Prevention of heterocyclic amine-induced DNA damage in colon and liver of ratsby different lactobacillus strains

Markus Zsivkovits, Kassie Fekadu, Gerhard Sontag1,Ursula Nabinger1, Wolfgang W.Huber, Michael Kundi2,Asima Chakraborty, Helmuth Foissy3 and SiegfriedKnasmuller4

Institute of Cancer Research, Borschkegasse 8a, A-1090 Vienna, 1Institute ofAnalytical Chemistry, Wahringer Strasse 38, A-1090 Vienna, 2Institute ofEnvironmental Hygiene, Kinderspitalgasse 15, A-1090 Vienna and3Department of Dairy Research and Microbiology, Agricultural University,Gregor Mendel Strasse 33, A-1180 Vienna, Austria

4To whom correspondence should be addressedEmail: siegfried.knasmueller@univie.ac.at

The aim of the present study was to investigate the impactof four different lactobacillus (LB) strains, namely Lacto-bacillus bulgaricus 291, Streptococcus thermophilus F4,S.thermophilus V3 and Bifidobacterium longum BB536,which are used for the production of yogurt, on the DNA-damaging effects of heterocyclic aromatic amines (HCAs).Male F344 rats were treated orally with HCA mix-tures containing 2-amino-1-methyl-6-phenylimidazo[4,5-b]-pyridine, 2-amino-3,8-dimethylimidazo[4,5-f ]quinoxaline,2-amino-3,4,8-trimethyl-3H-imidazo[4,5-f ]quinoxaline, 2-amino-9H-pyrido[2,3-b]indole and 2-amino-3-methyl-3H-imidazo[4,5-f ]quinoline, which were representative of theHCA contents found in fried beef (`beef mix') and chicken(`chicken mix'). Suspensions of LB were given by gavage tothe animals simultaneously with and at different time per-iods before administration of the HCAs. Subsequently, theextent of DNA migration was measured in colon and livercells in single cell gel electrophoresis (SCGE) assays. All fourstrains caused complete inhibition of DNA damage inducedwith beef mix after administration of 1 1010 LB cells/animal, whereas with chicken mix only marginal (non-significant) effects were seen. The inhibition of beef-induced DNA damage was dose dependent and was stillsignificant when 1 107 cells/animal were administered.Kinetics studies showed that the protective effects were stillsignificant when LB was given 12 h before the beef mix. Acomparison of the present results with chemical analyticaldata from in vitro experiments suggests that the strongreduction in DNA migration seen in the animals can beonly partly explained by direct binding effects. The resultsof the present study show that LB are highly protectiveagainst the genotoxic effects of HCAs under conditionswhich are relevant for humans and provide a possibleexplanation for the reduced colon cancer rates observedin some studies in individuals with either high LB counts in

their feces or with a high consumption of LB-containingfoods.

Introduction

Evidence is accumulating that heterocyclic aromatic amines(HCAs), which are formed from amino acids in meats duringcooking, might be involved in the etiology of human cancer(1,2). As a consequence, strong efforts have been made toidentify dietary factors which are protective towards HCAs.Approximately 600 complex dietary mixtures and individualcompounds have been investigated for DNA-protective andanticarcinogenic properties towards HCAs (reviewed in 3,4).Most of the data come from in vitro studies with bacteria andonly a few results from in vivo experiments with laboratoryrodents are available.

Of particular interest are dietary factors for which someevidence is available that their consumption is inverselyrelated to the incidence of colon cancer in humans. Typicalexamples are fibers (5), cruciferous vegetables (6) and lacticacid bacteria (7). It has been shown in a number of in vitromutagenicity experiments that lactobacilli prevent inductionof DNA damage and mutations by HCAs (reviewed in 8).Chemical analyses indicated that these protective effects are(at least partly) due to direct binding of the amines to cell wallcomponents of the bacteria (9,10). Only a few investigationson the effects of lactobacilli towards HCAs in laboratoryrodents have been published. Reddy and Rivenson (11) foundthat supplementation of the chow with 0.5% lyophilizedBifidobacterium longum caused pronounced inhibition of theincidence of 3-amino-1-methyl-5H-pyrido[4,3-6]indole (Trp-P-2)-induced liver and colon tumors (i.e. 80 and 100% reduction intumor frequencies, respectively), however, the concentrationof tryptophan pyrolysates in meats are much lower than thoseof other HCAs (12). More recently, Tavan et al. (13) reportedthat fermented milks containing either Streptococcus thermo-philus or Bifidobacterium animalis inhibited the induction ofpreneoplastic lesions [aberrant crypt foci (ACF) in the colon]and DNA migration caused by a HCA mixture containing2-amino-3-methyl-3H-imidazo[4,5-f ]quinoline (IQ), 2-amino-3,4-dimethyl-3H-imidazo[4,5-f]quinoline (MeIQ) and 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) (ratio 1:1:1).However, since not fermented milk but also unfermentedmilk elicited protective effects similar to those seen withfermented milk, the contribution of the bacteria to the protec-tive effects remains unclear. Furthermore, the uptake of radi-olabeled tryptophan-derived HCAs in the gut was reducedwhen the animals were treated simultaneously with lactobacilli(9,10,14).

