System. App!. Microbio!

' 21, 599-608 (1998)

__G_us_ta_vF_is_ch_er_V_er_lag_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

SYSTEI'v1ATIC AND
APPLIED MICROBIOLOGY

Characterization, Differentiation and Identification of Wild-type

Cellulose-synthesizing Acetobacter strains Involved
in Nata de Coco Production
E. B. BERNARDO, B. A. NEILAN, and I. COUPERWHITE 1
School of Microbiology and Immunology, University of New South Wales, Sydney, Australia
Received September 19, 1998

Summary
Cellulose-producing bacteria used for Nata de Coco production in the Philippines were isolated and
characterized. Two distinct wild-type strains (ITDI 2.1 and PA 2.2) were selected from thirty-eight isolates obtained. The two strains, and the rest of the cellulose producers, were established to be members
of the genus Acetobacter based on rapid phenotypic characterization. They were characterized and differentiated based on pellicle type and colony morphology, carbon-utilization pattern (Biolog assay),
amount of cellulose production and 16S rDNA sequence. The thick pellicle cellulose-producing strain
(ITDI 2.1) has a stable and consistent production of a very thick pellicle in any sugar-based, acidic medium, with a dry weight up to twenty-six times that of the thin pellicle cellulose-producing strain (PA 2.2)
after 7-8 days static batch culture. Species identification using 16S rDNA sequencing revealed that the
two strains belong to two different species of Acetobacter, Acetobacter xylinus and A. hansenii and may
be new subspecies under these species designation.
Key words: Acetobacter - cellulose - phenotype - 16s rRNA

Introduction
The genus Acetobacter is composed of Gram-negative,
rod-shaped, strictly aerobic, acidophilic bacteria able to
oxidize ethanol to acetic acid, and acetate and lactate to
carbon dioxide and water (DE LEY et al., 1984). Certain
members of the genus Acetobacter synthesize highly pure
extracellular polysaccharide called cellulose. To date,
there are three recognized species of Acetobacter, namely: A. xylinus (EUZEBY, 1997; formerly A. xylinum), A.
pasteurianus and A. hansenii (HOLT et al., 1994), that are
known to synthesize this polysaccharide (YAMADA, 1983;
GOSSELE et al., 1983; DE LEY et al., 1984). Cellulose from
Acetobacter resembles the crystalline structure and microfibrillar width of cellulose from plants and algae
(HOTCHKISS and BROWN, 1989; Ross et al., 1991). It was
characterized as highly crystalline, chemically pure,
metabolically inert, physically robust, highly absorbent,
and with less contaminating sugar and protein polymers
(HESTRIN and SCHRAMM, 1954; WHITE and BROWN,
1989; YAMANAKA et al., 1989; CANNON and ANDERSON,
1 To whom correspondence should be addressed. E-mail: l.Couperwhite@unsw.edu.au

1991; Ross et al., 1991). These properties distingiush it

from cellulose synthesized by other organisms. The extraordinary purity could potentially eliminate extensive
processing such as mercerization, pulping or delignification necessary in the production of paper products from
conventional sources such as timber (CANNON and ANDERSON, 1991; Ross et al., 1991).
In the Philippines, wild-type strains of Acetobacter
also produce this cellulose polymer known locally as
"nata". If the organism is grown in sugar-enriched coconut milk or water based medium, the product is called
"nata de coco". Nata de coco is one of the first reported
commercial food applications of bacterial cellulose
(LAPUZ et al., 1967; BROWN, 1989). Nata production in
the Philippines is a labor-intensive cottage industry, employing a non-aseptic, static batch culture system. In
spite of these limitations, it has attained commercial success and flourished into an important commodity export
earner in the Philippines. In the export market specifically to Japan, demand for nata de coco has surged in the
past few years, generating over 50 million US dollars export earnings for the Philippines from 1993 to 1997

600

E. B. BERNARDO et al.

(BETP, 1997). This demand was attributed primarily to

nata de coco's popularity and the probiotic effect or positive health benefits derived from the regular consumption of this fibrous low calorie food. The increase in consumer demand has also led to the rapid proliferation of
small and medium scale nata producers in many
provinces in the Philippines. Consequently, this has resulted in the escalation of production-related problems
most importantly on: a) poor control of the quality or
type of cellulosic nata needed in the export market; and
b) unstable rates of cellulose synthesis by the micro organism(s). These problems are attributed to a single or
combination of the following factors: seemingly heterogeneous microbial population in the starter culture, nonaseptic production process, and/or poor control of culture conditions for the growth of the microorganism(s).
Because of these different factors, it is not certain
whether only one strain of Acetobacter or a consortium
of microorganisms is present in the starter culture per
production batch. It is also unresolved whether a particular strain or group of strains is responsible for synthesis
of a particular type of nata, or if there are specific strains
responsible for production of a thick, tough nata pellicles, and other strains for thin, soft types. It is not known
whether rate of cellulose synthesis, by wild-type nata
producers, is strain dependent, or whether other genetic
factors dictate the stability of these microorganisms to
synthesize a particular type of nata pellicle, or at which
rate.
This study was initiated to address some of the issues
and problems previously mentioned. Cellulose-synthesizing isolates were characterized, differentiated and
grouped. Two phenotypically distinct strains were selected for further comparison and identification to reveal
strain variations that may dictate the quality or type, and
the rate of synthesis of the cellulosic nata pellicle.

