Mutation Research 412 1998.

323331

Adsorption of a hydrophobic mutagen to cereal brans and cereal bran dietary fibres
Philip J. Harris a,) , Vallappilakkandy K. Sasidharan b, Anthony M. Roberton a , Christopher M. Triggs c , Anthony B. Blakeney d , Lynnette R. Ferguson b
School of Biological Sciences, The Uniersity of Auckland, Auckland, New Zealand Cancer Research Laboratory, The Uniersity of Auckland, Auckland, New Zealand c Department of Statistics, The Uniersity of Auckland, Auckland, New Zealand The Bread Research Institute of Australia, P.O Box 7, North Ryde, NSW 2113, Australia
b a

Received 28 October 1997; revised 18 December 1997; accepted 19 December 1997

Abstract The abilities of brans from the cereals barley, oats, maize, rice, and wheat to adsorb in vitro the hydrophobic, environmental mutagen 1,8-dinitropyrene DNP. were investigated using a mutagenicity assay. These brans were obtained from known cultivars using defined milling conditions and were chemically characterised. The abilities of total and insoluble dietary fibre preparations obtained from these brans to adsorb DNP were also investigated. The predicted weight of each bran required to adsorb 50% of the added DNP was used to compare the adsorptive abilities of the different brans. The brans were ranked in the order most effective to least effective.: rice, wheat, maize, barley, and oats. The adsorptive abilities of the dietary fibre preparations were not significantly different from the bran from which they were prepared. However, if the dietary fibres cell walls. were the only components adsorbing the DNP, we would have expected the dietary fibre preparations to have adsorbed more DNP than the equivalent unextracted bran. This suggests that other components, probably starch, also adsorb DNP in the unextracted brans. It is not known why brans from different cereal species differ in adsorptive ability but the lignified cell walls in wheat bran may be important in conferring good adsorptive properties to this bran. The possible relationship between adsorptive ability and ability of the bran from a particular species to protect against colorectal cancer is discussed. q 1998 Elsevier Science B.V.
Keywords: Adsorption; Cereal bran; Colorectal cancer; Dietary fibre; 1,8-Dinitropyrene; Plant cell wall

1. Introduction Epidemiological and animal studies indicate that cereal brans in diets can protect against a variety of
) Corresponding author. School of Biological Sciences, The University of Auckland, Private Bag 92019, Auckland, New Zealand. Tel.: q64-9-373-7599 ext. 8366; fax: q64-9-373-7416; E-mail: p.harris@auckland.ac.nz

diseases common in the western world, including colorectal cancer. Furthermore, the cereal species from which the bran is derived appears to be important in determining the protective effects of the bran. For example, significant protection against colorectal cancer was found in most of the experiments in which rats were exposed to carcinogens and fed wheat bran throughout the experiment w1,2x. A single study using rice bran showed that this also protected

1383-5718r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved. PII S 1 3 8 3 - 5 7 1 8 9 8 . 0 0 0 0 3 - 5

324

P.J. Harris et al.r Mutation Research 412 (1998) 323331

against colorectal cancer in rats w2x. In contrast, animal studies using oat and barley brans have usually shown that instead of being protective these increased the incidence of colorectal cancer w35x. However, brans, even from the same cereal species, can vary in composition and physical properties. A major reason for this is because bran preparations produced by milling do not contain only bran layers as defined botanically. These bran layers consists of the outer tissue layers of cereal grains which overlie the starchy endosperm. During milling the bran layers split away from the starchy endosperm, but this separation is never perfect and thus the resulting bran preparation is contaminated, to a greater or lesser extent, by the starchy endosperm. The extent of contamination can be estimated by determining the starch content of the bran preparation as the bran layers contain no starch w6x. With wheat grains this separation is often good, whereas with oat grains it is usually poor and can result in oat-bran preparations with particularly high contents of starch w7,8x. Different cultivars of a cereal species can vary in how well the bran layers can be split from the starchy endosperm. For example, with hard wheat cultivars the separation is better than with soft wheat cultivars. Furthermore, the particle sizes of bran preparations can vary depending on the milling conditions and supplementing the diet of rats with wheat-bran preparations of varying particle sizes affected their colonic function w9x. The cell-wall or dietary fibre component of cereal brans may be important in the protection against colorectal cancer. Cereal brans are a good source of dietary fibre; indeed cereal brans are often erroneously described as dietary fibres despite dietary fibre making up less than half of many bran preparations. Furthermore, the composition and properties of plant cell walls can vary depending on the plant species and this may account, at least in part, for the different protective abilities of brans from different cereal species. One way the dietary fibre component of cereal brans may protect against colorectal cancer is by adsorbing carcinogens or promoters in the digestive tract. Providing the cell walls are not digested in the gastrointestinal tract, the carcinogens or promoters would be carried out of the body adsorbed to the cell walls. Thus, the effective concentrations available to initiate or promote cancerous changes in

