruit juices have shelf-life of at least 7 days at 5 C when processed at 70
o
C for 2 min..

References
Association of Analytical chemist. 1995. Official Methods of Analysis of AOAC
International Volume 1, Agricultural Chemicals; Contaminants; Drugs, 16
th
ed.,
Cunniff, P. (Ed.) Arlingtion, p. 8-9.
Carasek E and Pawliszyn A. 2006. Screening of Tropical Fruit Volatile Compounds Using
Solid-Phase Microextraction (SPME) Fibers and Internally Cooled SPME Fiber. J
Agric Food Chem. 26:8688-96.
Damar S. 2006. Processing of Coconut Water with High Pressure Carbon Dioxide
Technology. p. 91-93.
Dawson PL, Carl GD, Acton JC and Han IY. 2002. Effect of lauric acid and nisin-
impregnated soy-based films on the growth of Listeria monocytogenes on turkey
bologna. Poultry Science. 81:721-6.
Emken EA. 1994. Metabolism of dietary stearic acid relative to other fatty acids in human
subjects. Ame J Clinical Nutrition. 60:1023-8.
Izquiredo L, Velez C, Costell E, Orlando L, Nadal MI and Sendra, JM. 1993. The Effect of
Storage on The Constituents That Affet The Aroma of Valencia Orange. J Science
and Food Agri. 61:41-6.
Jaoui M, Kamens RM. 2003. Gas and Particulate Products Distribution from the
Photooxidation of -Humulene in the Presence of NO
x
, Natural Atmospheric Air
and Sunlight. J Atmospheric Chem. 46:29-54.
Jia M, Zhang QH and Min DB. 1998. Optimization of Solid-Phase Microextraction
Analysis for Head Space Flavor Compound of Orange Juice. J Agric Food Chem.
25:2744-7.
Kakac S and Liu H. 1998. Heat Exchanger: Selection Rating and Thermal Design. Second
Edition.
The 12
th
16 -18 June, 2011

142

Khunvirojpanich A. 2005. Reduction of Flavor Deterioration in Pasteurized Fruit and
Vegetable Juices by Ohmic Heating. (DM thesis). Bangkok, Thailand: King
Mongkut's University of Technology Thonburi. 98 p. Available from: King
Mongkut's University of Technology Thonburi, FDE 493.
Leizerson S and Shimoni E. 2005. Effect of Ultrahigh-Temperature Continuous Ohmic
Heating Treatment on Fresh Orange Juice. J Agric Food Chem.53:519-24.

Marmesat S, Mancha M, Ruiz-Mendez MV and Dobargenes MC. 2005. Performance of
Sunflower Oil with High Levels of Oleic and Palmitic Acids During Industrial
Frying of Almonds, Peanuts, and Sunflower Seeds. J Agric Food Chem. 82:505-
10.
Moser K, Kriwet K, Naik A, Kalia YN and Guy RH. 2001. Passive skin penetration
enhancement and its quantification in vitro. European Journal of Pharmaceutics and
Biopharmaceutics. 52:103-12.
Nishida R, Shelly TE, Whittier TS and Kaneshiro KY. 2000. -Copaene, A Potential
Rendezvous Cue for The Mediterranean Fruit Fly, Ceratitis capatata. J Chemical
Ecology. 26:87-100.
Rivas A, Rodrigo D, Martinez A, Barbosa-Canovas GV and Rodrigo M. 2006. Effect of
PEF and Heat Pasteurization on The Physical-Chemical Characteristics of Blended
Orange and Carrot juice. International Journal of Food Science and Technology.
39:1163-70.
Sastry SK. 1994. Ohmic Heating. Minimal Processing of Foods and Process Optimization.
Singh, R.P. and Oliveira, F.A.R. (Eds.). CRC Press. London,23.
Theophilus C, Fleischer C, Waigh D and Waterman PG. 1997. Bisabolene Sesquiterpenes
and Flavonoids from Friesodielsia enghiana. Phytocheraistry. 44: 315-8.
Tokimoto Y, Steinhaus M, Buttner A and Schieberle P. 2005. Odor-Active Constituents in
Fresh Pineapple (Ananas comosus) by Quantitative and Sensory Evaluation.
Bioscience Biotechnology Biochemistry. 69:1323-30.

The 12
th
16 -18 June, 2011

143

OB-48

Optimisation of Non-starch Polysaccharide (NSP) Extraction from the
Jelly Fig (Ficus awkeotsang Makino) Seeds by Response Surface
Methodology

Wensheng Lim
l,*
, Kelvin K.T Goh
l
, Lee Wah Koh
2
, Lin Kiat Saw
2
, Mann Na Loong
3

1
Massey University, Singapore Campus, Block T1A25, 500 Dover Road,Singapore139651;
2
School of Chemical & Life Sciences,Singapore Polytechnic,500 Dover Road,Singapore139651;
3
Food Innovation & Resource Centre,Singapore Polytechnic,500 Dover Road,Singapore 139651
* Corresponding author:lim_wensheng@sp.edu.sg

Abstract

A gelling polysaccharide extracted from Jelly fig (Ficus awkeotsang Makino) seeds are
traditionally used in Taiwan to make dessert jelly. The desserts are usually prepared in
small scales and sold in shops and food courts in many parts of Asia. To date, no literature
has reported the optimum conditions for extracting the non-starch polysaccharide (NSP)
from the seeds. The aim of this study was to determine the optimum NSP extraction
conditions from jelly fig seeds by using a series of sequential experimental designs. The
basis for achieving the optimized extraction conditions were maximum gel strength and
maximum extraction yield with minimum protein content. The significant effect of
extraction temperature (40, 50 and 60
o
C), pH (4, 5 and 6), time (10, 20 and 30 min), water
to seed ratio (40:1, 50:1 and 60:1) and their interactions on gel strength and extraction
yield were first identified using full factorial design. The results showed that all the four
factors significantly (p < 0.05) affected the gel strength and extraction yield. Steepest
ascent method was subsequently employed to determine separately the optimum region for
maximum gel strength and maximum extraction yield. Further optimization was
conducted according to the response surface methodology by using the central composite
design. Response surface analysis predicted that extraction temperature, pH and time of
32
o
C, pH of 6.1 and 15 min, respectively would give maximum gel strength of 0.818 N at
40:1 water to seed ratio. The analysis also showed that the maximum yield of 12.3 % and
a minimum protein content of 2.26 % could be achieved at water to seed ratio (12.4:1),
extraction temperature (37.7
o
C), pH (3.9) and time (33.4 min). Gel strength and yield
values obtained from actual experiment are in good agreement (p>0.05) with the predicted
values.

Keywords: Optimization, extraction, full factorial design, steepest ascent method,
response surface methodology

Introduction
Polysaccharides are widely used in food and non-food industries as stabilizers, thickening,
gelling agents, crystallization inhibitors, and encapsulating agents (Izydorczyk & Wang,
2005). The non-starch polysaccharides found in seeds are usually present in the seed coats
and cell wall materials of seed cotyledons and endosperms (Eskin, Ikeda, & Cui, 2007).

Ficus awkeotsang Makino is a unique evergreen woody vine distributed rampantly in
tropical and subtropical regions of Taiwan (Chua, Chou, Chan, & Tzen, 2007). The seeds
The 12
th
16 -18 June, 2011

144

from the fruits (usually called jelly fig) have been used to make a local dessert jelly
known as Ai-Yu-Tung in Taiwan for almost two hundred years (Suzuno, Kinugasa,
Nakahara, & Kawabata, 1997). The typical preparation procedure starts with placing the
seeds in a cotton bag. The bag is then submerged in cold water and manually rubbed. As
the bag is being squeezed and massaged, a slimy yellowish extract seeps through the bag.
After several minutes of washing, the extract is allowed to set into a gel under at room
temperature. In recent years, extraction methods of this jelly fig polysaccharide have been
reported. According to one of these reports, 10 g of freshly dried achene seeds were first
immersed in 60 volumes of tap water in a beaker accompanied by slow stirring for an hour.
Jelly fig aqueous extract was filtered through six layers of cotton cloth). This aqueous
extract spontaneously forms a gel at room temperature ( Jiang, et al., 2002.

Mol wt and sugar composition and proposed gelling mechanism cation-mediated gelation
Since 1930, the unique gel-forming property of the jelly fig extract has been the subject of
many chemical and physiochemical investigations (Miyazaki, et al., 2004). Due to its
unique characteristics, it has the potential to be utilised in the development of a controlled
drug release tablet (Miyazaki, et al., 2004).

To date, little information is known regarding the optimum extraction conditions for this
gelling polysaccharide from the jelly fig seeds. The objective of this study was to optimise
the NSP extraction from jelly fig seeds using full-factorial design, steepest ascent method
and response surface methodology (RSM) series of sequential experimental designs. RSM
helps to build empirical models and to define optimal performance in a complex data space
defined by several factors that would affect performance outcome. Such experiments
screen the appropriate data space, build empirical prediction equations using the significant
factors and describe the response surface around optimal performance (Williges, 2009).
Experimental design describes how to plan and conduct experiments in order to optimize
the response at different combinations of independent variables so as to obtain the
maximum amount of information with the lowest number of experiments (Walmsley &
Stoyanov, 2009).

In this present study, the influence of the four extraction conditions namely water to seeds
ratio, extraction pH, time and temperature and their interactions on NSP extraction yield
and gel strength were investigated using the full factorial design. The steepest ascent
method was then employed to approach the optimum area with the significant variables.
Lastly, the central composite design determines the optimum conditions to give maximum
gel strength and maximum NSP extraction yield with minimum protein content.

Materials
Jelly fig seeds harvested at Alishan in Taiwan were purchased from Zhong Yong Traders
Private Limited (Singapore). The seeds were stored at a dry and cool condition. Sodium
citrate was purchased from TTCA Co Pte Ltd (China); sodium hydroxide and hydrochloric
acid were purchased from Merck Pte Ltd (Germany); glucose standard, protein standard,
Folin and Ciocalteaus Phenol reagent and Biuret reagent were purchased from Sigma
Aldrich (USA).

The 12
th
16 -18 June, 2011

145

Extraction procedure
Jelly fig extract was extracted from whole seeds using deionised water. Water was
preheated in the water bath (WNB 14, Memmert, Germany) to a designated temperature
before the seeds were added. The temperature of the solution was monitored using a
thermometer (Testo 922, GmbH & Co, Germany) with a type-K thermocouple probe
(T53304) and kept steady within 0.1
o
C. The water and seed slurry was mixed
throughout the extraction period with an overhead stirrer (RW 20, IKA, USA), with a
propeller stirrer attached. The pH was monitored using the pH meter (827 pH lab,
Metrohm, Switzerland) and the pH was adjusted by 0.5 M hydrochloric acid and 0.5 M
sodium hydroxide to within 0.1. Separation of the seeds from the extract was achieved
by passing the slurry through a metal sieve and two layers of nylon cloth. The extracted
solution was prepared for yield, gel strength and protein analysis as explained in below
sections.

Determination of extracted yield
The NSP yield of jelly fig seeds was computed as the weight of the total solid content in
the extract over the weight of the wet seeds in percentage. 1 % sodium citrate was added to
the extract immediately after filtration to prevent gelation for ease of measurement. The
percentage solid content of the extract was measured using a moisture analyzer (MX-50, A
& D). The moisture analyzer was set at 105 C and approximately 5 g of extract used per
run with triplicates per sample.

Determination of protein content
Protein content was determined based on the micro Lowry, Onishi & Barr modification
method. 1 % sodium citrate was added to the extract immediately after extraction to
prevent gelation. Protein standard containing 100 mg/ml of bovine serum albumin (BSA)
was used. a calibration curve was constructed by dilution of the standard BSA solution to
concentration of 250, 500, 750 and 1000 g/ml. 0.2 ml of each concentration was pipetted
into a test tube. To each test tube, 2.2 ml of Biuret reagent was added; the tubes were
mixed well and allowed to stand at room temperature for 10 min. Then, 0.1 ml of Folin and
Ciocalteaus Phenol reagent was added to each tube. The tubes were mixed well and
allowed to stand at room temperature for 30 min. The absorbance was read using a
spectrophotometer (UV-1800, Shimadzu Asia Pacific Pte Ltd, Japan) at 700 nm against the
blank (0 g/ml). The samples were diluted to give a final protein concentration range
between 150 and 1000 g/ml and the subsequent preparation method was similar to the
procedures mentioned above. Triplicates were analysed per run.

Determination of gel strength
Gel strength was determined using the texture analyzer (TA-XT plus, Stable Micro
Systems Ltd, UK) with a cylinder shaped delrin tenon probe of 0.5 inch diameter. 45 g of
extracted solution was poured into each bottle and capped. The bottles were stored at 4 C
for a day. The gel strength measurements were made with the following test parameters:
pre-test speed, 2 mm/s; test speed, 1 mm/s; post-test speed, 3 mm/s; strain, 50 %; trigger
force, 0.2 N. Triplicates were measured per run.

The 12
th
16 -18 June, 2011

146

Experimental design
A series of sequential experimental designs were conducted to investigate the effect of four
extraction conditions (extraction temperature, pH, time and water to seed ratio) and
optimize two extraction conditions to achieve maximum yield with minimum protein
content and maximum gel strength. Literature studies and preliminary experiments
revealed that water to seed ratio (40:1 to 60:1); extraction temperature (40 C to 60 C), pH
(4 to 6) and time (10 min to 30 min) were supposed to have effects on NSP extraction. The
statistical designs and data analysis used in this paper were generated using commercial
statistical software, Minitab (version 15, Minitab Inc, USA).

1)Full factorial design (FFD)
A 2
4
FFD with three center points and two replicates per run was applied to determine the
significant effect of the four extraction conditions (extraction temperature, pH, time and
water to seed ratio) as well as their interactions. Preliminary experiments showed that
water to seed ratio has an infinite effect on gel strength. Therefore, the water to seed ratio
was fixed at 40:1. Hence, a 2
3
FFD with three center points was used for maximizing of gel
strength. In FFD, low and high level settings were coded -1 and 1 with the center point
coded as 0. The independent variables (extraction conditions) and range listed in Table 1
were determined from literature studies and preliminary experiments. Runs of center points
were included to identify the significance of curvature. Significant curvature indicates that
the optimum would be near or within the experimental region. The runs were randomized
to minimize systematic error. The significant of the effects were identified on the basis of
confidence levels above 95 % (P < 0.05).

Table 1 Coded and actual values of the independent variables in full factorial design
Independent variables Level of variables
-1 0 1
x
1
: water to seed ratio 40:1 50:1 60:1
x
2
: temperature (C) 40 50 60
x
3
: pH 4 5 6
x
4
: time (min) 10 20 30

2)Steepest ascent method
The steepest ascent method was applied in order to investigate the data space out of the
initial experimental region so as to locate a new experimental region that is nearer to the
optimum. This method investigates the data space along the path of steepest ascent until no
further increase in the responses (extraction yield and gel strength). The direction of the
ascent passes through the center point of FFD and is based on the ratio of the regression
coefficient,
i
of each significant independent variable and the smallest regression
coefficient. The increment of the ascent is based on the step size increase of the regression
coefficient ratio. The point where no further increase was observed will be used as the
center point for the next design.

3)Central composite design
RSM was applied using the central composite design to obtain a quadratic model. The
design consists of eight axial points to estimate the quadratic effects and seven center
points to assess the significance of the curvature. In addition to the coded levels in FFD,
The 12
th
16 -18 June, 2011

147

the high and low settings of the axial point were assigned 2 and -2 respectively. The central
composite design generated contains 31 runs. The data obtained from the runs were fitted
into a second degree polynomial equation as follows:

(1)

Y is the predicted response and the independent variables are x
1
, x
2
, x
3
and x
4
.The
regression coefficient
0
is a constant;
1
,
2
,

3
and
4
are the main linear coefficients;
12
,
13
,
14
,
23
,
24
and

34
are the interaction coefficients;
11
,
22
,

33
and
44
are the
quadratic coefficients. Analysis of variance (ANOVA) and regression analysis were used
for fitting the model and to examine the significance of the coefficient terms. The
adequacy of the model was checked using the R2 and R2 adjusted values. The optimisation
technique of the software was used for simultaneous optimisation of multiple responses or
of a single response. In this present study, two optimisation investigated were to obtain
maximum NSP extraction yield with minimum protein content and obtain maximum gel
strength.

Full factorial design and analysis
The influence of the four independent variables (water to seed ratio, temperature, pH and
time), their two-way interactions have on extraction yield and the significance of curvature
was investigated using the 2
4
FFD with three center points. Table 2 shows the four
independent variables, the coded level, the average response value of each run and the FFD
matrix generated from the software. First-order model was generated by fitting the data
using the regression analysis and analysis of variance (ANOVA) to examine the
significance of the effects and curvature. The p-values obtained were used to check the
significance of every coefficient. P-value of less than 0.05 indicates a significant
coefficient and vice-versa. The insignificant coefficients were eliminated from the model.
R
2
adjusted values were used as an indicator to compare between the models generated.
The R
2
adjusted value is a modified version of R
2
value. R
2
value is the proportion of the
variability in the response irrespective of the number of terms in the model. However, R
2

adjusted value takes into consideration the number of terms in the model and adjusts
accordingly. Hence, R
2
adjusted value was used to compare between models with the same
response data but with different numbers of terms. The model with the highest R
2
adjusted
value was determined to be the best model. The first-order model with extraction yield as
the response which gives the highest R
2
adjusted value of 86.2 % has the equation (coded
factors) as follows:

(2)

The model appeared to be adequate and adjusts well to the experimental data as only 9 %
of the total variation was not explained by the model (R
2
= 91 %). The model also shows a
non-significant lack of fit which indicates that the model was adequate within the range of
the four variables in the FFD. The curvature was significant which indicate that the
The 12
th
16 -18 June, 2011

148

optimum was near or within the current experimental design space. From Eq. (2), water to
seed ratio (x
1
) has the most effect on extraction yield followed by temperature (x
2
). Eq. (2)
also shows that water to seed ratio, temperature and pH have a negative effect and time has
a positive effect on extraction yield. Hence, decreasing water to seed ratio, temperature, pH
and increasing of time can move the current design space nearer to optimum.

Table 2 Full factorial experimental design with coded values of four independent variables with
average extraction yield of each run
Runs x
1
x
2
x
3
x
4

Water to seed
ratio
Temperature pH Time Extraction
yield (%)
1 -1 -1 -1 1 9.90
2 1 1 1 -1 4.60
3 0 0 0 0 7.69
4 1 1 1 1 5.23
5 1 -1 -1 -1 9.84
6 1 1 -1 -1 5.49
7 -1 1 -1 -1 9.11
8 -1 -1 -1 -1 7.77
9 0 0 0 0 8.93
10 1 -1 -1 1 8.11
11 -1 1 -1 1 8.74
12 -1 -1 1 1 9.53
13 1 1 -1 1 7.46
14 1 -1 1 1 6.68
15 -1 -1 1 -1 8.60
16 -1 1 1 -1 7.11
17 0 0 0 0 8.76
18 1 -1 1 -1 6.84
19 -1 1 1 1 9.88

Preliminary experiments show that water to seed ratio has an infinite effect on gel strength.
Therefore, the water to seed ratio was fixed at 40:1 in order to investigate the extraction
conditions for maximum gel strength. A 2
3
FFD with three center points was generated.
Table 3 shows the three independent variables, the coded level, the response value of each
run and the design matrix generated. Similarly, a first-order model was generated by fitting
the data obtained and regression analysis was conducted to determine the significance of
the effects. The first-order model with gel strength as the response which gives the highest
R
2
adjusted value of 99.1 % has the equation (coded factors) as follows:

(3)

The 12
th
16 -18 June, 2011

149

The model appeared to be adequate and adjusts well to the experimental data as only 0.6 %
of the total variation was not explained by the model (R
2
= 99.4 %). The model shows a
non-significant lack of fit which indicates that the model was adequate within the range of
the three variables in the FFD. The curvature was significant which indicate that the
optimum was near pr within the current experimental design space. From Eq. (3), pH (x
3
)
has the most effect on extraction yield followed by temperature (x
2
). Eq. (3) also shows
that temperature and time have a negative effect and pH has a positive effect on extraction
yield. Hence, decreasing temperature with time and increasing of pH can move the current
design space nearer to optimum.

Table 3 Full factorial experimental design with coded values of three independent variables with
average gel strength for each run
Runs x
2
x
3
x
4

Temperature pH Time Gel strength (N)
1 -1 -1 1 0.072
2 1 -1 -1 0.088
3 0 0 0 0.124
4 -1 -1 -1 0.066
5 1 -1 1 0.218
6 -1 1 1 0.581
7 0 0 0 0.125
8 -1 1 -1 0.723
9 1 1 -1 0.549
10 1 1 1 0.094
11 0 0 0 0.135

Steepest ascent method and result
The direction of steepest ascent for maximum extraction yield was determined by the beta-
coefficient of the four main effects in Eq. (2) and the direction of steepest ascent path for
maximum gel strength was determined by the beta-coefficient of the three main effects in
Eq. (3). The center point of the two FFD was taken as the origin for each of the ascent
path. Triplicates were conducted for each run in the steepest ascent method. The average of
the responses per run for maximum extraction yield and gel strength was shown in Table 4
and Table 5 respectively. From the results for extraction yield, region of optimum was
around the extraction conditions for run 2 as it shows a maximum yield of 11.06 %. For
results of gel strength, the region of optimum was around the extraction conditions for run
1 as it shows maximum gel strength of 0.651 N. The results confirmed that the region of
optimum for extraction yield and gel strength were near and within the design space of
FFD as indicated by the significance of the curvature.

The 12
th
16 -18 June, 2011

150

Table 4 Steepest ascent experiment with actual values of four independent variables with extraction
yield as the response
Run
Water to seed
ratio Temperature pH Time
Extraction
yield
(C) (min) (%)
1
23.4 34.4 3.7 30
10.14
2
15.4 29.7 3.3 33
11.06
3
10.1 26.5 3.1 35
10

Table 5 Steepest ascent experiment with actual values of three independent variables with gel
strength as the response
Run Temperature pH Time Gel strength
(C) (min) (N)
1 44.3 6.37 17.5 0.651
2 38.6 7.73 15 0.327
3 32.8 9.1 12.5 0.0490
4 27.1 10.5 10 0.0408
5 21.4 11.8 7.5 0.0423

Central composite design and analysis
A response surface design was employed to further optimize the extraction conditions.
Extraction conditions at the optimum point found by the steepest ascent method were used
as the center point for the response surface design. The response surface design was
conducted using the central composite design. The data obtained from the central
composite design were fitted into a second-order polynomial model. Two central
composite designs were used in order to predict optimum extraction conditions for
maximum yield with minimum protein content and for maximum gel strength.

1) Gel strength
A 20 runs central composite design consisting of 6 center points and 6 axial points was
generated to further optimize the extraction conditions for maximum gel strength (Table
6). The actual and coded values used in the central composite design are shown in Table 7.
A regression analysis was performed to obtain the estimated regression coefficient and
their significance determined by the p-value. The model shows a non-significant lack of fit
(p > 0.05) which implied that the model adequately described the functional relationship
between the main effects and the response. The model gives a good R-square value of 84.9
% and has the equation as follows:

(4)

The 12
th
16 -18 June, 2011

151

Table 6 Central composite design with coded values of three independent variables with average gel
strength for each run (coded level)
Runs x
2
x
3
x
4

Temperature pH Time Gel strength (N)
1 -1 -1 -1
0.316
2 -2 0 0
0.525
3 1 1 1
0.627
4 0 0 -2
0.316
5 0 2 0
0.512
6 0 0 0
0.681
7 0 0 0
0.571
8 -1 -1 1
0.652
9 2 0 0
0.735
10 0 0 0
0.544
11 0 -2 0
0.446
12 1 -1 -1
0.469
13 -1 1 -1
0.459
14 0 0 2
0.721
15 0 0 0
0.521
16 1 1 -1
0.478
17 -1 1 1
0.661
18 1 -1 1
0.616
19 0 0 0
0.562
20 0 0 0
0.597

Table 7 Coded and actual values of the independent variables in central composite design and
predicted optimum condition for maximum gel strength of 0.818 N

-2 -1 0 1 2 Optimum
x
2
: temperature (C)
32 36 40 44 48 32
x
3
: pH
5.4 5.7 6 6.3 6.6 6.1
x
4
: time (min)
5 7.5 10 12.5 15 15

Graphical analysis showing the effect of pH, time and temperature on gel strength is given
in Figure 1. Figure 1 shows an increasing extraction time results in increasing of gel
strength and increasing of pH up to around pH 6 also results in an increase of gel strength.
However, the decreasing of temperature results in increasing of gel strength. The optimum
extraction conditions to achieve maximum gel strength can be obtained using the
optimization plot. The optimised extraction conditions were at extraction temperature, pH
and time of 32
o
C, pH 6.10 and 15 min respectively, yielding gel strength of 0.818 N at
40:1 water to seed ratio.

The 12
th
16 -18 June, 2011

152

2) Extraction yield and protein content
A 31 runs central composite design consisting of 7 center points and 8 axial points was
generated to further optimize the extraction conditions for maximum extraction yield and
minimum protein content (Table 8). The actual and coded values used in the central
composite design can be seen in Table 9. Two models were obtained for two response
variables (extraction yield and protein content). Regression analysis was performed for the
two models to obtain the estimated regression coefficient and their significance determined
by the p-value. The models obtained for extraction yield and protein content show good
adequacy as both exhibit non-significant lack of fit (p > 0.05) and gives an R-square value
of 95.4 % and 81.2 % respectively. Both models give the equations as follows:

(5)

(6)

Table 8 Central composite design with coded values of three independent variables with average gel
strength for each run

Runs
x
1

x
2
x
3
x
4

Water to
seed ratio
yield (%)
Protein
content (%)
1 0 0 0 -2
9.50
4.94
2 -1 1 1 1
11.8
2.98
3 0 0 0 0
11.2
3.34
4 0 0 2 0
11.7
3.30
5 -1 -1 1 -1
10.1
5.48
6 1 -1 1 1
10.3
3.93
7 0 2 0 0
11.4
2.86
8 -1 -1 -1 1
10.1
2.88
9 -1 -1 -1 -1
9.71
3.65
10 1 1 -1 -1
10.1
3.74
11 0 0 0 0
11.1
2.77
12 1 1 1 -1
10.2
3.94
13 2 0 0 0
11.4
3.66
14 1 -1 -1 1
9.62
3.72
15 0 0 -2 0
10.5
4.49
16 1 1 -1 1
10.0
4.54

The 12
th
16 -18 June, 2011

153

Table 8 (Continued)

Runs
x
1

x
2
x
3
x
4

Water to
seed ratio
yield (%)
Protein
content (%)
17 0 -2 0 0
9.91
3.87
18 -1 1 1 -1
10.7
3.23
19 0 0 0 0
10.9
3.86
20 1 1 1 1
11.7
4.21
21 0 0 0 2
10.4
4.25
22 0 0 0 0
10.8
3.34
23 0 0 0 0
11.2
2.62
24 1 -1 -1 -1
10.3
4.34
25 -1 -1 1 1
11.1
3.91
26 -1 1 -1 1
11.1
3.89
27 1 -1 1 -1
9.09
5.18
28 0 0 0 0
11.3
3.03
29 0 0 0 0
11.1
2.77
30 -2 0 0 0
10.4
4.47
31 -1 1 -1 -1
10.4
4.99

Table 9 Coded and actual values of the independent variables in central composite design and
predicted optimum condition for maximum extraction yield (12.3 %) and minimum protein
content (2.26 %)


-2 -1 0 1 2 Optimum
x
1
: water to seed ratio
7.4 11.4 15.4 19.4 23.4
12.4:1
x
2
: temperature (C)
21.7 25.7 29.7 33.7 37.7
37.7
x
3
: pH
2.7 3 3.3 3.6 3.9
3.9
x
4
: time (min)
28 30.5 33 35.5 38
33.4

Contour plots showing the effect of water to seed ratio and time on extraction yield and
protein content are given in Figure 2. Figure 1 shows a decreasing in water to seed ratio
results in decreasing of protein content to a certain point and further decrease in water to
seed ratio results in a further increase in extraction yield. Increasing of extraction time also
results in decreasing of protein content to a certain extent and further increase in extraction
time results in further increase in extraction yield. The optimum extraction conditions to
achieve a maximum yield and minimum protein content were obtained by overlaying
contour plots generated by the two models as shown in Figure 3. The optimised extraction
conditions was at water to seed ratio, extraction temperature, pH and time of 12.4:1, 37.7
o
C, pH 3.9 and 33.4 min, respectively. It was predicted that at this extraction conditions,
yield of 12.3 % and protein content of 2.26 % can be achieved.
The 12
th
16 -18 June, 2011

154

The optimised pH (3.97) to achieve maximum yield were lower compared to the optimized
pH (6.1) required to achieve maximum gel strength. The reason for lower pH to achieve
higher extraction yield could be due to the acidic nature of the NSP. However, the
extraction pH will not fall below 3.5 as polysaccharide could hydrolyze under such
circumstances. The reason for higher pH to achieve stronger gel strength could be due to
the increase in binding affinity of the NSP to calcium ions under more alkaline condition.
The optimised extraction temperature to achieve maximum yield and maximum gel
strength were nearer to the lower limit as increasing of extraction temperature could result
in hydrolysis of polysaccharide.

3) Verification of predicted values
The optimised extraction conditions for maximum extraction yield with minimum protein
content were at water to seed ratio, extraction temperature, pH and time of 12.4:1, 37.7 C,
pH 3.9 and 33.4 min respectively. At such conditions, yield of 12.3 % and protein content
of 2.26 % can be obtained. Verification experiment of triplicates was conducted at these
conditions and yield of 11.2 % and protein content of 2.16 % were obtained. The result
shows that both values were in close agreement as there was no significant difference. The
optimised extraction conditions for maximum gel strength were at extraction temperature,
pH and time of 32
o
C, pH 6.10 and 15 min respectively. At such conditions, gel strength of
0.818 N was predicted to obtain.

Figure 1 Contour plot (A) and surface plot (B) showing the effect of pH and time on gel strength
(temperature = 32 C) and contour plot (C) and surface plot (D) showing the effect of pH and
temperature on gel strength (time = 15 mins).
The 12
th
16 -18 June, 2011

155

Figure 2 Contour plots the effect of water to seed ratio and time on: (A) extraction yield and (B) gel
strength.

Figure 3 Overlaid contour plots showing the effect of water to seed ratio and time on extraction yield
and protein content with the overlapping region giving the optimum extraction conditions.

at 40:1 water to seed ratio. Verification experiment of triplicates was conducted at these
conditions and gel strength obtained was 0.802 N. The result shows that the experimental
value was near the predicted value as there was no significant difference between the both
values.

Conclusions
In conclusion, optimisation of extraction conditions to achieve maximum yield with
minimum protein content and maximum gel strength were successfully carried out with a
series of sequential experimental designs (full factorial design, steepest ascent method and
central composite design). The four independent variables (extraction time, temperature,
pH and water to seed ratio) were found to have significant effect on extraction yield and
gel strength. The optimised extraction conditions to obtain a maximum yield of 12.3 %
with a minimum protein content of 2.26 % was at water to seed ratio, extraction
temperature, pH and time of 12.4:1, 37.7 C, pH 3.9 and 33.4 min respectively. On the
The 12
th
16 -18 June, 2011

156

other hand, optimised extraction conditions for water to seed ratio of 40:1 to obtain
maximum gel strength of 0.818 N was at extraction temperature, pH and time of 32 C, pH
6.10 and 15 min respectively. Verification experiments were conducted for both set of
extraction conditions and the experimental values show to have no significant differences
from the predicted values.

Acknowledgementsw
The author will like to acknowledge the financial support from Ministry of Education
Singapore (MOE). The author will also like to acknowledge the support from Mr Isaiah
Loong and statistical advice from Assoc. Prof. Hugh Morton.

References
Cui, S., Izydorczyk, M., & Wang, Q. (2005). Polysaccharide Gums Food Carbohydrates:
CRC Press.
Cui, S. W., Ikeda, S., & Eskin, M. N. A. (2007). Seed Polysaccharide Gums Functional
Food Carbohydrates: CRC Press.
Chua, A. C. N., Chou, W.-M., Chyan, C.-L., & Tzen, J. T. C. (2007). Purification, Cloning,
and Identification of Two Thaumatin-like Protein Isoforms in Jelly Fig (Ficus
awkeotsang) Achenes Journal of Agricultural and Food Chemistry, 55(18), 7602-
7608.
Jiang, C.-M., Lai, Y.-J., Lee, B.-H., Chang, W.-H., Wu, M.-C., & Chang, H.-M. (2002).
Changes in physico-chemical properties of pectin from jelly fig (Ficus awkeotsang
Makino) seeds during extraction and gelling. Food Research International, 35(1),
31-35.
Li, Y.-C., Chang, C.-T., Hsiao, E. S. L., Hsu, J. S. F., Huang, J.-W., & Tzen, J. T. C.
(2003). Purification and Characterization of an Antifungal Chitinase in Jelly Fig
(Ficus awkeotsang) Achenes. Plant Cell Physiol., 44(11), 1162-1167.
Miyazaki, Y., Yakou, S., & Takayama, K. (2004). Study on jelly fig extract as a potential
hydrophilic matrix for controlled drug delivery. International Journal of
Pharmaceutics, 287(1-2), 39-46.
Peng, C. C., Hsiao, E. S. L., Ding, J. L. C., & Tzen, J. T. C. (2005). Functional expression
in Pichia pastoris of an acidic pectin methylesterase from jelly fig (Ficus
awkeotsang). Journal of Agricultural and Food Chemistry, 53(14), 5612-5616.
Suzuno, H., Kinugasa, S., Nakahara, H., & Kawabata, A. (1997). Molecular Characteristics
of Water-soluble Polysaccharide Extracted from Jelly Fig (Ficus awkeotsang
Makino) Seeds. Bioscience, Biotechnology and Biochemistry, 61(9), 1491-1494.
Williges, R. (2006). Response Surface Methodology and Sequential Experimentation
International Encyclopedia of Ergonomics and Human Factors, Second Edition - 3
Volume Set: CRC Press.
Walmsley, A., & Stoyanov, K. (2009). Response-Surface Modeling and Experimental
Design Practical Guide To Chemometrics, Second Edition: CRC Press.

The 12
th
16 -18 June, 2011

157

OB-84

A Study of the Effectiveness of Supercritical Fluid Carbon Dioxide
in Bitter Almond Oil Extraction

Mingmiao Yu
1
, Chingshun Lan
1
, Horngliang Lai
2
, Jenshinn Lin
1
*

1
Department of Food Science, National Pingtung University of Science and Technology,
Pingtung, Taiwan 91201, ROC (Taiwan);
2
Department of Plant Industry, National Pingtung University of Science and Technology,
Pingtung, Taiwan 91201, ROC (Taiwan)
*Corresponding author: jlin@mail.npust.edu.tw

Abstract

Bitter almond oil, which is rich in aroma and unsaturated fatty acids, has healthful benefits
for preventing from cardiovascular diseases. Higher bitter almond oil could be obtained by
Soxhlet extraction, but n-hexane remnants are harmful for liver cell. For this purpose, the
objective of this research is to investigate how the oil content of bitter almond would be
extracted by the supercritical fluid carbon dioxide operated at different temperatures of an
indigenous supercritical fluid extractor (ISFE). The operating pressure was 5000 psi, and
the ethanol was used as a co-solvent. The results showed that extraction of bitter almond
oil was successful using this method in a considerable short period of time. The obtained
oil exhibited white color and strong aroma. In addition, the optimal operation conditions of
supercritical fluid extraction were obtained through the study. It was found at a
temperature of 46 and a retention time of 60 min. There was no difference between the
oils extracted by ISFE and Soxhlet extraction. Results of this research exhibited that the
bitter almond oil could be extracted effectively by supercritical uid carbon dioxide.

Key words: Bitter almond oil, aroma, Soxhlet extraction, supercritical fluid, carbon
dioxide

Introduction
Almond oil is a well known example which is used in cosmetics and for medical purposes
since many years (Hallabo et al. 1975), but also the fatty acid composition of other Prunus
kernels like peach, apricot, plum or cherry has been characterized (Helmy 1990; Femenia
et al. 1995; Johansson et al. 1997). The almonds also contain as much as about 50% oil. As
one of the most popular vegetable oils, it is rich in mono- and polyunsaturated fatty acids,
with oleic and linoleic acids as the major constituents, and contains the naturally occurring
Vitamins A, B1, B2, B6 and Vitamin E. This characteristic composition of the almond oil
makes it a valuable material for the food industry (Zhang et al. 2009). Depending on
variety, they contain approximately 50-150 pMol/g (dry weight basis) of the potentially
toxic cyanogenic glycosides amygdalin and prunasin (Tungel et al., 1990; Brimer et al.,
1993). However, recently, interest in amygdalin is gradually increasing due to a derivative
of amygdalin, that is, laetrile (vitamin B17), whose use as secondary cancer chemotherapy
and antineoplastic agent has been encouraged (Suchard et al., 1998; Tatsuma et al., 2000).

There are many methods of oil extraction with both mechanical and chemical separation
processes. Mechanical separation processes lack of efficiency with low yields and
chemical processes employ solvents such as hexane, which are harmful environment
The 12
th
16 -18 June, 2011

158

(Salgn 2008). Steam distillation is commonly used method to extract oil and its a simple
and easy maintains equipment often use this method in industry. High temperature steam
let volatile components come out and cool water can condense volatile components.
Finally, the oil and water are separation then we get high concentrations of almond oil. The
extraction temperature is too high to affect the oil quality, including oxidative damage, lose
their fragrance, which is the largest fatal problem. High-pressure cold pressing method is
also one of traditional oil extraction methods; it is the use of physical method. High
pressure equipment has simple structure and doesnt make almond oil heat damage.
Although cold press is better than other way but it takes a long time extraction and lack of
efficiency. Because it must go through a series of extracts of separation and purification,
until almond oil purity. Solvent extraction is the best method of lipid extraction, but it use
organic solvents, and produce large amount of solvent wastes. Organic solvent is
hazardous and flammable liquid organic solvent; it has potential toxic emissions during
extraction, and solvent is costly and high-purity required (Sahena et al. 2009).

Supercritical carbon dioxide extraction is new technology for extraction oil. This method
has the advantage of non-burning, non-pollution, non-residue, non-corrosive and there can
extract the oil from the natural world, with high purity and low cost. Carbon dioxide (CO
2
)
is the most commonly used supercritical fluid and its non-toxic, non-flammable and is
available at low cost with a high degree of purity. Its low critical constants (Tc = 31.2 ,
Pc = 1070 Psi) allow supercritical operation of thermally labile compounds. After
extraction, we reduce pressure at a certain temperature; extract and carbon dioxide will
separate. Sometimes we add ethanol to enhance the extraction of polar compound (Salgn
2008; Sanchez et al. 2009; Sahena et al. 2009; Mezzomo et al. 2010; Yilmaz et al. 2011;
Ozcana et al. 2011).

Materials
Bitter almond used in this study was purchased from Xing Hao Co., Ltd. (Kaohsiung,
Taiwan). Ethanol (95%) was purchased from Taiwan Tobacco and Liquor Corporation
(Pingtung, Taiwan). The CO
2
was purchased from Ching Shang Co., Ltd. (Kaohsiung,
Taiwan).

Extraction
An indigenous supercritical fluid extractor (ISFE) was used in this study. A schematic
diagram of the extractor was shown in Figure 1.

The 12
th
16 -18 June, 2011

159

Figure 1 T1: 7
o
C, T2: 50
o
C temperture control, P: pumps, E1: 2L extraction vessel, E2: 0.5L
extraction vessel, C: separator control, S1: separator vessel 1, S2: separator vessel 2.

Operation of ISFE
To operate the, semi-batch, indigenous supercritical fluid extractor, first the cooler was set
at a temperature of 7
o
C, and a heater was set at a temperature of 50
o
C. The temperature of
the cooler was used to control the temperature of the carbon dioxide, so it would be sent
out of the gas steel cylinder to the heater for vaporization. The vaporized carbon dioxide
would be pressurized and fed into the extraction vessel, where the wet sample volume of
80% had been placed into. Then, the temperature of the extraction vessel was controlled by
a cylindrical electrical heater around the vessel. The cylinder contains carbon dioxide was
open, and the desired flow rate was adjusted. Then, turn on the high pressure compression
motor, and adjust a pressure control to increase the pressure at a slow rate of 2-3 psi/sec.

The experimental method is the use of carbon dioxide and supercritical fluid extraction of
ethanol for the work, first of all the pressure slowly compressed pressure 5500 psi, 5500
psi is unknown at this time the best standards set pressure, the pressure from this set go for
the best extraction temperature. Because the temperature supercritical conditions of ethanol
in 42, then skip the carbon dioxide, 31.8, because the effect of ethanol co-solvent
effect, 42 below the temperature will not have to test. Therefore, the temperature
experiments started directly from more than 42, almond oil productions by the
experimental conditions, the temperature started increasing from 43, 47 in
temperature almond oil production began to decrease. Also adjust the pressure to find the
maximum extraction of the best barrel production, in which case the fluid within the
overall adhesive force and activity to achieve a certain balance. That above the optimal
range of temperature and pressure but reduce the total output, because the whole almond
oil molecules too large to be separated from carbon dioxide and ethanol co-solvent into a
single molecule, single molecule fatty acids have been hundreds of molecules of carbon
dioxide and ethanol Surrounded. The original carbon dioxide and ethanol molecules Brown
campaign, by the strong collision force becomes floating state. Carbon dioxide and ethanol
molecules adsorbed on the fatty acids, it will not necessarily impact the loss of the original
direction, must be floating the kinetic energy of the whole molecule. Temperature control
in the extraction process is to maintain the power of floating, the temperature is not
The 12
th
16 -18 June, 2011

160

affected by the extraction of reducing the effect of pressures in the extraction process will
be gradually reduced, which is semi-batch supercritical fluid extraction machine features.
Less than optimal output conditions, fatty acids with smaller molecular weight of more
output products, when compared with the best conditions, this time at the same time
extracting the amount of output would be less. Another output is greater than the best
conditions, the fatty acids with larger molecular weight of more output products, when
compared with the best conditions, when total extraction yield at the same time will be
less. The adjustment from 3500 to 5500 psi pressure conditions in the search, the
experiment is to test each additional 100 psi as a unit, and eventually found 4500 psi is the
largest almond oil output. At this time a total of carbon dioxide and ethanol soluble extract
maximum effect, the resulting ethanol extract is white with a very fine aroma of suspended
particles occurs, the net for 24 hours after the fine will be white gelatinous precipitate,
ethanol extract at this time Contains little aroma. Will be recorded in the literature
amygdalin dissolved in ethanol (Lapis Lazuli Light 2011; Tunpl et al. 1995), amygdalin
extracted by supercritical ethanol, the ethanol concentration in the amygdalin is saturated.
Wet in the process of raw materials level is 1/4 to 1/3 ranges, so the more humid the
amount of ethanol is more the amount of amygdalin was extracted, residual amount of
amygdalin in the almond in the less.

Preliminary tests of operation of an indigenous supercritical fluid extractor for bitter
almond oil were completed in this study. By subjective judgment on oil obtained, the
operating pressure of ISFE was 5000 psi, when the ethanol was used as a co-solvent. The
obtained oil exhibited white color and strong aroma, and shown in Figure 2.

Figure 2 Appearance of extracts of bitter almonds by ISFE

The optimal operation conditions of the indigenous supercritical fluid extractor, a
temperature of 46 and a retention time of 60 min, were obtained through different tests.
There was no difference between the oils extracted by ISFE and Soxhlet extraction. It was
found that the bitter almond oil could be extracted effectively by supercritical uid carbon
dioxide.

Absorption ethanol extraction would slowly move the oil of the bitter almonds inside
extraction vessel. There was a phenomenon of oil moving gradient shown in Figure 3. If
The 12
th
16 -18 June, 2011

161

the extraction time and the number of extraction were increased, the color of bitter
almonds became a pure white, and the left oil content was very low. For some reason, the
defatted bitter almonds were crushed into powder passing through an ASTM #80 sieve.
Crude almond powder and then by the second crushing process is shown in Figure 4.

Figure 3 Appearance of bitter almonds taken out of the extraction vessel with a moving oil gradient.

Figure 4 Appearance of the powder crushed with defatted bitter almonds.

Figure 5 Another picture showing different layers of extracts of bitter almonds using ISFE.

The upper pale yellow liquid shown in Figure 5 containing alcohol. The middle layer
showed white color with gelatinous texture. It might be alcohol-soluble proteins, although
the protein is only soluble in alcohol 50 in 75% ethanol. The white gelatinous extract will
be further examined in the future.

Conclusions
Using ethanol as a co-solvent, the extraction of bitter almonds oil by an ISFE was fount
successful. The extracts obtained need more experiments to verify its composition. Powder
obtained by crushing the defatted bitter almonds was found with fine texture, and would be
suitable for commercial usage.

The 12
th
16 -18 June, 2011

162

Acknowledgments
This research project is funded by the National Science Council, ROC-Taiwan. The project
number is NSC 99-2622-E-020-001-CC3. The food company, Ma Yu Shan is another co-
sponsor.

References
Brimer L, TunGel G, Nout MJR. 1993. Simple screening procedure for microorganisms to
degrade amygdalin. Biotechnology Techniques 7: 683-687.
Femenia A, Chen YC, Mulet A, Canellas J. 1995. Chemical composition of bitter and
sweet apricot kernels. J. Agric. Food Chem. 43: 356-361.
Hallabo SAS, El-Wakeil FA, Morsi M, Khairy S. 1975. Chemical and physical properties
of apricot kernel oil and almond kernel oil. Egypt. J. Food Sci. 3: 1-5.
Helmy HE. 1990. Studies on the pigments of some citrus, prune and cucurbit seed oils
when processed with or without cottonseed oil. J. Am. Oil Chem. Soc. 67: 376-380.
Johansson A, Laakso P, Kallio H. 1997. Characterization of seed oils of wild, edible Finish
berries. Z. Lebensm. Unters. Forsch. A 204: 300-307.
Mezzomo N, Mileo BR, Friedrich MT, Martinez J, Ferreira SRS. 2010. Supercritical fluid
extraction of peach (Prunus persica) almond oil: Process yield and extract
composition. Bioresource Technology 101:5622-32.
Ozcana MM, Unvera A, Erkanb E, Arslana D. 2011. Characteristics of some almond
kernel and oils. Scientia Horticulturae 127:330-3.
Salgn U. 2007. Extraction of jojoba seed oil using supercritical CO2+ethanol mixture in
green and high-tech separation process. Journal of Supercritical Fluids 39:330-7.
Sanchez VY, Cabanas A, Renuncio ARJ, Pando C. 2009. Supercritical fluid extraction of
peach (Prunus persica) seed oil using carbon dioxide and ethanol. Journal of
Supercritical Fluids 49:167-73.
Sahena F, Zaidul ISM, Jinap S, Karim AA, Abbas KA, Norulaini NAN, Omar AKM. 2009.
Application of supercritical CO2 in lipid extraction-A review. Journal of Food
Engineering 95:240-53.
Tatsuma T, Komori K, Yeoh HH, Oyama N. 2000. Disposable test plates with tyrosinase
and b-glucosidases for cyanide and cyanogenic glycosides. Analytica Chimica Acta
408: 233-240.
Tungel G, Nout MJR, Brimer L, GBktan D. 1990. Toxicological, nutritional and
microbiological evaluation of tempe fermentation with Rhizopus oligosponcs of
bitter and sweet apricot seeds. Int. J. Food Microbial. 11: 337-344.
Suchard JR, Wallace KL, Gerkin RD. 1998. Acute Cyanide Toxicity Caused by Apricot
Kernel Ingestion. Ann. Emergency Med, 32: 742-744.
Yilmaz EE, Ozvural EB, Vural H. 2011. Extraction and identification of proanthocyanidins
from grape seed (Vitis Vinifera) using supercritical carbon dioxide. Journal. of
Supercritical Fluids 55:924-8.
Zhang QA, Zhang ZQ, Yue XF, Fan XH, Li T, Chen SF. 2009. Response surface
optimization of ultrasound-assisted oil extraction from autoclaved almond powder.
Food Chemistry 116: 513-518.

The 12
th
16 -18 June, 2011

163

OB-122

Optimization of Extraction Conditions for Patin (Pangasius sutchi) Fish
Skin Gelatin and Measurement of Gel Strength

Mahmoodani,F.
1
, Sanaei Ardekani,V.
1
, See,S.F.
1
, Ghassem,M.
1
, Yusop, S. M.
1
,
Ibrahim, S.
2
, Babji, A.S.
1,*

1
Food Science Program, School of Chemical Sciences and Food Technology, Universiti Kebangsaan
Malaysia, 43600, Bangi, Selangor, Malaysia;
2
Fisheries Research Institute (FRI), Department of
Fisheries Malaysia, 11960 Batu Maung, Penang, Malaysia.
*
Corresponding author: daging@ukm.my

Abstract

Gelatin represents a major source of good quality protein biopolymer with many
applications in food, pharmaceutical, photographic and cosmetic industries. Fish skins are
major by-products of the fishery and aquaculture industries with high collagen content that
can be used to produce fish gelatin. Pangasius sutchi (Patin) is a popular freshwater fish in
Malaysia. To optimize the extraction of gelatin from Patin skin, a 2-step Response Surface
Methodology (RSM) involving a Central Composite Design (CCD) was applied. After
screening the results, it was suggested that 4 variables (concentration of sodium hydroxide
and acetic acid, extraction time, and temperature) had significant effects on gelatin
extraction. These key factors were selected as independent variables for the subsequent
optimization using RSM. According to the results of a 2-step optimization,
the optimum conditions for the highest values of response were at concentration of 0.16N
sodium hydroxide and 0.08N acetic acid and extraction time and temperature of 210 min
and 63.25C, respectively. The predicted response for these optimal extraction conditions
was 68.16% extracted gelatin yield. The results suggest that RSM is a great optimizing
methodology for optimal extraction conditions from Patin freshwater fish skin. The gelatin
obtained from P.sutchi also showed relatively good gel strength
(481.3 g) than those from bovine (380.7 g).

Keywords: Pangasius sutchi, fish skin gelatin, optimization, response surface
methodology, gel strength

Introduction

Gelatin is a water soluble proteinaceous substance prepared by processes, which involve
the destruction of the tertiary, secondary and to some extent the primary structure of native
collagens (Johnston-Bank 1990). It is a high molecular weight polypeptide and
an important hydrocolloid, which has proved popular with the general public and finds use
in a wide range of food products largely because of its gelling and thickening properties.
Gelatin is an important functional biopolymer widely used in foods to improve elasticity,
consistency and stability. It differs from other hydrocolloids because most of them are
polysaccharide, whereas gelatin is a digestible protein containing all the essential amino
acids except tryptophan. The amino acid composition particularly with respect to proline
and hydroxyproline can vary from specie to specie, as a result of exposure to a wide range
The 12
th
16 -18 June, 2011

164

of environmental conditions, particularly temperature (Ladislaus and others 2007). Gel
strength is one of the important properties of gelatin, and the purpose of gelatin was
determined by the range of gel strength values. Generally, fish gelatin has lower gel
strength than mammalian gelatin (Norland 1987). Previous researchers indicated that
warm-fish gelatin has physical properties more like beef and pork than that of cold-water
gelatin (Cho and others 2005). Rheological properties (gel strength, gelling and melting
points and viscosity) are associated with the contents of hydroxyproline and proline in
collagens of different species (Ilona and others 2004). The gel formation involves the
partial denaturation and aggregation of individual denatured gelatin molecules.
Hydroxyproline and proline play great role in aggregation of gelatin subunits (Johnston-
Banks, 1990; Kinsella and others 1994).

Gelatin can be obtained from the skin and bones of not only terrestrial animals but also
fish. The waste from fish processing after filleting can account for as much as
75% of the total catch weight. About 30% of such waste consists of skin and bones with
high collagen content that can be used to produce fish gelatin
(Gmez-Guilln and others 2002). Extraction of gelatin from fish skins may provide
an alternative source that is acceptable for kosher (Jewish) and halal (Muslim) products
and serve as an alternative for markets concerned about bovine spongiform
encephalopathy.

Pangasius sutchi, known as Patin, is a popular freshwater fish in Malaysia.
This fish species is also abundantly available in the Amazon River, in parts of Russia and
in other places of the world under different names (Abbas and others 2006). The aim of
this study was to optimize the conditions for extraction gelatin from Patin fish skin using a
2-step optimization. In the 1st stage (screening), the objective was to efficiently determine
the important control variables from a large collection of potential variables and in the 2
nd

stage (optimization), Response Surface Methodology was utilized.

Materials
Patin fish skins were obtained from a farm fish located in Pinang, Malaysia. The frozen
skins were stored at -20 C with a maximum storage of less than 2 months before use.
Mammalian gelatin extracted from the skin of bovine was purchased from Sigma Chemical
Co. All reagents used in this study were analytical grade.

Materials preparation
The frozen skins were thawed at 4 C for 24 h before removing any remaining flesh.
Patin skins were then cut into 2 2 cm pieces and then washed with tap water 3 times.
The cleaned fish skins were drained using cheesecloth.

Gelatin extraction
Screening
Based on preliminary experiments, a pre-treatment with an alkaline solution followed by
an acid solution was chosen for this study. Cleaned skin (30 g) were treated with NaOH
with varying OH
concentrations for varying times, and then the samples were drained and
rinsed with tap water 3 times. Samples were then treated with acetic acid with varying H
+

concentrations for varying times, depending on the design. The samples were then drained
The 12
th
16 -18 June, 2011

165

and rinsed with tap water 3 times. All the previous pretreatments were done at varying
temperatures (C), depending on the design. After that point, samples were mixed with
distilled water (the total ratio of skin/water, w/w, depending on the design) and extracted at
varying temperatures (C) for varying times (min).

Optimization
Cleaned skins (30 g) were treated with NaOH with varying OH
concentrations
(factor A, Normality) for 60 min, depending on the design, and then the samples were
drained and rinsed with tap water 3 times. Samples were then treated with varying
concentrations of acetic acid (factor B, Normality) for 60 min, depending on the design.
The samples were then drained and rinsed with tap water 3 times. All the pretreatments
above were done at 4C. After that point, samples were mixed with distilled water
(the ratio of sample/water, 1/8 w/w) and extracted at varying temperatures (factor C, C)
for varying times (factor D, min).

Experimental design
Design-Expert, Version 6 software was used for the RSM analysis (Design-Ease 2006).
Two stages were used for the optimization of the gelatin extraction: screening and
optimization. A 7-factor, 2-level fractional factorial screening design was used for
screening the extraction. This stage was used to select the 4 most important factors out of
an original 8 factors (pretreatment temperature, alkaline concentration and alkaline time;
acid concentration and acid time; extraction temperature and time, and the skin/water
ratio). This stage helped to determine which factors were significant for the gelatin
extraction (Araujo and Brereton 1996).

To optimize gelatin extraction, RSM with a 4-factor, 5-level Central Composite Design
(CCD) was adopted with hydroxyproline content as a response. Four factors from the
screening experiments were chosen as independent variables. After the conditions for the
desired range for the independent variables were set up, the RSM software would supply
many groups of optimized conditions during the optimization.

Determination of extraction yield
Gelatin yield was estimated by measuring hydroxyproline recovery by the method
described in (Anonymous 1978), with slight modifications. Dried gelatin (100 mg) was
placed into test tube, and mixed with 5 ml of 6N HCl. The sample solutions were
hydrolyzed for 12 h at 110 C using a dry bath. Absorbance of the solutions was measured
with a spectrophotometer at 558 nm. The hydroxyproline content of the sample solutions
was calculated from a calibration curve derived from standard using analytical grade
hydroxyproline purchased from Sigma Chemical Co.

Determination of proximate components
Moisture content (oven-drying procedure), crude protein (Kjeldahl method), fat content
(Soxhlet extraction) and ash content were estimated by the AOAC official method (AOAC
1990). The analyses were replicated three times.

Determination of gel strength
Gel strength was determined according to the AOAC official method 948.21 (Horwitz,
2000). Gelatin was dissolved with distilled water (6.67%, w/v) at 60 C for 30 min until
The 12
th
16 -18 June, 2011

166

completely dispersed and then kept at 7 C for 17 h. The gel strength was determined by
using the TAXT2 Texture Analyzer Stable Micro System with the following conditions;
plunger, 12.7 mm diameter; penetration depth, 4 mm; penetration speed, 2 cm/min.

To study a large number of factors efficiently, reduced factorial designs were employed.
The screening experiments can provide the information as to which factors are crucial to
the efficiency of extraction. After screening, the results suggested that 4 variables,
concentration of Sodium hydroxide, concentration of Acetic acid, extraction temperature
and extraction time have significant effects on yield (p<0.05), and these key factors were
then selected for the subsequent optimization using RSM. The 4 independent variables and
their range were set (Table 1) and 30 groups of extraction experiments were conducted
according to the CCD design (Table 2).

Table 1 Independent variables and their levels in the 4-factor, 5-level central composite design for
optimizing the extraction condition of Patin gelatin

Symbol Level
Independent variable Coded Uncoded 2 1 0 1 2
NaOH concentration (N) x1 X1 0 0.075 0.15 0.225 0.3
Acetic acid concentration
(N)
x2 X2 0.025 0.05 0.075 0.1 0.125
Extraction temperature
(C)
x3 X3 40 50 60 70 80
Extraction time (min) x4 X4 120 150 180 210 240

The 12
th
16 -18 June, 2011

167

Table 2 Predictive and experimental results of the central composite design for gelatin extraction from
Patin skin

Independent variable
Standard order x1 x2 x3 x4 Y (Exp) Y (Pred)
1 1 1 1 1 52.5345 47.61812
2 1 1 1 1 47.9833 50.95947
3 1 1 1 1 56.1504 55.64895
4 1 1 1 1 64.3774 60.51987
5 1 1 1 1 54.4326 58.49618
6 1 1 1 1 67.2577 64.22145
7 1 1 1 1 61.4802 59.28214
8 1 1 1 1 65.2553 66.53698
9 1 1 1 1 57.766 56.60435
10 1 1 1 1 56.811 54.84112
11 1 1 1 1 62.6776 61.5459
12 1 1 1 1 65.2558 61.31225
13 1 1 1 1 65.2275 64.91709
14 1 1 1 1 64.9163 65.53778
15 1 1 1 1 65.4699 62.61377
16 1 1 1 1 64.0156 64.76404
17 2 0 0 0 56.0427 58.52485
18 2 0 0 0 62.4507 64.01646
19 0 2 0 0 58.5385 58.38121
20 0 2 0 0 61.4331 65.6383
21 0 0 2 0 37.6587 42.88773
22 0 0 2 0 58.3987 57.21758
23 0 0 0 2 60.845 61.91516
24 0 0 0 2 66.1507 69.12845
25 0 0 0 0 62.5884 65.79935
26 0 0 0 0 67.0104 65.79935
27 0 0 0 0 66.5774 65.79935
28 0 0 0 0 67.7835 65.79935
29 0 0 0 0 68.537 65.79935
30 0 0 0 0 62.2994 65.79935

On the basis of the experimental data, the software generated prediction equations
for extraction yield in several formats such as linear, quadratic, cubic, etc. At least 1
significant model was obtained.

Development of response surface model
Regression analysis was employed to fit a full response surface model for every response
investigated including all linear (X
1
, X
2
, X
3
, X
4
), interaction (X
1
X
2
, X
1
X
3
, X
1
X
4
, X
2
X
3
,
X
2
X
4
, X
3
X
4
), and quadratic terms (X
1
2
, X
2
2
, X
3
2
, X
4
2
). To develop the fitted response
surface model equations, all insignificant terms (P >0.05) were eliminated and the fitted
models were shown in table 3.

The 12
th
16 -18 June, 2011

168

Table 3 Response surface model for gelatin extraction from Patin fish skin.

Response Quadratic polynomial model R
2
p-value
Extraction yield Y=65.80+1.81 X
2
+ 3.58 X
3
+ 1.80 X
4
3.94 X
3
2
1.81 X
1
X
2
0.9829 0.0003

Analysis of variance
The analysis of variance (ANOVA) was used to evaluate the significance of the quadratic
polynomial model equation. Any terms in the models with a large F-value and a small
P-value would indicate a more significant effect on the respective response variables (Yuan
and others 2008). Table 4 showed the ANOVA for the models that explained the response
of the dependent variables. The values of R
2
suggested that the models could explain a
high percentage of the variability in the observed data. Thus, the analysis of variance
showed that the predicted 2
nd
order models were statistically suitable. However, the lacks
of fits model were not significant (P >0.05) which indicated the optimum models for this
experiment.

Table 4 Analysis of variance (ANOVA) for response of the dependent variable (Y, %)
Sources Degree of freedom Sum of squares P value
Regression 14 1060.29 0.0003
Linear 4 510.30 0.0100
Square 4 447.30 0.0241
Interaction 6 102.69 0.3106
Residual 15 212.76
Lack-of-fit 10 176.70 0.1673
Pure error 5 36.06
Total 29 1273.05

Conditions for optimum response
For the optimization of gelatin extraction, the four independent variables were:
a concentration of 0.16 N of Sodium hydroxide, a concentration of 0.08 N of Acetic acid,
an extraction temperature of 63.25 C and an extraction time of 210 min.
The experimental yield of gelatin (68.16%) agreed well with the predicted value (67.99%)
obtained by the RSM.

Two three-dimensional response surface plots of the dependent variable (Y) against the
independent variables ([X
1
, X
3
] and [X
3
, X
4
]) were shown in Fig.1a and b. According to
Fig. 1a, the yield of extraction increased gradually as the NaOH Concentration increased,
while the yield of extraction increased sharply with an increase in temperature up to 65C,
beyond which, the yield of extraction decreased.

Enhanced in extraction time caused the yield of extraction raised gradually (Fig 1b).
Fig. 1 showed that the maximum yield observed was significant with the increase in
extraction temperature.
The 12
th
16 -18 June, 2011

169

a b

Figure 1 (a and b) Response surface plots for optimization of gelatin extraction from Patin skin
Chemical compositions of extracted gelatin from Patin

The chemical composition of the raw skin was 30.34% (1.11%) crude protein, 50.86%
(1.67%) moisture, 14.4% (1.88%) fat and 4.6 % (0.13%) ash. Furthermore, the gelatin
extracted from Patin was 92.19% crude protein, 2.85 % (1.14%) moisture,
3.55 % (0.01%) fat and 1.5% (0.01%) ash.

Gel strength
The gel strength of gelatin from the Patin skin, which is a tropical fish, was compared with
the mammalian gelatin. The gel strength of bovine gelatin was 380.7 Bloom. According to
the results of gel strength measured by (Choi and Regenstein 2000), fish gelatins showed
lower gel strength than mammalian skin gelatin. However, Patin fish skin gelatin in the
present study had a 481.3 bloom gel strength, which is remarkable for gel strength of fish
gelatin.

Discussion
Previous studies showed that, the catfish skin gelatin extraction with pretreatment by
alkaline only resulted in a dark-colored gelatin. Pretreatment with acid only has also been
reported and the acid only extraction resulted in some fish oil being left in the gelatin.
(Devictor and others 1995; Gomez-Guillen and others 2002; Gimenez and others 2005).
The alkaline and acidic pretreatments showed effects on removing noncollagenous proteins
with minimum collagen loss, excluding the effect of endogenous proteases on collagen,
causing a significant amount of swelling of fish skin and securing a high gelatin yield and
gel strength by destroying certain chemical crosslinkages present in the collagen with less
breakage of peptide bonds (Jamilah and Harvinder, 2002). It is suggested that the
combination of the 2 pretreatments provided a proper pH for extraction. Hence, based on
these results, a combination of the two pretreatments is applied in this study. The 2-step
RSM was used to optimize the extraction of gelatin from Patin fish (Pangasius sutchi)
skin. It was found that alkaline concentration, acid concentration, and extraction
temperature and time showed significant effects on extraction yield. With these production
conditions, the gelatin extracted also showed significantly higher gel strength than that of
bovine and other cold-water fish species.

References
Abbas KA, Sapuan SM, Mokhtar AS. 2006. Shelf life assessment of Malaysian Pangasius
sutchi during cold storage. Sadhana, 31:635-43.
X1: NaOH Con
X3: Extraction temp
X3: Extraction temp
X4: Extraction time
The 12
th
16 -18 June, 2011

170

Anonymous .1978. Meat and meat products-determination of L(-)hydroxyproline content
(reference method). International standard, ISO 3496(E).
AOAC .1990. In K. Helrich, Official method of analysis (15th ed.). Virginia: Association
of official analytical chemists, Inc.
Araujo PW, Brereton RG. 1996. Experimental design I: screening. Trends Anal Chem
15:2631.
Cho SM, Gu YS, Kim SB. 2005. Extracting optimization and physical properties of
yellowfin tuna (Thunnus albacares) skin gelatin compared to mammalian gelatins.
Food Hydrocolloid 19:2219.
Choi SS, Regenstein JM. 2000. Physicochemical and sensory characteristics of fish gelatin.
J Food Sci 65:1949.
Devictor P, Allard R, Perrier E, Huc A. 1995. Unpigmented fish skin, particularly from flat
fish, as a novel industrial source of collagen, extraction method, collagen and
biomaterial thereby obtained. U.S. patent 5,420,248.
Design-Ease. 2006. Design-Ease 7 manual with tutorials. Available from:
http://www.statease.com/.Minneapolis, MN, U.S.A.: Stat-ease Inc.
Gimenez B, Gomez-Guillen MC, Montero P. 2005. The role of salt washing of fish
skins in chemical and rheological properties of gelatin extracted. FoodHydrocolloid
19:9517.
Gomez-Guillen MC, Turnay J, Fernandez-Daz MD, Ulmo N, Lizarbe MA,
Montero P. 2002. Structural and physical properties of gelatin extracted from
different marine species: a comparative study. Food Hydrocolloid 16:2534.
Horwitz, W. 2000. Official methods of the association of official agricultural chemists
international (17th ed.). Gaithersburg: AOAC International.
Ilona K, Kaczorowski K, Piotrowska B, Sadowaska M. 2004. Modification of the
properties gelatine from skins of ballistic cod (Gadus morhua) with
transglutaminase.
J Food Chem. 86, 203209.
Jamilah B, Harvinder KG. 2002. Properties of gelatins from skins of fish-black tilapia
(Oreochromis mossambicus) and red tilapia (Oreochromis nilotica). J Food Chem.
77, 8184.
Johnston-Banks FA. 1990. Gelatine. In: Harris, P. (Ed.), Food Gels. Elsevier Science,
Essex, pp. 233289.
Kinsella JE, Rector DJ, Phillips LG. 1994. Physicochemical properties of proteins:
texturization via gelation, glass and film formation. In: Yada RY, Jackman RL,
Smith JL. (Eds.) Protein StructureFunction Relationship in Food. Blackie
Academic and Professional, London, pp. 121.
Ladislaus M, Kasankala YX, Weilong Y, Sun D, Hong Q. 2007. Optimization of gelatine
extraction from grass carp (Catenopharyngodon idella) fish skin by response
surface methodology. Bioresource Technology 98 (2007) 33383343.
Norland RE. 1987. Fish gelatin: technical aspects and applications. In S. J. Band (Ed.),
Photographic gelatin (pp. 266281). London: Royal Photographic Society.
Yuan Y, Gao Y, Mao L, Zhao J. 2008. Optimisation of conditions for the preparation of b-
carotene nanoemulsions using response surface methodology. J Food Chemistry,
107, 13001306.

The 12
th
16 -18 June, 2011

171

OB-169

Optimization Production of Instant Dry Yeast Using Mixture of Pineapple
Waste and Liquid Waste of Fermented Soybean Industries

Wignyanto
*
, Suprayogi

, Hendrix Y.S.

Department of Industrial Agricultural Technology, Faculty of Agricultural, University of Brawijaya
*
Corresponding author: mailhendrix@gmail.com

Abstract

The aim of this research is to utilize pineapple and fermented soybean waste industries as
substrate for the growth of yeast. Two stages of study were carried out. The first stage was
the proportion of substrate namely the pineapple extract and liquid fermented soybean (1 : 0, 1
: 0.5, 1 : 1, 1 : 1.5, 0 : 1), which are suitable for the growth of yeast. The second stage of
experiment was to optimize dextrin and lecithin concentration to produce high quality instant
dry yeast. Experimental design used was Central Composit Design with two factors: dextrin
concentrations (50%,60%,70%) and lecithin concentrations ( 0.5%, 1.0, 1.5%). Dextrin was
used as an encapsulating agent and lecithin was used as an emulsifier. The responses analyzed
were water content, solubility, cell viability, and ethanol concentration. Results obtained from
the first stage showed that the best media for the yeast growth was the ratio of pineapple waste
to liquid soybean waste of 1 : 0.5, incubation time of 12 hours, temperature of 30
o
C, agitation
of 150 rpm. The best yeast production was 4.35 x 10
8
cell/mL. Shortest adaptation time was 4
hours. The second stage experiment indicated that the optimal dextrin and lecithin
concentrations were of 63.59 % (w/v) and 1.57 % (w/v), respectively. The optimal water
content, solubility, cell viability, and ethanol concentration were 6.25%, 77.56%, 62.6%, and
4.21 %, respectively.