Materials and Methods

Isolation and selection of wild-type cellulose-synthesizing organisms: Wild-type cellulose-synthesizing microorganisms were
collected from the Philippines. The list and sources of the isolates are given in Table 1. Isolation of pure cultures from samples of spent liquor or starter cultures was done as follows.
From each representative sample, 1 ml was inoculated in triplicate into 15 ml Coconut Water Medium (CWM) containing
100 g sucrose, and 3.0 g K 2HP0 4 , per liter coconut water. Alternatively, Tomato Peptone Sucrose Salts Solution (TPSS) (GALLARDO-DE JESUS et aI., 1973) was used containing per liter:
500 ml of tomato decoction from 250 g fresh tomato, 5.0 g
peptone, 100 g sucrose, 3.0 g yeast extract, 0.2 g MgS04 .
7H zO, 1.0 g K 2HP0 4 , 0.1 g NaCI (NSRI, 1990). Fresh tomato
could be successfully substituted by canned tomato juice without added preservatives (100 mil-I). In both media, 24 g 1-1 of
agar was added whenever solid media was required. Glacial
acetic acid was used to adjust the pH to 4-4.5. Culture media
were sterilized in an autoclave at 121C for 10 minutes. Inoculated tubes were incubated at 30 C, under non-agitated condition for 4-7 days, or until an obvious pellicle was observed
floating on the clear liquid medium. From the positive tubes, 1
loop of inoculum was streaked onto TPSS agar plates. At ran-

dom, 2-3 colonies of the same morphology were isolated from

each plate and subcultered, again in TPSS broth, to check for
pellicle formation. This was repeated several times until pure
culture isolates were obtained. Pure cultures were maintained
on TPSS agar slants at 4 C and were passaged every 2-3
months. Other strains from culture collections were further
checked for purity and production of nata pellicle following the
steps mentioned and using the floating pellicle for initial subculturing. For long-term storage, each isolate was placed in TPSS
broth with 15% glycerol and stored at -70C.
Colony and pellicle morphology: The types of pellicle were
described on the basis of a) thickness: very thick (>1.27 cm/
tube), moderately thick (S;1.27 cm/tube) or thin 0.635 cm/
tube), and b) toughness: soft, firm or tough, based on visual inspection and the previously described "feel method" (GALLARDo-DE JESUS et aI., 1973). The colony was described according
to its color, shape, surface, size, edge, elevation, opacity and
general appearance.
Phenotypic identification and inoculum preparation: The
generic identification of the wild-type nata strains was determined using the rapid phenotypic tests for the genus Acetobacter (SWINGS, 1992). Initial identification to the species level was
also done using a combination of two schemes (CARR and PASSMORE, 1979 and SWINGS, 1992). Three reference strains, namely
Acetobacter xylinus ATCC 23768 and ATCC 53582, and A.
pasteurianus ATCC 10245 were used for comparison at the
generic level. At the species level, the three cellulose-producing
type species Acetobacter xylinus, A. pasteurianus and A. hansen ii, (HOLT et aI., 1994) and one reference strain A. xylinus
ATCC 53582 (ATCC, 1992) were used for comparison. Isolates
were routinely checked for pellicle formation. For the preparation of inoculum, slant cultures were streaked on agar plates to
obtain isolated colonies. One representative colony from each
strain was subcultured in a broth medium to check for pellicle
formation. Tubes positive for pellicle formation were shaken
and a loopful of culture from each tube was inoculated into
15 ml broth with exactly the same components as TPSS medium
but without tomato juice. Henceforth, this medium will be referred to as Peptone yeast extract sucrose salts medium (PYSS).
Two to three passages were performed to ensure actively growing cells. A 150 ml PYSS medium was inoculated with 10% inoculum from the tubes. Incubation was at 30C under static
condition usually 2-3 days, or until enough pellicle was formed.
After incubation, the pellicles were aseptically cut into small
pieces before they were homogenized with the culture broth in a
Stomacher for about 2-3 minutes, filtered through six layers of
cheese cloth, and centrifuged at 10,000xg for 15 minutes at
4 0C. The collected cells were washed twice with 0.85% saline
solution, re-suspended in the same solution and used as inoculum for specific tests previously mentioned. If individual
colonies were needed for specific tests, a loopful from positive
tubes was re-streaked on plates.
Identification of cellulose polymer and determination of cellulose yield: Four tests were used to identify the cellulose synthesized by cells or colonies of each strain. These tests were: (a)
formation of typical cartilaginous (GALLARDO-DE JESUS et aI.,
1973) cellulose pellicle floating on a clear medium under nonagitated condition (CANNON and ANDERSON, 1991), (b) appearance of bright blue threads when stained with Lugol's iodine
containing 60% H 2S0 4 (WARD, 1954), (c) fluorescence of
colony on agar medium incorporated with 0.02% calcofluor
white ST (CANNON and ANDERSON, 1991), and (d) insolubility
of pellicles in 2% boiling NaOH for 0.5-1 hr. (CANNON and
ANDERSON, 1991; BROWN, 1996).
To determine the cellulose yield, cellulose pellicles were harvested, boiled in 2% NaOH, dried (MASAOKA et aI., 1993), allowed to cool, weighed until constant weight and reported as

Wildtype Cellulose-sy nthesizing Acetobacter strains

dry weight per liter of culture medium. Selected strains used for
this assay were grown in 75 ml PYSS medium. Inoculum preparation and incubation followed essentially the protocols used
earlier with the suspended cells adjusted to OD 600 0.5 using 5%
inoculum per flask.
Characterization of strains using the BIOLOG test assay:
This assay utilizes Biolog GN (Gram-negative) Microplates
consisting of 95 biochemical tests. It was used to determine the
ability of each strain to utilize or oxidize a pre-selected panel of
carbon sources. The tests gave a characteristic pattern of
purple wells that constituted a "Metabolic fingerprint" of the
capabilities of the inoculated microorganisms (BIOLOG, 1993).
Well-isolated colonies from each strain were streaked onto several Tryptic Soy Agar (DIFCO Michigan, USA) plates with an addition of 40 g 1-1 glucose. Plates were incubated for 48 hr., at
30C to obtain a lawn of cells. Detailed inoculum preparation
and inoculation followed the procedures in the BIOLOG Instruction Manual (BIOLOG, 1993). Reading of plates was done
after exactly 48 hr., at 30C without shaking. A dendrogram
was constructed based on the results. The analysis was performed using the "UPGMA" algorithm using the Biolog Inc. Microlog 3.6 version for Gram-negative organisms (BIOLOG, 1993).
Characterization and identification of strains using 16S
rDNA sequence: Total DNA was obtained by the method of
JHINGAN (1992) and modified accordingly by TII.LETT and
NEILAN (1998). The integrity and concentration of resuspended
total DNA was checked and analyzed using agarose gel electrophoresis or by measuring the absorbance (260/280 nm) ratio
with a Beckman UV spectrophotometer. DNA extracted by this
method was used as template for PCR reactions.
16S rRNA gene amplifications were performed using
primers 27FI(UFP) and 1494Rc(URP) together with PCR
reagents as previously described (NEILAN, 1997). Thermal cycling was performed at 94C for 4 minutes followed by 30 cycles of 94 DC, 20 seconds; 50C 30 seconds; and 72 C, 2 minutes. After thermal cycling, mineral oil was removed from DNA
amplification reactions by chloroform; isoamylalcohol (23:1)
extraction. The aqueous phase was then purified using the
Wizard PCR purification system (Promega, Madison, WI ) to remove amplification reaction components including unincorporated primers and nucleotides. Approximately 100 ng of PCR