the gut mucosal cells would be lowered. In previous in vitro studies w1012x, we found that a range of dietary fibre plant cell wall. preparations with contrasting compositions all adsorbed the environmental, hydrophobic mutagen and carcinogen, 1,8-dinitropyrene DNP. w13x, but their abilities to do so varied substantially. Here we report the results of experiments in which we determined, using a mutagenicity assay, the adsorption of DNP to cereal bran preparations obtained using defined milling conditions from known cultivars of barley, maize, oats, rice, and wheat. An undefined oat-bran preparation obtained from a store was also examined. We also examined the adsorption of DNP to total dietary fibre and insoluble dietary fibre preparations obtained from these brans to determine whether components other than dietary fibre were involved in the adsorption.

2. Materials and methods 2.1. Bran preparations Bran preparations were obtained from the following cereals grown under field conditions in New South Wales, Australia: bread wheat Triticum aestium cv. Sunkota., hulless barley Hordeum ulgare cv. Bald Skinless., oats Aena satia cv. Cooba., rice Oryza satia cv. Amaroo., and maize Zea mays hybrid XL 80 from Diekalb Shand Seed.. All samples were cleaned using a Kice Mini-Aspirator Kice Industries, Wichita, KS, USA.. The oats and rice were hulled using a Satake TA 105 laboratory rice huller Satake Engineering, Hiroshima, Japan.. The rice was then milled in a McGill No. 3 Rice Miller Seedburo Equipment, 1022 West Jackson Blv., Chicago, IL, USA. and the bran was cleaned of broken rice by sieving or aspiration. Except for the oats, the other grains were milled with a belt drive, four strand experimental mill Allis Charmers, Milwaukee, WI, USA. using the short stand option of AACC method 26-21 w14x, but the wheat was given an extra break pass. All grains were tempered to 15% moisture for 18 h before milling AACC 26-10. w14x. The oats were milled on the same experimental mill using two passes through the first break rolls and three passes through the second

P.J. Harris et al.r Mutation Research 412 (1998) 323331

325

break rolls. Bran was collected as the material that did not pass an 18 wire sieve. Tempering water 1% wrw. was added to the oats 10 min before milling to toughen the bran. Where necessary, the particle size of bran fractions was reduced by milling in a Wiley mill to pass a 1 mm screen. 2.2. Isolation of dietary fibre preparations Total and insoluble dietary fibre preparations were isolated from the brans by modifications of the methods described by Prosky et al. w15,16x. Heat stable a-amylase Cat. No. A 3306., amyloglucosidase Cat. No. A 9913., and protease Cat. No. P 3910. were obtained from Sigma Chemical, St. Louis, MO, and were components of the Total Dietary Fiber Assay Kit TDF-100.. The total dietary fibre preparation was isolated by weighing bran 500 mg. dried over silica gel in a desiccator. into a centrifuge tube 250 ml capacity.. Sodium phosphate buffer 25 ml, 80 mM, pH 6.0. and a-amylase 50 m l. were added to the tube which was then capped and heated for 30 min, with gentle shaking at 5 min intervals, in a boiling water bath. After cooling to room temperature, the pH was adjusted to 7.5 with sodium hydroxide 5 ml, 275 mM., protease 2.5 mg. was added, and the solution incubated for 30 min at 608C with continuous shaking. After cooling again to room temperature, the pH was adjusted to 4.5 with hydrochloric acid 5 ml, 325 mM., amyloglucosidase 150 m l. was added, and the solution incubated for another 30 min at 608C with continuous shaking. Ethanol 140 ml, 95%, preheated to 608C. was added and after 60 min at room temperature the suspension was centrifuged 11,500 = g, 10 min.. The supernatant was discarded and the pellet washed three times by centrifugation as above. with 78% ethanol 20 ml, 78%. and then twice with 95% ethanol 10 ml., and twice with acetone 10 ml.. The residue total dietary fibre preparation. was allowed to dry in a stream of air and then in a vacuum desiccator over silica gel. The insoluble dietary fibre preparations were isolated in a similar way except after incubation with the amyloglucosidase, no 95% ethanol was added before the suspension was centrifuged as above.. The supernatant was discarded and the pellet washed three times by centrifugation as above. with water