The aim of the present study was to investigate the potentialprotective effects of four different lactobacillus strains cur-rently used for the production of yogurt under conditionswhich are relevant for humans. The experiments were carried

Abbreviations: ACF, aberrant crypt foci; AaC, 2-amino-9H-pyrido[2,3-b]indole; DiMeIQx, 2-amino-3,4,8-trimethyl-3H-imidazo[4,5-f ] quinoxaline;HCAs, heterocyclic aromatic amines; IQ, 2-amino-3-methyl-3H-imidazo[4,5-f ]quinoline; MeIQx, 2-amino-3,8-dimethylimidazo[4,5-f ]quinoxaline;PhIP, 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine; SCGE, single cellgel electrophoresis.

1913

Carcinogenesis vol.24 no.12 pp.19131918, 2003DOI: 10.1093/carcin/bgg167

out with mixtures of HCAs whose composition is character-istic of fried beef and chicken (15). The different strains oflactobacilli were administered to F344 rats either simulta-neously with or at different time periods before the HCAs.The protective effects were studied in liver and colon; theseorgans are the major targets for tumor induction by HCAs (16).In addition, the dose and time kinetics of the protective effectswere also studied. As an end-point we measured comet for-mation with the single cell gel electrophoresis (SCGE) assay,which is based on the determination of DNA migration in anelectric field. We showed earlier that induction of cometformation is an appropriate method to monitor DNA damage byHCAs in various inner organs of laboratory rodents (1719).We also used this approach successfully in antigenotoxicitystudies (20,21) and showed that prevention of HCA-inducedDNA migration by dietary compounds in colon and liver cellsof rats is paralleled by a reduction in the formation of preneo-plastic lesions in these organs (18,19,22,23).

Materials and methods

Chemicals and media

The HCAs were purchased from the Nard Institute Ltd (Amagasaki, Japan).Beef mix was composed as described by Keating and Bogen (15) andcontained PhIP (57.0%), 2-amino-3,8-dimethylimidazo[4,5-f ]quinoxaline(MeIQx) (26.3%), 2-amino-3,4,8-trimethyl-3H-imidazo[4,5-f ]quinoxaline(DiMeIQx) (4.7%), 2-amino-9H-pyrido[2,3-b]indole (AaC) (10.9%) and IQ(1.17%). Chicken mix was also prepared according to Keating and Bogen (15)and contained the same amines except for IQ (PhIP, 79.0%; MeIQx, 1.5%;DiMeIQx, 0.5%; AaC, 19.0%). For the animal experiments, these mixtureswere suspended in physiological saline and stored at 20C. Proteinase K,(R,R)-dithiothreitol, ethidium bromide, Triton X-100 and Tris were purchasedfrom Sigma Chemical Co. (St Louis, MO). Low melting point agarose andnormal melting point agarose were obtained from Gibco (Paisley, UK).Inorganic salts were from Merck (Darmstadt, Germany). Media for the deter-mination of the viability of the lactobacilli (MRS Agar, Reinforced ClostridalAgar and M17 Agar) were purchased from Oxoid Ltd (Basingstoke, UK).

Lactic acid bacteria

The lactobacillus strains (Lactobacillus bulgaricus 291, S.thermophilus F4,S.thermophilus V3 and B.longum BB536) were obtained from Danisco CultorGmbH (Niebull, Germany). The cells were stored deep frozen at 70C asconcentrated cultures. All strains are currently used for yogurt production.Immediately before the experiments, the cultures were defrosted in warmwater (40C) and suspended in sterile physiological saline. The viability ofthe cells was determined by plating (24): MRS agar was used for L.bulgaricus291, M17 agar for the two S.thermophilus strains and Reinforced ClostridalAgar for B.longum BB536 (25). Bifidobacterium plates were incubated in ananaerobic jar (Oxoid Ltd). To create anaerobic conditions, AnaerocultTM A(Merck) was used. All plates were incubated at 37C for 5 days in the dark.Subsequently the colonies were counted manually.

Animals and treatment

Male Fischer 344 rats (body weight 150 25 g) were purchased from Harlan-Winkelmann GmbH (Borchen, Germany) and housed in standard cages(2 animals/cage). Throughout the study, the animals were kept at constanttemperature and humidity (24 2C, 50 5%) with a 12 h light/dark cycle.

After 1 week of acclimatization, the animals were deprived of feed 16 hbefore oral treatment with the carcinogens (100 mg/kg body wt), but receiveddrinking water ad libitum. Negative controls received sterile physiologicalsaline only. The dose levels of and the exposure time to the HCA mixtureswere chosen on the basis of the results of previous experiments with HCAs(26). Freshly prepared suspensions of the different lactobacillus strains (2 ml/animal) were given by gavage either simultaneously with or (412 h) prior tothe carcinogens. The animals were killed 4 h after administration of thecarcinogens and liver and colon cells were isolated for the SCGE experimentsas described by Bradley et al. (27) and Brendler et al. (28), respectively.Isolated cells were resuspended in 50 and 20 ml of medium for liver andcolon cells, respectively, and cell viability was determined by the trypan bluemethod (29). Cells were used for the comet assays only when their viability was

80% (30). The total cell number was in the range 915 106 hepatocytes/mland 25 106 colonocytes/ml. Three animals were used per experimental point.