601

product and 10 pmol of previously described 16S rRNA gene

sequencing primers (NEILAN, 1997) were used to determine the
primary structure of the 16S rDNA. Automated DNA sequencing was performed using the PRISM cycle sequencing system
and the ABI373 sequencer (ABI, Foster City, CAl. Oligonucleotide primers were synthesized on a Beckman Oligo 1000
DNA synthesis system (Beckman, Fullerton, CAl and purified
by reverse phase chromatography.
DNA sequences were aligned using the programs Pileup
(GCG, Madison, Wisconsin) and the multiple sequence alignment tool from Clustal W (THOMPSON et al., 1994). Manual
confirmation of the sequence alignment was performed and
checked against both primary and secondary structure considerations of the 16S rRNA molecule. The aligned sequences were
applied to genetic distance and maximum parsimony methods
for phylogenetic inference. Ambiguous characters, where a deletion, insertion, or unidentified state was recorded for any strain,
were removed from the aligned data. For all multiple sequence
alignments and phylogenetic inference programs the input order
of taxa was randomized. Genetic distances were calculated
using the formula of Jukes and Cantor, where D = -3/4 In
(1-4/3 d) and d is the sequence dissimilarity (JUKES and
CANTOR, 1969). Phylogenetic inference protocols, DNADIST,
NEIGHBOR, DNAPARS, CONSENSE, and SEQBOOT were
supplied by the PHYLIP package (version 3.57c) (FELSENSTEIN,
1989).

Results
Morphological characterization and identification to
the generic level
A total of 38 cellulose producing isolates were isolated, purified and characterized. Of the 38 characterized
isolates, two strains designated as lTDI 2.1 (thick cellulose pellicle-producing strain) and PA 2.2 (thin cellulose
pellicle-producing strain) were selected for more detailed
characterization and identification. Formation of "nata"
cellulosic pellicle was observed in all 38 isolates. It is the

Fig. 1. Two types of cellulose pellicle floating on a clear broth medium: a) very thick, tough pellicle produced by strain ITDI 2.1, and
b) thin, soft pellicle produced by strain PA 2.2. The cells were incubated for 7 days at 30C in static condition. Arrows pointing to
the pellicles.

602

E. B. BERNARDO et al.

most important criterion used in the strain isolation and

selection process. Two distinct pellicle types are shown
in Fig. 1. The 38 isolates were initially grouped based on
the type of pellicle produced and on the morphology of
colony (Table 2). In Figures 2 a&b, two colony types exhibited by strains lTD! 2.1 and PA 2.2 are shown. To
verify the taxonomic position of the wild-type cellulose
producing organisms, rapid tests were done. Based on
these tests, all the cellulose-producing isolates were identified under the genus Acetobacter (Table not shown).
They all conformed to nine phenotypic features proposed by Swings which included the following: Gramnegative reaction, rod shaped cell, pellicle formation,
non-production of spores, motile or nonmotile, oxidative respiration, growth at pH 4.5, oxidation of ethanol
(2%) to acetic acid in neutral or acidic pH, and oxidation of acetate and lactate to CO 2 and H 2 0 (SWINGS,
1992).

Cellulose synthesis by the nata-producing strains

The presumptive evidence for the identification of
nata pellicle as cellulose was the production a floating
pellicle in any medium as mentioned earlier. All levels of
thickness of cellulose pellicles were able to withstand
boiling in 2% NaOH for up to 1 hour. Distinct blue cellulose microfibrils were observed microscopically after
staining with Lugol's iodine. Another proof that cellulose
and not just other polysaccharide was truly synthesized
by the cells was shown by the fluorescence of colonies on
agar plates containing 0.02 % calcofluor white (data not
shown). Fluorescent brighteners such as calcofluor white,
react with the glucan chains of cellulose by hydrogen
bonding and cause the bands of cellulose to fluoresce
brightly when viewed under ultraviolet light (HAIGLER,
1982). They have been used to confirm cellulose production (CANNON and ANDERSON, 1991).

Table 1. List of wild-type nata pellicle producing bacteria.

Isolate code

Source

Place obtained from

ITDI 2.1

Derivative of pure culture

ITDI Culture collection, DOST, Taguig, Manila

UPCC3

Derivative of pure culture

NSRI culture collection, UP Diliman, Quezon City

IN 103, IN 102

Derivative of pure culture

NSTA Food Inc., Quezon City

LB 01, 02, 06, 08, 12b, 13,

14a, 14b, 14c, 15a

mother liquor/derivative of
pure culture

IFST, U. P. Los Banos, Laguna province

AB2.1, AB2.3, BAl.1, PA

2.2, PB7, BA2.3

Mother liquor

Laguna province - Dr. Ric del Rosario

BTG 1.1, 6.1, 8.1, 9.2, 9.3

Mother liquor/spent liquor

Batangas province - Nata growers

CB 1.1,2.1,3.1

Mother liquor

Cubao, Quezon City - Nata growers

NY 1.1, 2.1, 2.2, 3.1

Mother liquor

Novaliches, Quezon City

Nata grower and processor

CUS 2, LBpl, LBnl,

LBMl, LBm2.1, LBpl.1

Mother liquor

Los Banos, Laguna; Cubao, Quezon City

Table 2. Groupings of the 38 nata-producing isolates based on colony and pellicle types.
Pellicle type
Colony type