10 ml., then twice with 95% ethanol 10 ml., and twice with acetone 10 ml.. The residue insoluble dietary fibre preparation. was dried as above. 2.3. Chemical characterisation of brans and dietary fibres The nitrogen content, which was used as a measure of the protein content, was determined by Kjeldahl digestion using the AACC method 46.10 w14x. Starch was determined by ball milling the samples in a Fritsch Pulverisette Fritsch, Idar-Oberstein, Germany. mill and then digesting them with thermostable a-amylase Novo Termamyl 120L. w17x followed by amyloglucosidase w18x. The glucose produced was determined using glucose oxidaserperoxidase w19x. Total 1 3, 1 4.-b-D-glucan was measured by the method of McCleary and Codd w20x using a kit from Megazyme International Ireland Bray Business Park, Bray, Wicklow, Ireland.. 2.4. Incubation of brans and dietary fibres with DNP in PBS All incubations were done in acid-washed, conical glass centrifuge tubes capacity 12 ml.. Brans and dietary fibres 0.1, 0.5, 1.0, 2.5, 5.0, and 10.0 mg. were weighed into the tubes and 1.96 ml of phosphate-buffered saline PBS. 20 mM sodium phosphate buffer, pH 6.5, containing 130 mM sodium chloride. added to each tube. Each tube stood for 30 min to allow the bran or dietary fibre to hydrate, then at time zero, 40 m l of a solution of DNP in DMSO 100 ng DNP, final DMSO concentration 2% vrv. was added and the contents of the tubes mixed using a vortex mixer. This gave concentrations expressed as ng DNP per mg of bran or dietary fibre. of 1000, 200, 100, 40, 20, and 10. The tubes were shaken 120 rpm. on an orbital shaker for 55 min at 378C, centrifuged 2500 = g for 5 min., and aliquots 3 = 50 m l. of the supernatants taken for the mutagenicity assay. Control incubations were also done in which DNP was incubated with no bran or dietary fibre. DNP associated with the bran or dietary fibre and with the tube walls was determined in the following way. The pellet was resuspended by vortexing 1 = 10 s. and the contents then tipped into a second tube. The first tube was rinsed with PBS 2 ml. and the

326

P.J. Harris et al.r Mutation Research 412 (1998) 323331

contents added to the second tube. This second tube was centrifuged 2500 = g, 5 min. and the supernatant discarded. DMSO 2 ml. was added to the pellet, the contents mixed using a vortex mixer 2 = 10 s., and the tube recentrifuged 2500 = g, 5 ml.. Aliquots 3 = 50 m l. were removed for the mutagenicity assay. This procedure measures the DNP associated with the bran or dietary fibre. The amount of DNP associated with the tube walls was determined by adding DMSO 2 ml. to the first tube, the contents mixed using a vortex mixer 3 = 10 s., and aliquots 3 = 50 m l. removed for the mutagenicity assay. The results were calculated as a percentage of the mutagenicity of DNP added initially. The latter was determined by mixing DMSO 1.96 ml. and a solution of DNP 40 m l containing 100 ng. in DMSO and assaying the mutagenicity of aliquots 3 = 50

m l.. Data for a minimum of two experiments were averaged. There was almost complete recovery of the DNP that was added. This showed that there were no other interactions between components of the bran or dietary fibre and the DNP. Thus, this justified using only the DNP recovered from the bran for the study.
2.5. Bacterial mutagenicity assay The Salmonella typhimurium plate incorporation assay used the strain TA98 as described by Maron and Ames w21x. To minimize day-to-day variability in tester strain sensitivity to mutagen, which is largely due to the exact growth phase of the culture, we grew cells for quantitative work as follows. A 1-ml vial 2 = 10 8 cells. was removed from storage at y808C and inoculated into bacterial complete medium 20 ml.. It was grown until 10-fold dilution