Single cell gel electrophoresis (SCGE) assay in vivo

The SCGE assay was performed according to Kassie et al. (18). DNA damagewas determined by measuring the comet tail lengths of the cells with afluorescence microscope (Nikon EFD-3, 125 magnification) connected to amonitor with a specific macro for the NIH public domain image analysisprogram (31). Three slides were prepared from each organ (18) and 50 cellswere analyzed from each slide (in total 150 cells/organ).

Statistical analysis

For SCGE assays, the distribution pattern of the tail lengths in the differenttreatment groups was analyzed as suggested by Tice et al. (30) and fitted to ag-distribution. The means of the tail lengths were calculated for each slide;means standard deviation and statistics were computed using results from3 animals/group. Comparisons of groups of animals were done by ANOVAbased on the means of three slides. Post hoc comparisons between experi-mental groups and control animals were done by Dunnett's test. For all testsP 0.05 was considered significant.

Results

The extent of DNA damage caused by beef mix and chickenmix in initial experiments in livers and colons is shown inFigure 1. Beef mix (Figure 1A and B) caused pronouncedeffects in both organs. DNA migration was far more pro-nounced in liver cells than in colonocytes. Figure 1C and Ddepicts the effect caused by chicken mix in colon cells underidentical experimental conditions. In this case only moderateDNA migration was seen (Figure 1C); the effect in the liverwas not significant (Figure 1D).

Figure 2A and C shows the impact of the different lacto-bacillus strains on DNA migration induced by either chickenmix (Figure 2A) in the colon or by beef mix in the colon(Figure 2B) and liver (Figure 2C). In all these experimentalseries, 1 1010 viable bacteria were administered per animal.With chicken mix, no significant decrease in DNA damagewas observed, however, with L.bulgaricus 291 a slight (~15%)

Fig. 1. Induction of DNA migration in colon cells and hepatocytes of F344rats by HCA mixtures representing fried beef (A and B) and fried chicken(C and D). The mixtures were composed according to Keating and Bogen (15).The animals were treated orally with the HCA mixtures. The total amountadministered was 100 mg/kg body wt. Four hours later, the animals were killedand the cells isolated from colons and livers and analyzed for comet formation.Bars indicate means SD of tail lengths measured with nine slides (3 animals/treatment group, 3 slides/organ, 50 cells/slide); numbers indicate thecorresponding tail moments. Ctrl, control animals; Mix, HCA mix (chickenmix or beef mix). Stars indicate statistical significance compared withuntreated controls, P 0.05.

M.Zsivkovits et al.

1914

decline in DNA migration was seen. In contrast, pronouncedprotection was found in all experiments with beef mix underidentical conditions. All four strains attenuated HCA-inducedDNA damage in both organs more or less completely.

Figure 3 shows the dose dependency of the protective effectof L.bulgaricus 291. The cells were administered in the con-centration range 1 1071 1010 viable cells/animal. It canbe seen in Figure 3A and B that the inhibition of DNA damagedepends on the number of bacteria. Administration of 1 108cells/animal led to an almost complete inhibition of HCA-induced DNA damage. With 1 107 cells/animal, DNAmigration was reduced by ~75% in colonocytes and by 89%in hepatocytes. Notably, the extent of comet formation inanimals which had been treated with 1 1011 cells/animalwas lower than that seen in untreated controls in both organs.In the liver, DNA damage was reduced by ~50% and in coloncells by 30%. A significant protective effect was also observedin both organs with B.longum at a low dose (107 cells/animal),which was similar to that measured with L.bulgaricus 291(data not shown).

Figure 4 depicts the time kinetics of the protective effects.These experiments were carried out with L.bulgaricus 291.

The animals received the bacteria either simultaneously withthe beef mix or at different time periods before the carcino-gens. Up to 4 h, complete protection was observed. When thebacteria were given 12 h prior to the HCAs, the reduction inDNA migration was still significant in both organs (69% in theliver and 45% in the colon). However, upon administration ofthe bacteria 24 h before the beef mix, no significant antigeno-toxic effects were detectable in either organ.