Keywords: Instant dry yeast, pineapple, fermented soybean waste

The 12
th
16 -18 June, 2011

172

OC : Malaysia Palm Oil Symposium
The 12
th
16 -18 June, 2011

173

OC-I

Palm Oil for Human Consumption:
Nutritional Considerations in Food Formulation

Tilakavati Karupaiah
1
, and Kalyana Sundram
2

1
Assistant Professor, Department of Nutrition & Dietetics, Faculty of Allied Health Sciences,
National University of Malaysia, Kuala Lumpur, Malaysia;
2
Deputy CEO, Malaysian Palm Oil Council, Kelana Jaya, Selangor, Malaysia

Abstract

Amongst vegetables oils, palm oil represents a challenge in food marketing because of the
perception that it is a saturated fat. The hyper-or hypocholesterolemic nature of dietary
fatty acid classes determines cardiovascular risk. Therefore the ideal fat compositional
model advocated by healthy dietary guidelines favor an increase in polyunsaturated
(PUFA) or monounsaturated fatty acids (MUFA) over saturated (SFA) fat content in the
diet. In food manufacturing, rheological properties and shelf and oxidative stabilities of
food products are of prime importance and determined by the solid fat formulation. For
many decades highly unsaturated vegetable oils through partial hydrogenation were able to
satisfy both industry needs and health concerns; and margarine over butter became the
staple on the breakfast table.

Today however trans a by-product of hydrogenation is the new bad fat on the block. With
trans there is a positive linear trend with lipoprotein levels and in fact considered more
hypercholesterolemic than SFA. With partial hydrogenation of vegetable oils no longer
favored, either inter-esterification or use of palm olein becomes the option for the food
industry. However, fatty acid composition, their predominance in various oils and fats, the
position occupied by individual fatty acids on the TAG molecule and the type and amount
of SFA in the diet are still important variables modulating lipoprotein metabolism. This
presentation discusses the nature of fat and cardiovascular risk factors with a particular
reference to palm olein.

The 12
th
16 -18 June, 2011

174

OC-II

Trans Free Solution Incorporating Palm and Palm Fractions

Mohd Suria Affandi Yusoff (PhD)

Sime Darby Research Sdn Bhd, Malaysia

Abstract

Major food producers in the world are working hard to develop alternative methods of
producing shelf-life stable vegetable oils. United State, Denmark and some other
developed countries have shown that eliminating partly hydrogenated vegetable oil and
replacing with other modified fats such as interesterfied , fractionated and even blended
fats does not results in a loss of palatability, availability or increase in cost of these fats.
These structured fats confer upon palm oil physical properties that allow it to be included
in food products especially in the formulation of semi-solid food products such as
margarine, shortening, vanaspati, and creamers. Utilising palm oil and its fractions in such
situations, could virtually eliminate their trans fatty acid arise from partial hydrogenation
process. We are optimistic that palm oil will be shown to be highly desirable and nutritive
edible oil that will continue to be sought after as a replacement for hydrogenated oils.

The 12
th
16 -18 June, 2011

175

OC-III

Palm Bioactive Ingredients for Nutraceutical Applications.

Dr. Syed Fairuz

MALAYSIAN PALM OIL BOARD

Abstract

The oil palm fruit has provided humans with various bioactive substances. Several lipid-
soluble antioxidants namely tocopherols, tocotrienols and carotenoids have been
incorporated into our daily diet through consumption of palm oil. Moreover, numerous
nutritional studies have been undertaken to further understand the role of these lipid-
soluble antioxidants in maintaining health status. Palm tocotrienols have been proven as
efficient free-radical scavenging antioxidants, and have been demonstrated to have
cholesterol-reduction effect in animals and humans. Additionally, several other
nutraceutical potentials such as anti-cancer, anti-ageing and neuroprotective effects have
been postulated for palm tocotrienols based on the current evidence, although their
absorption is still being investigated. Palm oil is also known for its substantial amount of
naturally-derived carotenoids, which most of them are presented as -carotene. As a
precursor of vitamin A, -carotene is widely incorporated into diets to prevent vitamin A
deficiency in many developing countries. Meanwhile it has been less known that the oil
palm fruit also contains potent water-soluble antioxidants. Recently a breakthrough
technology has been developed in harvesting a substantial amount of phenolic acids from
the bio-aqueous co-products of oil palm fruit. Based on emerging evidence that plant
phenolics are beneficial to health, several preliminary investigations in evaluating
physiological effects of oil palm phenolics (OPP) have been initiated. In-vitro and LDL-
oxidation tests have shown that OPP significantly acts as an effective antioxidant in
scavenging free-radical activity while studies in animal models showed that OPP has anti-
hypertensive, anti-cancer, anti-diabetic and anti-atherogenic effects. These groundwork
studies of OPP will provide new insights to enhance understandings in exploring future
nutraceutical effects of water-soluble palm bioactives.
The 12
th
16 -18 June, 2011

176

OC-IV

Positive Attributes of Palm Oil in Deep Frying Applications

Mohd Muslimin Hashim

Manager, Science and Environment Division
Malaysian Palm Oil Council, Kelana Jaya, Selangor, Malaysia

Abstract

Deep Frying is a high temperature frying carried out at a temperature of 185 200C. It is
an efficient method of heat transfer that allows quick cooking and adds flavor to the fried
food. At high-elevated temperature and in the presence of air and moisture, frying oil will
undergo several chemical changes including oxidation, polymerization and hydrolysis
among others. Stability at high frying temperature is the single most important attribute
for deep frying oil. In the fried food and snack food industry, Palm Oil is the preferred
choice for frying oil because it imparts superior shelf life to the final products due to its
high oxidative stability. Unlike the unstable polyunsaturated edible oils, palm oil does not
have to be hydrogenated to impart stability. Hence, it is naturally free of trans fatty acid.
Palm Oil also has balanced fatty acid content with equal ratio of saturated to unsaturated
fatty acids. The presence of natural antioxidants, tocopherol and tocotrienol further
contribute to the superior oxidative stability of palm oil. Another important attributes of
palm oil, which help to distinct it from others, is its bland taste. This helps to carry the
natural flavor of the food during frying process. The most important reason palm oil is the
preferred choice for deep-frying applications globally is because it is easily available at
anytime and is the most cost effective edible oil among many.

The 12
th
16 -18 June, 2011

177

OD : Food Safety and Quality Management
The 12
th
16 -18 June, 2011

178

OD-32
Microbiological Survey of Thai Exported Vegetables from Farm to
Export and Its Contamination between Washing Process
Pornpen Morakotjinda
1
, Warapa Mahakarnchanakul
1,*
, Donald W. Schaffner
2
, Nipa
Chokesajjawatee
3
, Siriporn Stonsaovapak
4
, Sudsai Trevanich
1
, Wipawadee Ontoum
1

1
Department of Food Science and Technology, Faculty of Agro-Industry, Kasetsart University, Bangkhen
Campus, Bangkok, 10900, Thailand;
2
Department of Food Science, Rutgers University, 65 Dudley Road,
New Brunswick, New Jersey 08901-8520, USA;
3
National Center for Genetic Engineering and
Biotechnology, 113 Phaholyothin Rd, Klong 1, Klong Luang, Pathumthani 12120, Thailand;
4
Institute of
Food Research and Product Development, Kasetsart University Bangkok 10900, Thailand

*
Corresponding author: fagiwpm@ku.ac.th
Abstract
Fresh produce is a major exported agricultural commodity to Europe, however over the last
five years fresh produce companies encountered the problem of Salmonella spp. and
Escherichia coli contamination on exported produce which caused the negative impact on
Thai business chain. The sources of contamination on vegetables were identified as farm
environments, equipments and poor practices in packing house, including washing process
without any application of sanitizers. Since there was no sanitizer and using of repeat
water, the microbiological survey showed Salmonella spp. and E. coli counts increased on
produce after washing ranged from 0.96 to 2.94 log CFU/g. Thus further experiment was
conducted to explicit the possibility of cross contamination when using repeat water
compared to water with and without 50 ppm sodium hypochlorite. Enterobacter
aerogenes was used as the surrogate of Salmonella spp, while basil and coriander were
used as the fresh produce model. After washing the inoculated basil and coriander in tap
water (without shaking for 5 min) resulted in 0.60 and 0.51 log unit reduction (from the
initial load 4.17 and 4.21 log CFU/g), respectively. E. aerogenes cells transferred to wash
water by 2.97 to 3.05 log CFU/ml. The second, third and fourth washing were done with
un-inoculated vegetables and E. aerogenes counts in repeat water (without chlorine) were
3.01, 3.01 and 3.02 log CFU/ml, respectively. E. aerogenes was transferred by repeat
water to un-inoculated basil by 2.23 to 2.58 log CFU/g. In case of coriander, the E.
aerogenes counts in repeat water were 3.05, 3.03 and 3.05 log CFU/ml, respectively, while
E. aerogenes was transferred to produce by 2.31 to 2.60 log CFU/g. If adding 50 ppm
sodium hypochlorite in wash water, it noticeably resulted in reduction of E. aerogenes in
inoculated basil and coriander by 1.25 and 1.03 log CFU/g, respectively. No detectable in
un-inoculated basil, un-inoculated coriander and repeat chlorinated water (remained free
chlorine 32.6 ppm) was observed. This experiment showed the adding chlorine as the
sanitizer was needed in washing process which remarkably kill pathogenic bacteria and
prevent the cross contamination between contaminated vegetables through the wash water
to clean vegetables.
Keywords: Fresh produce, washing process, cross contamination, pathogenic bacteria,
sanitizers
The 12
th
16 -18 June, 2011

179

Introduction
Fresh produce is one of the most exported agricultural products from Thailand. There is a
great potential market for health food. The range of microorganisms associated with
outbreaks linked to fresh vegetables encompasses bacteria, viruses and parasites. Mostly, it
was reported as bacteria particularly members of the Enterobacteriaceae. Of these,
Salmonella spp. and Escherichia coli O157 in sprouted seeds and fruit juices are of
particular concern. Fruits and vegetables normally carry a non-pathogenic epiphytic
microflora. However, many factors contribute to the microbiological contamination of
these products with pathogens. Contamination can arise as a consequence of cultivation
(soil, organic fertilizers and irrigation water), post-harvest (trimming, cutting, washing)
and transporting (Beuchat 1996; Doyle 2008; Keller 2009; Taormina 2009). Another
aspect contributing to the microbial risk for the consumer is consuming the raw vegetables
(European Commission 2002).

During 2005-2006, the fresh produce exported from Thailand, including coriander, curry
leaves and four varieties of basil was detected as Salmonella spp. contamination by the
Health Protection Agency (HPA) at the point of entry (Border inspection post) in London
and a pan-London retail (Surman-Lee 2008; Johannessen 2009). Moreover, in 2007, ready-
to-eat fresh herb (3,760 samples) was collected from retail premises in UK for determining
the presence or absence of Salmonella spp. and level of E. coli. Fifteen samples of Thai
exported fresh herb were collected and the results showed that 4 samples were
contaminated with E. coli which over limitation (2 log CFU/g) but absent of Salmonella
spp. (Elviss 2009). From that incidence of pathogen contamination in exported fresh
produce affected on economy of Thailand.

The microorganism decontamination by using antimicrobial agents in fruits and vegetables
was wildly reported (Han 2000; Keskinen 2009; Zhang 2009; Lpez-Glvez 2010).
Various antimicrobial agents can be used to reduce the microbial load on fruits and
vegetables. The most common antimicrobial agents used are chlorine-based compounds
with free chlorine concentrations of 50100 ppm (WHO/FAO 2008). In many researches,
surrogate microorganisms were used due to pathogen manipulation. For instance,
Enterobacter aerogenes was used since it is a non pathogenic surrogate with attachment
characteristics similarly to Salmonella spp. (Zhao 1998; Chen 2001). So, this work was
conducted to study 1) the microbiological quality of Thai exported fresh produce from
farm to export and 2) the contamination between washing process by using Enterobacter
aerogenes as model during washing process with and without sodium hypochlorite.
Collection of samples
The samples (n=330) were collected from farm environment (seeds, soil, irrigation water,
manure), packing house environment (gloves/hands, scissors, wash water, tables),
transporting process (basket, cover material) and vegetables (sweet basil and coriander) at
various step (after trimming, after washing, after transporting to factory and after
transporting to airport). The farms and factories were located in Nakhonpathom province,
Thailand.

The 12
th
16 -18 June, 2011

180

Microbial analysis
The environment and vegetable samples were analyzed the amount of Salmonella spp. by
using the PCR technique combination with MPN method. To count E. coli, the E.
coli/Coliform Count Plates Petrifilm (3M Petrifilm) was applied.

Strain isolation
To identify Salmonella spp., the samples were mixed with buffered peptone water (BPW,
Difco, and Detroit, Mich) and incubated at 37
o
C for 18 to 24 h. One ml. of enrich broth
was transferred to Rappaport-Vassiliadis broth (RV broth, Difco, Detroit, Mich) then
incubated at 42
o
C for 18 to 24 h. After that, the selective culture was streaked onto Xylose
Lysine Desoxycholate Agar and Hektoen Enteric Agar (XLD and HE, Difco, Detroit,
Mich) and incubated at 37
o
C for 18 h. The suspect colonies were tested with biochemical
test. Then, serological tests were performed when the results showed the positive for
Salmonella spp. at SAP laboratory (Thailand).

Bacteria strains and inoculum preparation
The nonpathogenic food-grade strain of Enterobacter aerogenes that is resistant to
nalidixic acid, which allows it to be quantified in the presence of other microorganisms in
food and of resident bacteria on the vegetables, was used in this work. E. aerogenes cells
were grown overnight (18 to 24 h) at 37
o
C in tryptic soy broth (Difco, Detroit, Mich.)
containing 50 g/ml. nalidixic acid (Sigma Chemical Co., St. Louis. Mo.). Cells were
harvested by centrifugation (Micro 7: Fisher Scientific, Pittsburgh, Pa.) at 5,000Xg for 3.5
min and washed three times in phosphate buffered saline (0.1 M, pH 7.2) (Fisher
Scientific). Cell pellets were resuspended in phosphate buffered saline. The initial
concentration of solution was 8 log CFU/ml. Finally, the bacteria suspension was adjusted
to 4-5 log CFU/ml.

Preparation of antimicrobial solution
Solution with 50 mg/l active chlorine was prepared by adding the appropriate volume of a
concentrated solution of sodium hypochlorite (NaClO) to water. pH of the sodium
hypochlorite solution was adjusted to 6.8 using hydrochloric acid (HCl). Washing
solutions were prepared one day before application. The initial free chlorine concentration
was measured by using an iodine-sodium thiosulfate redox titration (Oxidizer Kit 322,
Ecolab).

Background checking
For detecting Enterobacter aerogenes, Fresh basil and coriander were obtained from a
supermarket. Damaged leaves and stems were removed Representative 10 gram samples
each items were tested to confirm the absence of the test strain by added 90 ml of peptone
water and homogenized for 1 min in a masticator. Serially diluted technique was applied
and spread plated onto MacConkey agar (Difco) containing 50 g/ml nalidixic acid. Agar
plates were incubated at 37C for 24 h.

Contamination on fresh vegetables
Fresh basil (Ocimum basilicum) or coriander (Coriandrum sativum) was completely
submerging into the bacteria suspension (~10
4
for basil and ~10
5
for coriander CFU/ml)
and shaking by manual for 5 min at room temperature. After inoculation, the samples were
dried in a biosafety cabinet for 30 min. Inoculated air-dried samples were stored in a
The 12
th
16 -18 June, 2011

181

sterilized bag at 4C for 24 h before use. The appropriated level of E. aerogenes on basil
and coriander was 10
4
CFU/g.

To study the possibility of cross contamination between washing process during washing
process with and without sodium hypochlorite while basil and coriander were used as the
fresh produce model.

First washing process: Ten gram of inoculated basil and coriander were washed by
submerging into tap water or chlorinated water at room temperature for 5 min. After 5 min,
the wash water and the inoculated basil or coriander were taken and immediately
neutralized by adding 0.05% sterile sodium thiosulphate (Na
2
S
2
O
3
) by weight of
sample/volume of neutralize solution ratio 1:5 for stopping any residual bacteriostatic or
bactericidal activity for 1 min. After that 40 ml of peptone water was added mixed for 1
min. Then, the serially diluted technique was applied and plated onto MacConkey agar
containing 50 g/ml nalidixic acid and incubated at 37C for 24 h.

Second, third and forth washing process: Ten gram samples of un-inoculated basil or
coriander were washed by submerging into the reused water for 5 min. The wash water and
the un-inoculated basil or corianders were taken and immediately neutralized by adding
0.05% Na
2
S
2
O
3
for 1 min. After that 40 ml of peptone water was added mixed for 1 min.
Then, the serially diluted technique was applied and plated onto MacConkey agar
containing 50 g/ml nalidixic acid. Agar plates were incubated at 37C for 24 h.

From sweet basil farm environments, Salmonella spp. and E. coli were detected in each
source (Table 1). For Salmonella spp., the prevalence of Salmonella spp. was high in the
irrigation water and seed whereas the prevalence of E. coli was high in soil and seed. The
concentration of Salmonella spp. was high in soil and seed. About E. coli, the high
concentration was found from seed and soil. But the prevalence of E. coli was high in
soil and irrigation water. So, soil and irrigation water might be the sources of
contamination at farm environment.

Table 1 Salmonella spp. and E. coli concentration in sweet basil farm environments
Salmonella spp. E. coli Samples
Positive Number
(MPN/g,ml)
Positive Number
(log CFU/g,ml)
Soil
(n=15)
(5/15) <30 - 30 (11/15) 0 - 3.31
Manure
(n=15)
(4/15) <3 - 15 (1/15) 0 - 0.7
Irrigation water (n=15) (8/15) <3 - 15 (9/15) 0 - 2
Seed (n=3) (2/3) <30 - 30 (3/3) 4.81 - 4.86

For coriander farm environments (Table 2), the amounts of Salmonella spp. and E. coli
contamination in irrigation water, fertilizer and seed were detected in low concentration
because of theirs source. The underground water was used as the irrigation water,
moreover the composted manure was used as fertilizer. About coriander seed, the
commercial seed was used for cultivation. Then, the amounts of Salmonella spp. and E.
coli contamination were low. So, soil might be the source of E. coli contamination.
The 12
th
16 -18 June, 2011

182

Table 2 Salmonella spp. and E. coli concentration in coriander farm environments

Positive Number
(MPN/g,ml)
Positive Number
(log CFU/g,ml)
Soi l(n=15) (7/15) <3 15 (8/15) 0 - 3.00
Manure (n=15) (14/15) <3 15 (0/15) 0 - 0
Irrigation water(n=15) (7/15) <3 20 (0/15) 0 - 0
Seed (n=3) (0/3) <30 (0/3) 0 - 0

For sweet basil and coriander packing house and field environments, washing water
showed the high amounts of Salmonella spp. and E. coli (Table 3 and 4). It might be the
potential source of contamination because the wash water was not changed during process
and the disinfectant was not applied for reducing the microbial load in vegetables or
contamination between wash water and vegetables. Moreover, baskets for containing basil
and coriander after washing were contaminated with high level of Salmonella spp. and E.
coli. Additionally, Salmonella spp. and E. coli were found from table and gloves with high
level.

Table 3 Salmonella spp. and E. coli concentration in sweet basil packing house and field environments

Positive Number
(MPN/g, ml)
Positive Number
(log CFU/cm
2
,ml)
Tables
(n=6)
(6/6) 0.06 2.40 (4/6) 0 - 2.98
gloves
(n=8)
(7/8) <0.30 1.50 (6/8) 0 - 3.04
Scissors
(n=10)
(3/10) <0.30 1.60 (6/10) 0 - 1.48
Wash water
(n=3)
(3/3) <0.30 2.80 (3/3) 1.18 - 3.08
Baskets
(n=5)
(3/5) <0.30 2.80 (5/5) 1.70 - 2.56
Cover material
(n=5)
(5/5) 0.06 4.20 (1/5) 0 - 1.70

Table 4 Salmonella spp. and E. coli concentration in coriander packing house and field environments

Positive Number
(MPN/g, ml)
Positive Number
(log CFU/cm
2
,ml)
Table (n=10) (9/10) <0.06 5.80 (5/10) 0 - 3.76
Hand (n=10) (7/10) <0.06 0.56 (7/10) 0 - 3.76
Wash water (n=10) (9/10) <3.00 35 (9/10) 0 - 3.22
Baskets (n=5) (5/5) <0.06- 0.20 (5/5) 2.00 - 3.78
Cover material l(n=6) (4/6) <0.30 4.30 (6/6) 2.28 - 3.81

The concentration and prevalence of Salmonella spp. and E. coli contamination in sweet
basil and coriander were shown in figure 1 and 2, respectively. The amounts of Salmonella
spp. contamination in vegetables before washing were lower than after washing. It showed
The 12
th
16 -18 June, 2011

183

that the wash water could not only reduce the microbial level but also be the source of
contamination.

Figure 1 Concentration and prevalence of Salmonella spp. in vegetables

Figure 2 Concentration and prevalence of E. coli in vegetables

To identify the serotype of Salmonella contamination in sweet basil and coriander, the
results showed in Table 5. For sweet basil, Salmonella Hvittingfoss (group I) was found in
sample before washing. Singapore (group C) and also Weltevreden (group E) serotype
were found in exported sweet basil, as well as Aberdeen (group F) and Bovismorbificans
(group C) in gloves and table. And group J II 17: g,t: - (O: 17) was found from basket. For
coriander, Augustenborg (group C) were found in samples before washing, after soil
remove, after transporting to factory and in washing water. In addition Singapore (group
C) was isolated from exported coriander. Serotype Newport (group C), Albany (group C)
and IIIb 48: 1, V: 1, 5, 7 (O: 48) were found in coriander samples after washing, irrigating
water and table, respectively.
The 12
th
16 -18 June, 2011

184

Table 5 Isolation of Salmonella spp. from environments, sweet basil and coriander

The effect of washing by reused water on E. aerogenes contamination on vegetables and
wash water was shown in table 6. The initial E. aerogenes levels of inoculated basil and
coriander before washing were 4.17 0.14 and 4.21 0.22 log CFU/g, respectively.
After washing in tap water without shaking for 5 min, E. aerogenes counts in inoculated
basil and coriander were reduced by 0.63 and 0.59 log CFU/g, respectively. E. aerogenes
cells transferred to wash water by 2.97 (basil) and 3.05 (coriander) log CFU/ml. For
investigating the cross-contamination between wash water and herbs, 3 branches of un-
inoculated basil and coriander, each branch was washed in the reused water for 5 min
(previously washed water). The levels of E. aerogenes on un-inoculated basil and un-
inoculated coriander after washing in reused water for 5 min were reaching to 2.23 and
2.31 log CFU/g, respectively. After washing each 5 min, the second, third and fourth
washing process were done with un-inoculated vegetables, E. aerogenes counts in repeat
water (without chlorine) were 3.01, 3.01 and 3.02 log CFU/ml, respectively with no
significant difference (=0.05). E. aerogenes was transferred by repeat water to un-
inoculated basil by 2.23 (second) to 2.58 (fourth) log CFU/g. In case of coriander, the E.
aerogenes counts in the second, third and fourth repeat water were 3.05, 3.03 and 3.05 log
CFU/g, respectively with no significant difference (=0.05). When adding 50 ppm sodium
hypochlorite in wash water, noticeably resulted in reduction of E. aerogenes in inoculated
basil and coriander by 1.25 and 1.03 log CFU/g, respectively, was shown. No detectable
in un-inoculated basil, un-inoculated coriander and repeat chlorinated water was observed.
Since the available chlorine still remains in wash water (remained available chlorine 32.6
ppm).

Type of vegetables
Sources Serotype
Salmonella Aberdeen (group F) Hands
Salmonella Bovismorbificans (group C)
Salmonella Aberdeen (group F) Tables
Salmonella Bovismorbificans (group C)
Basket Salmonella II 17 :g,t:- (O:17)(group J)
Sweet basil before washing Salmonella Hvittingfoss (group I)
Salmonella Singapore (group C)
Sweet basil
Exported sweet basil
Salmonella Weltevreden (group E)
Irrigation water Salmonella Albany (Group C)
Wash water Salmonella Augustenborg (Group C)
Table Salmonella IIIb 48 : 1, V : 1, 5, 7 (O:48)
Salmonella Augustenborg (Group C) Coriander before washing
Salmonella IIIb 48 : 1, V : 1, 5, 7 (O:48)
Coriander after soil removing Salmonella Augustenborg (Group C)
Coriander after washing Salmonella Newport (Group C)
Coriander at factory Salmonella Augustenborg (Group C)
Coriander
Exported coriander Salmonella Singapore (Group C)
The 12
th
16 -18 June, 2011

185

Table 6 E. aerogenes on inoculated herbs, un-inoculated herbs and wash water with and without
sodium hypochlorite
Mean SD
Enterobacter aerogenes population (log CFU/g,ml)
Antimicrobial agent Samples
Initial load 5 min First Second Third
Inoculated basil
4.17 0.14 3.54 0.28 - - -
Uninoculated basil
ND - 2.23 0.32 2.27 0.35 2.58 0.25
No
Wash water
ND 2.97 0.10 3.01 0.16 3.01 0.13 3.02 0.22
Inoculated basil
4.17 0.14 2.92 0.12 - - -
Uninoculated basil
ND - ND ND ND
50 ppm Sodium hypochlorite
Wash water
ND ND ND ND ND
Inoculated coriander
4.21 0.22 3.70 0.52 - - -
Uninoculated coriander
ND - 2.31 0.28 2.25 0.13 2.60 0.18
No
Wash water
ND 3.05 0.05 3.05 0.01 3.03 0.04 3.05 0.14
Inoculated coriander
4.21 0.22 3.18 0.07 - - -
Uninoculated coriander
ND - ND ND ND
50 ppm Sodium hypochlorite
Wash water
ND ND ND ND ND
ND: not detected (Detection limit: wash water = 1 CFU/ml, vegetables = 0.1 CFU/g).
The results indicated that washing vegetables with 50 ppm sodium hypochlorite could
prevent the cross contamination between wash water and un-inoculated vegetables. Kim
and others (2009) showed that iced can be carrier by ice contaminated with the pathogen or
by transferred from lettuce surfaces via melting ice. Chlorine washing is enough for
preventing E. aerogenes transference from inoculated vegetables to wash water and from
inoculated vegetables to un-inoculated vegetable via wash water. In addition, chlorinated
water is not enough to eliminate E. aerogenes attaching on vegetables but it showed in
more efficient pathogen removing on the contaminated vegetables. Adding chlorine as the
sanitizer was needed in washing process for killing pathogenic bacteria and preventing the
cross contamination between contaminated vegetables through the wash water to clean
vegetables.
References
Annous BA, Solomon EB, Niemira BA. 2006. Biofilms on fresh produce and difficulties in
decontamination. Food Quality. 2006. 13:80-4.
Beuchat LR. 1996. Pathogenic microorganisms associated with fresh produce. J Food Prot.
59(13):204-16.
Doyle MP, Erickson MC. 2008. Summer meeting 2007 the problems with fresh produce:
an overview. J Appl Micro. 105(2):317-30.
Elviss NC, Little CL, Hucklesby L, Sagoo S, Surman-Lee S, Pinna ED, Threlfall EJ. 2009.
Microbiological study of fresh herbs from retail premises uncovers an international
outbreak of salmonellosis. Int J Food Microbiol. 134(1-2):83-8.
European Commission. 2002. European Commission (EC): Risk profile on the
microbiological contamination of fruits and vegetables eaten raw, Report of
scientific committee on food, SCF CS FMH SURF Final 29 April 2002.
Available from: www.europa.eu.int/comm/food/fs/sc/scf/out125_en.pdf. Accessed
Feb 5.
The 12
th
16 -18 June, 2011

186

Han Y, Sherman DM, Linton RH, Nielsen SS, Nelson PE. 2000. The effects of washing
and chlorine dioxide gas on survival and attachment of Escherichia coli O157: H7
to green pepper surfaces. Food Microbiol. 17(5):521-33.
Johannessen, GS. and Cudjoe KS. 2009. Regulatory issues in Europe regardingfresh fruit
and vegetable safety. In: The produce contamination problem. San Diego:
Academic. p 331-52.
Keller SE. 2009. Microbial contamination of fresh produce. In: Intentional and
unintentional contaminants in food and feed. American Chemical Society. 25-45.
Keskinen LA, Burke A, Annous BA. 2009. Efficacy of chlorine, acidic electrolyzed water
and aqueous chlorine dioxide solutions to decontaminate Escherichia coli O157:H7
from lettuce leaves. Int J Food Microbiol. 132(2-3):134-40.
Kim JK, Harrison MA. 2009. Surrogate selection for Escherichia coli O157:H7 based on
cryotolerance and attachment to romaine lettuce. J Food Prot. 72:1385-91.
Lpez-Glvez, F, Allende A, Selma MV, Gil MI. 2009. Prevention of Escherichia coli
cross-contamination by different commercial sanitizers during washing of fresh-cut
lettuce. Int J Food Microbiol. 133(1-2):167-71.
Lpez-Glvez F, Gil MI, Truchado P, Selma MV, Allende A. 2010. Cross-contamination
of fresh-cut lettuce after a short-term exposure during pre-washing cannot be
controlled after subsequent washing with chlorine dioxide or sodium hypochlorite.
Food Microbiol. 27(2):199-204.
Surman-Lee S, Murphy N, Pathak K, Clements J, Pinna ED. 2008. A pan London study of
the microviological quality of fresh herbs. Abstract 262 at Health Protection 2008.
Available from: www.hpa-events.org.uk/hp2008. Accessed Aug 28.
Taormina PJ, Beuchat LR, Erickson MC, Ma L, Zhang G, Doyle MP. 2009. Transfer of
Escherichia coli O157:H7 to iceberg lettuce via simulated field coring. J Food Prot.
72(3):465-72.
World Health Organization and Food and Agriculture Organization of the United Nation
(WHO/FAO). 2008. Microbiological hazards in fresh leafy vegetables and herbs :
Meeting report. Microbiological risk assessment series 14. Geneva. 151p.
Zhang, G, Ma L, Phelan VH, Doyle MP. 2009. Efficacy of antimicrobial agents in lettuce
leaf processing water for control of Escherichia coli O157:H7." J Food Prot.
174(72):1392-7.
Zhao, P, Zhao T, Doyle MP, Rubino JR, Meng J. 1998. Development of a model for
evaluation of microbial cross-contamination in the kitchen. J Food Prot. 61:960-3.
The 12
th
16 -18 June, 2011

187

OD-58
Perceptions of European Gatekeepers towards
Thai Fruit and Coffee Products with EU Geographical Indication
Rungsaran Wongprawmas
1,*
, Maurizio Canavari
1
, Rainer Haas
2
, Daniele Asioli
1

1
Department of Agricultural Economics and Engineering, Alma Mater Studiorum-University of Bologna,
Bologna, Italy;
2
Department of Economics and Social Sciences Institute of Marketing & Innovation,
University of Natural Resources and Applied Life Sciences, Vienna, Austria
*
Corresponding author: rungsaran.wongprawmas80@gmail.com

Abstract
This study is aimed at exploring perceptions of European gatekeepers towards renowned
Thai fruit and coffee products protected by geographical indication (GI) and factors
influencing purchasing decision of gatekeepers towards food products imported from
Thailand. Sixteen qualitative interviews with distribution channel gatekeepers were
administered in Austria, Italy and Switzerland in 2010. The interviewees are food
distribution practitioners and experts and are key informants for imported fruits and coffee
in Europe and they were asked for an opinion about recognition of Thai GIs in the EU
system. Content analysis and concept mapping were used to analyze data. Results show
that Thai GIs products might be interesting for European gatekeepers, but the GI attribute
alone might not be sufficient to ensure the product is successful. Support of consistent
information and promotion campaigns and fulfillment of other gatekeepers requirements
of both products and suppliers are also necessary. Eight major factors have been identified,
which influence European gatekeepers decision to purchase imported food products:
quality, price, food safety, environmental aspects, social aspects, business relationship,
consumer awareness and preference, and competitors. Results are useful to develop
appropriate managerial marketing strategies to introduce these GI products into the EU
market.

Keywords: Geographical indications, Thai fruits and coffee, gatekeepers, perception,
factors influencing purchasing decision

Introduction
Thailand is one of the main producers of agricultural products and is recognized as a
producer of great unique and tasty products (Chomchalow and others 2008). In 2007,
Thailand produced 8.95 million metric tons of fruit, which is 1.61% of the world fruit
production (FAO 2009). During last few years Thailand has been trying to penetrate the
European market, since it is the largest single market with expanding demand for ethnic
products and with high premium price for quality produce (GTZ 2010). Thai fruit export is
mainly represented by tropical fruit (e.g. longans, mangoesteen, durians, etc.). The value of
EU fruit import from Thailand steadily grew during last eight years, from 216.64 million
euro in 2002 to 274.65 million euro in 2009, now representing 1.88% of total fruit trade
(elaborated from EU Market access database 2010). This is a small share in comparison to
other trading partners of the EU. Coffee trade between Thailand and the EU is mainly
represented by green bean coffee export to the EU and (semi) finished coffee products
The 12
th
16 -18 June, 2011

188

import from the EU. Thailands share of the EU coffee imports ranged only between 0.01-
0.26% in the period 2002-2006 (ECF 2009).

Thai government is currently looking for strategies to improve quality standard and
differentiation of products so as to enhance the competitiveness of Thai exports, since
Thailand cannot compete on price with competitors like China and Vietnam anymore. In
addition, there is growing interests in the world market toward quality food products,
especially high quality local and traditional food specialties, which may pave the way for
specialty products from less developed countries to gain access to lucrative international
markets (Canavari and others 2010). As a result, Thai government endorsed policies and
regulations - the Geographical Indication Act of B.E. 2546 (2003); and Thai
Geographical Indication Logo Approval B.E. 2008 (Jaovisidha 2003) in order to protect,
foster, and promote Thai quality products and culinary recipes. They expected that
Geographical Indications (GIs) labels will help consumers to recognize quality of the
products by linking them with geographical characteristic and culinary heritage of
Thailand. Thus, they aim to enhance the credibility of products, improve market access,
promote typical products, and finally lead to sustainable economic and social development
in a specific territory (Bramley and Kirsten 2007).

As of June 2010, 35 product names are registered as GIs in Thailand, 29 of which are from
Thailand, 5 from the EU and one from Peru. Registration is pending for 34 more products
(Department of Intellectual Property 2010). The most represented product categories are:
fruit (23%), handicrafts (23%), rice (20%), alcohol (17%), and coffee (6%). Furthermore,
during last few years, Thailand applied to register some Thai products under the EU
regulation, namely: Thung Kula Rong-Hai Thai Hom Mali Rice, Doi Chaang Coffee
and Doi Tung Coffee (DOOR database, European Commission 2010). Motivation is the
belief that Thai GIs will be perceived as higher quality products and will gain a better
position in the EU market if recognized by the EU.

The importance of GI labels in the European market
In the last decades, European consumers became more thoughtful about their purchasing
decision, especially about product quality and food safety as well as environmental and
social impacts associated with food production and marketing (Bredahl and others 2001).
One of the quality cues frequently used in the EU market especially in wine and food
products is related to product origin. The EU has decades long history of GIs labeling and
currently two types of GIs are available for food products: protected designation of origin
(PDO) and protected geographical indication (PGI). Producers and companies have been
using GIs label to signal quality to the consumer and expected that they will perceive GIs
as a quality attribute adding value to the product. Several studies in the literature indicated
that perceived quality associated with the intrinsic attribute of PDO products have a
positive and significant influence on European consumers buying intentions (Fandos and
Flavin 2006) and they are willing to pay higher prices for PGI label in fresh meat
(Loureiro and Mc Cluskey 2000). Hence, GIs might signal quality to the consumer and
create a sense of affection to consumers, but only when it deals with a high quality product
and consumers recognize its superior quality, using GI labels as a brand.

The 12
th
16 -18 June, 2011

189

GIs products from third countries in the European market
Many successful cases of GIs products value-enhancement are available in Europe,
especially in Southern Mediterranean countries. However, most of these products are
produced in small local areas inside Europe, which were already famous among consumers
(Barjolle and Sylvander 1999). For Thai GIs products this is not the case, since consumers
might not be able to perceive their value because of lack of knowledge, awareness, and
familiarity with products (Boccaletti 1999). In addition, successful value-enhancement
strategies for GI products do not depend only upon origin but also upon effective value-
adding strategies used to promote the product (Bardajand others 2009). Vandecandelaere
and others (2009) highlighted that third countries that would like to develop fruitful GIs
schemes should be aware of a number of requirement to be achieved in advance, such as
traceability along the supply chain, responsibility and accountability of producers,
processors and marketers, as well as clear and sound marketing strategies.

Issues addressed in this study
Since introduction of Thai GIs into the EU market is still in the initial stage, products have
to be chosen by European gatekeepers first. Gatekeepers are food channel members who
have the power to select products to be in the market. Even though they are expected to
merely select products according to their perception of consumers demand, in many cases
they may also exert a great influence on product ranges in the market due to their power to
make food buying decisions on behalf of food importing and distribution companies that
supply millions of end consumers (Knight and Gao 2005).

Hence, this research focuses on gatekeepers perceptions toward Thai GIs fruits and coffee.
We aim at gaining deeper understanding of the attitudes of gatekeepers and the possibility
to use GI labels to foster the competitiveness of Thai fruits and coffee products in the EU
market. This research aims to (1) explore how European gatekeepers perceive GIs Fruit
and Coffee products from Thailand, (2) explore potential and barriers for GIs Fruit and
Coffee products from Thailand in the European markets and (3) explore the key factors
that influence purchasing decision of European gatekeepers.

Exploratory research based on qualitative approach is adopted in this study. This approach
makes us able to deal with complexity and rich diversities of the gatekeeper perceptions
and it gives room to generate hypotheses for further research, although it cannot be
generalized to all food chain members and channels (Myers 2009). A semi-structured
interview schedule was designed to serve as a non-binding outline of the discussion with
respondents, following the three research aims mentioned above. The schedule contains a
series of open-ended questions introducing wide topics and inducing the informant to raise
salient issues which he or she thinks are important and relevant to the topic of interest
during the conversation (Myers 2009).

Selection of respondents
Purposive non stochastic sampling was applied to recruit participants in this study to
retrieve information from persons who have knowledge and might be able to highlight the
relevant problems or issues on a specific topic (Trochim 2006). In addition, the snow-ball
sampling procedure was also applied.

The 12
th
16 -18 June, 2011

190

A list of European operators was created on the basis that those listed are expert and
professional practitioners in European food distribution (importers, wholesalers, retailers,
practitioners and experts) who already have experiences dealing with Thai or Asian
products in Europe. We interviewed sixteen out of thirty-five selected contacts ten in
Italy, five in Austria and one in Switzerland.

The characteristics of the interviewed companies were the following: thirteen respondents
were importers and distributors of fruit and food products and three respondents were
researchers and experts on agri-food marketing and European fruit markets. Among
importers and distributors, six companies were fruit and/or vegetable distributors, four
companies were specialty shops, and the remaining were representatives of large retail
companies. Among researchers and experts, two were specialized in organic products only.
Given the limited geographical scope covered by this sample, conclusions and hypotheses
drawn by means of this research are most likely to be valid just in the Austrian and Italian
situations.

Interview procedure
Sixteen semi-structured interviews were administered during March June 2010. The
semi-structured interview schedule was sent to respondents in advance. Personal
interviews duration ranged between 30-60 minutes. They were conducted in English, in
Italian and in Thai according to respondents preference. Thirteen interviews were face-to-
face and three interviews were administered by telephone. Interviews were recorded if the
respondent agreed upon, and the interviewer took note of important information and
observed context-specific elements during the interview. Immediately after the interview
was administered, the interviewer prepared a summary report based on notes and on the
recorded conversation, when available.

Data analysis
Information from the summary reports, together with available transcription and comments
were analysed through a content summarizing procedure, aimed at describing the
phenomenon and at presenting the most interesting elements arising from each interview,
in order to gain an extensive overview of informants attitudes toward the topic (Downe-
Wamboldt 1992). From the analysis and synthesis of relations among factors affecting
purchasing decisions a conceptual map was created, in order to visually describe the
concepts mentioned by key informants and connections among them (Novak and Caas
2008).

Results
Perception of European gatekeepers toward Thai GI fruit and coffee products
It appeared that European consumers and gatekeepers are aware of the GIs concept mainly
as it relates to local or European products, since GIs products strongly link to territory and
traditional local culture. European consumers have already developed positive perceptions
toward these products and the GIs label can therefore be used as a differentiation tool.
However, when dealing with Thai GIs products, respondents made many and various
comments.

The 12
th
16 -18 June, 2011

191

Most of the gatekeepers thought that the GIs label may not be able to add value to Thai
fruit and coffee products because consumers do not have enough information about or
experience with these products:
This business is too small in quantity. For me, you have to work with quality freshness,
information. Because for me, in my opinion, when you have many fruits, you use PGI to
differentiate. This is small market, so it cannot be differentiated.(Importer and
Distributor in Italy)

On the contrary, other gatekeepers maintain that the use of a well-known quality label such
as European public brands for GIs could represent an advantage for Thai GIs products
because of confidence and trust that European consumers ascribe to them:
If you have an origin certification of the product, it could be a way to enforce the
knowledge, the national of the product. For the European market, I think it could be
interesting, if you have this kind of certification between the European and Thai at the
same level of quality assurance, origin and traceability and so on. I think that it could be a
possibility especially for fruits. (Marketing researcher in Italy)

Some gatekeepers and marketing researchers thought that the GIs label might be useful as
a mediator of trust to assure the quality and food safety of Thai products:
Certainly, PGI would help us to have warranties and also suppliers. They would enter
more easily the market. It would be an advantage for consumers, too. (Importer and
Distributor in Italy)

Although GIs certification is not a food safety guarantee, they argued that it could be seen
as such in the eyes of the public. This is due to the fact that gatekeepers and consumers
tend to be more sensitive to imported food than domestic products, due to food scares and
general unfamiliarity; hence, they tend to look for any sort of certification and may regard
this as a sign of food safety and quality. Gatekeepers also mentioned that traceability
systems are crucial for products entering the EU market.

One respondent opined that GIs label would not be useful for Thai products because he
gave more value on his quality private brand than to GIs certification. This inferred that
GIs label, which is one of many quality labels, might be a competitor of quality private
brands (retailer brand); hence, large retailers might have less of an inclination to stock
products with other quality labels in their stores. This could be especially true for coffee,
but it might not be the case for Thai fruits since they are basically totally different from
fruits produced in the EU.

Opportunities and Challenges of Thai Geographical Indications products
Information and promotion campaigns are necessary to support Thai GIs product
registration efforts, since European consumers and gatekeepers are unfamiliar with these
products and cannot distinguish the differences between them and other similar products in
terms of quality and taste.

The interviewees underlined the following conditions required for Thai GIs products to be
successful in the European market: (1) products must be a specialty fruit with outstanding
quality, (2) exporters should provide correct information or story behind products about
The 12
th
16 -18 June, 2011

192

landscape, production area, production and work condition and (3) food safety certification
according to the EU regulation must be provided.

Some of the interviewees suggested that Organic and Fair Trade labels used in
combination with the Thai GIs label might be useful to enhance the competitiveness of
products, although this could be costly and difficult to achieve. Furthermore, usefulness of
combining GIs labels with other quality labels relies upon clear consumer understanding of
the separate value of product attributes communicated by each label. It is also important to
synchronize the Thai GIs label with the European one (mutual recognition) to maintain
consistency in labeling.

Another challenge for Thai GIs products mentioned by gatekeepers is consumers
attachment to own local food and culinary tradition. This is especially relevant in
Mediterranean countries like Italy, where a well defined and renowned national food
culture dominates and, as a consequence, a limited demand for foreign products arises.

Factors influencing purchasing decisions of European gatekeepers
Eight major factors appear to influence European gatekeepers decisions to purchase
imported food products: quality, price, food safety, environmental aspect, social aspect,
business relationship, consumer awareness and preference, and competitors. These factors
are illustrated in the conceptual map analyzing gatekeepers views (Figure 1). However, it
seems that the elements of trust and reliability are the most prominent ones with regard to
decision to import food products. In light of this, country image might influence the
psychic distance between customers and exporters. Additionally, cultural distance may
hinder the successful market penetration of Thai products, as they may be perceived by
European consumers as simply too strange, alien and unfamiliar to be desirable.

Figure 1 Factors influencing imported food products purchasing decision of the EU gatekeepers.
Solid arrow represents positive effect and Hollow arrow represents inhibitory effect
The 12
th
16 -18 June, 2011

193

Suggested distribution channels for Thai GIs products
There were three potential channels suggested by gatekeepers: specialty shops, large
retailers, and restaurants/SPAs. They explained that specialty shops are a good distribution
channel because they are run by experts with experience with similar products. These
experts can be a better vehicle of information on products and methods of preparation and
consumption to consumers. Limitations of this channel include the fact that it will be only
a niche market, however, the consumers in this channels tend to have a larger intention to
purchase these products than other channels. Large national and international retailers are
thought to be a promising channel for Thai GIs products, since it can move high volumes
of products, has wider access to mass consumers, and employs better marketing strategies.
However, producers and operators interested in distributing products through this channel
may face difficulties in maintaining bargaining power, since premium prices and stricter
regulations are required. Thai restaurants and spas were mentioned as channels to display
and perhaps provide a first impression of Thai products to European consumers. Agri-
tourism markets were also mentioned as an innovative channel by one respondent.

Suggested marketing strategies for Thai GIs products
Gatekeepers suggested some potential marketing strategies to introduce Thai GIs products:
spread of information through public relations and communication initiatives; show-casing
the region of the products and telling the story behind them; assuring product safety and
guaranteeing quality; demonstrating products and letting consumers try them; developing
joint export platform for Thai cuisine and fruits; starting with pilot products which are
typical, high quality, and without environmental and social problems; differentiating Thai
GIs products from other products by quality, healthiness, and packaging; offering
promotion to gatekeepers; and selecting the proper distributors to channel Thai GIs
products. Learning lessons from the outcomes of marketing strategies used by both success
and failure cases in the EU market, could also be beneficial.

Discussion
According to European gatekeepers Thai GIs products might be interesting, but the GI
attribute alone cannot enhance competitiveness of Thai GIs fruits and coffee. Other
attributes of products and suppliers, such as quality, price, food safety, environmental
aspect, social aspect, business relationship, consumer awareness and preference, and
competitors, also have an impact on purchasing decision of the gatekeeper. All respondents
highlighted that information and communication are crucial issues for Thai GIs products.
This conforms with the study of Boccaletti (1999) that GIs products from abroad should
have good communication, quality indication that this product has traditional
characteristics, and other relevant characteristics to consumers requirement. The
originality of a typical local area can lead to a differentiation of the product only if
consumers recognize its value. This highlights that niche marketing through origin-labeling
may require an extensive awareness campaign so as to capture the benefits associated with
differentiation of products (Bramley and others 2009).

The study outcomes suggest that GIs seem to act more as a mediator of trust in the ability
to assure quality and safety of Thai food products, rather than as an attribute able to
enhance the product image per se. This is consistent with Knight and others (2007) who
found that country of origin is related to confidence and trust of products rather than to
country image itself.
The 12
th
16 -18 June, 2011

194

Nevertheless, GIs label might be useful as an attribute to foster the perceived quality of
Thai GIs products, both as an intrinsic and an extrinsic cue. This may transform credence
quality attributes of products into search attributes (Fandos and Flavin 2006).
Furthermore, the respondents also insisted that Thai operators and producers should
improve quality control and traceability system in order to maintain high quality of
products, which will lead to differentiation of products and better market access later.
Therefore, high and consistent quality standard, well-known products and consumers
knowledge are essential properties for GIs products to be successful.

Acknowledgements
This research was performed in the framework of the project Pro-GIs: Intellectual
Property Right extension & Geographical Indications protection for the benefit of EU-
Thailand Trade, co-funded by the programme "Thailand - EC Cooperation (TEC)
Facility" of the European Union.

References
Bardaj I, Irizoz B, Rapn M. 2009. Protected Geographical Indications and integration
Into the Agribusiness System. Agribusiness 25(2):198214.
Barjolle D, Sylvander B. 1999. Some Factors of Success for Origin Labelled Products in
Agri-Food Supply Chain in Europe: Market, International Resources and
Institutions. In: 67th EAAE Seminar: The socio-economics of origin labelled
products in agrifood supply chains: spatial, institutional and coordination aspects;
1999 Oct 28-30, Le Mans, France.
Boccaletti S. 1999. Signaling quality of food products with Designations of Origin:
Advantages and limitations. In: World Food and Agribusiness Congress; 1999 Jun
13-14, Florence, Italy.
Bramley C, Kirsten JF. 2007. Exploring the Economic Rationale for Protecting
Geographical Indicators in Agriculture. Agrekon 46:69-93.
Bramley C, Bienabe E, Kirsten J. 2009. The Economics of Geographical Indications:
Towards a Conceptual Framework for Geographical Indication Research in
Developing Countries. In WIPO. The Economics of Intellectual Property. p 109-49.
Bredahl ME, Northen JR, Boecker A, Normile MA. 2001. Consumer Demand Sparks the
Growth of Quality Assurance Schemes in the European Food Sector. In Economic
Research Service, U.S. Department of Agriculture, Agriculture and Trade Report.
Economic Research Service/USDA/WRS-01-1. Changing Structure of Global Food
Consumption and Trade. p 90-102.
Canavari M, Galanti G, Haas R, Wongprawmas R. 2010. Geographical Indications:
outlook on the European and Thai systems and overview of EU gatekeepers
perceptions towards GI fruit and coffee products proceeding from Thailand.
Dipartimento di Economia e Ingegneria agrarie, Alma Mater Studiorum-University
of Bologna, Bologna, Italy - Bioagrico-op scrl, Casalecchio di Reno (BO), Italy -
Thai-Italian Chamber of Commerce, Bangkok, Thailand. 117 p.
Chomchalow N, Somsri S, Na Songkhla P. 2008. Marketing and Export of Major Tropical
Fruits from Thailand. Assumption University Journal of Technology 11(3):133-43.
Department of Intellectual Property. 2010. The Registration of Geographical Indication.
Available from: www.ipthailand.go.th/ipthailand/index.php Accessed Aug 1, 2010.
The 12
th
16 -18 June, 2011

195

Downe-Wamboldt B. 1992. Content analysis: method, applications, and issues. Health
Care for Women International 13(30): 31321.
[ECF] European Coffee Federation. 2009. European Coffee Report 2008. Available from:
www.ecfcoffee.net/index.php. Accessed Jun 20, 2010.
European Commission. 2010. Agriculture and Rural Development. DOOR Database.
Available from: http://ec.europa.eu/agriculture/quality/door/list.html. Accessed
Aug 3, 2010.
EU Market access database. 2010. Statistic Database. Available from:
http://mkaccdb.eu.int/mkaccdb2/statistical_form.htm. Accessed Jun 18, 2010.
Fandos C, Flavin C. 2006. Intrinsic and extrinsic quality attributes, loyalty and buying
intention: an analysis for a PDO product. British Food Journal 108(8): 646-62.
[FAO] Food and Agricultural Organization. Statistics Division. 2009. The FAO Statistical
Yearbook 2009. Available from: www.fao.org. Accessed Jul 30, 2010.
Jaovisidha S. 2003. Protection of Geographical Indications Thailands Perspective. In:
the EU-ASEAN Workshop on Geographical Indication: A way into the market
Hanoi; 2003 Oct 7 8; Hanoi.
Knight JG, Gao H. 2005. Country of Origin and Confidence in Quality of Imported Foods
in China. Available from:
http://eprints.otago.ac.nz/188/1/COOandQualityChina.pdf. Accessed May 20,
2010.
Knight JG, Holdswort D, Mather D. 2007. Country-of-origin and choice of food imports:
an in-depth study of European distribution channel gatekeepers. Journal of
International Business Studies 38:107-25.
Loureiro ML, Mc Cluskey JJ. 2000. Assessing Consumer Response to Protected
Geographical Identification Labeling. Agribusiness 16(3):309-20.
Myers M. 2009. Qualitative Research in Business & Management. London: SAGE
publications Ltd. 284 p.
Novak JD, Caas AJ. 2008. The theory underlying concept maps and how to construct and
use them. Florida Institute for Human and Machine Cognition Pensacola Fl.
Available from: http://cmap.ihmc.us. Accessed Aug 2, 2010.
[GTZ] The Deutsche Gesellschaft fr Technische Zusammenarbeit. 2010. Thai Fruit and
Vegetable Case Study. Available from:
www.deza.admin.ch/ressources/resource_en_164930.pdf. Accessed Mar 14, 2010.
Trochim WM. 2006. The Research Methods Knowledge Base. 2nd ed. Available from:
www.socialresearchmethods.net/kb/. Accessed May 20, 2010.
Vandecandelaere E, Arfini F, Belletti G, Marescotti A. 2009. Linking people, places and
products, A guide for promoting quality linked to geographical origin and
sustainable Geographical Indication. FAO Rome. 194 p.

The 12
th
16 -18 June, 2011

196

OD-80
The Development of Halal Foods in Indonesia

Umar Santoso

Department of Food & Agricultural Product Technology, Faculty of Agricultural Technology, Gadjah Mada
University , Bulaksumur, Yogyakarta 55281, Indonesia. E-mail: umar_santoso@yahoo.com

Abstract

Halal food are foods that are permissible to consume by Muslim according to Islamic
dietary law, this term is opposite to haram, i.e., the foods that are forbidden or not allowed
to consume by Muslim.

Although Indonesia is the greatest Muslim population in the world with more than 210
million Muslim, however, the accomplishment of national concerns on halal foods just
appeared remarkably since 1989, especially after the Assessment Institute for Food, Drugs
and Cosmetics Indonesian Council of Ulama (AIFDC ICU, or LPPOM-MUI) was
established. Since then, halal certification activities to foods industries increases rapidly as
the awareness and demand for halal products of Muslim people increases. The demand for
halal foods is huge and this leads Indonesia to become a great and lucrative market for
halal food business. The fast development of halal food industry is supported by the highly
developed halal certification system. In this country, halal assurance system (HAS) set up
by company is a basic prerequisite for application of getting halal certificate. Indonesia
also has a big chance to take part in the halal food business competition in the global
market. In the current presentation, an outlook and prospect of halal food industry in
Indonesia, and halal certification system including halal standard and procedure of halal
certification, are discussed. The constraints and strategy of halal food industry
development in this country are also discussed.

Keywords: Halal foods, Indonesia, food industry, halal certification, Muslim

Introduction

Halal is an Arabic word. Literally, halal means allowed, permitted, or lawful. The opposite
of this world is haram, i.e., prohibited or unlawful. Halal food is foods that are allowed or
permitted to consume by Muslim according to syariah - the Islamic law. Actually, the
perfect guidance and concept for Muslim in relation to consuming foods is Halaalan-
Thayyiban, or halal and thayyib. While halal means lawful, the material kind or category
of which is decided absolutely by God (stipulated in the Holy Book), thayyib means good,
nutritious, wholesome, and safe, the criteria of which can be examined by food science
approach.

Halal concept can be seen from some perspectives. From Islamic perspective, halal means
as an obligatory for Muslim to consume only the halal foods; from the business
perspective, halal concept means a good opportunity to produce and trade halal food
products to Muslim people that are as a lucrative market. In addition, halal concept can
also be seen from scientific point of view, i.e., as a vast opportunity to perform researches
including researches to develop methods for material authentication and to produce
material substitute that are halal to replace non-halal materials.
The 12
th
16 -18 June, 2011

197

Although Indonesia is the greatest Muslim population in the world with more than 210
million muslim, however, the accomplishment of national concerns on halal foods just
appeared remarkably since 1989, especially after the Assessment Institute for Food, Drugs
and Cosmetics Indonesian Council of Ulama (AIFDC ICU, or LPPOM-MUI) was
established. Afterward, halal certification activities to food industries increase rapidly as
the awareness and demand to halal products of Muslim people increases. The demand for
halal foods is huge and this leads Indonesia to become a great and lucrative market for
halal food business. The fast development of halal food industry is supported by the highly
developed halal certification system. The purpose of this paper is to share information and
experience on the development of halal food industry and halal certification system in
Indonesia.

Global Halal Food Market
The world Muslim population today is estimated 1.6 1.8 billion, about one fifth of the
total world population. The Muslim population spread in over 145 countries with most
dense are in Middle East, South East Asia, and some countries in Africa. The average
population growth is 2.9%. The demand for halal foods is huge, and the estimated global
demand for these products is US$ 625 billion annually with the growth of 10%. The
estimated demand for halal foods will be US$2.1 trillion in 2015 (Shield, 2009). The
Muslim population in some countries is shown in Table 1. With Muslim population of
about 210 million (87% of total population), Indonesia is an enormous market for halal
foods. The halal food transaction in Indonesia is estimated US$10 billion annually, with
the growth approximately 7 to 10%.

Population increase in Muslim countries has outpaced supply. Western countries have
recognized the potential to supply Muslim countries in order to increase their export
growth or to supplement falling export volumes. Governments and food industry
organizations have made halal a major focus to grow their export of their agricultural
sectors. In todays modern world the positioning of halal in the global market place is
moving beyond its traditional religious base (Shield, 2009). Muslim consumer demand is
foods that meet both religious and dietary standards. It is better to note that halal foods is a
universal concept and now non-Muslim people are also interested due to the connotation
that halal foods are more nutritious, hygienic, and wholesome.
Middle East and South East Asia are countries with dense Muslim population, however,
their need for halal food is mainly supplied by Australia, New Zealand, and Brazil.
Australia exported 187,000 tons of halal meat valued at 673 million AUD to the countries.
Halal meat exports now represent 10% of Australian meat exports with the majority of
these exports going to the Middle East and South East Asia.

Some South East Asian Countries now is making effort to develop their halal food industry
to strengthen their share in the global market of halal foods. While Indonesia is the largest
Muslim population in the world, developing halal food industry in this country is important
due to two benefit targets, i.e., to protect domestic consumers from consuming non-halal
products, and to gain earnings from export of halal food products.

The Establishment of AIFDC ICU (LPPOM MUI)
Although Islam is a religion followed by the majority of Indonesian population and the
history of Islam in this country has dated back since centuries ago, but concerns on halal
The 12
th
16 -18 June, 2011

198

food just appeared remarkably since 1988. In that year, there was a sensitive issue based on
a publication about food products on sale that contained lard (pork fat), which Muslim
consumers refrain consuming the products resulting most companies discontinued their
production. This national issue affected not only the spiritual tranquility of Muslim
consumers but also threatened the stability of national economy. To respond this situation,
the Government of Indonesia requested the Indonesian Council of Ulama (ICU) (Majelis

Table 1 Muslim population in some countries

Country
(Asia, Africa)
1

Muslim
population
(Million)
Country
(Europe and USA)
2

Muslim population
(Million)
Indonesia 203 UK 1.6
Pakistan 174 France 6.0
India 161 Germany 3.05
Bangladesh 145 Netherland 0.9
Egypt 79 Spain 0.6
Nigeria 78 Turkey 68
Turkey 74 Albania 2.2
Iran 74 Russia 27
Algeria 34 USA 8
Morocco 32
Source:
1)
Dahlan, W, 2009. Global halal food market and future role of halal science & technology. A Paper
presented at the International Seminar on the Global Prospect of Safety and Halal Products
Analysis. Organized by Faculty of Pharmacy UGM, Yogyakarta, October 19, 2009.

2)
Shield, N., 2009. Halal food industry : a Global outlook. A Paper presented in ASEAN Food
Conference, Brunei Darussalam, October 23-26, 2009.

Ulama Indonesia, MUI) as the renowned organizing body to overcome this crisis. As a
result, supported by some Muslim scholars and academician MUI established the
Assessment Institute for Food, Drugs, and Cosmetics (AIFDC) (Lembaga Pengkajian
Pangan, Obat-obatan dan Kosmetika, LPPOM) on January 6, 1989 in Jakarta (Husein,
2009). LPPOM is an institution that assists MUI as an authoritative halal certifying body in
Indonesia. The LPPOM members are competent scientists with various disciplines
including chemistry, biochemistry, food science & technology, veterinary, agro-industry
and so on.

In 1995 MUI issued decree on the permission of Provincial MUIs in Indonesia to establish
a Provincial LPPOM. In following years, some Provincial LPPOM-MUIs were established
including West Java, East Java, Central Java, Yogyakarta Special Region, West Sumatra,
South Sulawesi, Bali, and so forth. Up to the present time there are 28 Provincial LPPOM-
MUIs have been established.

Vision and Mission of AIFDC-ICU (Anonymous, 2010)
The vision of the Assessment Institute for Food, Drugs, and Cosmetics (AIFDC-ICU)
(LPPOM MUI) is to become a trusted halal certifier in Indonesia and also worldwide to
give tranquility to Muslim ummah (society) and to become the world halal center which
extend information, solution, and halal standard admitted in national and international
level. The mission are: 1) to make and develop halal auditing system, 2) to perform halal
The 12
th
16 -18 June, 2011

199

certification for products spread and consumed by Muslim society, 3) to educate and aware
the society to consume halal product, and 4) to give complete and accurate information
about halal status of products from all point of view.

Importance of Halal Certification
The halal status of food products distributed on market should be indicated legally. It is not
so difficult to control the halalness of foods that are prepared in household, but, for the
food products from manufacturers it is not so easy to determine the halalness of the
product due to complex processing. Technology development has made it possible that
prohibited (non-halal) materials be used as ingredients, additives, or processing aid in
various processed foods. Consequently, there are many foods containing mixed halal and
non-halal material making the halal status of the products unclear or doubtful (syubhat,
mashbooh).

In order to ascertain their halal status, all processed foods whose halal status are considered
to be doubtful, have to be assayed through a halal certification process. Halal certificate is
a written fatwa of Indonesian Council of Ulama (Majelis Ulama Indonesia, MUI) that
certifies the halalness of a product in accordance to Islamic law and is issued based on the
assessment and audit by LPPOM MUI. Halal certificate is a requirement for a license
from the authorized government institution National Agency for Drugs and Foods
Control (BPOM - RI) to attach a halal label in product package. To get halal certificate, a
company must set up and implement Halal Assurance System (HAS), that ensures the
continuity of halal production process during holding the certificate.

Halal Assurance System (HAS)
Halal Assurance System (HAS) is a system designed, applied, and maintained by a
company holding halal certificate to assure the sustainability of halal production process,
hence the halalness of its products continuously and consistently. The objective of HAS is
to keep the sustainability of halal production process and management in order to assure its
halalness according to the rule stated by LPPOM-MUI. Halal Assurance System is a part
of company management policy and must be documented. As a management system, the
main components of HAS are halal policy (commitment), planning, doing
(implementation), monitoring and evaluation, and, corrective action as a cycle (Figure 1).

Halal Standard
The halal standard of LPPOM MUI is mainly based on Quran (Holy Book of Islam) and Al
Hadist (the tradition of the Prophet Muhammad
mpbh
). Halal food is food or food stuff that
is not contaminated with haram or non-halal material or najis (dirt). The haram materials
are carrion (dead animals), blood, pork, animals slaughtered in name other than Allah, and
alcoholic beverages, and also some materials considered as haram as mentioned in Al
Hadist.
The 12
th
16 -18 June, 2011

200

Figure 1 Cycle of Halal Assurance System

Halal Certification Procedure
A company which will apply for halal certificate must fill in an application form provided
by LPPOM-MUI for registration. In the same time, the company must set up Halal
Assurance System (HAS) and officially assign persons to be a Halal Team of the company
or Internal Halal Auditors whose responsible in ensuring the halal production by
implementing HAS.

The completed application form attached with all supporting documents including HAS
document be submitted to LPPOM MUI Secretariat to be checked for completeness.
Then, LPPOM MUI will assign halal auditors team to inspect or audit to the companys
plants

at running time of processing. The result of auditing are evaluated and discussed in
LPPOM meeting. Auditing result which is not complete yet will be informed to the
company, and which is already complete and conformed with the requirements will be
submitted to the Fatwa Committee of MUI meeting to get decision of the halal status. The
Fatwa Committee of MUI can reject the report of auditing results if it is not met the whole
requirements stipulated. Halal Certificate is then issued by MUI after the halal status has
been determined by Fatwa Commission. The validity of the Certificate is 2 years upon the
date of issuance, and three months before expiry date the company should make renewal
halal certification according the policy of LPPOM - MUI. Figure 2 shows the diagram of
halal certification process. The scope of audit process covers company management in
ensuring halal production (implementation of HAS), the documentation, materials, process
and facilities in the plant and material or product sampling when necessary.

Achievements of LPPOM MUI
Many programs and activities of LPPOM-MUI have been made. As a halal certifying
body, the main mission is to issue halal certificate to the companies which desire to apply.
The first halal certificate was issued in 1994, and since then the number of company which
apply to halal certification increases sharply. Up the present time thousands of product
items distributed in Indonesia have been halal certified. The data of number of company,
Corrective
action
Planning
Doing
Monitoring &
Evaluation
Halal Policy
(Commitment
)
The 12
th
16 -18 June, 2011

201

number of halal certificate issued, and product items that have been halal certified by
LPPOM-MUI from 2005-2010 is shown in Table 2. The list of halal products certified by
LPPOM-MUI National Level and Provincial LPPOM-MUIs can be seen through
www.halalmui.org. A bimonthly of Jurnal Halal is also published by LPPOM-MUI as a
media for information and communication on halal which also contains the list of halal
certified products.

International Cooperation
On the globalization era, implementation of halal certificate is not only applied in domestic
level. Many material or ingredients used by local industry are imported from abroad, and
so cooperation with other parties or council is needed to ease the process of certification.
Creating international standard on halal for foods, drugs and cosmetics is an urgent
necessity. However, many halal certification bodies do not have human resources and
auditing standard as expected. It is logic that halal-haram matter becomes global issues
that must be implemented according to Islamic law. Until now, LPPOM-MUI had given
training about halal certification to some overseas halal certification bodies (Anonymous,
2010).

In 1999 World Halal Council was established in Jakarta as a communication center among
international halal certification bodies. In this forum, halal certification problems, halal
auditing standard, and similarity point of view about many things related with halal

Figure 2 Diagram of Halal Certification Procedure

Planning
Registration HAS
Audit
Audit Evaluat ion
(LPPOM-MUI meet ing)
Fatwa Commission of
MUI
Halal Certificat e
The 12
th
16 -18 June, 2011

202

Table 2 Number of company, halal certificate issued, and product items that have been halal
certified by LPPOM-MUI, Year 2005-2010

Year Number of Company Number of Halal
Certificate
Number of product items
2005 414 969 2.408
2006 443 1.123 12.533
2007 488 1.013 8.636
2008 548 921 10.242
2009 353 470 10.550
2010 619 641 21.834
Source : Republika, January 21, 2011.

Accelerating the Development of Halal Food Industry
To accelerate of the development of halal food industry five parties i.e., industry (food
producers), government, certifier body, and academician should actively play their role and
make a good cooperation. Food scientists from university can support through many
programs related with halal foods. For this, some universities have established Halal
Science Center, e.g., Chulalongkorn University - Thailand (Dahlan, 2009), University
Putra Malaysia (UPM) - Malaysia (Che Man, 2008), and IPB-Bogor, Indonesia (Anas
Fauzi, 2009).