Isolate codes

lTD! 2.1

B
UPCC3, PB7, LB 14a, BA 2 .3, NY 1.1, NY 2.1

IN 103, CB 1.1, CB 2 .1, CB 3.1, NY 2.2, NY 3.1, BTG 9.2, LBpl

LB 14b, LB 15a, LB ml, LB m2.1, LB nl

LB 01, LB 02, LB 06, LB 08, LB 12b, LB 13, LB 14c, LB p2.1,

IN 102, BA 2.1, CUS 2, BTG 6.1, BTG 1.1, BTG 8.1, BTG 9.3

PA 2.2, AB 2.1, AB 2.3

Type A - thick, smooth and tough; Type B - moderately thick, smooth, firm to tough; Type C - thin, soft net-like appearance; Type 1
- smooth, circular, convex, shiny, tough, whitish to off-white; Type 2 - smooth, circular, convex, shiny, soft to firm, cream; Type 3 smooth, irregular, shiny, soft to firm, cream, release sticky exudates; Type 4 - rough, circular, shiny to dry, soft to firm, cream; Type
5 - rough/smooth, circular, shiny to dry, soft to firm, brownish.

Wildtype Cellulose-synthesizing Acetobacter strains

603

Fig. 2. Two types of colony morphology: a) smooth colony type 1 by strain lTD! 2.1, and b) rough colony type 5 by strain PA 2.2.
Cells were streaked on PYSS plates and incubated at 30 DC for 7 days.
II OJ

B08

A TJlmu\

Biolog grouping of isolates

Te

53582

LB 06
LB 15a

P\
B

J
J

LBp 1
LBI4

lJP(T 3

A t1lmu
BT 81
B1 61

23768

LBml
l B 12b
LBn l 1
LBp21
BT 93

-1L.._ _ _ _ __

102 5

I: cnlt
I' jllIlJrc\(en

Fig. 3. Dendrogram of cellulose producing strains based on

BIOLOGTM assay. Capital letters opposite each horizontal line
represent groupings of the different strains, followed by the isolate code names. The scale below the dendrogram is in units of
taxonomic distance, which means that, one unit of distance is
one test different, for non-variable specimen, and with variable
reaction being weighted non-linearly.

Biolog test assay revealed differences among the 38

wild-type isolates, specifically between the two selected
strains. Results of the 95 Biolog biochemical tests (data
not shown) indicated that each isolate has its own unique
set of biochemical fingerprints for carbon utilization. One
hundred percent of the strains, including three ATCC reference strains oxidized a-D-glucose, whereas, 92 % of the
isolates oxidized methyl pyruvate. Fructose and D-mannitol were both oxidized by 88% of the isolates with
81.7%,69%, and 42 % of the strains oxidizing D-L-Iactic
acid, glycerol and D-gluconic acid, respectively. Only the
members of the D, E and F groups, as shown in the constructed dendrogram (Fig. 3) oxidized formic acid.

Cellulose production and species identification of the

two strains
Strains lTD! 2.1 and PA 2.2 were singled out and selected from the rest of the 38 isolates because of the very
distinct phenotypic differences between the two. In order
to support the visual difference in pellicle thickness, the
amount of cellulose produced by the two, strains was
quantified. lTD! 2.2 strain produced up to 26 times more
cellulose as dry weight than PA 2.2, regardless of the carbon sources or incubation times in a batch static fermenTable 3. Comparison in the amount of cellulose produced in
terms of dry weight between the two Acetobacter strains.
Strain
lncub.
(days)

10% Sucrose
7-S
Average (g 1- 1)

7% Sucrose
7-S
Average (g 1- 1)

10% Suc + Fruc

S
Average (g 1-1 )

lTD! 2.1

12.53 0.37

S.1S0.25

6.93 0.69

PA2.2

0.472

0.79

0.01

0.05

1.14

0.09

604

E. B. BERNARDO et al.

Table 4. Phenotypic characteristics of the two wild type Acetobacter strains at the species level.
Differentiating features
Formation of
Water-soluble brown pigment on GYC++
medium
Gluconic acid
Dhiydroxyacetone from glycerol
Growth on carbon sources:
Ethanol
Methanol
Sodium acetate
Growth on L-amino acids in the presence
of mannitol as carbon source:
L-asparagine
L-glutamine
Growth in 30% D-glucose
Growth in 10% ethanol
Catalase
Ubiquinone type" "
G+C Content (mol%)**
16S rDNA sequence

lTDI 2.1

PA2.2

ATCC
53582

A. xylinus"

A. pasteurianus':

A. hansenii'

+
+

+a
+

+a

da
+

d
d

+
+
nd

+
+
nd

+
+
nd

+
nd
nd
D, This study

+
nd
nd
D, This study

+
nd
nd
nd

+
+
+
QI0
55-63
D"''''

d
+
Q9
53-63
D"""

+
nd
58-63
D':""

LEGENDS: D - determined; nd - not determined; d - 11-89% strains +, a = 5-ketogluconic acid

* Based on Bergey's Manual of Determinative Bacteriology, 1994; ,,':. Based on SWINGS, 1992; ,.",:. Based on SIEVERS et aI., 1994;
++ DELEY et ai., 1984
Acidiphilium multivorum AIU301 (AB0067 I I)
62

Acidocella sp. GS 19h (X9 I 797)

Acidosphaera rubrifaciens HS-AP3 (D86512)
Acetobacter liquefaciens IFOl2388 (X75617)
Acetobacter diazotrophicus ATCC49037 (X75618)
Acetobacter sp. PA2.2
Acetobacter hansenii NCIB8746 (X75620)
Acetobacter sp.ITDl2.1
Acetobacter oboediens LTH2460 (AJOO 1631)

51

Acetobacter intermedius TF2 (Y!4694)

Acetobacter xylinum NCIB 11664 (X756!9)

91

Acetobacter europaeus DES1! (Z2!936)

Acetobacter pomorum LTH2458 (AJOOI632)
Acetobacter pasteurianus LMD22.! (X71863)
75

Acetobacter aceti NCIB862! (X74066)

Gluconobacter Jrateurii IF03264 (X82290)

74
100

Acetobacter methanolicus MB58 (X77468)

Acidomonas methanolica IMETI 0945 (D30770)

L-_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

O.O!