Table 1 The content of starch, nitrogen and 1 3,1 4.-b-glucans in the bran and bran dietary fibre preparations Bran or bran dietary fibre Barley Bran Total dietary fibre Insoluble dietary fibre Oats (c. Cooba) Bran Total dietary fibre Insoluble dietary fibre Oats (commercial) Bran Total dietary fibre Insoluble dietary fibre Maize Bran Total dietary fibre Insoluble dietary fibre Rice Bran Total dietary fibre Insoluble dietary fibre Wheat Bran Total dietary fibre Insoluble dietary fibre ND, Not determined. Starch % DW. 38.6 2.4 2.5 Nitrogen % DW. 2.2 3.4 2.2 1 3,1 4.-b-Glucans % DW. 4.1 7.0 4.1

4.2 5.4 7.2

3.7 4.7 6.2

6.8 16.4 7.7

70.8 ND ND

3.4 ND ND

2.2 ND ND

8.0 1.8 0.1

1.1 0.7 0.6

0.2 0.2 0.2

11.5 7.1 8.2

2.4 2.6 2.7

0.1 0 0

16.3 15.3 11.8

2.7 1.9 1.8

1.7 2.6 2.7

P.J. Harris et al.r Mutation Research 412 (1998) 323331

327

in fresh medium gave an absorbance of 0.110.12 at 654 nm, which took approximately 3 h. DNP levels in experiments were measured by assaying mutagenicity and each measurement was performed in triplicate on at least two separate occasions. There is a linear relationship between the amount of DNP added to each plate and the number of revertant colonies over the dose range 05 ng w10x. The maximum amount of DNP added to each plate in experiments with brans or dietary fibres was 2.5 ng which gave approximately 1500 revertant colonies. The background number of revertant colonies was approximately 65. Positive controls containing 4-nitro-o-phenylene diamine 100 m grplate. were included in all experiments to test the bacterial strain response. Results of experiments in which DNP was incubated with bran, or dietary fibre were expressed as a percentage of the mutagenicity of DNP added initially.

bran cv. Cooba.. With the 1 3, 1 4.-b-glucans, in barley and oats cv. Cooba. the total dietary fibre preparation had a higher content than the insoluble dietary fibre preparation or the unextracted bran.. In wheat and maize the 1 3, 1 4.-b-glucan content of the total and insoluble dietary fibre preparations was similar. 3.2. The adsorption of DNP to the brans and dietary fibre preparations The percentages of added DNP adsorbed by different weights of the brans and dietary fibre preparations were transformed to stabilize variance and regression lines were fitted relating logit percentage adsorbed. to log weight of bran or dietary fibre preparation. The highest percentage of DNP was adsorbed by the largest weights of the brans and dietary fibre preparations and decreased with decreasing weight. This is illustrated for rice bran and

3. Results 3.1. Compositions of the bran and dietary fibre preparations The contents of starch, nitrogen and 1 3, 1 4.-b-glucans are shown in Table 1. The starch content of the bran preparations ranged from 4.2% for the oat bran cv. Cooba. to 70.8% for the commercial oat bran. This indicates that the commercial oat bran was extensively contaminated with starchy endosperm. The nitrogen content of the brans ranged from 1.1% for maize bran to 3.7% for the oat bran cv. Cooba. indicating that the maize bran had the lowest and the oat bran the highest protein content. The 1 3, 1 4.-b-glucan content of the brans ranged from 0.1% for the rice bran to 6.8% for the oat bran cv. Cooba.. With the exception of the oat bran cv. Cooba., the starch content of the total and insoluble dietary fibre preparations was lower than that of the equivalent unextracted bran. The nitrogen content of the total and insoluble dietary fibre preparations from the maize and wheat brans was lower than that of the content of the equivalent unextracted brans, but this was not so for the preparations from barley and oat

Fig. 1. The effect of increasing weights of rice bran and rice bran dietary fibres on the percentage of 1,8-dinitropyrene DNP. adsorbed. DNP 100 ng. was incubated in phosphate-buffered saline 2 ml. with various weights of bran or bran dietary fibre and samples were taken after 55 min for mutagenicity assays. The mutagenic activities associated with the bran or bran dietary fibre were measured. Values two experiments. represent total numbers of revertant colonies minus those from negative controls. expressed as a percentage of the numbers obtained for a comparable sample of DNP diluted into DMSO and plated immediately. The data for these experiments were transformed and regression lines were fitted relating logit percentage adsorbed. to log weight of bran or dietary fibre. I s Bran preparation; ^ s total dietaryfibre preparation; and `s insoluble dietary-fibre preparation.