Table 3 List of list of some halal certifying bodies

Asia Australia and
New Zealand
Europe USA Latin
America and
others
Jabatan Kemajuan
Islam Malaysia
(JAKIM),
Malaysia
Australia Halal
Food Service
(AHFS)
Association for the
Inspection,
Certification of Food
& Supplies
(GIMDES), Turkey
IFANCA (Islamic
Food & Nutrition
Council of
America),
Chicago
Central Islamica
Brasileira de
Alimentos Halal
(CIBAL), Brazil
Majlis Ugama
Islam Singapore
(MUIS),
Singapore
The Adelaide
Mosque Islamic
Society of South
Australia Inc.
Halal Control e K American Halal
Foundation (AHF)
Islamic
Dissemination
Center for Latin
America
(CDIAL), Brazil
Islamic Dawah
Council of the
Philippine (IDCP)
Halal Sidiq
Services, Australia
Halal Feed, Food
Inspection,Authority,
(HFFIA)
Halal Transaction
of Omaha
Halal Monitoring
Authority
(HMA), Canada
Bahagian
Kawalan
Makanan Halal,
Jabatan Hak
Ehwal Syariah,
Kementrian
Ugama Hal Ehwal
Ugama, Brunei
Darussalam
Supreme Islamic
Council of Halal
Meat in Australia
Inc. (SICHMA)
Halal Food
Authority England
Islamic
Information
Center of America
Islamic Society
North America
(ISNA)

The 12
th
16 -18 June, 2011

203

Table 3 (Continued)

Asia Australia and
New Zealand
Europe USA Latin
America and
others
The Central
Islamic
Committee of
Thailand
(CICOT)
The Islamic
Coordinating
Council of Victoria
(ICVV)
Islamic Food
Council of Europe,
(IFCE)
Islamic Service of
\America (ISA)

Office on Muslim
Affars , The
Philippines
Islamic Council of
Western Australia
Taipe Grand
Mosque, Taiwan
Islamic Association
of Kattaning,
Australia
Perth Mosque
Incorporated
The Assessment
Institute for Food,
Drugs, &
Cosmetics
(AIFDC)
(LPPOM MUI),
Indonesia.
Al Kautsar Halal
Food Inspection
(AL KAHFI)
Total Quality Halal
Correct (TQHQ)

Source: Anonymous, 2010. Indonesia Halal Directory 2010. The Assessment Institute for Food, Drugs, &
Cosmetics Indonesian Council of Ulame (AIFDC-ICU)) (LPPOM MUI), Jakarta, Indonesia.

References
Anas Fauzi, M., 2009. Quality assurance of halal products in Indonesia. A Paper presented
at the International Seminar on the Global Prospect of Safety and Halal Products
Analysis. Organized by Faculty of Pharmacy UGM, Yogyakarta, October 19, 2009.
Anonymous, 2010. Indonesia Halal Directory 2010. The Assessment Institute for Food,
Drugs, & Cosmetics Indonesian Council of Ulame (AIFDC-ICU)) (LPPOM
MUI), Jakarta, Indonesia.
Che Man, B. Yaacob, 2008. Current research on halal food authentication. A paper
presented in 2
nd
International Halal Research Symposium, IMT-GT, IPB
Convention Center, Bogor.
Dahlan, W, 2009. Global halal food market and future role of halal science & technology.
A Paper presented at the International Seminar on the Global Prospect of Safety
and Halal Products Analysis. Organized by Faculty of Pharmacy UGM,
Yogyakarta, October 19, 2009.
Hakim, L. and Syamsu, K., 2008. Launching of HAS as an international quality system.
Proceeding of Workshop on Animal Derivatives. LPPOM MUI, Bogor, 3 July
2008.
Hosen, M. N, 2009. Guidelines for Halal Assurance System. International Training for
Halal Auditors. LLPOM MUI, Jakarta.
Republika, January 21, 2011. The abundance of halal food products (Indonesian) (Report
of Indah Wulandari).
Shield, N., 2009. Halal food industry : a Global outlook. A Paper presented in 11
th
ASEAN
Food Conference, Brunei Darussalam, October 20-23, 2009.
The 12
th
16 -18 June, 2011

204

OD-149
Application of Modified Drop Plate Technique (MDPT) and Logistic
Model to Optimize Non-Selective Substrate for Salmonella Typhi
Resuscitation
Juthamas Khueankhancharoen
1
and Aluck Thipayarat
1,*
1
Department of Food Engineering, Faculty of Engineering, King Mongkuts University of
Technology Thonburi, Bangkok 10140, Thailand
*
Corresponding author: athipaya@yahoo.com

Abstract
The use of modified drop plate technique (MDPT) for aerobic cell enumeration together
with sigmodial-type model simulation was proposed as an effective method to select
suitable substrates (i.e., carbon and nitrogen sources) to maximize Salmonella typhi
growth. This technique was able to accelerate the optimization of medium recipe that
allows a microorganism to grow at its best using a suitable enrichment medium. The
modified drop plate utilized only 0.3 ml of agar substrate and 5 l of inoculums. The
incubation conditions were identical to the spread plate technique (37
o
C); however, the
detection time was significantly reduced from 18-24 h to only 8-10 h. The faster time to
detect colony was assisted by the use of digital microscope and image analysis software.
The comparison of the total plate count obtained from the modified drop plate (MDPT) and
the spread plate technique (SPT) showed the equivalent number of final colony counts.
Fast enumeration of forming colonies during aerobic plate cultivation synergized by the
use of mathematical model (i.e., logistic model) enabled fast analytical time taken to
optimize non-selective enrichment substrate for Salmonella typhi growth. Several key
kinetics parameters, including maximum specific growth rate (
max
), were extracted and
compared among different enrichment broths. The results indicated that TSB and BPW
supported better growth of Salmonella than LB and NB. It was indicated by higher
maximum specific growth rate (
max
) and larger achievable final cell concentration.
Perhaps, the buffer system was able to stabilize the acidity allowing the cells to grow at its
most optimal condition.
Keywords: Salmonella spp., specific growth rate, colony image analysis, logistic model,
non-selective enrichment broth

Introduction
Salmonella is widely considered as one of the most serious pathogenic contaminants
present in products derived from poultry, eggs and meat. Occasionally this lethal pathogen
can cause widespread disease outbreaks that have been documented in many countries all
over the world (Stacy and others 2003; Luciane and others 2007). Centers for Disease
Control and Prevention of US (CDC) recently reported Salmonella outbreaks responsible
for approximately 43,000 cases of food-transmitted salmonellosis (lost 2.7 billion$
annually) (CBCNEWS 2009). Perhaps, high death tolls contribute to the lack of
inexpensive and fast Salmonella test kits (Pallav and Williams 1994). The popular standard
analytical method (e.g, ISO 6579, Bacteriological Analytical Manual, etc.) still requires
long analytical time usually a minimum of 4 days and several additional days for the
The 12
th
16 -18 June, 2011

205

confirmation of presumptive results (USFDA 2007; Purnendu 1997). The general protocol
entails numerous enrichment and isolation steps; none of them is considered absolute. In
the actual industrial food safety application, the majority of the food samples analysed
return negative results. There are demands for disruptive measures in reducing analytical
time and providing cost effective protocols. Also the development should extend to
confirmation of potentially positive results (John and others, 1995).

The standard detection protocol of Salmonella requires three essential steps: pre-
enrichment, selective enrichment and plating. The key factor to increase the change to
detect Salmonella especially injured cell resulting from processing stresses such as heat,
cold, or dehydration, is the development of effective non-selective enrichment. A variety
of enrichment substrates have been recommended for resuscitation of the injured cells
(Lesley and others 1952; Gary and others 1975; Edel and Kampelmacher 1973). In
particular, lactose broth and buffered peptone water were among a few recommendations
suggested for the pre-enrichment of sublethally injured salmonellae from dried and frozen
food products as well as frozen meat samples (North 1961; Edel and Kampelmacher 1973).
In this study, new technique (MDPT) was proposed for fast Salmonella typhi enumeration
and detection. The new protocol can also be used to optimize growth substrates and study
the growth kinetics of Salmonella cultivation.

Microorganism and preparation of inoculum
Salmonella typhi was obtained from the culture collection of Department of Microbiology,
King Mongkuts University of Technology Thonburi, Thailand. Cultures were maintained
by monthly transfer on tryptic soy agar (TSA, Difco) slants and stored at 4+1C. After two
successive transfers of the stock culture on TSA at 37C for 24 h, the activated culture was
inoculated into 100 ml of tryptic soy broth (TSB, Difco) and incubated at 37C for 16 h at a
shaking speed of 200 rpm. The cell-free supernatant was recovered by centrifugation
(8000xg, 4 C, 10 min), washed twice with 0.01 M phosphate buffered saline (PBS, pH
7.4). The washed cells, suspended and diluted in PBS to ca log 3 CFU/ml, were used as the
inoculum.

Bacterial cell enumeration and validation in micro-scale cultivation
Salmonella typhi activated pure culture were recovered in 100 ml TSB containing in
baffled flask at 37C for 16 h. Cells were harvested, washed, suspended and diluted in
normal saline to ca log 3 CFU/ml as inoculums. These initial washed cell suspensions were
enumerated by both the conventional spread plate method and proposed modified drop
plate technique on tryptic soy agar (TSA). The cultivation volumes were reduced from
normal volume of 100 ml (spread plate) to only 5 l and the agar utilization per sample
was around 0.3 ml on agar in the cover of 96-well plate. All the samples were kept in a hot
air incubator at 37+2C. Colony expansions and cell amplification on the cover of 96-well
microplate were detected and captured using a reflected light microscope equipped with a
CCD camera (Figure 1) during incubation along 24 h compared with the traditional spread
plate method. Also, the washed cells were prepared at different initial cell concentrations
(0-8 log CFU/ml) in order to enumerate by micro-scale technique in comparison with the
spread plate method.
The 12
th
16 -18 June, 2011

206

Figure 1 Image acquisition prototype of modified drop plate technique (MDPT)

Comparison of standard enrichment broths
Various standard enrichment media for Salmonellae (nutrient broth; NB, lactose broth; LB,
buffered peptone water; BPW and tryptic soy broth; TSB), as specified by BAM, AOAC,
FSIS, ISO etc., were evaluated to determine the effect of different media recipes on growth
profiles of S. typhi at 37C. Growth curves were determined after a specific period of
cultivation along 24 h using 2 enumeration methods (spread plate and modified drop plate
technique).

Determining the growth kinetics using mathematical model
Mathematical model as sigmoid 4 Parameter (equation 1) has been successfully employed
to predict the growth kinetics in batch cultivation including maximum specific growth rate
(
max
), inflection time (t
i
), etc. (Jeffrey 2003; Vijay and others 2009).

(1)

where: x(t) is the (CFU/ml) of cell concentration at time, t; x
o
is the initial concentration
(CFU/g); b is maximum relative growth rate at t=N (h
1
); N is time at which the absolute
growth rate is maximum (h). The parameter b can be used to define the specific growth
rate of the growth;
max
(McMeekin and others 1987.).

Bacterial cell enumeration and validation in micro-scale cultivation
Micro-scale cultivation was developed to substitute the conventional cell count method
(i.e., spread plate technique (SPT)). This technique was modified from the drop plate
technique with the goal to fasten the colony count estimation and lower the associated
analytical expense. This modified drop plate technique (MDPT) was compared against the
common spread plate colony count. To verify the technique, S. typhi were prepared at
approximately 9 log CFU/ml and applied to both MDPT and spread plate experiments for
cell count. The SPT utilized the conventional Petri dish format whereas the MDPT made
A stereomicroscope
A bottom
illumination system
A top
illumination
Sample
holder
A digital camera
The 12
th
16 -18 June, 2011

207

use of the 96-well plates. The results showed that when applying the modified drop plate
technique (MDPT), the time to detect Salmonella colonies was shorten to within 8 h
(Figure 1). At this particular hour, no colony detection was detected in the spread plate
experiment using human visualization. It took up to 16 h for the spread plate colonies to be
visually detected and 18-24 h to achieve steady cell count. Approximately, the detection
time of the SPT was 8-14 hours lagged behind that of the MDPT.

Figure 2 Number of Salmonella typhi in log CFU/ml enumerated by 2 methods (SPT and MDPT)
during incubation time for 24 h (Values in the bar with the same color but with different
lowercase letters differ significantly while the bar within at any incubation time with
different uppercase letters differ in significantly according to Duncans multiple range test
(p< 0.05)).

However, the colony count at the 8
th
hour from the MDPT was significantly less than the
final approachable number (Figure 2). The MDPT required at least 10 hours for the colony
count to be stabilized statistically (Table 1). Noted that this particular experiment was
performed using active cells, the actual time for steady cell count must be investigated case
by case. The final cell counts from the MDPT and the spread plate technique using the
same stock culture was not statistically different at 95% confident level.

The comparison of the SPT and MDPT detection on different concentrations of Salmonella
was performed and the results were used to construct a standard curve as shown in Figure
2. The stock cultures were prepared from 1 to 8 log CFU/ml. The MDPT returned
practically identical cell counts as performed by the SPT. The slope of the standard curve
was 0.9985 and the relative coefficient (r
2
) was 0.9793. The variation of cell count values
was very narrow except at the very low concentration. At the 2 log CFU/ml, the cell count
variation was fairly larger than the other concentrations since the MDPT was performed in
the proximity of its lower detection limit. At the 1 log CFU/ml, the MDPT exceeded its
detection limit; hence, the method showed zero cell count. The inoculums size of the
MDPT was only 5 l as opposed to 100 l in the SPT. There was a two orders of
magnitude difference inherently associated with the sample size. This MDPT was
previously applied to enumerate various strains of bacteria, for example Listeria spp. and
Salmonell anatum (Suchanan 2009).
b, B
c, B c, B c, B c, A b, A
a, A
a, A a, A a, A
b, A c, A b, A c, A
a, A a, A
The 12
th
16 -18 June, 2011

208

Table 1 Comparison between the proposed modified drop plate technique (MDPT) under constructed
light microscope and conventional spread plate method (SPT) detected by human visualization

MDPT SPT
Enumeration
method

Time to detect 10-12 h 18-24 h
Number of S. typhi*
(log CFU/ml)
9.85+0.25
A
9.77+0.18
A

*
Values in the same row with different letters differ significantly according to Duncans multiple range test (p
< 0.05).

Figure 3 Comparison of a conventional spread plate (SPT and modified drop plate technique
(MDPT) method in Salmonella typhi enumeration on TSA after incubation at 37 C
Validation of MDPT and SPT on pre-enrichment broth for Salmonella resuscitation
In Figure 4, the MDPT and SPT was applied and compare side by side to capture the
growth profile of Salmonella enriched in different standard media, including nutrient broth
(NB), lactose broth (LB), buffered peptone water (BPW) and tryptic soy broth (TSB). The
results confirmed the early finding that both MDPT and SPT generated growth profiles
with very high correlation. The cell count numbers at any given time were not significantly
different (p0.05) in all media treatments.
The 12
th
16 -18 June, 2011

209

Incubation time (h)
0 5 10 15 20 25 30 35
N
o
.

o
f

S
.

t
y
p
h
i

(
l
o
g

C
F
U
/
m
l
)
0
2
4
6
8
10
NB (SPT)
NB (MDPT)

Incubation ime (h)
0 5 10 15 20 25 30
N
o
.

o
f

S
.

t
y
p
h
i

(
l
o
g

C
F
U
/
m
l
)
0
2
4
6
8
10
LB (SPT)
LB (MDPT)

Incubation ime (h)
0 5 10 15 20 25 30
N
o
.

o
f

S
.

t
y
p
h
i

(
l
o
g

C
F
U
/
m
l
)
0
2
4
6
8
10
BPW (SPT)
BPW (MDPT)

Incubation time (h)
0 5 10 15 20 25 30
N
o
.

o
f

S
.

t
y
p
h
i

(
l
o
g

C
F
U
/
m
l
)
0
2
4
6
8
10
TSB (SPT)
TSB (MDPT)

Figure 4 The growth of Salmonella typhi in the presence of various standard enrichment media (A:
NB; B: LB; C: BPW and D: TSB) at 37C along incubation time of 24 h enumerated by 2
methods (Spread plate vs MDPT)

Generally, both protocols gave every small standard deviation; however, the exponential
phase intrinsically possessed higher variation of cell count due to the asynchronous nature
of the stock culture. This large variation subsided as the cell growth reaching its stationary
phase. The only non-agreeable data were at the early stage of the growth period. MDPT
returned no colony estimation whereas the SPT detected a few colonies in the order of 2
log magnitude. Again this happened because the detection limit of the MDPT was around 2
log CFU/ml as explained earlier. Nevertheless, the detection range of the rest of the growth
profile was distant away from this critical limit and this methodology was able to apply
without major cell count errors.

Study of non-selective growth kinetics for Salmonella
Applying the logistic model to simulate the growth profiles, this sigmoidal-type model fit
well to the unique s-shape characteristics excluding the death phase (Figure 5). The growth
kinetic of Salmonella on different traditional enrichment broths (i.e. NB, LB, BPW and
TSB) were estimated, including the maximum cell density, the maximum specific growth
rate, and the inflection time (Figure 4). The 6 log-scale increment from initial cell loadings
(2 log CFU/ml) to 5 log CFU/ml of the final cells was achieved within 16 h. The rapid
growth during pre-enrichment was desirable; it promoted the target cell survival over the
strong toxicity during the selective enrichment Chen and others (1993). Preferably, the pre-
(A) (B)
(C) (D)
The 12
th
16 -18 June, 2011

210

enrichment should result in the proliferation of cell population at least 5 log CFU/ml prior
to the selective enrichment step.
Incubation time (h)
0 5 10 15 20 25 30
G
r
o
w
t
h

(
l
o
g

C
F
U
/
m
l
)
0
2
4
6
8
10
Nutrient broth (NB)
Lactose broth (LB)
Buffered peptone water (BPW)
Tryptic soy broth (TSB)

Figure 5 The growth of Salmonella typhi in the presence of various standard enrichment media (NB,
LB, BPW and TSB) at 37C along incubation time of 24 h

However, most standard protocols (e.g. BAM, ISO, etc.) prolonged the incubation time to
16-24 h or overnight to ensure ample cell density (DAoust and others 1992). Suitable cell
density for subsequent selective agar plating was the subtle factor to determine appropriate
enrichment period. The concerns over the most effective choice of non-selective broth for
Salmonella resuscitation was less critical Doyle (1989). Figure 4 showed all standard
media returned similar growth profile. Many studies compared several nonselective
media also showed a comparable prolifeability (DAoust and Maishment 1979). The more
relevant dilemma was the concern of microbial population presented in real food products.
After sublethal processing, it most likely not only includes normal cells but also contains
injured microorganisms (stressed, sublethally or reversibly injured) (Ray 1979; Wu 2008).
The resuscitation and detectability of these injured cells was more pertinent to optimize the
requirement of the non-selective enrichment.

Figure 6 summarized the key kinetics parameter extracted from the logistic function. The
similar growth profiles except for the LB treatment produced no statistical differences in
max
, X
max
and to. The effect of LB was the alteration of acidity because S. typhi belongs to
one of the lactose-positive Salmonella strains (Kunz and Ewing 1965). When there was
lactose presented in enrichment broth, these microorganisms can utilize lactose as energy
source and then produce organic acids lowering broth pH. Hoffman and others (1997)
demonstrated the potential resuscitation of Salmonella in the media containing lactose was
significantly less when the final pH reached 4.4. Without lactose, the final pH decreased to
only 5.7 and Salmonella grew to higher cell density.

The 12
th
16 -18 June, 2011

211

Figure 6 The kinetics parameters (A: X
max
; B:
max
and C: t
0
) of Salmonella typhi in the presence of
various standard enrichment media (NB, LB, BPW and TSB) at 37C along incubation time
of 24 h

Conclusion
The study of different pre-enrichment media for Salmonella pointed to the replaceablity of
any standard non-selective substrates. Among the selected generic media used, LB should
be the less preferable medium due to the promotion of acid production from lactose. This
intensive study was facilitated by the application of the MDPT developed. The micro-scale
cultivation allowed the efficient use of media and supply. Also the detection time was
significantly reduced using digital microscopic detection of growing colony. The MDPT
was successfully validated to the SPT showing high repeatability and good accuracy.
References
Allen LA, Pasley SM, and Pierce MSF. 1952. Conditions affecting the growth of
Bacterium coli on bile salts media. Enumeration of this organism in polluted water.
Journal of General Microbiology 7: 257-67.
Bissonnette GK, Jezeski JJ, McFeters GA and Stuart DG. 1975. Influence of
environmental stress on enumeration of indicator bacteria from natural waters.
Applied Microbiology 29: 186-94.
The 12
th
16 -18 June, 2011

212

Borowsky LM, Schmidt V and Cardoso M. 2007. Estimation of Most Probable Number of
Salmonella in Mince Pork Samples. Brazilian Journal of Microbiology 38:544-6.
Brewster JD. 2003. A simple micro-growth assay for enumerating bacteria. Journal of
Microbiological Methods 53:77 86.
CBCNEWS. 2009. Recall in U.S. salmonella outbreak expands to more than 125 products.
Available from: http://www.cbc.ca/news/. Accessed Nov 11.
Chen H, Fraser ADE and Yamazaki H. 1993. Evaluation of toxicity of salmonella selective
media for shortening the enrichment period. International Journal of Food
Microbiology 18: 151-9.
Doyle MP (Ed.). 1989. Salmonella. In: Foodborne Bacterial Pathogens. Marcel Decker Inc.
New York. p 384-6.
DAoust JY and Maishment C. 1979. Pre-enrichment conditions for effective recovery of
Salmonella in foods. Journal of Food Protection 44: 369-74.
DAoust JY, Sewel AM and Warburton DW. 1992. A comparison of standard cultural
methods for the detection of foodborne Salmonella. Journal of Food Microbiology
16: 41-50.
Edel W and Kampelmacher EH. 1973. Comparative studies on the isolation of "sub-
lethally injured" salmonellae in nine European laboratories. Bullatin W.H.O. 48:
167-74.
Favrin SJ, Jassim, SA and Griffiths MW. 2003. Application of a novel immunomagnetic
separation-bacteriophage assay for the detection of Salmonella enteritidis and
Escherichia coli 0157:H7 in food. International Journal of Food microbiology
85:63-71.
Hoffman CA, Wagner DE and Tatini S. 1997. Comparative study of recovery of
Salmonella javiana from mozzarella cheese by two official analytical procedures.
Journal of food protection 60: 1493-6.
Juneja VK, Melendres MV, Huang L, Subbiah J, Thippareddi H. 2009. Mathematical
modeling of growth of Salmonella in raw ground beef under isothermal conditions
from 10 to 45
o
C. International Journal of Food Microbiology 131:10611.
Kunz LJ and Ewing WH. 1965. Laboratory infection with a lactose-fermenting strain of
Salmonella typhi. Journal of Bacteriology 89: 1629.
McMeekin TA, Chandler RE, Doe PE, Garland CD, Olley J, Putro S and Ratkowsky DA.
1987. Model for combined effect of temperature and salt concentration/ water
activity on the growth rate of Staphylococcus xylosus. Journal of Applied
Naraipitak S. 2009. Optimization and growth kinetics of Listeria spp. in Micro-cultivation
Environment for Rapid Detection (A Special Research Project). Bangkok: King
Mongkuts University of Technology Thonburi. 74 p. Available from
KingMongkuts University of Technology Thonburi.
North WR. 1961. Lactose pre-enrichment method for isolation of Salmonella from dried
egg albumen. Its use in a survey of commercially produced albumen. Applied
Oggel JJ, Nundy DC, Ann Zebchuk P and Shaw SJ. 1995. Reliability of the modified semi-
solid RappaportVassiliadis agar and the modified 12 test system for detection of
Salmonella in poultry feeds. Journal of Food Protection 58: 98101.
Patel PD and Williams DW. 1994. Evaluation of commercial kits and instruments for the
detection of foodborne bacterial pathogens and toxins. In: Patel PD, editor. Rapid
The 12
th
16 -18 June, 2011

213

analysis techniques in food microbiology. Chapman & Hall. Glasgow. Scotland. p
60 103.
Ray B. 1979. Methods to detect stressed microorganisms. Journal of Food Protection 42:
346-55.
[USFDA] United State Food and Drug Administration. 2007, Bacteriological Analytical
Manual (BAM): Chapter 5 Salmonella. Available from: http://www.cbc.ca/news/.
Accessed September 22.
Vasavada PC. 1997. Advances in pathogen detection. Food Testing and Analysis 3(2): 18
20, 223, 47.
Wu VCH. 2008. A review of microbial injury and recovery methods in food. Food

The 12
th
16 -18 June, 2011

214

OD-162

Enhancement of Growth Kinetic of E. coli using Sodium Pyruvate in
Tryptic Soy Broth (TSB)

Pattarin Supanivatin and Aluck Thipayarat*
Department of Food Engineering, Faculty of Engineering, King Mongkuts University of Technology
Thonburi, Bangkok 10140, Thailand
*
Corresponding author: (athipaya@yahoo.com)

Abstract
This research proposed an effective protocol of E. coli growth in non-selective enrichment
using sodium pyruvate supplement. Multiplication of cell mass efficiently during primary
enrichment is the key to successful foodborne pathogen detection in food manufacturing.
In this research, the growth kinetics of E. coli under various Tryptic Soy Broth (TSB) and
sodium pyruvate concentrations were investigated. Cells were grown on micro-scale
reactors (approximately 200 L). The TSB concentrations were varied at 0.25, 0.5, 1.0, 2.0
and 3.0X strength of the original TSB recipe and sodium pyruvate were supplemented to
0.25X TSB from 0 to 15 mg/l. Culture samples were collected to delineate batch growth
curves at 352C in a hot-air incubator. Logistic model was utilized to simulate Sigmoidal
nature of batch cultivation. The maximum specific growth rates eluted from the logistic
model were used to compare the most suitable TSB and sodium pyruvate concentrations to
maximize E. coli growth during liquid enrichment. The result showed that logistic model
was able to describe the growth profile of E. coli very well. Sodium pyruvate as a quickly-
absorbed carbon source supplement had minimal effect in promoting E. coli multiplication.
The maximum specific growth rate indicated that only low concentration of TSB (as low as
0.25X) was required to sustain growth of E. coli in liquid enrichment. Basically, with the
modified TSB E. coli can grow 2 to 8 log CFU/ml within 10 h. Higher cell density in the
final stage of non-selective enrichment can heighten the detect ability of E. coli in the
subsequent steps in the overall protocol of E. coli detection analysis.
Keywords: E. coli, logistic model, non-selective enrichment, sodium pyruvate,
specific growth rate

Introduction
Verotoxigenic Escherichia coli (VTEC), widely-recognized as important enteric
pathogens, can inflict very serious illnesses in humans (Baker and others 1999). In
particular, the O157 serogroup has been reported to cause many serious food-borne
outbreaks. Transmission of this pathogen occurs primarily through the consumption of
contaminated foods (e.g., undercooked beef, raw milk, raw vegetables, fermented meats,
etc.). The development of fast detection of this lethal strain is of paramount benefit to food
manufacturers and public health microbiologists alike. Since E. coli O157 is highly
resilient to harsh manufacturing conditions, especially in fairly high acidic environment
(Blackburn and McCarthy 2000).
Reliably and effective methods have been widely sought after to detect and/or enumerate
both stressed and healthy Escherichia coli O157 from a noisy background of outnumbering
competing bacteria. The ability of methods to isolate injured cells is even more important
The 12
th
16 -18 June, 2011

215

when considering E. coli 0157 due to the potential low infections dose of this
microorganism. The direct use of selective media imposes a profound consequence to the
overall detection methodology. The growth of injured E. coli 0157 has been shown to be
inhibited by the use of certain selective agents; hence, false negative result ensues
(Weagant and Bound 2001; Kudo and others 2000). This may explain why the methods,
which are usually developed initially for clinical samples, sometimes perform poorly on
food samples. This paper proposed to study and understand the background knowledge of
non-selective enrichment step to resuscitate and increase the likelihood to detect injured
cells from processing such as heating, freezing, low pH and chemical preservation
(Blackburn and McCarthy 2000; Everis 2001).

In this experiment, E. coli culture was prepared to reach 10
9
CFU/ml in shake flasks. The
initial cell concentration of E. coli was 10
2
-10
3
CFU/ml by using serial dilution technique.
One hundred micro-liters of E. coli cultures were inoculated into eppendorf containing
0.86% normal saline 900 l.

Growth of E. coli at various TSB concentrations
Standard TSB were prepared in test tube by having various concentrations at 3X, 2X, 1X,
0.5X, 0.25X of TSB for 5 ml per each concentration. Desired initial cell of E. coli
concentration at 10
2
-10
3
CFU/ml was injected into test tube that containing TSB then
mixes them by using vortex mixer. And the final cultivation volume in 96-microwell plate
was 200 l. The cultivation temperature was controlled at 352
o
C. The total colony
forming units were observed in the entire well surface.

Effect of sodium pyruvate using 0.25X TSB
TSB 0.25X were empirically selected from prior experiment to investigate the effect of
pyruvate supplement. The concentration of sodium pyruvate was varied at 0.25X, 0.5X,
1X, 2X, and 3X. The micro-cultivation procedure and conditions followed the previous
protocol.

Determining the maximum specific growth rate using mathematical model
Several mathematical models have been successfully employed to predict the growth
kinetics in batch cultivation (Brewster, 2003; Juneja and Marks, 2006). Only six generic
models were excerpted from numerous literatures to represent E. coli growth (Meyer and
others 1999; Mitchell and others 2004; Saeaung and others 2010; Saeaung and
Boonyaprapasorn 2010).

The 12
th
16 -18 June, 2011

216

b
x x
e
a
y y
) ( 0
0
1

+
+ =
b
x
x
a
y y
|
|
\
|
+
+ =
0
0
1
(
(
=
+
b
b x x
c c
e a y
) 2 ln (
/ 1
0
1
b
e
x x
ae y y
)
0
(
0
+ =
b b
b
x c
ax
y y
+
+ =
0
( )
c
bx
e a y y

+ = 1
0
Table1 Typical mathematical models representing batch growth curve of bacteria

Sigmoidal Model Equation
Sigmoid

Logistic

Welbull

Gompertz

Hill

Chapman

where: y(x): Cell density (log CFU/ml) at any given time (x in hour)
y
o
: Initial inoculums density (log CFU/ml)
x
o
: Chracteristic function associated with incubation time in hour.
a, b, and c; Curve-fitting coefficients

Statistical analysis
All data were analyzed at p<0.05 for significant values by ANOVA and Duncans multiple
range tests (Statistical Analysis System).

Results
Microbial growth model
A broad spectrum of mathematical models has been developed to describe microbial
growth in the last two decades (Mc Meekin and others 1993). One of the original ideas was
to mathematically describe the sigmoid nature of isothermal growth curves in a closed
habitat. Since then, its application has been expanded to cover a wide range of rate models,
empirical models and population dynamics based models. Prediction of microbial
population of contaminants can be performed using appropriate models. Since most foods
become inedible or unsafe to eat long before they have reached the stationary phase,
The 12
th
16 -18 June, 2011

217

models that only accurately simulate the sigmoid part of the growth curve generally suffice
for food systems. Most likely, the mortality stage of these microbial contaminants was
irrelevant. Figure 1 revealed contrasted different mathematical model (Sigmoid, Logistic,
Weibull, Gompertz, Hill and Chapman) and showed their performances to capture the
growth profiles of E. coli on regular strength TSB using micro-cultivation volume (200 l)
at 352C.

Figure 1 Comparison of different growth models to simulate the growth kinetic of E. coli on TSB
All models were able to simulate the growth curve identifying the adaptation periods (e.g.,
a short or long lag phase), adjust to a sudden growth in the exponential phase, and
capture the asymptotic nature in a stationary phase. Several authors have reported that a
few selections of mathematical models were able to describe the sigmoid isothermal
microbial curves with a very similar degree of fit judged by the same statistical criteria as
in this paper (Simpon and others 2006; Marks 2008).

Table 2 Summary of model expressions, estimated kinetic parameters and goodness of fit
Curve-fitting coefficients
Model R
2
avg
RMSE
a b c
(1) Sigmoid 0.998 0.103 8.17 1.67 NA
(2) Gompertz 0.996 0.175 7.74 2.25 NA
(3) Hill 0.995 0.206 7.68 3.68 5.46
(4) Logistic 0.995 0.206 7.68 -3.68 NA
(5) Chapman 0.994 0.220 7.61 0.44 7.38
(6) Weibull 0.991 0.301 9.18 131.90 42.21
NA: Not application

The summary of mathematical expressions showed the closed forms of mechanistic growth
models (Table 2). In all growth equations, models disregard the death phase and terminate
as the cell density reaches the maximum cell concentration (
max
). The variations of
Time (h)
0 5 10 15 20 25 30
N
u
m
b
e
r

o
f

E
.

c
o
l
i

(
l
o
g

C
F
U
/
m
l
)
0
2
4
6
8
10
gompertz 4parameter
hill 4parameter
chapman 4parameter
b
Time (h)
0 5 10 15 20 25 30
N
u
m
b
e
r

o
f

E
.

c
o
l
i

(
l
o
g

C
F
U
/
m
l
)
0
2
4
6
8
10
sigmoid 4parameter
logistic 4parameter
weibull 4parameter
a
The 12
th
16 -18 June, 2011

218

mathematical expressions represent various forms of the sigmoidal functions derived from
the mixture of empirical and semimechanistic equations. The simplest form in this list
(Table 2) containing the useful physical meanings was the sigmoid model. The model was
derived from of the original mechanistic differential equation describing the exponential
growth of multiplying organisms (Eq. 1). This simple and well-represented model is
widely used that increases without bounds. In mathematical terminology, the specific
growth rate () is a function of cell concentration x(t) at any given time. For the Sigmoid
equation, the specific growth rate is described in negative feedback term (or (x)
illustrated in Eq. 2) that slows the maximum specific growth rate (
max
) as the cell
concentration approaching the maximum cell density (x
max
)(Mitchell and others 2004).
x
dt
dx
= Eq. 1
( )
|
|
\
|
=
max
max
1
x
x
x Eq. 2
Notice that the feedback term in Eq. 2 ( ) x is close to
max
at the early stage of incubation
or small cell density. The specific growth rate then again approaches zero as cell density
enters the stationary phase or x reaches the maximum final cell density. Thus, the growth
rate begins at the maximum specific growth rate where cells are multiplied exponentially.
But then it decreases to zero as the cell density approaches its limit, producing an S-shaped
(sigmoidal) growth trajectory.
It is possible to solve the logistic differential equation above to obtain an analytic
(algebraic) solution. The analytical solution to the logistic differential equation can be
shown as follows:

( )
b
x x
e
a
y y
0
1
0
+
+ = Eq. 3
This specific growth rate function eliminated the requirement of any additional observable
variables (i.e., substrate concentration) as in the conventional Monod function. The
equation (Eq. 3) provides meaningful kinetic terms to biological system, including initial
cell density, inflection time, maximum specific growth rate, and maximum asymptotic cell
density. The y
0
represents the amount of initial inoculation of E. coli. The parameter is the
difference between the maximum and initial cell density (or y
max
y
o
) in log CFU/ml. The
inflection time (x
o
) was determined from the time (h) at the inflection point where cell
density equals to the average between maximum and the minimum values (or (cell
density
max
+ cell density
min
)/2). The key intrinsic growth characteristic was the maximum
specific colony growth rate (
max
) in h
-1
estimated from
b
1
. Despite the fact that all
equations agreed well with the E. coli growth as reflected by coefficient of determination
and low RMSE, only the sigmoid model was utilized to estimate kinetics of E. coli growth
in the subsequent experiment.
Growth of E. coli at various TSB concentrations
The logistic model was applied to investigate the effect of nutrient concentration of the
TSB and study its effect on the growth characteristic. The changes of TSB strength varying
The 12
th
16 -18 June, 2011

219

the concentration from 0.25X to 3X only marginally affect the E. coli growth profiles
(Figure 2). The high strength of TSB at 3-fold concentration showed slight lag during the
exponential phase. Presumably, the high hypotonic stress and less free water imposed
unfavorable condition for cell growth.
The similar growth profiles occurred on the broad spectrum of TSB concentrations even at
very low TSB concentration (i.e., 0.25X treatment) suggesting adequate nutrients sufficient
to maintain proper growth of E. coli during non-selective enrichment. The benefit of
medium cost reduction (as much as 4-fold less usage) can be related to industrial
application where constant enrichment practice performs on the daily basis. The estimated
values of the maximum specific growth rate and the inflection time showed significant
differences at the 2X and 3X treatments (Table 2). The TSB concentrations from 0.25X to
1X did not statistically affect the growth performance of E. coli.

Figure 2 Comparing the growth kinetic of E. coli between typical concentration and upper standard
of TSB concentration (e.g. 0.25X, 0.5X, 1X, 2X and 3X)

Table 3 The effect of TSB concentration on the growth of E. coli
No. of E. coli (log cfu/ ml)
Treatments
max
r
2
t
0
(h)
Initial cell
count
Final cell
count
1) 0.25X 0.61
a
0.00 09970.001 4.87
b
0.04 1.770.02 8.870.01
2) 0.50X 0.61
a
0.00 0.9980.001 4.84
ab
0.02 1.820.05 9.150.00
3) 1.00X 0.62
a
0.01 0.9980.000 4.78
a
0.03 1.740.06 9.190.01
4) 2.00X 0.59
b
0.01 0.9970.000 4.90
b
0.00 1.780.00 9.320.03
5) 3.00X 0.46
c
0.00 0.9920.001 5.42
c
0.05 1.850.00 9.470.02
a,b,c
values in a column with different superscripts are significantly difference at P<0.05.
Effect of sodium pyruvate using 0.25X TSB
As the end product of glycolysis, pyruvate serves as a cytosolic substrate for mitochondrial
oxidation in the Krebs cycle. Several literatures supported the use of pyruvate to supply
Time (h)
0 5 10 15 20 25 30
N
u
m
b
e
r

o
f

E
.

c
o
l
i

(
l
o
g

C
F
U
/

m
l
)
0
2
4
6
8
10
TSB 0.25X
TSB 0.50X
TSB 1.00X
TSB 2.00X
TSB 3.00X
The 12
th
16 -18 June, 2011

220

carbon requirement for quick colony growth and cell recovery (Abdel-Hamid and others
2001). Particullarly, pyruvate had been applied to increases the number of E. coli 0157:H7
cells after exposed to freezing and heating (6.Czechowicz and others 1996 and Mizunoe
and others 1999). The 0.25 X TSB was selected to minimize the excessiveness of available
nutrient and accentuate the effect of pyruvate component. However, the application of
sodium pyruvate in this case only showed minimal improvement to boost E. coli growth
(Figure 3). The nutrients provided in the TSB cocktail outperformed the effect of the
pyruvate supplement. The use of only sodium pyruvate solution was not able to maintain
the growth of E. coli and substantially undermined the growth kinetics as well as the final
achievable cell concentrations. The summary of extracted growth kinetics corresponded
well with the findings in Figure 3. The logistic model provided a good fit with the actual
data and was appropriate to describe the sigmodal characteristics of microbial batch growth
kinetics.

Figure 3 Comparing the growth kinetic of E. coli between typical concentration and upper standard
of TSB concentration (e.g. 0.25X, 0.5X, 1X, 2X and 3X).

Table 4 The effect of sodium pyruvate concentration on the growth of E. coli.
Treatments
max
r
2
t
0
(h) No. of E. coli(log cfu/ ml)
TSB
Sodium
pyruvate

Initial cell
count
Final
cell count
0.00X 1.00X 0.39
c
0.02 0.9890.005 5.91
d
0.15 1.700.00 5.880.00
0.25X 0.00X 0.61
b
0.00 0.9970.001 4.86
a
0.03 1.770.02 8.870.01
0.25X 0.59
b
0.01 0.9950.001 5.39
b
0.08 2.130.03 8.850.00
0.50X 0.60
b
0.01 0.9950.001 5.54
bc
0.12 2.220.09 8.930.04
1.00X 0.58
b
0.01 0.9990.001 5.44
bc
0.04 2.170.02 9.000.04
2.00X 0.69
a
0.02 0.9990.000 5.54
bc
0.03 2.270.01 8.760.04
3.00X 0.58
b
0.00 0.9980.001 5.61
b
0.02 2.260.00 8.860.06
a,b,c
values in a column with different superscripts are significantly difference at P<0.05.
Conclusion
Logistic model was simple and suitable to simulate batch growth kinetic of E. coli
cultivation. The straight-forward derivation from mechanistic growth model provided a
The 12
th
16 -18 June, 2011

221

wealth of useful kinetic parameters for growth optimization and comparison. The reduction
of TSB concentration was possible to lower enrichment cost while maintaining comparable
growth characteristics as achieved by the original TSB concentration. The TSB reduction
as low as 4-fold of the original strength was performed on active cells and no growth
deterioration was observed. The pyruvate supplement at any concentrations on the minimal
TSB concentration did not improve growth performance of E. coli.

Acknowledgement
The authors are grateful to have strong industrial partnership with Buono (Thailand) Co.,
Ltd. The authors would like to express sincere appreciation to Mrs. Achara Jiamthaworn
and Mr. Chettha Supparasuwat for their kind collaboration through the course of our
research.

Special thanks are extended to Thai Research Coucil (TRC) for their graduate scholarship
support or The Thailand Research Fund- Master research Grants (TRF- MAG).

References
Abdel-Hamid AM, Attwood MM, Guest JR. 2001. Pyruvate oxidase contributes to the
aerobic growth efficiency of Escherichia coli. Microbiology 147: 1483- 1498.
Baker M, Eyles R, Bennett J, Nicol C, Wong W, Nephrologist P, Garrett N. 1999.
Emergence of verotoxigenic Escherichia coli (VTEC) in New Zealand. New
Zealand Public Health Report. 6(2): 9-16.
Blackburn CDW, McCarthy JD. 2000. Modifications to methods for the enumeration and
detection of injuired Escherichia coli 0157:H7 in foods. International Journal of
Food Microbiology 55: 285-290.
Czechowicz SM, Santos O, Zottola EA. 1996. Recovery of thermally-stressed Escherichia
coli O157: H7 by media supplemented with pyruvate. International Journal of Food
Everia L. 2001. Injured bacteria in foods. Nutrition and Food Science 31 (2): 84-87.
Kudo YH, Ikedo M, Kodaka H, Nakagawa H, Goto K, Masuda T, Konuma H, Kojima T,
Kumagai S. 2000. Selective Enrichment with a Resuscitation Step Isolation of
Freeze-Injuired Escherichia coli O157:H7 from Foods. Applied and Environmental
Microbiology 66 (7): 2866-2872.
Marks PB. 2008. Status of Microbial Modeling in Food Process Models. Comprehensive
Reviews in Food Science and Food Safety 7: 137-143.
McMeekin TA, Baranyi J, Bowman J, Dalgaard P, Kirk M, Ross T, Schmid S, Zwietering
MH. 2006. Information systems in food safety management. International Journal
of Food Microbiology 112: 181-194.
Meyer PS, Yung JW, Ausubel JH. 1999. A Primer on Logistic Growth and Substitution
The Mathematics of the Loglet Lab Software. Technological Forecasting and
Social Change 61 (3): 247-271.
Mitchell DA, Meien OF, Krieger N, Dalsenter FDH. 2004. A review of recent
developments in modeling of microbial growth kinetics and intraparticle
phenomena in solid state fermenter. Biochemical Engineering Journal 17: 15 26.
MiZunoe Y, Wai SN, Takada A, Yoshida SI. 1999. Restoration of culturability of
starvation-stressed and low-temperature-stressed Escherichia coli 0157 cells by
using H
2
O
2
-degrading compounds. Arch Microbiol 172: 63-67.
The 12
th
16 -18 June, 2011

222

Saeaung W. Boonyaprapasorn A. 2010. Comparison of mathematical models to describe
nutrient limitation of E. coli colony expansion on TSA agar. Proceedings of
International Conference for a Sustainable Greater Mekong Subregion. 199 202p.
Saeaung W, Boonyaprapasorn A, Thipayarat A, editors. 2010. Digital image detection of
E. coli micro image to accelerate aerobic plate count analysis. Proceedings of Food
Innovation Asia Conference. 119 127p.
Simpon SS, Corradini MG, Normand MD. 2007. Estimating microbial growth parameters
from non-isothermal data: A case study with Clostridium perfringens. International
Journal of Food Microbiology 118: 294-303.
Statistical Analysis System. 1983. SAS Introductory Guide. SAS Institute. Cary. NC. 93p.
Weagant SD, Bound AJ. 2001. Evaluation of techniques for enrichment and isolation of
Escherichia coli O157:H7 from artificially contaminated sprouts. International
Journal of Food Microbiology 71: 87-92.

The 12
th
16 -18 June, 2011

223

OD-205

The Assessment of Food Safety Practices
Among Foodservice Operators in Puncak Alam, Selangor, Malaysia

Hashim Fadzil Ariffin
*
, Nor Alifah Ahmad, Ahmad Safwan Abd Aziz,
Nor Azshafida Mohd Azman, and Chemah Tamby Chik

Faculty of Hotel and Tourism Management, Universiti Teknologi MARA
40450 Shah Alam, Selangor, Malaysia
*
Corresponding author: hashim@salam.uitm.edu.my

Abstract

This research tried to i) examine the relationship between prior knowledge and experience
and the assessment of food safety practices, ii) identify the program effectiveness that
affects the assessment of food safety practices , and iii) investigate the management
involvement in the assessment of food safety practices through foodservice management.
Data was gathered from 200 restaurant managers, hawkers, stall handlers and bazaar
handlers in Puncak Alam, Selangor Malaysia using a survey questionnaire. Data then was
analyzed using Regression Analysis in SPSS version 17 to interpret the relationship
between dimensions of social demographic, safety program effectiveness, management
involvement and food safety practices. The hypotheses were expected to be positively
proven and accepted. First of all, this study was purposely done to specify the assessment
of food safety practices. Next, this study is beneficial to foodservice operators because they
should know how food safety practices be used in the operations. Foodservice operators
can properly practice the food safety in order to improve quality of foods and services.
Finally, foodservice operators awareness in food safety practices is an importance factor
to reduce food poisoning problem. It is found that there is a relationship between the
variables. Management involvement is the strongest factor beside prior knowledge and
experience, and safety program effectiveness. This area must be extensively studied
especially by scholars and foodservice operators, as it will give an added advantage to
attract new customers as well as to retain current customers.

Keywords: Food safety practices, program effectiveness, management involvement

The 12
th
16 -18 June, 2011

224

OD-263
Thermal Resistance of Local Isolates of Staphylococcus aureus
Ratih Dewanti-Hariyadi
1,2,*
, Juli Hadiyanto
2
and Eko Hari Purnomo
1,2

1
Southeast Asia Food Agricultural Science and Technology (SEAFAST) Center;
2
Department of Food Science and Technology, Faculty of Agricultural Technology,
Bogor Agricultural University, Bogor 16002, Indonesia
*
Corresponding author: ratihde@ipb.ac.id

Abstract
Food poisoning is an important indicator for food safety status in any country. Report from
The National Agency for Drugs and Food Control of Republic of Indonesia (BPOM)
showed that 40.87% of food poisoning occurred in Indonesia was associated with
homemade food. Staphylococcus aureus probably is an important pathogen contributing to
the food poisoning cases in Indonesia, because this pathogen is a natural flora that lives in
human body and could contaminate food due to poor sanitation and hygienic practices.
Generally, growth of S. aureus can be prevented by temperature modification such us
refrigeration and heating. Since most of Indonesian foods are heavily heated, it is
interesting to know whether most processing could actually inactivate a large number of S.
aureus and whether the pathogens isolated from local sources are tolerant to heat. This is
important because the risk of having S. aureus surviving in cooked food is even worse
since they could eventually produce enterotoxins. The goal of this research is to evaluate
the heat resistance of several S. aureus isolated from ready to eat (RTE) Indonesian
traditional foods. The study was conducted by inoculating 1 ml of a late log phased S.
aureus culture into 9 ml of heating menstruum (Trypticase Soy Broth) at 52, 53, 54, and
56
o
C for 5, 7, 10, and 15 minutes. S. aureus surviving from the heating process was
enumerated on Baird Parker Agar (BPA) media containing egg yolk tellurite after
incubation for 48 hours at 35
o
C. Thermo tolerance parameters, i.e. D and Z values were
estimated using standard regression analysis based on log linier models. The result was
used to estimate the adequacy of various cooking methods for several RTE Indonesian
traditional foods. The D
53
, D
54
, D
55
, and D
56
values of local isolates of S. aureus were
19.47- 64.59 min, 13.42 23.8 min, 6.59 14.3 min and 5.17-8.78 min, respectively. The
thermal inactivation of S. aureus followed rst order kinetics with r
2
values of 0.92-0.99.
The Z values calculated in this study ranged from 3.37 to 6.06
o
C. These values were within
the range of reported Z values for most non-spore forming bacteria (4 6
o
C). This study
provided data on the thermal resistance of S. aureus isolated from Indonesia and validated
that heating commonly applied in cooking of RTE traditional foods could reduce
Staphylococcus aureus to up to 6.9x10
6
log cycle. However, common practices following
heating of certain foods may allow recontamination, thus handling of RTE foods after
cooking is very important for the management of this pathogen.

Keywords: Staphylococcus aureus, D-value, Z-value, thermal resistance

The 12
th
16 -18 June, 2011

225

Introduction
Staphylococcus aureus is an important foodborne pathogen worldwide and has been linked
to various foodborne disease outbreaks. The ability the bacterium to grow in food
containing salt up to 20% or Aw as low as 0.83 as well as its ability to produce different
kinds of enterotoxins (Adams and Moss 2005) have been thought to play a role in causing
food intoxication in processed food such as pasteurized milk, cream-filled bakery etc.

In Indonesia, bacterial pathogens were the main cause of foodborne outbreaks between
2007-2010. Based on limited data, the report also suggested that ready-to-eat food
produced in home was accountable for 40% of the outbreaks (BPOM 2010). Although
Staphylococcus aureus was not singled out as the main causative agent in the outbreaks, it
was very likely that the pathogen was responsible for some of the outbreak because of poor
implementation of sanitation and hygiene program observed in various food vendors. S.
aureus has been reported to grow well in several ready-to-eat (RTE) Indonesian traditional
foods such as chicken soup, stir fry green bean and rice cooked in coconut milk (nasi uduk)
(Dewanti-Hariyadi and others 2008). Dwintasari (2010) and Apriyadi (2010) reported that
S. aureus can be isolated from hands of workers in street vendors, nasi uduk and shredded
chicken in several vendors around Bogor area. Several factors can be attributed to the
finding of these bacteria in such foods. First, although most Indonesian traditional foods
are wll cooked, post processing handling may lead to contamination. Secondly, common
household practice to store ready-to-eat at rom temperaturee may support bacterial growth
and subsequent toxin producton. Thirdly, reheating which is a comon practice may be
inadequate and or may not be effective since the heat stable enterotoxin may have already
been produced.

S. aureus in non-spormer bacteria which could easily be killed by heat. Although data of
thermal resistance of S. aureus have neen reported worldwide, it is not known whether S.
aureus isolated from RTE Indonesian traditional foods which generally receive long
cooking time has similar heat resistance.

The objectives of this study was to obtain information on the thermal resistance (D and z
values) of several S. aureus previously isolated from RTE Indonesian traditional foods and
use the information to evaluate thermal adequacy of several cooking methods commonly
applied in food vendors.

Inoculum preparation S. aureus used in this study were AS2 (isolated from shredded
chicken), NU3 (isolated from nasi uduk) obtained by Apriyadi (2010) and ATCC 25923 as
a control. Individual isolate was grown in Tryptose Soy Broth (TSB) at 35C for 24 h to
reach late log phase. The culture containing ca. 1.0x10
8
-1.0x10
9
CFU/ml was used as
inoculum to achieve the desired initial concentration in the heating menstruum.

Preparation of Heating Menstruum
The heating menstruum was 9 ml TSB which was ppreviously sterilized at 121
o
C for 15
minutesThermal Resistance TestingSets of glass tubes containing the heating menstruum
were placed in different waterbath set at 53, 54, 55, dan 56C. When the heating
menstruum reached the desired temperatures, one mililiter of overnight culture of S.
aureus was inoculated into the glass tubes containing the heating menstruum such that the
The 12
th
16 -18 June, 2011

226

initial counts were ca. 1,0x10
7
-1,0x10
8
CFU/ml. The menstruum in the tubes was allowed
to be heated for 5, 7, 10, and 15 minutes. Enumeration of S. aureus surviving the heating
was carried out on Baird Parker Agar (BPA) containing egg yolk tellurite (Bennet and
Lancette 2001) after incubation at 35C for 48 h. The number of S. aureus surviving were
plotted against the heating times to yield a curve of rate of inactivation at four different
temperatures, i.e. 53, 54, 55, dan 56
C. Based on the curve, the D values, i.e. time

(minutes) at certain temperatures to reduce the number of S. aureus by 1 log cycle can be
calculated from the equation D = -1/slope. A Thermal Death Time (TDT) curve was made
to establish the relationship between D (minutes) with temperatures (
o
C). The Z values,
i.e. temperature intervals to reduce D by 1 log cycle was also determined from the curve.

Assessing Thermal Adequacy of Several RTE Indonesian Traditional Foods
Assessment of the thermal adequacy of various RTE foods was conducted through a
survey to determine the common practice and time of cooking generally applied by food
vendors and measurement of product internal temperatures during cooking. The
respondents were 16 food vendors surrounding Darmaga campus area, Bogor. The
adequacy of thermal process applied in the surveyed food vendors was assessed by
extrapolating the Z value equation to obtain D
T
(D values at at the cooking temperatures
applied).

D values of local isolates of Staphylococcus aureus
The data of heat resistance of local isolates of S. aureus are presented in Figure 1. The D
values of S. aureus isolate AS2 originated from shredded chicken at 53, 54, 55 and 56
C were
19.471.33; 13.420.13; 6.590.85, and 5.170.26 minutes, respectively. S. aureus NU3,
isolated from nasi uduk had D values of 64.59 2.95, 23.83 0.80, 14.30.78, and 8.780.92
minutes at 53, 54, 55, and 56
C. Meanwhile, the control ATCC isolate had D values of

22.00 1.02, 15.31 1.16, 11.120.52, and 7.531.76 minutes at the above temperatures. The
data suggest that the heat resistance of the three isolates varied. The linear equation of the
logarithmic decrease of the bacterial number had r
2
values of 0.92-0.98.

S. aureus NU3 had D
53
, D
54
, D
55
, D
56
values higher than AS2 and ATCC 25923. This
strain was isolated from nasi uduk which was generally maintained in rice thermos at 40-
60
o
C (Dewanti-Hariyadi and others 2008). Storage at the temperature range probably
contribute to the increased heat resistance (Jay 2000).

The D
53,
D
54
, D
55,
dan D
56
values of Staphylococcus aureus AS2, NU3, dan ATCC 25923
were higher than that reported by Walker and Harmon (1966) who concluded that the heat
resistance of S.aureus S-18 and B-120 in phosphate buffer was lower than those in milk.
They concluded that the isolates had lower D values in the phosphate buffer, possibly due
to protection by nutrient in milk. Jay (2000) also stated that nutrient such as carbohydrate,
proteins, fat, soluble solid as well as aw and pH strongly influence cell damage during
heating. In general carbohydrate, proteins, fat and soluble solid protect bacterial cell from
heating. Heat resistance increases with increase in the above nutrient content. Since this
study was conducted in TSB, it is thought that the nutrent protection is similar to milk.
Additionally, the difference in the heat resistance of isolates used in this study and those of
Walker and Harmon (1966) could also be attributed to strain difference (Jay, 2006).

The 12
th
16 -18 June, 2011

227

Figure 1 Log decrease of the number of S. aureus heated at 53, 54, 55, 56
C.

The D
55
values of AS2 and ATCC 25923 isolates were lower than the D
55
value of S.
aureus cocktail reported by Kennedy and others (2005) i.e.13.0 minutes. Parente and
Mazzatura (1991) also reported the heat resistance of S. aureus BP3 and S. aureus 237
The 12
th
16 -18 June, 2011

228

isolated from goat milk. Isolate BP3 had also lower D
55
value (3.30 minutes), while isolate
237 had similar D
55
(10.60 minutes) to those of NU3, AS2 and ATCC 25923.Z Values of
Staphylococcus aureus AS2, NU3, and ATCC 25923

The sensitivity of D values to temperature changes is expressed as Z-value, i.e. changes of
temperatures to change D value by 1 log cycle or 90% (Toledo 1991). The Z values of S.
aureus AS2 were 4.74-5.10C, S. aureus NU3 were 3.37-3.7C whileATCC 25923 were
5.59-6.06C (Figure 2). The results showed that S. aureus NU3 had the lowest Z thus the
D value of the isolate was more sensitive to temperature changes than that of AS2 or
ATCC 25923. Figure 2 suggests that S. aureus NU3 is more heat resistant than AS2 and
ATCC 25923 at temperatures less than 56
o
C. However, this is not always happen when
heating temperature changes. Analysis of the Z values suggested that the intercept
between Z curves of NU3, AS2 and ATCC 25923 occured at 57.6C, 55.9C and 50.3C.
Two microorganisms have the same heat resistance at the interception of the Z curve due
to the same D-values (Toledo 1991). Our results suggests that NU3 and AS2 isolates have
the same heat resistance at 57.6C. At temperatures below 57.6C, NU3 isolate is more
heat resistant than that of AS2; however at temperatures above 57.6C, NU3 isolate is less
heat resistant than that of AS2.
The 12
th
16 -18 June, 2011

229

Figure 2 Z-value curves of S. aureus AS2, NU3, dan ATCC 25923
The Z values of S.aureus isolates in this study varies and similar to previous studies
reported by Stumbo (1973) for S. aureus in pasteurized food i.e. 4.6-6.7C. Isolates AS2
and ATCC 25923 had Z values within the range of those reported by Stumbo (1973).
However, the Z value of NU3 (3,3-3,37C) was lower than that of Stumbo (1973). Eden
and others (1977) reported a Z value of 9.46C, while Kennedy and others (2005)
concluded that a Staphylococcus aureus cocktail had Z value ranging from 7.70 to 8.0C.
The Z values reported in this study were a lot lower than those reported by Kennedy and
others suggesting that these local isolates had a higher sensitivity toward heat. However
the Z values of local isolates of S. aureus were similar to other nonsporming bacteria in
protein-rich heating menstruum such as chicken broth, TSB etc, i. e. 5
o
C. Table 1 showed
the Z values reported in this study as compared to other pathogens (Table 1).
The 12
th
16 -18 June, 2011

230

Table 1 Z values of Staphylococcus aureus AS2, NU3,and ATCC 25923 as compared to other
pathogens in protein-rich heating menstruum
Micrroorganism Heating menstruum Z (C)
Staphylococcus aureus AS2 TSB 4,74-5,10
Staphylococcusaureus NU3 TSB 3,37-3,7
Staphylococcus aureus ATCC 25923 TSB 5,59-6,06
Campylobacter jejuni
a
chicken broth 5,81
Salmonella
b
chicken broth 5,35
Listeria monocytogenes
b
chicken broth 5,11
Salmonella

typhimurium
c
chicken broth 5,80
Salmonella enteritidis
c
chicken broth 5,86
Yersinia enterolitica
d
minced beef 5,1
S. epidermidis
e
chicken broth 7,46
Escherichia coli O-157
f
breaded pork patties 5,43
a
(Blankenship and others 1982),
b
(Murphy

and others 2004),
c
(Jenuja and other 2001),
d
(Bolton
and others 2000),
e
(Bertolatti and others 2001),
f
(Osaili and others 2007)
Evaluation of Thermal Process Adequacy of Several RTE Indonesian Traditional
Foods Sold in Food Vendors
The results of the survey of 16 food types from 16 food vendors is presented in Table 2. In
general traditional RTE foods were heated at temperatures above 70
o
C. The foods were
either boiled , steamed, grilled, fried, stirfried or received a combination of two cooking
methods.

Using the equation obtained in the Z curves, extrapolation was carried out to determine D
73

and D
92
values
.
Extrapolation at 92
o
C was conducted to simulate boiling, while 73
0
C was
used to simulate stir friyng and grilling. The results of extrapolation was used to assess the
thermal process adequacy.

The 12
th
16 -18 June, 2011

231

Table 2 Survey of cooking practices and temperature f several RTE traditional foods during cooking

Vendor Name of Food

Cooking Methods

Temperature of
product during
heating
1 Semur jengkol boiling for 1 h 92
o
C
2 Beef soup boiling beef for 3 h, storage at RT, cutting
into cubes, mixing with broth
92
o

C
3 Meatball soup boiling for 1 h 96
o
C
4 Steamed coconut milk-
rice
boiling for 30 min, steaming for 30 min,
storage at 50
o
C
82
o
C
5

Steamed rice boiling for 30 min, steaming for 25 min,
storage at RT
83
o
C
6 Chicken opor boiling for 1 h 95
o
C
7 Vermicelli salad boiling of vermicelli, mixing with fresh
chilli-peanut sauce
89
o
C
8 Cooked vegetable salad boiling vegetable, mixing with fresh chilli-
peanut sauce
89
o
C
9 Siomay steaming continuously 86
o
C
10 Grilled chicken boiling for 2 h, storage at RT, grilling 5-6
minutes
73
o
C
11 Grilled fish grilling for 10 min 73
o
C
12 Fried coconut chicken frying in shredded coconut until brown 95
o
C
13 Fried tempe frying for 2-3 min 98
o
C
14 Fried potato frying for 2-3 min 98
o
C
15 Stir fried green bean stir frying for 5 min 73
o
C
16 Stir fried eggplant stir frying for 5 min 73
o
C

Table 3 shows the extrapolated D
73
dan D
92
. The extrapolated values suggest that cooking
food at 73C for 0,00006-0,011 minute could reduce Staphylococcus aureus by one log
cycle. Similar effect could also be obtained by cooking at 92C for 1,5x10
-10
- 1,93 x10
-6

minute

.

Table 3 Extrapolated D
73
and D
92
values of S. aureus AS2, NU3, and ATCC 25923
Isolate D
73
(min) D
92 (min)
NU3 (1) 0,0002 1,62x10
-9
NU3 (2) 0,00006 1,5x10
-10

AS2 (1) 0,001 1,25 x10
-7

AS2 (2) 0,002 4,4 x10
-7

ATCC 25923 (1) 0,006 1,93 x10
-6

ATCC 25923 (2) 0,011 8,70 x10
-6

Boiling food at 92C for an hour could reduce 6,9x10
6
log cycle of the bacteria, meanwhile
stir-frying at 73C for 5 minutes decreases 454,5 log cycle of S. aureus. Assuming that the
initial count of nasi uduk is 1,0x10
3
CFU/g (Hartini 2001), a serving size of nasi uduk of
100 g would contain 1.0x10
5
CFU S. aureus and boiling at 92C for 1 hour or stir
frying at 73C for 5 minutes could reduce S. aureus to very low number (<1/10
449,5
).

The
results suggested that the likelihood of S. aureus to present in the RTE Indonesian
traditional food after cooking was very low and compliance with various guideline that
called for a maximum S. aureus of 1x10
2
CFU/g (Shapton and Shapton 1993) or 0-5x10
3

CFU/g(BPOM 2009) are easy to achieve.

The 12
th
16 -18 June, 2011

232

Although the cooking process of all RTE Indonesian traditional foods provide an adequate
heating for inactivation of S. aureus, several practices may allow recontamination of the
foods. For beef soup for examples, prolonged storage and cutting the beef into cubes may
allow the re-entry of S. aureus or other pathogens as well as spoilage bacteria. Vegetable
and vermicelli for the salad are boiled, however mixing with fresh chilli-peanut sauce may
introduce bacteria. Grilled chicken also needs precaution such that storage time after
boiling prior to grilling is short (< 2 h) to avoid growth although the grilling should be able
to decrease bacterial number substantially.

Table 4. Inactivation of S. aureus due to cooking commonly practiced for RTE traditional foods

Cooking method Temperature Time Decrease in S.
aureus number
(log cycle)
Note
Boiling 92
o
C 1 h 6.9 x 10
6
Boiling is effective in
reducing S. aureus, however
some products were mixed
with fresh sauce or stored at
RT for prolonged period
prior to serving thus permit
recontamination
Stirfrying/grilling 73
o
C 5 min 454.5 Stir frying is effective in
reducing S. aureus,
recontamination may occur
during storage at RT

Conclusion
This study found that S. aureus isolated from RTE Indonesian traditional foods had
thermal resistance similar to that reported from S. aureus isolated elsewhere, with D
53
,
D
54
, D
55
, and D
56
values of 19.47- 64.59 min, 13.42 23.8 min, 6.59 14.3 min and 5.17-
8.78 min, respectively. The Z values of S. aureus isolates ranged from 3.37 to 6.06
o
C,
which were within the range of reported Z values for most non-spore forming bacteria.
The study also reported that heating commonly applied in cooking of RTE Indonesian
traditional foods could significantly eliminate Staphylococcus aureus. Therefore, post
cooking practices became very critical because recontamination may occur.