Escherichia coli*

Fig. 4. Phylogenetic affiliations between

strains lTDI 2.1 (Ace. No. AF062475), PA
2.2 (Ace. No. AF062474), documented
species of the genus Acetobacter, and other
closely related bacteria as indicated by near
complete 16S rRNA gene sequence alignment. The phenogram was reconstructed
from the pairwise distance matrix using the
neighbor-joining method (SAITOU and NEI,
1987). An alignment of 1430 unambiguous
characters, between positions 27 and 1492
(E. coli numbering), was used to calculate genetic distances (JUKES and CANTOR, 1969).
The scale represents 1 base substitutions per
100 nucleotide positions. Strain designations
are indicated after the species name and
database (GenBank, EMBL, RDP) and accession numbers are shown in parentheses.
" The branched length for E. coli has been
halved.

Wildtype Cellulose-synthesizing Acetobacter strains

~O\~~~~~~~o~,-<,-<~~o~~

0\
,-<

~~~~~~~~~~~~~~~~~~
~~~~~~~~~~~~~~~~~~

ooOoooo'-<~~~~Ooo~N~~oo~

00
,-<

NrlrlrlNNrlrlrl~NrlrlNNrlrl

0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\
~,-<~N~O\~No\'-<~o\~Noo~

,-<

O,....;~~~~MMM~~~~~~~

0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\
~~~~ooOoo~'-<~N~~~N

~
~

OO~~~~MM~~~~~~O\

0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\
~oo~~NNO\~N~~oooo~

,-<

OO~~~~MM~~~~~~

0\0\0\0\0\0\0\0\0\0\0\0\0\0\
~~NO~~~No\N~Noo

~
~

,....;,....;~~~~~~~~~~~

0\0\0\0\0\0\0\0\0\0\0\0\0\
~~~~~oo~~~'-<N~

,-<

,....;N~~~~~~~~~O\

0\0\0\0\0\0\0\0\0\0\0\0\
~oo~~~oooo~~~~

,....;N~~~~~~~~~

0\0\0\0\0\0\0\0\0\0\0\

,,.....
...

0\0\0\0\0\0\0\0\0\0\

O~OO\~~~~o\N

NM~~~~~~~oO

~O~~O~O\~~

,....;M~~~~~~~

0\0\0\0\0\0\0\0\0\

605

tation (Table 3). The two strains were also subjected to

additional tests to further identify them at the speCies
level and the results were tabulated in Table 4.
Identification of the two wild-type strains by 16S
rONA sequence
Confirmation of species identification for strains lTD!
2.1 and PA 2.2 was achieved by using the complete 16S
rRNA sequences (accession numbers AF062475 and
AF062474, respectively). The resulting 16S rDNA sequences determined for both strains were compared with
those in the Genbank, EMBL, and Ribosomal Database
Project databases and revealed a high sequence similarity
to previously published complete 16S rRNA gene sequences for the current 7 Acetobacter species (SIEVERS,
1994). Sequence identity values shown in Table 5 were
obtained from which a phylogenetic tree was constructed. Based from these sequence similarity values, revealed
evolutionary relationships between strains IDTI 2.1 and
PA 2.2 with other Acetobacter species and related bacteria (Fig. 4). Based on the tree, strain lTD! 2.1 is most
closely related to A. xylinus and A. europaeus with a sequence similarity value of 98.7 for both. On the other
hand, strain PA 2.2 is most closely related to strain lTD!
2.1 and A. hansenii with a sequence similarity value of
98.4 and 98.2, respectively.

~O~N-N~N

rlr000000000r--.:OO

0\0\0\0\0\0\0\0\
~~oo~~~~

00

,....;M~~~~OO

0\0\0\0\0\0\0\
O~~~~~

NMoOoOoOoO

0\0\0\0\0\0\
N~~N~

NMo\o\o\

0\0\0\0\0\

0\,-<00

""';MO\O\

0\0\0\0\
ooN~

>....

O:"l

.~

'8

,.....~

'Vi

~
0

rlr0o\

0\0\0\

0\ 0\
~

0\

Discussion
It has always been speculated that the starter culture
used in producing cellulose nata pellicle is composed of a
heterogeneous population of microorganisms primarily
of different Acetobacter strains. But, there has not been
enough data to support this claim. Variations and differences exhibited by the 38 isolates in this study, based on
pellicle and colony types, indicated that there was heterogeneity in starter cultures used in nata de coco production. This may account for the observed batch to batch
differences in the type and volume of nata pellicle produced by the commercial producers. The two selected
strains, lTD I 2.1 and PA 2.2, explicitly demonstrated
these differences. Thick pellicle-forming strain lTD! 2.1,
showed a colony morphology distinct from the thin pellicle-forming strain PA 2.2. Colony type 1 morphology,
demonstrated by the thick cellulose pellicle-forming
strain, and the rest of the smooth colony forming isolates, is very different from a typical cellulose colony previously described. A typical colony produced by a cellulose synthesizing Acetobacter strain is described as initially smooth and spheroid and becoming rough, crinkled and flattened after 8 days incubation (DE LEY et aI.,
1984). In contrast, the colony produced by strain PA 2.2,
designated as colony type 5 and all the rest of rough
colony types in this study, conforms to the standard description of colony morphology for cellulose-producing
organisms except that, typically, these colonies were
raised not flattened after prolonged incubation. These results agreed with earlier works published on nata pelli-

606

E.

