Lactic Acid Bacteria Isolated from Dairy Products InhibitGenotoxic Effect of 4-Nitroquinoline-1-oxide in SOS-Chromotest

Giovanni Cenci1, Jone Rossi2, Francesca Trotta1, and Giovanna Caldini1

1 Dipartimento di Biologia Cellulare e Molecolare, Università di Perugia, Italy2 Dipartimento di Scienze degli Alimenti, Università di Perugia, Italy

Received: August 15, 2002

Summary

Antigenotoxic activity against 4-nitroquinoline-1-oxide (4-NQO) of lactic acid bacteria isolated fromcommercial dairy products was studied using SOS-Chromotest. The supernatants from bacteria-genotox-in co-incubations in general exhibited a strong suppression on SOS-induction produced by 4-NQO on thetester organism Escherichia coli PQ37 (sfiA::lacZ). High genotoxicity inhibition (>75%) was found for31/67 of the examined bacteria and the maximum values of some strains within the species were as fol-lows: Lactobacillus casei, 99.1%; L. plantarum, 93.3%; L. rhamnosus, 93.4%; L. acidophilus, 90.9%; L. delbrueckii subsp. bulgaricus, 85.7% and Bifidobacterium bifidum, 89.6%; Strains with low antigeno-toxicity (5–60%) were evidenced in both L. acidophilus and L. delbrueckii subsp. bulgaricus, whereassome inactive strains were found only in L. casei and L. delbrueckii subsp. bulgaricus. Cell exposure to100 °C for 15 min prevented antigenotoxicity and no effect was evidenced for cell-free spent media. Theactive strains survived at 0.1 mM 4-NQO exposure and generally presented some relevant functionalproperties, such as tolerance to bile (0.5%) or acid environment (pH 2.0) and adherence to Caco-2 ente-rocytes. Antigenotoxicity was always associated with modification of the 4-NQO absorbance profile.

Key words: Lactic acid bacteria – dairy products – probiotics – 4-nitroquinoline-1-oxide – antigenotoxi-city – SOS-Chromotest

Introduction

The health benefits attributed to probiotic bacteria,the lactic acid bacteria (LAB) in particular, are their utili-ty in maintaining gastrointestinal microbiota perfor-mance, both in the human and zoo-technical fields [11,12, 27, 17]. Different from some intestinal microfloracomponents, chiefly anaerobic bacteria, which can be im-plicated in gut carcinogenesis, either by producing en-zymes that convert precarcinogens to active carcinogensor by expressing other risk factors (e.g. mucolytic activi-ties) [15, 16, 28], probiotics besides balancing the gut mi-crobiota and reinforcing host immunity also seem effec-tive against cancerogenesis [42, 6]. They exert a markedinfluence on gut microbial metabolism, by reducing theactivity of bacterial enzymes, such as β-glucuronidase, β-glucosidase, and nitroreductase, which are implicated inthe metabolism of toxic compounds [14, 26, 9].

Evidence has been accumulated, showing that thesebacteria could also be responsible for reducing in vitrothe activity of some genotoxic compounds, such as ni-troarenes, nitrosamines, aflatoxins, PAH, heterocyclic

aromatic amines [13, 15, 31, 23, 34]. This aspect is gain-ing interest with the increasing demand for functionalfoods, especially dairy products such as yoghurts and fer-mented milks containing lactobacilli and bifidobacteria[40]. The chemical-biological interactions inherent ingenotoxin deactivation are not yet completely evidencedand so there are some aspects which are not quite clear.The studies available are relatively few and not reliable[23]. However different features may be presumed for theaction of some probiotic strains against genotoxic com-pounds. The more relevant are: (i) competition with bac-teria which convert precarcinogens into carcinogens, (ii)direct inhibition of tumorogenesis by their metabolites,(iii) mutagen binding on cell components (cell wall, ex-opolysaccharides, etc) (iv) mutagen bioconversion [20,38, 33, 37, 7].