References
Adams MR, Moss MO. 2005. Food Microbiology 2
nd
Edition. United Kingdom: The
Royal Society of Chemistry.
Apriyadi TE. 2010. Risk of Staphylococcus aureus in ready to eat Indonesian traditional
foods and evaluation of its presence in nasi uduk. Skripsi. Available from: Faculty
of Agricultural Technology, Bogor Agricultural University (in Bahasa Indonesia).
Bennett RW, Lancette GA. 2001. Staphylococcus aureus . In Bacteriological Analytical
Manual. Association of Official Analytical Chemist, Arlington, VA.
Bertolatti D, Steven JM, Warren BG, Colin WB. 2001. Thermal inactivation of
antimicrobial resistant gram-positive cocci in chicken meat: D and Z value
determinations. Int. J. Environ. Health Res 11:257 266.
Blankenship LC, Craven SE. 1982. Campylobacter jejuni survival in chicken meat as a
function of temperature. J.Appl. Environ. Microbiol. 44: 88-92.
Bolton DJ, McMahon CM, Doherty AM, Sheridan JJ, McDowell DA, Blair IS, Harrington
D. 2000. Thermal inactivation of Listeria monocytogenes and Yersinia
The 12
th
16 -18 June, 2011

233

enterocolitica in minced beef under laboratory conditions and in sous-vide prepared
minced and solid beef cooked in a commercial retort. J. Appl. Microbiol. 88: 626-
632.
BPOM. 2010. Annual report of foodborne disease outbreaks 2009. Unpublished data. The
National Agency for Drug and Food Control (NADFC), Republic of Indonesia.
BPOM. 2009. Regulation on food contaminants. The National Agency for Drug and Food
Control (NADFC), Republic of Indonesia.
Dewanti-Hariyadi R, Rawendra R, Dewi SP. 2008. Growth of S. aureus in Indonesian
Traditional Dishes at Storage Temperatures Practiced by Households. Poster
Presentation at Conference of Asian Food and Nutrition Safety, Cebu, Phillippine,
November 5-6, 2008, ILSI-SEA.
Dwintasari V. 2010. Growth of Staphylococcus aureus in shredded chicken and its
correlation with cleanness of employee and handling practices in vendors. Skripsi.
Available from: Faculty of Agricultural Technology, Bogor Agricultural University
(in Bahasa Indonesia).
Hartini PB. 2001. Study on microbiogical safety of food sold in campus canteen. Skripsi.
Available from: Faculty of Agricultural Technology, Bogor Agricultural University
(in Bahasa Indonesia).
Jay JM. 2000. Modern Food Microbiology 6
th
Edition. Gaithersburg, Maryland: Aspen
Publishers, Inc.
Jenuja VK, Eblen BS, Ransom GM. 2001. Thermal inactivation of Salmonella spp. in
chicken broth, beef, pork, turkey, and chicken: Determination of D- and Z-values.
J.of Food Sci. 66 :1.
Kennedy K, Blair IS, McDowel DA, Bolton DJ. 2005. An investigation of the thermal
inactivation of Staphylococcus aureus and the potential for increased
thermotolerance as a result of chilled storage. J. Appl. Microbiol. 99: 12291235
Murphy RY, Osaili, T, Duncan, LK, Marcys, JA, 2004. Thermal inactivation of Salmonella
and Listeria monocytogenes in ground chicken thigh/leg meat and skin. J. Poultry
Sci. 83:1218-1225
Osaili TM, Griffin CL, Martin EM, Beard BL, Keener AE, Marcy JA. 2007. Thermal
inactivation of Escherichia coli O157:H7, Salmonella, and Listeria monocytogenes
in breaded pork patties. J. Food Sci. 7 :2
Parente E, Mazzatura A. 1991. Growth and heat resistance of Staphylococcus aureus in
goat milk. Ital. J. Food Sci. 3: 27-37.
Shapton DA, Shapton NF. 1993. Principles and practises for the safe processing of foods.
Oxford, Great Britain: Butterworth-Heineman Ltd.
Stumbo CR. 1973. Thermobacteriologi in food processing. New York: Academic Press.
Toledo RT. 1991. Fundamentals of food process engineering 3
rd
. Georgia: Springer
Walker GC and Harmon LG. 1966. Thermal resistance of Staphylococcus aureus in milk,
whey and phosphate buffer. Appl. Microbiol. 14: 584-590

The 12
th
16 -18 June, 2011

234

OE : Driving Trends in Aquatic Food Market
The 12
th
16 -18 June, 2011

235

OE-16
Effect of Kraft Lignin on Rheological Properties and Protein
Aggregation of Fish Protein-Based Bioplastic
Yotsavimon Sakunkittiyut
1
, Thiranan Kunanopparat
2,*
, and Suwit Siriwattanayotin
1
1
Department of Food Engineering, Faculty of Engineering, King Mongkuts University of Technology
Thonburi, Bangmod, Tungkru, Bangkok 10140, Thailand;
2
Pilot Plant Development and Training Institute, King Mongkuts University of Technology
Thonburi, Bangmod, Tungkru, Bangkok 10140, Thailand
*
Corresponding author: thiranan.kun@kmutt.ac.th

Abstract
Protein waste from surimi industry is an interesting source to produce the bioplastics.
However, a main drawback of protein-based bioplastics is a narrow window processing of
extrusion due to protein aggregation via an exchange of thiol/disulfide bond. Kraft lignin
(KL) is obtained as a major industrial waste material of paper industries. It is known as
radical scavenger. It may interfere with the protein aggregation and reduce the blend
viscosity. Therefore, the objective of this work was to study the effect of KL on
rheological properties and protein aggregation of fish protein (FP)-based bioplastic. Firstly,
FP powder was blended with 30% glycerol and 0-70% KL. Then, the blends were thermal-
molded by compression molding. The viscosity of KL/FP mixtures was determined by
capillary rheometer. Protein aggregation and protein molecular weight of bioplastics were
respectively determined by protein solubility and sodium dodecyl sulfate polyacrylamide
gel electrophoresis (SDS-PAGE). Free radicals of bioplastics were measured by electron
spin resonance. KL addition resulted in a decrease in blend viscosity at processing
temperature. The rheology of KL/FP bioplastics followed a shear thinning behavior.
Therefore, the presence of KL in the mixture seemed to improve the flow properties. In
addition, KL addition increased protein solubility in SDS buffer, indicating a decrease of
protein molecular weight. However, molecular weight changes in the studied range and
radical scavenging effect between KL and FP were not observed.
Keywords: Bioplastic, free radical, protein aggregation, kraft lignin, viscosity

Introduction
Plastic packaging has come into widespread use because its good mechanical properties.
However, plastic is non biodegradable which poses a serious ecological problem.
Therefore, packaging from agricultural resources such as carbohydrates, starch, and
proteins is receiving considerable attention (Jerez and others 2007; Verbeek and Van Den
Berg 2010).
Blending polymers with other materials can aid in the development of new products with
better performance such as flow ability, mechanical properties (Torres-Giner and others
2009). The eco-friendly materials often make from natural polymers which are waste
products of agriculture and industry. Therefore, blending fish protein (FP) and Kraft lignin
(KL) which are respectively byproducts of surimi and paper industry may be an effective
method to produce a new value-added product with better properties.
The 12
th
16 -18 June, 2011

236

Fish waste is byproduct generated from surimi industry. In processing, it causes a
lot of wastes in water which contains unprocessed fish (Piyadhammaviboon and
Yongsawatdigul 2009). Although the nutritional values of these wastes are fairly high,
these useful resources have been mainly used as animal feed with low value (Bourtoom
and others 2009). Moreover, improper disposal of these wastes may cause pollution and
emit an oensive odour. In order to manage this problem, the protein waste is an
interesting source to produce the bioplastics.
However, a major drawback of protein-based bioplastics is a narrow window processing on
using conventional methods such as extrusion and injection molding. The problem of
protein-based in extrusion process when using high screw speed caused extrudate rupture
was reported (Pommet and others 2003; Ullsten and others 2006). During thermal
processing, protein has aggregation via sulfhydryl-disulfide interchange reactions, leading
to an increase of the covalent bonds and high viscosity. Ullsten and others (2006) reported
that salicylic acid allowed an enlargement of protein extrusion window. To delay protein
aggregation, free radical scavenger actions of salicylic acid were proposed and were
confirmed by electron spin resonance analysis.
Lignin is a natural polymer and a main component in plants (Shen and others 2008) which
usually derived from three phenolic molecules by a dehydrogenating polymerization
involving radical coupling. Beside, KL is the byproduct of the alkaline pulping process
from pulp and paper industry. It is a polyphenolic thermoplastic compound which is known
for its radical scavenging properties (Thielemans and others 2002). KL may interfere with
the protein aggregation and may decrease viscosity of protein due to its free radical
scavenging properties. Therefore, the objective of this work was to study the effect of KL
on rheological properties and protein aggregation of FP-based materials. Rheological
properties of KL/FP blends were determined by capillary rheometer. Protein aggregation
and protein molecular weight of KL/FP materials were determined by protein solubility
and sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), respectively.
In addition, free radicals of KL/FP materials were measured by electron spin resonance
(EPR).
1. Materials
Threadfin bream (Nemipterus sp.) were purchased from Prachauthid 61 market. (Bangkok,
Thailand). Kraft lignin (KL) was obtained from Raja engineering. Co. Ltd. (Bangkok,
Thailand).
2. Preparation of fish protein
Threadfin bream (Nemipterus sp.) were gutted and headed. After that, the fish mince were
washed, chopped and dried in hot air oven at 50
o
C for 5 hours and 40
o
C for 24 hours.
Finally, the dried fish protein (FP) was ground less than 425 m. The proximate
compositions of FP powder were determined as described by AOAC (1996). Protein, fat,
ash and moisture contents were 84.932.82, 4.280.76, 2.770.16 and 6.860.15 %
respectively.
2.1 Preparation of fish protein/Kraft lignin materials
Materials in this study contain a mixture of FP: KL: glycerol in a weight ratio ranging from
70:0:30 to 0:70:30.
The 12
th
16 -18 June, 2011

237

2.2 Mixing process
Fifty grams of KL/FP and glycerol were mixed in an internal mixer (Plasti-corder W50,
Brabender, Duisburg, Germany) at temperature 80 C, 100 rpm for 15 minutes.
2.3 Compression molding process
Ten or twenty-ve grams of the blend were placed in a square mould (9 9 cm) and
thermo molded at 100 C for 15 min in a Hydraulic Press Machine (20 T., SMC TOYO
METAL Co., Ltd., Thailand). A pressure of 1 ton was directly applied to the sample in the
mould. The thickness of material is approximately about 1 or 2 mm depending on initial
weight of blend.
3. Characterization of KL/FP materials
3.1 Rheological properties
After mixing, the melt ow behavior of KL/FP blends were determined by capillary
rheometer (capillary rheometer RHEO-TESTER 2000, Germany) with a 2-mm capillary
die and a length-to-diameter ratio of 15. Measurements were carried out at 140 C under a
shear rate ranging from 10 to 1000 s
-1
. Apparent viscosity is plotted against shear rate. The
Rabinowitsch correction was applied to account for the inuence of shear thinning in the
calculation of the shear rate and corresponding viscosity, and the Bagley correction,
corresponding to the adjustment for excess pressure drop at the die entrance, was applied
by using three capillaries with the same radius but different length/radius ratios. A simple
mathematical expression describing the relationship between viscosity and shear rate is
= K
n-1
(1)
where the consistency (K) corresponds to the viscosity value for a shear rate of s
-1
and
the power-law index (n) characterizes the deviation from the Newtonian behavior.
3.2 Protein solubility
Protein solubility was determined according to the method of Kunanopparat and others
(2009). Briefly, the protein powder (26.67 mg) was stirred for 80 min at 60 C in the
presence of 20 ml of 0.1 M sodium phosphate buffer (pH 6.9) containing 1% sodium
dodecyl sulfate (SDS). The SDS-soluble protein extract was recovered by centrifugation
(50 min at 15000g and 20 C), and 1000 l was used to determine protein content. The
pellet was suspended in 5 ml of SDS-phosphate buffer containing 20mM dithioerythritol
(DTE). After it was shaken for 60 min at 60 C, the extract was sonicated for 3 min. These
treatments brought insoluble protein from the pellet into the solution. After centrifugation
(50 min, 15000 g, 20 C), a part of the supernatant was determined protein content using
the Kjeldahl method (Kjeltec System-Tecator, Sweden).
3.3 SDS-PAGE
Molecular weight of materials was studied on SDSPAGE according to the method of
Laemmli (1970). Protein samples were solubilized in an SDSsolution containing 10 ml of
10% SDS, 5% 2-mercaptoethanol, 20% glycerol, 0.5 M TrisHCl (pH 6.8) and Trace
bromophenol blue 0.0003 g. The supernatant was used for gel electrophoresis. Stacking gel
and separating gel were made of 4% (w/v) and 15% (w/v) polyacrylamide, respectively.
The amount of protein loaded onto the polyacrylamide gel was 15 mg. The Precision Plus
Protein
TM
standard (10-250 kDa) as a marker was also injected at the loading of 20 l in
The 12
th
16 -18 June, 2011

238

the left lanes of the gel and the electrophoresis was carried out at 200 V for about 60 min.
After electrophoresis, gels were stained and distained.

3.4 Electron Spin Resonance (ESR) Spectroscopy
The radicals formed on material after treatments were identified according to the method
of Ullsten and others (2006) and quantified by electron spin resonance spectroscopy.
Firstly, samples were collected immediately after mixing and molding, and were put into
liquid nitrogen. In testing, the samples were placed into ESR glass containers. These were
immediately transferred to a liquid nitrogen container, in order to inhibit further reaction
before the spectromagnetic investigation. The complete cycle lasted for approximately 2
min. ESR spectra (first derivatives) were recorded for the frozen samples using a JEOL at
77 K (1 mW microwave power and 0.5 mT modulation amplitude). The g values were
determined by standardization with , - diphenyl--picryl hydrazyl (DPPH). The g value
is calculated from the relationship h = gB, where h is Planks constant (6.63
10
-34
J s), is the microwave frequency (9.4 GHz, measured by a frequency counter),
is the Bohr magnetron (9.2710
-24
A m
2
), and B is the magnetic field (G). Spectra were
recorded during runs.
1. Rheological properties of KL/FP blends
Viscosity of FP blends with 0-70% KL was determined by using capillary rheometer which
is a technique whereby a sample undergoes extrusion through a die of defined dimensions.
The test sample was pushed by a piston driven at temperature of 140
o
C through a capillary
die and measured its rheological properties like shear rate and shear viscosity. The data
was obtained from steady shear measurement as shown in Figure 1.
The shear viscosity of KL/FP blend decreased when shear rate increased. The shear
viscosity of the blends was recorded in the shear rate range of 10-1000 s
-1
. However, the
testing region for FP blend with 0% KL could not exceed 500 s
-1
with the same amount of
sample. Thus, the melt of the blend exhibited more stability under high shear. The addition
of KL resulted in a decrease of shear viscosity with an increase of KL content.

The 12
th
16 -18 June, 2011

239

Figure 1 Viscosity of FP blends with 0-70% KL contents determined by capillary rheometer at 140
o
C
of die temperature
The relationship between the shear viscosity and shear rate can be described by power law
equation according eq. (1). Power-law index (n) and flow behavior consistency (K) were
calculated as shown in Table 1. The power-law index of FP-based material is 0.17. Values
for the power index for protein ranging from 0.12 to 0.78 have been reported in the
literatures (Orliac and others 2003; Bengoechea and others 2007). All samples represented
shear thinning behavior and their ability to flow. Power-law index and the consistency
coefficient decreased when KL content increased. This may be explained by a low
molecular weight of KL (MW= 2000-5000) (Pouteau and others 2003; Silva and others
2009) compared to FP (MW= 8-600 kDa) (Lodha and Netravali 2005; Hernandez-
Izquierdo and Krochta 2008). In addition, KL may increase chain-mobility at high
temperature and high shear rate Therefore, the presence of KL in the mixture seemed to
improve the flow properties substantially. The rheological properties of FP-based
biomaterial strongly affected by the behavior of KL and were expected to play an
important role in any process developed to produce bioplastics from FP.
Table 1 Power-law model parameters of KL/FP blends
Sample Consistency (Pa.s
n
) n
0% KL 14569321442 0.17 0.01
20% KL 105266 4981 0.02 0.03
40% KL 47727 3976 0.14 0.01
60% KL 28459 428 0.03 0.00
70% KL 30235 117 0.0 0.00

The 12
th
16 -18 June, 2011

240

2. Protein aggregation of KL/FP materials
2.1 Protein solubility
Solubility of protein bioplastics before and after processing is often regarded as a good
indicator of cross-link formation during processing (Verbeek and Van Den Berg 2010). In
order to study the effect of KL on protein aggregation, changes in solubility of protein in
SDS phosphate buffer were investigated. In this testing, protein was dissolved in the SDS
buffer. This protein fraction is called SDS-soluble protein. Then, DTE as a reducing agent
was added to the remaining protein in order to cleavage disulfide bonds. This SDS-
insoluble protein was then dissolved in SDS-buffer.
Protein content of SDS-soluble and SDS-insoluble fraction of KL/FP molded at 100C is
shown in Table 2. After mixing and compression molding process, SDS soluble fraction of
FP powder decreased dramatically from 64 to 18% compared to FP-based material with
0% KL. This was attributed to heat treatment induced cross-linking during the molding
process that led to an increase in covalent cross linking of protein by disulfide bonds,
hydrogen bonds and hydrophobic interactions (Pommet and others 2005; Kunanopparat
and others 2009). Total protein recovery was less than 100% in all samples. This may be
explained by the original high protein of FP powder and the high processing temperature.
The addition of KL resulted in an increase of the SDS-soluble protein and total protein
recovery compared to FP material without KL. These results indicated a decrease of
protein molecular weight of samples, suggesting the effect of KL on protein aggregation.
Table 2 Protein content (%) of SDS-soluble, SDS-insoluble fraction and total protein recovery of FP
powder and KL/FP materials molded at 100C.
Sample SDS-soluble
protein fraction
SDS-insoluble
protein fraction
Total protein
recovery (%)
Fish protein 648 244 88
0% KL 182 83 26
20% KL 211 112 33
40% KL 234 123 35
60% KL 285 93 37

Protein molecular weight
Protein molecular weights of KL/FP materials molded at 100 C were measured using 15%
SDS-PAGE gel. Figure 2 (a) and (b) show respectively molecular weight of SDS-soluble
and SDS-insoluble protein of FP/KL materials. Molecular weight of FP powder ranges
from 18 to 250 kDa in both SDS-soluble and insoluble. After compression molding, FP-
based material presented only the low intensity band at 50-75 kDa for SDS-soluble and at
37-50 kDa for SDS-insoluble. Band of materials with 20-60% KL contents presents the
same trend as materials with 0% KL. Therefore, no significant change of molecular weight
of protein was observed in the studied molecular weight range (10-250 kDa)
when KL was added in the FP-based materials.
The 12
th
16 -18 June, 2011

241

(a) (b)
Figure 2 Protein molecular weight of SDS-soluble (a) SDS-insoluble (b) of FP powder and FP
materials with 0-70%KL contents.

2.3 Free radicals
As KL is known as free radical scavenger, it may interfere with the protein aggregation by
capturing free radicals during processing FP-based material. Therefore, electron spin
resonance (ESR) signals of KL/FP materials were investigated at 77 K. ESR spectra
corresponded to relatively stable free radicals of KL/FP materials after mixing at 80 C and
compression molding at 100 C are shown in Figure 3(a) and (b), respectively. KL/FP
materials after mixing presented ESR spectra same as after compression molding. The
signal intensity of FP-based material (0%KL) indicated the nitrogen or nitroxyl radicals at
g = 2.00 (Saeed and others 1999; Rebello and Schaich 1999). However, sulfur centered
radicals such as thiol and disulfide radicals were either absent or their signal was hidden in
the nitrogen or nitroxyl signal. For KL-based material (70%KL), high intensity is observed
at g = 2.00 corresponded aromatic radicals (Czechowski and others 2004). Therefore, ESR
results indicated that no radical scavenging effect between KL and FP during mixing and
molding process was observed.
The 12
th
16 -18 June, 2011

242

Figure 3 Free radicals of KL/FP materials after mixing at 80C (a) and molding at 100C (b)

In conclusion, KL is an alternative additive to enlarge the protein thermal processing
window. KL can improve the rheological properties of protein by decreasing the blend
viscosity at high shear rate. This might be associated with the low molecular weight of KL
and the effect of KL on protein aggregation
Acknowledgements
The authors are very grateful to office of the National Research Council of Thailand for
financial support in this research.
References
Bengoechea C, Arrachid A, Guerrero A, Hill SE, Mitchell JR. 2007. Relationship between
the glass transition temperature and the melt flow behavior for gluten, casein and
soya. J Cereal Sci 45(3):275-284.
Bourtoom T, Chinnan MS, Jantawat P, Sanguandeekul R. 2009. Recovery and
characterization of proteins precipitated from surimi wash-water. LWT Food Sci
Technol 42(2):599-605.
Czechowski F, Golonka I, Jezierski A. 2004. Organic matter transformation in the
environment investigated by quantitative electron paramagnetic resonance (EPR)
spectroscopy: studies on lignins. Spectrochim Acta A 60(6):1387-1394.
Hernandez-Izquierdo VM, Krochta JM. 2008. Thermoplastic Processing of Proteins for
Film FormationA Review. J Food Sci 73(2):R30-R39.
Jerez A, Partal P, Martnez I, Gallegos C, Guerrero A. 2007. Protein-based bioplastics:
effect of thermo-mechanical processing. Rheol Acta 46(5):711-720.
Kunanopparat T, Menut P, Morel MH, Guilbert S. 2009. Modification of the Wheat Gluten
Network by Kraft Lignin Addition. J Agr Food Chem 57(18):8526-8533.
The 12
th
16 -18 June, 2011

243

Laemmli UK. 1970. Cleavage of Structural Proteins during the Assembly of the Head of
Bacteriophage T4. Nature 227(5259):680-685.
Lodha P, Netravali AN. 2005. Thermal and mechanical properties of environment-friendly
green plastics from stearic acid modified-soy protein isolate. Ind Crop Prod
21(1):49-64.
Orliac O, Silvestre F, Rouilly A, Rigal L. 2003. Rheological Studies, Production, and
Characterization of Injection-Molded Plastics from Sunflower Protein Isolate. Ind
Eng Chem Res 42(8):1674-1680.
Piyadhammaviboon P, Yongsawatdigul J. 2009. Protein cross-linking ability of
sarcoplasmic proteins extracted from threadfin bream. LWT - Food Sci Technol
42(1):37-43.
Pommet M, Redl A, Guilbert S, Morel MH. 2005. Intrinsic influence of various plasticizers
on functional properties and reactivity of wheat gluten thermoplastic materials. J
Cereal Sci 42(1):81-91.
Pommet M, Redl A, Morel MH, Domenek S, Guilbert S. 2003. Thermoplastic processing
of protein-based bioplastics: chemical engineering aspects of mixing, extrusion and
hot molding. Macromol Symp 197(1):207-218.
Pouteau C, Dole P, Cathala B, Averous L, Boquillon N. 2003. Antioxidant properties of
lignin in polypropylene. Polym Degrad Stabil 81(1):9-18.
Rebello CA, Schaich KM. 1999. Extrusion chemistry of wheat our proteins: II.
Sulfhydryl-disulde content and protein structural changes. Cereal Chem
76(5):756763.
Saeed S, Fawthrop SA, Howell NK. 1999. Electron spin resonance (ESR) study on free
radical transfer in fish lipidprotein interaction. J Sci Food Agr 79(13):1809-1816.
Shen Q, Zhang T, Zhu MF. 2008. A comparison of the surface properties of lignin and
sulfonated lignins by FTIR spectroscopy and wicking technique. Colloid Surface A
320(1-3):57-60.
Silva EABd, Zabkova M, Arajo JD, Cateto CA, Barreiro MF, Belgacem MN, Rodrigues
AE. 2009. An integrated process to produce vanillin and lignin-based polyurethanes
from Kraft lignin. Chem Eng Res Des 87(9):1276-1292.
Thielemans W, Can E, Morye SS, Wool RP. 2002. Novel applications of lignin in
composite materials. J Appl Polym Sci 83(2):323-331.
Torres-Giner S, Ocio MJ, Lagaron JM. 2009. Novel antimicrobial ultrathin structures of
zein/chitosan blends obtained by electrospinning. Carbohyd Polym 77(2):261-266.
Ullsten NH, Gllstedt M, Johansson E, Grslund A, Hedenqvist MS. 2006. Enlarged
Processing Window of Plasticized Wheat Gluten Using Salicylic Acid.
Biomacromol 7(3):771-776.
Verbeek CJR, Van Den Berg LE. 2010. Extrusion Processing and Properties of Protein-
Based Thermoplastics. Macromol Mater Eng 295(1):10-21.

The 12
th
16 -18 June, 2011

244

OE-313

Comparative Evaluation of Divalent Cations on Conformational Changes
of Actomyosin Extracted from Fresh and Freeze-Thawed Tilapia

Sornchai Sinsuwan, Prapassorn Lumpongchat, Pornpimol Sungperm, and
Jirawat Yongsawatdigul
*

School of Food Technology, Institute of Agricultural Technology,
Suranaree University of Technology, Nakhon Ratchasima 30000, Thailand
*
Corresponding author: jirawat@sut.ac.th

Abstract

Conformational changes of natural actomyosin prepared from fresh and freeze-thawed
tilapia (Oreochromis niloticus) in the presence of Ca
2+
and Mg
2+
were investigated. There
was no difference in Ca
2+
-ATPase activity of actomyosin extracted from both samples.
Denaturation temperatures (T
d
) of actomyosin extracted from fresh fish were higher than
those from freeze-thawed counterparts. A decrease in T
d
was observed as Ca
2+
or Mg
2+

concentration increased from 0 to 100 mM. Ca
2+
or Mg
2+
enhanced protein aggregation at
40 C. Myosin heavy chain (MHC) and actin (ATN) tended to form large aggregates in
the presence of 100 mM Ca
2+
or Mg
2+
. Both hydrophobic interactions and disulfide
linkages were mainly involved in actomyosin aggregation as evident by a decrease in
surface hydrophobicity (S
o
) and total sulfhydryl groups. Ca
2+
at 20 mM or Mg
2+
at 100
mM significantly increased gel strength of washed tilapia mince when set at 40 or 60 C,
regardless of freeze-thaw cycle applied. Divalent ions, Ca
2+
and Mg
2+
, could be used to
improve textural properties of fish actomyosin extracted from either fresh or frozen fish.

Keywords: Actomyosin, tilapia, calcium, magnesium, conformation

The 12
th
16 -18 June, 2011

245

OG : Innovative Fermented Foods and
Functional Ingredients
The 12
th
16 -18 June, 2011

246

OG-119

Optimization of Pyrazine Production from Newly Isolated
Bacillus sp. by Submerged Culture

Nopmaneerat Leelawiwat
1
, Pinpinut Adiluchayakorn
1
, Patcharakamol Sutasanawichanna
1
,
Suwattana Pruksasri
2
and Arunsri Leejeerajumnean
1,
*

1
Department of Food Technology, Faculty of Engineering and Industrial Technology, Silpakorn University,
Nakorn Pathom, Thailand;
2
Department of Biotechnology, Faculty of Engineering and Industrial
Technology, Silpakorn University, Nakorn Pathom, Thailand
*
Corresponding author: arunsri@su.ac.th, cin_cinderella_cin@hotmail.com

Abstract

Pyrazines, using for savory flavour industry, are normally produced by solid state
fermentation of soybeans but the finished products have the problems with beany flavour.
The aim of the research was to optimize pyrazine production from various C-sources and
initial pH of medium by submerged culture. The highest pyrazine production culture,
Bacillus sp. was screened from 30 cultures isolated from Thua-Nao and natto, Thai and
Japanese alkaline fermented soybeans. Spores suspension of Bacillus sp. was inoculated in
250 ml flasks containing 100 ml of GYP medium and the flasks were put on the shaker 200
rpm at room temperature. The parameters involved in pyrazine production including initial
pH, time-course of fermentation and types of C-source were investigated. The initial pH of
the medium was varied from 6.0-8.5. Glucose, sucrose and fructose were used as C-
sources. Liquid samples were taken every 12 h from 0 to 120 h during the fermentation.
Then the volatile compounds or pyrazines from the liquid samples were absorbed by
CAR/PDMS fiber and analyzed by GC-MS. The result found that pyrazines could be
synthesized more efficiently at the initial pH 8.5 and which giving the total pyrazines
formation of 7,695 mg L
-1
after 72 h of incubation at 30 1C. Among three types of C-
source, fructose gave the highest pyrazine production. So this culture had a very high
potential for pyrazine production.

Keywords: Pyrazine, Bacillus sp., submerged culture, flavour, optimization

Introduction
Pyrazines are heterocyclic, nitrogen-containing molecules which found in a wide variety of
foodstuffs. It is important as flavor components in many kinds of foods especially roasted
fried and grilled food. The flavor characteristics of pyrazines could be generally described
as nutty, roasty and toasty dependent on their forms (Masuda and Mihara 1988).
Tetramethylpyrazine (TTMP) (Fig. 1), one of the common alkylpyrazines, was widely used
in the food industry as flavor ingredient (Seitz 1994).

Several chemical methods of pyrazine synthesis via the Maillard reaction and Strecker
degradation have been established. However, consumers prefer natural products even
though they are generally much more expensive than the corresponding chemical
compounds, for example, natural TTMP is 3,737.50 $/ kg while chemical TTMP is 339.30
$/kg (Sigma, USA 2011)

The 12
th
16 -18 June, 2011

247

In recent decades, numerous methods of pyrazine synthesis have been developed. Pyrazine
production by microbial process from natural raw materials becomes more appropriated for
its bio or nature properties (Schrader 2007). Several application of pyrazine production
was explored in solid state and submerged fermentations by Bacillus sp. (Besson and
others 1997; Larroche and others 1999; Xiao and others 2006), Corynbacterium
glutamicum mutant (Demain and others 1967) and Lactococcus lactis sub sp. lactis biovar.
diacetyllactis FC1 (Kim and others 1994). The highest TTMP production of 7.46 and 7.34
g l
-1
in flask and fermenter experiments by feed strategy which based on the stimulating
effect of ammonium phosphate was obtained by Bacillus subtilis CCTCC M208157 under
the conditions of stable pH, good oxygen supply, a favorable incubation temperature and
rich medium (Zhu and Xu 2010).

Fig. 1 Chemical structure of tetramethylpyrazine (TTMP) (Eugene 1994)

Several pieces of evidences have indicated an important role of medium composition in
supporting TTMP production by the Bacillus sp. strains (Besson and others 1997; Xiao and
others 2006; Zhu and others 2010), and kinds of methods were accordingly developed for
enhanced TTMP production, such as enrichment of the medium with precursor 3-hydroxy-
2-butanone (HB), the employment of soytone and vitamin supplements and accumulation
of precursor HB endogenously from glucose. However, little work has concentrated on the
starter culture isolated from natural sources, such as Thua-Noa and natto for pyrazine
production.

The purpose of this investigation was to screen the starter culture from natural alkaline
fermented soybean products and optimize pyrazine production from various C-sources and
initial pH of medium from the selected starter culture. All of these experiments were done
in flasks.

Chemical
Yeast extract (YE), glucose and fructose were purchased from Merck (USA), peptone was
from Himedia Laboratories (USA), diammonium phosphate (DAP) was from Ajax
Finechem (USA) All others chemicals used in this investigation were analytical grade
available commercially.

Microorganism screening
Natto or Thua-Noa products from many department stores in Bangkok were purchased.
Natto or Thua-Noa samples from each brand were suspended in sterile peptone water
(1:10), then put in stomacher at high speed for 1 min and microorganisms in the suspension
were isolated on plate count agar by pour plate technique. The agar plate were incubated at
35C for 24 h at, then the single colony was streaked onto nutrient agar slant and incubated
at 35C for 24 h. Spores suspension were prepared by streaking on nutrient agar adding
The 12
th
16 -18 June, 2011

248

0.005 g l
-1
of MgSO
4
.7H
2
O and 0.005 g l
-1
of MnSO
4
.4H
2
O which favored for spore
germination. After incubation at 36C for 48 h, 2 ml of sterile water was rinsed on
medium. The spore suspension was collected into sterile test tube and put into 80 C water
bath for 10 min, centrifuged and diluted with 10 ml of sterile water, respectively. The spore
suspension was stored at 5 C for further screening.

Media and shake flask batch culture
The starter culture is used in the form of spore suspension. The screening medium was
GYP which composed of glucose, yeast extract, peptone, and di-ammonium phosphate.
Before autoclaving, the pH of the medium was adjusted to 6.0 6.5 7.0 7.5 8.0 and 8.5.
Glucose and the others ingredients were autoclaved separately at 121 C for 15 min. The
spore suspension from selected culture was inoculated into the fermentation medium (GYP
medium) and incubated with shaking (200 rpm) at room temperature. Then, the samples
were taken from each treatment for analysis. Different culture conditions, such as initial
pH (6.0, 6.5, 7.0, 7.5, 8.0 and 8.5) and carbon sources (glucose fructose and sucrose as
same weight as each others called sucrose (w) and sucrose as same carbon atoms as each
others called sucrose (c)) were tested.

Analytical methods
Cell growth was measured by spectrophotometer at a wavelength of 660 nm and calculated
from the optical density (OD
660
) with a linear correlation. The pH of the fermentation
medium was adjusted to desired pH with 10 mol l
1
of NaOH and 10 mol l
1
of HCl. Each
treatment of medium was sampled every 12 h until fermented time reached to 120 hr. The
pH value of sampling solution was measured by pH meter. The volatile compounds were
directly analyzed using a GC-MS system (Hewlett-Packard HP 6890, USA) equipped with
a wall-coated open tubular (WCOT) fused silica 30 m x 0.32 mm x 0.25 m (thickness) of
chemically bonded polysiloxane low bleed phase (HP-5 MS), split/splitless model CP-
3800 injector. Mode of MS detector was electron impact (EI) mode was generated at 70eV,
mass range 40400 m/z. The volatile compound was collected by headspace sampling,
solid phase micro-extraction with CAR/PDMS fiber (Dajanta and others 2011). The
collected condition was as followed: fiber thickness 75m; maximum temperature 320C;
operating temperature 250-210C; conditioning temperature 300C. The volatile
compounds were trapped with fiber for 40 min at 50C. The operation conditions were as
follows: helium was used as the carrier gas with flow rate 1.8 ml min
-1
; the injector port
splitless mode at 280 C for 3 min, the column oven was kept constant 45 C for 3 min,
and then programmed to 210 C with a temperature increase of 6 C min
-1
. The
temperature maintained at 210 C for 10 and finally, post run at 240C for 5 min. The
volatile compounds were identified by Mass Spectroscopy with National Institute of
Standards and Technology (NIST) database through Saturn mass spectra library search.
The compounds were tentatively identified by comparing their mass spectra with those
contained in the MS data system. The quantity of volatile compound was analyzed by
internal standard method and 2,4,6-trimethyl pyridine was used as the internal standard.
The linear retention index of volatile compounds was calculated with standard alkane C 8-
20.

The 12
th
16 -18 June, 2011

249

The starter culture
The starter culture in this study was a Bacillus sp. isolated from natto or Thua-Noa bought
from supermarkets in Bangkok, Thailand. This newly isolated Bacillus sp., produced the
highest concentration of pyrazines comparing to the others cultures used in this study (data
not shown). Pyrazine production in flask experiments was conducted to search for the
initial pH of medium and carbon source. Organic nitrogen source was found to be
necessary for healthy growth and the accumulation of TTMP (Xiao and others 2006). A
number of complex nitrogenous materials, including YE and peptone were used as
components of medium. The ammonia source was found to be another important
ingredient and diammonium phosphate ((NH
4
)
2
HPO
4
) was added because it showed the
best result for pyrazine production due to its buffering capacity and keeping constant pH
through the production process (Zhu and Xu 2010).

Effects of initial pH on cell growth and TTMP fermentation in flasks
The optimized pH conditions in flask fermentations were tested by adding 1.0% (v/v) of
Bacillus spores (10
7
CFU ml
-1
) in 100 ml medium which contained in a 250-ml Erlenmeyer
flask. The initial pH was adjusted to 6.0, 6.5, 7.0, 7.5, 8.0 or 8.5. The flasks were shaken
(200 rpm) at room temperature. Oxygen supply had a significant effect on the
performance of the fermentation process. The culture with a higher oxygen transfer rate
(rotation speed 200 rpm) produced higher levels of TTMP and precursor HB. Zhu and
others (2010) reported that the production of TTMP and precursor HB was only 0.17 and
6.6 gl
-1
, respectively at a lower oxygen transfer rate (rotation speed 120 rpm). The cell
growth of each pH showed in Fig.2. The results showed that the initial pH had an effect on
cell growth particularly when growing for longer time (Fig.2).

Figure 2 Growth curve of the starter culture at different initial pH

The pH value of medium was decreased because of acidification accompanied bacterial
growth and carbon consumption (Fig. 3). The rate of pH drop was slow because of the
buffering agent contained in the medium. Optimum pH for pyrazine production was 8.5
(Fig. 4) and the pyrazines were produced at 7,695 mgl
-1
. In the batch fermentation, quick
TTMP accumulation began at the middle of the fast-growth phase and deceased when the
growth ended. The main pyrazine was tetramethylpyrazine (Fig. 5) but the acidic volatile
The 12
th
16 -18 June, 2011

250

by-product of the glucose metabolism was acetic acid. Its concentration reached a peak
during the growth phase, but soon declined to a low level in the stationary phase.

Figure 3 pH value of medium through the pyrazine production process at different initial pH

Figure 4 Total pyrazines produced by starter culture initial pH 8.5

The 12
th
16 -18 June, 2011

251

Figure 5 Types of pyrazine found in the medium at initial pH 8.5, and fermentation time of 72 h

The optimum initial pH of medium for cell growth and pyrazine production was 8.5. The
pH of medium was decreased due to aciduric by-products through fermentation process, at
72 h of fermentation process pH dropped to 6.7. Pyrazine was able to produce at this pH,
however, when the initial pH was below 8.5, pyrazine production decreased. The
concentration of TTMP was low in the culture with an initial pH of 6.0-7.0 and increased
with pH towards neutrality. So, the initial medium pH 8.5 was chosen as a suitable medium
in the next experiment.

Effects of carbon source on cell growth and TTMP fermentation in flasks
The carbon source showed little effect on cell growth and pH of medium slowly decreased
through the pyrazine production process (Fig.6). Among all of the carbon sources tested in
this study including glucose fructose and sucrose, the results showed that using different
sugar as carbon source gave different total pyrazine contents (Fig. 7). Fructose was the best
substrate for pyrazine production (168.8 mg l
-1
) while sucrose performed poorly (Fig. 7).
The 12
th
16 -18 June, 2011

252

Figure 6 Average pH values of medium with different carbon sources, glucose, fructose, sucrose (c)
and sucrose (w) in pyrazine production process

Figure 7 Total pyrazine contents produced from medium with different carbon sources

In this study, all experiments were done at room temperature to search for the suitable
initial pH and C- sources for pyrazine production. The temperature was changed by day
and night. This might cause non appropriate for pyrazine production. Zhu and others
(2010) reported that pyrazine fermentation process was strictly temperature dependent
owing to the strong dependence of enzymatic activity and cellular maintenance on
temperature. Increases in the culture temperature from 30 to 37C resulted in significant
enhancements of TTMP and HB production about 2.5-fold and 33% improvement,
respectively. When the fermentation temperature increased above 37C, cell degradation
probably became dominant over the growth process. This would result in a breakdown in
cellular regulatory mechanisms related to metabolism with a resultant decrease in TTMP
production (Zhu and others 2010). So, the fermentation temperature should be constantly
controlled.

The 12
th
16 -18 June, 2011

253

The mechanism of microbial TTMP formation has not yet been fully explored. Although
the hypothesis that pyrazines were synthesized from acetoin and ammonia (Adachi and
others 1964), and has been supported by several studies (Demain and others 1967; Kim
and others 1994; Besson and others 1997; Larroche and others 1999; Xiao and others
2006). The related enzymes which catalyze the reactions in which the precursor HB is
transformed into TTMP have not yet characterized. Pyrazine formation by nonenzyme-
catalyzed reactions of acyloins with ammonia under mild conditions has also been
demonstrated (Rizzi 1988). Future research efforts in this field should be focused on the
supplement nitrogen source to increase pyrazine contents from this starter culture and
upscale to fermentor, which temperature and oxygen supply can be controlled.

In summary, the selected starter culture from natto which produce flavor compounds,
pyrazines. The newly isolated strain was found to be capable of producing a high level of
total pyrazine contents. Preliminary optimization experiments revealed that neutral pH and
a suitable culture temperature were necessary for better precursor HB accumulation
leading to a high production of TTMP. This strain possesses promising characteristics for
utilization in an economically feasible and environmentally acceptable fermentation
process with the potential for industrial applications.

Acknowledgements
The authors gratefully acknowledge the financial support of this work by TRF and Mighty
International co., Ltd. (Thailand). We are also grateful to Department of Food Technology,
Faculty of Engineering and Industrial Technology, Silpakorn University for analytical
equipments.

References
Adachi T, Kamiya H, Kosuge T. 1964. Studies on the metabolic products of Bacillus
subtilis. IV. determination and mechanism of formation of tetramethylpyrazine.
Yakugaku Zasshi 84:545548.
Besson I, Creuly C, Gros JB, Larroche C. 1997. Pyrazine production by Bacillus subtilis in
solid-state fermentation on soybeans. Appl Microbiol Biotechnol 47:489495.
Dajanta K, Apichartsrangkoon A, Chukeatirote E. 2011. Volatile profiles of thua nao, a
Thai fermented soy product. Food Chemistry 125:464470.
Demain AL, Jackson M, Trenner NR. 1967. Thiamine-dependent accumulation of
tetramethylpyrazine accompanying a mutation in isoleucine-valine pathway.J
Bacteriol 94:323326.
Eugene WS. 1994. Fermentation Production of Pyrazines and Terpenoids For Flavors and
Fragrances. In: Gabelman A, Editor. Bioprocess Production of Flavor, Fragrance,
and Color Ingredients. New York: John Wiley and Sons. pp. 95-104.
Kim KS, Lee HJ, Shon DH, Chung DK. 1994. Optimum conditions for the production of
tetramethylpyrazine flavor compound by aerobic fed-batch culture of Lactococcus
lactis subsp. lactis biovar. diacetylactis FC1. J Microbiol Biotechnol 4:327332.
Larroche C, Besson I, Gros JB. 1999. High pyrazine production by Bacillus subtilis in
solid substrate fermentation on ground soybeans. Process Biochem 34:667674.
Masuda H, Mihara S. 1988. Olfactive properties of alkylpyrazines and 3-substituted 2-
alkylpyrazines. J Agric Food Chem 36:5847.
Rizzi GP. 1988. The biogenesis of food-related pyrazines. Food reviews. International
4:375-400.
The 12
th
16 -18 June, 2011

254

Schrader J. 2007. Flavours and fragrances: chemistry, bioprocessing and sustainability. In:
Berger RG (ed). Springer, Berlin.
Seitz EW. 1994. Bioprocess production of flavor, fragrance, and color ingredients. In:
Gabelman A (ed). Wiley, New York.
Xiao ZJ, Xie NZ, Liu PH, Hua DL, Xu P. 2006. Tetramethylpyrazine production from
glucose by a newly isolated Bacillus mutant. Appl Microbiol Biotechnol 73:512
518.
Zhu B-F, Xu Y. 2010. A feeding strategy for tetramethylpyrazine production by Bacillus
subtilis based on the stimulating effect of ammonium phosphate. Bioprocess
Biosyst Eng 33:953959.
Zhu BF, Xu Y, Fan WL. 2010. High-yield fermentative preparation of tetramethylpyrazine
using an endogenous precursor approach by Bacillus sp. J Ind Microbiol Biotechnol
37:179186.

The 12
th
16 -18 June, 2011

255

OG-158

An Innovative Natural Dietary Fibre (Nata) Product Made from Solid
Waste of Pineapple Concentrate Industry using Acetobacter xylinum

Sri Kumalaningsih
*
, Wignyanto, Hendrix Y.S., Wardanu A.D

Departement of Industrial Agricultural Technology, Faculty of Agricultural,
University of Brawijaya, Indonesia
*
Corresponding author: kumalaningsih@yahoo.com

Abstract

A large amount of solid waste occurs during the course of pineapple concentrate industry.
Processing of this solid waste for the production of nata using A. xylinum is considered
beneficial not only reducing extended pollution problem but also producing a natural
dietary fibre. Two stages of experiments were carried out. The first stage aimed to find out
the best pH for each type of solid waste. A nested factorial design with two factors was
used. The first factor is the type of waste (skin, pulp, and mixed skin and pulp) and the
second factor is the pH (4.0; 4.5; 5.0). The second stage of experiment aimed to find out
the combination treatment between initial sucrose concentration (0.5%; 7.5%; and 10 %)
and the ammonium sulphate concentration (0.3%; 0.5%; and 0.7%) for the production of
Nata de Pina. Feasibility study of a small Nata plant design which could be implemented in
small scale industry is carried out. The best substrate is the mixture of skin and pulp and
the initial pH is 4.5. The yield of the product 53%, the thickness is 5 mm; the weight of
nata is 120 g (wet basis). Further study indicated that the best combination treatment is the
addition of sucrose 5% and ammonium sulphate 0.7%, the thickness is 5.67mm, the weight
is 217 g. An increase in the height of substrate by 50 % increased the thickness and the
wet weight as well as the yield. The feasibility study was carried out based on the
assumption on the production capacity of 425 kg/circle or 5728 cup/day. The operational
cost is $ 66,320.75 with the capital investment is $ 21,963.59. The selling price is $ 0.134;
pay back period is 2 years and one month.

Keywords: Nata, pineapple waste, Acetobacter xylinum

Introduction
In Indonesia, pineapple (Ananas comusus (L) Merr.) is the most important horticultural
crops which gained consumer acceptance. According to Central Bereau of Statistic (l996),
pineapple production in East Java in 2004 reached 444.507 tons. About 70% of the total
productions are consumed as fresh product and 30% is processed into various food
products such as jam, jelly, beverage and concentrate, which produced a large amount of
solid waste such as skin, core etc. (Tahir, 2008).

At Pontianak, West Kalimantan Province, pineapple is processed into concentrate puree
with total production capacity of 30 tons/hour producing about 15 19,5 tons solid waste
consisting of skin and also pulp(Sianipar and others 2006) This solid waste has not been
utilized for the production of high economical value product but is used as animal feed of
low price. The use of this solid waste as raw material for the production of nata de pina by
The 12
th
16 -18 June, 2011

256

using Acetobacter xylinum is considered beneficial not only reducing pollution problem
but also producing dietary fibre food

Nata de Pina is microbial cellulose which become as a new type of biopolymer. Previously
nata is produced from coconut water and consumed as dietary food. The manipulation of
intrinsic factors such as pH, sucrose, and nitrogen compound may affect the yield of nata
(Misgiyarta, 2011). Several bacteria such as Acetobacter aceti, Sarcina, Agrobacterium,
Rhizobium, and Acetobacter have shown the ability to synthesize cellulose from waste
material. However Acetobacter xylinum has the best ability to produce cellulose (Tsuchida
and Yoshinaga 1997). Further study reported by Skinner and Cannon (2000) indicated that
Acetobacter xylinum is an aerobic soil that ferment carbohydrate to vinegar. This organism
synthesize and extrude fibril or cellulose to form a mat yielding edible product which so
called nata. Fermentation production and application for the production of nata has been
documented by Chwla and others (2009), and concluded that several factors such as the
type of substrate and other intrinsic factors should be searched prior to implement at
industrial scale. The purpose of this study was to find out the best initial pH for each type
of solid waste and the nutrient such as initial sucrose and ammonium sulphate
concentration for the production of high yield of nata de pina and a feasibility study of the
nata plant design profitable carried out at small scale industry.

Microorganism and culture conditions
Strain of A. xylinum was obtained from Agricultural Technology Laboratory, Tanjung
pura, West Kalimantan Province. The organisms were grown and maintained on nutrient
agar and stored at 5
o
C at Agricultural Technology, division Brawijaya University East
Java. A. xylinum cells were grown in Nutrient broth for 24 hours at 30
o
C, harvested by
centrifugation. The sedimented cells were washed three times in a 0.1 % peptone solution,
resuspended to the desired concentration and used immediately in the experiments.

Research Methods
a) Substrate Preparation
The solid waste consisting of skin and pulp was separated then added with mineral water in
a ratio of 1 : 1 then press and extracted. The extracted juice was then boiled for 10 minutes,
cooled and placed in container ( 21 x 27 x 8 cm) and then covered with clean paper. Each
container was filled with 250 ml extracted juice. The temperature of the extracted juice
was controlled around 30 35
o
C and the pH value was adjusted into pH 4.0; 4.5; 5.0 with
the addition of acetic acid.

Starter Preparation
The cells of A.cylinum were grown in 660 ml medium containing the best substrate
incubated at room temperature (25-30 C) for 7 days.

Extracted Solid Pineapple Waste Preparation:
1. Laboratory Scale Experiment
a. Preparation media from pineapple skin
The pineapple fruit was peeled. The skin was collected and chopped, then blended
and added with mineral water in a ratio of 1 : 1 (v/v)
b. Preparation of mixed substrate
The 12
th
16 -18 June, 2011

257

The skin and the solid waste containing of pulp were blended in a proportion of 1 : 1 (w/w)
and added with mineral water, extracted and use as substrate

The solid pineapple waste was obtained from the pineapple concentrate industry West
Kalimantan Province. This solid waste was extracted with mineral water in a ratio of 1 : 1.

b) Experimental design
First stage experiment
A nested factorial design was used. The type of substrate i.e. extract of skin, extract
of mixed skin and pulp and pulp as the main factor and the second factor was pH
4.0; 4.5; 5.0.
Second stage of experiment
The best treatment obtained from the first experiment was used as substrate for the
second stage experiment. The extract was added with sucrose at three level
concentrations (5%; 7.5%; 10%) and ammonium sulphate at three levels (0.3%;
0.5%; 0.7%). The best result obtained from the first and second stage of experiment
was then being used for making nata de pina plant design which could be
implemented at small scale industry.

c) Scale up experiment
The idea of scaling up is to increase the size of volume and amount of plate used in the
process. The volume of substrate is increased to 10 times, from 200 ml to 2.0 L/plate and
the amount of plate is 9 times. Total production capacity is 10 kg of solid waste and the
volume of extracted substrate become 18 L. The best treatment obtained from the first and
second stage of experiment i.e. mixed solid waste extract, added with 5% sucrose; 0,7%
ammonium salt and the pH is adjusted to pH 4,5; placed in plastic container incubated at
room temperature ( 30
o
C).

d) Feasibility Study
The Production of Nata de Pina at small scale industry.

Production Capacity
The processing of Nata de Pina is designed into two groups. The first group is the
production of whole Nata and the Second group is the processing of whole Nata into
finished product.

Raw Material Handling The Raw Material was obtained from the concentrate of pineapple
industry. After being sortated, the refined waste was weighed and diluted with water in a
ratio of 1 : 1

Extraction About 500 kg diluted waste is then extracted using hydraulic extractor, the
extracted juice is than boiled at 100
0
C for two ( 2 ) hours and then added with 5%
sucrose, and 0,7% Ammoniun sulphate and agitated to homogenize all the ingredient, and
then cooled and the pH is adjusted to pH 4,5 with acetic acid.

Fermentation The extracted solid waste is then inoculated with starter of Acetobacter
xylinum containing 2,0x10
7
cell/ ml placed in plastic container of 31x23x3cm size covered
The 12
th
16 -18 June, 2011

258

with clean paper, incubated at room temperature ( 27 30
0
C ) for 10 ( ten )
days.

Harvesting The Nata de Pina produced is harvested and soaked for 24 hours in clean water
to eliminate the sour taste, and then cut into small pieces for further processing to make
beverage drink.

First stage of experiment
As stated by Anastasia and Afrianto (2008) pineapple fruit of giant variety used in this
experiment has thick skin, about 15 20% of fruit consisting of skin, which still contained
a large amount of fibre, sucrose, glucose, mineral and vitamin, where as the pulp as waste
containing organic acid. These entire valuable compound may support the synthesized of
cellulose. Result obtained from the first experiment is shown in Table 1 and Table 2.

1. Chemicals Characteristic of the Extracted Pineapple Solid Waste.
The Analytical results presented in Table 1 shows that the substrate of Extracted solid
waste made from mixture of skin and pulp contains high crude fibre 15,43%, and valuable
compounds.

Table 1 The chemical compositions of the pineapple solid waste extract.
Composition Value
Crude Fibre 15,43 %
Glucose 3,65 %
Citric Acid 0,61 %
Sucrose 8,87 %
Mineral 0,115 %
Vitamin A 29,0 ( SI )
Vitamin B 0,08 mg/ 100gr

The use of extracted mixed skin and solid waste pulp was providing a promising substrate,
by the adjustment the media at pH 4.5 producing the highest thickness, wet weight and the
yield of nata as shown in Table 2.

Table2. Effect of type of solid waste and pH on the thickness, weight, and yield of nata de pina.

Type of solid waste pH Thickness (mm) Wet Weight (g) Yield (%)
Skin 4,0 0,49 10,00 2,86
4,5 1,97 51,00 14,57
5,0 1,13 29,33 8,38
Skin & pulp 4,0 1,26 35,00 10,00
4,5 2,37 84,67 24,19
5,0 1,00 30,33 8,67
Pulp 4,0 1,35 28,33 8,10
4,5 2,31 61,67 17,62
5,0 1,15 24,33 6,95

Apparently the use of mixed skin and pulp at pH 4.5 cause the media has pronounced and
environmental condition which favored the organism to grow. The formation of acid
during fermentation resulted in lowering the pH level and hence inhibited the wild
The 12
th
16 -18 June, 2011

259

organism to grow. It is therefore this treatment has the highest thickness of and obviously
influences the wet weight and the yield after 10 days of incubation (the yield is in the range
of 2.66 24.19 %).

The formation of nata during incubation is shown in Figure 1. The microbial cellulose
formed in the mixed substrate is higher than that of the substrate which containing waste of
extracted pulp only. Apparently the crude fibre presence in the media has the potential role
to distribute the nutrition during fermentation. On the other hand crude fibre contains
various micro elements such as Fe, Mn, Mg, which may support the activity of enzyme
(Misgiyarta, 2011).

Figure 1 The Formation of Nata during incubation

Second stage experiment
The microbial cellulose produced is proportionate to fermentation time. The organisms
start to synthesize the cellulose after 2 (two days) this evident indicated that the substrate
has high nutrition value, although there is no micro particles as stated by Chng and
Muhammad (2003) (Fig.l). Sucrose and ammonium sulphate concentrations affected the
thickness of Nata. Statistical analysis indicated that there is no significant interaction
between treatments. The results are given in Table 3.

Table 3. Effect of sucrose and ammonium sulphate concentration on the thickness of Nata

Ammonium Sulphate (%)
0,3 0,5 0,7 Sucrose (%)
Thickness
5 4,67 4,9 5,67
7,5 3,50 3,73 3,67
10 1,83 2,17 2,17

From Table 3, it could be seen that sucrose and ammonium salt concentration has affected
individually. The best sucrose concentration is 5%. Increasing the sucrose concentration
caused the media become viscous and inhibited the organisms to grow. Presumably
sucrose concentration up to 5 % could stabilize the existence of carbon source thus the
microbial cellulose synthesize could be sustained, as stated by Mashudi (1993). Further
Days
Thickeness
(mm)
The 12
th
16 -18 June, 2011

260

study to improve the thickness, weight, and yield of nata by increasing the height of
substrate was shown in Table 4.

Table 4 The Effect of height of substrate on the thickness wet, weight and yield of Nata

Height of Media
( cm )
Thickness
( mm )
Weight
( g )
Yield
( % )
1,5 5 190 54,29
3,0 15 270 98,5
4,5 16 275 99,9

The height of media also affected the production of Nata; the maximum height is 3 cm of
the reactor vessel. Presumably the bigger the amount of substrate resulted in the more
valuable compound and dissolved oxygen which support the extended of cellulose
production. Increasing the content of volume of media has no significant effect. These
results were then used for scale up experiment.

Fibre content
The sucrose concentration affected the fibre content but the addition of ammonium
sulphate has no significant effect. From Table 5 above it is found that there is no
significant interaction between sucrose and ammonium salt concentration on the fibre
content of nata de pina. Table 5 shows the fibre content of nata after various treatment and
times.

Table 5 Effect of sucrose and ammonium sulphate concentrations on the fibre content of nata (%).

Ammonium Sulphate (%) Sucrose (%)
0,3 0,5 0,7 Total
5 2,77 3,05 3,55 3,12
7,5 2,37 2,27 2,47 2,37
10 1,35 1,69 2,29 1,78

The highest amount of fibre content is obtained in treatment containing 5% sucrose and
0,7% Ammonium Sulphate, About 3,12% (w/w) of fibre is achieved. This figure indicate
that nata de pina produced in this study is higher than nata de coco, which only contain
2.5% fibre as stated by Lipi (2005).

Scale Up Experiment
Result obtained from the scale up experiment indicated that increase volume of extracted
media from 200 ml to 2 L or total volume of media is 18 L resulted in an increase in the
yield of Nata. This may be due to the height of the media that used is larger than media at
laboratory scale. Apparently the bigger or larger the fermentation container increase the
amount of dissolved oxygen, As stated by Tantration and others ( 2005 ), the dissolved
oxygen presence in the media may increase the glucomic acid of the media which support
the organism to produce Nata.

During fermentation, the extrinsic and intrinsic factors play important roles. According
Mashudi (1993), the medium used to grow microorganisms must be able to transform and
meet the needs of carbon compounds and energy, as an element of macro-nutrients such as
P, K and Mg, micronutrients and growth factors such as vitamins and amino acids.
The 12
th
16 -18 June, 2011

261

Minerals and vitamins needed in the media establishment as micro-elements of nata
materials required but in small amounts. Minerals and vitamins will be used by the bacteria
in it is life. Besides, the medium components that must comply with the requirements of
microbes, before being used should be in a state of sterile medium, not overgrown by other
microorganisms that are not desired.

From the scale up experiment the Nata de Pina produced is 10.6 kg and the thickness is 7
11 mm, the weight is 487.560 gram, and the yield is 64% which higher than that obtained
from the laboratory experiment which is only 54%.

Financial Analysis
Production Process The production of nata de pina is designed as a continuous process
which divided into two part of processing. The first is the production of nata de pina at
large scale prior to package and the second part is the production of nata in plastic cups
containing 100 ml of nata/cup. Production capacity is 462.51 kg nata/cycle or 5728
cups/day, or the actual capacity is 425 kg/ cycle.

Capital Investment
Total = USD 21,722
Operational Cost = USD 3,259
Fixed Capital = USD 18,464
Feasibility indicators Feasibility indicators of nata de pina industry indicated by the value
of IRR is 56.80%, NPV of USD 64,382; 3.7 profitability index, payback period of 2 years
1 month; breakeven point at the production level of 33.39% and net B / C 1.39. These
results show this project is feasible.

Conclusion
Skin waste can be used as growth media for the production of nata de pina. The best
substrate is the mixture of skins and pulp and the initial pH 4.5. The yield of the product
64%, the thickness is 12 mm. Analysis result of the industry shows this project is feasible.

References
Anastasia and Afrianto.2008. Nata de Sewed Quality in Different Citrus Addification.
Proceeding of National Science and Technology II 2008. University of Lampung.
Lampung
Biro Pusat Statistik, 2007, Survei Pertanian Produksi Buah-buahan di Jawa, BPS, Jakarta-
Indonesia.
Chng and Muhammad.2003. Evaluation and Optimization of Microbial Cellulose (nata)
Production Using Pineapple Waste as Substrate. Chemical Engineering. UTM.
Malaysia.
Chawla, P.R. Ishwar B.B, Shrikant A. S and R.S. Sighhal. 2009. Microbial Cellulose:
Fermentative Production and Applications. Journal of Food Technology,
Biotechnology . India.
Manoi, F.2007. Penambahan Ekstrak Ampas Nenas sebagai medium campuran pada
pembuatan nata de cashew. Balai Penelitian Tanaman Obat dan Aromatik.
Mashudi. 1993. Mempelajari pengaruh penambahan sumber nitrogen dengan berbagai
konsentrasi pada pembuatan nata de coco. Skripsi jurusan teknologi pangan dan
gizi, Fateta, IPB.Bogor
The 12
th
16 -18 June, 2011

262

Misgiyarta, 2011. Teknologi Pembuatan Nata de Coco, Balai Besar Penelitian dan
Pengembangan Pascapanen Pertanian.
http://pascapanen.litbang.deptan.go.id/media/berita/misgiyarta-natadeCoco.pdf.
Acces date: March 10
th
2011.
Pambayun. R. 2002. Teknologi Pengolahan nata de coco. Kanisius. Jogjakarta
Tahir, Sri Sumarsih and Shinta Dwi Astuti. 2008. Kajian penggunaan limbah buah nenas
local (ananas comosus, l) sebagai bahan baku pembuatan nata. Makalah Seminar
Nasional Kimia XVIII, Jurusan Kimia FMIPA UGM Yogyakarta.
Tantration ,S., Tammarate,W. Krusong, Bhattarakosol, Phunsri. 2005. Effect of dissolved
oxygen on cellulose production by Accetobacter sp. Journal Sci. Res. Chula Univ.
30. Pp. 179-186
Tsuchida, T and Yoshinaga, F. 1997. Production of Bacterial Cellulose by Agitation
Culture System. Pure & Appl, Chem., Vol.69, No. 11, pp.2453-2458

The 12
th
16 -18 June, 2011

263

OG-219

Fate of Yoghurt Bacteria in Functional Ice Cream in the Presence of
Soy Extract Powder as Prebiotic

Hermanto
1
, Masdiana Padaga
2,*

1
Faculty of Animal Husbandry, Brawijaya University, Indonesia;
2
Veterinary Science Program, Brawijaya University, Indonesia
*
Corresponding author: mpadaga@gmail.com

Abstract

Ice cream is one of much popular dairy-based processed products. Functional ice cream is
an innovative ice cream products made by adding a source of probiotics and prebiotics into
the ice cream mix (ICM) to improve the nutritional value and health benefit of the product.
In this study, Soy Extract Powder (SEP) as prebiotic was incorporated in elevated
concentrations (0, 4, 6, and 8%) into ice cream prepared from yoghurt and low fat ice
cream mix (ICM) to produce functional ice cream. The effect of SEP substitutes on the
growth of yoghurt bacteria as well as on nutritional value of the ice cream was evaluated.
The results showed that yoghurt bacteria could grow in ice cream mix before incubation
and increased in numbers after incubation. SEP was not able to protect the bacteria cell
during the process of aging, agitation and hardening, which illustrated from the bacterial
population. There was no significant difference effect (P>0.05) on the total LAB after
ICM incubation, aging, after agitation and after hardening of the yoghurt ice cream.
However, SEP as prebiotic could promote the growth of yoghurt bacteria in the frozen
product. The functional ice cream contained high numbers of probiotic bacteria (8.30 log
cfu/ml) in comparison to the probiotic ice cream with the standard formula without the
addition of SEP that harboured lower counts of bacteria (7.5 log cfu/ml). The best quality
functional ice cream contained 8.8% fat, 38.2 mg lysine, 6.3 mg methionine, 5.1 mg
cystine, 3.14 % fibre and 8.30 log cfu/mL of probiotic bacteria, which produced by
subtitution of 8% SEP.

Keywords: Soy extract powder, functional ice cream, probiotic bacteria, prebiotic

Introduction
In recent years consumers have directed their demand for foods toward functional product
since this associated with health benefit. The benefits of consuming foods containing
probiotic cultures and prebiotic ingredients have been demonstrated in many scientific
fields (Sanders 2009, Sanders and Klaenhammer 2001). Functional ice cream is the
development of dairy products made by adding a source of
probiotics and prebiotics into ice cream mix (ICM) in order to gain the functional
properties of the product. The health benefits of probiotic dairy products have been well
documented (Granato, Branco, Cruz, Faria and Shah 2010)
Yogurt is a healthy and nutritious food and has many forms including drinkable (liquid) or
solid, low fat or fat free, fruity or cereal flavored, and (Tamime and Robinson, 2000;
McKinley, 2005). Substitution of yoghurt into ice cream mix (ICM) which contain lactic
acid bacteria (LAB) have drawn the interest of food industries because of the functional
properties and the Generally Recognized as Safe (GRAS) status of LAB ( Lara Matia-
Merino 2007)
The 12
th
16 -18 June, 2011

264

Commonly, yoghurt fortified with skim milk powder, whey protein concentrate or isolate ,
and other milk protein-based. Other non dairy protein-based ingredients have gained
acceptance in order to increase total solids in fat-free or low-fat yogurts (Tamime and
Robinson 2000). Utilization of soy ingredients in the food industries have been on the
rise. This is because of the knowledge of the recent health effects and associated with their
high nutritional quality especially with respect to protein and amino acids (Gandhi 2006).
There are few studies regarding the soy fortified dairy based yoghurts and soy protein
isolates fortification of ice cream . Fortification of yoghurt ice cream with this ingredient
will improve the textural quality of the product including firmness, viscosity, and
functional ingredients provide health benefits (Hekmat and McMahon 1997). Lecithin in
the soy ingredient, as emulsifiers, can increase the viscosity and stability of ice cream,
refined texture and extend the melting time (Belitz and Grosch, 1999, Samato and others
2007). Soy ingredient also contains considerable high content of oligosaccharide that is a
non digestible fiber which facilitates the growth of LAB when incorporated into yoghurt
(de Vrese and Schrezenmeir, 2008).

International Dairy Federation (IDF) suggests that a minimum of 10
7
cells/g should be
alive in the product at a time of consumption to be able the bacteria to exert a positive
health effect (Sultana et al 2000). The bacteria number in dairy food should be at a
minimum level of 10
6
CFU/g or the daily intake should be about 10
8
CFU/g in order to
compensate reduction of the bacteria population during the passage through the gut (Shah
2007).

The objective of this study was to incorporate the soy extract powder (SEP) as prebiotic
into yoghurt ice cream formula to produce functional ice cream and to evaluate the
growth of yoghurt bacteria in the presence of soy ingredient. The physical, chemical, and
sensory properties of the products were also studied.

Ice cream Ingredients and Starter Culture
Ingredients were used: fresh milk obtained from local farmer, soy extract powder (SEP),
diet sugar, CMC, skim milk powder and egg purchased from local market, yoghurt starter
culture obtained from Animal Product Technology Laboratory, Faculty of Animal
Husbandry, Brawijaya University. All dried ingredient were stored at 5C until use. Starter
culture was kept refrigerated and subcultured before used.

Yoghurt Making
Yoghurt was made from fresh whole milk according to the method of Tamime and Robinson
(2000),. The milk was pasteurised at 61-63 C for 30 minutes, cooled down to 43 C and
inoculated with 4% of yoghurt bacteria starter culture. The inoculated milk was then
incubated at 43 C for 4 hours when the pH reached 4,7. After incubation, yogurts were
immediately cooled in an ice water bath and stored at 5C until used for ice cream
ingredient.

Ice cream mix formulation and processing
Ice cream mixes were prepared with and without the addition of soy extract powder (SEP)
as shown in Table 1. The ice cream mix (before incorporating yoghurt) consisted of
CMC 0,2%, egg yolk 1,5%, sucrose 15%, skim milk , and SEP (0%, 4%, 6 %, 8% w/w)
The 12
th
16 -18 June, 2011

265

which produce ICM with 38% dry matter as suggested by Marshall et al (2003) . The
ingredients were reconstituted at 50C and homogenized using hand mixture for 15 min.
The mix was then pasteurized at 86C for 15 s, cooled to 45C. The yoghurt were mixed
and gently stirred into the ICM at a weight ratio of 50% : 50% and incubated at the same
temperature for 4 hour. The yoghurt ICM was then aged at the temperature 4C for 12 h.
The ice cream mix was then placed into the ice cream maker and agitated for 30 min. The
ice cream was filled into container and frozen for 24 hour.

Tabel 1 Ice Cream Mix (ICM) composition with SEP treatment
Composition SEP treatment Ingredient
0% (gram) 4%

(gram) 6%

(gram) 8% (gram)
Fresh milk
Skim milk powder
Egg yolk
CMC
Sucrose
SEP
68,83
15,47
1,50
0,20
14,00
------
68,69
11,61
1,50
0,20
14,00
4,00
68,62
9,68
1,50
0,20
14,00
6,00
68,55
7,75
1,50
0,20
14,00
8,00
Total 100 100 100 100

Analytical methods
Chemical analysis
Total solids of the ICM and fat content of the ice cream were determined as described by
AOAC (2000). pH values were measured using a pH-meter at 20 C. The ice cream
qualities including Viscosity determined using the method of Moeerfard and Teharani
(2008), meltdown according to Nelson and Trout (1965), Overrun as described by Muse
and Hartel (2004). The best quality ice cream was analyzed for amino acid profile using
HPLC (AOAC, 2000) and fiber (Anonymous, 2000).

Microbiological analysis
Lactic acid bacterial (LAB) count during the production of ice cream was determined
according to Vanderzant and Splittstoesser (1992) using MRS agar for ICM before
fermentation, ICM after fermentation, after aging, agitation and ice cream after hardening.

Data analysis
The LAB count, viscosity, meltdown, overrun and fat content data were tested by analysis
of variance (ANOVA)

Results
Growth of yoghurt bacteria as affected by Soy Extract Powder (SEP)
The viable cells of Lactic acid bacteria (LAB) were counted in every stage of yoghurt ice
cream production. All samples, supplemented or not with SEP, exhibited counts between
10
7
and 10
8
cfu/gram, which is a necessary condition for a fermented product to be
considered as probiotic. Incorporation of SEP into yoghurt ice cream gives no significant
effect (P>0.05) on total count of LAB in the ICM before fermentation, after fermentation,
after aging, agitation and hardening of the yoghurt ice cream (Table 2). However, there
was an increase in number of LAB in the product of ice cream (Fig 1).

The 12
th
16 -18 June, 2011

266

Tabel 2 Effect of Soy Extract Powder (SEP) on the Growth of Yoghurt Bacteria in Functional Ice
Cream

Total Numbers of yoghurt bacteria (log cfu/ml) SEP
TREATMENT
(%)
ICM
(before
fermentation)
ICM
(after
fermentation)
Aging Agitation Hardening
0% 7,66
a
8,28
b
7,78
c
7,42
d
7,56
e

4% 7,53
a
8,17
b
8,42
c
7,66
d
7,73
e

6% 7,63
a
9,02
b
8,34
c
7,98
d
8,07
e

8% 7,72
a
9,41
b
9,19
c
8,45
d
8,30
e

Note: Superscript represents statistically no significant differences (P> 0.05) in the same column
respectively according to Duncans Multiple Range test.

Figure 1 Total Numbers of yoghurt bacteria ( log cfu/g) during Ice Cream Production
1. ICM (before fermentation), 2. ICM (after fermentation), 3. Aging,
4. Agitation, 5. Hardening

Characteristics of Functional Ice Cream supplemented with SEP
Supplementation of ICM with SEP gave a significant difference effect (P<0.05)
on viscosity, meltdown, overrun and fat content of the ice cream (Table 3). SEP could
improve qualities of the functional ice cream, as compare to the ice cream prepared from
basic ingredient (without SEP). The ice cream with the highest amount of SEP has the
lowest fat content which is appeal to consumers.

Tabel 3 Characteristics of Functional Ice cream as affected by Soy Extract Powder

Characteristics SEP
Treatment (%) Viscosity
(Poise)
Meltdown
(minute/50 gram)
Overrun
(%)
Fat content
(%)
0% 7,45
a
57,35
a
33,78
ab
10,31
b

4% 7,59
ab
63,74
b
30,50
a
9,78
ab

6% 8,01
c
66,80
c
38,74
bc
9,26
a

8% 7,96
bc
65,94
bc
42,49
c
8,83
a

Superscript a, b d and c in the same column indicate significance different (P<0,05)

The 12
th
16 -18 June, 2011

267

Amino acid profile and quality of the Functional ice cream
The best quality functional ice cream obtained from the ICM supplemented with 8% SEP
(Table 4). When this product evaluated for the amino acid profile, it is proved that SEP
increased lysine and methionine content of the product. As well as, fiber content was
higher in the ice cream supplemented with SEP. This indicated that SEP improved
functional properties of the ice cream.