Discussion

The present findings are the first which give clear evidence forDNA-protective effects of lactobacilli used for yogurt produc-tion against DNA damage caused by HCAs in organs whichare targets of tumor induction by these compounds. In earlierinvestigations it was shown that lactobacilli reduce chemicallyinduced DNA migration and precarcinogenic lesions incolon cells, but in these experiments model compoundswhich are not present in the human diet were used, such as1-methyl-3-nitro-1-nitrosoguanidine, 1,2-dimethylhydrazine,N-methyl-N-nitrosourea and azoxymethane (32,33). Tavan

Fig. 2. Effect of different lactobacillus strains on DNA damage induced by chicken mix in colon cells (A) or by beef mix in colon cells (B) and hepatocytes (C).The animals were treated orally with suspensions of the individual lactobacillus strains (1 1010 viable cells/animal suspended in 2 ml of physiological saline).Subsequently, the rats received the HCA mixtures by gavage (100 mg/kg body wt). Four hours later, comet induction was measured as described in Materials andmethods. Bars indicate means SD of tail lengths measured with nine slides (3 animals/treatment group, 3 slides/organ, 50 cells/slide); numbers indicate thecorresponding tail moments. Ctrl, control animals; Mix, DNA migration in animals which were treated with the HCA mix (chicken mix or beef mix) alone;B.I., combined treatment (mix and B.longum BB536); S.t.1, combined treatment (mix and S.thermophilus F4); S.t.2, combined treatment (mix and S.thermophilusV3); L.b., combined treatment (mix and L.bulgaricus 291). Stars indicate statistical significance compared with HCA-treated animals, P 0.05.

Fig. 3. Effect of administration of different doses of L.bulgaricus 291 on the extent of DNA migration caused by beef mix in colon (A) and liver cells (B).The experiment was carried out essentially as described in the legend to Figure 2, but different amounts of bacteria were administered immediately beforeadministration of the HCA mix. The x-axis indicates the numbers of viable cells administered per animal. Bars indicate means SD of tail lengths measuredwith nine slides (3 animals/treatment group, 3 slides/organ, 50 cells/slide); numbers indicate the corresponding tail moments. Ctrl, control animals; Mix,HCA mix (beef mix) alone; hatched columns, combined treatment (beef mix and L.bulgaricus 291). Stars indicate statistical significance compared withHCA-treated animals, P 0.05.

Prevention of DNA damage by lactobacilli

1915

et al. (13) also used the Comet technique to study potentialprotective effects of fermented milk, but the interpretation ofthe results is not conclusive. Although a reduction in cells withcomets was observed, this effect was not paralleled by adecrease in undamaged cells. Furthermore, a decrease in HCA-induced DNA migration was also seen with unfermented milkin this study. Therefore, it is not clear if and to what extentlactobacilli contributed to the protective effects.

In the present experiments we used HCA mixtures whichreflect the composition of these compounds in fried beef andfried chicken. It is likely that the weak induction of DNAdamage by the chicken mix (Figure 1C and D) in both organsis due to its high content of PhIP, which causes only marginaleffects in the Comet assay in rats (34). In contrast, pronouncedinduction of DNA migration was seen with beef mix, whichcontains higher amounts of quinolines and quinoxalines suchas IQ, which is a potent inducer of comets in liver and coloncells of laboratory mice and rats (1719). Pronounced preven-tion of comet formation was seen with different lactobacillusstrains (Figure 2A and C) only with the beef mix, not with thechicken mix. This suggests that the protective effects of lacto-bacilli depend on the composition of the HCAs in the food.

Our findings indicate that a reduction in DNA migration bylactobacilli takes place under conditions which are relevant forhumans. With the lowest dose used in the present experiments(107 cells/animal; see Figure 3A and B), highly significantprotective effects were still seen with L.bulgaricus 291 andB.longum BB536 in both organs. This is a new observation, asin other in vivo studies with HCAs (11,13,14) higher concen-trations of lactobacilli were given to the animals, whichexceeded the daily consumption levels in humans. Assumingan equal distribution of the administered bacteria in the gastro-intestinal tract (1025 ml total volume), the administra-tion of 107 cells/animal will lead to an average concentrationof 0.51 106 cells in the digestive tract of the animals.Consumption of high quality yogurts (containing 108lactobacilli/ml) by humans (250 ml/day) will lead to aconcentration of ~106 cells/ml in the human gastrointestinal

tract (estimated total volume 25 l). In this context, it is notablethat several commercial yogurts contain considerably lowernumbers of lactobacilli (i.e. in the range 105107/ml) (for areview see 35).

Our data also show that the reduction in DNA damageoccurs not only when the lactobacilli are administered simul-taneously with the cooked food mutagens but also when thebacteria are given up to 12 h before the HCAs (Figure 4A and B).This observation suggests that protective effects can alsobe expected in humans when lactobacillus-containing diaryproducts are consumed several hours prior to fried meats.