328

P.J. Harris et al.r Mutation Research 412 (1998) 323331

rice bran dietary fibres in Fig. 1. With increased weights of all the brans and dietary fibre preparations, the percentage of DNP in the supernatant remained approximately constant, but the percentage adsorbed to the tube walls decreased data not shown.. The effectiveness of each bran and dietary fibre preparation in adsorbing DNP can be summa-

rized by two parameters characterizing each fitted line. The response of percentage DNP adsorption to different weights of added bran or dietary fibre preparation was measured by the slopes of the regression lines. The overall effectiveness of the brans and dietary fibre preparations can be expressed in a number of ways, such as the intercept of the regres-

Table 2 Percentage of added DNP adsorbed to brans and bran dietary fibres after incubation of a solution of DNP in phosphate-buffered saline for 1 h in the presence of different weights of brans or bran dietary fibre Bran or bran dietary fibre Oats (commercial) Bran Total dietary fibre Insoluble dietary fibre S.E.D. Oats (c. Cooba) Bran Total dietary fibre Insoluble dietary fibre S.E.D. Barley Bran Total dietary fibre Insoluble dietary fibre S.E.D. Maize Bran Total dietary fibre Insoluble dietary fibre S.E.D. Wheat Bran Total dietary fibre Insoluble dietary fibre S.E.D. Rice Bran Total dietary fibre Insoluble dietary fibre S.E.D. S.E.D. for brans only
a

Slope of line a

% DNP adsorbed 0.1 mg

b

Predicted weight to adsorb 50% added DNP

c

10 mg

0.32 0.41 0.37 0.040

20 6.3. 16 19 4.8

51 6.6. 55 57 4.9

8.4 5.9 4.9 1.9

0.32 0.33 0.31 0.030

22 6.5. 20 21 3.6

55 6.5. 53 52 3.7

5.2 6.5 7.8 61.7

0.52 0.40 0.46 0.066

12 5.1. 17 12 6.7

59 6.5. 56 54 8.2

4.9 5.5 6.9 1.7

0.42 0.36 0.34 0.045

17 5.9. 22 22 5.6

59 6.5. 59 58 5.5

4.1 3.5 4.1 1.6

0.79 0.50 0.49 0.111

7 3.9. 13 15 12.1

73 5.9. 61 63 13.4

2.9 4.2 3.4 1.8

0.65 0.52 0.44 0.058 0.074

13 5.2. 18 22 7.3

74 5.8. 70 73 6.8

2.0 1.9 1.3 1.6 1.67

The data were transformed and regression lines were fitted relating logit percentage adsorbed. to log weight of bran or bran dietary fibre. These are the slopes of the regression lines. b Percentage of added DNP adsorbed by 0.1 mg of bran or bran dietary fibre. c Percentage of added DNP adsorbed by 10 mg of bran or bran dietary fibre. Numbers in parentheses are SEDs for brans only.