Table 4. The best quality ice cream supplemented with 8% SEP as compared to control

Quality Control
(without SEP)
Ice cream
+ 8% SEP
Viscosity (Poise) 7,45 7,96
Meltdown (minute/50 gram) 57,35 65,94
Overrun (%) 33,78 42,49
Fat content (%) 10,31 8,83
Fiber (%) 2,62 3,14
Lysine (mg/100g) 34,0 38,2
Methionine 5,6 6,3
LAB (cfu/g) 7,56 8,30

Discussion
Survival of yoghurt bacteria in the food matrix depends on several factors, such as
temperature, pH, oxygen , inhibitor and presence of competing non starter microorganisms
(Granato, Branco, Nazzaro, Cruz and Faris 2010). Even though statistically SEP did not
give significant effect, Table 2 and Figure 1 show that there was a marked increase in
numbers of yoghurt bacteria during fermentation of ICM at 45C. These conditions are
optimal conditions for growth of lactic acid bacteria (Isleten and Yuceer, 2006). All ICM
and ice cream supplemented or not with SEP had viable cell counts between 78 log cfu/g.
The ice cream without SEP had the lowest population of yoghurt bacteria. This could be
attributed to the presence of SEP that contains oligosaccharides such as starchyose and
raffinose which may stimulate the growth of yoghurt bacteria during the production
process (El-Sayed and others 2002). It may also SEP act as a cryoprotectant that protects
the bacteria during exposing to low temperature (Downey 2003). At the end of the
process, during hardening the bacterial population decreased due to freezing temperature ,
however, functional ice cream supplemented with 8% SEP contain more than 8 log cfu/g
yoghurt bacteria. This is sufficient for a product to ensure colonization of the host intestine
and the ensuring health benefits or the product considered as probiotic (Sultana et al 2000).

With regard to physical and chemical qualities of the ice cream, Table 3 shows that the
product posses a good structures. The addition of SEP may increase total solid contents
that affect the viscosity. Moreover SEP facilitate the growth of yoghurt bacteria resulted in
coagulation of casein that affect the viscosity and may be attributed to the release of
extracellular polysaccharides by yoghurt bacteria which give a thick body and high
viscosity. SEP also contained lecithin. According Moeerfard and Tehrani (2008) the use
of lecithin in the ice cream can increase the viscosity.

Meltdown of the ice cream ranged from 57.35 to 66.80%. The results showed that the use
of SEP increase viscosity in the ICM that may affect meltdown of the ice cream at room
temperature. Addition of SEP also increases carbohydrate or solid materials that increase
the viscosity of the Ice cream that eventually affect meltdown (Muse and Hartel, 2004).
The 12
th
16 -18 June, 2011

268

Meltdown of ice cream generally influenced by ingredients added to ICM such as
stabilizer, emulsifier, mineral, sugar, as well as total solid of the end product,
processing and storage conditions (Isleten and Yuceer 2006) . In addition, SEP also contain
fat that affect meltdown of the ice cream (Koxholt and Eisenmann and Hinrichs 2001)
The functional ice cream had overrun ranged between 33.78-42.49%. Supplementation of
SEP to ICM improved the ice cream overrun . This due to protein in the SEP form a good
fat protein emulsion that affect air trapping in the emulsion and influenced overrun of the
ice cream The ice cream with a good overrun can reduce the formation of large
ice crystal sizes in ice cream which will affect the texture, meltdown and mouthfeel of the
product (Marshall et al 2003)

Lecithin in the SEP also acts as an emulsifying agent (Samato, Maenbuchi, Miyazaki,
Kohno, as protect the membrane proteins from damaged during processing
(Ismunandar 2006). Hirotsuka and Kito 2007) that capable
of forming a good fat structure and air distribution in the ice cream thereby extending
the time of melting of the ice cream, as well

Soy extract powder significantly reduced fat content of ice cream which is improved the
functional properties of the product. The fat content in regular and super premium
ice cream products is between at least 5% and a maximum of 20% (El Owni and Khater
2009) . Fat content in the ice cream have a very important role in maintaining
the stabilization of the ice cream structure, especially during storage (Alvarez and
others 2005).

Addition of SEP increased lysine and methionine content in the ice cream
from 34.0 mg to 38.2 mg and from 5.6 mg to 6.3 mg, respectively (Table4). SEP is a good
source of amino acid of plant origin (Dudek 2001) . As a functional ice cream, the product
also contained 3% food fiber that was higher as compared to the control. This indicated
that supplementation of SEP could improve fiber content of the ice cream. The fiber will
also facilitate the growth of yoghurt bacteria in the ice cream that may maintain the
functional properties of the ice cream as probiotic ice cream. In conclusion,
the ice cream in this study can be categorized as functional ice cream because it
contains probiotic bacteria and prebiotic from SEP. The growth of yoghurt bacteria in the
ice cream was facilitated by the presence of SEP.

References
Alvarez V.B., C. L. Wolters, Y. Vodovotz and T. Ji. 2005. Physical Properties of Ice
Cream Containing Milk Protein Concentrates. J Dairy Sci. 88: 862-871.
Anonymous. 2000. Official Methods of Analysis of AOAC International. 17
th
Ed. AOAC
International. Maryland. USA.
Beal, C., F. Fonseca, and G. Corrieu. 2000. Resistence to Freezing and Frozen Storage of
Streptococcus thermophillus is Realted to Membrane Fatty Acid Composition. J.
Dairy Sci. 84:2347-2356.
Downey, G. 2003. Effects of Cryoprotectant Mixtures on Physical Properties of Frozen
and Thawed Pureed Cooked Potatoes: Some Introductory Studies.
Dudek, S. G. 2001. Nutrition Essentials for Nursing Practice (4th ed.). Philadelphia.
Lippincott.
Gibson, G.R. and C.M. Williams. 2000. Functional Foods. CRC Press. England.
The 12
th
16 -18 June, 2011

269

Hawab, H.M. 2004. Pengantar Biokimia. Edisi Revisi. Bayumedia Publishing. Malang.
Hekmat, S and D.J. McMahon. 1992. Survival of Lactobacillus acidophilus and
Bifidobacterium bifidum in Ice Cream for Use as a probiotic Food. J. Dairy Sci.
75:1415-1422.
Isleten, M. and Y.G. Yuceer. 2006. Effects of Dried Dairy Ingredients on Physical and
Sensory Properties of Nonfat Yogurt. J. Dairy Sci. Vol. 89 No. 8, 2006.
Koxholt, M.M.R., B. Eisenmann and J. Hinrichs. 2001. Effect of the fat globule sizes on
the meltdown of ice cream. J. Dairy Sci., 84: 317.
Marshall, R.T and W.S Arbuckle. 1996. Ice Cream. Fifth edition. Chapman and Hall. New
York.
Moeerfard, M. and M.M. Teharani. 2008. Effect of some Stabilizer on the Physcochemical
and Sensory Properties of Ice Cream Type Frozen Yogurt. American-Eurasian J.
Agric.& Environ. Sci.,4(5): 584-589.
Muse, M.R. and R.W. Hartel. 2004. Ice Cream Structural Elements that Affect Melting
Rate and Hardness.J.Dairy Sci.,87:1-10.
Patel, M.R, R.J Baer and M.R. Acharya. 2006. Increasing Protein Content of Ice Cream.
American Dairy Science Association. J. Dairy Sci. 89: 1400-1406.
Samato, M.,M. Maenbuchi,.C. Miyazaki,.M. Kohno.,M. hirotsuka, and M. Kito. 2007.
Abundant Protein Associated with Lechitin in Soy Protein Isolate. Food
Chemistry. 102: 317-322.
Shah, N.P. 2002. Functional Foods From Probiotics and Prebiotics. J. Food Tech. 55 (11):
46-52.
Tamime, A.Y., and R.K. Robinson. 2000. Yoghurt Science and Technology. CRC Press.
Oxford, New york, Toronto.
Vanderzant, C. and D.F. Spittstoesser. 1992. Compendium of Methods for the
Microbiological Examination of Foods. 3
th
Edition. EPS Group Inc. USA.
Yukuchi, H., T. Goto and S. Okonogi. 1992. The Nutritional and Physiological Value of Fermented
Milk and Lactic Milk Drinks. Elsevier Applied Science, Chapter 10: p 226-227. England
The 12
th
16 -18 June, 2011

270

OH : Novel Food Processing & Packaging
The 12
th
16 -18 June, 2011

271

OH-79

Stability of Green Tea Catechins during Steamed Bread Processing

Victoria K. Ananingsih, Ethel S.P. Chng, Weibiao Zhou
*

Food Science and Technology Programme, c/o Department of Chemistry,
National University of Singapore, 3 Science Drive 3, Singapore 117543
*
Corresponding author: Telephone: +65 6516 3501; Email: chmzwb@nus.edu.sg

Abstract

Addition of green tea extract (GTE) could enhance nutritional quality of steamed bread and
render it a functional food. It is well known that tea catechins are vulnerable to degradation
and subject to epimerization, in which both temperature and food matrix play important
roles. It was also previously found that (-)-epigallocatechin gallate (EGCG) underwent
thermal degradation and epimerization simultaneously during bread baking. However,
stability issues of tea catechins in a high humidity system such as dough during steaming
have not been addressed. This research studied the stability of the two most abundant tea
catechins, namely EGCG and (-)-epicatechin gallate (ECG), during steamed bread
processing. The temperature of bread skin increased towards the steaming temperature (i.e.
100
o
C) faster than that of bread crumb. During steaming, the moisture contents of the skin
and the crumb increased, following first-order dynamics. It was found that the degradation
and epimerization of catechins in both the skin and the crumb followed first-order reaction
kinetics and their rate constants were in accordance with the Arrhenius equation.
Furthermore, the results indicated that EGCG and ECG underwent thermal degradation and
epimerization simultaneously during the steamed bread processing.

Keywords: Green tea; catechins; steamed bread; stability; EGCG; ECG

1. Introduction
Addition of green tea extract (GTE) would make steamed bread a functional food that
provides health benefits beyond basic nutrition. Green tea is produced from leaves of
Camellia sinensis and contains polyphenolic compounds among which the dominated ones
are known as tea catechins. There are four main green tea catechins, namely (-)-
epigallocatechin gallate (EGCG), (-)-epicatechin gallate (ECG), (-)-epigallocatechin
(EGC), and (-)-epicatechin (EC) (Wang and Helliwell 2000). It is believed that tea
polyphenols can act as protective agents against cardiovascular diseases and cancers
(Hodgson and Crofta 2010; Linka and others 2010). The tea polyphenolic compounds also
work as antioxidants and anti-inflammatory agents (Iriti and others 2010). Scientific
evidences of potential beneficial effects from tea chemical constituents are emerging and
would make the consumption of green tea and tea derived products impactful on health
promotion among individuals and the public.

Tea catechins are subject to thermal degradation and epimerization during processing.
Epimerization was found to be a change of the chemical structures of tea catechins from
2R,3R (2,3-cis, epi- form) to 2S,3R (2,3-trans, nonepi- form). The major change appeared
to be epimerization from epistrucutred catechins to the nonepistructured ones (Chen and
Chan 1996; Wang and Helliwell 2000).

The 12
th
16 -18 June, 2011

272

In a study by Wang and others (2006) on the kinetics of the thermal stability of tea
catechins, EGCG and ECG were taken as markers. Their respective epimers are (-)-
gallocatechin gallate (GCG) and (-)-catechin gallate (CG). Comparing EGCG and ECG, as
well as GCG and CG, their chemical structures differ in the presence of a hydroxyl group
at the C-5 position of the B-rings of EGCG and GCG. The reversible epimerization
reactions between the pair catechins of EGCG & GCG and ECG & CG are illustrated in
Figures 1 and 2.

Figure 1 Epimerization between EGCG (on the left) and GCG (right)

Figure 2 Epimerization between ECG (on the left) and CG (right)

Wang and others (2008a) established mathematical models for the stability of EGCG in
aqueous systems. It was also found that during bread baking EGCG underwent thermal
degradation and epimerization simultaneously, which all followed first-order reaction
kinetics and their rate constants followed the Arrhenius equation (Wang and others,
2008b).

At present, there are scientific reports in the literature that describe the stability of green
tea catechins in assorted products such as canned and bottled tea drinks and baked bread
(Bazineta and others 2010; Wang and others 2008b). However, stability issues of tea
catechins in a high-humidity processing environment with semi-solid food matrix e.g.
wheat-flour dough during steaming, have not been addressed. It is therefore of interest to
examine their stability in steamed bread. The aim of this research was to investigate the
stability of two most abundant tea catechins, namely EGCG and ECG, and to develop a
mathematical model for their stability during steamed bread processing. The ability to
predict their profile in the food product may allow for future optimization of the processing
conditions of the product.

Materials
Steamed bread flour (protein content 7.9%) and instant dry yeast (Saccharomyces
cerevisiae) were purchased from Gim Hin Lee Pte Ltd. (Singapore). Fine sugar and salt
were purchased from Phoon Huat & Co. Pte Ltd. (Singapore). Green tea extract (GTE) was
purchased from Pure Herbal Remedies Pte Ltd. (Singapore), which was made from green
The 12
th
16 -18 June, 2011

273

tea (Camellia sinensis) leaves harvested in Guangxi (China). Methanol (HPLC grade) was
purchased from Tedia Company Inc. (USA) and formic acid from Merck & Co., Inc.
(Germany).

Preparation of steamed bread
Steamed bread fortified with 0.50% GTE was prepared. The ingredients of dough included
1 kg flour, 550 ml water, 10 g dry instant yeast, 10 g salt, 10 g sugar and 5.0 g GTE.
Hemispherical dough pieces (50 2 g each) were made and proofed for 30 min at 40C
and 85% RH. They were then steamed for up to 60 min at 100C.

High-Performance Liquid Chromatography (HPLC) analysis
Freeze-dried samples, ca. 1 g were weighed, then 50 ml of an aqueous mixture of 70%
methanol with 29.7% (deionized) water and 0.3% formic acid was added. Extraction was
conducted in a water bath at 70.0C for 45 min. Micro filtration with filter pore size of 0.45
m was performed for the final extract before injection. Tea catechins were analyzed using
a reversed-phase HPLC system. The separation system consisted of a C18 column, a
gradient elution system of methanol/water and formic acid, and a photodiode array
ultraviolet detector. The mobile phase consisted of 0.1% formic acid in methanol (eluent
A) and 0.1% formic acid in water (eluent B). A gradient system was applied as follows: 0-
10 min, 10% B; 10-28 min, linear gradient from 10-30% B; 28-35 min, linear gradient
from 30-45% B; 35-45 min, linear gradient from 45-60% B; 45-50 min, 60% B; 50-55 min,
linear gradient from 60-100% B. The post run time was 5 min. The flow rate was 0.5
ml/min. The sample injection volume was 20 l. Detection was made at 275 nm.

Temperature analysis
Type T thermocouples were placed in the steamer, at dough surface (i.e. skin), as well as in
between dough surface and center (i.e. crumb) to record the temperature profiles at these
positions during steaming.

Moisture content analysis
Briefly, ca. 2 g of each sample type was weighed to a pre-dried and cooled metal dish.
These dishes were placed in an oven at about 100C for overnight. At the end of drying,
the sample was immediately transferred to a desiccator and weighed after reaching room
temperature.

Modeling approach
Matlab 7.10.0 R2010a (The MathWorks, Inc., Massachusetts, U.S.) software package was
used for modeling the moisture profiles. It was also used to model the reaction kinetics
between the pairs of {EGCG, GCG} and {ECG, CG}, thereby obtaining the desired kinetic
parameters. Degradation and epimerization reactions of EGCG and ECG followed first-
order kinetics and the rate constant (k) of the reactions complied with Arrhenius equation
(Wang and others 2006, 2008b). Root-mean-squared error (RMSE) between the
experimental and modeled (i.e. predicted) values was taken as a measure of the model
quality.

The 12
th
16 -18 June, 2011

274

Temperature and moisture profiles during steaming
Figure 3 shows the temperature profile during 60 min of steaming. The initial temperature
was highest for the steamer, followed by the skin, then the crumb, as expected. After about
10 min, the temperature of the crumb was close to the steam temperature.

Figure 3 Temperature profiles during 60 min of steaming

The moisture profiles of the skin and crumb of steamed bread were determined
experimentally as shown in Figures 4.

Figure 4 Moisture profiles during 60 min of steaming

The moisture content of both the skin and crumb increased with increasing steaming time,
which ranged from 40 to 43%. The increase of moisture content could be considered as
small. Furthermore, the increase was greater in the skin than in the crumb. The release of
water vapour by the steamer during steaming created a saturation environment inside the
steamer. Contact of bread with the vapour could lead to moisture adsorption by the bread.
It was mentioned in a study on steamed wheat grains that steam taken up by the grain was
ultimately absorbed as liquid water. The grain moisture content was also shown to vary
exponentially with time (Stapley and others 1999). The surface region could be postulated
to absorb more water vapour than the internal region, thus explaining the greater increase
in the skin than in the crumb.

The 12
th
16 -18 June, 2011

275

While the moisture profiles in Figure 4 were plotted using the experimental data, a first-
order exponential model could be developed to express the relationship between the
moisture content and steaming time, as follows:

where M is moisture content (in %), M
i
is the initial moisture content, M
f
is the final
moisture content, r is a rate constant, and t is time (in min). The model parameters
concerning M
i
, M
f
, and r are listed in Table 1.

Table 1 Model parameters for moisture profiles
Model parameters M
i
, % M
f
, % r R
2 a
RMSE
b

Skin 40.8293 43.4417 0.3627 0.9931 0.0990
Crumb 40.8333 42.0178 0.2538 0.9856 0.0645
a
Coefficient of determination;
b
Root-mean-squared error

The R
2
values indicating the degree of fit were all above 0.98, suggesting that the
relationship between the moisture content and steaming time was indeed exponential. In
view that tea catechins were mainly contained in the aqueous phase of the bread system, a
model for the moisture content in steamed bread is essential for the later modeling of the
stability of catechins, because changes in the moisture profiles would effectively change
the medium environment of the catechins and thus their stability.

Thermal degradation of tea catechins
The concentrations of individual catechins, i.e. [EGCG], [GCG], [ECG], and [CG] were
expressed in mg/kg steamed bread. The retention of total catechins in the skin and crumb,
represented as [EGCG + GCG] and [ECG + CG] vs time, are shown in Figures 5 and 6,
respectively.

Figure 5 Retention of total catechins in the skin (A) {EGCG, GCG} pair; (B) {ECG, CG} pair.

(A) (B)
R
2
= 0.98
RMSE = 7.39
R
2
= 0.96
RMSE = 6.23
The 12
th
16 -18 June, 2011

276

Figure 6 Retention of total catechins in the crumb (A) {EGCG, GCG} pair; (B) {ECG, CG} pair.

Both [EGCG + GCG] and [ECG + CG] in the skin and crumb decreased exponentially
over time. The loss of catechins could be due to thermal degradation. In addition to these
known reactions, their loss could also be due to other possibilities such as interacting with
wheat proteins via hydrogen bonding and scavenging of GS* radicals that were initiated
during SH/SS interchange reactions (Wang and others 2004).

The modeling results had the R
2
values all above 0.92 and the RMSE values ranged from
6.23 to 23.80. Therefore, the modeled values could be considered well in agreement with
the experimental results. The high R
2
values and relatively low RMSE values could further
verify that the degradation of tea catechins followed first-order kinetics and that the
corresponding reaction rate constant k
y
complied with the Arrhenius equation. Assuming
the activation energy Ea for {EGCG, GCG} and {ECG, CG} pairs as 42.78 and 41.58
kJ/mol respectively (Wang and others 2006), the kinetic parameters k
y
and A were
calculated from the models and they are all listed in Table 2.

Table 2 Kinetic parameters for the degradation of tea catechins

Model parameters k
y
E
a
, kJ/mol A
{EGCG, GCG} 0.0023 42.78 2.2616 10
3
Skin
{ECG, CG} 0.0030 41.58 2.0376 10
3

{EGCG, GCG} 0.0021 42.78 2.0325 10
3
Crumb
{ECG, CG} 0.0028 41.58 1.8545 10
3

A in the Arrhenius equation refers to the frequency of collision between molecules in the
proper orientation for reaction to occur. Hence, the stability or reactivity of tea catechins
was said to be structure dependent (Wang and others 2006). From Table 2, it is shown that
A was greater in the pair of {EGCG, GCG} than {ECG, CG}, suggesting that EGCG and
GCG were less stable or more reactive than ECG and CG. In a study on the antioxidative
activity of tea catechins, the order was reported as EGCG being higher than ECG. It was
suggested that the antioxidative activity would increase with increasing number of
hydroxyl groups and when a hydroxyl group existed at the C-5' position (Kumamoto and
Sonda 1998). Hence, an additional hydroxyl group at the C-5 of B-ring of EGCG and
GCG could increase their reactivity, as compared to ECG and CG with only two adjacent

(A)
R
2
= 0.92
RMSE = 23.80
R
2
= 0.95
RMSE = 10.82
(B)
The 12
th
16 -18 June, 2011

277

hydroxyl groups at the C-3 and C-4. Based on structural features of the catechins and
results of A in Table 2, it could be concluded that the stability of EGCG and GCG was
lower than ECG and CG. This would also mean a higher frequency of successful collisions
led to faster loss or lower retention rate of EGCG and GCG.
In comparison, the values of A were higher in the skin than in the crumb. This could be
understood using the results from the moisture analysis, which revealed higher moisture
content in the skin than in the crumb. Firstly, a greater increase of percent moisture could
increase the mobility of catechins, allowing more catechins to react. Higher moisture
content in the skin possibly led to higher frequency of collisions between catechins to form
products and greater reactivity. Secondly, it was suggested that lower moisture content
could confine the collision of molecules associated with degradations (Wang and others
2008b). It follows that higher moisture content could provide more space for higher chance
of successful collisions. Due to higher moisture content, there could be lower stability and
faster loss of tea catechins in the skin.

Epimerization of tea catechins
The stability profiles of catechins in the skin and crumb, represented as [EGCG], [GCG],
[ECG], and [CG] vs time, respectively, are shown in Figures 7 and 8.

Figure 7 Stability profiles of catechins in the skin (A) EGCG; (B) GCG; (C) ECG; (D) CG.

600
700
800
900
1000
1100
1200
0 5 10 15 20 25 30 35 40 45 50 55 60
Min
m
g

E
G
C
G

/

k
g

s
t
e
a
m
e
d

b
r
e
a
d
Experimental
Modeled

140
160
180
200
220
240
260
280
300
0 5 10 15 20 25 30 35 40 45 50 55 60
M in
m
g

G
C
G

/

k
g

s
t
e
a
m
e
d

b
r
e
a
d
Experimental
Modeled

(A) (B)
R
2
= 0.98
RMSE = 10.12
200
250
300
350
400
450
500
0 5 10 15 20 25 30 35 40 45 50 55 60
m
g

E
C
G

/

k
g

s
t
e
a
m
e
d

b
r
e
a
d
Experimental
Modeled
5 0
5 5
6 0
6 5
7 0
7 5
8 0
8 5
9 0
9 5
1 0 0
0 5 1 0 1 5 2 0 2 5 3 0 3 5 4 0 4 5 5 0 5 5 6 0
M in
m
g

C
G

/

k
g

s
t
e
a
m
e
d

b
r
e
a
d
Exp e r ime nta l
M o d e le d

R
2
= 0.97
RMSE = 0.84
R
2
= 0.97
RMSE = 5.68
(C) (D)
R
2
= 0.94
RMSE = 5.30
The 12
th
16 -18 June, 2011

278

Figure 8 Stability profiles of catechins in the crumb (A) EGCG; (B) GCG; (C) ECG; (D) CG.

The actual skin and crumb temperature during the 60 min steaming provided sufficient
energy for epimerization. As [EGCG] and [ECG] decreased, the concentrations of their
epimers i.e. [GCG] and [CG] increased. Both the decrease and increase appeared to be
exponential with time. The results were consistent with the description of a typical
epimerization process in which nonepi-structured counterparts increased while epi-
structured catechins decreased (Wang and others 2006; Sharma and Zhou 2011).

It can be seen in Figures 7 and 8 that the concentrations of GCG and CG increased to a
maximum level at around 20 to 40 min, before decreasing as the steaming continued to 60
min. It could be deduced that for the first 20 to 40 min, epimerization from EGCG to GCG
was predominant, leading to increases of [GCG] and [CG]. With a prolonged steaming to
60 min, thermal degradations could result in decreases of [GCG] and [CG]. The findings
also agreed with the observation of a maximum for the formation of GCG in aqueous
solutions (Wang and others 2006). Overall, the results support the necessity of establishing
kinetic models that take into account simultaneous occurrence of degradation and
epimerization.

As stated earlier, the E
a
values of epimerization of EGCG to GCG and of its reverse
epimerization were assumed to be 117.59 and 84.15 kJ/mol, respectively. On the other
hand, the E
a
values of epimerization of ECG to CG and of its reverse epimerization were
assumed to be 119.25 and 96.22 kJ/mol, respectively (Wang and others 2006). The kinetic
parameters for epimerization of the catechins in the skin are all listed in Table 3. The
modeling results verify that the epimerization of tea catechins in the skin followed first-
600
700
800
900
1000
1100
1200
0 5 10 15 20 25 30 35 40 45 50 55 60
M in
m
g

E
G
C
G

/

k
g

s
t
e
a
m
e
d

b
r
e
a
d
Experimental
Modeled

1 4 0
1 6 0
1 8 0
2 0 0
2 2 0
2 4 0
2 6 0
2 8 0
3 0 0
0 5 1 0 1 5 2 0 2 5 3 0 3 5 4 0 4 5 5 0 5 5 6 0
M in
m
g

G
C
G

/

k
g

s
t
e
a
m
e
d

b
r
e
a
d
Exp e r ime nta l
M o d e le d

(A) (B)
R
2
= 0.98
RMSE = 9.72
R
2
= 0.95
RMSE = 5.62
20 0
25 0
30 0
35 0
40 0
45 0
50 0
0 5 1 0 1 5 2 0 25 30 35 40 45 50 5 5 6 0
M in
m
g

E
C
G

/

k
g

s
t
e
a
m
e
d

b
r
e
a
d
Exp erime nta l
Mod eled

5 0
5 5
6 0
6 5
7 0
7 5
8 0
8 5
9 0
9 5
1 0 0
0 5 1 0 1 5 2 0 2 5 3 0 3 5 4 0 4 5 5 0 5 5 6 0
M i n
m
g

C
G

/

k
g

s
t
e
a
m
e
d

b
r
e
a
d
E xp e r ime nt a l
M o d e le d

R
2
= 0.96
RMSE = 1.11
R
2
= 0.96
RMSE = 5.91
(C) (D)
The 12
th
16 -18 June, 2011

279

order kinetics and that the rate constants of the epimerization complied with the Arrhenius
equation. The assumptions in regard to the E
a
values of the epimerization were valid too.

Table 3. Kinetic parameters for the stability profiles of tea catechins in the skin and crumb

Model parameters k E
a
, kJ/mol A
EGCG 0.0328 117.59 9.8409 10
14

GCG 0.1106 84.15 6.8416 10
10

ECG 0.0216 119.25 1.1059 10
15

Skin
CG 0.0900 96.22 2.7359 10
12

EGCG 0.0280 117.59 8.2323 10
14

GCG 0.0955 84.15 5.8158 10
10

ECG 0.0206 119.25 1.0309 10
15

Crumb
CG 0.0881 96.22 2.6311 10
12

When comparing the epi-structured catechins with the nonepi-structured, EGCG and ECG
had a greater A than GCG and CG, respectively. The results imply that EGCG and ECG
were less stable or more reactive than GCG and CG, respectively. It was suggested that
EGCG and ECG in the 2,3-cis form could have bigger steric hindrance than GCG and CG
in the 2,3-trans form, respectively, therefore exhibiting higher rate of collisions during
thermal reactions (Wang and others 2008b; Guo and others 1999).

The result of A being higher in the skin than in the crumb, was similar to that obtained
earlier for the degradation. Hence, moisture content could well be a factor that affected the
reactivity of the catechins differently in the skin and crumb of the steamed bread. As
explained earlier, higher moisture content in the skin could increase the frequency of
successful collisions, giving a higher A value. As a result, it could be concluded that tea
catechins in the skin were less stable and lost at a faster rate.

Conclusion
Mathematical models for predicting the stability of EGCG and ECG during steamed bread
processing were developed and validated. The kinetic models accounted for not only the
simultaneous degradation and epimerization of the catechins, but also the changes in
temperature and moisture profiles during the steaming. It was found that the degradation
and epimerization of catechins in both the skin and crumb followed first-order reaction
kinetics and the rate constants were in accordance with the Arrhenius equation.

Acknowledgements
The first author is grateful to the financial support from the Directorate General of
Indonesian Higher Education in sponsoring her Ph.D. study. Financial support from
Singapore Ministry of Education through Academic Research Fund Tier 1 grant R-143-
000-404-112 is also acknowledged.

References
Bazineta L, Araya-Fariasa M, Doyena A, Trudelc D, Ttuc B. 2010. Effect of process unit
operations and long-term storage on catechin contents in EGCG-enriched tea drink.
Food Research International 43:1692-1701.
Chen ZY, Chan PT. 1996. Antioxidative activity of green tea catechins in canola oil.
Chemistry and Physics of Lipids 82:163-172.
The 12
th
16 -18 June, 2011

280

Guo Q, Zhao B, Shen S, Hou J, Hu J, Xin W. 1999. ESR study on the structure antioxidant
activity relationship of tea catechins and their epimers. Biochimica et Biophysica
Acta 1427:13-23.
Hodgson JM, Crofta KD. 2010. Tea flavonoids and cardiovascular health. Molecular
Aspects of Medicine 31:495-502.
Iriti M, Vitalini S, Fico G, Faoro F. 2010. Neuroprotective Herbs and Foods from Different
Traditional Medicines and Diets. Molecules 15:3517-3555.
Kumamoto M, Sonda T. 1998. Evaluation of the antioxidative activity of tea by an oxygen
electrode method. Bioscience, Biotechnology and Biochemistry 62:175-177.
Linka A, Balaguera F, Goel A. 2010. Cancer chemoprevention by dietary polyphenols:
Promising role for epigenetics. Biochemical Pharmacology 80:1771-1792.
Sharma A, Zhou W. 2011. A stability study of green tea catechins during the biscuit
making process. Food Chemistry 126:568-573.
Stapley AGF, Landman KA, Please CP, Fryer PJ. 1999. Modelling the steaming of whole
great grains. Chemical Engineering Science 54:965-975.
Wang H, Helliwell K. 2000. Epimerisation of catechins in green tea infusions. Food
Chemistry 70:337-344.
Wang R, Zhou W. 2004. Stability of tea catechins in the breadmaking process. Journal of
Agricultural and Food Chemistry 52:8224-8229.
Wang R, Zhou W, Wen RH. 2006. Kinetic study of the thermal stability of tea catechins in
aqueous system using a microwave reactor. Journal of Agricultural and Food
Chemistry 54: 5924-5932.
Wang R, Zhou W, Jiang X. 2008a. Reaction kinetics of degradation and epimerization of
epigallocatechin gallate (EGCG) in aqueous system over a wide temperature range.
Journal of Agricultural and Food Chemistry 56:2694-2701.
Wang R, Zhou W, Jiang, X. 2008b. Mathematical modeling of the stability of green tea
catechin epigallocatechin gallate (EGCG) during bread baking. Journal of Food
Engineering 87:505-513.

The 12
th
16 -18 June, 2011

281

OH-85

Converting Empirical Tests into Fundamental Tests: A Case Study Using
the Biaxial Bubble Inflation Test

SM Goh

Curtin University of Technology Sarawak, CDT 250, 98009 Miri, Sarawak, Malaysia.
aaron.goh@curtin.edu.my

Abstract

The bubble inflation test is used to test the mechanical performance of a variety of
materials including packaging and foodstuff like bread dough. The test involves blowing a
sample into a bubble until the sample bursts. The measurements of the bubble pressure
and volume are then used as indicators of the material performance. However, these
measurements remain empirical numbers since they are not the fundamental properties of
the materials. The ability to obtain fundamental properties will allow better prediction of
the material performance for example, it would be possible to use numerical simulations
to determine if a dough will provide quality bread, or whether a packaging will fail during
usage. The objective of this study was to develop a method to extract the fundamental
stress-strain properties from the bubble inflation test. The method is based on an inverse
analysis using the Levenberg-Marquardt algorithm and the finite element method. The
method was used to identify non-linear viscoelastic constitutive properties from the bubble
pressurepiston displacement relationship. The inverse method was evaluated using
numerical experiments and it was found that good estimates of the viscoelastic properties
could be obtained using only one set of bubble pressurepiston displacement data.

Keywords: Ciscoelastic, inflation, rheology, inverse, burst

Introduction
The bubble inflation test and its variants have been used to characterise the mechanical
properties of a large range of soft materials including packaging, rubber films, polymer
melts, food materials and tissues (e.g. Bloksma 1957, Kokelaar and others 1996,
Charalambides and others 2001a,b, Feliu-Baez and others 2001). The test is attractive
because it approximates the stress conditions attained in practice for these materials. For
example, the test is popular for characterising the behaviour of dough as it is perceived to
simulate the expansion of the air cells in the dough during the baking process
(Charalambides and others 2001b). The test has also been adopted as a standard test in
monitoring the quality of packaging and seal strengths (Feliu-Baez and others 2001).

In the bubble inflation test, a sample is held at a certain part and then inflated to a bubble
using pressurised air. The pressurised air can be achieved via, for example, a piston
mechanism. Typically, the pressure inside the bubble and the bubble volume or the time
elapsed are measured experimentally. These measurements are then used as an indicator to
compare between different samples.

The greatest drawback of the bubble inflation test is that the measurements are empirical
numbers. This means that the measurements are arbitrarily defined and their values are
dependent on how the tests are conducted. This drawback is especially important for
The 12
th
16 -18 June, 2011

282

materials that are typically subjected to the test such as rubbers or dough. These materials
exhibit complex, non-linear, viscoelastic behaviour which means that their behaviour
depends on the amount of deformation and the time that is taken to deform them. In other
words, the stresses in these materials exhibit both a strain dependent as well as a time
dependent component.

Since the measurements are empirical, the bubble inflation test does not provide
fundamental properties. The ability to extract fundamental rheological properties from the
bubble inflation test will significantly increase the value of the test. Not only can the
empirical numbers be used as in current practice, the fundamental properties can also be
used to predict, based on physical laws, the material behaviour in practice. For example,
the properties of dough can be used to simulate the expansion of gas cells in bread, with
different properties leading to different volumes of bread. In addition, the seal strengths in
a packaging can be ascertained independently of the packaging design.

Attempts to extract the fundamental rheological properties from the bubble inflation test
measurements have been done before. Bloksma (1957) developed an analytical solution
which relates the stress-strain relationship to the pressure-volume measurements. However,
recent studies (Reuge and others 2001, Charalambides and others 2001b) have suggested
that the assumptions regarding the bubble geometry in Bloksmas solution are inaccurate
and that it is necessary to measure the bubble geometry experimentally to calculate the
stress and strain. Some of these geometrical measurements, such as the thickness of the
material at the pole of the bubble, are tedious and very difficult to obtain. In addition, the
analytical solution does not consider the large variation of the strain rate as the bubble
expands (Goh and others 2004). Attempts to calculate analytically packaging pouch
behaviour during a restrained burst test have also not yielded satisfactory results (Feliz-
Baez and others 2001) since the geometries of the test is not easily accounted for in the
theoretical manner.

With the advancement of computer technologies, inverse methodologies have been
increasingly used to identify material properties from non-standard experimental data
where the deformation field is not homogeneous and analytical solutions are not available.
In the inverse method, the stress-strain properties are changed in an iterative manner until
the predicted mechanical response matches the experimental measurements. The inverse
method has been applied to the bubble inflation test technique previously (Rachik and
others 2001). However, in that work, only elastic behaviour was examined and time-
dependent effects were not included.

The aim of the current work was therefore to apply an inverse methodology to obtain strain
and time dependent viscoelastic properties from measurements obtained from the bubble
inflation test. The method was developed and tested via a set of numerical experiment
using the properties of bread dough as a case study.

Finite Element Model
The finite element analyses were performed using the commercial package ABAQUS. The
geometry of the model was based on the actual experiments in a previous study
(Charalambides and others 2001b). The material was modeled to be in contact with a gas
The 12
th
16 -18 June, 2011

283

which was enclosed in a cylinder and driven with a piston at a predetermined speed. The
material was modeled using axisymmetric, eight-noded elements while the gas was
modeled using bi-nodal hydrostatic elements. Both the cylinder and piston were modeled
as rigid bodies. The simulations were performed assuming a piston speed of 25 mm/min, a
total piston displacement of 300 mm and time duration of 12s. From the simulations,
values of bubble pressure and piston displacement were obtained from the ABAQUS code
variables PCAV and U2 respectively. These variables relate to experimental measurements
of pressure in the bubble and piston displacement.

The inverse methodology pursued in this work was based on matching the bubble
pressurepiston displacement predicted from guesses for the material properties to
experimental data. The experimental data here refer to exact finite element
computations of PCAV and U2 using known values of material properties.

Material Model
The stress-strain behaviour of the material was described by:
( ) ( ) ( ) t g t
0
, = (1)
where is the stress at true strain and time t . The strain dependent function, ( )
0
,
has dimensions of stress and was approximated by the Mooney-Rivlin hyperelastic
potential:
( ) ( )
|
\
|
+ =
2 01 10 0
1
2
C C
where
10
C and
01
C are constants and is the stretch ratio. The time dependent function,
( ) t g , is dimensionless and was modeled using the Prony series, which consists of N
number of exponential functions:
( )

=
|
|
\
|
+ =
N
i i
i
t
g g t g
1
exp

where
i
are time constants and
g and
i
g are dimensionless constants. Since the Prony
series contain a large number of unknowns, it was first fitted with a modified Power law
equation using Microsoft Excel before the inverse method was applied. The modified
Power law equation is:
( ) ( )
n
r
t
t
g g t g

|
|
\
|
+ + = 1 1
where n is the power law exponent and
g and
r
t are constants.

Inverse Analysis
The inverse method was used to determine the two strain dependent parameters,
10
C and
01
C , and the three time dependent parameters, n ,
r
t and
g . The Levenberg-Marquardt
algorithm as described in Schnur and Zabaras (1992) was utilized and implemented in a
Python script which interfaced with the ABAQUS software to perform the finite element
simulations and with Microsoft Excel to fit the Prony series to the modified power law
equation at each iteration.

Numerical experiments were conducted to evaluate the feasibility of the inverse analysis.
Finite element computations of the bubble pressure-piston displacement based on known
The 12
th
16 -18 June, 2011

284

parameters were used as the experimental data. These experimental data were computed
using the material parameters from Charalambides and others (2006), with the Prony series
converted into the modified power law equation resulting in constants as shown in Table 1
(input values). These input values were the exact solutions that were to be determined
through the inverse methodology.

All five material parameters,
10
C ,
01
C , n ,
r
t and
g ,were subjected to identification from

the inverse method. Four sets of initial guess values were used as shown in Table 1.

Results
Figure 1 shows how the inverse analysis iteratively led to a converged solution in the
material properties. Initially the initial guesses would give a large discrepancy in the
bubble pressure-piston displacement relationship, but with increasing iterations, the
properties converged to values that did not vary much with further iterations.
For all the four sets of initial guesses, the final predicted bubble pressure-piston
displacement relationships matched the experimental data well (Figure 2) except for Case 3
where the solution did not converge to the experimental data. For Cases 1, 2 and 4, the
predicted bubble pressure-piston displacement relationships were very similar, but the
actual values of the time dependent parameters were different from each other (Table 1).
These values result in slightly different stress-strain relationships as shown in Figure 3.

Table 1 Summary of values for the different cases in the numerical experiments. The units are kPa for
10
C and
01
C and seconds for
r
t . Both n and
g have no units.

Case No. Input values Initi al guesses Final values from i nverse method Iterat
i ons
C
10
C
01
n t
r
g
?
C
10
C
01
n t
r
g
?
C
10
C
01
n t
r
g
?

1 1.214 0.000 -0.468 0.396 0.000 1.0 0.01 -0.500 1.000 0.003 1.214 0.000 -0.596 0.666 0.003 16
2 1.214 0.000 -0.468 0.396 0.000 1.0 0.01 -0.700 0.250 0.010 1.216 0.000 -0.566 0.387 0.084 20
3 1.214 0.000 -0.468 0.396 0.000 1.0 0.01 -0.300 4.000 0.010 1.005 0.001 -0.458 1.680 0.000 20
4 1.214 0.000 -0.468 0.396 0.000 25.0 0.01 -0.300 0.250 0.010 1.215 0.000 -0.554 0.421 0.061 11

0
20
40
60
80
100
120
140
160
180
200
0 100 200 300
p
r
e
s
s
u
r
e
i
n
b
u
b
b
l
e
(
P
a
)
piston displacement (mm)
iteration 2
iteration 5
iteration 20

Figure 1 Intermediate results in the inverse analysis showing the predictions of the bubble pressure-
volume relationship based on the material properties calculated at different iterations for
Case 3.
The 12
th
16 -18 June, 2011

285

0
20
40
60
80
100
120
140
160
180
200
0 100 200 300
p
r
e
s
s
u
r
e
i
n
b
u
b
b
l
e
(
P
a
)
piston displacement (mm)
experiment
Case 1
Case 2
Case 3
Case 4

Figure 2 Final results showing the predictions of the bubble pressure-volume relationship based on
the material properties calculated from the inverse analysis.
0
1
2
3
4
5
6
7
8
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
s
t
r
e
s
s
(
k
P
a
)
strain
input values
Case 1
Case 4
0.25/min
25/min
from inverse method using
bubble pressure-time data

Figure 3 Stress-strain properties based on the final results of the inverse analysis for Case 1 and 4.
Markers represent the input stress-strain properties.

Discussion
The results obtained here suggest that it is possible to apply the inverse method to the
bubble inflation test to obtain fundamental viscoelastic material properties. Only data
describing the pressure in the bubble and the piston displacement were necessary and data
from one piston speed appeared to give sufficient material estimation. However, accurate
identification of the material parameters was found to be dependent on the initial guesses
of the parameters. In addition, the solutions obtained may not necessarily be unique despite
them giving very similar bubble pressure-piston displacement relationships. Further work
is being undertaken to improve the inverse algorithm so that a higher accuracy can be
achieved.

Acknowledgements
Discussion with Dr Maria Charalambides is much appreciated.
The 12
th
16 -18 June, 2011

286

References
Bloksma AH. 1957 A calculation of the shape of the alveograms of some rheological
model substances. Cereal Chemistry 34: 126-136.
Charalambides MN, Wanigasooriya L and Williams JG. 2001a Biaxial deformation of
dough using bubble inflation technique II: numerical modelling. Rheologica Acta
41: 541-548.
Charalambides MN, Wanigasooriya L and Williams JG. 2001b Biaxial deformation of
dough using bubble inflation technique I: experimental. Rheologica Acta 41: 532-
540.
Charalambides MN, Wanigasooriya L, Williams JG, and others. 2006 Large deformation
extensional rheology of bread dough. Rheologica Acta 46: 239-248.
Feliu-Baez R, Lockhart HE, Burgess G. 2001 Correlation of peel and burst tests for
pouches. Packaging Technology and Science 14:63-69
Goh SM, Charalambides MN, Williams JG. 2004 Determination of the constitutive
constants of non-linear viscoleastic materials. Mechanics of Time-Dependent
Materials 8: 255-268.
Kokelaar J.J., van Vliet T. and Prins A. (1996) Strain hardening properties and
extensibility of flour and gluten doughs in relation to breadmaking performance.
Journal of Cereal Science, 24:199-214
Rachik M, Schmidt F, Reuge N and others. 2001 Elastomer biaxial characterisation using
bubble inflation technique. I: Numerical investigation of some constitutive models.
Polymer Engineering and Science 41: 532-541
Reuge N, Schmidt FM, Le Maoult Y and others. 2001 Elastomer biaxial characterisation
using bubble inflation technique. I: Experimental investigations. Polymer
Engineering and Science 41: 522-531
Schnur DS, Zabaras N. 1992 An inverse method for determining elastic material properties
and a material interface. International Journal for Numerical Methods in
Engineering 33: 2039-2057

The 12
th
16 -18 June, 2011

287

OH-148

Volatile Compounds as Quality Indicators of Fresh Chicken and Possible
Application in Intelligent Packaging

Chompoonoot Rukchon
1
, Sudsai Trevanich
2,3
, Tunyarut Jinkarn
1,3
and
Panuwat Suppakul
1,3,

1
Department of Packaging and Materials Technology, Faculty of Agro-Industry, Kasetsart University,
Bangkok, Thailand 10900;
2
Department of Food Science and Technology, Faculty of Agro-Industry,
Kasetsart University, Bangkok, Thailand, 10900;
3
Center for Advanced Studies for Agriculture and Food
(CASAF), Kasetsart University, Bangkok, Thailand 10900
Corresponding author: fagipas@ku.ac.th

Abstract
Fresh chicken wing sticks have been studied the relationship between the numbers of
microorganisms and the amount of volatile compounds via microbiological analysis and
determination of volatile compounds, respectively. The sensitivity of the dye mixtures as
pH indicator has been assessed by observing with unaided eyes, and measuring with pH
meter and UV-Visible spectrophotometer. Volatile compounds produced by
microorganisms were total volatile basic nitrogen (TVB-N). Especially, an increase in the
amount of TVB-N was in relation to an increase in the numbers of both total aerobic
bacteria and Pseudomonas spp., in turn, a linear relationship between metabolites and
microbial spoilage. The TVB-N concentration would increase within the package of fresh
chicken wing sticks under refrigerated and temperature abuse conditions throughout the
time storage. A TVB-N concentration rapidly increased, in chicken packages stored at 10
C, as a temperature abuse condition. Consequently, TVB-N can be employed as quality
indicator of fresh chicken wing sticks. The dye mixture in yellow color (pH = 6.20, Abs =
1.229) was gradually change to yellow-green (pH = 6.33, Abs = 1.216, green (pH = 6.48,
Abs = 1.091), purple (pH = 7.08, Abs = 0.713) and blue (pH = 7.39, Abs = 0.468) as
increasing the amount of 10 mM ammonia, from 0 to 20 L. Possible application in
intelligent packaging technology as a freshness indicator label has also been demonstrated.
Keywords: Volatile compounds, Metabolites, Fresh chicken, Spoilage, Freshness Indicator
Introduction
Consumer demand for mildly preserved, minimally processed, easily prepared and ready-
to-eat fresher foods together with the globalization of the food business, and the
logistics of distribution from processing centers pose major challenges for food quality
and safety (Vermeiren and others, 1999; Sonneveld, 2000). There is great interest among
the food industry, retailers, consumers rights watchdogs, and food safety controlling
bodies in developing accurate, cost-effective, rapid, reliable, non-invasive and non-
destructive methods or devices to evaluate real-time freshness of food products. An
alternative concept to meet this requirement is the development of intelligent packaging in
the form of a food spoilage indicator to monitor freshness status (Nopwinyuwong and
others, 2010). Intelligent packaging can be defined as a mode of packaging that is capable
of carrying out intelligent functions (such as detecting, sensing, recording, tracing,
communicating, and applying scientific logic) to facilitate decision making to extend shelf
The 12
th
16 -18 June, 2011

288

life, enhance safety, improve quality, provide information, and warn about possible
problems (Yam and others, 2005).
Poultry meat is a highly perishable food and it usually deteriorates within 1 week of
slaughtering, regardless of storage chill systems, and such spoilage is largely due to
dierent types of microorganisms, including bacteria, such as Pseudomonas spp.,
Shewanella putrefaciens and yeasts, depending on the initial microbiological quality of the
poultry carcasses (Russell and others, 1996). In the case of aerobic storage, Pseudomonads
and yeasts are the main microorganisms prevailing (Sundheim and others, 1998; Hinton
and others, 2002). Equally important is the fact that Pseudomonas spp. (including the
human pathogens, Ps. aeruginosa, and Ps. uorescens that are rarely implicated in food
borne disease outbreaks) have been linked with the spoilage of fresh poultry (Allen and
others, 1997; Ellis and Goodacre, 2001; Kelley and others, 1998).
It is well established that the formation of amines, including non-volatiles, such as the
biogenic amines (BAs), and volatile amines (VAs), such as trimethylamine nitrogen
(TMA-N) and total volatile basic nitrogen (TVB-N), is primarily a consequence of the
enzymic decarboxylation of specic amino acids due to microbial enzyme activity
(Bardcz, 1995; Halsz and others, 1994; Ruiz-Capillas and Jimnez-Colmenero, 2004).
TVB-N may potentially be considered as a freshness marker or as an indicator of quality
deterioration in meat, since only small concentration of BAs are found in fresh food, while
the concentration of most BAs usually increases during storage (Rokka and others, 2004).
Reliable methods for assessing the microbiological quality and/or freshness of meat would
benet both consumers and the meat industry (Dainty, 1996). Quantifying chemical
changes could thus provide information on the degree of spoilage, and a number of
chemical indicators have been proposed to assess meat quality including volatile bases,
nucleotides breakdown products, volatile acidity, biogenic amines (BAs) etc. (Halsz and
others 1994; Veciana-Nogus and others 1997).
Protein as main composition in poultry meat is continuously broken down by
microorganisms and finally become a variety of amines including non-volatiles, such as
the biogenic amines, and volatile amines, such as dimethylamine and trimethylamine and
total volatile basic nitrogen (TVB-N). Consequently, these compounds can be employed as
quality indicator of fresh chicken during storage. According to Smolander (2003), color
change of pH dyes (e.g. bromothymol blue, bromophenol blue, bromocresol purple, methyl
red, bromocresol green, methyl orange, methyl yellow, phenol red etc.) can be employed to
detect acidic-basic volatile compounds and displayed an irreversible change in visual
appearance. The objectives of this study are aimed at investigating for the relationship
between the amount of volatile compounds and the numbers of microorganisms and for
sensitivity of the dye mixtures as pH indicator and at assessing for its possible application
in intelligent packaging.

Experimental setup
Chicken wing sticks have been purchased from Betagro Food Company Limited. and
transported to the laboratory within 1 h of purchase. Samples were used in each series of
experiments. First, 585 g of Chicken wing stick samples were aseptically placed into
sterilized 1,000 mL Erlenmeyer flasks. Samples were stored at 4 C and 10 C and
periodically analyzed for product quality during storage.
The 12
th
16 -18 June, 2011

289

Microbiological analysis
Chicken wing sticks were examined for levels of total aerobic bacteria and Pseudomonas
spp. Duplicate samples from each treatment were aseptically opened on the sampling days;
then a 25 g portion of chicken wing sticks was aseptically transferred to a sterile stomacher
bag. Next, 225 mL of 0.1 M sterile sodium phosphate buffer solution (pH 7.0) was added,
and homogenized for 1 min by a Stomacher 400 laboratory blender (Seward, UK). A
series of decimal dilutions was carried out according to recommended microbiological
protocols (Spencer and Ragout de Spencer, 2001). In order to determine total aerobic
bacteria and Pseudomonas spp., 1 mL of each appropriate dilution was pour plated in
duplicate on plate count agar (for total aerobic bacteria) and 0.1 mL of each appropriate
dilution was spread plated in duplicate on cetrimide fusidin cephaloridine agar (for
Pseudomonas spp.). Total aerobic bacteria plates were incubated aerobically for 2 day at
37 C (American Public Health Association, 1984), while Pseudomonas spp. Plates were
incubated for 2 day at 25 C (Mead and Adams, 1977) .Colonies were counted and
reported as log CFU (colony forming units) g
1
.
Determination of total volatile basic nitrogen content
Total volatile basic nitrogen (TVB-N) content was determined using the Conway
microdiffusion assay as described by Ng (1987). A sample (2 g) was added to 8 ml of 4%
trichloroacetic acid (TCA) (w/v) and ground well. Then, it was stand for 30 min at ambient
temperature with occasional grinding, followed by filtration through a filter paper
(Whatman No. 41). The filtrate was kept at 4 C. This filtrate referred to as sample
extract (1 mL) was placed in the outer ring of Conway apparatus. The inner ring solution
(1% boric acid containing the Conway indicator) was then pipetted into the inner ring. To
initiate the reaction, K
2
CO
3
(1 mL) was mixed with sample extract. The Conway unit was
closed and incubated at 37 C for 60 min. The inner ring solution was then titrated with
0.02 M HCl until the green color turned to pink. The concentration of total volatile bases
nitrogen was expressed as mg nitrogen/100 g sample as described by Ng (1987).
Indicator fabrication
Indicator solution
Indicator solution with a concentration of 5% (v/v) in de-ionized water was prepared by
mixing bromothymol blue (0.04%, w/v) in de-ionized water and phenol red (0.04%, w/v)
in de-ionized water in a ratio of 3:2.
Indicator coating
Indicator coating, with a concentration of 30% (v/v), contained: a binder, such as
methylcellulose (3%, w/v); a plasticizer, such as polyethylene glycol-400 (1%, w/v) in
distilled water; and indicator solution with a concentration of 1% (v/v) in ethanol (50%,
v/v). To obtain the desired coating solution, the mixture was homogenized at a speed of
10,000 rpm until complete dissolution of the binder had been achieved. This solution was
then degassed in a CREST 275D ultrasonic water bath for 10 min.
Indicator label
Indicator label was obtained by casting indicator coating onto nylon/LLDPE film to obtain
a thickness of 145 m; label was then dried at room temperature for 24 h.
The 12
th
16 -18 June, 2011

290

Color changes of indicators caused by ammonia
Color changes of the indicator solution due to contact with NH
3
were studied by enclosing
10 mL of colorimetric mixed-dye-based solution in test tubes (20 mL). Ammonia with the
concentrations of both 1 mM and 10 mM was added into test tubes in the ranges of 0-200
L and 0-20 L, respectively. As pH indicator, it has been also observed with unaided eyes
and measured with pH meter. In addition, ammonia with the concentration of 10 mM
Optical spectrum of these indicator solution samples was recorded with a T60 UV
spectrophotometer (PG Instruments, UK).
Color changes of indicator label due to contact with NH
3
were studied by enclosing
indicator labels in gas-tight 25 mL vials. Ammonia with the concentration of 150 mM was
added into vials in the range of 0-10 L. The color change of the colorimetric mixed-dye-
based indicator was observed with unaided eyes.
Chemical and microbiological changes in fresh chicken
Total volatile basic nitrogen (TVB-N) was markedly detected in fresh chicken, and its level
increased with storage time (Fig. 1). At lower temperature, more TVB-N is adsorbed into
the food matrix, while at higher temperature less is adsorbed. For gaseous specie, its
corresponding vapor pressure is very high. Generally, it was found that the vapor pressure
of TVB-N increases more and more rapidly with rise of temperature. Moreover, a dynamic
change in TVB-N level was observed, which could be related to the growth of
microorganisms (Fig. 1 and Fig. 2). Proliferation of the microflora contributing to spoilage
changes as seen by increased TVB-N level coincided with the detection of trimethylamine
(TMA-N) (data not shown). Pseudomonas spp. which were expected to be a specific
spoilage organism (SSO) were able to continuously grow under these conditions.
Initially, SSO are present in low quantities and constitute only a minor part of the natural
microflora. During storage, SSO generally grow faster than the remaining microflora and
produce the metabolites responsible for off-odors and off-flavors, and finally cause sensory
rejection. The cell concentration of SSO at rejection may be called the minimal spoilage
level, and the concentration of the metabolites that correspond to spoilage can be used as
an objective chemical spoilage index (CSI) (Dalgaard, 1993).
This resulted in 6.99 log CFU g
-1
and 7.10 log CFU g
-1
of total aerobic bacteria, and in
6.17 log CFU g
-1
and 6.08 log CFU g
-1
of Pseudomonas spp., stored at 4 C and 10 C for
4.5 d and 3.5 d, respectively, which did not exceeded the limit of acceptability at the end of
storage. The limit of acceptability was based on the onset of food spoilage, which was
considered to be 7.00 log CFU g
-1
of both total aerobic bacteria and Pseudomonas spp.
(Senter and others 2000; Balamatsia and others 2007). Thus, shelf lives of fresh chicken,
stored at 4 C and 10 C, were approximately 4.5 d and 3.5 d, respectively.
A clear correspondence was found between the microbiological quality of fresh chicken
(protein-based) and the level of metabolites. Total volatile basic nitrogen formation took
place evenly during the storage period; however, its formation was dependent upon
temperature. Higher storage temperature had a more conspicuous effect on the formation
of TVB-N, which accumulated especially at the end of the storage period (Fig. 1). This
correlation is in agreement with the findings of Balamatsia and others (2007), which firstly
reported that trimethylamine (TMA-N) and total volatile basic nitrogen (TVB-N) could be
employed as potential chemical indicators in monitoring the microbial quality of fresh
chicken meat during chill storage under aerobic and modified atmosphere packaging
The 12
th
16 -18 June, 2011

291

(MAP) conditions, and the findings of Rokka and others (2004), which showed a clear
relationship between the microbiological quality of poultry (protein-based) and the total
amount of biogenic amines.

Figure 1 Total volatile basic nitrogen found in chicken wing sticks at 4 C( ) and 10 C ( ).

Figure 2 Total aerobic bacterial count and Pseudomonas spp. count found in Chicken wing sticks at
4C( ) and 10 C ( ).

Color changes of indicators
The dye mixture in yellow color (pH = 6.20, Abs = 1.229) was gradually change to yellow-
green (pH = 6.33, Abs = 1.216, green (pH = 6.48, Abs = 1.091), purple (pH = 7.08, Abs =
0.713) and blue (pH = 7.39, Abs = 0.468) as increasing the amount of 10 mM ammonia,
from 0 to 20 L, as shown in Fig. 3.
When NH
3
was added to vials containing colorimetric mixed-dye-based indicator solution,
a visual color change of the solution from yellow to blue was detected. The most
remarkable change in the absorption spectra after 30 min of reaction time took place at
high absorption peaks in the wavelength range of 430-615 nm (Fig. 4). Bromothymol blue,
which shifts from acidic form (yellow, pH 5.8) to basic form (blue, pH 7.6), results in
maximum lambda (
max
) shifting from 430-435 nm to 615-618 nm .Phenol red, which
shifts from acidic form (yellow, pH 6.8) to basic form (red, pH 8.4), results in maximum
The 12
th
16 -18 June, 2011

292

lambda (
max
) shifting from 430-435 nm to 557-560 nm (Sabnis, 2008). As a mixed-dye-
based indicator, yellow changed to blue in relation to a shift in the absorption peak from
430-435 nm to 615-618 nm when exposed to levels of 150 mM NH
3
from 0-10 L, as
depicted in Fig. 5. This vivid color spectrum of mixed-dye indicator solution is in
agreement with Wallach (1996), who reported that a mixed indicator could enhance an
expansion of the range of color change, as compared with a single indicator. The main
purpose for applying colorimetric mixed-dye-based indicators to food packaging is to
easily and reliably monitor the level of food spoilage of packaged food products in a non-
destructive manner during distribution and retail sale.

Figure 3 Absorbance and pH of pH dye solution depended on the amount of ammonia 1 mM ( ) and
10 mM ( ).

Figure 4 Absorption spectra of pH indicator solution. Ammonia (10 mM): A = 0 L, B = 2 L, C = 4
L, D = 7 L, E = 8 L, F = 11 L, G = 13 L, H = 16 L and I = 20 L.

Figure 5 Color change of pH indicator label as increasing volume of 150 mM NH
3
from 0 to 10 L.
Film
[NH
3
] 0 L 2.5 L 5 L 7.5 L 10 L

The 12
th
16 -18 June, 2011

293

Conclusions
This study revealed that microbial spoilage of fresh chicken wing sticks resulted in an
increase in metabolites as a linear relationship. Bromothymol blue and phenol red mixture
could be employed as a tool for monitoring microbial spoilage. This pH indicator
demonstrated a possible application in intelligent packaging as food spoilage indicator.
However, further research is necessary to comprehensively assess the efficiency of the
indicator and the types of food that can benefit from this indicator packaging material.
Acknowledgements
This work was partially supported by a fund for the promotion of research at the Center for
Advanced Studies for Agriculture and Food (CASAF), Kasetsart University. The authors
express their thanks and appreciation for these supports. Author Rukchon would like to
thank Miss Atchareeya Nopwinyuwong for her support towards the experiments.
References
Allen CD, Russell SM, Fletcher DL. 1997. The relationship of broiler breast meat color
and pH to shelf-life and odor development. Poult Sci 76(7):1042-6.
American Public Health Association. 1984. In: Speck ML. editor. Compendium of
Methods for the Microbiological Examination of Foods (second ed), Washington,
DC : APHA.
Balamatsia CC, Patsias A, Kontominas MG, Savvaidis IN. 2007. Possible role of volatile
amines as quality-indicating metabolites in modified atmosphere-packaged chicken
fillets: Correlation with microbiological and sensory attributes. Food Chem
104(4):1622-8.
Bardcz S. 1995. Polyamines in food and their consequences for food quality and human
health. Trends Food Sci Technol 6(10):341-6.
Dainty R. 1996. Chemical/biochemical detection of spoilage. Int J Food Microbiol
33(1):19-33.
Dalgaard P. 1993. Evaluation and Prediction of Microbiological Fish Spoilage. PhD
thesis, Royal Veterinary and Agricultural University. Copenhagen, Denmark.
Ellis DI, Goodacre R. 2001. Rapid and quantitative detection of the microbial spoilage of
muscle foods: current status and future trends. Trends Food Sci Technol
12(11):414-24.
Halsz A, Barth A, Simon-Sarkadi L, Holzapfel W. 1994. Biogenic amines and their
production by microorganisms in food. Trends Food Sci Technol 5(2):42-9.
Hinton Jr A, Cason JA, Ingram KD. 2002. Enumeration and identification of yeasts
associated with commercial poultry processing and spoilage of refrigerated broiler
carcasses. J Food Prot 65(6):993-8.
Kelley TR, Pancorbo OC, Merka WC, Barnhart HM. 1998. Antibiotic resistance of
bacterial litter isolates. Poult Sci 77(2):243-7.
Mead GC, Adams BW. 1977. A selective medium for the rapid isolation of Pseudomonas
associated with poultry meat spoilage. Br Poult Sci 18(6):661-70.
Ng CS. 1987. Determination of trimethylamine oxide (TMAO-N), trimethylamine (TMA-
N), total volatile basic nitrogen (TVB-N) by Conway's method. In: Hasegawa H.
editor. Laboratory manual on analytical methods and procedures for fish and fish
products. Singapore: Marine Fisheries Research Department, Southeast Asian
Fisheries Development Center. pp. 18.
The 12
th
16 -18 June, 2011

294

Nopwinyuwong A, Trevanich S, Suppakul P. 2010. Development of a novel colorimetric
indicator label for monitoring freshness of intermediate-moisture dessert spoilage.
Talanta 81(3):1126-32.
Rokka M, Eerola S, Smolander M, Alakomi H-L, Ahvenainen R. 2004. Monitoring of the
quality of modified atmosphere packaged broiler chicken cuts stored at different
temperature conditions: B. Biogenic amines as quality-indicating metabolites. Food
Control 15(8):601-7.
Ruiz-Capillas C, Jimnez-Colmenero F. 2004. Biogenic amines in meat and meat products.
Crit Rev Food Sci Nutr 44(7-8):489-99.
Russell SM, Fletcher DL, Cox NA. 1996. Spoilage bacteria of fresh broiler chicken
carcasses. Poult Sci 75(12):2041-7.
Senter SD, Arnold JW, Chew V. 2000. APC values and volatile compounds formed in
commercially processed raw chicken parts during storage at 4 and 13 C under
simulated temperature abuse conditions. J Sci Food Agric 80(10):1559-64.
Smolander M. 2003. The use of freshness indicators in packaging. In: Ahvenainen R,
editor. Novel Food Packaging Techniques. Cambridge: Woodhead Publishing Ltd.
p. 127-143.
Sonneveld K. 2000. What drives (food) packaging innovation? Packag Technol Sci
13(1):29-35.
Spencer JFT, Ragout de Spencer AL. 2001. Food Microbiology Protocols. Totowa:
Humana Press. 495 p.
Sundhein G, Sletten A, Dainty RH. 1998. Identification of Pseudomonads from fresh and
chill stored chicken carcasses. Int J Food Microbiol 39(3):185-94.
Veciana-Nogus MT, Marin-Font A, Vidal-Carou MC. 1997. Biogenic amines as
hygienic quality indicators of tuna relationships with microbial counts, ATP-related
compounds, volatile amines and organoleptic changes. J Agric Food Chem
45(6):2036-41.
Vermeiren L, Devlieghere F, van Beest M, de Kruijf N, Debevere J. 1999. Developments
in the active packaging of foods. Trends Food Sci Technol 10(3):77-86.
Wallach DFH. 1996. Methods and devices for detecting microbial spoilage in food
products. US Patent No. 6,495,368.
Yam KL, Takhistov PT, Miltz J. 2005. Intelligent Packaging: Concepts and Applications. J
Food Sci 70(1):R1-R10.

The 12
th
16 -18 June, 2011

295

OH-173

Effect of Protein Addition on Qualities of Gluten-Free Rice Crackers

Ratanatriwong, P.A.
*
, Nammakuna, N., Thanasukarn, P., Suwansri, S

Department of Agro-Industry, Faculty of Agriculture Natural Resources and Environment,
Naresuan University, Phitsanulok Thailand 65000
*
Corresponding author: rikaja@yahoo.com, 66-55-962743 (Tel), 66-55-962703(Fax)

Abstract

This research determined the effect of protein sources on improving qualities of gluten-free
rice crackers. Two protein types including soy protein isolate and pea protein isolate were
added at 2.5%, 5.0% and 10% w/w (total rice flour basis) to a formulated-flour blend.
Physical properties, dough rheology and microstructure of samples were compared with
three controls including non-protein-added formula, 100% rice flour and 100% wheat
flour. Rice dough had the highest storage modulus (G) followed by those of non-protein-
formula, pea-protein, soy-protein and wheat dough. Wheat dough was the most elastic and
easily-handling whereas rice dough was dry and tough. Adding protein helped facilitate the
rice dough handling process, and soy-protein dough was less tough than pea-protein.
Wheat dough illustrated starch granules embedded in matrix structure. The structures of
protein-added dough were more evident than those of rice and non-protein controls that
starch granules in wrapped in thicker layers, which agreed with the dough rheology.
Addition of protein significantly increased moisture content and water activity of all
treatments (p<0.05). They were closer to wheat cracker than other controls (p<0.05).
Addition of pea proteins slightly darkened the color shade as compared to the non-protein
control, but their colors more acceptable than control rice cracker (p<0.05). Texture of
control rice cracker was crumbly while wheat cracker was light-crisp and brittle, and non-
protein formula was slightly hard. Adding both protein sources helped improve rice
cracker texture to be closer to wheat cracker. They were perceived as more brittle and crisp
than non-protein control. Thus, adding protein sources in gluten-free rice cracker helped
improve dough characteristic, resulting in more desirable product qualities.

Keywords: Gluten free, rice cracker, product quality, soy protein, pea protein

Introduction
Wheat is one of the most common food allergens. Wheat allergy is frequently confused
with celiac disease. People with wheat allergy have an IgE-mediated response to wheat
protein, but they may tolerate other grains (FAAN
TM
, 2011). Children with wheat allergy
sometimes outgrow this disease before reaching their adulthood. On the other hand, celiac
disease is a digestive disease with a permanent adverse reaction to gluten, which is a
protein primarily found in wheat, rye, barley and sometimes oats (NDDIC, 2008).
Individuals with celiac disease will not lose their sensitivity to gluten, and must be on
restrict gluten-free diets. Rice, potato, soy or corn flours become alternative in gluten-free
diets. Rice is a major corps of Thailand, and rice production is higher than wheat due to
climatic conditions. Thus, this is an opportunity for snack industries to add their own twists
for gluten-free snacks such as rice crackers using rice flour as wheat flour. In addition, it is
highly reasonable in terms of marketing and cost aspects. However, the absence of gluten
in rice flour presents the technical difficulties in rice cracker production since gluten plays
The 12
th
16 -18 June, 2011

296

an important role in bakery products by being the most important structure forming
protein. Many have tried to mimic gluten properties in wheat flour by using hydrocolloids
or other protein sources. Various hydrocolloids were used with dairy protein and oat -
glucan to improve gluten-free bread (Lazaridou and others 2007). Different protein sources
such as dairy or soybean proteins were used in gluten-free products as a nutritional source
(Marco and Rosell, 2008b). Soybean or pea proteins were added to modify the mechanical
properties such as increasing the elastic modulus of a rice-protein blend dough (Marco and
Rosell, 2008a). Various hydrocolloids were added in a formulated flour blend, composed
of rice flour and pregelatized tapioca starch, to improve rice cracker qualities (Nammakuna
and others 2010). Despite these successes, the application of those ingredients for gluten-
free rice cracker is very limited. Thus, this research focused on the effects of protein source
addition and their optimal addition amount on gluten-free rice crackers.