It has been proposed that the antimutagenic effects of lacto-bacilli against HCAs which have been found in a number ofearlier in vitro mutagenicity studies (36) are due to directbinding of the amines (reviewed in 37). In earlier chemicalanalytical studies, in which individual HCAs were incubatedwith lactobacillus strains in vitro, maximal binding with thecompounds contained in the beef mix was in the range4060%, with stronger binding (8095%) only being observedfor tryptophan derivatives (14,37). In vivo studies withtryptophan-derived HCAs showed that lactobacilli preventtheir absorption; oral administration of different strains torats reduced the blood levels of 3-amino-1,4-dimethyl-5H-pyrido[4,3-6]indole (Trp-P-1) in the range 4062% (10); inanother study with radiolabeled Trp-P-2, a significant reduc-tion in radioactivity was seen in different inner organs of micewhen the animals received Lactobacillus acidophilus orB.longum (14). The latter compounds were not contained inthe presently used HCA mixtures as they were not detectedin fried meats in a number of studies (12). In a preliminarystudy aimed at developing a reliable HPLC protocol withcoulometric electrode array detection, the same HCA mixand dose was used as in the present study and a rat was treatedsimultaneously with 109 cells of L.bulgaricus 291 (42). Simul-taneous administration of the bacteria led to a substantialdecrease in the urinary excretion of the different amines, thestrongest effect being seen with MeIQx and DiMeIQx (40 and47%, respectively). The respective values for IQ, PhIP and

Fig. 4. Effect of the administration time on the protective effects of L.bulgaricus 291 against DNA migration caused by beef mix. The animals were treatedwith a suspension (2.0 ml) of L.bulgaricus 291 cells (1 1010 viable cells/animal in physiological saline) for different time periods before they received the HCAmix (100 mg/kg body wt). Four hours after administration of the beef mix, the animals were killed and comet formation was measured in colon cells (A) andhepatocytes (B). The x-axis indicates the time of administration of the L.bulgaricus 291 (0, simultaneous administration with the HCA mix; 4, 12 and 24, hours ofL.bulgaricus 291 administration before the HCA mix). Bars indicate means SD of tail lengths measured with nine slides (3 animals/treatment group, 3 slides/organ, 50 cells/slide); numbers indicate the corresponding tail moments. Ctrl, control animals; Mix, HCA mix (beef mix) alone; hatched columns, combinedtreatment (beef mix and L.bulgaricus 291). Stars indicate statistical significance compared with HCA-treated animals, P 0.05.

M.Zsivkovits et al.

1916

AaC were 30, 21 and 4%. A similar result was obtained in anexperiment with B.longum BB536. These observations suggestthat lactobacilli reduce in particular the bioavailablility ofquioxalines. In this context it is also notable that some evi-dence exists that lactobacilli reduce the bioavailablity of HCAsin humans as consumption of strains used for yogurt produc-tion led to a significant reduction in urinary mutagenicityinduced by consumption of fried beef (43,44).

Persistence of the protective effect over several hours(Figure 4A and B) indicates that, apart from direct bindingeffects which were found in a number of in vitro experimentswith various lactobacilli, including the strains used in thepresent experiments (37,42), additional indirect mechanismsmay play a role. It has been shown that lactobacilli adhere tointestinal mucosa cells in vitro and in vivo (3840,45) and it isconceivable that this feature affects the uptake of HCAsthrough the intestinal barrier.

Not much is known about the metabolism of HCAs bylactobacilli. It has been reported that living lactobacilli reducethe mutagenic activity of the amines to a greater extent thanheat-inactivated cells (for a review see 37). This observationindicates that the compounds are metabolically detoxified andit was postulated that the amines react with short chain organicacids produced by the bacteria (46). In a recent study, theformation of 7-hydroxy-IQ by different representatives ofthe human intestinal microflora was investigated and no evi-dence for the formation of this metabolite by lactobacilli wasfound, neither were any other metabolites of this aminedetected after in vitro incubation with lactobacillus strains(46; C.Humblot, personal communication).

The antigenotoxic effects seen in the present experimentswith lactobacilli are stronger than those observed in earlierSCGE studies with rats in which IQ was used as a model HCA.The maximal effect which we found with cruciferous vege-tables and prebiotics such as inulin and oligofructose didnot exceed 50% of the chemically induced DNA migration(18,22,23). In the case of the cruciferous vegetables, it is likelythat induction of detoxifying phase II enzymes accounts for theprotective effects. That this mechanism accounts for the pro-tective effects of the lactobacilli can be excluded. In the pre-sent study a strong decrease in HCA-induced DNA migrationwas seen 4 h after administration of the bacteria and manystudies on the time kinetics of enzyme induction in rodentsshow that neither the activities of enzymes involved in thedetoxification of HCA, such as glutathione S-transferase andglucuronosyltransferase, nor their mRNA expression isincreased within this short time period (see for example4750). The inhibition of activating phase I enzymes, whichis an important mechanism of chemoprotection against HCAs(4), seems to play no or only a minor role. Tavan et al. (13) fed5 108 lactobacilli (either B.animalis or S.thermophilus) torats and found no change in EROD activities in liver and colon;MROD activity was not altered in the colon, whereas a slight(1520%) decrease was seen in hepatic tissue.