0723-2020/02/25/04-483 $ 15.00/0

System. Appl. Microbiol. 25, 483–490 (2002)© Urban & Fischer Verlag http://www.urbanfischer.de/journals/sam

Abbreviations: LAB – lactic acid bacteria; GI – genotoxicity inhi-bition; 4-NQO – 4-nitroquinoline-1-oxide

The aim of the present study was to characterise forantigenotoxic activity against 4-NQO, a nitroaromaticgenotoxin [10], a selection of lactobacilli and bifidobacte-ria isolated from commercial dairy products – yoghurt,bio-yoghurt, fermented milk, mozzarella cheese – or fromstarter culture collection. In this context, it was impor-tant to define if antigenotoxicity can be attributed to sin-gle strains or fairly generalised to defined taxonomicgroups, and if the involved mechanisms are similar or dif-ferent within and between species. Another objective wasto study the presence of some safety and probiotic prop-erties (antibiotic-resistance, tolerance to bile salts andacidity, capability of colonising the intestinal tract), inrepresentative isolates having antigenotoxic activity, con-sidering that the above prerogatives are not necessarilypresent in the strains currently used in commercial dairyproducts [17, 18].

The study was carried out using SOS-chromotest, astandardised short term assay for detecting DNA damag-ing agents [30].

Materials and Methods

Bacterial strainsA total of 67 single colonies of LAB were selected randomly

from agar plates and examined for purity. 45 strains were iso-lated from different lots of dairy products available in Italiansupermarkets (the names of the producers are not reported)and 22 strains derived from the culture collection of our labo-ratory. The strains and their origin are listed in Table 1. Isola-tion was obtained in microaerophylic (LAB) or anaerobic (bifi-dobacteria) conditions using respectively, Campygen or Gas-pack (Oxoid) in jar, and MRS agar (Oxoid) medium at 30–40 °C for 24–48 hours. Characterisation was performed bytraditional tests (colony and cell morphology, Gram stain ap-pearance, catalase reaction, carbon source fermentation, acidand gas production from glucose, reduction of 1% NO3, pro-duction of NH3, growth at 15 and 45 °C and in 2, 4, 6%NaCl) and by API 50CHL kit with APILAB Plus 4.0 software(BioMérieux) [23]. The strains were stored at –20 °C (glycerol90%) and subcultured 2–3 times on MRS before further exper-iments.

484 G. Cenci et al.

Table 1. Lactic acid bacteria used in this study and their origin.

Species and strains n Origin Industry

Lactobacillus delbrueckii subsp. bulgaricus 15 yoghurt AV2Z2, V2Z3, V2Z4, V2Z5, V2Z6, V2Z8, V2Z9, V5Z5, V5Z6, V5Z7,V5Z11, V2c, V5X5, 6a, 6b

Lactobacillus delbrueckii subsp. bulgaricus 1 yoghurt B2b

Lactobacillus delbrueckii subsp. bulgaricus 3 yoghurt C5a, 5b, 5c

Lactobacillus delbrueckii subsp. bulgaricus 2 yoghurt D1a, 1b

Lactobacillus delbrueckii subsp. bulgaricus 3 our culture collectionJ87, J88, J89

Lactobacillus casei 10 fermented milk B5H1, 5H2, 5H3, 5H4, 5H5, 5H6, 5H7, 5H8, 5H9, 5H10

Lactobacillus casei 3 fermented milk EC1, C2, C3

Lactobacillus casei 3 bio-yoghurta AV5Z4, V5Z9, V5Z10

Lactobacillus casei 1 bio-yoghurta B2a

Lactobacillus rhamnosus 4 mozzarella cheese FJ10, J30, J42, J54

Lactobacillus rhamnosus 2 our culture collectionJ61, J62

Lactobacillus acidophilus 15 our culture collectionJ71, J72, J76, J77, A1, A2, A3, A4, A5, A7, A8, A9, A11, A43, A44

Lactobacillus plantarum 3 mozzarella cheese FJ1, J25, J40

Bifidobacterium bifidum 2 our culture collectionJ91, J92

a yoghurt supplemented with probiotic strains.

Assessment of potential probiotic strains in vitro Antibiotic resistance. Antibiotic susceptibility was deter-

mined by diffusion disc method [2] using Mueller-Hinton Agar(Oxoid). The multiple antibiotic resistance index was expressedas the ratio between the number of observed resistances for agiven strain and the total number of antibiotics tested.

Effect of bile salts and pH on growth and survival. The effectson growth were evaluated comparing optical density at 620 nmon MRS broth (Oxoid) after 24–48 h incubation (anaerobically)at 37 °C in the presence of the stressors (0.3 and 0.5% ox-bilesalts, Oxoid; pH 2.0, 2.5, 3.0, 3.5 and 4.0, adjusted with HCl)with that of control cultures in the same medium. The survivalunder the different conditions was determined after 1 h incuba-tion at 37 °C (106 cells ml–1) and plating on MRS agar (Oxoid).