Flour and ingredients
All protein-added rice cracker formulation used a mixed flour blend of rice flour variety
Surin1 (Chaijalearn Co., Ltd., Bangkok, Thailand, waxy rice flour variety RD6 (Jewhoksang
Co., Ltd., Lampang, Thailand) and pregelatinized tapioca starch (Bangkok Starch Co., Ltd.,
Bangkok, Thailand). Controls were non-protein-added mixed flour formula (FF), 100% rice
flour (RF) and 100% wheat flour (WF). The remaining ingredients purchased from local
market were salt, sugar, milk powder, palm oil, glucose syrup, baking powder, margarine,
dry yeast, ammonia powder, lecithin, flavor better milk, vanilla, corn flour and tapioca
flour. Two protein sources of soy protein isolate (F.A. Unity Co., Ltd, Bangkok, Thailand)
and pea protein isolate (Mighty food Thailand Co., Ltd, Bangkok Thailand) were added at 2.5%,
5.0% and 10.0% (total rice flour basis) in treatment samples.

Cracker preparation
Cracker dough of either treatments with adding protein sources or controls were prepared
by mixing 50% of flour, sugar and water in a mixer (Kitchen-aid model, USA), adding
yeast, blending and proofing dough at 25-30
o
C for 60 min, resulting first-portion dough.
The other portion of mixture consisted of margarine, baking powder, ammonia powder,
lecithin, salt, milk powder, palm oil, glucose syrup, flavor better milk and vanilla were
well blended before adding the previously-proofed dough and the rest of flour. The
mixture was blended in a mixer follow by proofing dough at 25-30
o
C for 60 min. Proofed
dough was kneaded, sheeted, layered, cut to a cracker size of 5x5x0.2 cm, baked in the
oven at 200
o
C for 7-8 minute and was cooled at 25-30
o
C for 60 min. Samples were packed
in sealed bags and stored at room temperature until being determined.

Quality determination of dough and cracker
The measurements of moisture content and water activity (a
w
) of samples were determined
in three replicates using a moisture meter model Sartorius MA40 (Sartorius, Inc.,
Goettingen, Germany) and an water activity meter model Novasina RS 200 (Novasina,
Axair Ltd., Pfaffikon, Switzerland), respectively. Texture of samples was determined in ten
replicates using a Texture Analyzer model QTS25 (Brookfield Engineering Labs., Inc.
MA. USA). Color was determined in CIE system (L*, a*, b*, hue angle and chroma) by a
color reader model CR-10 (Konica Minolta sensing Inc., Osaka, Japan).

The dough microstructure was studied using a scanning electron microscope Model
The 12
th
16 -18 June, 2011

297

1455VP (Leo Electric Systems, Cambridge, UK). Prior to SEM study, cracker dough
samples was cut to size 1010 mm, freeze dried, and kept in a desiccators until further use.
Dough sample was mounted on a slide and separately placed on a sample holder using
double-sided scotch tape. The internal structure was faced upward and sputter-coated with
gold (2 min, 2 mbar) before being transferred to the microscope where it was observed in
vacuum at an accelerating voltage of 5 kV.

Statistical Analysis
The qualities of samples were analyzed in two replicates. The dough microstructure,
physical and chemical properties of protein source-treated samples were compared with
non-protein-added controls. Experimental design in this research was a Completely
Randomized Design (CRD). The main effect was protein sources type. Two protein types
with three levels each were used. All treatments were duplicates. Data was statistically
analyzed by Duncans Multiple Range test (DMRT) at 95% confidence level.

Addition of protein significantly increased the moisture content and water activity values
of all treated samples to be much closer to that of wheat cracker than other controls
(p<0.05) (Table 1). The desirable color of cracker was golden-yellowish, which was
perceived from wheat crackers whereas the control rice cracker was considered too pale.
The color parameters showed significant differences among samples (p<0.05). As
observed, the shade of pea-protein-added samples was darker than soy-protein samples and
other rice- and non-protein controls. This agreed with darkened bread crust found by
adding protein (Mezaize and others, 2009).

Table 1 Physical and chemical properties of rice crackers with added proteins.

Samples L* Chroma Hue angle a
w
Moisture
content (%)
FF 55.192.078
a
33.813.678
a
73.374.830
bc
0.160.00
f
3.430.12
g
RF 51.482.080
bc
31.462.46
b
76.322.100
a
0.150.01
f
3.520.50
g
WF 44.020.370
e
31.454.17
b
71.447.582
cd
0.220.00
e
5.540.03
cd
FF+SP 2.5% 53.712.247
ab
28.782.696
c
77.102.166
a
0.320.00
a
7.010.22
a
FF+SP 5.0% 51.661.598
bc
28.881.863
c
77.201.763
a
0.220.01
e
5.760.16
c
FF+SP 10.0% 50.961.842
cd
30.682.072
bc
75.731.953
ab
0.220.00
e
5.340.13
de
FF+PEA 2.5% 52.785.150
bc
29.856.186
bc
74.390.486
ab
0.250.00
c
5.060.33
ef
FF+PEA 5.0% 50.353.800
cd
34.152.647
a
70.421.189
d
0.230.00
d
5.030.14
f
FF+PEA 10.0% 48.890.512
d
35.561.760
a
68.763.450
d
0.270.00
b
6.090.17
b
Note: Different letters in the same column indicated statistical differences (p0.05).
FF, RF and WF were formulated-flour, rice-flour and wheat-flour controls, respectively.

The 12
th
16 -18 June, 2011

298

Table 2 Effect of protein sources on texture of rice crackers with added proteins.

Samples
Fracture force Hardness
FF 397.367.33
bc
459.036.06
ab
RF 107.33.050
e
216.893.04
c
WF 554.053.30
a
537.738.08
a
FF+SP 2.5% 256.765.00
d
306.714.74
abc

FF+SP 5.0% 263.019.70
d
296.336.09
abc
FF+SP 10.0% 221.342.15
d
337.054.44
abc
FF+PEA 2.5% 268.741.53
d
252.765.04
bc
FF+PEA 5.0% 308.756.75
cd
375.073.50
abc
FF+PEA 10.0% 433.735.80
b
521.743.65
a

Note: Different letters in the same column indicate statistical differences (p0.05).
FF, RF and WF were formulated-flour, rice-flour and wheat-flour controls, respectively.

a b c
d c
Figure 2 Effect of protein sources on the dough microstructure of rice crackers with added protein
(a) FF+soy protein isolate 10.0%, (b) FF+pea protein isolate 10.0%, (c) formulated flour
control (FF), (d) rice flour control (RF) and (e) wheat flour control (WF) ( 700).
The 12
th
16 -18 June, 2011

299

According to Table 2, significant differences in texture characteristics were found among
samples (p<0.05). Wheat cracker is used as positive control since it is desirable for
consumer. Its texture was lightly crisp and brittle. This was because wheat dough was
elastic and able to entrap gas (Lazaridou and others 2007), resulting in high puffiness (data
not shown) with many inner layers. In contrast, texture of control rice cracker was
crumbly with no layers as indicated by its fracture force and hardness being the lowest
(Table 2). This is because rice protein is different from wheat gluten that it does not have a
cohesive dough structure. The texture of rice crackers is mostly caused from starch
gelatinization, which lacks of good viscoelastic properties so it was not strong enough to
entrap gas produced in the food system (Sozer, 2009). The control from mixed flour blend
(FF) contained pregelatized-tapioca starch in the flour blend that facilitated dough
formation from more starch gelatinization (Sozer, 2009), consequently resulting in better
texture than rice control. However, the texture of non-protein-added control still had fewer
layers and was less puffy, which caused it being harder than wheat crackers, despite its
lower values of fracture forces and hardness (Table 2). Besides wheat crackers, the
protein-added samples were perceived as relatively more brittle and lightly crisp than
other controls because they had more layers and more puffy. The result agreed with Sozer
(2009) that addition of proteins helped starch granules adhere to one another, and water
was more distributed through the system because of polymeric structure of proteins
(Sivaramakrishnan and others 2004).

In addition, it was clearly seen in the dough microstructure that the structures of protein-
added dough were more evident than those of rice and non-protein controls (Fig 2). Starch
granules in soy-protein- and pea-protein-added dough were wrapped in layers (Fig. 2a-2b),
which helped entrap gas during baking, whereas the structure of non-protein- and rice
dough were indistinguishable, and starch granules were relatively loosely adhere to one
another (Fig. 2c-2d). This explained why better cracker textures were obtained from
protein-added crackers. As expected, the wheat dough microstructure illustrated starch
granules of both A- and B-types embedded in gluten matrix structure that helped entrap
water and gas in dough system (Kim and Huber, 2010). Thus, adding protein sources in
gluten-free rice cracker helped improve dough characteristic which consequently resulted
in improving rice cracker textures.

The effect of frequency on the viscoelastic behavior of rice cracker dough with protein
addition compared with controls was shown in Fig. 3. For all samples, the storage modulus
G (elastic property) was greater than the loss modulus G (viscous property). The control
rice dough had the highest elastic modulus value (G). This probably represented a more rigid
structure than an elastic structure (Weipert, 1990). The result agreed with Sozer (2009). A
decrease in storage modulus value was observed in pasta dough with more gelatinized rice
flour because of an increase in water absorption and starch granules swelling (Sozer,
2009). Likewise, wheat cracker dough showed the lowest storage modulus value because
gluten has a unique viscoelastic property with more water holding capacity and network
formation, resulting in a more elastic dough than rice (Marco and Rosell, 2008b; Sozer,
2009). With protein additions, the storage modulus values were lower than control rice
dough, showing the improvement of dough elasticity. This was probably due to the
polymeric structure of protein that facilitated the water holding capacity and water
distribution in dough. The result also agreed with research done by many (Lazaridou and
others 2007; Marco and Rosell, 2008b; Nammakuna and others 2009, Sozer, 2009) that dough
The 12
th
16 -18 June, 2011

300

viscoelastic properties could be improved by addition of protein and gums because they
enhanced the viscoelastic properties of polysaccharides in the aqueous medium.

Figure 3 Effect of different protein sources on rheological properties of rice cracker based on the
frequency sweep test.
Note: FF, RF and WF were formulated-flour, rice- and wheat cracker controls, respectively.

Conclusion
The effect of protein sources on the improvement of rice cracker qualities were illustrated
by dough rheological properties, dough microstructure, physical and chemical properties,
and texture of rice cracker samples. Adding proteins in rice cracker help facilitate dough
formation and improve dough viscoelastic property so protein-added dough could hold
more water and entrap gas in the system, resulting in rice cracker samples with more
desirable qualities.

Acknowledgements
This research was funded by the Thailand Research Fund (Grant number: MRG 5380217).
The In-kind supports including rice flour (Sin Salee Snack & Biscuit Ltd., Part. Bangkok,
Thailand), soy protein isolates (F.A. Unity Co., Ltd, Bangkok, Thailand) and pea protein isolate
(Mighty food Thailand Co., Ltd, Bangkok) were gratefully acknowledged.

The 12
th
16 -18 June, 2011

301

References
[FAAN] The Food Allergy & Anaphylaxis Network. 2011. Wheat Allergy:Available from:
http://www.foodallergy.org/page/wheat-allergy. Accessed March 3.
Lazaridou A., Duta D., Papageorgiou M., Belc N and Biliaderis C.G. 2007. Effects of
hydrocolloids on dough rheology and bread quality parameters in gluten-free
formulations. J Food Eng 79:1033-1047.
Marco and Rosell, 2008a. Effect of different protein isolates and transglutaminase on rice
flour properties. J Food Eng 84:132139.
Marco and Rosell, 2008b. Functional and rheological properties of protein enriched gluten
free composite flours. J of Food Eng 88:94103.
Nammakuna, N., Suwansri, S., Thanasukan, P. and Ratanatriwong, P. 2010. Effects of
hydrocolloids on quality of rice crackers made with mixed-flour blend. As. J. Food
Ag-Ind. 2(04):780-787.
Rosell, C. M., Rojas, A. J. and Benedito, B. C. 2001. Influence of hydrocolloids on dough
rheology and bread quality. J Food Hydrocolloid 15:75-81.
Sivaramakrishnan, H.P., Senge, B. and Chattopadhyay, P.K. 2004. Rheological properites
of rice dough for making rice bread. J Food Eng 62:37-45.
The National Digestive Diseases Information Clearinghouse (NDDIC). 2008. Celiac
Disease. National Institutes of Health (NIH). United States Department of Health
and Human Services. NIH Publication No. 084269, September 2008.
Weipert, D. 1990. The benefits of basic rheometry in studying dough rheology.
Cereal Chem 67(4):311-317.

The 12
th
16 -18 June, 2011

302

OH-306

Methanol Reduction in Fruit Winemaking Process

Ming-Chang Wu
1
, Chien-Hao Chen
3
, Ping-Hsiu Huang
1
, Yuh Tai Wang
2,*

1
Department of Food Science, National Pingtung University of Science and Technology,
1, Hsueh Fu Road, Nei-Pu Hsiang, 91201 Pingtung, Taiwan, ROC;
2
Life Science Center, Hsing Wu College, No. 101, Sec.1, Fen-Liao Road, Lin-Kou, Taipei 244, Taiwan;
3
Department of Food and Beverage management, National Kaohsiung University of
Hospitality and Tourism, Taiwan, ROC
*
Corresponding author: yuhtai@yahoo.com

Abstract

Using commercial pectic enzyme (CPE), a complex enzyme including pectin methyl
esterase (PME) (EC 3.1.1.11), polygalacturonase (PG) (EC 3.2.1.15), and pectin lyase (PL)
(EC 4.2.2.2), in fruit wines and spirits manufacturing has resulted a problem regarding its
excess production of methanol in the final products due to methyl esters hydrolysis in
polygalacturonide linkage of the pectin by PME during winemaking. Since methanol
intake which is resulted from drinking fruit wines and the distilled liquors that contain
methanol may cause blindness and even death, it is recommended in the manufacturing
process of wines and liquors to improve the safety of wines and liquors by reducing
methanol production. In this report, the methods utilized in methanol reduction, include
addition of PME inhibitor from jelly fig (Ficus awkeotsang) seeds, coumaric acid and
gallic acid, and utilization of PL for CPE replacement, did reduce the methanol content of
fruit wines during winemaking.

Keywords: Methanol, winemaking, pectin methyl esterase, pectin lyase, gallic acid,
coumaric acid.

Introduction
Pectins are structural heteropolysaccharides that exist mainly in the middle lamella and
primary cell walls of higher plants (Oakenfull and Sctt 1984; Northcote 1972). The
presence of pectins causes problems in the food industry due to the increase of viscosity in
fruit juices, thus impeding the processes of filtration, concentration, and clarification of
fruit juice (Silvia and others 1996; Blanco and others 1997; Nedjma and others 2001; Lima
and others 2002; Fachin and others 2003; Moyo and others 2003).
Pectic substances could be digested using pectic enzymes produced by bacteria, fungi and
yeast (Bai and others 2004; Jayani and others 2005). The most common commercial pectic
enzyme (CPE) used in food industry is from microbial sources such as Aspergillus niger,
Penicillium dierckxii, Kluyveromyces marxianus, and Penicillium griseoroseum (Fawole
and Odunfa, 2003; Gummadi and Panda 2003). It is a complex enzyme which contains
pectin methyl esterase (PME), polygalacturonase (PG), and pectin lyase (PL). It plays an
important role in juice extraction and winemaking process, shortens the time of
maceration, setting, and filtration, consequently, it increases the yield and quality such as
flavor, transmittance, viscosity and pigments, enhances the extraction of anthocyanins and
flavanols, and reduces skin contact time (Lao and Lopez-Tamames 1999; Muoz and
others 2004; Wu and others 2007; Hou and others 2008). Moreover, it is reported that the
addition of CPE for ciders (Massiot and others 1994) or wines (Servili and others 1992;
The 12
th
16 -18 June, 2011

303

Frenkel and others 1998; Revilla and Gonzlez-SanJos 1998; Cabaroglu 2005; Wu and
others 2005, 2007) causes an increase in methanol level during the mashing, fermentation
and aging stages.

Methanol is an undesirable product in the processing juice, wine and spirit beverages,
which are resulting from de-esterification of pectins, become low-methoxyl pectins by
PME. Methanol content can be enhanced by the cooperation of PME and PG during juice
clarification (Massiot and others 1997). Methanol from fruit, vegetables, fermented
beverages, and foods sweetened with aspartame may be ingested by methanol breaks down
metabolism in the gastrointestinal tract (Newsome 1993). Therefore, the safety limit or
maximum permitted concentration of methanol in wines and distilled fruit spirits is fixed in
regulations for avoiding the risk to consumers health as the following: 35 mg of
methanol/100 mL in wine of Brazil (Silva and others 2007), 10 g of methanol/L of ethanol
for EU (Paine and others 2001), 150 mg/L for white wine and ros wine and 300 mg/L for
red wine in Intl. Office of Vine and Wine (Cabaroglu 2005), 2000 mg of methanol/L of
ethanol for wine and its spirit (equal to 200 mg/L from 10% ethanol containing), and 1000
mg methanol/L of ethanol for the other kinds of alcoholic beverages (equal to 100 mg/L
from 10% ethanol containing) except fruit wine and its spirit in Taiwan (Wu and others
2007).

Following the government policy of relaxing the control of tobacco and liquor sales in
Taiwan, farmers are allowed to brew fruit wines in order to transform agriculture products
into product with higher value added so that to raise their standard of living. Regardless of
the manufacturing materials and various techniques used, fruit wines and the distilled
liquors contain methanol. According to the results of Wang and others (2004), some
commercial wines, alcohol-containing drinks and liquors in Taiwan were found to have
methanol content exceeding the standard limits, especially when it is based on alcohol
content. In addition to, the presence of methanol in grape pomace stored until distillation
affects the aromatic characteristics of the distillate, therefore, reduction of methanol by
controlling methanol formation can enhance the quality of grappa (Zocca and others 2007).
Therefore, better understanding of methanol formation and its controlling methods will
benefit to the manufacturing process of winemaking industry and the marketing of fruit
wines and spirits.

Methanol Toxicity in Alcoholic Beverages
Methanol is a colorless volatile compound with a mild alcoholic odor. The molecular
formula of methanol is CH
3
OH, which produces toxicity in human body mainly as a result
of drinking adulterated liquors and is readily absorbed by ingestion and inhalation and,
more slowly, through the skin. In the body, methanol is metabolized in the liver, and
converted first to formaldehyde and then to formate (Lamiable and others 2004). The
methanol toxicity is mainly caused by an increase in formic acid and the resultant
metabolic acidosis. High levels of formate after excessive methanol intake can cause
severe toxicity and even death.

The impure liquors flooded the markets and caused many cases of blindness and even
death in locations such as China, Brazil, El Salvado, Turkey and Taiwan, there is high
correlation between localities and death rate (Nesime and others 2003). Thus, the lethal
toxicity of methanol could not be overlooked.
The 12
th
16 -18 June, 2011

304

For example, ingestion of more than 4 g methanol could produce toxic symptoms such as
headache, nausea, vomiting, and dizziness. Ingestion of methanol more than 10 g might
cause blindness and ingestion of more than 24 g might lead to death. In the case of spirit,
methanol is less toxic to the body due to the competition between high ethanol content and
methanol in the spirit which results in the inhibition of methanol metabolism (Wang and
others 2004; Azmak 2006). The toxicity of aldehyde is 33 times higher than methanol and
formic acid is six times more toxic than methanol (Plaziac and others 2003). Furthermore,
the metabolite of methanol is even more toxic than methanol itself.

Methanol Production in Fresh Juice
CPE has been widely used in manufacturing processes to increase juice extraction and
decrease pectin hazes in fruit juice making. It is found that methanol can be produced from
the hydrolysis of methyl ester groups in pectin by endogenous PME and CPE contained
PME and occurs naturally at a low level in fresh fruit juices and most of the alcoholic
beverages (Wu and others 2005; Hou and others 2008) and increases to 3.5 times higher in
fermented pomace at day 20 after crushing (Zocca and other 2007).
Nevertheless, methanol production could be influenced by factors such as, but not limited
to pectin content, pectin esterification, PME enzymatic activity (especially activity at low
pH of fruit juice), temperature and duration of manufacturing processing and ripeness of
fruits (Arbaisah and others 1996; Deuel and Stutz 1985; Wu and others 2002; Diaz and
others 2007). The level of methanol production varies in different categories of fruits and
vegetables. The production of methanol is positively associated with enzyme activities of
PME and PL, while negatively associated with total pectin content in fruit juices.
Specifically, the production of methanol is positively associated with PME activity, pH
value and titratable acidity, while negatively associated with PG activity in vegetable
juices (Hou and others 2008).

Methanol Production during Winemaking Process
Methanol in fruit wines and their distilled liquors originates mainly from brewery process
in the winemaking industry. During the winemaking process, CPE also plays an important
role in improving extraction of color, flavor, filtration efficiency, and clarification of fruit
wines. However, CPE creates a problem as well by methanol production during crushing in
the early stage of manufacturing process.

Wu and others (2005) carried out a model system to monitor methanol production after
various concentrations of PME were added to the solution containing a fixed concentration
of pectin (1%). They found that the more PME was added, the more methanol was
produced. Revilla and Gonzalez-SanJose (1998) used four different kinds of pectic
enzymes to examine the reason why methanol production was increased and found that
enzymatic treatment produced higher methanol levels than the control in the final wine.
Higher methanol content was related to higher pectolytic activity, which is reflected in
lower residue pectin content, and to their higher PME activity (Revilla and Gonzalez-
SanJose 1998).

Furthermore, types of grapes (fermentation with grapes with or without fruit skin),
physiological condition of grapes, filtrate fermentation process, fermentation temperature
and pretreated pectic enzymes are all reported to affect methanol production (Cabaroglu
2005). For example, red wine contains about 113.7 mg/L of methanol whereas white wine
The 12
th
16 -18 June, 2011

305

contains 58.5 mg/L of methanol. The winemaking process of white wines is different from
red wines because grape juices were used instead. That is to say, most of the interaction
between pectin in the grape skin and PME were removed. As a result, the methanol content
in the white wine is lower. Liu (2009) found that the methanol content of strawberry wine
is reduced 54-57% when replacing fruit must with fruit juice. It is speculative to attribute
high methanol content in the wine to using fruits with skin intact which provides a longer
duration of interaction between pectin of fruits and their own PME. That is to say that
methanol occurs naturally at a low level in fruit must and fruit juice, which is mainly
derived from intrinsic PME in fruits acting on PME during fermentation. Consequently,
fruit juice extraction method used and the added PME to the juice during fermentation
process influence greatly on methanol production.

Methods for Methanol Reduction in Fruit Winemaking Process
The production of methanol is positively associated with the enzyme activities of PME and
PL in fruit juices (Hou and others 2008). PMEs are found to be inhibited by three kinds of
inhibitors from fruits and vegetables as follows: 1) some polysaccharides (about 200 kDa)
in potato that are mainly consisted of uronic acids (McMillan and Prombelon 1995); 2) a
glycoprotein from kiwi fruit that displays remarkable inhibition on PMEs from tomato,
orange, apple, banana and potato sources (Balestrieri and others 1990); 3) thermally stable
PME inhibitor (PMEI) from jelly fig (a woody vine in Taiwan) is able to reduce the
activities of PMEs from citrus, jelly fig achenes, tomato, apple, asparagus and guava (Jiang
and others 2002b) and banana, and especially the rubbery fruit which acts against banana
and pea pod PMEs (Wu and others 2002, 2005). Consequently, the discussion of this study
incorporates three methods of reducing methanol in winemaking process. The two of them
are PME inhibition methods for the utilization of CPE and the other one is the PL
application method in which the commonly used CPE is replaced by the PL.

The addition of PMEI from jelly fig with CPE for methanol reduction
The most promising application of PMEI is that PMEI can be added into the fresh fruit
juice prior to very mild thermal treatment in the production of minimally processed juice at
a quality level unachievable by the conventional pasteurization process (Castaldo and
others 1991; Jiang and others 2002a). Wu and others (2005) also reported that PMEI can
be applied as food additive in winemaking industry in order to reduce the methanol content
in fruit wine.

Although methanol content of carambola wines increased slowly reaching 250 ppm at day
14 fermentation without the addition of pectic enzymes, in the presence of added pectic
enzymes methanol increased rapidly and reached 400 ppm which is equivalent to 3333
mg/L ethanol, a value higher than the legal standard limit posing toxicity. The addition of
PMEI from jelly-fig extract to fruits must reduced methanol content amount about 58 ppm
which is equivalent to 483 ppm/L of ethanol. Thus, the addition of PMEI from jelly-fig
extract to fermentation process has a great advantage to food processing industry in
reducing methanol production.

The addition of phenolic acids with CPE for methanol reduction
Five different kinds of phenolic acids such as gallic acid, coumaric acid, caffeic acid,
cinnamic acid, and ferulic acid were found to inactivate PME (Hou and others 2008).
Enzyme activities of PME and PL within CPE decreased significantly when 0.1 mg/L of
The 12
th
16 -18 June, 2011

306

these acids were added and produced more effect on PME. These phenolic acids, however,
exert no inhibition on PG activity even when 0.5 mg/L of phenolic acids were added.

As expected from the inhibitory effect on PME, the addition of phenolic acids with CPE to
decrease methanol production in fruit wines occurred. Methanol content of grape puree
increased gradually to a plateau at day 6 fermentation. After 15-d fermentation, the
methanol level reached 30.2 ppm in the control group while the methanol level increased
to 56.67 ppm in the CPE added group. The group with CPE and gallic acid or the group
with CPE and coumaric acid had methanol profile similar to the control group, being 807
mg/L and 795 ppm, respectively for the gallic and coumaric acid groups. Furthermore,
addition of 0.2 mg/L gallic acid or coumaric acid caused an increase in total phenolic
content from 197444 in CPE adding group to 2440255 mg/L in CPE/gallic acid group
and 2513104 mg/L in CPE/coumaric acid group (Hou and others 2008).

Phenolic compounds have been an intense discussion area for research in recent years due
to their potential beneficial effects on human health. Moreover, they have been widely
studied for their strong antioxidant capacity and cell functions regulation (Rasool and
others 2010). Phenolic acids such as gentisic acid, gallic acid, ferulic acid, and coumaric
acid are reported to be able to significantly induce phase II hepatic antioxidant enzyme and
increase the antioxidant status of liver (Yeh and Yen 2006). The phenolic compounds are
secondary plant metabolites that exist in the skin, seed, and flesh of grapes and are
extracted into wines (especially red) during the process of vinification (Vias and others
2000). Total polyphenolic extracts from red wine are found to have some inhibitory effect
on reducing the 1,2-dimethylhydrazine-induced colon carcinogenesis in F344 rats (Femia
and others 2005).

In relation to gallic acid which is found in several different vegetable sources is an
important natural antioxidant and possesses anti-allergic, anti-inflammatory, anti-
mutagenic and anti-carcinogenic activity (Lin and others 1998; Fukumoto and others
2000). Regarding coumaric acid, it is reported to exhibit an effective decrease in the
oxidative DNA damage in rat colonic mucosa (Guglielmi and others 2003) and cytotoxcity
to nine human cancer cells such as K562, MGC-803, HL60, SH-SY5Y, SW1116, SMMC-
7221, SW480, HepG2 and KB (Yao and others 2010). Gallic acid and coumaric acid were
also found to enhance the antioxidant capacity of wine while they were added in the
winemaking process (Hou and others 2008). Therefore, gallic acid or coumaric acid
addition in winemaking process is a potential method for reducing methanol content in
order to improve wine quality.

The addition of PL in place of CPE for methanol reduction
Although CPE plays an important role in the winemaking process for extraction,
clarification, and filtration of fruit juice and wine must to increase the yield and quality of
fruit juice and wine in the characteristics such as pigment, flavor, transmittance, and
viscosity (Lao and Lopez-Tamames 1999; Wu and others 2007). In comparison to the
control group, 5.0 fold and 2.8 fold increase in the methanol content of wine after 15 days
of fermentation were found when CPE and PME/PG were added to the grape musts during
winemaking. Nevertheless, the addition of PL (isolated from CPE) to grape musts does not
cause an increase in methanol content (Wu and others 2007). The reduction in PG activity
The 12
th
16 -18 June, 2011

307

was also reported to cause a decrease in the methanol content of wine. That is, the
cooperation of PG with PME will greatly increase the production of methanol.

Liu (2009) found that pectin in alcohol insoluble solid (AIS) and high methoxyl pectin in
AIS (HM-AIS) can inhibit PME activity without inhibition of PL and PG activities.
Therefore, five methanol reduction methods including CPE/AIS, CPE/HM-AIS, CPE/jelly-
fig PMEI, PL (the same PL activity within CPE), and CPE/centrifugation (CPE reacts for
60 min and then centrifuged at 3,500 g for 10 min, 4
o
C) for strawberry winemaking,
made from strawberry juice, were analyzed for the methanol reducing ability in
comparison to control group (adding CPE). All of the five methods expressed methanol
reduction ability in strawberry wine. Adding PL yielded the best methanol reduction with a
result of 84.3 g methanol/mL wine, 31.6% of the control group (267.016.8 g/mL). The
significant outcome is followed by CPE/centrifugation 65.1 g/mL (24.4%), CPE/jelly-fig
PMEI 45.9 g/mL (17.2%), CPE/HM-AIS 26.7 g/mL (10.0%), and CPE/AIS 22.7 g/mL
(8.5%) (Liu 2009). Namely, about 31.6% reduction in methanol content of wine was found
when replacing CPE with PL in strawberry winemaking. The removing of CPE containing
particles by centrifugation that formed after CPE reaction exhibited 24.4% inhibition in
methanol production. Addition of PME inhibitor including AIS, HM-AIS, and jelly-fig
PMEI exhibited 8.5% to 17.2% reduction in methanol content.

Regarding CPE/centrifugation method, it revealed 24.4% reduction in methanol content yet
the yield of strawberry wine was 71% (Liu 2009). No significant reduction in the yield of
wine was observed between PL group and CPE group (control).

In other words, some methanol was produced in wine of the PL-addition groups than CPE-
added group due to the elimination of PME activity from CPE. Excluding methanol
content, the use of PL for wine fermentation resulted in better physicochemical properties
such as transmittance, lightness, and redness as that prepared in the control group (Wu and
others 2007; Liu 2009). Therefore, adding CPE without PME or using PL in place of CPE
has a great potential advantage in manufacturing fruit wines in terms of both clarity and
safety of wines.

Conclusion
There are reasons for CPE to play an important role in fruit juice and wine manufacturing
industries: (1) the addition of CPE to fruits and fruit during maceration improves extraction
of fruit juice, color, and flavor; (2) CPE shortens the time required for filtration and
clarification of fruit juice and wines. However, in both merits CPE could also increase
methanol production to the excess of legal standard limit.
In reviewing the methods for methanol reduction, it is found that the addition of PMEI,
gallic acid or coumaric acid with CPE or using PL for CPE replacement are effective in
reducing methanol content of fruit wines. Therefore, it is expected for winemaking
industry to utilize the above methods in manufacturing process for the purpose of reducing
methanol content in fruit wines and spirits in the future.

The 12
th
16 -18 June, 2011

308

Reference
Azmak D. 2006. Methanol related deaths in Edirne. Leg Med 8:39-42.
Bai ZH, Zhang HX, Qi HY, Peng XW, Li BJ. 2004. Pectinase production by Aspergillus
niger using wastewater in solid state fermentation for eliciting plant disease
resistance. Bioresour Technol 95:49-52.
Balestrieri C, Castaldo D, Giovane A, Quagliuolo L, Servillo L. 1990. A glycoprotein
inhibitor of pectin methylesterase in kiwi fruit (Actinidia chinensis). Eur J Biochem
193:183-187.
Blanco P, Sieiro C, Diaz A, Reboredo NM, Villa TG. 1997. Grape juice biodegradation by
polygalacturonases from Saccharomyces cerevisiae. Int Biodeterior Biodegradation
40:115-118.
Cabaroglu T. 2005. Methanol contents of Turkish varietal wines and effect of processing.
Food Control 16:177-181.
Castaldo D, Lovoi A, Quagliuolo L, Servillo L, Balestrieri C, Giovane A. 1991. Orange
juices and concentrates stabilization by a proteic inhibitor of pectin methylesterase.
J Food Sci 56:1632-1634.
Fachin D, Van Loey AM, Nguyen BL, Indrawati IV, Hendrickx ME. 2003. Inactivation
kinetics of polygalacturonase in tomato juice. Innov Food Sci Emerg Technol 4:135
142.
Fawole OB, Odunfa SA. 2003. Some factors affecting production of pectic enzymes by
Aspergillus niger. Int Biodeterior Biodegradation 52:223-227.
Femia AP, Caderni G, Vignali F, Salvadori M, Giannini A, Biggeri A, Gee J, Przybylska
K, Cheynier V, Dolara P. 2005. Effect of polyphenolic extracts from red wine and 4
OH-coumaric acid on 1,2-dimethylhydrazine-induced colon carcinogenesis in rats.
Eur J Nutr 44:79-84.
Frenkel C, Peters JS, Tieman DM, Tiznado ME, Handa AK. 1998. Pectin methylesterase
regulates methanol and ethanol accumulation in ripening tomato (Lycopersicon
esculentum) fruit. J Biol Chem 273:4293-4295.
Fukumoto LR, Mazza G. 2000. Assessing antioxidant and prooxidant activities of phenolic
compounds. J Agric Food Chem 48:3597-3604.
Guglielmi F, Luceri C, Giovannelli L, Dolara P, Lodovici M. 2003. Effect of 4-coumaric
acid and 3,4-dihydroxybenzoic acid on oxidative DNA damage in rat colonic
mucosa. Br J Nutr 89:581-587.
Gummadi SN, Panda T. 2003. Purification and biochemical properties of microbial
pectinases-a review. Process Biochem 38:987-996.
Hou CY, Lin YS, Wang YT, Jiang CM, Wu MC. 2008. Effect of storage conditions on
methanol content of fruit and vegetable juices. J Food Comp Anal 21:410-415.
Jayani RS, Saxena S, Gupta R. 2005. Microbial pectinolytic enzymes: A review. Process
Biochem 40:2931-2944.
Jiang CM, Li CP, Chang HM. 2002a. Influence of pectinesterase inhibitor from jelly fig
(Ficus awkeotsang Makino) achenes on pectinesterases and cloud loss of fruit
juices. J Food Sci 67:3063-3068.
Jiang CM, Li CP, Chang JC, Chang HM. 2002b. Characterization of pectinesterase
inhibitor in jelly fig (Ficus awkeotsang Makino) achenes. J Agric Food Chem
50:4890 4894.
Lamiable D, Hoizey G, Marty H, Vistelle R. 2004. Acute methanol intoxication. EMC
Toxicol Pathol 1:7-12.
The 12
th
16 -18 June, 2011

309

Lao C, Lopez-Tamames E. 1996. Grape pectic enzyme treatment effect on white musts and
wines composition. J Food Sci 61:553-556.
Lima AS, Alegre RM, Meirelles AJA. 2002. Partitioning of pectinolytic enzymes in
polyethylene glycol potassium phosphate aqueous two-phase systems phosphate
aqueous two-phase systems. Carbohydr Polym 50:63-68.
Lin JK, Lin CL, Liang YC, Lin-Shiau SY, Juan IM. 1998. Survey of catechins, gallic acid,
and methylxanthines in green, oolong, pu-erh, and black teas. J Agric Food Chem
146:3635-3642.
Liu WT. 2009. Effect of different enzyme treatments on the reduction of methanol content
in strawberry wine. [Masters thesis], Pingtung: National Pingtung University of
Science and Technology. 106 p. Available from: National Digital Library of Theses
and Dissertations in Taiwan, Taipei, Taiwan.
Massiot P, Le Qur JM, Drilleau JF. 1994. Biochemical characteristics of apple juices and
fermented products from musts obtained enzymatically. Fruit Process 4:108-113.
Massiot P, Perron V, Baron A, Drillean JF. 1997. Release of methanol and
depolymerization of highly methyl esterified apple pectin with an
endopolygalacturonase from Aspergillus niger and pectin methylesterases from A.
niger from orange. LWT-Food Sci Technol 30:697-702.
McMillan GP, Prombelon MCM. 1995. Purification and characterization of a high pI
pectin methylesterase isozyme and its inhibitor from tubers of Solanum tuberosum
subsp. tuberosum cv. Katahdin. Physiol Mol Plant Pathol 46:413-427.
Moyo S, Gashe BA, Collison EK, Mpuchane S. 2003. Optimising growth conditions for
the pectinolytic activity of Kluyveromyces wickerhamii by using response surface
methodology. Int J Food Microbiol 85:87-100.
Muoz O, Seplvedab M, Schwartzb M. 2004. Effects of enzymatic treatment on
anthocyanic pigments from grapes skin from chilean wine. Food Chem 87:487-490.
Nedjma M, Hoffmann N, Belarbi A. 2001. Selective and sensitive detection of pectin lyase
activity using a colorimetric test: application to the screening of microorganisms
possessing pectin lyase activity. Anal Biochem 291: 290-296.
Nesime Y, Hasan A, Ahmet T, Sermet K. 2003. Fatalities due to methyl alcohol
intoxication in turkey: an 8-year study. Forensic Sci Int 131:36-41.
Newsome R. 1993. Sugar substitutes. In: Altschul AM, Editor. Lowcalorie foods
handbook. Vol 8. New York: Marcel Dekker Inc. p 139-170.
Northcote DH. 1972. Chemistry of the plant cell wall. Annual Review: Plant Physiol
23:113-132.
Oakenfull D, Sctt A. 1984. Hydrophobic interaction in the gelation of high methoxyl
pectins. J Food Sci 49:1093-1098.
Paine A, Davan AD. 2001. Defining a tolerable concentration of methanol in alcoholic
drinks. Hum Exp Toxicol 20:563-568.
Plaziac C, Lachapelle P, Casanova C. 2003. Effects of methanol on the retinal function of
juvenile rats. Neurotoxicology 24:255-260.
Rasool MK, Sabina EP, Ramya SR, Preety P, Patel S, Mandal N, Mishra PP, Samuel J.
2010. Hepatoprotective and antioxidant effects of gallic acid in paracetamol-
induced liver damage in mice. J Pharm Pharmacol 5: 638-643.
Revilla I, Gonzlez-SanJos ML. 1998. Methanol release during fermentation of red grapes
treated with pectolytic enzymes. Food Chem 63:307-312.
The 12
th
16 -18 June, 2011

310

Servili M, Begliomini AL, Montedoro G, Petruccioli M, Fedrerici F. 1992. Utilization of a
yeast pectinase in olive oil extraction and red wine making processes. J Sci Food
Agric 58:253-260.
Silvia AC, Ruth MA, Navarro DCA, Osvaldo C. 1996. Study of variables involved in
fungal pectic enzyme fractionation by immobilized metal ion affinity
chromatography. Process Biochem 31:81-87.
Vias P, LpezErroz C, Marin-Hernndez JJ, Hernndez-Cordoba M. 2000. Determination
of phenols in wines by liquid chromatography with photodiode array and
fluorescence detections. J Chromatogr A 871:85-93.
Wang ML, Wang JT, Choong YM. 2004. A rapid and accurate method for determination
of methanol in alcoholic beverage by direct injection capillary gas chromatography.
J Food Compost Anal 17:187-196.
Wu JS, Wu MC, Jiang CM, Hwang YP, Shen SC, Chang HM. 2005. Pectinesterase
inhibitor from jelly fig (Ficus awkeotsang Makino) achenes reduces methanol
content in carambola wine. J Agric Food Chem 53:9506-9511.
Wu MC, Tseng KC, Huang TH, Chang HM. 2002. Pectinesterase inhibitor in rubbery
banana (Musa sapientum L.). J Food Sci 67:1337-1340.
Wu MC, Jiang CM, Huang PH, Wu MY, Wang YT. 2007. Separation and utilization of
pectin lyase from commercial pectic enzyme via highly methoxylated cross-linked
alcohol-insoluble solid chromatography for wine methanol reduction. J Agric Food
Chem 55:1557-1562.
Yao H, Li S, Hu J, Chen Y, Huang L, Lin J, Li G, Lin X. 2010. Chromatographic
fingerprint and quantitative analysis of seven bioactive compounds of scutellaria
barbata. Planta Med [Epub ahead of print]
Yeh CT, Yen GC. 2006. Induction of hepatic antioxidant enzymes by phenolic acids in rats
is accompanied by increased levels of multidrug resistance-associated protein 3
mRNA expression. J Nutr 136:11-15.
Zoccaa F, Lomolinoa G, Curionia A, Spettolia P, Lante A. 2007. Detection of pectin
methylesterase activity in presence of methanol during grape pomace storage. Food
Che 102:59-65.
The 12
th
16 -18 June, 2011

311

OH-314

The Kinetics of Virgin Coconut Oil Deterioration during Storage at
Different Temperatures

Ambagan, Norberto G.
*

Food Processing Division, Industrial Technology Development Institute (ITDI)
Department of Science and Technology (DOST), PHILIPPINES
*
Corresponding author: norberto.ambagan@gmail.com

Abstract

The chemical kinetics of virgin coconut oil (VCO) deterioration during storage at different
temperatures (40
O
C, 50
O
C and 60
O
C) was determined. Five (5) samples of VCO from
different production methods (fermentation with heat, fermentation without heat,
centrifuge, dry-process and enzymatic) were collected. The moisture content of the VCO
samples were altered into three levels: low (L), moderate (M) and high (H). The free fatty
acid (FFA) and peroxide value (PV) of the original (O) and moisture-altered (L, M, H)
samples stored different temperatures (40
O
c, 50
O
C and 60
O
C) conditions were evaluated on
scheduled intervals. The kinetics of free fatty acid formation follows a first-order model
while that of peroxide formation follows a zero-order model. The results showed that the
rate of free fatty acid formation (k
FFA
) increases as MC and T
s
increases while the rate of
peroxide formation (k
PV
) increases only as T
s
increases and not affected by changes in MC.
Based on these results, an empirical model for the combined effects of initial MC (MC
i
),
initial FFA (FFA
i
) and storage temperature (T
s
) on k
FFA
was derived. This model can be
used to predict the shelf-life (based on FFA) of the VCO stored at a given condition (T
s
)
with known values of MC
i
and FFA
i
.

Keywords: Virgin coconut oil, kinetic, deterioration, storage temperature

The 12
th
16 -18 June, 2011

312

OJ : Trends in Food Research : ASEAN Perspective
(Graduate Papers)
The 12
th
16 -18 June, 2011

313

OJ-40

Changes of Physiochemical and Enzyme Quality in Longkong Fruit
during Four Different Weeks of on Tree Maturation

Karthikeyan Venkatachalam, Mutita Meenune
*

Department of Food Technology/Postharvest Technology Innovation Center, Faculty of Agro-
Industry, Prince of Songkhla University, Hat Yai, Thailand
*
Corresponding author: mutita_meenune@yahoo.co.uk, Phone: +66-7428-6335, Fax: +66- 7455-8866

Abstract

Longkong (Aglaia dookkoo Griff.) is a non climacteric and tropical fruit and majorly
cultivated in peninsula of Thailand. The small and round shaped longkong has a faintly
aromatic smell with sweet and sour taste. Generally the optimum harvesting period of
longkong is started from 13 to 16 weeks of after anthesis. Changes in physiochemical and
enzyme quality were measured during 4 different weeks of on tree maturation. The colour,
such as lightness (L*), redness (a*) and yellowness (b*) were measured. The a* and b*
were significantly affected (p<0.05) but the L* was affected non-significantly (p>0.05).
The L* and b* were decreased and conversely a* was increased during on tree maturation.
Weight of the fruit was increased throughout the stage, at the end it was 24.93 g (p<0.05)
and the changes in diameter also were found (p>0.05). Chemical quality such as total
soluble solids (TSS), pH, titratable acidity (TA) and total sugar (TS) and reducing sugar
(RS) was significantly affected throughout the period (p<0.05). TSS, pH, TS and RS was
increased while TA was found to be decreased. Poly phenol oxidase (PPO), peroxidase
(POD) and phenyl alanine ammonia lyase (PAL) enzymes were measured and they were
significantly affected throughout the optimum period (p<0.05). PPO was increased
throughout the maturation. POD and PAL were increased at 13
th
and 14
th
weeks of
maturation and after that it decreased.

Keywords: Longkong, maturation, physiochemical, enzyme, quality

The 12
th
16 -18 June, 2011

314

OJ-45

Impact of Some Additives on the Stability of Microbial Transglutaminase
from Providencia sp. C1112

Chaiwut Bourneow
1
, Soottawat Benjakul
1
, Aran H-Kittikun
2,*

1
Department of Food Technology, Faculty of Agro-Industry;
2
Department of Industrial Biotechnology,
Faculty of Agro-Industry, Prince of Songkhla University, Hat Yai, Thailand
* Corresponding author: aran.h@psu.ac.th

Abstract

Microbial transglutaminase (MTGase) is a transferase that has been applied to many
protein containing foods to improve their texture. The stability is an important feasibility of
enzyme to achieve the economic trading. Providencia sp. C1112 was cultivated to produce
MTGase in SPY medium at 37C, 150 rpm for 24 h. MTGase from the culture supernate
was prepared by precipitation (50-80% (NH
4
)
2
SO
4
), dialysis and ion exchange
chromatography (SP-Sepharose). The MTGase was formulated with different additives
including glycerol, maltodextrin, sorbitol and xylitol. Glycerol and maltodextrin showed
the effective performance to stabilize the MTGase activity at 20% (w/v) during storage at
37C. Twenty % Glycerol with 0.2% glutathione showed the most effectiveness on
stabilizing of MTGase during storage at 4C for 3 months.

Keywords: Glutathione, microbial transglutaminase, maltodextrin, Providencia sp.,
stabilizer

Introduction
Enzymes are biocatalysts that have been important used in many fields. Enzymes may be
easily denatured by change of the environmental conditions (Michiaki and others 1997).
Therefore, the resistance of enzyme toward extreme deterioration is one of the most
important criteria for commercialization and industrial application (Renate and Ulrich
1999; Daumantas and others 1999).

The stabilization of enzyme could be retained by addition of some protective agents such
as salts, polyethylene glycol, glycerol, dextran, bovine serum albumin, polyethylenimine,
osmolytes (e.g. betaine), organic solvents (isoctane, cyclohexane, chloroform, benzene)
and sugars (e.g. mannitol, sorbitol, xylitol, inositol, erythritol) (Noriko and others 1999;
Yuji and others 1984; Lozano and others 1994; Misset 1992; Combes and others 1987).
The selection of the appropriate additive depends on the nature of the enzyme (Costa and
others 2002).

Microbial transglutaminase (MTGase, EC 2.3.2.13) is known as the amine transferase or
protein glutamine -glutamyl transferase (Motoki and Seguro 1998). MTGase catalyzes the
formation of isopeptide by the exchange of an acyl group from -carboxyamide (-
C(O)NH
2
) group of proteins containing glutamine residue, an acyl donor, to a variety of
acyl acceptor which results in cross-linking of protein (Motoki and Seguro 1998). MTGase
has captured interest due to its attractive potential application in food industries (Motoki
and Seguro 1998), immobilization of enzymes (Josten and others 1999) and textile
industries (Cortez and others 2004).
The 12
th
16 -18 June, 2011

315

Screening and production of MTGase are important studies for investigation of the novel
MTGase. MTGase was reported in cultures of Streptoverticillium mobaraense and
Streptomyces sp. (Neilson 1995). MTGase could be highly produced by Sv. ladakanum
(Ashie and Lanier 2000), Sv. cinnamoneum (Duran and others 1998), and S. lividans (Lin
and others 2006). Ajinomoto Co., Inc., produces the MTGase not only from
Streptoverticillium species but also from other genus such as Bacillus (Barros and others
2003; Zhu and others 1996). Providencia sp. C1112 was classified nearly to P. stuartii, a
bacteria in the biosafety level 1 (Li and others 2007). Providencia sp. C1112 was isolated
from wastewater pond of seafood factories in Songkhla, Thailand. Providencia sp. C1112
possessed as a dominant MTGase producer.

However, preservation of the enzymes during storage is also important. Therefore, this
study aimed to investigate the effect of protective agents on preservation of the MTGase
activity produced by Providencia sp. C1112 during storage.

Preparation of MTGase
Providencia sp. C1112 (NBRC106720) was cultivated in SPY medium (2.0% starch, 2.0%
peptone, 0.2% yeast extract, 0.2% Mg
2
SO
4
, 0.2% KH
2
PO
4
and 0.2% K
2
HPO
4
, pH 6.5) and
incubated at 37C, 150 rpm for 24 h. The cell pellets were removed by centrifugation at
7,500g for 10 min at 4C. The supernatant was precipitated using 50-80% (NH
4
)
2
SO
4

stepwise-precipitation, dialyzed in 200 mM citrate-phosphate buffer (pH 6.5) and eluted
through SP-Sepharose using 0.2-0.4 M NaCl in 200 mM citrate-phosphate buffer (pH 6.5).
The obtained MTGase was dialyzed, freeze dried and diluted in 200 mM citrate-phosphate
buffer (pH 6.5) to 2 U/ml.

Effect of some additives on MTGase stability
MTGase solution was incorporated to glycerol, maltodextrin, sorbitol and xylitol at various
concentrations in the presence and absence of 0.2% glutathione in 200 mM citrate-
phosphate buffer (pH 6.5) and kept at 37C for 12 h. The samples were taken to monitor
the residual MTGase activity using hydroxamate assay (Folk and Cole 1965). The most
effective additives were chosen for long term preservation of MTGase.

Effect of glycerol and maltodextrin on long term stability of MTGase
MTGase stabilized in (i) the liquid form of glycerol (20% w/v) and (ii) the freeze-dried
powder form of maltodextrin (20% w/v) with and without glutathione (0.2% w/v) was
prepared. The prepared enzyme was kept at 4C and the residual MTGase activity was
monitored during storage for 90 days.

Determination of MTGase activity
The sample (100 l) was mixed with 200 l of 200 mM citrate buffer (pH 6.0), 25 l of
125 mM hydroxylamine, 25 l of 12.5 mM glutathione and 75 l of 37.5 mM CBZ-Gln-
Gly. The reaction mixture was incubated at 37C for 1 h. Thereafter, an equal volume of
5% ferric chloride in 15% TCA solution was added to the reaction mixture and centrifuged
at 7,500 g for 10 min. The absorbance of the supernatant was measured at 525 nm. One
unit of MTGase activity was defined as the amount of enzyme needed to produce 1 mol
of hydroxamic acid per min (Folk and Cole 1965)

The 12
th
16 -18 June, 2011

316

Each experiment was set in duplicate and the sample was collected in triplicate for
analysis. The data was analyzed using the analysis of variance (ANOVA) and mean
comparison was run by Duncans multiple range test using a SPSS package (SPSS 10.0 for
windows, SPSS Inc, Chicago, IL).

Glycerol, maltodextrin, sorbitol, and xylitol were often used to preserve proteins and
enzymes. Table 1 showed the effectiveness of these additives on stabilization of MTGase
solution from Providencia sp. C1112 during storage at 37C for 12 h. Glycerol and
maltodextrin at 20% could stabilize MTGase activity higher than sorbitol and xylitol at all
concentrations.

Table 1 Effect of some additives on the residual relative MTGase activity after incubating at 37C, 12 h
Relative activity (%)
Additive
2.5% (w/v) 5% (w/v) 10% (w/v) 20% (w/v)
Glycerol 96.26+1.67
abB
92.83+2.97
aB
99.66+1.74
bB
103.64+3.83
cC

Maltodextrin 104.67+2.21
aC
103.08+3.01
aC
104.31+5.69
abB
109.30+1.23
bD

Sorbitol 72.90+1.45
aA
77.26+5.31
abA
90.03+2.69
cA
77.26+2.31
bA

Xylitol 75.70+2.21
aA
77.76+2.58
aA
90.34+3.38
bA
95.64+2.02
bB

Control* 40.02+1.20
The different superscripts in the same row (a-c) and column (A-D) denoted the significant different (P<0.05)
* Citrate-phosphate buffer

Moreover, the addition of 0.2% glutathione could improve the effectiveness of sorbitol and
xylitol as showed in Table 2. However, addition of glutathione to maltodextrin or glycerol
could not help to increase MTGase activity. Based on the residual MTGase activity,
therefore, glycerol and maltodextrin were used for further long term preservation of
MTGase at 4C.

Table 2 Effect of glutathione on the residual relative MTGase activity after incubating at 37C, 12 h
Relative activity (%)
Additive
2.5% (w/v) 2.5% (w/v) 2.5% (w/v) 2.5% (w/v)
Glycerol 94.66+2.53
aB
97.26+2.86
aC
98.84+1.28
bB
109.62+1.31
bD

Maltodextrin 101.38+2.54
aC
99.48+1.91
aC
105.81+1.24
bC
102.01+1.25
aC

Sorbitol 83.00+2.53
aA
86.17+0.65
aB
93.95+3.48
bA
93.77+0.65
bA

Xylitol 80.22+0.40
aA
80.47+1.27
aA
93.32+1.57
bA
97.26+1.59
cB

Control* 40.87+0.95
0.2% glutathione 54.81+0.81
The different superscripts in the same row (a-c) and column (A-D) denoted the significant different (P<0.05)
* Citrate-phosphate buffer

During long term storage of MTGase at 4C, the addition of glutathione could synergist to
20% glycerol and maltodextrin (in the powder form) for retaining of MTGase activity. The
addition of 0.2% glutathione with glycerol or maltodextrin could increase the MTGase
stability during storage (Figure 1). However, preservation of MTGase in glycerol although
without glutathione was better than maltodextrin with glutathione.
The 12
th
16 -18 June, 2011

317

Figure 1 The protective effect of maltodextrin and glycerol with glutathione on retained MTGase
activity during storage at 4C.

The result suggested that MTGase could preserved in 20% glycerol with 0.2% glutathione
which could retained MTGase nearly 80% of the initial activity at 4C for 3 months.

Acknowledgement
The authors are grateful to the Graduate School, Prince of Songkhla University for their
financial support.

References
Ashie INA, Lanier TC. 2000. Transglutaminase Seafood Processing. New York, USA:
Marcel Dekker Inc. p. 344.
Barros SLH, Assmann F, Zachia AMA. 2003. Purification and properties of a
transglutaminase produced by a Bacillus circulans strain isolated from the Amazon
environment. Biotechnol. Appl. Biochem. 37:295-299.
Combes D, Yoovidhya T, Girbal E, Willemot RM, Monsan P. 1987. Mechanism of
enzyme stabilization. Annu. N. Y. Acad. Sci. 501:59-62.
Cortez J, Bonner PLR, Griffin M. 2004. Application of transglutaminases in the
modification of wool textiles. Enzym Microb. Technol. 34:64-72.
Costa SA, Tzanov T, Carneiro AF, Paar A, Gbitz GM, Cavaco-Paulo A. 2002. Studies of
stabilization of native catalase using additives. Enzym Microb. Technol. 30:387-
391.
Daumantas M, Charles W, Tong VP, Chandra G, Rex L. 1999. Protection of enzymes by
aromatic sulfonates from inactivation by acid, and elevated temperatures. J. Mol.
Catal. B: Enzym.:21-36.
Duran R, Jungua M, Schmitter JM, Gancet C, Gaulas P. 1998. Purification,
characterisation, and gene cloning to transglutamianse from Streptoverticillium
cinnamoneum CBS 683.68. Biochimie 80:313-319.
Folk JE, Cole PW. 1965. Structural requirements of specific substrates for guinea pig liver
transglutaminase. J. Biol. Chem. 240:2951-2960.
R
e
l
a
t
i
v
e

a
c
t
i
v
i
t
y

(
%
)

The 12
th
16 -18 June, 2011

318

Josten A, Meusel M, Spencer F, Haalk L. 1999. Enzyme immobilization via microbial
transglutaminase: A method for the generation of stable sensing surface. J. Mol.
Catal. B: Enzym. 7:58-66.
Li XZ, Hauer B, Rosche B. 2007. Single-species microbial biofilm screening for industrial
applications. Appl. Microbiol. Biotechnol. 76:12551262.
Lin YS, Chao ML, Liu CH, Tseng M, Chu WS. 2006. Cloning of the gene coding for
transglutaminase from Streptomyces platensis and its expression in Streptomyces
lividans. Process Biochem. 41:519-512.
Lozano P, Combes D, Iborra JL. 1994. Effects of polyols on -chymotrypsin
thermostability a mechanistic analysis of the enzyme stabilization. J. Biotechnol.
35:9-18.
Michiaki M, Koji K, Kazuo K. 1997. Effects of polyols and organic solvents on
thermostability of lipase. J. Chem. Technol. Biotechnol. 70:188-192.
Misset O. 1992. Stability of industrial enzymes. Proceedings of an International
Symposium held in Maastricht, The Netherlands, pp. 22-25.
Motoki M, Seguro K. 1998. Transglutaminase and its use for food processing. Trends Food
Sci.Technol. 9:204-210.
Neilson PM. 1995. Reactions and potential industrial applications of transglutaminase.
Review of literature and patents. Food Biotechnol. 9:119-156.
Noriko M, Miok K, Tado K. 1999. Stabilization of L-ascorbic acid by superoxide
dismutase and catalase. Biosci. Biotechnol. Biochem. 63:54-57.
Renate UH, Ulrich AJM. 1999. The concept of the unfolding region for approaching the
mechanisms of enzyme stabilization. J. Mol. Catal. B: Enzym. 7:125-131.
Yuji I, Hidroyuki N, Kosunobu T, Takyuki Y, Anutosh RS, Yuji S. 1984. Ester synthesis
catalyzed by polyethylene glycol modified lipase in benzene. Biochem. Biophy.
Res. 122:845-850.
Zhu Y, Rinzema A, Tramper J, Bol J. 1996. Medium design based on stoichiometric
analysis of microbial transglutaminase production by Streptoverticillium
mobaraense. Biotechnol. Bioeng. 50:291298.

The 12
th
16 -18 June, 2011

319

OJ-46

Characterization of Coconut Milk Products by Electronic Nose and
Descriptive Sensory Analysis

Saowapark Wattanapahu, and Thongchai Suwonsichon
*

Department of Product Development, Faculty of Agro-Industry, Kasetsart University,
50 Phahol Yothin Road, Chatuchak, Bangkok 10900
*
Corresponding author: fagitcs@ku.ac.th

Abstract

Coconut milk is one of the most popular products in Thailand, South Asia and Pacific
region and exported to several regions. The quality attributes of coconut milk products are
aroma, flavor and taste which were characterized by chemical and sensory methods. The
objectives of investigations were to characterize aromatic properties of various coconut
milk products by descriptive sensory analysis and instrumental electronic nose (E-nose)
analysis as well as the comparison of classifying milk products based on their aromatic
properties from both analyses. Eighteen coconut milk samples consisted of two fresh, four
pasteurized, four UHT, five sterilized and three spray dried forms were collected and
determined. Result of descriptive sensory analysis by eight highly trained panelists from
KUSCR (Kasetsart University Sensory and Consumer Research unit) showed that there
were twelve attributes which contributed to overall coconut milk, overall sweet, freshness,
coconut oil, cooked, nutty, soya milk, vegetable oil, synthetic, metallic and rancid. All
twelve sensory attributes were significantly different amongst coconut milk samples
(p<0.05).While result of twelve metal oxide E-nose sensors demonstrated significantly
different among samples (p<0.05). Multivariate techniques which were the hierarchical
cluster analysis (HCA) and principal component analysis (PCA) were also applied to
analyze both data from sensory and E-nose as well as categorization. Results demonstrated
that PCA reduced twelve aromatic sensory attributes into two independent principal
components (PCs), which accounted for 62.62% of the explained variance. These two PCs
with cluster analysis could classify twelve samples into 5 groups. While PCA reduced
twelve metal oxide E-nose sensors into two independent principal components (PCs),
which accounted for 95.14% of the explained variance. When two new PCs and CA were
applied to E-nose data, samples were also categorized into 5 groups. Both categorizations
were in agreement and also related to tempering process of samples. Therefore, there is a
potential to use E-nose to analyze aromatic attributes of coconut milk in food industry and
its result is quite related to sensory analysis.

Keywords: Coconut milk, descriptive sensory analysis, electronic nose

Introduction
Aromatic property is one of the important factors of coconut milk qualities as well as,
consumer acceptance. Descriptive sensory analysis is an effective method to evaluate the
quality of aromatic attributes of this product. However, it is timeconsuming and requires
highly trained panelists. Recently, the electronic nose (E-nose), instrumental method, has
been used to measure the volatile compounds like human olfactory system. It is rapid
techniques detecting the aromatic compound and has been applied to study in food
The 12
th
16 -18 June, 2011

320

products such as wine (Berna and others 2008) and tea (Dulta, and others 2003). The
objectives of this study were to evaluate aromatic attributes of various coconut milk
products by descriptive sensory analysis and E-nose and to compare results of classifying
coconut milk products from both analyses.

Sample preparation: Eighteen available commercial samples consisted of two fresh (F1 to
F2), four pasteurized (P1 to P4), four UHT (U1 to U4), five sterilized (C1 to C5) and three
powder (S1 to S3) samples. They were purchased in 2010 from supermarkets located in
Bangkok, Thailand.

E-nose measurement: 2 ml of each sample were placed and sealed in 10 ml headspace vial.
Samples were incubated at 50 C and shake at 1000 rpm for 15 min before injecting 2500
l of the saturated headspace air into E-nose (Fox 3000, Toulouse, France). The carrier gas
flow rate was 300 ml/min and the runtime, purge time and delay time were 120 s, 180 s,
and 1080 s, respectively. Twelve metal oxide sensors of E-nose (LY2/G, LY2/gCT,
LY2/gCTL, LY2/LG, P10/1, P10/2, T30/1, T30/2, PA/2, LY2/AA, LY2/GH, and P40/1)
were used to detected volatile compounds. The changing ratios of conductivity between the
zero gas and volatile gas from 12 sensors of each samples were collected by Alpha soft
2004 version 9.0 (Alpha M.O.S., Toulouse, France) for further statistical analysis.

Sensory analysis: Eight highly trained panelists from Kasetsart University Sensory and
Consumer Research Center (KUSCR) participated in this study. 25 ml of each sample at 50
C were poured into plastic cup with a cover lid and immediately monadic served to
panelists. Panelists developed twelve aromatic attributes for describing the intensity of
aromatic compounds of coconut milk samples. These attributes were overall coconut milk,
overall sweet, freshness, coconut oil, cooked, nutty, soya milk, vegetable oil, synthetic,
metallic and rancid. The15 point interval scale with 0.5 increments, where 0 = attribute not
detected and 15 = attribute extremely strong was used in this study.

Data analysis: The principal component analysis (PCA) was applied to analyze and
establish the relationship of aromatic variables from sensory data and E-nose data. While
the hierarchical cluster analysis (HCA),Wards method, was used to identify clusters of
sample having the similar aromatic characteristics. All data analysis were performed by
the XLSTAT software version 7.0, Addinsoft, USA.

Results
Figure 1a showed the E-nose radar plot result of 12 sensors detecting the changing ratios
of conductivity of eighteen coconut milk samples. The LY2/LG, LY2/G, LY2/AA,
LY2/GH, LY2/gCTL, and PA/2 sensors were associated with ammonia, amine, HS,
alkane, hydrocarbon, and ester. The sensors showed significantly different among the
samples. PCA and HCA were used to analyze the ratios of conductivity data. Figure 1b
showed that the first two PCs could be explained 95.14 % of the total variation. PC1 was
accounted for 80.16 % of the variation and highly correlated to LY2/G, LY2/gCT,
LY2/gCTL, LY2/LG, P10/1, LY2/AA, T30/1, T30/2, and P40/1. The PC1 was named as
lack of freshness component. While the PC2 was accounted for 14.99 % of the variation
and highly correlated to PA/2, P10/2 and LY2/GH, It was referred as other oil impurities
component such as soy bean oil. Based on HCA, coconut milk samples could be
The 12
th
16 -18 June, 2011

321

categorized into five clusters as shown in Fig1. b. Classification among the cluster were
related to types and/or heating processes of coconut milk.

Figure 1 Radar plot from 12 sensors of electronic nose measurement (a) and bi plot of coconut milks
on PC axes using PC1 and PC2 for describe electronic nose sensors and grouping (b).

II
I
III
IV
V
b
The 12
th
16 -18 June, 2011

322

Figure 2 Bi plot of coconut milks on PC axes using PC1 and PC2 for describing sensory attributes
and grouping.

Figure 2 showed the bi-plot results of sensory data analysis. The first two PCs explained
62.62 % of the total variation. The PC1 was accounted for 44.17 % of the variation and
highly positive correlated to freshness, overall sweet, overall coconut milk, but negative
correlated to rancid, metallic, synthetic, nutty, and cooked aromatic. This PC1 was named
as degree of aromatic freshness. PC2 was accounted for 18.45 % of the variation and
highly correlated to high intensity of caramelized, coconut oil, vegetable oil, and soya milk
aromatics. It was named as caramelized and oily aroma. Result of cluster analysis of
sensory data as shown in Fig. 2 categorized samples into five clusters which also related to
their types and/or heating processes of coconut milk samples.

Discussion
Comparison results as shown Fig. 1 and 2, both categorization from E- nose and
descriptive sensory analysis were in agreement and also related to types and tempering
processes. Therefore, E-nose could be practical for quality control and product
development in coconut products manufacturing. It could provide the direction for
manufacturer monitoring the changing of volatile compounds in the tempering process or
during the storage time.

Reference
Berna AZ, Trowell S, Cynkar W, Cozzolino D. 2008. Comparison of metal oxide-based
electronic nose and mass spectrometry-based electronic nose for prediction of red
wine spoilage. J. Agri Food Chem. 56: 3238-44.
Dutta R, Hines EL, Gardner JW, Kashwan KR, Bhuyan A. 2003. Tea quality prediction
using o tin oxide-based electronic nose: an artificial intelligence approach. Sens
Actual. 94: 228-37.

The 12
th
16 -18 June, 2011

323

OJ-51

Characterization of Lactic Acid Bacteria and Coagulase Negative
Staphylococci Isolated From Raw Meat and Cured Meat Products

Nattaneewan Charoenrak, and Suree Nanasombat
*

Department of Biology, Faculty of Science, King Mongkuts Institute of Technology Ladkrabang,
Bangkok 10520, Thailand.
*
Corresponding author: Tel: 662-3298000 ext.6264; Fax: 662-3298412
E-mail address: knsuree@kmitl.ac.th

Abstract

The aims of this study were to screen lactic acid bacteria (LAB) and coagulase negative
staphylococci (CNS) from 50 samples of raw beef, raw pork, bacon, Nham and Saigoyisan
(Thai fermented pork sausages) to select the most suitable strains for use as meat starter
cultures. The 403 LAB and 250 Micrococcaceae isolates were characterized. The 184 LAB
isolates inhibited growth of at least one of 11 indicator organisms by agar spot test.
Twenty-five LAB isolates were nitrate reductase positive and amino acid decarboxylase
negative which may not accumulate biogenic amine in fermented foods. After further
characterization, three LAB isolates (the isolate S0602 isolated from Nham, the isolates
T0903 and T0904 isolated from Saigoyisan) with ability to produce bacteriocins and
highest resistance to lowest pH (pH 1.5) of hydrochloric and lactic acid and highest
concentration of sodium chloride (11% w/v) and bile salts (4-5% w/v) were identified as
Lactobacillus plantarum S0602, Lactobacillus brevis T0903 and L. brevis T0904. The only
one of 19 selected CNS isolates with catalase, nitrate and nitrite reductase positive was the
isolate SC0903 isolated from raw beef. This bacterium was amino acid decarboxylase
negative and was identified as Staphylococcus xylosus. Then, L. plantarum S0602, L.
brevis T0904 and S. xylosus SC0903 were selected to study the depletion of sodium nitrite
in culture media. Among all bacterial strains tested, L. plantarum S0602 caused the highest
decreasing of residual nitrite in the culture medium added with 1,000 mg/l during
fermentation at 37C.

Keywords: Coagulase negative staphylococci, lactic acid bacteria, Micrococcaceae,
cured meat product, starter culture

Introduction

Lactic acid bacteria (Lactobacillus, Pediococcus) and Micrococcaceae (Staphylococcus,
Micrococcus) are often used as meat starter cultures. Lactic acid bacteria (LAB) produce
inhibitory substances and improve sensory quality in final products (Leroy and others
2006). Coagulase-negative Staphylococcus (CNS) are members of family Micrococcaceae.
They are gram-positive, catalase-positive cocci (GCC) (Jay and others 2005) CNS are a
group of the most important microorganisms used as starter cultures. They play an
important role in development of color, flavor and aroma in cured meat products (Gtterup
and others 2007). Some Staphylococcus species are found as normal microflora in animals
(Jay and others 2005) S. saprophyticus and S. xylosus have been found as dominating CNS
in a southern Italian fermented sausage (Coppola and others 2000). They are commonly
used as commercial starters. Selection of suitable LAB and CNS strains are based on some
important criteria to be used as meat starter cultures such as ability to produce bacteriocins,
The 12
th
16 -18 June, 2011

324

nitrate reductase, nitrite reductase, inability to produce amino acid decarboxylase and some
probiotic properties. Therefore, the aims of this study were to screen, characterize and
identify new suitable LAB and CNS strains from raw meat and cured meat products to be
used as functional starter cultures.