We found in earlier animal experiments with cruciferousvegetables that the prevention of HCA-induced DNA migra-tion in colon and liver measured in SCGE assays is associatedwith a decrease in the frequency of preneoplastic lesions(GSTP foci in the liver and ACF in the colon) in laboratoryanimals (18,19,23). This indicates that the prevention of DNAdamage as seen in the present experiments leads to an inhibi-tion of HCA-induced cancer formation in liver and colon.Since the lactobacilli caused pronounced effects under

conditions which are relevant for humans in the presentstudy, our results provide a possible explanation for the inverseassociation between consumption of yogurt, excretion of highamounts of lactobacilli and the incidence of colon cancer inhumans which has been found in some epidemiological studies(32,51).

Acknowledgements

The authors are grateful to Dagmar Riedler and Dr Karl Mazzucco for theirhelp in the realization of the animal experiments and thank Bernd Haschke(Molkerei Alois Muller GmbH & Co, Aretsried, Germany) for providing thebacterial strains. The experimental work described in this article was realizedwithin the frame of a coordinated EU project (HC-AMINES to S.K.; QLK-1-1999-01197).

References

1. Augustsson,K. and Steineck,G. (2000) Cancer risk based on epidemiolo-gical studies, In Nagao,M. and Sugimura,T. (eds), Food BorneCarcinogenesis: Heterocyclic Amines. Wiley, New York, NY,pp. 332347.

2. Sugimura,T., Nagao,M. and Wakabayashi,K. (2000) Complex factorspertinent to human hazard risks. In Nagao,M. and Sugimura,T. (eds), FoodBorne Carcinogenesis: Heterocyclic Amines. Wiley, New York, NY,pp. 349359.

3. Dashwood,L.H. (2002) Modulation of heterocyclic amine-induced muta-genicity and carcinogenicity: an `A-to-Z' guide to chemopreventiveagents, promoters and transgenic models. Mutat. Res., 511, 89112.

4. Schwab,C.E., Huber,W.W., Parzefall,W., Hietsch,G., Kassie,F., Schulte-Hermann,R. and Knasmuller,S. (2000) Search for compounds that inhibitthe genotoxic and carcinogenic effects of heterocyclic aromatic amines.Crit. Rev. Toxicol., 30, 169.

5. Jacobs,D.R., Slavin,J. and Marquart,L. (1995) Whole grain intake andcancer: a review of the literature. Nutr. Cancer, 24, 221229.

6. Steinmetz,K.A. and Potter,J.D. (1996) Vegetables, fruit and cancerprevention: a review. J. Am. Diet. Assoc., 96, 10271039.

7. Wollowski,I., Rechkemmer,G. and Pool-Zobel,B.L. (2001) Protectiveeffects of probiotics and prebiotics in colon cancer. Am. J. Clin. Nutr.,73 (suppl.), 451S455S.

8. Knasmuller,S., Steinkellner.,H, Majer,B.J., Nobis,E.C., Scharf,G. andKassie,F. (2002) Search for dietary antimutagens and anticarcinogens:methodological aspects and extrapolation problems. Food Chem. Toxicol.,40, 10511062.

9. Zhang,X.B. and Ohta,Y. (1991) Binding of mutagens by fractions of thecell wall skeleton of lactic acid bacteria on mutagens. J. Dairy Sci., 74,14771481.

10.Zhang,X.B. and Ohta,Y. (1993) Microorganisms in the gastrointestinaltract of the rat prevent absorption of the mutagen-carcinogen 3-amino-1,4-dimethyl-5H-pyrido[4,3-b]indole. Can. J. Microbiol., 39, 841845.

11.Reddy,B.S. and Rivenson A. (1993) Inhibitory effect of Bifidobacteriumlongum on colon, mammary and liver carcinogenesis induced by2-amino-3-methylimidazo[4,5-f ]quinoline a food mutagen. CancerRes., 53, 39143918.

12.Felton,J.S., Jagerstad,M., Knize,M.G., Skog,K. and Wakabayashi,K.(2000) Contents in foods, beverages and tobacco. Nagao,M. andSugimura,T. (eds), Food Borne Carcinogenesis: Heterocyclic Amines.Wiley, New York, NY, pp. 529.

13.Tavan,E., Cayuela,C., Antoine,J.M., Trugnan,G., Chaugier,C. andCassand,P. (2002) Effects of dairy products on heterocyclic aromaticamine-induced rat colon carcinogenesis. Carcinogenesis, 23, 477483.

14.Orrhage,K.M., Annas,A., Nord,C.E., Brittebo,E.B. and Rafter,J.J. (2002)Effects of lactic acid bacteria on the uptake and distribution of the foodmutagen Trp-P-2 in mice. Scand. J. Gastroenterol., 37, 215221.

15.Keating,G.A. and Bogen,K.T. (2001) Methods for estimating heterocyclicamine concentrations in cooked meats in the US diet. Food Chem.Toxicol., 39, 2943.

16.Ohgaki,H. (2000) Carcinogenicity in animals and specific organs,Rodents. In Nagao,M. and Sugimura,T. (eds), Food BorneCarcinogenesis: Heterocyclic Amines. Wiley, New York, NY,pp. 198228.