Viability. Cell viability after genotoxin-cell co-incubation wasdetermined by plate count (MRS agar, Oxoid) or using the fluo-rescent Syto9 (Live/Dead BacLight L-7012, Molecular Probes).

ChemicalsThe genotoxin 4-nitroquinoline-1-oxide (4-NQO) (CAS no.

56–57–5) was obtained from Sigma. A stock solution was ob-tained in DMSO (1 mg/ml) and diluted in water before the tests.The substrates o-nitrophenyl-β-D-galactopyranoside (ONPG)and p-nitrophenylphosphate (PNPP), for colorimetric evaluationof β-galactosidase and alkaline phosphatase respectively, werepurchased from Sigma.

Genotoxin-cells coincubationLAB were cultured in MRS broth (30–40 °C overnight, 24h),

washed (6,000 × g , 15 min) and resuspended in physiologicalsaline until 108–109 cell ml–1. The 4-NQO genotoxin was addedin a final concentration of 0,1 mM and co-incubation was main-tained at 37 °C for 150 min (under shaking) according to Caldi-ni et al. [7]. After co-incubation the residual genotoxic activitywas determined on filtered (0.45 µm Sartorius membrane) su-pernatants. Experiments were also carried out using heat treatedcells (100 °C , 15 min), according to Lankaputhra and Shah[24]. In co-incubation experiments the optical density at 620 nmof cell suspensions were near 1.00. To determine whether the in-hibitory effect on 4-NQO activity was extracellular, genotoxinwas also co-incubated with cell-free supernatants from station-ary phase cultures.

Genotoxicity and genotoxicity inhibition assay4-NQO genotoxicity and its residual activity after cell co-incu-

bation were carried out using SOS-chromotest. The test is basedon the use of the genetically modified Escherichia coli, PQ37strain, in which the lacZ is under the control of the sfiA gene(sfiA::lacZ) [35, 36]. Genotoxicity in samples was detected mea-suring the activation of SOS-response of tester organism by evalu-ating β-galactosidase induction and alkaline phosphatase (consti-tutive) expression. These enzyme activities were used to calculatethe SOS induction factor (IFSOS). 4-NQO was used as positivecontrol for the test without metabolic activation at a final concen-tration of 3.2 nmol/ml with saline as negative control. Details aregiven in our previous study [7]. The experiments were repeated2–3 times and determinations were done in triplicate.

Genotoxin analysis after cell co-incubationUV-visible spectra were used for analysis of the remaining

genotoxin or bioconversion products in the supernatants.

In vitro adhesion assayLAB adhesion to the intestinal Caco-2 cell line was studied

according to the protocol described by Lee et al. [25]. Briefly, 1ml of overnight LAB culture (108–109 CFU × ml–1) was added to

15-days Caco-2 monolayers in microwells (BD Falcon) contain-ing 1ml of Dulbecco’s modified Eagle’s medium. After 1 h incu-bation at 37 °C, the wells were washed with phosphate buffer(pH 7.2), methanol fixed and gram stained. The number of bac-teria adhering to about 1000 Caco-2 were evaluated by 20 to 60random microscopic fields. Experiments were performed in trip-licate and results expressed as bacteria adhering to 100 Caco-2.

Statistical analysisResults were presented as the average of two-three indepen-

dent experiments with one standard deviation when appropriateor by median. The differences between groups were determinedusing Student’s t test for independent samples.

Results

Strains description

The strains, isolated from dairy products or obtainedfrom starter culture collection, belong prevalently to thespecies Lactobacillus delbrueckii subsp. bulgaricus (infollowing Figures and text referred as L. delbrueckii), L. casei, L. acidophilus, and to a lesser number L. rham-nosus and Bifidobacterium bifidum. Strains frequencyand their origin are reported in Table 1.

Inhibition of genotoxin activity by live cells

The supernatants of genotoxin-LAB co-incubationproduced a significant suppression of β-galactosidase ac-tivity of tester PQ37 with respect to levels commonlyfound in the presence of genotoxin. During our experi-ments, in the positive controls (3.2 µM 4-NQO) the meanunits of β-galactosidase were 32.9 ± 7.4 and those of al-kaline phosphatase 15.2 ± 4.3 (n = 25). The same activi-ties in negative controls (saline) were 3.4 ± 0.8 and 30.2 ±6.4, respectively. Inhibition of SOS induction factor wasused to assay in vitro antigenotoxic effect on 4-NQOproduced by LAB-co-incubation, and results were ex-pressed as percent of genotoxicity inhibition (GI) with re-spect to positive controls (Figure 1).