Material and Methods
Physicochemical and microbiological analysis raw meat and cured meat products
Samples of raw beef, raw pork, bacon, Nham and Thai fermented sausages (Saigoyisan)
(10 samples each) were purchased in Bangkok and Samutprakarn, in Thailand. Water
activity (a
w
) and pH values of all samples were measured by using water activity meter
(Aqualab Series 3TE) and pH meter (Testo 205), respectively. Residual nitrite contents
were measured using to the method presented by Kirk and Sawyer (1991).
Total microbial and Micrococcaceae counts in all samples were determined by spiral
plating onto Plate Count Agar (PCA) and Mannitol Salt Agar (MSA) respectively. Total
LAB counts were determined by spread plating onto deMan Rogosa Sharpe agar (MRS,
Difco) containing 0.5% CaCO
3
. Typical colonies of Micrococcaceae on MSA (opaque
white or cream colonies) and LAB colonies with clear zone on MRS agar were randomly
picked and purified.

Screening of lactic acid bacteria
Inhibitory activities of all selected LAB isolates were assessed by agar spot test against 11
indicator organisms including Bacillus cereus DMST 5040, Escherichia coli DMST
42112, Listeria monocytogenes DMST 11256, Pseudomonas fluorescens DMST 20076,
Salmonella typhimurium DMST 0562, Staphylococcus aureus TISTR 118, Vibrio
parahaemolyticus OYI, Yersinia enterocolitica DMST 9380, Enterococcus faecium TISTR
1283, Lactobacillus plantarum TISTR 050 and Pediococcus acidilactici TISTR 051
(Schillinger and Lcke 1989). The isolates inhibiting at least one of indicator bacterium
were selected for characterizations such as tolerance of hydrochloric acid, lactic acid,
sodium chloride and bile salts (Chung and others 1999), production of catalase (Bennett
and Lancette 2001), nitrate reductase (Papamaloni and others 2002), nitrite reductase
(Salome and others 2009), amino acid decarboxylase (Bover-cid and Holzapfel 1999) and
bacteriocin production by agar well diffusion (Schillinger and Lcke 1989).

Screening of coagulase negative Staphylococcus
Differentiation of Staphylococci and Micrococci were performed by testing acid
production from glucose, acid production from glycerol, susceptibility to furazolidone and
lysostaphin resistance to bacitracin and oxidase production (Baker 1984). The isolates
which were considered as Staphylococcus could produce acid from glucose and glycerol.
They were oxidase negative and susceptible to furazolidone and lysostaphin, but not resist
to bacitracin. All Staphylococcus isolates were selected for coagulase test (Bennett and
Lancette, 2001), production of catalase (Bennett and Lancette 2001), nitrate reductase
(Papamaloni and others 2002), nitrite reductase (Salome and others 2009) and amino acid
decarboxylase (Bover-cid and Holzapfel 1999).

The 12
th
16 -18 June, 2011

325

Identification of lactic acid bacteria and coagulase negative Staphylococcus
All isolates were tested for the characteristics of gram reaction and cell morphology.
The selected LAB isolates were tested for carbon dioxide production from glucose and
growth at different conditions such as growth at 10C and 45C, growth in the presence of
6.5 % and 18% NaCl and growth at pH 4.4 and 9.6 to identify at genus level (Axelsson
2004). The selected LAB isolates were then characterized by their carbohydrate
fermentation patterns using the API-50 CH system (bioMrieux, France), while the CNS
isolates were identified at species level by the API Staph identification system
(bioMrieux, France).

Sodium nitrite depletion by selected lactic acid bacteria and coagulase negative
Staphylococcus isolates
This experiment was performed according to the method of Yan and others (2008). Briefly,
the selected LAB and CNS isolates (1x10
7
cells/ml) were inoculated into 10 ml MRS and
TSB broth containing 1000 mg/l of sodium nitrite, respectively. The MRS broth was
overlaid with paraffin oil. All samples were incubated at 37C for 4 days. The turbidity
(OD
620
nm), pH values and residual nitrite content (Kirk and Sawyer 1991) of all culture
media were measured at day 0, 1, 2, 3 and 4 of incubation.

Results
Physiochemical and microbiological characteristics of raw meat and cured meat
products
Raw meat samples had slightly acidic pH (5.30-5.98), while bacon, Nham and fermented
sausage samples had lower pH values (4.32-5.56). The pH values of raw pork samples
(5.30-5.52) were lower than those of raw beef samples (5.50-5.98). Among all samples,
raw beef and pork had the highest a
w
values (0.991-0.997). Residual nitrite contents in raw
meat and cured meat products were detected in acceptable range (0.71-23.26 mg/kg)
(Table 1). Total microbial counts and total Micrococcaceae counts in raw beef and pork
were found in higher number, compared to other samples. However, these raw meat
samples contained slightly lower LAB counts than those found in Nham and bacon (Table
1). The LAB isolates (403 isolates) and Micrococcaceae isolates (250 isolates) were
purified and stored at 4C for further characterization.

Table 1 Physiochemical characteristics of raw meat and cured meat products

Physicochemical characteristics Type of samples
(number of samples) a
w
pH residual nitrite (mg/kg)
Raw beef (10) 0.991-0.995 5.50-5.98 0.71-1.41
Raw pork (10) 0.992-0.997 5.30-5.52 0.35-1.06
Bacon (10) 0.979-0.989 4.98-5.56 0.35-23.26
Nham (10) 0.964-0.972 4.45-5.20 1.40-4.56
Fermented sausages (10) 0.958-0.975 4.32-4.85 1.80-4.93
Total (50) - - -

Raw meat had slightly acidic pH. This is probably due to accumulation of lactic acid as a
result of anaerobic glycolysis after slaughtering (Feiner 2006). Final pH values of meat
may correlate with glycogen contents. Blixt and Borch (2001) found that 100g pork loin
(pH 5.35) had higher glycogen content than 100g beef loin (pH 5.45). In this study, Nham
and fermented sausages samples had lower pH values because they probably contained
organic acids as fermentation products. Normally, carbohydrate added in fermented meat
products at the concentration of 0.5-0.7% can reduce pH values to lower than 5.0 (Toldr
and others 2001). In addition, the low a
W
value of nham and fermented sausage may be the
The 12
th
16 -18 June, 2011

326

result of water releasing during fermentation. Visessanguan and others (2006) reported that
nham samples fermented for at least 36 h had pH values of 4.6 and their weight decreased
due to releasing of water during fermentation. In this study, residual nitrite content in all
raw and cured meat samples was not exceeded the standard of the Ministry of Public
Health, Thailand (125 mg/kg). However, the department of disease control, Thailand
reported toxicity of sodium nitrite in patients with excessive intake of 3,137 mg/kg nitrite
in sausages. In this study, high microbial counts in raw pork and beef indicated poor
sanitary condition during slaughtering and carcass handling of raw meat.

Screening, biochemical characterization and identification of lactic acid bacteria
The 184 LAB isolates (45.66% of 403 isolates) had antibacterial activity by inhibiting at
least one of 11 indicator organisms (Table 2). The LAB isolates (25 of 184 isolates) which
were catalase and nitrate reductase positive, but amino acid decarboxylase negative were
selected to test the resistance to acid, sodium chloride and bile salts (Table 3). Ten LAB
isolates (40% of 25 isolates) had high ability to tolerate the lowest pH of HCl and lactic
acid at pH 1.5 and the highest concentration of sodium chloride (11%) and bile salts (5%).
They were selected to identify at genus level. These isolates were shown to be
Enterococcus (the isolates B0407, B0410, T0701, T0704 and T0901), Lactococcus (the
isolates B0708 and C0604) and Lactobacillus (the isolates S0602, T0903 and T0904).

Table 2 Microbiological characteristics of raw meat and cured meat products

Microbiological characteristics Type of
samples
(number
of samples)
Total
microbial
counts (CFU/g)
Total LAB
counts (CFU/g)
Selected
LAB
(isolates)
LAB with
antimicrobial
activity (isolates)
Total
Micrococcaceae
counts (CFU/g)
Selected
Micrococcaceae
(isolates)
Raw beef (10) 1.2x10
7
-1.8x10
9
7.8x10
5
-9.3x10
7
100 21 2.2x10
4
-9.6x10
6
80
Raw pork (10) 3.7x10
6
-1.7x10
9
7.0x10
5
-4.0x10
7
70 43 1.8x10
4
-1.0x10
7
90
Bacon (10) 2.7x10
6
-5.4x10
7
1.1x10
6
-1.2x10
7
100 40 <2,500 0
Nham (10) 3.2x10
7
-1.1x10
8
2.5x10
6
-6.7x10
8
70 47 5.0x10
3
-2.9x10
4
30
Fermented
sausages (10)
1.2x10
5
-1.6x10
7
1.7x10
4
-2.3x10
7
50 33 4.5x10
3
-1.2x10
4
50
Total (50) - 403 184 250

The selected LAB isolates could inhibit growth of indicator bacteria as they may produce
some inhibitory substances such as organic acids, bacteriocins, hydrogen peroxide and
others. In addition, these isolates could tolerate hydrochloric and lactic acid at low pH level
and high concentration of bile salts. The probiotic bacteria are known to have benefit to the
consumer health. They play an important role in balancing microflora in the gastrointestinal (GI)
tract. These bacteria have been reported to inhibit the growth of undesirable bacteria, reduce the
disease caused by infection in the GI tract, stimulate the body's immune system, inhibit growth of
cancer cells, and produce enzymes that are essential to the body. To gain these health benefits,
these bacteria need to survive passage through the GI tract. One of the important characteristics
of probiotic bacteria is their resistance to all harmful acid and bile salts in the gut. In this study, the
selected LAB which can tolerate to acid, NaCl and bile salts may be suitable for used as starter
cultures (Fuller 1989; Saarela and others 2000).

The 12
th
16 -18 June, 2011

327

Table 3 Hydrochloric acid, lactic acid, sodium chloride and bile salt tolerance of lactic acid bacteria
isolated from raw meat and cured meat products.

Source of
isolates
LAB isolates pH
a
of HCl

pH
a
of lactic
acid
NaCl

(%w/v)
b

Bile salts
(%w/v)
b

C0403, C0609 1.5 1.5 11.0 3.0
C0604 1.5 1.5 11.0 5.0
C0605 1.5 4.0 7.0 5.0
C0608 1.5 1.5 11.0 1.5
C0610 1.5 > 5.0 < 6.0 5.0
Raw beef
C0708 1.5 3.0 7.0 5.0
P0509 2.0 2.0 10.0 3.0
P0510, P0804,
P0806
1.5 1.5 11.0 5.0
Raw pork
P0706, P0708 2.5 2.0 10.0 2.0
B0407, B0410 1.5 1.5 11.0 5.0
B0708 1.5 1.5 10.0 2.0
B0709 2.5 1.5 10.0 2.0
B1002 2.0 1.5 10.0 2.0
Bacon
B1010 2.0 1.5 9.0 2.0
Nham S0602 1.5 1.5 11.0 4.0
T0701, T0901,
T0904
1.5 1.5 11.0 5.0
Fermented
sausage
T0704, T0903 1.5 1.5 11.0 5.0

a
The lowest pH of acid at which the isolates were able to grow.
b
The highest concentration of sodium chloride and bile salts at which the isolates were able to grow.

Screening, biochemical characterization and identification of coagulase negative
Staphylococcus
The results of differentiation tests of 250 Micrococcaceae isolates showed that 19 isolates
(7.6%) were Staphylococcus and 231 isolated (92.4%) were Micrococcus. All of
Staphylococcus isolates were CNS (Table 4).

Table 4 Differentiation of Staphylococcus and Micrococcus isolates in family Micrococcaceae

Number of isolates Source of
Isolates
Micrococcaceae
(isolates) Staphylococcus
a
Micrococcus
b

Raw beef 80 8 72
Raw pork 90 4 86
Bacon 0 0 0
Nham 30 4 26
Fermented sausage 50 3 47
Total 250 19 231
a
acid production from glucose and glycerol, furazolidone and lysostaphin susceptibility, bacitracin
resistance and oxidase negative
b
non acid production from glucose and glycerol, furazolidone and lysostaphin resistance, bacitracin
susceptibility and oxidase positive

These 19 Staphylococcus isolates were coagulase negative, but catalase and nitrate
reductase positive, while 18 isolates were amino acid decarboxylase positive. The only
Staphylococcus isolate which showed amino acid decarboxylase negative was the isolated
SC0903 (Table 5). These isolates were identified as S. xylosus (8 of 19 selected CNS
isolates) S. saprophyticus (2 isolates), S. lentus (2 isolates), S. cohnii ssp. urealyticum (3
isolates), S. cohnii ssp. cohnii (2 isolates), S. haemolyticus (1 isolate), and S. lugdunensis
The 12
th
16 -18 June, 2011

328

(1 isolates). Only the isolate SC0903 (isolated from raw beef) was selected to identify
using the API staph kit. It was identified as S. xylosus (99.9% Identity).

All Staphylococcus isolates were catalase and nitrate reductase positive. Catalase is
generally present in all aerobes and facultative anaerobes. This enzyme mediates
degradation of hydrogen peroxide (Krckel 1995). Nitrite reductase is also an essential
enzyme for preservation, color development and aroma formation. Nitrate in the medium
will be reduced into nitrite by action of this enzyme. It also helps to recycle nitrite by
converting this compound into nitrate in the sequential reaction with myoglobin (Toldr
and other 2001). In this study, S. xylosus SC0903 could produce both nitrate and nitrite
reductase. Gtterup and others (2007) reported that some strains of Staphylococcus isolated
from cured meat products (S. saprophyticus, S. simulans, S. sciuri and S. carnosus) could
produce both nitrate and nitrite reductase. The commercial API Staph system showed high
accuracy in the identification of isolated SC0903 which is in agreement with the study of
Cunha and others (2004).

Table 5 Catalase, nitrate reductase, nitrite reductase and amino acid decarboxylase activity from
selected Staphylococcus isolates

Sample Isolates Number of isolates
Catalase Nitrate
reductase
Nitrite
reductase
amino acid
decarboxylase activity
Raw beef SC0204, SC0206
SC0209, SC0608
SC0609, SC0903
SC0904, SC0907
8 8 1
a
7
Raw pork SP0202, SP0204
SP0205, SC0207
4 4 0 4
Bacon 0 0 0 0 0
Nham SS0301, SS0802
SS0804, SS0902
4 4 0 4
Fermented
sausage
ST0901, ST0704
ST1001
3 3 0 3
Total - 19 19 1 18
b

a
The only one isolate with nitrite reductase positive was the isolate SC0903.
b
The only one isolate with amino acid decarboxylase negative was the isolate SC0903

Sodium nitrite depletion by selected lactic acid bacteria and coagulase negative
Staphylococcus isolates

In this study, three bacterial isolates, Lactobacillus plantarum S0602, Lactobacillus brevis
T0904 and Staphylococcus xylosus SC0903 were selected to study the depletion of sodium
nitrite in culture media added with 1,000 mg/l sodium nitrite during fermentation at 37C.
The result showed that at the beginning of fermentation, all media containing 1,000 mg/l
sodium nitrite and inoculated with those three bacteria as starter cultures had residual
nitrite concentration in the range of 130.20-134.48 mg/l. After incubation for 24 h, nitrite
concentration in MRS broth inoculated with L. plantarum S0602 (19.03 mg/l) were
significantly lower than those of other treatments (P<0.05). The culture medium with L.
brevis T0904 had decreased nitrite concentration of 37.28 mg/l, while the medium with S.
xylosus SC0903 contained 132.34 mg/l residual nitrite after 24 h fermentation. Then, the
amount of nitrite in culture medium of all treatments slightly decreased until day 4 of
fermentation (Figure 1a).
The 12
th
16 -18 June, 2011

329

Fermentation time (days)
0 1 2 3 4 5
R
e
s
i
d
u
a
l

n
i
t
r
i
t
e

c
o
n
c
e
n
t
r
a
t
i
o
n

(
m
g
/
l
)
0
20
40
60
80
100
120
140
(a)
0 1 2 3 4 5
O
D

6
2
0

n
m
0.0
.5
1.0
1.5
2.0
2.5
3.0
(b)
0 1 2 3 4 5
p
H
3.0
4.0
5.0
6.0
7.0
8.0
(C)

Figure 1 Change of residual nitrite content (a), optical density (b), and pH values (c) in culture media
with 1,000 mg/l sodium nitrite (TSBN or MRSN) or without sodium nitrite (TSB or MRS
broth): TSB inoculated with Staphylococcus xylosus SC0903 (), TSBN inoculated with S.
xylosus SC0903 (), MRS broth inoculated with Lactobacillus brevis T0904 (), MRSN broth
inoculated with Lactobacillus brevis T0904 (), MRS broth inoculated with L. plantarum
S0602 (), MRSN inoculated with L. plantarum S0602 ().

Considering the optical density at 620 nm after incubation for 24 h, L. plantarum S0602 in
the culture medium grew rapidly with the highest OD
620nm
(2.080) followed by L. brevis
T0904 (1.823) and S. xylosus SC0903 (0.239). Then, OD
620 nm
slightly changed when the
fermentation time increased until 4 days of fermentation (Figure 1b). The depletion of
sodium nitrite was concomitant with the increasing of OD
620 nm
and the decreasing of pH
value of culture medium. As seen in Figure 1c, the pH values of culture medium inoculated
with L. plantarum S0602 and that with L. brevis T0904 decreased rapidly from 6.20-7.50
at the beginning of fermentation to 3.64-4.50 after 24 hours of fermentation, while the pH
value of the culture medium inoculated S. xylosus SC0903 had higher pH values (5.60-
7.19). Then, the pH values of all treatment media changed slightly until the end of
fermentation (Figure 1c).

In conclusion, two bacteriocin- producing LAB strains, L. plantarum S0602 and L. brevis
T0904 with probiotic properties could potentially be used as functional starter cultures in
fermented meat products. In addition, S. xylosus SC0903 could also be used in products as
a starter bacterium. These bacterial strains could provide significant health benefit and
The 12
th
16 -18 June, 2011

330

enhance safety of the products as they may inhibit the growth of pathogenic bacteria and
decrease the residual nitrite concentration in products.

Acknowledgement
The authors would like to thank the Faculty of Science, King Mongkuts Institute of
Technology Ladkrabang for financial support of this research.

References
Axelsson L. 2004. Lactic acid bacteria: classification and physiology. In: Salminen S,
Wright AV, and Ouwehand A, editors. Lactic acid bacteria: microbiological and
functional aspects. New York: Marcel Dekker, Inc. p 1-66.
Baker JS. 1984. Comparison of various methods for differentiation of staphylococci and
micrococci. J Clin Microbiol 19: 875-9.
Bennett RW, Lancette GA. 2001. Staphylococcus aureus. Bacteriological analytical
manual (BAM). Available from: www.fda.vov/Food/ScienceResearch/Laboratory
Methods/BacteriologicalAnalyticalManualBAM/ucm071429.htm. Accessed Nov
12, 2010.
Blixt Y, Borch E. 2002. Comparison of shelf life of vacuum-packed pork and beef. Meat
Sci 60: 371-8.
Bover-Cid S, Holzapfel WH. 1999. Improved screening procedure for biogenic amine
production by lactic acid bacteria. Int J Food Microbiol 53: 33-41.
Chung HS, Kim YB, Chun SL, Ji GE. 1999. Screening and selection of acid and bile
resistant bifidobacteria. Int J Food Microbiol 47: 25-32.
Coppola S, Mauriello G, Aponte M, Moschetti G, Villani F. 2000. Microbial succession
during ripening of Naple-type salami, a southern Italian fermented sausage. Meat
Sci 56: 321-9.
Cunha MLRS, Sinzato YK, Silveir, LV. 2004. Comparison of methods for the
identification of coagulase negative staphylococci. Mem Inst Oswaldo Cruz 99(8):
855-60.
Gtterup J, Olsen K, Knchel S, Tjener K, Stahnke LH, Mller JKS. 2007. Relationship
between nitrate/nitrite reductase activities in meat associated staphylococci and
nitrosylmyoglobin formation in a cured meat model system. Int J Food Microbiol
120(3): 303-10.
Fuller R. 1989. Probiotics in man and animals. J Appl Bacteriol 66(5): 365-78.
Jay JM, Loessner MJ, Golden DA 2005. Modern Food Microbiology, 7th ed. New York:
Springer Science+Business Media Inc. 790 p.
Kirk RS, Sawyer R. 1991. Pearsons composition and analysis of foods. Essex: Longman
scientific & technical. 708 p.
Krckel L. 1995. Bacteria fermentation of meats. In: Campbell-Platt G, Cook PE. editors.
Advances in fermented meats. Great Britain: Blackie academic and professional. p
69-101.
Leroy F, Verluyten J, Vuyst LD. 2006. Functional meat starter cultures for improved
sausage fermentation. Int J Food Microbiol 106: 270-85.
Papamaloni E, Kotzekidou P, Tzanetakis N, Litopoulou-Tzanetaki E. 2002.
Characterization of Micrococcaceae isolated from dry fermented sausage. Food
Microbiol 19(5): 441-9.
Saarela M, Mogensen G, Fondn R, Mtt J, Mattila-Sandholm T. 2000. Probiotic
bacteria: safety, functional and technological properties. J Biotech 84: 197-215.
The 12
th
16 -18 June, 2011

331

Salome JP, Amuta R, Jagannathan, P, Josiah JJ, Berchmans S, Yegnaraman V. 2009.
Electrochemical assay of nitratrate and nitrite reductase activities of Rhizobium
japonicum. Biosens Bioelectron 24(12): 3487-91.
Schillinger U, Lcke FK. 1989. Antibacterial activity of Lactobacillus sake isolated from
meat. Appl Environ Mirobiol 55(8): 1901-6.
Toldr F, Sanz Y, Flores M. 2001. Meat fermentation technology. In: Hui YH, Nip WK,
Rogers RW, Young, OA, editors. Meat science and applications. New York:
Marcel Dekker. p 538-58.
Visessanguan W, Benjakul S, Smitionont T, Kittikun C, Thepkasikul P, Panya A. 2006.
Change in microbiological, biochemical and physic-chemical properties of Nham
inoculated with different inculum levels of Lactobacillus curvatus. LWT-Food Sci
Technol 39(7): 814-26.
Yan PM, Xue WT, Tan SS, Zhang H,Chang XH. 2008. Effect of inoculating lactic acid
bacteria starter cultures on the nitrite concentration of fermenting Chinese paocai.
Food Contr 19(1): 50-5.

The 12
th
16 -18 June, 2011

332

OJ-66

Antiphotooxidative Effect of Ascorbic Acid Microemulsion
in Virgin Coconut Oil

Ambar Rukmini
1
, Sri Raharjo
2,*
, Pudji Hastuti
2
, and Supriyadi
2

1
Department of Food Technology and Nutrition, Faculty of Agricultural Technology,
Widya Mataram University, Yogyakarta, Indonesia;
2
Department of Food and Agricultural Product Technology, Faculty of Agricultural Technology, Gadjah
Mada University, Yogyakarta, Indonesia
*
Corresponding author: sraharjo_ugm@yahoo.com

Abstract

Virgin coconut oil (VCO) is susceptible to quality deterioration due to photooxidation.
This reaction can be inhibited by addition of singlet oxygen quencher (SOQ). Naturally
present SOQs in the VCO were not effective for inhibiting photooxidation. Ascorbic acid
was reported as an effective SOQ in aqueous solutions. However, due to its very poor
solubility in non-aqueous media, ascorbic acid cannot be easily dispersed in hydrophobic
medium such as VCO. Water-in-oil (w/o) microemulsion system may be used to overcome
this problem. The objective of this study was to determine the effectiveness of ascorbic
acid microemulsion to inhibit photooxidation of VCO. The w/o microemulsions were
prepared by mixing deionized water, surfactants, and VCO. Ascorbic acid was
incorporated into the microemulsion formula at level of 0.5, 0.75, 1, 1.5, 2, 3, 4, 5, and
10% to determine its maximum loading capacity. Another set of microemulsions were also
prepared to contain 50, 100, 150, 200, and 250 ppm of ascorbic and mixed with VCO. The
same level of ascorbyl palmitate, TBHQ, and BHA were added into VCO and used as
comparison. All of these samples were subsequently subjected to photooxidation under
fluorescent light exposure for up to 8 hours. Peroxide values and p-anisidine values of
photooxidized samples were measured at 1 hour interval. The result indicated that at the
level of 1% ascorbic acid showed the maximum solubility of ascorbic acid in w/o
microemulsion. This microemulsion was highly stable against photooxidation. At the level
of 250 ppm, ascorbic acid effectively inhibited photooxidation of VCO as compared with
the others antioxidants. This study confirmed that the highly hydrophilic SOQ such as
ascorbic acid can be successfully incorporated into w/o microemulsion and the addition of
ascorbic acid microemulsion effectively inhibited photooxidation of VCO during storage
under fluorescent light.

Keywords: Photooxidation, singlet oxygen quencher, ascorbic acid, virgin coconut oil,
microemulsion

The 12
th
16 -18 June, 2011

333

Introduction
Virgin coconut oil (VCO) is susceptible to quality deterioration due to photooxidation.
Photooxidation occurs when there are light, triplet oxygen, and photo-sensitizer (Min and
Boff 2002). Chlorophyll and its derivatives is common sensitizer that acts as promoters of
photooxidation in vegetable oils (Choe and Min 2006). This reaction can be inhibited by
addition of singlet oxygen quencher (SOQ). Naturally present SOQs in the VCO were not
effective for inhibiting photooxidation. Ascorbic acid reportedly is an effective SOQ in
aqueous solutions (Bodannes and Chan 1979; Jung and others 1995a, b). However, due to
its very poor solubility in non-aqueous media, ascorbic acid cannot be easily dispersed in
hydrophobic medium such as VCO. Water-in-oil (w/o) microemulsion system may be used
to overcome this problem. On the other hand, Lee and Jung (2010) reported that the
synthetic antioxidants, especially TBHQ, have a strong singlet oxygen quenching ability.
Because of its several advantages such as transparent, thermodynamically stable, low
viscosity, easy formulation (low interfacial tension and almost spontaneous formation) and
high surface area (high solubilisation capacity), the use of water-in-oil (w/o)
microemulsion to include ascorbic acid may be compatible in the VCO system. The
objectives of this study were (1) to determine the solubility of ascorbic acid in w/o
microemulsion, and the stability of ascorbic acid microemulsion against photooxidation;
and (2) to evaluate the antiphotooxidative effect of ascorbic acid microemulsion in VCO as
compared with common synthetic antioxidants.

Materials
A freshly prepared VCO obtained from a local VCO producer was filtered in the presence
of anhydrous Na
2
SO
4
to obtain the dry VCO and then used as continuous phase. Low HLB
surfactant, i.e. sorbitan monooleate (Span 80) and medium HLB surfactant, i.e. sorbitan
monolaurate (Span 20) were purchased from Sigma Chemical Company (St. Louis, MO),
and high HLB surfactant, i.e. polyoxyethylene sorbitan monolaurate (Tween 20) was
purchased from Merck (Darmstadt, Germany). Deionized water was used as disperse
phase. Ascorbic acid and ascorbyl palmitate were purchased from Sigma Chemical
Company (St. Louis, MO), whereas TBHQ and BHA were purchased from Merck
(Darmstadt, Germany), and all chemicals used were of analytical grade.

Preparation and photooxidative stability test of ascorbic acid microemulsion
For the preparation of the ascorbic acid microemulsion the following procedure was
employed. A predetermined amount of ascorbic acid (0.5, 0.75, 1, 2, 3, 4, 5, or 10%) was
first dissolved in the deionized water. This solution was subsequently added with
surfactant mixtures which consist of 16.6% Tween 20, 15.0% Span 20, and 68.4% Span 80
and mixed on a hot-plate stirrer (SRS 710 HA, Advantec, Japan) at medium speed and the
temperature was kept at 402C. After 10 minutes, the dry VCO was added dropwise while
stirring at high speed for up to 10 minutes.

To evaluate the photooxidative stability of ascorbic acid microemulsion, a portion of 20
mL ascorbic acid microemulsions were placed in 30 mL transparent serum bottles with
rubber caps. Photooxidation was performed under accelerated condition using fluorescent
lights with intensity of approximately 4000 lux and samples were exposed to that light for
up to 5 hours at room temperature (301C). Peroxide values (PV) of the samples were
measured according to the method proposed by the AOCS Official Method (AOCS 2004).
The 12
th
16 -18 June, 2011

334

Photooxidation experiments
To evaluate the antiphotooxidative effect of ascorbic acid microemulsion in VCO, a
portion of 20 mL VCO samples containing ascorbic acid microemulsion were placed in a
number of 30 mL transparent serum bottles with rubber caps. The levels of ascorbic acid in
VCO were arranged on 50, 100, 150, 200, or 250 ppm. The same levels of ascorbyl
palmitate (AP), TBHQ, and BHA were mixed into the VCO and used as comparison. All
of these samples were subsequently subjected to photooxidation under accelerated
condition using fluorescent light exposure (4000 lux) for up to 8 hours at room temperature
(301C). The peroxide values (PV) and p-anisidine values (p-AnV) of photooxidized
VCO samples were measured at 1 hour interval. PV and p-AnV were determined according
to the method proposed by the AOCS Official Method (AOCS 2004). The PV and the p-
AnV were used to calculate the total oxidation (TOTOX) of the samples. According to
AOCS (2004), the use of PV and p-AnV analysis together gives a more complete picture of
total oxidation than each test separately. The equation is TOTOX = 2 PV + p-AnV. The
low TOTOX values indicate good stability.

Solubility of ascorbic acid in w/o microemulsion and its photooxidative stability
Incorporation of ascorbic acid at the level of higher than 1% (w/w) produced slightly turbid
or cloudy solution with visible precipitations of ascorbic acid (Figure 1).
According to Pakpayat and others (2009), incorporation of ascorbic acid in the formulated
microemulsion leads to destabilization of the microemulsion structure and fail to obtain
transparent isotropic systems as expected. In this study, at the level of 1% ascorbic acid or
higer give slightly cloudy solution with ascorbic acid precipitate in appearance. Thus, it
indicated that the maximum solubilisation of ascorbic acid in the w/o microemulsions was
1%. They showed the maximum ability of w/o microemulsion in retaining the ascorbic
acid without phase separation (Figure 1). This ascorbic acid microemulsion formula
subsequently was used on photooxidation stability test.

Figure 1 Ascorbic acid microemulsion photographs (arrows indicate ascorbic acid precipitates).

The 12
th
16 -18 June, 2011

335

The photooxidative stability test of ascorbic acid microemulsions indicated that ascorbic
acid which was incorporated in the w/o microemulsion system could play as an effective
SOQ and inhibit the photooxidation reaction. As reported by Bodannes and Chan (1979),
ascorbate can scavenge singlet oxygen, superoxide and peroxide. In comparison with the
said study, this experiment resulted in a similar trend which was indicated by the zero
values of PV in ascorbic acid microemulsions, even after they were exposed to the
fluorescent light. This result indicated that the ascorbic acid microemulsion resistant to the
photooxidation reaction.

Antiphotooxidative effect of ascorbic acid microemulsion in VCO
The TOTOX value measures both hydroperoxides and their breakdown products, and
provides a better estimation of the progressive oxidative deterioration of fats and oils
(Shahidi and Zhong 2005). The TOTOX values of light exposed VCO samples containing
various levels of ascorbic acid microemulsion, AP, TBHQ, or BHA was showed in Figure
2.

Figure 2 TOTOX value of light exposed VCO containing various levels of ascorbic acid
microemulsion (AA, ), ascorbyl palmitate (AP, ), TBHQ (), or BHA ().

Duration of light exposure (hours)
T
O
T
O
X

v
a
l
u
e

50 ppm 100 ppm
200 ppm 150 ppm
250 ppm

The 12
th
16 -18 June, 2011

336

Light exposed VCO samples containing ascorbic acid microemulsion showed very
significantly lower TOTOX value (P < 0.01) than that of VCO containing AP, TBHQ, or
BHA on all levels of antioxidants. Intense of light exposure with relatively high intensity
(4000 lux) was very effective to initiate photooxidation. The light energy from that light
exposure was absorbed by chlorophyll which implied in VCO and caused the chlorophyll
to become excited.

The hydroperoxides which were produced during light exposed VCO samples had more
polar characteristics than the triacylglycerols composed of VCO. They were able to diffuse
to the more polar zone and interacted with the compounds present in the aqueous phase
(Decker and others 2010). The more polar zone in this study was the interface layer of
microemulsions which were contained of ascorbic acid in their microdroplets. Because of
the reason, the hydroperoxides were quenched easily by the ascorbic acid as compared
with the lipophilic antioxidants (TBHQ or BHA) or the oil-soluble ester antioxidant (AP).
Moreover, ascorbic acid also could act as singlet oxygen quencher (SOQ) and prevented
the alkyl radical formation that initiated the oxidation chain reactions. It was resulted in
lower TOTOX value of VCO containing ascorbic acid microemulsion as compared with
those synthetic antioxidants.

Although ascorbic acid was hydrophilic antioxidant, ascorbic acid microemulsion more
effectively acted as antiphotooxidative agent in VCO than the lipophilic or oil-soluble ester
antioxidants. The capability order of those antioxidants on preventing photooxidation of
VCO were ascorbic acid microemulsion > TBHQ > AP > BHA.

References
AOCS. 2004. Official Methods and Recommended Practices of the AOCS. 5th ed. Illinois:
American Oil Chemists Society.
Bodannes RS, Chan PC. 1979. Ascorbic acid as a scavenger of oxygen singlet. Febs Letters
105(2):195-196.
Choe E, Min DB. 2006. Mechanisms and factors for edible oil oxidation. Comp Rev Food Sci
and Food Saf 5:169-186.
Decker EA, Alamed J, Castro IA. 2010. Interaction between polar components and the degree of
unsaturation of fatty acids on the oxidative stability of emulsions. J Am Oil Chem Soc
87: 771-780.
Jung MY, Kim SK, Kim SY. 1995a. Riboflavin-sensitized photodynamic UV spectrophotometry
for ascorbic acid assay in beverages. J Food Sci 60:360-363, 368.
Jung MY, Kim SK, Kim SY. 1995b. Riboflavin-sensitized photo-oxidation of ascorbic acid
kinetics and amino acid effects. Food Chem 53:397-403.
Lee JH, Jung MY. 2010. Direct spectroscopic observation of singlet oxygen quenching and
kinetic studies of physical and chemical singlet oxygen quenching rate constants of
synthetic antioxidants (BHA, BHT, and TBHQ) in methanol. J Food Sci 75(6):C506-
C513.
Min DM, Boff JM. 2002. Chemistry and reaction of singlet oxygen in food. Comp Rev Food Sci
and Food Saf 1:58-72.
Pakpayat N, Nielloud F, Fortune R, Tourne-Peteilh C, Villarreal A, Grillo I, Bataille B. 2009.
Formulation of ascorbic acid microemulsions with alkyl polyglycosides. Eur J Pharm
72:444-452.
Shahidi F, Zhong Y. 2005. Lipid Oxidation: Measurement Methods. In Shahidi, editor. Baileys
Industrial Oil and Fat Products. 6th ed., vol.5. New Jersey: John Wiley & Sons.

The 12
th
16 -18 June, 2011

337

OJ-69

Banana (Musa sapientum var. Cavendish) Flour as Wheat Flour
Extender in Selected Bakery Products

Maria Ligaya T. Braganza
1,*
,Edilborga H. Tolentino

1
School of Food Science and Technology, The Philippine Womens University
*
Corresponding author: joy.braganza@yahoo.com

Abstract

Banana flour made from fresh, mature green Cavendish variety was used to partially
replace wheat flour in soft bun (pan de sal) and doughnut in concentrations ranging from
a minimum of 10% to a maximum of 50% using the bakers percentage (i.e. partial
replacement of wheat flour with banana flour was based on the total wheat flour weights in
the formulation) using the no time dough method. No adverse effects were observed on the
over-all baking and sensory qualities of the baked products. A savings of 3-5% for soft bun
and 27% for doughnut was obtained in terms of direct material costs. Moreover, the
inclusion of banana flour effectively increased the theoretical Vitamin C, potassium and
dietary fiber contents of the baked goods. Approximately 3-5 pieces of soft bun and
doughnut can satisfy of the average per day Recommended Dietary Allowance for
Filipinos for Energy and Specific Nutrients (RENI) for children ages 7-9 years old. On the
other hand, approximately 10-13 pieces of the same products can satisfy at least of the
US Recommended Dietary Allowance (RDA) for potassium and the Dietary Reference
Intake (DRI) for dietary fiber , respectively.

Keywords: Banana flour, wheat flour extender, no time dough method, level of inclusion
Cavendish banana

Introduction
The banana fruit is a healthy, nutritious commodity which contains 74% water, 23%
carbohydrates, 1% protein and 0.5% fat. Without its peel, it is a good source of Vitamin
B6, potassium and fiber (http://www.extento.hawaii.edu). Moreover, it has no sodium and
cholesterol and is a great source of Vitamin C and magnesium and contains three natural
sugars sucrose, fructose and glucose giving an instant, sustained and substantial boost of
energy (http://www.antioxidant-fruits.com). Potentially, it can be processed and preserved
to expand its market value, such as, puree from ripe fruits for use in ice cream, yogurt,
cake, baby foods and nectar; sliced and canned in syrup for use in fruit salads and as
toppings; sun dried banana crispy; and, fermented to produce vinegar and alcoholic
beverage.

A new product with commercial value is the banana flour which can be used as a mixture
for various cakes and breads. But since it does not contain gluten, it could not be used as
the main ingredient but rather mixed with wheat flour in the production of quality baked
products (http://www.ffc.agnet.org/library).. The main objective of the study was to extend
the utilization and increase the consumption of banana flour in two selected bakery
products, i.e. soft bun and doughnut. Specifically it sought answers to the following:
The 12
th
16 -18 June, 2011

338

1. What is the practicality of using of using banana flour as wheat flour extender in
selected bakery products?
2. What is the minimum and maximum levels of inclusion of banana flour in the
standard bread formulation.?
3. What are the effects of partial replacement of wheat flour with banana flour on the
over-all product acceptability; theoretical nutrient content and direct material cost of the
baked products?

Product Formulation
Banana flour made from fresh mature green Cavendish variety were made to replace part of
the wheat flour requirements in the basic recipes for soft bun (pan de sal) and doughnut in
concentrations ranging from 10% to 50% of eight (8) experimental lots including the control
samples using the no time dough method (http://www.seabeecook.com).

Weigh ingredients separately Combine except lard Mix at low speed (#1:5-10 mins)
Scale dough Rest (20 mins) Mix until elastic Add lard
(30 gm/pc) covered with plastic (speed #2)

Grease pan Arrange dough Proof (1-1.5 hrs) Bake Cool
(440
o
F;15-20 mins)

Figure 1 Flow Chart for the Preparation of Banana Soft Bun Using the No Time Dough Method

Table 1 Percentage Formulation of Soft Bun (Bakers Percentage)

Raw Materials Control Lot 1
10%
BF
Lot 2
15%
BF
Lot 3
20%
BF
Lot 4
25%
BF
Lot 5
30%
BF
Lot 6
40%
BF
Lot 7
50%
BF
Bread flour 100 90 85 80 75 70 60 50
Banana Flour 0 10 15 20 25 30 40 50
Total 100 100 100 100 100 100 100 100
Water 65 65 65 65 65 65 65 65
R . Sugar 24 24 24 24 24 24 24 24
Lard 7 7 7 7 7 7 7 7
R. Salt 2 2 2 2 2 2 2 2
Instant Yeast 1.7 1.7 1.7 1.7 1.7
1.7

1.7
1.7
Dough Improver 0.3 0.3 0.3 0-3 0.3
0.3

0.3
0.3
Total 100 100 100 100 100 100 100 100

The 12
th
16 -18 June, 2011

339

Table 2 Percentage Formulation of Doughnut (Bakers Percentage)

Raw Materials Control Lot 1
10%
BF
Lot 2
15%
BF
Lot 3
20%
BF
Lot 4
25%
BF
Lot 5
30%
BF
Lot 6
40%
BF
Lot 7
50%
BF
Bread flour 100 90 85 80 75 70 60 50
Banana Flour 0 10 15 20 25 30 40 50
Total 100 100 100 100 100 100 100 100
Water 48.5 48.5 48.5 48.5 48.5 48.5 48.5 48.5
R . Sugar 13.7 13.7 13.7 13.7 13.7 13.7 13.7 13.7
Butter 13.7 13.7 13.7 13.7 13.7 13.7 13.7 13.7
Egg 10.5 10.5 10.5 10.5 10.5 10.5 10.5 10.5
Milk powder 8.4 8.4 8.4 8.4 8.4 8.4 8.4 8.4
R. Salt 2 2 2 2 2 2 2 2
Instant Yeast 1.6 1.6 1.6 1.6 1.6 1.7 1.7 1.7
Baking Powder 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6
Total 100 1009 100 100 100 100 100 100

Acceptability Tests
Sensory evaluation using the 7- and 9-point hedonic scales by thirty (30) untrained
laboratory panel composed of faculty members, college students and non-teaching
personnel from the Cluster of Food Science, Tourism and Hospitality Management.

Weigh ingredients & combine Add water Add shortening Rest

Rest (25 mins) Sheet Round into a ball Mix until gluten develops

Cut Deep fat fry (light brown) Roll in granulated sugar

Figure 2. Flow Chart for the Preparation of Banana Doughnut Using the No Time Dough Method

Theoretical Nutrient Content of Experimental Products
Banana flour samples were subjected to proximate analysis (i.e. moisture, protein,
potassium, total dietary fiber and Vitamin C) at the National Food Authority (NFA) Food
Development Center. Results of analysis together with the use of the Food and Nutrition
Research Institute (FNRI) Philippine Food Composition Tables were used in the
calculation of the theoretical nutrient contents of the baked products.

Direct Material Costs of Most Acceptable Formulations
The following assumptions were used in computing the direct material costs of the
products under study:
Banana Flour = PhP 18.50/kg
Bread Flour = PhP 39/kg
All Purpose Flour = PhP 42/kg
Mark up = 20%

The 12
th
16 -18 June, 2011

340

Table 3. Direct Material Costs of Most Acceptable Baked Products

Acceptable Lots Direct Material
Cost
Total Cost (with
20% mark up)
Yield/Cost Per Piece
Soft Bun, Control Php 56.90 PhP 68.30 52 pieces at 30 grams per piece/
PhP 1.31per piece
Soft Bun w3ith 10% banana
flour
PhP 54.90 PhP 64.90 52 pieces at 30 grams per
piece/PhP 1.25 per piece
Soft bun with 15% banana flour PhP 52.90 PhP 63.50 52 pieces at 30 grams per
piece/PhP 1.22 per piece
Doughnut, control PhP 147.70 PhP 177.24 50-55 pieces at 35 grams per
piece/PhP 3.22-3.54 per piece
Doughnut with 10% banana
flour
PhP 147.25 PhP 176.70 50-55 pieces at 35 grams per
piece/PhP 3.21-3.53 per piece
Doughnut with 15% banana
flour
PhP 146.08 PhP 175.30 50-55 pieces at 35 grams per
piece/PhP 3.19 3.51 per piece

Acceptability Tests
Sensory evaluation results showed products with 10%- 40% banana flour were all
acceptable obtaining mean panel scores ranging from 5.2 6.2 interpreted as good and
very good for the sensory attributes of appearance, aroma, flavor and texture. The lot
containing 10% and 15% banana flour got the highest mean panel scores indicating that
these were the most preferred by the taste panelists. However, experimental lots
containing 40% banana flour exhibited sourish flavors and odors, as well as, darker colors
compared with the lots containing lesser amounts of banana flour. Products with 50%
banana flour were totally unacceptable to the taste panelists.

Theoretical Nutrient Content of Baked Products
A 30-gram piece of soft bun with 10-15% banana flour can theoretically provide 87-91
kcal, 8-9 mg. Vitamin C, 18-27 mg. potassium and 15 mg. calcium. On the other hand, a
35-gram piece of doughnut with the same amounts of banana flour theoretically contains
96-99 kcal, 2.8 gm. protein, 25-26 ug. Vitamin A, about 10 mg. Vitamin C, 29 mg. calcium
and 18 - 27 mg. potassium. In addition, both products are able to satisfy from 4% - 11% of
the daily US RDA for potassium and from 1.4 to 3.4% of of the US daily Dietary
Reference Intake (RDI) for various age groups ( i.e. 1-9 years).

Direct Material Cost
At 10% and 15% wheat flour replacement, the total direct material costs were
approximately US$ 1.50 to US$3.00, for the soft bun and doughnut, respectively. Average
yields of 52 pieces at 30- gram and 35-gram per piece of soft buns and doughnuts were
obtained.

Conclusion and Recommendation
a. Conclusions
Banana flour can practically replace part of the wheat flour requirements for soft bun (
pan de sal) and doughnut from a minimum of 10% to a maximum of 30% without any
adverse effect on the over-all quality of the baked products.
The 12
th
16 -18 June, 2011

341

In terms of sensory attributes, the optimum level of incorporation of banana flour ranged
from 10% - 15% for both baked products. Flavor and odor were enhanced with the
addition of banana flour thereby further increasing their acceptability.

Approximately 5% and 27% savings in terms of direct material costs can be obtained by
replacing wheat flour with 10%-15% banana flour in soft buns and doughnut.

Partial replacement of wheat flour with 10%-15% banana flour can increase the potassium,
Vitamin A and dietary fiber contents of both baked products.

Approximately 3-5 pieces of soft buns and 10-13 pieces of doughnut with 10%-15%
banana flour replacement can satisfy at least of the RENI, RDA and DRI for Filipino
children ages 7-9 years old.

b. Recommendations:
In addition to soft bun and doughnut, the use of banana flour as wheat flour extender in
other equally popular baked products can be explored. Likewise, its use can be further
extended for pasta and noodle making.

Since the nutrient composition was only computed theoretically, the experimental products
can be subjected to proximate and specific nutrient analyses to determine their actual
nutrient contribution.

The baked products with banana flour can be subjected to glycemic index analyses to
determine their health and wellness benefits particularly for nutrition-related lifestyle
diseases like diabetes and android obesity.

References
Food and Nutrition Research Institute (FNRI). 2000. Nutritional guidelines for Filipinos.
rev. ed. Manila: Department of Science and Technology (DOST).
_____. 2001. Philippine nutrition facts and figures. Manila: DOST. 109 p.
Guzman, Matilde P. and others. 1986. Basic foods for Filipinos. Rev. ed. Manila:Merriam
and Webster, Inc. 455 p.
UNICEF. Oct. 2000. Fact sheet on nutrition-related lifestyle diseases in the Philippines.
FFTC Publication Database. Processing of Banana Flour. Available from
www.fftc.agnet.org/library. Accessed Aug. 18,2009.
HONcode. Homepage. Dietary Fiber:Daily Reference Intakes (DRIs) for Fiber, RDA.
Available from www.dietaryfiber.food.com/fiber-rda.php. Accessed Jan. 12,2010.
SeabeeCook.com Homepage. 1916 Bread Making Methods. Available from
www.seabeecook. Accessed Sept. 2, 2009.
The 12
th
16 -18 June, 2011

342

Appendix A. Analytical Laboratory Test Results of Banana Flour

Analysis Results Remarks
Moisture, g/100g 9.8 Gravimetric@105 degree C (Official Methods of AOAC
International, 18
th
ed. 2005)
Protein ((N x 6.25), g/100g 4.8 Kjeldahl (Official Methods of AOAC International, 18
th
ed. 2005)
Potassium Content
(mg/100g)
943 Official Method 975.03 (Official Methods of Analysis of AOAC
International .18
th
ed. 2005)
Total Dietary Fiber
Content (g/100g)
5.46 Official Method 985.29 (Official Methods of Analysis of AOAC
International. 18
th
ed. 2005)
Vitamin C content (mg
ascorbic acid/100 g(

28.6

Official Method 967.21 (Official Methods of Analysis of AOAC
International. 18
th
ed. 2005)
Note: Moisture and protein contents analyzed by the Sentro sa Pagsusuri, Pagsasanay at Pangangasiwang
Pang Agham at Teknolohiya Corp., Mandaluyong City, Phils.
Potassium , Total Dietary Fiber and Vitamin C contents analyzed by the Food Development Center,
National Food Authority, Taguig, Philippines

Appendix B. Soft Buns and Doughnuts with 15% Banana Flour

The 12
th
16 -18 June, 2011

343

OJ-131

The Inhibitory Activity of Cinnamon Essential Oil on Fish Spoilage
Bacteria: Laboratory Media and Food System Assay

O. M. Bahurmiz
1, 3,*
, R. Ahmad
1
, N. Ismail
1
, and S. F. Sulaiman
2

1
Food Technology Division, School of Industrial Technology, Universiti Sains Malaysia, 11800 Penang,
Malaysia;
2
School of Biological Sciences, Universiti Sains Malaysia, 11800 Penang, Malaysia;
3
Faculty of Environmental Sciences and Marine Biology, Hadhramout University of Science
and Technology, Mukalla, Yemen
*
Corresponding author: osan79@hotmail.com

Abstract

The inhibitory effect of cinnamon essential oil (CEO) on Pseudomonas spp. associated
with microbial spoilage of tilapia was investigated in liquid media and fish fillet. The first
assay was carried out using two different media; nutrient broth and fish extract model.
Various concentrations of CEO were introduced to each medium, followed by a
Pseudomonas inoculation of approximately 4.5 log CFU/ml. The medium-containing tubes
were then incubated at 4C and the bacterial growth was recorded for 10 days. In the
application assay, fillets of freshly caught tilapia were dipped into CEO-containing
solutions prior to chilled storage, 4C/10 days. Treated fillets were then subjected to
microbiological and sensory analysis. Results from the in vitro assay showed MIC value of
0.250 l ml
-1
. The growth of Pseudomonas spp. in fish extract was effectively retarded by
CEO, but with markedly lower activity than that in the broth medium. On the other hand,
considerable higher concentrations of CEO (about 10 MIC value onward) were required
for fish fillet to retain an inhibitory effect comparable to that achieved in laboratory
medium. However, in means of fish preservation, these levels are generally encouraging,
especially as they showed no adverse effect on the consumer acceptability of treated
tilapia. These findings indicated that CEO could be a promising natural antimicrobial
agent for controlling spoilage bacteria and extending the shelf-life of chilled fish.

Keywords: Antimicrobial, cinnamon essential oil, Pseudomonas spp., broth, fish
The 12
th
16 -18 June, 2011

344

OJ-133

Effect of Multi Level Ethanol Leaching on Physico-Chemical Properties
of Konjac Flour (Amorphophallus Oncophyllus)

Simon Bambang Widjanarko
*
, Anni Faridah and Aji Sutrisno

Department of Food Technology, Faculty of Agricultural Technology,
Brawijaya University, Malang, Indonesia
*
Corresponding author: Areng-areng Timur 8 Dadaprejo, Junrejo Batu 6532
Tel/Fax:+62341460518; Email: simonbw@ub.ac.id

Abstract

Common konjac flour (CKF) is unfit for human consumption due to the high presence of
calcium oxalate. This causes CKF to have an acrid taste, therefore it needs further
purification. The objective of this research is to reduce the amount of calcium oxalate as
well as to increase its glucomannan content, viscosity and the degree of whiteness CKF by
applying a multi-stage ethanol leaching processes under an ultrasound-assisted extraction
(UAE). The results showed that after UAE, treatment combination of the 3
rd
stages of
leaching process for 25 minutes had the best results - higher glucomannan and viscosity
than untreated flour, and this the best treated flour had physico-chemical characteristics
as follows: 87.83% glucomannan, 0.63% calcium oxalate, 49.09 the degree of whiteness,
and 8200 c.Ps viscosity. Multi-stage ethanol leaching CKF by an UAE showed promise in
reducing acrid flavor of CKF.

Keywords: Physico-chemical properties, multi-stage ethanol leaching, konjac flour,
ultrasound.

Introduction
Konjac or a local name porang is a native plant to Indonesia (Widjanarko, 2010) and it
has been known as a plant with medicinal properties (Harrison, 2009). Nowadays, the
processing of konjac is done with up-to-date technique. Common konjac flour (CKF) is
processed as follows: a tuber is sliced, sun-dried, ground and purified mechanically to
reduce tachiko or tobico (dust from the konjac tuber). The konjac flour powder is still
unpalatable due to the high content of Ca-Oxalate. The konjac flour needs further
purification processes by boiling with salt solutions and precipitated out with adding
ethanol solution and then dried under vacuum or freeze dried. By this extraction technique,
konjac glucomannan (KGM) is obtained (Anonymous, 2009). The main problem of
producing KGM with this method, according to Nishida and Hasimas method and Ohtsuki
method (Sugiyama and others 1972), a collected dried material is difficult to be grinded
into a powder and it is insoluble in water. Our own experience prove (Unpublished reports)
that KGM prepared by boiling CKF in aluminum sulphate, sodium chloride and
trichloroacetic acid under maceration and ultrasound-assisted extraction (UAE) and then it
was precipitated with ethanol solution, a collected dried material is difficult to be grinded
and is not soluble in water (Perdani and Widjanarko, 2009, Meta and Widjanarko, 2010,
Erryana and Widjanarko, 2010, Diak and Widjanarko, 2010).

The 12
th
16 -18 June, 2011

345

To overcome the above problems, the konjac flour is purified by multi-stage ethanol
leaching processes without boiling with salt solutions, prior to be precipitated with ethanol.
The multi-stage ethanol leaching on CKF was originally introduced by Sugiyama and
others (1972), but it took three days in extracting crude konjac flour with three times of 50
v/v% ethanol, and then followed by rinsing residual flour with 80 v/v% ethanol. The
collected dried flour was apparently similar to the original flour or CKF, which was
soluble in cold water. To shorten the period of times in purifying CKF as described by
Sugiyama and others (1972), the ultrasound-assisted extraction is chosen. To remove the
acrid taste and other impurities that adhere on the surface of CKF granules, it is washed by
several leaching stages with increasing concentrations of ethanol solution gradually, during
leaching processes.

Power ultrasonic, having frequencies between 20 kHz and 100 kHz, is now well-known to
have significant effects on the rate of various physical and chemical processes. Cleaning
and extraction are the more developed applications and a large variety of ultrasound baths
and transducers exist for laboratory use. The most interesting effect of ultrasound-based
operational unit is less processing time and the increase of product quality (Chemat and
others 2009). The use of high power ultrasound in the extraction of functional and
bioactive component of plant materials has been reported by many researchers: Salvia
officianalis (Salisofa and others 1997), bioactive from plant materials (Vinatoru and others
1997), bioactive from herbs (Vinatoru, 2001), steroids and triterpenoids (Schinor and
others 2004), ginseng root (Yang and Zhang, 2008) and citrus peel (Ma and others 2009).
Ultrasound-assisted alkaline extraction was reported better and faster in comparison to
conventional alkaline extraction in isolating lignin from wheat straw ( Sun and Tomkinson,
2002).

The beneficial effects of ultrasound waves on extraction or cleaning processes are
attributed to the formation and asymmetrical collapse of micro cavities of the vicinity of
cell walls leading to the generation of micro jets rupturing the cells, especially on the
surface area of the cells. It is also thought that the pulsation of bubbles causes acoustic
streaming which improves mass transfer rate by preventing the solvent surroundings the
plant tissues from getting saturated and therefore enhancement of convection currents. The
objective of this experiment is aimed to report the effect of multi-stage ethanol leaching at
different dipping stage and period by UAE on the physico-chemical properties of purified
konjac flour (PKF).

MATERIAL AND METHODS
Konjac tuber with outer diameter were between 19 and 25 cm, weight 3 kg0,2, was
collected from a konjac farmer at Sumberbenda Village, Saradan district, Madiun Regency,
Indonesia. All the chemicals used were analytical grade, while the water was glass
distilled. Commercial konjac flour (Made in USA) is bought through on line trading.

Experimental Design
Factorial Complete Randomized Design (FCRD) with two factors used in this experiment.
First factor was extraction stage (1, 2 and 3), and second factor was extraction period (5,
15 and 25 minutes). Stage 1, CFK was dipped in 200 ml 40% ethanol and the leaching
processes is under ultrasonic bath. Stage 2, CFK was dipped in 200 ml 40% ethanol under
ultrasound-assisted, filtered, then followed by 200 ml 60% ethanol solution, and stage 3,
The 12
th
16 -18 June, 2011

346

again in 200 ml 40%, then 200 ml 60%, finally in 80% ethanol with an ultrasound-assisted
extraction. There were 9 treatment combinations which were 3 stages of leaching processes
times 3 period of extraction. Each treatment was replicated two times. Therefore, there
would be 18 samples. Control was untreated konjac flour and commercial konjac flour
(KGM). Multi attributes method as described by Zeleny (1982) was used to determine the
best treatment of purifying CKF and this best treatment will be compared to untreated and
a commercial KGM.

Sample Preparation
Stage 1, crude or common konjac flour (CKF) was weighed 25 gram per sample,
transferred into a 500 ml beaker glass. 200 mL 40% ethanol solution was added into the
beaker, then an ultrasonic bath (type RETSCH URG) at room temperature was applied at
the power of 50 kHz for 5, 15 and 25 minutes. The extracted CKF slurry was sieved and a
residue was dried at 40
o
C in an oven for 36 hours until reached constant weight and
analyzed. Stage 2, sample was prepared as stated at stage 1, then it was leached again in
200 ml 60% ethanol solution under the ultrasound-assisted extraction, prior to be dried.
Stage 2 residue was dried until reached constant weight at 40
o
C in an oven for 24 hours.
Stage 3, sample was prepared as noted at stage 2, and finally it was leached in 200 ml 80%
ethanol solution under ultrasound-assisted extraction, prior to be dried at 40
o
C in an oven
for 12 hours. Therefore we had 9 samples and each samples replicated twice. Each
samples, including control and a commercial KGM were stored in airtight containers until
it was used for further analysis.

The best treated flour, untreated flour and a commercial KGM will be analyzed for
physico-chemical properties such as: glucomannan, viscosity, degree of whiteness, calcium
oxalate, moisture content, rehydration period and total yield.

Analysis
Glucomannan content
Glucomannan content was determined using the method as described by Peiying and others
(2002).

Viscosity measurement
Viscosity of 1% konjac flour solution was determined using spindle needle 1 at 60 rpm of
Brookfield Viscometer (Brookfield LD IV) at room temperature as described by Peiying
and others (2002)

Degree of whiteness
Degree of whiteness sample was determined using color reader (Minolta CR-100)
according to the procedure noted by Koeswara (2009).

Calcium oxalate content
The oxalate content was determined using the method originally developed by Ukpabi dan
Ejidoh, (1989).

Moisture content
Moisture content was determined according to the standard method (AOAC, 1984)

The 12
th
16 -18 June, 2011

347

Rehydration Period
Rehydration period only for 3 samples include the best treatment of PKF, untreated konjac
flour and a commercial KGM was determined as described by Japanese Patent 63 68054
(Ohashi and others1988).

Total yield
Total yield of purified flour was determined according to the method as described by
Sudarmadji and others (1997).

Statistical analysis was carried out using Excel software. Significance of differences was
defined at P<0.01 using Duncan Multiple Range Test (DMRT) or Least Significance
Difference (LSD). The best treatment used in this experiment was determined by multiple
attributes as described by Zeleny (1982).

RESULTS AND DISCUSSION
Glucomannan and Viscosity of PKF
Shimizu and Shimahara (2004) claimed that ethanol 80 v/v% can dissolve an inorganic
materials adhered on the surface of a konjac flour granula, therefore the ash content of
treated flour was lower than control. Takigami (2000) also reported that the ash content,
total protein and total fat were significantly lower for an ethanol-purified konjac flour
than crude konjac flour. The lower non-glucomannan materials (tachiko) present a
purified konjac flour, the higher the glucomannan content.

Table 1 shows that CKF dipped in ethanol solutions at stage 3 for 25 minutes showed the
highest glucomannan content than other treatments. This data is supported by data reported
by Widjanarko and others (2011) that the highest glucomannan content of PKF was
indicated by dipping CKF with ethanol solution at stage 3 for 4 hours by maseration (data
not shown). Chan (2009) mentioned that alcohol can dissolves impurities adhered on
konjac flour granules.

Whereas dipping CKF at stage 1, from 5 to 25 minutes, the glucomannan content ranged
52.27%1.57 60.11%1.45 were relatively lower than control (71.83%0.00). Arifin
(2001) reported glucomannan content of crude konjac flour was 64.98%, meanwhile
Widjanarko and others (2011
b
) reported glucomannan of crude konjac flour was 64.36%
0.01. This could be due to the loss of dissolving some of glucomannan granules into the
water fractions of 40% ethanol solutions, during the dipping period. The presence of
hydroxyl (-OH) groups and mannose and glucose units of KGM is well known as
described by Zhang and others (2001) and Widjanarko and others (2011
a
). Therefore
KGM is easily dissolved in water and leached during period of purification with aqeous
ethanol solution.

The purity of konjac flour increases as the dipping stage and period in ethanol solution
increases, and the less impurities present in PKF. Finally, the best PKF treatment showed
the highest viscosity amongst other treatments. Table 1 shows that the highest viscosity
(8.200 cPs 282.84) of PKF was indicated by dipping stage 3 for 25 minutes. In general,
rheological gels behaviour depend on molecular structure of a gelling agent. Relationship
between the elastic properties of gels and molecular weight were studied for gelatin (Ward
The 12
th
16 -18 June, 2011

348

and Saunders, 1955), agarose (Watase and Nishinari, 1983), kappa-carrageenan (Rochas
and others 1990), CMC (Nishinari and others 1997).

Table 1. Means of glucomannan content and viscosity of PKF affected by different dipping stage and
period with multi-stage ethanol leaching by UAE.

Dipping stage Dipping period (minute) Glucomannan (%)* Viscosity
(c.Ps)**

control - 71.83 4900
1
+)
5 60.11 a 4450 c
15 55.84 a 2700 b
25 52.27 a 1550 a
2
+)
5 64.65 c 5000 cd
15 69.82 d 5750 d
25 69.37 d 5750 de
3
+)
5 71.91 d 7000 e
15 85.43 e 7250 ef
25 87.83 f 8200 f
Notes: *DMRT 1% (4.34-5.06); **DMRT 1% (829.28 966.29);
Control flour was untreated common konjac flour (CKF); 1
+)
CKF was dipped in 40% ethanol,
then dried in an oven at 40
0
C for 36 hours. 2
+)
CKF was dipped in 40% ethanol solution, then
dipped in 60%, finally dried at 40
0
C for 24 hours, and 3
+)
CKF was dipped in 40%, then 60% and
finally in 80% ethanol solution, under ultrasound-assisted extraction, and finally dried at 40
0
C for
12 hours.

Elastic modulus of gel increases as the increase of molecular weight of gels. The
noteworthy of the relation between the glucomannan content and viscosity of PKF is
shown at Table 1, whereby PKF viscosity increases as the glucomannan content increases.
This data was corroborated with data Yoshimura and Nishinari (1999), showed that LM4
fraction of KGM (highest MW) had dynamic viscoelasticity of about 2300 Pa, whilst LM1
fraction (low MW) had 41 Pa of dynamic viscoelasticity.

Degree of Whiteness of PKF
Means of degree of whiteness of PKF effected by dipping stage and period are presented at
Figure 1 A and Figure 1 B, respectively. Analysis of variance of treatment combinations
between dipping stage and period were not very significant different at P<0.01.

Figure 1 Means of the degree of whiteness of PKF at different dipping stage (A) and dipping period
(B) with multi-stage ethanol leaching by ultrasound-assisted extraction.

The 12
th
16 -18 June, 2011

349

Generally, degree of whiteness of PKF at dipping stage 3 (Figure 1 A) was very significant
(at P> 0.01) higher than dipping stage 1, but it was not very significant (at P> 0.01) than
dipping stage 2. Whereas, PKF was dipped in ethanol solution for 25 minutes (Figure 1 B),
was significant (at P>0.05) higher than dipping period for 5 minutes. Chemat and others
(2009) observed by SEM studies, distinguishable physical changes on the extracted matrix
of caraway seeds without treatment and ultrasound-assisted extraction (UAE). After 30
minutes of UAE at 20
o
C, a huge perforation of the particles external surface and some
waste materials is dispersed. While untreated sample showed that the cell walls seem to be
thicker and intact. Therefore, extracting PKF with multi-stage ethanol solution at dipping
stage 3, revealed the highest degree of whiteness (49.25) compared to dipping stage 1
(46.33, Figure 1 A). Furthermore, Povey and Mason (1998) reported that UAE exhibited
stronger mechanical effects on extracted plant materials, whereby ultrasound waves
penetrates more power towards plant tissues matrix than classical extraction technique.
All treated konjac flour samples showed lower degree of whiteness than the untreated,
which was 51.15. Because, a color reader measures the degree of lightness (L value) of
PKF and meanwhile, the colour of CKF from Indonesia looks light brown. Takigami
(2000) reported that crude konjac flour from Japan had degree of whiteness 66-68, whilst
purified flour had 73. Whereas China konjac flour showed degree of whiteness of 69.9 for
ordinary flour and 68.9 for purified flour. Naturally, konjac tuber from Japan and China are
white in colour, and Indonesian konjac tuber is between orange to reddish in colour.

Calcium Oxalate of PKF
Calcium oxalate present in konjac tuber (Amorphophallus konjac) and cocoyam causes
irritant to skin, scratchiness in mouth and throat. The existence of this natural compounds
in wild tubers hindered the use of full potential of these tuber (Bradbury, 1988; Iwuoha
and Kalu, 1995). The lowest calcium oxalate (0.61%) exhibited by treated flour with multi-
stage ethanol leaching by UAE at dipping stage 3, whereas the dipping period for 25
minutes had the lowest calcium oxalate content (0.64%, Figure 2). Therefore, the results
indicated that the dipping stage 3 resulted the lowest calcium oxalate than the dipping stage
1 and 2. The longer period of dipping CKF in ethanol solution, the lower ca-oxalate
content in PKF. None published data reported in relation to calcium oxalate content of
PKF with multi-stage ethanol leaching by UAE.

Calcium oxalate content of control flour was 2.11%, and the calcium oxalate content of
CKF was dipped for 25 and 5 minutes under UAE was 0.64% and 1.61%, respectively.
The fact that the flour was dipped for 5 minutes showed higher amount of calcium oxalate
content than the one was dipped for 25 minutes in ethanol solution (Figure 2). Perdani and
Widjanarko (2009) reported that crude konjac flour contained 2.27 % calcium oxalate, in
the meantime, konjac flour was purified by boiling with 10% aluminium sulphate solution
and then precipated with 80% ethanol (technical grade) solution had 0.12% calcium
oxalate. This data is much lower than data at Figure 2, because of both those end product
are not similar. The former was purified konjac flour and the latter was KGM flour.

The 12
th
16 -18 June, 2011

350

Figure 2 Means of calcium oxalate of PKF at different dipping stage (A) and dipping period (B) with
multi-stage ethanol leaching by UAE.

Moisture Content and Total Yield of PKF
Commercial konjac flour should meet the standard of grade konjac flour as stated by
Anonymous (2006) that its moisture content is 10%. The presence of hydroxyl groups of
ethanol molecules create hydrogen bonds amongst hydroxyl groups within ethanol
molecules at -186
0
C. These hydrogen bonds cause pure crystall ethanol becomes
hygroscopic. Similarly chemical properties of concentrated ethanol solutions (96%) was
also hygroscopic (Anonymous, 2010). Therefore, ethanol solution used in this experiment,
which were 40, 60 and 80%, also absorb water from the treated flour. Hence, 80% ethanol
solution will absorbs less water than 40 and 60% ethanol solutions from CKF. This is due
the polarity of 40% ethanol solution was more polar than 60% and 80% ethanol as well.
On contrary, konjac flour is water-soluble non-ionic polysaccharide consists of mannose
and glucose units (Kaname and others 2003). Consequently, this CKF granule will
absorbs more water during ethanol leaching processes under UAE at 40% ethanol than 60
and 80% ethanol solution. Ultimately, CKF granules swell bigger in dilute ethanol solution
than the more concentrates of ethanol solution. Therefore it needs longer times to dry the
flour was dipped in 40% ethanol solution than the flour was dipped in 60 and 80% ethanol
solution, at 40
o
C in an oven, (looked at sample preparation).