P.J. Harris et al.r Mutation Research 412 (1998) 323331

329

sion line or the percentage DNP adsorbed by 0.1 mg or by 10 mg added. When the slopes of the fitted regression lines for the brans were compared Table 2., there were significant differences between them p - 0.001.. The shallowest slope was shown by the two oat brans, and the steepest by the wheat bran. Thus, for each added unit of weight, wheat bran was the most effective in adsorbing DNP, changing the percentage adsorbed by a factor of 2.5 compared with the least responsive, oat bran. For any given weight of bran the different brans differed in the percentage DNP adsorbed. When 0.1 mg of bran was added, wheat bran, barley bran and rice bran adsorbed less DNP than the two oat brans p - 0.001.. Because of the differences in the slopes given by the different brans, when 10 mg of bran was added, the wheat and rice brans adsorbed approximately 15% more than the other brans p - 0.001.. For each fitted line both of the characterising parameters, the slope and the percentage DNP adsorbed by 0.1 mg or 10 mg can be combined into a single measure: the predicted weight required to adsorb 50% of the added DNP. This predicted weight can be used to compare the different brans and dietary fibre preparations. For the brans the weight ranged from 8.4 mg for oat bran commercial. to 2.0 mg for the rice bran Table 2.. Three brans, the two oat brans and the barley bran were less effective, whereas the brans of maize, wheat and rice were more effective p - 0.001.. For a particular bran, comparison of the fitted regression lines for the unextracted bran, the total dietary fibre and the insoluble dietary fibre showed no differences for the two oat brans, the maize bran and the barley bran, p values 0.14, 0.07, 0.13 and 0.12, respectively. Table 2.. For both the wheat and rice bran, the fitted regression lines for the total and insoluble dietary fibre preparations were not different p values 0.64 and 0.89, respectively., but the unextracted brans were more responsive to differences in the weight added, having steeper slopes than the dietary fibre preparations p values 0.011 and 0.003, respectively.. For the wheat bran the difference in slope between unextracted bran and the two dietary fibre preparations meant that when 0.1 mg was added the total and the insoluble dietary fibre preparations adsorbed more DNP than unextracted

bran whereas when 10 mg was added this was reversed, although the differences were not formally significant. For the rice bran, when 0.1 mg was added the difference in slope resulted in a difference in percentage DNP adsorbed of 7% whereas when 10 mg was added there was no difference, although the differences were not formally significant. For both the rice bran and the wheat bran, in addition to the other brans, the predicted weight to achieve 50% adsorption of the added DNP was not significantly different between the unextracted bran and the two dietary fibre preparations. 4. Discussion Our results, obtained using a mutagenicity assay, showed that bran preparations from different cereal species adsorbed the hydrophobic mutagen and carcinogen DNP to different extents. The rice bran showed the greatest ability to adsorb DNP and oat bran showed the least. Furthermore, the total and insoluble dietary fibre preparations prepared from these brans showed adsorptive abilities similar to the unextracted brans. Dietary fibre preparations from wheat bran have also been shown to be much more effective than dietary fibre preparations from fruits and vegetables in adsorbing the heterocyclic amine carcinogen 2-amino-3,8-dimethylimidazo w 4,5f xquinoxaline MeIQx. w22,23x. In our experiments, if the cell walls were the only bran components adsorbing the DNP, then especially with those bran preparations with a high starch content, we would have expected the dietary fibre preparations to have adsorbed significantly more DNP than the equivalent unextracted bran. This suggests that other components, probably mostly starch, also adsorbs DNP in the unextracted brans. Indeed, we have found that granular potato starch adsorbed DNP significantly unpublished data, this laboratory.. It is interesting that rice and wheat brans, which showed most DNP adsorption, were also the brans that gave protection against colorectal cancer in rat experiments w1,2x. In contrast, oat and barley bran, which showed the least DNP adsorption, were not protective w35x. Wheat bran has also been shown in two human intervention studies to have a protective effect on the occurrence and size of adenomatous polyps w24,25x.

330

P.J. Harris et al.r Mutation Research 412 (1998) 323331

We have previously shown in vitro that solublefibre polysaccharides, including 1 3,1 4.-bglucans, maintained DNP in aqueous solutions and decreased its adsorption to insoluble cell wall components w26x. Moreover, many soluble-fibre polysaccharides enhance carcinogenesis in a carcinogen fed rat model w27x. One reason we considered for the poor adsorptive ability of oat and barley bran was that our chemical analyses indicated the presence in these brans of water-soluble 1 3,1 4.-bglucans. Oat and barley grains are known to be rich sources of 1 3,1 4.-b-glucans w6x. However, our results showed that the insoluble dietary fibre preparations from oats and barley brans did not adsorb DNP more readily than the equivalent total dietary fibre preparations. Our previous studies showed that lignified cell walls strongly adsorbed DNP w11x. Wheat bran is known to contain lignified cell walls; the walls of the outer and inner pericarp gave a positive histochemical reaction for lignin with phloroglucinol-HCl w28x. These lignified cell walls may thus account for the superior adsorption by wheat bran of DNP as well as MeIQx. We have no information about whether the cell walls of brans of other cereal species are lignified or not. Thus, we do not know if lignification of the cell walls of the bran layers accounts for the differences in adsorption of DNP among the brans of different cereal species. The lignified cell walls of the outer and inner pericarp tissue of wheat bran have been found to be very resistant to degradation in experiments in which wheat bran was incubated in vitro with human faecal bacteria w29x. These lignified pericarp cell walls were recovered intact in the faeces of volunteers fed wheat bran w30x. Furthermore, Ryden and Robertson w22x found that the ability of a dietary fibre preparation from a coarse wheat bran had increased ability to adsorb MeIQx after incubation in vitro with human faecal bacteria. Thus, wheat bran may protect against colorectal cancer by these lignified cell walls adsorbing carcinogens or promoters in the gastrointestinal tract and carrying these compounds out of the body on the undegraded cell walls. The gut mucosal cells will thus be exposed to lower concentrations of carcinogens or promoters. In conclusion, we have established that brans from different cereal species differ in their abilities