17.Sekihashi,K., Yamamoto,A., Matsumura,Y., Ueno,S., Watanabe-Akanuma,M., Kassie,F., Knasmuller,S., Tsuda,S. and Sasaki,Y.F. (2002)

Prevention of DNA damage by lactobacilli

1917

Comparative investigation of multiple organs of mice and rats in the cometassay. Mutat. Res., 517, 5375.

18.Kassie,F., Rabot,S., Uhl,M., Huber,W., Qin,H.M., Helma,C., Schulte-Hermann,R. and Knasmuller,S. (2002) Chemoprotective effect of gardencress (Lepidium sativum) and its constituents towards 2-amino-3-methyl-imidazo[4,5-f ]quinoline (IQ)-induced genotoxic effects andcolonic preneoplastic lesions. Carcinogenesis, 23, 11551161

19.Kassie,F., Uhl,M., Rabot,S., Grasl-Kraupp,B., Verkerk,R., Kundi,M.,Chabicovsky,M., Schulte-Hermann,R. and Knasmuller,S. (2003)Chemoprevention of 2-amino-3-methyl-imidazo[4,5-f ]quinoline (IQ)induced colonic and hepatic preneoplastic lesions in the F344 rat bycruciferous vegetables administered simultaneously with the carcino-gen. Carcinogenesis, 24, 255261.

20.Uhl,M., Darroudi,F., Seybel,A., Steinkellner,H., Kassie,F., Grasl-Kraupp,B., Mersch-Sundermann,V. and Knasmuller,S. (2001)Development and use of new experimental models for the identificationof DNA-protective and anticarcinogenic plant constituents. In Kreft,I.(ed.), Molecular and Genetic Interactions Involving Phytochemicals.COST-916 Action, Ljubljana, Slovenia, pp. 2135.

21.Kassie,F., Mersch Sundermann,V., Edenharder,R. et al. (2003) Devel-opment and application of test methods for the detection of dietaryconstituents which protect against heterocyclic aromatic amines.Mutat. Res., 523/524, 183192.

22.Uhl,M. (2002) Einfluss von Inhaltsstoffen von Brassicacen auf DNA-schadigende und kanzerogene Effekte von heterozyklischen aromatischenAminen. PhD thesis, University of Vienna.

23.Humblot,C., Kassie,F., Nugon-Baudon,L., Knasmuller,S. and Lhoste,E.F.(2003) Diet modulates the genotoxicity of IQ (2-amino-3-methylimidazo[4,5-f ]quinoline) in rats associated with a human faecal flora.J. Chromatogr., in press.

24.Bast,E. (1999) Bestimmung der Zellzahl und Zellmasse einzelligerMikroorganismen. In Bast,E. (ed.), Mikrobiologische Methoden: EineEinfuhrung in grundlegende Arbeitstechniken. Spektrum AkademischerVerlag, Heidelberg, Germany.

25.Gerten,B. (1993) Nahrboden. In Gerten,B. (ed.), OXOID Handbuch 5.Aktualisierte Auflage. R. Meyer, Dusseldorf, Germany.

26.Kassie,F., Rabot,S., Kundi,M. Chabicovsky,M., Qin,H.-M. andKnasmuller,S. (2001) Intestinal flora plays a crucial role in thegenotoxicity of the cooked food mutagen 2-amino-3-methylimidazo[4,5-f ]quinoline (IQ). Carcinogenesis, 22, 17211725.

27.Bradley,M.O., Dysart,G., Fitzsimmons,D., Harbach,P., Lewin,J. andWolf,G. (1982) Measurement by filter elution of DNA single and doublestrand breaks in rat hepatocytes: effects of nitrosamines and g-irradiation.Cancer Res., 42, 25922597.

28.Brendler,S., Tompa,A., Hutter,K.H., Preussmann,R. and Pool-Zobel,B.L.(1992) In vivo and in vitro gentoxicity of several N-nitrosamines inextrahepatic tissues of the rat. Carcinogenesis, 13, 25352441.

29.Lindl,T. and Bauer,J. (1994) Zell und Gewebekultur: Einfuhrung in dieGrundlagen sowie ausgewahlte Methoden und Anwendungen. GustavFischer Verlag, Stuttgart, Germany.

30.Tice,R.R., Agurell,E., Anderson,D. et al. (2000) Single cell gel/cometassay: guidlines for in vivo and in vitro genetic toxicology testing.Environ. Mol. Mutagen., 35, 206221.

31.Helma,C. and Uhl,M. (2000) A public domain image analysis program forthe single cell gel electrophoresis (comet) assay. Mutat. Res., 446, 915.

32.Wollowski,I., Ji,S.T., Bakalinsky,A.T., Neudecker,C. and Pool-Zobel,B.L.(2001) Bacteria used for the production of yoghurt inactivate carcinogensand prevent DNA-damage in the colon cells. J. Nutr., 129, 7782.

33.Bolognani,F., Rummey,C.J., Pool-Zobel,B.L. and Rowland,I.R. (2001)Effect of lactobacilli, bifidobacteria and inulin on the formation ofaberrant crypt foci in rats. Eur. J. Nutr., 40, 293300.