As it was stated at the sample preparation that the drying period of the flour which is
dipped in 40% ethanol, was 32 hours at 40
o
C, 24 hours for 60% and 12 hours for the
flour is dipped in 80% ethanol solution. Konjac flour is unstable when it is dried at high
temperature, so 40
o
C by oven drying was chosen to avoid such damage on dried flour.
James (1999) noted that the period of drying process to determine moisture content from
plant materials is, if it is dried by an oven, vacuum oven, microwave oven or other
methods of drying and reaches a constant weight. Therefore, the final moisture content of
those treated flour and untreated flour are nearly the same.

A B
The 12
th
16 -18 June, 2011

351

Table 2 Means of moisture content and total yield of PKF were affected by different dipping stage and
period with multi-stage ethanol leaching by UAE.

Dipping stage Dipping period
(minute)
Moisture content
(%)
Total yield (%)

Control
1

-
5

11.76 0.02
12.44 0.06

-
95.03 3.79
15 12.35 3.41 94.32 1.77
25 12.05 0.23 94.03 0.51
2 5 12.45 2.54 97.25 1.77
15 13.54 0.23 96.89 0.23
25 15.94 4.14 95.44 0.41
3 5 14.22 0.87 96.21 0.61
15 11.11 1.67 95.46 2.12
25 13.80 0.90 93.99 1.66

Total yield of treated flour was not effected with by dipping stage and period (Table 2).
Total yield of final dried flour were in between 93.99%1.66 and 97.25%1.77. Meta and
Widjanarko (2010) reported that total yield of glucomannan extracted with water
containing 10% trichloroacetic acid (TCA) by a transducer UAE, then precipitated with
80% ethanol is between 67.4 and 68.6%. The differences amongst those data are mainly
due to the method of purification of konjac flour. The former was not using gelling agent
(TCA), therefore the end product is PKF, and the latter used TCA as gelling agent to
precipitate impurities of CKF, and the end product is KGM, but the end product is difficult
to be grinded and it is difficult to dissolve in water.

Physico-chemical properties of best treated flour, untreated and a commercial KGM
The best treated flour is determined by giving a maximum or minimum ideal value at all
physico-chemical properties which are tested based on the analysis of multiple attribute
(Zeleny, 1982). Maximum ideal values are set to glucomannan, viscosity and degree of
whiteness of the best treated flour and minimum ideal value is set to calcium-oxalate
content of the best treated flour. So, the best treated flour should have physico-chemical
properties as follows: It has higher glucomannan content, viscosity and degree of
whiteness and lower calcium oxalate content than other treatments used in this experiment.
Using multiple attribute method as described by Zeleny (1982), it has been concluded that
the best treated flour, if CKF is dipped in ethanol solution with dipping stage 3 for 25
minutes by UAE. The results showed that glucomannan content and viscosity of the best
treated flour increased, meanwhile, the calcium-oxalate content decreased, but the degree
of whiteness was lower than control. In this experiment the changes in physico-chemical
attributes on treated samples are expected.

The 12
th
16 -18 June, 2011

352

Table 3. Means of physico-chemical properties of best treated flour, untreated and a commercial KGM

Type of konjac flour
Item Best Treated
*)
Untreated
*)
A commercial KGM
*)

Glucomannan (%) 87.83 71.83 92.51
Ca-Oxalate (%) 0.63 2.11 0.08
Degree of whiteness 49.09 51.15 62.86
Viscosity (c.Ps) 8200 4900 11000
Rehydration period (Sec.) 910 1320 40
*) Data is taken in duple.

Glucomannan content of the best treatment was higher (87.83%) than control (71.83%),
but relatively lower than a commercial KGM (92.51%). Peiying and others (2002) reported
that konjac flour standard for top grade quality, it should have glucomannan content
70%. Whereas ca-oxalate of the best treated flour is lower than the untreated, but the
lowest ca-oxalate was indicated by KGM (Table 3). Bradbury (1988) repored that several
oxalate fractions are soluble in water at room temperature, and Ca-oxalate present in tobico
will be leached during ethanol leaching processes. All impurities including ca-oxalate
attached on top of konjac flour granules surface will be washed away during ethanol
leaching processes and stay with ethanol solution waste. Tobico is a very light fractions of
konjac flour which contains impurities such as: Ca-oxalate, protein, starch etc) and the size
of particle is less than 1x10
-2
mm in diameter (Shimizu and others 1973), this light and
very small particle will be swept into ethanol solution.

The best treated flour was more viscous than the untreated flour, but its viscosity was still
lower than KGM. Huang and others (2002) noted that increasing KGM concentration,
shortened gelling time as the viscosity increases. The degree of whiteness between the
best treated flour and untreated flour remained nearly the same, and the highest value of
the degree of whiteness was noted by KGM powder. Syaefullah (1990) claimed that the
flesh color of Indonesian konjac tuber is reddish yellow due to the present of natural
pigment in konjac tuber. Rehydration period of KGM was only for 40 seconds, while the
best treated flour was 910 seconds, the longest time to dissolve in water was shown by the
untreated flour. This reveals by SEM studies of KGM granules is between 75.39 and 156.5
um and those granules of both the best treated and untreated flour were between 295.5
and 400.8 um (Widjanarko and others 2011
b
). The smaller the granule size, the quicker the
granule to dissolve in an aqueous solution.

Conclusion
Multi-stage ethanol leaching by UAE can increase glucomannan content, viscosity and
reduce calcium oxalate of PKF. The best treatment in purifying CKF, if CKF was dipped
in ethanol solution is at dipping stage 3 for 25 minutes by UAE. The best purified konjac
flour (PKF) had the highest glucomannan and viscosity than the untreated flour, but the
degree of whiteness was lower than the untreated. Moisture content and total yield of
treated konjac flour were not affected by treatments combinations of dipping stage and
period with multi-stage ethanol leaching by UAE. Rehydration period of KGM was the
fastest (40 sec.) than the best treated flour (910 sec.) and the untreated flour (1320 sec.)

The 12
th
16 -18 June, 2011

353

Acknowledgement
This work was supported by Postgraduate Research Project, Directorate of Higher
Education, Ministry of Education, Republic of Indonesia, University of Brawijaya Annual
Budget. Rector Decree, Number: 039/SK/2010, February 17, 2010.

References
Anonymous. (2006). What Is Konjac. http://www.konjacbaoji.com/properties.htm.
Retrieved date, March 6, 2010.
Annonymous. (2009). Konjac Mannan Gel Powder. Leaflet GfN Wirstoffe Vertriebs
GmbH. Page 1.www.Gfn-selco.de. info@Gfn-selco.de.
Anonymous, (2010). Ethanol. http://www.wikipedia.com. Retrieved date, July 26, 2010
AOAC, (1984). Official Methods of Analysis. Association of Official Analysis Chemistry.
Washington
Arifin, M. A., (2001). Method of drying konjac chips by mechanical machinery to improve
konjac chip quality. Master thesis of Postharvest Technology Study Programme,
Postgraduate Programme. Institute of Agricultural Bogor.
Bradbury, O.H. (1988), Chemistry of Tropical Root Crops: Significance for Nutrition and
Agriculture In The Pacific, Chemistry Department Australian Centre for
International Agricultural Research, Canberra.
Chan, A.P.N. (2009). Konjac Part I. Cultivation to Commercialization of Component.
http://www.worldfoodscience.org/cms/?pid=1003556. Retrieved date, November 3,
2010.
Chemat, F., V. Tomao, and M. Virot. (2009). Ultrasound-Assisted extraction in food
analysis. In Semih Otles. Eds. 2009. Handbook of Food Analysis Instruments.
CRC Press. Taylor & Francis Group, N.Y. pp. 85-99.
Diak, K.M.P. and Widjanarko, S.B. ( 2010). Studies on different concentration of
Trichloroacetic acid in producing Glukomannan from common konjac flour
(Amorphophallus Oncophyllus). Unpublised report. Depart. Of Food Technology,
Faculty of Agricultural Technology. Brawijaya University Malang, Indonesia.
Erryana, I.F., and Widjanarko, S.B. (2010). Studies on different ethanol concentration in
producing Glukomannan from common konjac flour (Amorphophallus
Oncophyllus). Unpublised report. Depart. Of Food Technology, Faculty of
Agricultural Technology. Brawijaya University Malang, Indonesia.
Harrison, H. (2009). Information about konjac root. Http://www.eHow.com/
/facts_5804458_information-konjac-root.html. Retrieved November 6, 2010.
Huang L, Takahashi R, Kobayashi S, Kawase T, Nishinari K. 2002. Gelation behavior of
native and acetylated konjac glucomannan. Biomacromolecules, Nov-Dec;
3(6):1296-1303.
Iwuoha, C.I. and F.A. Kalu. (1994). Calcium Oxalate and Physico-Chemical Properties of
Cocoyam (Colocasia esculenta and Xanthosoma sagittifolium) Tuber Flours as
Affected by Processing. Department of Food Science and Technology, Federal
University of Technology, PMB 1526, Owerri, Nigeria.
James, C.S. (1999). Analytical chemistry of Foods. An Aspen Publication. p. 178.
Kaname, K., Kohsaku, O., Kenichi, H., Ryuichi., O., Takaya, S., and Kei, M. (2003).
Constitution of Konjac glyucomanan; Chemical analysis and 13C NMR
Spectroscopy. Carbohydrate Polymer. 53:183-189.
Koeswara, S. (2009). Ebook on Food : Konjac and its food processing product
characteristic. http://ebookpangan.com Retrieved date, August 1, 2009.
The 12
th
16 -18 June, 2011

354

Ma, Y., J. Chen, D. Liu and X., Ye.(2009). Simultaneous extraction of phenolic
compounds of citrus peel extracs: Effect of Ultrasound, J. Ultrasonics
Sonochemistry, 16:57-62
Meta, A and Widjanarko, S.B. (2010). Extraction of konjac glucomannan (KGM) from
common konjac (Amorphophallus oncophyllus) flour (CKF) by Ultrasonification.
Studies on the effect of proportion of konjac flour and duration period of extraction.
Unpublised report. Depart. Of Food Technology, Faculty of Agricultural
Technology. Brawijaya University Malang, Indonesia.
Nishinari, K. (1997). Rheological and DSC study of sol-gel transition in aqueous
dispersions of industrially important polymers and colloids. ColloidPolymer
Science., 275, 1093-1107.
Ohashi,S., Shelso, G.J., Moirano D., Arthur L. and Drinkwater, W. L. (1988).
FreePatentOnline. Clarified konjac glucomannan. Japanese Patent 63-68054 (Mar.
26).
Perdani, G. C. and Widjanarko, S.B. (2009). Studies on the effect of salt solutions and
Trichloroacetic acid in purifying common konjac (Amorphophallus oncophyllus)
flour on physico-chemical characteristics of konjac glucomannan (KGM).
Unpublised report. Depart. Of Food Technology, Faculty of Agricultural
Technology. Brawijaya University Malang, Indonesia.
Peiying, L., Zhang, S., Zhu, G., Chen, Y., Quyang, H., Han, M., Wang, Z., Xiong, W., and
Peng, H. (2002). Professional Standart of The Peoeple Republic of China for
Konjac Flour. NY/T : 494-2002
Povey, M and Mason, T.J. (1998). Ultrasound in Food Processing, Povey, M. and Mason,
T.J. (Eds.), Blackie Academic &Professional, London.
Rochas, C., Rinaudo, M., and Landry, S. (1990). Role of molecular weight on the
mechanical properties of kappa carrageenan gels. Carbohydrate Polymer., 12:255-
266.
Salisofa M. Toma, S., and Mason, T.J. (1997). Comparation of conventional and
ultrasonically assisted extractions of pharmaceutically active compounds from
Salvia offininalis, Sonochem. 4:131-134.
Schinor , E.C., Salvador, M.J., Turatti, I.C.C., Zucchi, O.L.A.D. and Dias, D.A. (2004).
Comparison of classical and ultrasound-assisted extractions of steroids anf
triterpenoids from three Chresta spp. Ultrason, Sonochem. 11:415-421.
Shimizu,M. and H. Shimahara. (2004). Method Of Selective Separation Of Konjac Fluor
From The Tubers Of Amorphophallus Konjac.
http://codex.foodnara.go.kr/lib/base_down.jsp?dr=eu&fn=konjac%2 gum.pdf.
Retrieved date May 14, 2006.
Sudarmadji, S. B, Haryono and Suhardi. (1997). Procedure of Food Analysis. Liberty
Press. Jogjakarta.pp. 50-85.
Sugiyama, N., S. Shimara, and T. Ando. 1(972). Studies on Mannan and Related
Compounds I. The Purification of Konjac Mannan. Bulletin Chem. Soc. Of Japan.
45:561-56
Sun, R.C. and J. Tomkinson. (2002). Comparative study of lignin isolated by alkali and
ultrasound-assisted alkali extractions from wheat straw, J. Ultrasonics
Sonochemistry, 9: 85-93.
Takigami, S., (2000). Konjac Mannan. Dalam G.O. Phillips; and P.A. Williams, Eds.
Handbook of Hydrocolloids, pp. Woodhead. Cambridge.
The 12
th
16 -18 June, 2011

355

Ukpabi U.J and Ejidoh, J.I. (1989). Effect of deep oil frying on the oxalate content and the
degree of itching of cococyam (Xanthosoma and Colocasia spp). Technical paer
presented at the 5
th
Annual Conference of The Agricultural Society of Nigeria,
Federal University of Technology, Owerri, Nigeria, 3-6 Sept.
Vinatoru, M., Toma, M., Radu, O., Filip, P.I., Lazurca, D and Mason, T.J. (1997). The use
of ultrasound for the extraction of bioactive principles from plant materials,
Ultrason. Sonochem. 4:135-139.
______, M. (2001). An overview of the ultrasonically assisted extraction of bioactive
principles from herbs. Ultrason, Sonochem. 303-313.
Ward, A.G. and Saunders, P.R. (1955). Mechanical properties of degraded gelatines,
nature, 176:26-29.
Watase, M., and Nishinari, K. (1983). Rheological properties of agarose gels with
different molecular weight, Rheol. Acta, 22: 580-587.
Widjanarko, S.B. (2010). Processing, marketing, and konjac based food products.
Http://simonbw.lecture.ub.ac.id/files/2010/03/KONJAC-SBW-PROCESS.pdf.
pp.1-3. Retrieved date, October 25, 2010.
_____, S.B., Nugroho, A and Estiasih, T. 2011
a
. Functional interaction components of
protein isolates and glucomannan in food bars by FTIR and SEM studies. African
Journal of Food Science Vol 5.(1), pp. 12 - 21, January.
______. S.B., Faridah, A and Sutrisno, A. 2011
b
. Effect Of MultiStage Ethanol Leaching
By Maseration On PhysicoChemical Properties Of Konjac Flour
(Amorphophallus Oncophyllus). Unpublished report. Depart. Of Food Technology,
Faculty of Agricultural Technology. Brawijaya University, Malang, Indonesia.
Yang, Y. and F. Zhang. (2008). Ultrasound-assisted extraction of rutin and quercetin from
Euonymus alatus (Thunb) Sieb, J.Ultrasonic Sonochemistry, 15:308-313.
Yoshimura, M. and K. Nishinari. (1999). Dynamic viscoelastic study on the gelation of
konjac glucomannan with different molecular weights. Food Hydrocolloids 13:
227-233.
Zeleny, M. 1982. Multiple Criteria Decision Making. Mc Graw-Hill. New York. pp. 60 -
75.
Zhang, H., Yoshimura, M., Nishinari, K. Williams, M.A.K., Foster, T.J. and I.T. Norton.
(2001). Gelation behaviour of konjac glucomannan with different molecular
weight. Biopolymers. 59:38-50.
The 12
th
16 -18 June, 2011

356

OJ-134

Preparation of Inulin Powder from Dioscorea esculenta Tuber with
Foam Mat Drying Method

Eni Harmayani
1,*
,

Sri Winarti
2
and Rudi Nurismanto
2

1
Department of Food and Agricultural Technology, Faculty of Agricultural Technology, Gadjah Mada
University, Yogyakarta, Indonesia. E-mail: eniharmayani@yahoo.com;
2
Department of Food Technology, Faculty of Industrial Technology, UPN Jl Rungkut Madya,
Surabaya, Indonesia, 60294. E-mail : swin_tpupn@yahoo.com
*
Corresponding author: eniharmayani@yahoo.com

Abstract

Inulin is a natural nondigestible fructooligosaccharide that occurs as a reserve carbohydrate
in several plants. Currently, chicory is an excellent source of inulin. However, chicory is a
plant that grown primarily in Europe. Lesser yam (Dioscorea esculenta) is one of local
tuber that has potential as a source of inulin. Our previous study indicated that higher
amount of inulin could be extracted from Dioscorea esculenta as compared with the other
tubers. This research was conducted to study the production of inulin powder from lesser
yam using foam mat drying method. The specific objective of the study was to examine the
effect of filler types (dextrin, maltodextrin, Na-CMC) and foaming agent concentration (2-
10% of egg white) on the yield and characteristic of the inulin powder. Inulin was prepared
by the following method. Lesser yam was mixed with hot water at 90 C (1:20), decanted
for 1 hour, filtered and the filtrate was freezed at -20 C. After thawing, the filtrate was
centrifuged at 1500 rpm for 15 min. The precipitate was mixed with filler and foaming
agent, then dryed at 60 C for 6 hours. Inulin powder was characterized for its solubility,
water content and purity. The inulin purity was determined by HPLC with Aminex HPX-
87c column. Data were analyzed using Analysis of Variance (ANOVA) and further tested
using Duncan't Multiple Range Test (DMRT). The results showed that combination of
dextrin as filler and 2% egg white produced highest yield of inulin powder (10.20%) with
optimum precipitation temperature at -20 C. The water content in inulin powder was
10.77%, with 99.97% solubility and 85.83% purity.

Keywords : Inulin, Dioscorea esculenta, powder, foam mat drying, preparation

The 12
th
16 -18 June, 2011

357

OJ-152
The Meat Quality of Japanese Quail as Affected by
Probiotic Administration

U. Kalsum
*
, H. Soetanto, Achmanu, and O. Sjofjan

Faculty of Animal Husbandry, Brawijaya University, Malang East Java- Indonesia.

*
Corresponding author: kalsum2008@gmail.com

Abstract

This study was conducted to ascertain the effect of adding probiotic into feed on the quality
of Japanese quails meat. Lactobacillus salivarius from Japanese quail intestine was used
as probiotic. A randomized block design was used to assign four levels of L. salivarius
administration; R0 = Standard feed without probiotic, R1 = 0.2 % probiotic, R2 = 0.4 %
probiotic and R3 = 0.6 % probiotic, and four hen day production levels; (1) 14.28 %; (2)
19.64 %; (3) 25.43 %, and (4) 35.71 %. The results showed that adding probiotic did not
alter significantly feed intake, carcass weight, carcass percentage, water holding capacity
(WHC) and tenderness of meat.

Keywords: Meat quality, Japanese quail, and probiotic

Introduction
Quail meat has a high prospect to be developed in addition to delicious, rich in protein, low
in fat, and safe for health. Recent development shows that the demand for poultry products
is not just for protein consumption but also for safety. This is due to information about
anti-biotic residue in poultry products that cause adverse effects not only for animal but
also for consumers. One effort to replace these antibiotics is by use of the probiotic, since
they are able to balance the composition of microbes in the digestive tract. It is expected
that administering probiotic in livestock is a solution to the problem of residues of these
antibiotics.

The first phase of Ph.D. research has been successfully carried out the isolation and
characterization of several strains of microbes from the quail digestive tract and the genus
of Bacillus was found. The genus can produce several enzymes such as protease (Libertina
and others 2009) and amylase (Wardhani and others 2009), therefore they potentially can
be exploited as a source of probiotic. They had some dominant probiotic microbes, one of
them identified as Lactobacillus salivarius (Kalsum and others 2008).

The use of this probiotic as feed additive needs to be explored, especially to determine the
quality of probiotic produced and its effect on optimize the work of the natural microbes in
the digestive tract. Kalsum and others (2010) reported that endogenous probiotic such as
L.salivarius can be used to improve the condition of the digestive tract of quail. Therefore
to increase feed efficiency by improving the condition of the digestive tract of animal, so
they can produce nutritious food products and safe for our health. In addition, Afdora and
others (2010) reported that L. Salivarius isolated from quail intestine can product
antibacterial compounds that could inhibit the growth of microbial pathogens such as
Salmonella typhimurium, Eschericia coli of human origin, and E. coli of bird origin.

The 12
th
16 -18 June, 2011

358

It needs to be studied for the use of endogenous probiotic for quail and production of stable
probiotic, also a biological test of the ability of probiotic to improve quail performance
especially about the proper dose. Therefore it needed a more study on production of
endogenous probiotic as feed additive to improve the nutritional value and digestibility.
Further research will be assessed how the relationship between the administrations of
endogenous probiotic as feed additive for quail on meat quality. So it will be a model of
feeding management more practical, economical and sustainable through the use of
endogenous probiotic as feed additives for quail.

The research was conducted at the Field Laboratory, Faculty of Animal Husbandry.
Production of probiotic was conducted at the Laboratory of Microbiology, Faculty of
Mathematics and Natural Sciences. The meat quality analysis was done at the Laboratory
of Animal Product Technology Faculty of Animal Husbandry Brawijaya University,
Malang, East Java-Indonesia.

Endogenous microbe isolated from Japanese quail intestine was used as probiotic. It was
cultured in MRS for 48 hours. Species of endogenous microbe (Lactobacillus salivarius)
was confirmed with the API 50 CH test Medium Kit (Biomerieux) and API 50 CHL
medium. To determine feasibility as a probiotic candidate needed several tests such as pH
test, pathogen test, bile salts test, etc. The best probiotic produced in the first phase of
Ph.D. research was used as feed additive. For biological test 63 days old of 128 tails quails
were raised in 16 plots cage system with 40 x 40 x 25 cm
3
per plot.

A randomized block design was used to assign four levels of probiotic administration; R0
= standard feed without probiotic, R1 = standard feed with 0.2 % probiotic, R2 = standard
feed with 0.4 % probiotic and R3 = standard feed with 0.6 % probiotic, and four hen day
production levels; (1) 14.28 %; (2) 19.64 %; (3) 25.43 %, and (4) 35.71 %. The
composition and nutrient content of feeds were metabolic energy by 2700.09 kcal/kg,
protein by 20%, fat by 3.48%, crude fiber by 4.53%, 2.62% of calcium and 0.42% of
phosphor.

Results
Result of biological test using endogenous probiotic (Lactobacillus salivarius) as feed
additive for quail included in table 1.

Table 1 Effect of Probiotic on Feed Intake, Carcass Weight, Carcass Percentage, Water Holding
Capacity and Tenderness of Meat
Treatments
Feed
intake (g)
Carcass
weight (g)
Carcass
percentage (%)
Water Holding
Capacity
Meat tenderness
(mm/g/s)
Ro 18.29
81,04 58,15 68,68 0.15
R1 17.45 85,05 57,19 63,90 0.10
R2 18.74 89,06 58,44 69,06 0.10
R3 18.74 83,39 56,68 71,05 0.10

The 12
th
16 -18 June, 2011

359

Discussion
Feed intake
Analysis of variance showed that the use of probiotic in the diet did not affect significantly
(P> 0.05) on feed intake. This is due to the nutrient content of standard feed was similar in
composition (iso-energi), each treatment was only distinguished by the addition of a
probiotic as feed additive of 2 to 6 grams per kilogram of feed. Metabolism energy of feed
is one of several factors that affect on feed intake of poultry. According to Ferket and
Gernat (2006) that energy demand and energy content in poultry feed would affect
consumption. Poultry will eat more if metabolism energy of feed is poorly; because poultry
will consumption feed for supply their energy demand. It will stop consumption feed if
their energy demand have completed for maintenance and growth or production.

Carcass weight
The endogenous probiotic as feed additive that used in this study was the lactic acid
bacteria identified as Lactobacillus salivarius. The use of this microbe as probiotic was
supported by Jans (2007) that most of probiotic used as feed additive for broiler was
Lactobacillus bacteria and Bifidobacterium, and other types of fungi such as Aspergillus
niger and Aspergillus oryzae.

Statistical analysis showed that the endogenous treatment of the used of probiotics in the
diet did not significantly affect on carcass weight. The results were consistent with
Chumpawadee (2009) that the use of probiotics such as Saccharomyces cerevisiae by
1x10
8
organisms/kg feed did not given a significant impact on quail performance.

Carcass percentage
Analysis of variance showed that the use 0.6% of L. salivarius in feed was not given a
significant influence on the carcass percentage. The results are consistent with Kalsum
(2007) that the use of Candida utilis as probiotic cultures in fermented dry poultry waste
and cassava did not show significantly affect on the percentage of carcass in quail.

WHC (Water Holding Capacity)
The used 6 grams of L. salivarius per kilogram of diet was not given a significant impact
on WHC of quail meat. This meant that the nutrients in feed were preferred to the
establishment of meat protein. According to Kalsum and others (2010) that probiotic as a
producer of single-cell protein have a higher capacity in assimilating number compounds
of carbon and nitrogen than other types of microorganisms, so they produced more
compounds that were needed for the formation of meat protein. According to Van
Koevering (1995) that WHC meat value were influenced by the quality of protein in feed.
Contrary to Homma and Shinohara (2004) that the used of commercial probiotic such as
Bacillus cereus toyoi as feed additive for quail was able to reduce abdominal fat and fat
percentage of meat, and alter protein percentage of meat.

Meat tenderness
Based on Variance analysis, the use 0.6% of endogenous probiotic in feed did not
significantly affect on quail meat tenderness. This was due to the low cross linking of meat
filament so the water trapped in the tendon was increases and it had a high WHC.
According to Wheeler and others (1999) that the increase in meat tenderness could be
The 12
th
16 -18 June, 2011

360

reflected by the water content and WHC was high and the swelling properties of the meat
fibers.

References
Afdora PT, Ardiyarti T, Osfar S, Kalsum, U, editors. 2010. Proceedings of International
Conference on Global Resource Conservation (ICGRC 2010); 2010 July 8
th
;
Department of Biology, Faculty of Sciences, Brawijaya University, Malang,
Indonesia. 322 p.
Chumpawadee S, Chinrasri O, Santaweesuk S. 2009. Effect of Dietary Inclusion of
Cassava Yeast as Probiotic Source on Growth Performance and Carcass Percentage
in Japanese Quails. Pakistan J of Nutrition 8(7):1036-9.
Ferket PR, Gernat AG. 2006. Factors That Affect Feed Intake of Meat Birds. J Poult Sci
5(10):905-11.
Homma H, Shinohara T. 2004. Effects of probiotic Bacillus cereus toyoi on abdominal fat
accumulation in the Japanese quail (Coturnix japonica). J Animal Sci 75(1):37-41.
Jans D. 2007. Probiotics for Animal Nutrition. FEFANA, EU Association of Feed
Additives Procedures.
Kalsum U, editors. 2007. Proceeding International Conference on Molecular Biology of
Life Sciences. Central Laboratory of Life Science; 2007 Nov 19-21; Brawijaya
University, Malang. 247 p.
Kalsum U, Sofjan O, Suryanto D, Aulaniam, editors. 2008. Proceedings International
Research Seminar and Exhibition on Isolation, Selection and Characterization
Microbe from Intestinal Tract of Quail as Source of Probiotic; 2008 Nov 7-8;
Muhammadiyah Malang University, Indonesia.
Kalsum U, Osfar S, Dedi S, editors. 2010. Proceedings International Conference on Global
Resource Conservation (ICGRC 2010); 2010 July 8
th
; Brawijaya University,
Malang, Indonesia.
Libertina I, Tri A, Osfar S, Umi K, editors. 2009. Proceedings International Research
Seminar and Exhibition International Symposia on The Recent Advances of
Microbiology in Health, Bioindustry, Agriculture and Environment; 2010 Nov 20-
21; PERMI Surabaya, Indonesia.
Van Koevering MT, Gill FN, Owens DR, Dolezal HG, Strasia CA. 1995. Effect of Time
on Performance of Feedlot Steers, Carcass Characteristics and Tenderness and
Composition of Longissimus Muscles, J Animal Sci 73:21-8.
Wardhani WK, Tri A, Osfar S, Umi K, editors. 2009. Proceedings International Research
Seminar and Exhibition International Symposia on The Recent Advances of
Microbiology in Health, Bioindustry, Agriculture and Environment; 2010 Nov 20-
21; PERMI Surabaya, Indonesia.
Wheeler JL, Savell JW, cross HR, Lunt DK, Smith SB. 1999. Effect of Postmortem
Treatments on the Tenderness of Meat. J Animal Sci 77: 3677-86.
The 12
th
16 -18 June, 2011

361

OJ-196

Antioxidative Activity of Protein Hydrolysates from Bambarra
Groundnut (Voandzeia subterranea)

Vilailak Klompong
1,*
, Natawan Mingmolee
1
, Frukaree Bintaleb
1
and Soottawat Benjakul
2

1
Department of Food Science and Technology, Thaksin University, Papayom, Phattalung 93110 Thailand,
2
Department of Food Technology, Faculty of Agro-Industry, Prince of Songkhla University,
Hat Yai, Songkhla 90112 Thailand
*
Corresponding author: klompong@hotmail.com

Abstract

Antioxidative activity of protein hydrolysates from Bambarra groundnut (Voandzeia
subterranea) protein, hydrolyzed by Protamex (HP) and Flavourzyme (HF) with different
degrees of hydrolysis (DH) were investigated. As the DH increased, DPPH radical
scavenging activity and ferric reducing antioxidant power of HP decreased (p<0.05). For
HF, DPPH radical scavenging activity decreased (p<0.05) but no difference was observed
for ferric reducing antioxidant power (p>0.05), as the DH increased. Metal chelating
activity, ABTS radical scavenging activity and hydroxyl radical scavenging capacity of
both HP and HF increased with increasing DH (p<0.05). HP generally had a higher
(p<0.05) DPPH radical scavenging activity, ferric reducing antioxidant power, ABTS
radical scavenging activity and hydroxyl radical scavenging capacity than HF at the same
DH tested. However, HF showed a higher (p<0.05) metal chelating activity than HP. HP
and HF at 200 ppm were able to delay the formation of thiobarbituric acid (TBA) in
lecithin liposome system. At high DH, HP and HF generally, possessed the stronger
antioxidative activity than at lower DHs (p<0.05). HP possessed the stronger antioxidative
activity than HF (p<0.05), however -tocopherol at 200 ppm showed the greatest
antioxidative activity in the system. The results reveal that antioxidative activity of protein
hydrolysates from Bambarra groundnut were determined by the DH and by the enzyme
types used.

Keywords: Protein hydrolysate, antioxidant, Bambarra groundnut, degree of hydrolysis

Introduction
Bambarra groundnut is indigenous nut of the southern part of Thailand. Although
Bambarra groundnut has been used in local area as food, the fully potential use has not
been executed. Hydrolysis processes have been used to convert native protein with no
functionality into more marketable and acceptable forms. Enzymatic hydrolysis potentially
influences the antioxidative activity of fish protein hydrolysate (Klompong and others
2007). Therefore, the objective of this study was to produce protein hydrolysate from
Bambarra groundnut with various DHs using two different enzymes and to study their
antioxidative activities.

Bambarra groundnut (Voandzeia subterranea), was obtained from Phatthalung province
area. The groundnut was peeled manually, baked at 60 C for 24 h. Thereafter, the
groundnut was ground using a grinder. The ground Bambarra groundnut was defatted and
de-flour. The Bambarra groundnut protein obtained was subjected to enzymatic hydrolysis
The 12
th
16 -18 June, 2011

362

using Protamex and Flavourzyme by the pH-stat method as described by Adler-Nissen
(1986). DH was then calculated as follows:
DH (%) = BN
b
x 100
Mp h
tot

Log
10
(enzyme concentration) vs. DH was plotted. From the regression equation, the
enzyme concentrations required to hydrolyze Bambarra groundnut protein to obtain the
desired DHs (1, 3, 5%) were calculated. The protein hydrolysates were subjected to
analyses for antioxidative activities, including 2,2-Diphenyl-1-picryhydrazyl (DPPH),
ABTS and hydroxyl radical scavenging activity, ferric reducing antioxidant power
(FRAP), metal chelating activity and antioxidative activities in lecithin liposome system.

Effect of enzyme type and concentration and time on degree of hydrolysis of
Bambarra groundnut protein hydrolysate

Figure 1 Degree of hydrolysis (DH) of Bambarra groundnut hydrolyzed with Protamex (a) or
Flavourzyme (b) at different concentrations (0.25% ( ), 0.5% ( ), 1% ( ) and 2%
( )

Hydrolysis of Bambarra groundnut using Protamex or Flavourzyme was carried out by the
pH-stat method (Fig. 1). Rapid hydrolysis was found within the first 10 min. Thereafter, a
slower rate of hydrolysis was observed up to 120 min. The typical hydrolysis curve was
also reported for Pacific whiting solid wastes (Benjakul and Morrissey 1997).
Additionally, the result showed that peptide bonds were more likely cleaved in the
presence of a higher amount of enzyme (Klompong and others 2007a). With the same
protein substrate and the same amount of enzyme, HP showed a higher DH than did HF
over the entire hydrolysis time. Except for at the enzyme concentration of 2%, HF showed
the higher DH than did HP. That might be due to the difference in enzyme substrate ratio.
The higher DH observed with HP indicates higher proteolytic activity of Protamex towards
Bambarra groundnut proteins, compared to Flavourzyme. Generally, endo-peptidase,
including Protamex, show higher activities than does exo-peptidase such as Flavourzyme.
Thus, the susceptibility to hydrolysis of Bambarra groundnut proteins depends on the type
of enzyme used.

b
a
The 12
th
16 -18 June, 2011

363

Figure 2 Relation between DH and log enzyme concentration of Protamex ( ) and
Flavourzyme ( ) in Bambara groundnut. Different amounts of enzyme were added
to the Bambara groundnut protein. The reaction was run for 60 min at pH 7, 50 C.

When log
10
of enzyme concentration and DH were plotted, a linear relationship was
observed as showed in Fig. 2. From the regression, the exact concentration of enzyme
required to hydrolyze Bambarra groundnut proteins to obtain the specific DH could be
calculated.

Antioxidative activity of protein hydrolysates from Bambarra groundnut protein,
hydrolyzed by Protamex (HP) and Flavourzyme (HF) with different DHs (1, 3, 5% DH)
were depicted in Fig. 3. As the DH increased, DPPH radical scavenging activity and ferric
reducing antioxidant power of HP decreased (p<0.05). For HF, DPPH radical scavenging
activity decreased (p<0.05) but no difference was observed for ferric reducing antioxidant
power (p>0.05), as the DH increased. Metal chelating activity, ABTS radical scavenging
activity and hydroxyl radical scavenging capacity of both HP and HF increased with
increasing DH (p<0.05). HP generally had a higher (p<0.05) DPPH radical scavenging
activity, ferric reducing antioxidant power, ABTS radical scavenging activity and hydroxyl
radical scavenging capacity than HF at the same DH tested. However, HF showed a higher
(p<0.05) metal chelating activity than HP.

The 12
th
16 -18 June, 2011

364

Effect of degree of hydrolysis (DH) and enzyme type on antioxidative activity of
Bambarra groundnut protein hydrolysate

Figure 3 DPPH (a), ABTS (b) and hydroxyl radical scavenging activity (c), FRAP (d) and
metal chelating activity (e) of Bambarra groundnut protein hydrolysate produced
using Protamex () and Flavourzyme () with different DHs. Bars represent
standard deviation from triplicate determinations. Different letters within the same
parameter indicate the significant differences (p<0.05).

Antioxidative activity of protein hydrolysates depends on the proteases and hydrolysis
conditions employed (Jun and others 2004). During hydrolysis, a wide variety of smaller
peptides and free amino acids is generated, depending on enzyme specificity (Kristinsson
and Rasco 2000). Changes in size, level and composition of free amino acids and small
peptides affect the antioxidative activity (Wu and others 2003). The result reveals that the
Bambarra groundnut hydrolysates potentially contained substances which were electron
donors and could react with free radicals to convert them to more stable products and
terminate the radical chain reaction. Additionally, peptides in hydrolysates could chelate
the prooxidants, leading to decreased lipid oxidation. From the result, the peptides in HP
and HF could act both as primary and secondary antioxidants.

a
b
c
d
e
The 12
th
16 -18 June, 2011

365

Figure 4 The formation TBA in lecithin liposome systems containing Bambarra groundnut
protein hydrolysate using Protamex (a) and Flavourzyme (b) at the DH of 1%
( ), 3% ( ) and 5% ( ), -tocopherol at 200 ppm ( ) and the control
( ). Bars represent standard deviation from triplicate determinations.

HP and HF at 200 ppm were able to delay the formation of thiobarbituric acid (TBA) in
lecithin liposome system (Fig. 4). At high DH, HP and HF generally, possessed the
stronger antioxidative activity than at lower DHs (p<0.05). HP possessed the stronger
antioxidative activity than HF (p<0.05), however -tocopherol at 200 ppm showed the
greatest antioxidative activity in the system. The results reveal that antioxidative activity of
protein hydrolysates from Bambarra groundnut were determined by the DH and by the
enzyme types used. A longer induction period indicates a stronger antioxidative activity
(Wu and others 2003). -Tocopherol is a donor antioxidant (reductant) that will react with
peroxyl radical. Thus, it served as a chain breaking antioxidant, inhibiting the propagation
cycle of lipid peroxidation. Klompong and others (2007b) found that fish protein
hydrolysates showed antioxidative activity in liposome systems and hydrolyzing with
Alcalase showed the lower antioxidative activity in the liposome systems than did
hydrolyzing with Flavourzyme.

From the result, it could be concluded that the DH of Bambarra groundnut protein
hydrolysate was governed by amount and type of enzymes used and hydrolysis time.
Bambarra groundnut protein hydrolysates possessed antioxidative activities and
antioxidative activities depend on the proteases employed and DH. Therefore, Bambarra
groundnut protein hydrolysates could be used as natural antioxidants.

Acknowledgements
We would like to thank department of Food Science and Technology, Thaksin University
for financial support.

References
Adler-Nissen J. 1986. Enzymic hydrolysis of food proteins. New York: Elsevier Applied
Science Publishers.
Benjakul S, Morrissey MT. 1997. Protein hydrolysates from Pacific whiting solid wastes. J
Agric Food Chem. 45: 3423-30.
Jun SY, Park PJ, Jung WK, Kim SK. 2004. Purification and characterateriztion of an
antioxidative peptide from enzymatic hydrolysate of yellowfin sole (Limanda
aspera) frame protein. Eur Food Res Technol. 219: 20-6.
The 12
th
16 -18 June, 2011

366

Klompong V, Benjakul S, Kantachote, D, Shahidi F. 2007a. Antioxidative activity and
functional properties of protein hydrolysate of yellow stripe trevally (Selaroides
leptolepis) as influenced by the degree of hydrolysis and enzyme type. Food Chem.
102:1317-27.
Klompong V, Benjakul S, Kantachot D, Hayes, KD, Shahidi F. 2007b. Comparative study
on antioxidative activity of yellow stripe trevally protein hydrolysate produced
from Alcalase and Flavourzyme. Int J Food Sci Technol. 43: 1019-26.
Kristinsson HG, Rasco BA. 2000. Fish protein hydrolysates: production, biochemical, and
functional properties. Crit Rev Food Sci Nutr. 40: 43-81.
Wu HC, Chen HM, Shiau CY. 2003. Free amino acids and peptides as related to
antioxidant properties in protein hydrolysates of mackerel (Scomber austriasicus).
Food Res Inter. 36: 949-57.

The 12
th
16 -18 June, 2011

367

OJ-231

New Generation of Innovative Healthy Foods to Address Some of the
Major World Health Challenges

Vijay Jayasena
1,2,*
,

and Syed M. Nasar-Abbas
1,2

1
Food Science and Technology, School of Public Health, Curtin Health Innovation Research Institute,
Curtin University, Perth, Australia;
2
Centre for Food and Genomic Medicine, Western Australian Institute for Medical Research, Perth, Australia
*
Corresponding Author: V.Jayasena@curtin.edu.au

Abstract

Malnutrition and obesity may be two opposite but major food related problems in the
world. There are about 900 million undernourished people in the world. Asia, which
accounts for about 60% of the total, is the most affected region and the major cause is the
lack of balanced protein in the diet. Protein-rich foods are less affordable to a major
segment of the population in the region. Obesity, type-2 diabetes and cardiovascular
diseases (CVD) pose a major health challenge in the world. Nutritionally poor foods e.g.
low in dietary fibre, high in oil, high in starch and low in protein plays a key role in these
health problems. A range of commonly consumed foods including bread, pasta, instant
noodles, cookies and chips were developed by incorporating 10-50% lupin flour. Lupin
flour is high in protein (40%) and dietary fibre (28%) and also a good source of micro
nutrients such as Ca, P, K, Fe and Zn. Addition of lupin flour substantially increased the
nutritional value of foods. Incorporation of 20% lupin in pasta, instant noodles and cookies
resulted in up to 55% increase in protein and up to 160% increase in dietary fibre contents
without deteriorating the consumer acceptability. The crisps prepared by using lupin flour
had similar attractiveness to potato and other type of chips but contained 300% more
protein, six times more dietary fibre, six times less starch and half of the fat than traditional
potato chips.

Keywords: Malnutrition, obesity, lupin, protein, healthy foods

Introduction
Malnutrition is one of the biggest health challenges in the world especially in the Asian
and African countries. According to the FAO (2010), about 28% population in Africa, 22%
in South Asia and 14% in Southeast Asia (Thailand 16%, Indonesia 13%) is suffering from
malnutrition. Children are the most victims of malnutrition as 46% of the children in the
South Asia are undernourished (India 47%, Pakistan 40%, Bangladesh 48%) which is
almost double than that of sub-Saharan Africa (24%). Malnutrition could lead to coronary
heart disease, stroke, respiratory infections, diarrhoeal diseases and prematurity and low
birth weight (Black 2003; Gariballa 2000; King and others 2008; Victora and others 2008).
The resistance and immunity of human body are weakened by undernourishment and
hunger. Malnutrition also affects educational attainment that ultimately impacts on
productivity and economic growth. The high price of protein rich foods, which is
unaffordable by some of the population in these countries, is one of the major reasons for
the malnutrition. Development of low cost high protein foods is paramount in addressing
these issues.
The 12
th
16 -18 June, 2011

368

Obesity, diabetes and cardiovascular diseases are also some of the biggest health problems
in the world. According to World Health Organisation (WHO 2009) obesity has reached
epidemic proportions globally with more than 1 billion adults overweight and at least 300
million of them are clinically obese. Obesity and overweight pose a major risk for chronic
diseases, including Type-2 diabetes, cardiovascular disease (CVD), hypertension, stroke
and cancer. In addition to the lack of physical exercises, food (low in dietary fiber, high in
oil, high in starch and low in protein) plays a key role in the current health problems. Most
of the population in developed countries consumes less than the Recommended Dietary
Intake (RDI) of dietary fiber which is one of the major causes of common health problems.
Although the problems are quite opposite in developed and developing countries, a
common solution could be found in lupin-based foods. Lupin (Lupinus angustifolius) is a
grain legume similar to soybean that can be grown in marginal agricultural conditions
(Jayasena and Quail 2004). Lupin flour has high nutritional value. It contains about 40%
protein, 28% dietary fibre and 6% fat contents and has low glycaemic index (GI) due to
little or no starch content. Lupin flour, due to its low cost, GM free status, high protein,
low fat, high dietary fibre, low carbohydrate and low GI, has a great potential in
incorporation into food formulations. High dietary fibre content of lupin flour is one of the
most important factors in terms of developing healthy diets to reduce the incidences of
cardiovascular diseases, diabetes and obesity. An increased consumption of dietary fibre in
daily diet has been recommended by nutritionists to improve health.

Independent studies in Europe and Australia have revealed many health benefits of lupin
based foods including lowering cholesterol and blood glucose levels (Feldheim 1998; Hall
and others 2005; Petterson 1998). Addition of lupin in foods reduce GI that results in
lowered insulinaemic index which is thought to be benefit in the prevention of type 2
diabetes and cardiovascular diseases (Ludwig 2002). Lupin containing foods can help
reduce obesity by increasing satiety and reducing energy intake (Archer and others 2004;
Lee and others 2006).

Keeping in mind the valuable properties of lupin, a series of experiments were conducted
to investigate the incorporation of lupin flour in pasta, cookies, muffins and crisps to
improve their nutritional quality without deteriorating their physical and sensory
properties.

Commercially available Australian sweet lupin flour was utilised in preparing pasta,
cookies, muffins and crisps (a product alternative to potato and/or corn chips). Lupin flour
was incorporated at 20% level in pasta, cookies and muffins and at 50% level in crisps.
The food samples were prepared using standard laboratory based methods. Control for
pasta, cookies and muffins were formulated with no added lupin flour whereas commercial
potato chips were considered as the control for lupin crisps. The foods were assessed for
nutritional quality, physical properties of colour and texture and sensory acceptability.

Moisture, protein and dietary fibre contents
AOAC International (2000) methods were used to determine moisture (method 925.10)
and protein (method 950.36) contents. Dietary fibre contents were calculated on the basis
of dietary fibre contents of raw ingredients.
The 12
th
16 -18 June, 2011

369

Instrumental colour measurement
Colour was measured using Minolta spectrophotometer CM-508i (Minolta Co. Ltd. Japan)
and expressed as L
*
(lightness), a
*
(+ a
*
= redness, a
*
= greenness) and b
*
(+ b
*
=
yellowness, b
*
= blueness) colour coordinates according to the methods specified by the
equipment manual. The instrument was equipped with a pulsed xenon arc lamp as light
source, a silicon photodiode array detector and has the illumination/measurement area of
11mm.

Textural Analysis
Texture analyses of the food samples were carried out using a TA.XT2i Texture Analyser
(Stable Micro System Ltd, United Kingdom) by applying methods given in AACC (2000).

Sensory evaluation
The sensory evaluation was carried out using standard methods. Panellists were asked to
express their liking/disliking using the 9-point hedonic scale with 1 = dislike extremely, 2
= dislike very much, 3 = dislike moderately, 4 = dislike slightly, 5 = neither like nor
dislike, 6 = like slightly, 7 = like moderately, 8 = like very much and 9 = like extremely.
The samples were evaluated for colour, taste/flavour, texture and overall acceptability. The
number of panellist varied from 10 trained panellists to 50 untrained panellists depending
upon their availability at the time of the product development.

Statistical analysis
Data were analysed for one-way analysis of variance (ANOVA) using SPSS 17.0. Since
sensory evaluation data was non-parametric, Kruskil Wallis test was used for data analysis.
Statistical significance was established at p 0.05.

Improvement in the nutritional value
Incorporation of lupin flour substantially increased the nutritional value of the foods.
Compared with the control samples, incorporation of lupin flour at 20% level in pasta and
cookies resulted in an increase of 54% and 85% in protein (Fig 1) and 160% and 210% in
dietary fibre contents (Fig. 2), respectively.

Figure1 Changes in protein contents as a result of
incorporating 50% lupin flour in crisps
and 20% in pasta and cookies.

Figure 2 Changes in dietary fibre contents as a
result of incorporating 50% lupin flour
in crisps and 20% in pasta and cookies.
The 12
th
16 -18 June, 2011

370

Similarly, crisps prepared by incorporating 50% lupin flour in the formulation contained 3
times more protein and 6 times more dietary fibre as compared to commercial potato chips
(Figs. 1 & 2). Lupin based crisps also absorbed less oil (almost half of that of potato chips) and
contained significantly higher amounts of micro nutrients (Table 1). Lupin crisps 100g provide
the recommended dietary intake of 55-65% of dietary fibre, 35-50% of protein and around
30% of micronutrients such as potassium, magnesium, phosphorus and iron (Table 1).

Table 1. Nutritional comparison of lupin based crisps with commercial potato chips

Lupin crisps RDI available out of 100g
lupin crisps (%)

Nutrients

Potato
chips
contents

Contents % Difference from
potato chips
Men
(age 31-50)
Women
(age 31-50)

Energy (kJ)

2200

1300

-40

Protein (g) 8 23 +185 35 50
Fat (g) 32 16 -50
Dietary Fibre (g) 3 17 +460 55 65
Potassium (mg) 1200 690 -56 18 24
Magnesium (mg) 50 104 +108 27 35
Calcium (mg) 25 55 +120 5 5
Phosphorus (mg) 130 204 +57 21 21
Iron (mg) 1.0 2.5 +150 34 15
Zinc (mg) 0.35 1.7 +485 12 21

The results are in agreement to those of other studies in which protein, dietary fibre and
mineral contents increased substantially by lupin flour incorporation into spaghetti and
other pasta products (Martinez-Villaluenga and others 2007; Rayas-Duarte and others
1996; Torres and others 2007). Bread prepared by adding defatted lupin flour at 6% level
significantly increased the fibre content (Mubarak 2001). Incorporation of lupin flour in
pasta at 15% level not only increased the total protein contents but also the biological value
of protein (Lampart Szczapa and others 1997).

High protein foods are required in many developing countries where a considerable
proportion of the population lives under protein malnutrition. Millions of children are
suffering from severe chronic malnutrition which has been linked to high morbidity and
mortality (UNICEF 2006). Lupin incorporated convenient foods such as pasta, cookies and
crisps that contain high amount of protein would provide an effective way of improving the
nutrition status of the population without increasing the cost.

Obesity, which has reached epidemic proportions globally and considered as a major
contributor to the global burden of chronic disease and disability (WHO 2008), can be
dealt with by using lupin incorporated foods. Due to the low GI value, high dietary fibre
and high protein contents, lupin added bread helped reduce obesity by increasing satiety
and reducing energy intake (Lee & others 2006). Similarly, lupin added pasta and cookies
can help reduce obesity and will provide more choice for health conscious individuals.

The 12
th
16 -18 June, 2011

371

Effect of lupin incorporation on the colour
Lupin flour is naturally yellower than wheat flour owing to its high carotenoids contents
(Mohamed and Rayas-Duarte 1995; Wang and others 2008). Replacement of wheat flour
with lupin flour in the formulation would affect the colour of the product. Colour analysis
using Minolta chromameter showed that in case of dry pasta the colour coordinates L
*
, a
*

and b
*
representing lightness, redness and yellowness, respectively, increased with the 20%
lupin flour incorporation. However after cooking lightness and yellowness of the pasta
containing 20% lupin flour showed no significant differences compared to the control
(Table 2). It was due to the colour leaching of lupin pasta in the boiling water. Similar
results were found in lupin fortified instant noodles where yellow colour imparted by lupin
flour addition was significantly reduced after the cooking process (Jayasena and others
2010b).

In case of cookies there were no significant colour changes between the control and
samples containing 20% lupin flour. It may be due to browning caused by baking process
that masked any slight colour changes caused by lupin addition. Muffins demonstrated an
increase in greenness and yellowness with the lupin flour incorporation.

Table 2 Effect of 20% lupin addition on the colour

Minolta colour values Sample Lupin incorporation
L* a* b*

0% (Control)

56.72 b

1.52 b

34.45 b

Dry Pasta
20% 58.85 a 3.76 a 36.84 a

0% (Control)

77.42 a

-1.44 b

25.50 a

Cooked Pasta
20% 76.35 a 0.80 a 25.87 a

0% (Control)

66.01a

8.53a

40.09a

Cookies
20% 65.53a 8.75a 40.73a
0% (Control) 74.67a -1.28b 21.28b Muffins
20% 72.90a 3.16a 35.58a
Means in a column (within a product) with same letters are not significantly different (P0.05)

Effect of lupin flour incorporation on the textural properties
Textural properties of pasta are very important in terms of its quality. The breaking
strength of the dry pasta is important in terms of packaging, transport and storage. The
breaking strength decreased with lupin flour incorporation at 20% level. (Table 3). As
gluten strength and quality determine the strength of dry pasta (Smewing 1997) the
reduction in breaking strength may be due to decrease in gluten concentration with the
increase in lupin flour. Lupin flour is deficient in gluten. The strength of lupin
incorporated pasta may be increased by adding gluten or other additives such as food
gums. Hydrocolloids or non-starch polysaccharides (gums) can be used to mimic the
properties of gluten and result in firmness of structure (Lazaridou and others 2007).
Development of a network structure with addition of gums contribute to overall pasta
firmness (Edwards and others 1995).

The 12
th
16 -18 June, 2011

372

Table 3 Effect of lupin flour incorporation on physical properties of pasta

Lupin flour
incorporation

Breaking strength
(dry, g)
Stickiness
(cooked, g)
Firmness
(cooked, g)
Cooking
time (min)
Cooking
loss (%)
0% (control) 94790a 50456a 3194a 12.50.5 a 6.60.3 a
20% 75251b 43158a 31415a 12.00.5 a 7.10.6 a
Results are expressed as means SD, n=6
Means with different superscripts within a column are significantly different (p0.05)

Textural characteristics of cooked pasta are prime concern with firmness and stickiness
playing major roles in the acceptability by the consumers. A sticky pasta is generally
unacceptable. The data presented in Table 3 on the firmness and stickiness of cooked pasta
reveal that the samples prepared by incorporating 20% lupin flour had no significant
difference (p0.05) to the control sample in terms of firmness (F
max
) and stickiness. There
were also no significant changes in terms of cooking time and coking losses between
control sample and the sample containing 20% lupin flour.

Effect of lupin incorporation on the sensory properties
The sensory evaluation results (Table 4) reveal that 20 % lupin addition has no significant
effect on the sensory score for colour (consumer likeness of colour) of pasta and cookies.

Table 4 Effect of lupin incorporation on the sensory properties

Samples Lupin incorporation Colour Taste Texture Overall acceptability
0% (Control) 7.0 a 7.4 a 7.2 a 7.1 a
Pasta
20% 6.5a 6.2 a 6.5 a 6.7 a
0% (Control) 7.3a 7.5a 7.2a 7.2a
Cookies
20% 6.9a 7.0a 6.8a 7.0a
0% (Control) 5.4b 6.7a 6.2a 6.3a
Muffins 20% 6.8a 6.5a 6.2a 6.4a
Means in a column (within a product) with same letters are not significantly different (P0.05)

Muffin colour, however, was improved due to natural yellow colour imparted by lupin
flour. A yellowish colour of muffins was preferred by the judges over the control sample.
Studies by other researchers revealed similar effects when legume flours were added in the
pasta. Pasta samples supplemented with 10-30% cowpea flour demonstrated a substantial
improvement in colour as compared to samples prepared from soft wheat flour (Bergman
and others 1994).

The samples prepared by adding 20% lupin flour in pasta and cookies were equally liked
by the panellists for taste, texture and overall acceptability as compared with the control
samples (Table 4). Our results are supported by those of previous studies that have shown
that incorporation of lupin at certain levels in other food products did not reduce their
acceptability. Addition of lupin to replace up to 10% bakers white wheat flour in bread did
not affect appearance, flavour, texture and general acceptability (Hall & others 2005).
While addition of lupin flour at 20% level had no significant effect on overall acceptability
of instant noodles, the colour liking improved significantly (Jayasena & others 2010b).
The 12
th
16 -18 June, 2011

373

Lupin can be incorporated up to 40% in tofu (Jayasena and others 2010a) and up to 60%
in tempe (Jayasena and others 2007) without affecting its overall acceptability.

References
AACC 2000. Approved Methods of the American Association of Cereal Chemists.
American Association of Cereal Chemists. Minnesota.
AOAC 2000. Official Methods of Analysis of AOAC International. AOAC International.
Gaitherburg.
Archer, BJ, Johnson, SK, Devereux, HM Baxter, AL 2004. Effect of fat replacement by
inulin or lupin-kernel fibre on sausage patty acceptability, post-meal perceptions of
satiety and food intake in men. Brit. J. Nutr. 91(4): 591-599.
Bergman, CJ, Gualberto, DG Weber, CW 1994. Development of a high-temperature-dried
soft wheat pasta supplemented with cowpea (Vigna unguiculata L. Walp). Cooking
quality, color, and sensory evaluation. Cereal Chem. 71: 523-527.
Black, RE 2003. Zinc deficiency, infectious disease and mortality in the developing world.
J. Nutr. 133: 1485S-1489S.
Edwards, NM, Biliaderis, CG Dexter, JE 1995. Textural characteristics of wholewheat
pasta and pasta containing non-starch polysaccharides. J. Food Sci. 60(6): 1321-
1324.
FAO 2010. The State of Food Insecurity in the World. Food and Agriculture Organization
of the United Nations. Rome.
Feldheim, W 1998. Sweet lupin flour-a healthy asset. Int. J. Food Ingredients. 5: 25-26.
Gariballa, SE 2000. Nutritional factors in stroke. Brit. J. Nutr. 84: 5-17.
Hall, RS, Thomas, SJ Johnson, SK 2005. Australian sweet lupin flour addition reduces the
glycaemic index of a white bread breakfast without affecting palatability in healthy
human volunteers. Asia Pacific J. Clin. Nutr. 14(1): 91-97.
Jayasena, V, Kardono, L, Quail, K Coorey, R 2007. Manufacturing of lupin tempe. In
Parker, W. Agribusiness Crop Updates pp. 61-62). Department of Agriculture and
Food, Western Australia. Perth.
Jayasena, V, Khu, WS Nasar-Abbas, SM 2010a. The development and sensory
acceptability of lupin-based tofu. J. Food Qual. 33(1): 85-97.
Jayasena, V, Leung, PPY Nasar-Abbas, SM 2010b. Effect of lupin flour substitution on
the quality and sensory acceptability of instant noodles. J. Food Qual. 33: 709727.
Jayasena, V Quail, K 2004. Lupin: A legume with a future. Food Beverage Asia. 16-21.
King, DA, Cordova, F Scharf, SM 2008. Nutritional aspects of chronic obstructive
pulmonary disease. Proc. Am. Thoracic Society: 5, 519-523.
Lampart Szczapa, E, Obuchowski, W, Czaczyk, K, Pastuszewska, B Buraczewska, L
1997. Effect of lupine flour on the quality and oligosaccharides of pasta and crisps.
Nahrung/Food 41(4): 219-223.
Lazaridou, A, Duta, D, Papageorgiou, M, Belc, N Biliaderis, CG 2007. Effects of
hydrocolloids on dough rheology and bread quality parameters in gluten free
formulations. J. Food Eng. 79: 10331047.
Lee, YP, Mori, TA, Sipsas, S, Barden, A, Puddey, IB, Burke, V, Hall, RS Hodgson, JM
2006. Lupin-enriched bread increases satiety and reduces energy intake acutely
[Erratum: 2007 Apr., v. 85, no. 4, p. 1166.]. Am. J. Clin. Nutr. 84(5): 975-980.
Ludwig, DS 2002. The glycemic index: physiological mechanism relating to obesity,
diabetes and cardiovascular disease J. Am. Med. Assoc. 287: 2414-2423.
The 12
th
16 -18 June, 2011

374

Martinez-Villaluenga, C, Urbano, G, Porres, JM, Frias, J Vidal-Valverde, C 2007.
Improvement in food intake and nutritive utilization of protein from Lupinus albus
var. multolupa protein isolates supplemented with ascorbic acid. Food Chem.
103(3): 944-951.
Mohamed, AA Rayas-Duarte 1995. Composition of Lupinus albus. Cereal Chem. 72(6):
643-647.
Mubarak, AE 2001. Chemical, nutritional and sensory properties of bread supplemented
with lupin seed (Lupinus albus) products. Nahrung/Food. 45(4): 241-245.
Petterson, DS 1998. Composition and food uses of lupins. In Gladstones, JS, Atkins, CA
Hamblin, J. Lupins as crop plants: biology, production and utilization pp. 353-384).
Cab International. Wallingford.
Rayas-Duarte, P, Mock, CM Satterlee, LD 1996. Quality of spaghetti containing
buckwheat, amaranth, and lupin flours. Cereal Chem. 73(3): 381-387.
Smewing, J 1997. Analysing the texture of pasta for quality control. Cereal Foods World
42(1): 8-12.
Torres, A, Frias, J, Granito, M, Guerra, M Vidal-Valverde, C 2007. Chemical, biological
and sensory evaluation of pasta products supplemented with -galactoside-free
lupin flours. J. Sci. Food Agric. 87(1): 74-81.
UNICEF 2006. Children's issues: Nutrition. Avaialble from:
http://www.unicef.org/india/children_2356.htm. Accessed Mar 10, 2009.
Victora, CG, Adair, L, Fall, C, Hallal, PC, Martorell, R Richter, L 2008. Maternal and
child undernutrition: consequences for adult health and human capital. The Lancet.
371: 340-357.
Wang, SF, Errington, S, Sweetingham, M, Yap, HH Smith, L 2008. Studies on
carotenoids from lupin seeds. In Palata, JA Berger, JD. Proceedings of 12th
International Lupin Conference pp. 198-202). International Lupin Association.
Fremantle, Western Australia.
WHO 2008. Obesity and overweight. Avaialble from: http://www.who.int/dietphysicalactivity/
publications/facts/obesity/en/. Accessed Feb 16, 2009.

The 12
th
16 -18 June, 2011

375

OJ-255

Stability of Ascorbic Acid during Thermal and High Pressure Processing

Minh Thuy Nguyen
1,*
, Indrawati Oey
2
, Marc Hendrickx
3

1
Can Tho University, Viet Nam;
2
University of Otago, New Zealand;
3
Katholieke Universiteit Leuven, Belgium
*
Corresponding author: nmthuy@ctu.edu.vn

Abstract

The stability of ascorbic acid (AA) has been investigated under varying conditions, such as
temperatures (80C to 100C), pressures (100 to 750 MPa), pH (5 and 7) and in presence
of oxygen (0.25 mM). A biphasic model was selected to describe the time-dependent
changes due to thermal treatment of AA. Oxygen was found to be very important in
affecting AA stability. Changes in temperatures from 80 to 100C accelerate AA
degradation, pH had different effect on the aerobic and anaerobic thermal degradation of
AA and seemed not clearly effect on the degradation during high pressure thermal
treatments. From the combined pressure-temperature experiments performed for longer
time, AA was stable at pressures up to 700 MPa and temperatures up to 70C after pressure
build up and no further degradation of AA was observed up to 100 min of treatment.

Keywords: Ascorbic acid, stability, temperature, high pressure, oxygen

Introduction
Ascorbic acid is known to be thermolabile. It could be degraded during food processing
even though it has the profitability to preserve foods by the strength of its reducing
property. In order to evaluate the process impact on vitamin C stability, the degradation of
AA in model system under isothermal and isobaric-isothermal conditions was studied on a
kinetic basis. Moreover, the influence of pH and molar ratio between oxygen and AA
concentration on AA stability was evaluated.

Temperature and pressure stability of ascorbic acid (AA) in different model systems (i.e.
differences in pHs, oxygen concentrations) was studied on a kinetic basis. The samples
were subjected to isothermal-isobaric treatments for predefined time intervals in thermostat
small sample holders or in multivessel-HP equipment. The ascorbic acid concentrations
were measured using RP-HPLC method. In the presence of oxygen, it assumes that AA
degrades under aerobic and anaerobic conditions and each degradation follows first-order
kinetics (Eq. 1). The temperature dependence of the degradation rate constant (k) is
estimated as activation energy (E
a
) using the Arrhenius equation (Eq. 2).

C
t
=C
a
exp(-k
a
t)+C
an
exp(-k
an
t) (1)
(
(
|
|
\
|
+ =
T T R
E
k k
ref T
a
ref
1 1
) ln( ) ln(
(2)
C
a
: the proportion of the aerobic degradation (% or mM); C
an
the proportion of the
anaerobic degradation (% or mM); k
a
the aerobic degradation rate constant and k
an
the
anaerobic degradation rate constant.
The 12
th
16 -18 June, 2011

376

Results
Thermal stability of ascorbic acid: the effect of molar ratio between oxygen and AA on
the vitamin stability in acetate buffer (0.2 M, pH 5.0) and phosphate buffer (0.1 M, pH 7.0)
during thermal treatment was studied. The initial concentration of soluble oxygen in the
sample was around 0.25 mM. The decrease in AA concentration as a function of time
occurred in two phases, e.g. at 80C (Figure 1). At low AA concentration, the most
degradation of AA occurred under aerobic condition, e.g. AA (50 g/ml0.28 mM)
completely consumed the initial oxygen concentration (0.25 mM). At high AA
concentration, AA was further degraded under anaerobic condition. Similar degradation
behaviors were observed at other degradation temperatures and pHs.

Figure 1 Effect of AA concentrations 50 g/ml ( ); 100 g/ml (o); 250 g/ml ( ); 500 g/ml ( ) on AA
degradation in phosphate buffer (0.1 M, pH 7.0) at 80C

Thermal stability of AA in all systems was investigated at various constant temperatures
from 80 to 100C. As expected, increasing temperature accelerates the degradation (Table
1), the degradation accelerates by increasing temperatures (above 80C), but temperature
does not influence the proportion of AA degradation under aerobic (C
a
) and anaerobic
conditions (C
an
). When the oxygen is completely consumed, anaerobic AA degradation
occurs and this anaerobic degradation occurs more slowly than the aerobic AA
degradation. Using the estimated kinetic parameters, the biphasic model (Eq. 1) was used
to predict the residual AA concentrations.

Table 1 Effect of temperature on AA (100 g/ml) stability in phosphate buffer (0.1 M, pH 7.0
a
)
T (C) k
a
(min
-1
) k
an
(x 10
-3
min
-1
)

C
a
(%) C
a
(mM)
b
C
an
(%) C
an
(mM)
80
90
100
0.200.08
c

0.410.11
0.450.12
3.693.87
12.904.22
24.205.49
36.297.23
47.535.97
50.336.18
0.250.05
0.310.04
0.350.04
64.177.38
52.895.26
51.345.58
0.450.05
0.340.03
0.360.04
E
a

(kJ.mol
-1
)
44.418.8
r
2
= 0.85
103.218.0
d

r
2
= 0.97

/

/

/

/
a
The initial oxygen concentration was 8 ppm (0.25 mM),
b
Calculated using MW of AA of 176.13 Da
c
Asymptotic standard error of non linear regression,
d
Standard error of linear regression.

Effect of pH
The kinetics of the oxidative breakdown of AA in liquid are complex. The obtained results
from our studies show that at low AA concentration (100 g/ml0.57 mM), the thermal
degradation of AA occurred mainly under aerobic condition due to the oxidation reaction.
In this case, AA aerobic degradation at pH 7 was higher than at pH 5 and afterwards, it was
slow down or almost stabilized during anaerobic condition (Figure 2). From the
thermodynamic point of view, the reaction is favored in alkaline media because of the
decrease in ascorbic acid redox potential (Manuel de Villena and others 1989) and the
oxidation process is impeded in acidic environment and at low temperature.
The 12
th
16 -18 June, 2011

377

Figure 2 Effect of pH on temperature stability of AA (100 g/ml) at 100C in acetate buffer (0.2 M,
pH 5.0) ( ) and phosphate buffer (0.1 M, pH 7.0) (o)

Pressure stability of ascorbic acid: At higher pressure levels (above 100 MPa), no further
AA degradation was found during a 15 min of treatment. The similar finding was also
found for different temperatures (40 and 70C). At constant pressure, increasing
temperature from 40 to 70C enhances the AA degradation (Figure 3).
0
1
2
3
0 100 200 300 400 500 600 700
Pressure (MPa)
R
e
s
i
d
u
a
l

A
A

c
o
n
c
.

(
m
M
)

Figure 3 Effect of temperature 40C ( ) and 70C ( ) on pressure stability (15 min) of AA (500 g/ml)
in acetate buffer (0.2 M, pH 5.0) in presence of initial oxygen concentration of 0.25 mM

In all cases, AA showed very stable during combined pressure (up to 700 MPa) and
temperature (up to 70C) treatments. No further degradation of AA was observed within
100 min of pressure treatment time. By increasing temperature (up to 80C) in
combination with pressures from 600 to 750 MPa, the AA degradation was found.

Effect of pH
The results show that pH seems to give no significant effect on the pressure stability of
AA.

Conclusion
Degradation of AA in aqueous solution due to heat treatment was observed and a biphasic
model is adequate to describe the time-dependent changes due to thermal treatment. High
pressure has no/little effects on covalent bonds resulting in limited loss e.g. in vitamin
content. However, AA was degraded at temperatures above 70C combined with pressures
above 700 MPa. The effect of pH on the pressure stability of vitamin C was not clearly
observed probably due to pH instability of the buffer solutions.

References
Manuel De Villena FJ, Martin AA, Polo Dez LM, Prez Prez R. 1989. Kinetic
determination of ascorbic acid in fruit juices and pharmaceutical preparations.

Full Oral

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Full Oral

Uploaded by

Copyright:

Available Formats

The 12

C. Based on the curve, the D values, i.e. time

C. Meanwhile, the control ATCC isolate had D values of

g ,were subjected to identification from

Corresponding author: fagipas@ku.ac.th

You might also like