B. BERNARDO

et al.

cle-forming isolates of Aeetobaeter from several sources,

where colony, pellicle type and yield variations were described (LAPUZ et a!., 1967; GALLARDO-DE JESUS et a!.,
1973). In another study, variations were observed among
6 strains of A. xylinus in terms of colony morphology
but not on pellicle type (VALLA, 1991).
The morphological characteristics, plus an array of
other phenotypic tests, were sufficient to identify the two
selected strains and the rest of the cellulose-synthesizers
as members of the genus Aeetobaeter. The dendrogram
generated by the Biolog test results (Fig. 4) showed different subgroupings illustrating differences in carbon utilization among isolates. It also indicated that there exists
different species of Aeetobaeter, with regards to nata
producing wild-type organisms. The distinct clustering of
lTD! 2.1 from PA 2.2 and the other isolates is clearly indicative that each strain utilizes different types of carbon
substrates provided in the biolog assay kit. Despite the
subgroupings, all isolates, including the three ATCC
strains were clustered and arose from a common point of
origin, forming a coherent genus. The two outgroups
Escherichia coli and Pseudomonas {luoreseens were clustered distinct from the Aeetobaeter lineage. Both of these
outgroups are Gram-negative and non-acidophilic,
whereas Acetobaeter is a Gram-negative and acidophilic
genus.
The tests, which used the scheme by CARR and PASSMORE (1979) and SWINGS (1992) to identify the two
strains at the species levels, were not very useful for
strain/species differentiation and identification. When
compared with the three cellulose synthesizing type
species of Acetobacter, both strains lTD! 2.1. And PA
2.2 appeared to have met the minimal requirements for
A. hansenii (GOSSELE and SWINGS, 1985), but the fact
that both are strong producers of gluconic acid from Dglucose may also indicate that they may belong under A.
xylinus. 16S rRNA gene sequences generated for the two
selected strains however confirmed species differences
and led to the identification of the two strains. Sequence
similarities higher than 94.2 % for both strains confirmed
their identification as members of the genus Acetobacter
(SIEVERS et a!., 1994). Strain lTD! 2.1 shares the same
percent similarity (98.7) to two type species, A. xylinus
and A. europaeus, and also have a very near percent similarity to A. intermedius (98.6) and A. oboediens (98.5).
The latter two species however, were not compared in
detail with strain lTD! 2.1 since both are non-cellulose
producers (SOKOLLEK et a!., 1998). Another new species
of Acetobacter, A. pomorum (SOKOLLEK et a!., 1998),
shows a lower percent 16S rRNA similarity with strains
lTD! 2.1 and PA 2.2 since again, it is a non-cellulose producer and can grow in 3% ethanol with up to 4% acetic
acid. It is possible for ITD! 2.1 to share the same percent
similarity to two species, A. xylinus and A. europeaeus,
since these two are the closest related species based on
16S rDNA sequences (shared 99.6% similarity, SIEVERS
et a!., 1994) and on DNA-DNA hybridization and phenotypic characteristics (SIEVERS et a!., 1992). Although
some strains of A. europaeus synthesize cellulose, strain
ITD! 2.1 is classified here to be under A. xylinus and not

under A. europaeus because ITDI 2.1 does not have a requirement for acetic acid for growth, whereas A. europaeus has an absolute requirement for acetic acid in
the growth medium (SIEVERS and TEUBER, 1995). On the
other hand, strain PA 2.2 having 98.2 % sequence similarity to A. hansenii, and having met the minimal phenotypic and biochemical requirement of this type species, is
proposed to be under this species designation. The derived percent similarities of the two strains ITD! 2.1
(98.7%) and PA 2.2 (98.2%), to two different type
species A. xylinus and A. hansen ii, respectively, strongly
supported the earlier assumption that these isolates were
distinctly different and may represent a novel and previously undescribed species of cellulose-synthesizing Acetobaeter. Speciation with Acetobacter, based on 16S
rDNA sequence similarities, has been warranted with as
little as 0.3 % sequence dissimilarity.
It was necessary to elucidate the species identification
of the two selected strains in order to establish which of
the three cellulose-producing species: A. xylinus, A. pasteurianus and A. hansenii they may belong. In the past,
there were significant changes in the nomenclature and
identification of cellulose producing Aeetobacter species
(DE LEY and FRATEUR, 1974; GOSSELE, 1983; YAMADA,
1983; DE LEY et a!., 1984). In the case of the nata-forming organism, the first correct identity, based on morphological and biochemical characteristics, was established
as belonging to Acetobacter xylinus subsp. Xylinus sensu
DE LEY and FRATEUR (LAPUZ et a!., 1967; GALLARDO-DE
JESUS et a!., 1973). Another nata bacterium obtained
from the Philippines was identified as Acetobaeter hansenii (GOSSELE and SWINGS, 1985) based on numerical
analysis of 177 phenotypic features (GOSSELE et a!.,
1983). This nata bacterium differed in a few characteristics from strain PA 2.2. It does not produce ketogluconic
acids from D-glucose and it has strong ketogenesis (dihydroxyacetone formation) towards polyalcohols whereas
the opposite in both tests were manifested by strain PA
2.2 (Table 4).
The reported cellulose dry weights for the two strains
were obtained from sucrose and fructose as carbon
sources where they grow favorably. Strain ITDI 2.1 can
also produce cellulose balls in shake culture condition
similar to A. xylinus subsp. sucrofermentans. In preliminary tests, it has been observed to produce an average
un optimized production yield of 2.50 and 3.70 g 1-1 in
sucrose and fructose, respectively, as compared to 4.4 g
1-1 optimum production yield of A. xylinus subsp. suerofermentans from Corn Steep liquor-Fructose medium
(TOYOSAKI et a!', 1995). Phylogenetic ally, strain lTD! 2.1
cannot be compared with A. xylinus subsp. sucrofermentans since there was no published 16S rRNA sequence
data for this organism. However, a few phenotypic differences were seen. ITDI 2.1 oxidizes fructose but not
lactose, produces abundant gluconic acid from glucose,
cannot grow in 10% ethanol but can grow minimally in
lactose, and shows growth in L-asparagine and L-glutamine in the presence of mannitol as the carbon source.
A. xylinus subsp. sucrofermentans showed the opposite
traits in all cases.