to adsorb the hydrophobic mutagen and carcinogen DNP. These results may be related to the different abilities of brans from different cereal species to protect against colorectal cancer. Our results emphasize the need in epidemiological experiments to identify the cereal species from which brans were obtained; simply indicating the amount of bran consumed is insufficient.

Acknowledgements We are grateful to Mr. Mark E. Watson and Ms. Margrit Martin for technical assistance. The work was financially supported by the Provisional Grand Rights of New Zealand and the Auckland Cancer Society.

References
w1x K. Watanabe, B.S. Reddy, J.H. Weisberger, D. Kritchevsky, Effects of alfalfa, pectin and wheat bran on azomethane or methylnitrosourea-induced colon carcinogenesis in F-344 rats, J. Natl. Cancer Inst. 63 1979. 141145. w2x W.S. Barnes, J. Maiello, J.H. Weisberger, In vitro binding of the food mutagen 2-amino-3-methylimidazo-w4,5-f xquinoline to dietary fibers, J. Natl. Cancer Inst. 70 1983. 757760. w3x L.R. Jacobs, J.R. Lupton, Relationship between colonic luminal pH, cell proliferation, and colon carcinogenesis in 1,2-dimethylhydrazine treated rats fed high fiber diets, Cancer Res. 46 1986. 17271734. w4x A. McIntyre, P.R. Gibson, G.P. Young, Butyrate production from dietary fibre and protection against large bowel cancer in a rat model, Gut 34 1993. 386391. w5x G.H. McIntosh, L. Jorgensen, P. Royle, The potential of an insoluble dietary fiber rich source from barley to protect from DMH-induced intestinal tumors in rats, Nutr. Cancer 19 1993. 213221. w6x B.A. Stone, Cereal grain carbohydrates, in: R.J. Henry, P.S. Kettlewell Eds.., Cereal Grain Quality, Chapman & Hall, London, 1996, pp. 251288. w7x E.F. Caldwell, M. Dahl, R.B. Fast, S.E. Seibert, Hot cereals, in: R.B. Fast, E.F. Caldwell Eds.., Breakfast Cereals and How They Are Made, American Association of Cereal Chemists, St. Paul, MN, 1990, pp. 243272. w8x N.L. Kent, A.D. Evers, Technology of Cereals: An Introduction for Food Science and Agriculture, 4th edn., Pergamon, Oxford, UK, 1994. w9x L.R. Ferguson, P.J. Harris, Particle size of wheat bran in relation to colonic function in rats, Food Sci. Technol. 30 1997. 735742.