Based on these data we evidenced that 46.3% of thestrains (31/67) are active and produce GI > 75%. Activestrains were found in all considered species: L. casei, L. rhamnosus,, L. acidophilus, L. delbrueckii, L. plan-tarum, B. bifidum. For all the species, except L. del-brueckii, the activity of some isolates was greater than90% and regarded the following strains: L. casei 5H6(99.1%), 5H4 (96.8%), 5H10 (94.2%); L. plantarumJ25 (93.3%), J40 (90.6%); L. rhamnosus J30 (93.4%),J40 (93.2%), J10 (92.8%); J62 (90.8%), J61 (90.0%), L. acidophilus A10 (90.9%). Only in L. acidophilus andL. delbrueckii the median GI was lower than 75%.

Some completely inactive strains were found in L. casei and L. delbrueckii. Within the species no sub-stantial differences were observed in relation to the indus-try from which the dairy product was obtained.

LAB survival in the presence of genotoxin

It was observed that viability of the strains withantigenotoxic activity against 4-NQO was maintained

Antigenotoxicity of Lactic Acid Bacteria 485

after genotoxin co-incubation. In fact, cell survival gener-ally remained above 92%, with the exception of L. aci-dophilus strains which was near 75% (Figure 2). Cell via-bility of the above strains was also confirmed by the fluo-rescent Syto9 assay.

Antigenotoxicity hypotheses

In order to evaluate some hypotheses about LAB-geno-toxin interaction the following aspects were studied.

The effect of killed cells. As shown in Figure 3 thehigh LAB antigenotoxicity against 4-NQO observed withlive cells was strongly reduced (L. rhamnosus and L. aci-dophilus) or completely inhibited (L. casei and B. bi-fidum) when cultures were heat treated before their co-in-cubation with genotoxin.

The role of spent culture supernatants. The spent su-pernatants recovered from the stationary phase culture oftwo strains (L. casei C1, L. rhamnosus J54) with highantigenotoxic activity were ineffective when co-incubatedwith 4-NQO. Moreover the typical 4-NQO spectrumwas revealed by UV-vis analyses after co-incubation.

The effect of washing fluid. No residual genotoxic ac-tivity was evidenced in washing fluids collected from livecells of active strains (L. casei 5H8, C1; L. acidophilusJ72, J77; L. plantarum J1, L. delbrueckii J87) co-incubat-ed with genotoxin and no 4-NQO profile was detected byspectrophotometric analyses on the examined samples.

In vitro assessment of potential probiotic properties

For LAB strains which demonstrated high in vitro ac-tivity against 4-NQO, the following characteristics relat-ed to probiotic activity have been considered (Table 2).

Antibiotic susceptibility. All the isolates, with few ex-ceptions, presented relevant resistance patterns and weresusceptible to ampicillin and imipenem and resistant toaztreonam, nalidixic acid, sulfonamide and streptomycin.The mean multiple antibiotic resistance indexes withinspecies were: 0.56 for L. casei, 0.90 for L. plantarum,0.66 for L. rhamnosus, 0.89 for L. delbrueckii, 0.71 forL. acidophilus and 0.31 for B. bifidum.

Effect of pH and bile salts on growth and survival.Generally an acid environment influenced growth al-though survival was not affected. In general pH ≤ 3.5 in-hibited LAB growth and so in these conditions the cellyield was very small (about 20% compared with controlculture). Nevertheless some exceptions were observed:the growth of L. acidophilus strains (A1, J76, J77) wasinhibited at pH ≤ 4.0, whereas B. bifidum J91 grew alsoat pH = 3.0.

Instead, the tested levels of bile salts did not influencegrowth (A1 was one exception) or strain survival. Therewas no difference in cell yield between control culturesand those with bile. No difference in behaviour was evi-denced for the isolates belonging to the same species.

Adhesion to Caco-2 line. The ability to adhere toCaco2 cells was demonstrated for all the 10 examined

486 G. Cenci et al.

Fig. 1. In vitro antigenotoxicity of LAB strains towards 4-NQOin the SOS-chromotest. Strains were co-incubated with genotox-in at 37 °C for 150 min. Relative percent inhibition was calcu-lated from residual activity (SOS induction factor) evaluated onsupernatants in relation to that of positive control. Each pointrepresents a mean value from 2–3 determinations for each strainand the horizontal bar indicates the median of species. Thestrain number is shown in parenthesis.