Wildtype Cellulose-synthesizing Acetobacter strains

In view of the unique characteristics of strains lTDl

2.1 and PA 2.2, in terms of colony and pellicle morphologies, carbon oxidation patterns, cellulose production abilities, growth in different carbon sources and 16S
rRNA sequence data, it is probable that these two strains
belong to previously undescribed subspecies of A. xylinus and A. hansen ii, respectively.
This study has shown that strain or species variations
may indeed dictate the quality and type of nata pellicle
production. The observed phenotypic differences among
the different isolates specifically on the two selected
strains, lTD! 2.1 and PA 2.2, supported further by the
significant difference in cellulose production yield may
lead us to conclude that synthesis of nata pellicle is strain
dependent. The 16S rDNA sequence analysis confirmed
that the two wild-type cellulose producing strains belonged to two different Acetobacter species.
Acknowledgments
Strain lTD! 2.1 was a generous gift from the Microbiology
Division Culture Collection Center, of the Department of Science and Technology (DOST) in the Philippines. The thin cellulose-producing strain, PA 2.2, isolated in this study, was obtained from a mother liquor from Laguna province kindly given
by Dr. Ric del Rosario from the Institute of Food Science and
Technology (IFST), University of the Philippines in Los Banos.
Thanks are also due to the entire nata de coco producers/growers and friends who generously provided the mother liquor or
mixed cultures used in this study. The first author gratefully acknowledges an Australian government sponsored scholarship
through AusAID.

References
ATCC: Bacteria, pp. 1-389. In: American Type Culture collection Catalogue of Bacteria and Bacteriophages (R. GHERNA
and P. PIERTA) Rockville, Maryland 1992.
BETP: Nata de coco Exporters and Suppliers' (1993 to 1997)
Statistics. Manila, Philippines, Philippine Bureau of Export
and Trade Promotion (1997).
BIOLOG: Biolog Instruction Manual. Hayward, CA: Biolog Inc.
1993.
.
BROWN, M. ].: Bacterial Cellulose, pp. 145-151. In: Cellulose:
Structural and Functional Aspects (J. E KENNEDY, G. O.
PHILLIPS and P. A. WILLIAMS) Chichester, West Sussex, England, Ellis Horwood Limited 1989.
BROWN, R. M . ].: The Biosynthesis of cellulose. ]. M. S. - Pure
and Appl. Chern. A33 (10), 1345-1373 (1996).
CANNON, R. E. and ANDERSON, S. M .: Biogenesis of bacterial
cellulose. Crit. Rev. Microbiol. 17 (6 ), 435-447 (1991).
CARR, ]. G. and PASSMORE, S. M.: Identification of acetic acid
bacteria. In: Identification Methods for Microbiologists (E
A. SKINNER, and D . LOVELOCK) London, Academic Press
1979.
DE LEY,]. and FRATEUR,].: Genus ACETOBACTER, Part 7 . GramNegative Aerobic Rods and Cocci, pp. 21 7-285. In: Bergey's
Manual of Determinative Bacteriology (R. E. BUCHANAN and
N. E. GIBBONS), 8th Edition, Baltimore, USA, The Williams
and Wilkins Co. 1974.
DE LEY, ]., GILLIS, M. and SWINGS, ].: Family VI. Acetobacteriacaea ., pp. 267-278. In: Bergey's Manual of Systematics Bacteriology., Vol. 1 (N. R. KREIG, and j. G. HOLT) Baltimore,
Williams and Wilkins, Co. 1984.