P.J. Harris et al.r Mutation Research 412 (1998) 323331 w10x P.J. Harris, A.M. Roberton, H.J. Hollands, L.R. Ferguson, Adsorption of a hydrophobic mutagen to dietary fibre from the skin and flesh of potato tubers, Mutation Res. 260 1991. 203213. w11x A.M. Roberton, L.R. Ferguson, H.J. Hollands, P.J. Harris, Adsorption of a hydrophobic mutagen to five contrasting dietary fiber preparations, Mutat. Res. 262 1991. 195202. w12x L.R. Ferguson, A.M. Roberton, R.J. MacKenzie, M.E. Watson, P.J. Harris, Adsorption of a hydrophobic mutagen to dietary fiber from taro Colocasia esculenta., an important food plant of the South Pacific, Nutr. Cancer 17 1992. 8595. w13x F.J. De Serres, T. Matsushima, Mutagenesis and carcinogenesis by nitropyrenes and cancer chemotherapeutics, Mutation Res. 164 1986. 38. w14x AACC, American Association of Cereal Chemists Approved Methods, 8th edn., The Association, St. Paul, MN, USA, 1983. w15x L. Prosky, N.-G. Asp, I. Furda, J.W. de Vries, T.F. Schweizer, B.F. Harland, Determination of total dietary fiber in foods and food products: collaborative study, J. Assoc. Off. Anal. Chem. 68 1985. 677679. w16x L. Prosky, N.-G. Asp, T.F. Schweizer, J.W. deVries, I. Furda, Determination of insoluble, soluble, and total dietary fiber in foods and food products: interlaboratory study, J. Assoc. Off. Anal. Chem. 71 1988. 10171023. w17x I.L. Batey, Starch analysis using thermostable alphaamylases, Starch 34 1982. 125128. w18x P. Aman, K. Hesselman, Analysis of starch and other main constituents of cereal grains, Swedish J. Agric. Res. 14 1984. 135140. w19x A.B. Blakeney, N.K. Matheson, Some properties of the stem and rice pollen starches of rice, Starch 36 1984. 265269. w20x B.V. McCleary, R. Codd, Measurement of 13., 14.b-D-glucan in barley and oats: a streamlined enzymic procedure, J. Sci. Food Agric. 55 1991. 303312. w21x D.M. Maron, B.N. Ames, Revised methods for the Salmonella mutagenicity test, Mutation Res. 113 1983. 173215. w22x P. Ryden, J.A. Robertson, The effects of fibre source and

331

w23x

w24x

w25x

w26x

w27x

w28x

w29x

w30x

fermentation on the apparent hydrophobic binding properties of wheat bran preparations for the mutagen 2-amino-3,8-dimethylimidazow4,5-f xquinoxaline MeIQx., Carcinogenesis 16 1995. 209216. P. Ryden, J.A. Robertson, The consequences of fruit and vegetable fibre fermentation on their binding capacity for MeIQx and the effects of soluble fibre sources on the binding affinity of wheat bran preparations, Carcinogenesis 16 1995. 17111716. J.J. De Cosse, H.H. Miller, M.L. Lesser, Effects of wheat fiber and vitamins C and E on rectal polyps in patients with familial adenomatous polyposis, J. Natl. Cancer Inst. 81 1989. 12901297. R. MacLennan, F. MacRae, C. Bain, D. Battistutta, P. Chapuis, H. Gratten, J. Lambert, R.C. Newland, M. Ngu, A. Russell, Randomized trial of intake of fat, fiber, and beta carotene to prevent colorectal adenomas. The Australian Polyp Prevention Project, J. Natl. Cancer Inst. 87 1995. 17601766. P.J. Harris, A.M. Roberton, M.E. Watson, C.M. Triggs, L.R. Ferguson, The effects of soluble-fiber polysaccharides on the adsorption of a hydrophobic carcinogen to an insoluble dietary fiber, Nutr. Cancer 19 1993. 4354. L.R. Jacobs, Influence of soluble fibers on experimental colon carcinogenesis, in: D. Kritchevsky, C. Bonfield, J.W. Anderson Eds.., Dietary Fiber Chemistry, Physiology, and Health Effects, Plenum, New York, 1990, pp. 389401. P.B. Schwarz, W.H. Kunerth, V.L. Youngs, The distribution of lignin and other fiber components within hard red spring wheat, Cereal Chem. 65 1988. 5964. B.J.H. Stevens, R.R. Selvendran, C.E. Bayliss, R. Turner, Degradation of cell wall material of apple and wheat bran by human faecal bacteria in vitro, J. Sci. Food Agric. 44 1988. 151156. F.R. Dintzis, L.M. Legg, W.L. Deatherage, F.L. Baker, G.E. Inglett, R.A. Jacob, S.J. Reck, J.M. Munoz, L.M. Klevay, H.H. Sandstead, W.C. Shuey, Human gastrointestinal action on wheat, corn, and soy hull branpreliminary findings, Cereal Chem. 56 1979. 123127.