Fig. 2. The survival of LAB strains after co-incubation with 3.2µM 4-NQO at 37 °C for 150 min. Cells were washed three-times, resuspended in saline and suitable dilutions plated onMRS agar. Mean values ± standard deviation; the strain numberis shown in parenthesis. Bars marked with the same letter werenot significantly different by t test.

Antigenotoxicity of Lactic Acid Bacteria 487

Fig. 3. Antigenotoxic effect against 4-NQO of live and heattreated cells of LAB strains belonging to different species.

Table 2. Antibiotic resistance pattern, growth and survival on MRS broth at different pH or in the presence of bile and adherence onCaco-2 cells of representative LAB strains with antigenotoxic activity against 4-NQO.

Species Resistance patterna Growthb/Survivalc at Adher-and ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– enced

strainpH bile (%)

(%)––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– –––––––––––––––––––––––––––4 3.5 3 2.5 2 0.3 0.5

L.casei5H1 Azt Tob Gen [Cef] Sul Kan Ami Str Nov +/+ ±/+ –/+ –/+ –/± +/+ +/+ nd5H6 Azt Tob Gen Nal Sul Kan Ami Str Nov +/+ ±/+ –/+ –/+ –/± +/+ +/+ W5H10 Azt Tob Gen Nal Sul Kan Ami Str [Nov] +/+ ±/+ –/+ –/+ –/± +/+ +/+ WC1 Azt [Tob] Gen Nal Sul Kan Ami Str Nov +/+ ±/+ –/+ –/+ –/± +/+ +/+ WC2 Azt Tob [Gen] Nal Sul Kan Ami Str Nov +/+ ±/+ –/+ –/+ –/± +/+ +/+ ndC3 Azt Tob Gen Nal Sul Kan Ami Str Nov +/+ ±/+ –/+ –/+ –/± +/+ +/+ HL. plantarumJ25 Azt [Clo] Tob Gen Nal Sul Kan Ami Str Nov ±/+ –/+ –/+ –/± –/± ±/+ ±/+ HJ40 Azt Tob Tet Gen Nal Sul Kan Ami Str Nov +/+ ±/+ –/+ –/± –/± +/+ +/+ ndL. rhamnosusJ54 Azt Tob Tet [Gen] Nal Sul Kan Ami Str [Nov] +/+ ±/+ –/+ –/± –/± +/+ +/+ HJ61 Azt Tob Gen Nal Sul Kan Ami Str Nov +/+ ±/+ –/+ –/± –/± +/+ +/+ ndJ62 Azt Tob Tet Gen Nal Cef Sul Kan Ami Str Nov +/+ ±/+ –/+ –/± –/± ±/+ +/+ HJ42 Azt Tob Cep Gen Nal Sul Kan Ami Str Nov +/+ ±/+ –/+ –/± –/± +/+ +/+ HJ10 Azt Tob Nal Sul Kan Str +/+ ±/+ –/+ –/± –/± +/+ +/+ ndL. acidophilusA1 Azt Tob Gen Nal Sul Kan Ami Str Nov –/+ –/+ –/+ –/± –/± –/± –/± HJ76 Azt Tob Tet Cep Gen Nal [Cef] Sul Kan Ami Str Nov –/+ –/+ –/+ –/± –/± +/+ +/+ ndJ77 Azt Tob Tet Cep Gen Nal Cef Sul Kan Ami Str Nov –/+ –/+ –/+ –/± –/± +/+ +/+ HL. delbrueckiiJ87 Azt Tob Tet Cep Gen Nal Cef Sul Kan Ami Str Nov +/+ –/+ –/+ –/± –/± ±/+ ±/+ ndB.bifidumJ91 Azt Tob Tet Cep Gen Nal Cef Sul Kan Ami Str Nov +/+ +/+ +/+ –/± –/± +/+ +/+ nd

a Determined by agar diffusion method using the following antibiotics (mg per disc): Amikacin (Ami, 30), Aztreonam (Azt, 30),Cephalotin (Cep, 30), Cephoperazone (Cef, 30), Chloramphenicol (Chl, 30), Gentamicin (Gen, 10), Kanamycin (Kan, 30), Nalidixicacid (Nal, 30), Novobiocin (Nov, 5), Rifampin (Rif, 30), Streptomycin (Str, 10), Sulfonamide (Sul, 25), Tetracycline (Tet, 30),Tobramycin (Tob, 10). All the strains were susceptible to Ampicillin (10) and Imipenem (10).

b Evaluated by optical density at 620 nm, after 24–48 h incubation at 37 °C on MRS broth: –, no growth; ± and +, under and above70% of control, respectively.

c Determined by plating on MRS agar after 1 h incubation of 106 cells ml–1 from stationary phase: ± (102–105 cfu ml–1) and + (> 105 cfu ml–1).d Bacteria/100 Caco-2 : W- weakly (50–600); H - high (>1000).nd, not determined.