607

EUZEBY, J. P.: Revised nomenclature of specific and subspecific

epithets that do not agree in gender with the generic names
that end in -bacter. Int. j. Syst. Bacteriol. 47 (2), 585 (1997).
FELSENSTEIN, ].: PHYLIP: Phylogeny inference package. Cladistics. 5,164-166 (1989).
GALLARDO-D E JESUS, E., ANDRES, R. M. and MAGNO, E. T.: A
study on the isolation and screening of microorganisms for
the production of diverse-textured nata. Philip. j. Sci 100 (1),
41-49 (19 73).
GOSSELE, E, SWINGS, j., KERSTERS, K. and DE LEY,].: Numerical
analysis of phenotypic features and protein gel e1ectrophoregrams of a wide variety of Acetobacter strains. Int. ]. Syst.
Bacteriol. 33, 65-81 (1983).
GOSSELE, E and SWINGS, ].: Identification of nata-producing
bacterium as Acetobacter hansenii. Philip. J. Sci. 114,
179-182 (1985).
HAIGLER, C. H. : The alteration of cellulose assembly in Acetobacter xylinum by fluorescent brightening agents, direct dyes
and cellulose derivatives, The University of North Carolina
at Chapel Hill, 0153 (1982).
HESTRIN, S. and SCHRAMM, M.: Synthesis of cellulose by Acetobacter xylinum. II. Preparation of freeze-dried cells capable
of polymerizing glucose to cellulose. Biochem. j. 58,
345-352 (1954).
HOLT, j. G., KREIG, N. R., SNEATH, P. H., STANLEY, j. T. and
WILLIAMS, S. T.: Bergey's Manual of Determinative Bacteriology. Baltimore, MD .: Wiiliams and Wilkins 1994.
HOTCHKISS, A. T. j. and BROWN, R. M.: Evolution of the cellulosic cell wall in the Charophyceae, pp. 591-609. In: Cellulose and Wood - chemistry and technology. (c. SCHUERCH)
New York, john Wiley and Sons, Inc. 1989.
jHINGAN, A. K.: A novel technology for DNA isolation. Meth.
Mol. and Cell. BioI. 3,15-22 (1992).
JUKES, T. H. and CANTOR, C. R.: Evolution of protein
molecules, pp. 21-132. In: Mammalian Protein Molecules,
vol. 3 (H. N. MUNRO) New York, N. Y., Academic Press, Inc.
1969.
LAPUZ, M. M., GALLARDO, E. G. and PALO, M. A.: The nata organism: cultural requirements, characteristics and identity.
Philip. J. Sci. 96, 91-109 (1967).
MASAOKA, S., OHE, T. and SAKATO, N.: Production of cellulose
from glucose by Acetobacter xylinum. j. Ferment. Bioeng. 75
(1),18-22 (1993).
NEILAN, B. A., JACOBS, D., DEL DOT, T., BLACKALL, L. L.,
HAWKINS, P. R., Cox, P. T. and GOODMAN, A. E.: rRNA sequences and evolutionary relationships among toxic and
nontoxic cyanobacteria of the genus Microcystis. Int. ]. Syst.
Bacteriol. 47 (3), 696-697 (1997).
NSRI: Catalog of Culture Collections, Natural Sciences Research Institute, University of the Philippines in Diliman,
Quezon City, Philippines (1990).
Ross, P., MAYER, R. and BENZIMAN, M.: Cellulose biosynthesis
and function III bacteria. Microbiol. Rev. 55 (1), 35-58
(1991).
SAITOU, N. and NEI, M.: The neighbor-joining method: a new
method for reconstructing phylogenetic trees. Mol. BioI.
Evol. 4, 406-425 (1987).
SIEVERS, M., LUDWIG, W. and TEUBER, M.: Phylogenetic positioning of Acetobacter, Gluconobacter, Rhodopila and
Acidiphilium species as a branch of acidophilic bacteria in
the a-subclass of Proteobacteria based on 16S ribosomal
DNA sequences. System. Appl. Microbial. 17, 189-196
(1994).
SIEVERS, M., SELLMER, S. and TEUBER, M .: Acetobacter europeus
sp. nov., a main component of industrial vinegar fermentors
in Central Europe. System. Appl. Microbiol. 15, 386-392
(1992).

608

E. B. BERNARDO et al.

SIEVERS, M. and TEUBER, M.: The microbiology and taxonomy

of Acetobacter europeus in commercial vinegar production,
pp. 84S-95S. In: Microbial Fermentations: Beverages, Foods
and Feeds. (R. G. BOARD, D. JONES and B. JARVIS) Oxford,
UK, Soc. Appl. Bacteriol., Blackwell Science Ltd. 1995.
SOKOLLEK, S., HERTEL, C. and HAMMES, W.: Description of Acetobacter oboediens sp. nov. and Acetobacter pomorum sp.
nov., two new species isolated from industrial vinegar fermentation. Int. J. of Syst. Bacteriol. 48, 935-940 (1998).
SWINGS, J.: The Genera Acetobacter and Gluconobacter, pp.
2268-2286. In: The Prokaryotes (A. BALOWS, H. G. TRUPER,
M. DWORKIN, W. HARDER, and K.-H. SCHLEIFER), 2nd Ed.
New York, Springer-Verlag, Inc. 1992.
THOMPSON, J. D., HIGGINS, D. G. and GIBSON, T. J.: CLUSTAL
W: improving the sensitivity of progressive multiple sequence
alignment through sequence weighting, position specific gap
penalties and weight matrix choice. Nucl. Acids Res. 22,
4673-4680 (1994).
TILLETT, D. and NEILAN, B. A.: Small-scale preparation of the
single-copy bacterial artificial chromosome (BAV) vector:
pBeloBAC11. Biotech. 24, 568-572 (1998).
TOYOSAKI, H., NARITOMI, T., SETO, A., MATSUOKA, M., TSUCHIDA, T. and YOSHINAGA, E: Screening of bacterial celluloseproducing Acetobacter strains suitable for agitated culture.
Biosci. Biotech. Biochem. 59 (8), 1498-1502 (1995).
TOYOSAKI, H., KOJIMA, Y., TSUCHIDA, T., HOSHINO, K., YAMADA,
Y. and YOSHINAGA, E: The characterization of an acetic acid
bacterium useful producing bacterial cellulose in agitation
cultures: The proposal of Acetobacter xylinum subsp. suro-

fermentans subsp. nov. J. Gen. Appl. Microbiol. 41, 307-314

(1995).
VALLA, S.: Cloning of genes involved in cellulose biosynthesis in
Acetobacter xylinum. pp. 245-257. In: Biosynthesis and
Biodegradation of Cellulose, (c. H. HAIGLER and P. J.
WIEMER) New York, Marcel Dekker, Inc. 1991.
WARD, K. J.: Occurrence of Cellulose. Chapter II, pp. 9-27. In:
Cellulose and Cellulose Derivatives, Part I (H. E. OTT, M.
SPURLIN and M. W. GRAFFLIN) London, Interscience Publishers Ltd. 1954.
WHITE, D. G. and BROWN, R. M.: Prospects for the commercialization of the biosynthesis of microbial cellulose, pp.
573-590. In: Cellulose and Wood - Chemistry and Technology (c. SCHUERCH) New York, John Wiley and Sons, Inc.
1989.
YAMADA, T. : Acetobacter xylinus sp. nov., nom. rev., for the cellulose-forming and cellulose-less, acetate-oxidizing acetic
acid bacteria with the Q-10 system. J. Gen. Appl. Microbiol.
29,417-420 (1983).
YAMANAKA, S., WATANABE, K., KITAMURA, N., IGUCHI, M., MITSUHASHI, S., NISHI, Y. and URYU, M.: The structure and mechanical properties of sheets prepared from bacterial cellulose. J. Material Sci. 24, 3141-3145 (1989).

Corresponding author:
1. COUPERWHITE, School of Microbiology and Immunology, University of New South Wales, Sydney, 2052 NSW, Australia
e-mail: LCouperwhite@unsw.edu.au