34.Zisvkovits,M (2002) Der Einfluss von Milchsaurebakterien auf dieGenotoxizitat von heterozyklischen aromatischen Aminen (HCAs).Thesis, Vienna, Austria.

35.Kailasapathy,K. and Chin,J. (2000) Survival and therapeutic potential ofprobiotic organisms with reference to Lactobacillus acidophilus andBifidobacterium spp. Immunol. Cell Biol., 78, 8088.

36.Orrhage,K., Sillerstrom,E., Gustafsson,J.-A., Nord,C.E. and Rafter,J.(1994) Binding of mutagenic heterocyclic amines by intestinal and lacticacid bacteria. Mutat. Res., 311, 239248.

37.Knasmuller,S., Steinkellner,H., Hirschl,M.A., Rabot,S., Nobis,E.C. andKassie,F. (2001) Impact of bacteria in dairy products and of the intestinalmicroflora on the genotoxic and carcinogenic effects of heterocyclicamines. Mutat. Res., 480, 129138.

38.Muscettola,M., Massi,L., Tanganelli,C. and Grasso,G. (1994) Effects oflactobacilli on interferon production in young and aged mice. Ann. N. Y.Acad. Sci., 717, 226232.

39.Coconnier,M.H., Klaenhammer,T.R., Kerneis,S., Bernet,M.F. andServin,A.L. (1992) Protein mediated adhesion of Lactobacillus acidophilusBG2FO4 on human enterocyte and mucus secreting cell lines inculture. Appl. Environ. Microbiol., 58, 20342039.

40.Bernet,M.F., Brassart,D., Nesser,J.R. and Servin,A.L (1994) Lactobacillusacidophilus LA 1 binds to cultured human intestinal cell lines andinhibits cell attachment and cell invasion by enterovirulent bacteria.Gut, 35, 483489.

41.Gerbl,U., Cichna,M., Zsivkovits,M., Knasmuller,S. and Sontag,G. (2003)Determination of heterocylcic aromatic amines in beef extract, cookedmeat and rat urine by HPLC with coulometric eleectrode array detection,J. Chromatogr., in press.

42.Hayatsu,H. and Hayatsu,T. (1993) Suppressing effect of Lactobacilluscasei administration on the urinary mutagenicity arising from friedground beef in the human. Cancer Lett., 73, 173179.

43.Lidbeck,A., Overvik,E., Rafter,J., Nard,C.E. and Gustafsson,J.A. (1992)Effects of Lactobacillus acidophilus supplements on mutagen excretionin feces and urine of humans. Microbiol. Environ. Health, 5, 5967.

44.Greene,J.D. and Klaenhammer,T.R. (1994) Factors involved in theadherence of lactobacilli to human Caco-2 cells. Appl. Environ.Microbiol., 60, 44874494.

45.Sreekumar,O. and Hsono,A. (1998) The heterocyclic amine bindingreceptors of Lactobacillus gasseri. Mutat. Res., 421, 6572.

46.Humblot,C., Combourieu,B., Vaisanen,M.-L. et al. (2003) Metabolism of2-amino-3-methyl-3H-imidazo[4,5-f]quinoline (IQ) by the humandigestive flora and by bacteria isolated from it. In 2nd InternationalWorkshop on Heterocyclic Aromatic Amines (abstract), Graz, Austria.

47.Lamb,J.G. and Franklin,M.R. (2000) Early events in the induction of rathepatic UDP-glucuronosyltransferases, glutathione S-transferase andmicrosomal epoxide hydrolase by 1,7-phenanthroline: comparison witholtipraz, tert-butylhydroxyanisole and tert-butylhydroquinone. Drug.Metab. Dispos., 28, 10181023.

48.Kim,S.G., Nam,S.Y., Kim,C.W., Kim,J.H., Cho,C.K. and Yoo,S.Y. (1997)Enhancement of radiation-inducible hepatic glutathione-S-transferasesYa,Yb1,Yc1 and Yc2 gene expression by oltipraz: possible role inchemoprotection. Mol. Pharmacol., 51, 225233.

49.Cho,M.C. and Kim,S.G. (1998) Differential induction of rat hepaticmicrosomal epoxide hydrolase and rGSTA2 by diazines: the role ofcytochrome P450 2E1-mediated metabolic activation. Chem. Biol.Interact., 116, 229245.

50.Siess,M.-H., Mas,J.-P., Canivence-Lavier,M.-C. and Suschetet,M. (1996)Time course of induction of rat hepatic drug metabolising enzymeactivities following dietary administration of flavonoids. J. Toxicol.Environ. Health, 49, 481496.

51.Stillwell,W.G., Skipper,P.L. and Tannenbaum,S.R. (2002) Biomonitoring:methods and analysis, In Nagao,M. and Sugimura,T. (eds), FoodBorne Carcinogenesis: Heterocyclic Amines. Wiley, New York, NY,pp. 319331.

Received May 13, 2003; revised August 14, 2003; accepted August 20, 2003

M.Zsivkovits et al.

1918