Table 3. Modification of 4-NQO absorbance profile after co-in-cubation with LAB strains with different antigenotoxicity degree.

4-NQO Wavelength of maximum Genotoxicityco-incubated with absorbance (nm) inhibition (%)

none 365 –L. casei C1 347 89.0L. casei C1 a 365 0L. casei 2a 365 0L. plantarum J25 353 93.3L. plantarum J1 365 16.9L. acidophilus J76 353 85.8L. acidophilus J76 a 365 0L. acidophilus A10 334 90.9L. acidophilus A2 364 14.3L. delbrueckii subsp. 353 85.7bulgaricus J87L. delbrueckii subsp. 365 0bulgaricus J87 aL. delbrueckii V2Z6 365 4.6B. bifidum J92 353 89.6B. bifidum J92 a 365 0

a heat treated cells

LAB with antigenotoxic activity. Results show that 7strains adhered with a high degree (>1000 bacteria/100Caco-2) and 3 to a lower extent (50–100 bacteria/100Caco-2).

UV-visible spectra

Evident modification of spectroscopic properties of 4-NQO were observed after co-incubation with live cellsof LAB with antigenotoxic activity, but not with heattreated cells of the same bacteria or with live cells ofstrains which were unable to reduce 4-NQO genotoxicity.In all cases the modification of 4-NQO spectroscopiccharacteristics was evidenced by a shift of the maximumabsorbance peak to a lower wavelength. In relation tothis characteristic three different behaviours were ob-served for examined strains with maximum of ab-sorbance shift from to 365 nm to 353, 347 or 334 nm, re-spectively (Table 3).

Discussion

The study evaluated in vitro the presence and extent ofanti-genotoxic activity of different lactic acid bacteria inrelation to the scarcity of information about the be-haviour of strains belonging to different species, widelypresent in dairy foods and used as probiotics. The majorindications in this regard are in fact inherent to epidemio-logical researches related to assumption of dairy productsand to some laboratory studies, which evidenced the pro-tective role of lactobacilli and bifidobacteria in rat coloncancer incidence [41, 17, 27]. The choice of 4-NQO as ref-erence genotoxin is justified since it is a potent carcino-gen, effective on microbial mutagenesis by strand scissionof DNA and production of charge transfer adducts [32],and a direct agent in the short term SOS-chromotest forgenotoxicity evaluation [29] with no need for metabolicS9-mix activation.

The data presented in this study indicate that the po-tential genotoxic role of 4-NQO can be strongly reducedby in vitro co-incubation with LAB and bifidobacteriafrom dairy products belonging to the six consideredspecies: L. casei, L. plantarum, L. rhamnosus, L. aci-dophilus, L. delbrueckii, B. bifidum. This result is impor-tant because LAB antigenotoxicity against 4-NQO hasbeen previously described for L. acidophilus and B. bi-fidum [24]. Nevertheless the prerogative under study re-garded 46.3% strains of examined sample and genotoxic-ity inhibition was not a prerogative of all isolates butonly of some, so that in a few species such as L. casei, L. acidophilus and L. delbrueckii two different activitygroups were evidenced, comprising active and moderate-ly-inactive strains, respectively. In L. acidophilus and L. delbrueckii a wide range of genotoxicity inhibitionwas observed, showing that the mean responses from allisolates were very low and certainly not representative ofthe potential activity expressed by some strains. A dis-crete number of L. acidophilus and L. delbrueckii withmoderate activity were found, whereas completely inac-

tive isolates were evidenced only in L. casei and L. del-brueckii.

In all cases LAB anti-genotoxicity was related to (i)high cell survival in the presence of genotoxin and (ii)modification of spectroscopic properties of the genotoxinin the water phase after co-incubation. These characteris-tics were not observed when genotoxin was co-incubatedwith inactive strains or with heat treated cells of activestrains. Different from what was observed for other dietarycarcinogens with different structure, e.g. benzo(a)pyrene,aflatoxin B1, imidazole derivatives [4, 19], no binding of 4-NQO to LAB cells was observed. All the above pointslead us to presume that LAB antigenotoxicity against 4-NQO may be directly or indirectly referred to LABmetabolic activity.

In particular, the behaviour evidenced could be ex-plained as an interaction with bacterial metabolites, forexample thiol peptides and polysaccharides produced byLAB [5, 39, 8], but in this study no activity was detectedon cell-free supernatants of spent media. Therefore thehypothesis of a biotransformation by LAB of the geno-toxin to inactive compound(s), as evidenced for Bacillusstrains where 4-aminoquinoline was found [7] is possible.An attempt to confirm the validity of this hypothesis byGC/MS, analysis of surnatants of representative strains isin course. The interpretations are coherent with the as-sumption that the antimutagenicity of cultured milk hasbeen properly attributed to the presence of LAB [1]. Inour study genotoxicity inhibition was observed using cul-tures from stationary phase, and it was produced rapidly.Different from what we observed for Bacillus [7], theLAB anti-genotoxicity was strain dependent, and so itcould be related to the molecular diversity usually ob-served among LAB strains [3, 21]. Further investigationscould give interesting information in this direction.

In conclusion our experimental results about inherentLAB antigenotoxicity seem to give additional credit tothis functional property of lactobacilli and bifidobacteriastrains present in commercial dairy products, even if theyare not labeled as probiotics by manufacturers. It is im-portant to note that for the strains which inhibit 4-NQOactivity, both the in vitro acidity and bile tolerance, andthe ability to adhere to enterocyte cultures have been evi-denced.

AcknowledgementsThis work has been carried out in the scope of “Pro-

getto di Ateneo Interarea 05”, financed by Universitàdegli Studi di Perugia, Italy.

References

1. Abdelali, H., Cassand, P., Sousotte, V., Koch-Bocabeille,B., Narbonne, J. F.: Antimutagenicity of components ofdairy products. Mutat. Res. 331, 133–141 (1995).

2. Bauer, A. W., Kirby, M. M., Sherris, J.T., Turck, M.: An-tibiotic susceptibility testing by a standard single discmethod Am. J. Clin. Pathol. 45, 493–496 (1966).

3. Ben Omar, N., Ampe F., Rainbault, M., Guyot, J. P., Tail-liez, P.: Molecular diversity of lactic acid bacteria from

488 G. Cenci et al.

cassava sour starch (Colombia). System. Appl. Microbiol.23, 285–291 (2000).

4. Bolognani, F., Rumney, C. J., Rowland, I. R.: Influence ofcarcinogen binding by lactic acid-producing bacteria ontissue distribution and in vivo mutagenicity of dietary car-cinogens. Food Chem. Toxicol. 35, 535–545 (1997).

5. Borob’eva, L. L., Cherdyntseva, T. A., Abilev, S. K.: An-timutagenic action of bacteria on mutagenesis induced by4-nitroquinoline-1-oxide in Salmonella typhimurium.Mikrobiologiia. 64, 228–233 (1995).

6. Burns, A. J., Rowland, I. R.: Anti-carcinogenicity of pro-biotics and prebiotics. Curr. Issues Intest. Microbiol. 1,13–24 (2000).

7. Caldini, G. Trotta, F., Cenci, G.: Inhibition of 4-nitro-quinoline-1-oxide genotoxicity by Bacillus strains. Res.Microbiol. 153, 165–171 (2002).

8. Dal Bello, F., Walter, J. Hertel, C., Hammes, W. P.: Invitro study of prebiotic properties of levan-type ex-opolysaccharides from lactobacilli and non-digestible car-bohydrates using denaturating gradient gel electrophore-sis. System. Appl. Microbiol. 24, 232–237 (2001).

9. Dunne, C., O’Mahony, L., Murphy, L., Thornton, G.,Morrisey, D., O’Halloran, S., Feeney, M., Flynn, S.,Fitzgerarld, G., Daly, C., Kiely, B., O’Sullivan, G. C.,Shanahan, F., Collins J. K.: In vivo selection criteria forprobiotic bacteria of human origin: correlation with invivo findings. Am. J. Clin. Nutr. 73/2, 386S–392S (2001).

10. Fann, Y. C., Methos-Dickey, C. A., Winston, G. W., Sygu-la, A., Rao, D. N. R., Kandiiska, M. B., Mason, R. P.: En-zymatic and nonenzymatic production of free radicalsfrom the carcinogens 4-nitroquinoline N-oxide and 4-hy-droxylaminoquinoline N-oxide. Chem Res. Toxicol. 12,450–458 (1999).

11. Fuller, R.: Probiotics in man and animal. J. Appl. Bacteri-ol. 66, 365–378 (1989).

12. Fuller, R., Gibson, G. R.: Modification of the intestinalmicroflora using probiotics and prebiotics. Scand. J Gas-troenterol. 222, 28S–31S, 1997

13. Goldin, B. R.: In situ bacterial metabolism and colon mu-tagens. Annu. Rev. Microbiol. 40, 367–369 (1986).

14. Goldin, B. R., Gorbach, S. L.: The effect of milk and Lac-tobacillus feeding on human intestinal bacterial enzymeactivity. Am. J. Clin. Nutr. 39, 756–761 (1984).

15. Goldin, B. R.: Intestinal microflora: metabolism of drugand carcinogens. Ann. Med. 22, 43–48 (1990).

16. Gorbach, S. L., Goldin, B. R.: The intestinal microfloraand the colon cancer connection. Rev. Infect. Dis. 12/2,252S–261S (1990).

17. Gorbach, S. L.: Probiotics and gastrointestinal health.Am.J.Gastroenterol. 95, 2S–4S (2000).

18. Haller, D., Colbus, H., Gänzle, M. G., Scherenbacher, P.,Bode, C., Hammes, W. P.: Metabolic and functional prop-erties of lactic acid bacteria in the gastro-intestinal ecosys-tem: a comparative in vitro study between bacteria of in-testinal and fermented food origin. Syst. Appl. Microbiol.24, 218–226 (2001).

19. Haskard, C. A., El-Nezami, H. S., Kankaanpää, P. E.,Salminen, S., Ahokas, J. T.: Surface binding of aflatoxinB1 by lactic acid bacteria. Appl. Environ. Microbiol. 67,3086–3091 (2001).

20. Hirayama, K., Rafter, J.: The role of lactic acid bacteria incolon cancer prevention: mechanistic considerations. An-tonie van Leeuwenhoek. 76, 391–394 (1999).

21. Holzapfel, W. H., Haberer, P., Geisen, R., Bjiorkroth, J.,Schillinger, U.: Taxonomy and important features of pro-biotic microorganisms in food and nutrition. Am. J. Clin.Nutr. 73/2: 365S–373S (2001).

22. Klein, G., Pack, A., Bonaparte, C., Reuter, G.: Taxonomyand physiology of probiotic lactic acid bacteria. Int. J.Food Microbiol. 41, 103–125 (1998).

23. Knasmüller, S., Steinkellner, H, Hirschl, A. M., Rabòt, S.,Nobis, E. C., Kassie, F.: Impact of bacteria in dairy pro-duicts and of the intestinal microflora on the genotoxicand carcinogenic effects of heterocyclic aromatic amines.Mutat.Res. 480, 129–138 (2001).

24. Lankaputhra, W. E. V., Shah, N. P.: Antimutagenic prop-erties of probiotic bacteria and organic acids. Mutat. Res.397, 169–182 (1998).

25. Lee, Y. K., Lim, C. Y., Teng, W. L., Ouwehand, A. C., To-muola, E. M., Salminen, S.: Quantitative approach in thestudy of adhesion of lactic acid bacteria to intestinal cellsand their competition with enterobacteria. Appl. Environ.Microbiol. 66, 3692–3697 (2000).

26. Ling, W. H., Korpela, R., Mykkanen, H., Salminen, S.,Hanninen, O.: Lactobacillus strain GG supplementationdecreased colonic hydrolytic and reductive enzyme activi-ties in health female adults. J. Nutr. 124, 18–23 (1994).

27. Marteau, P. R., de Vrese, M., Cellier, C. J., Schrezenmeir,J.: Protection from gastrointestinal diseases with the useof probiotics. Am. J. Clin. Nutr. 73/2: 430S–436S (2001).

28. McBain, A. J., Macfarlane, G. T.: Ecological and physio-logical studies on large intestinal bacteria in relation toproduction of hydrolytic and reductive enzymes involvedin formation of genotoxic metabolites. J. Med. Microbiol.47, 407–416 (1998).

29. Mersch-Sundermann, V., Kevekordes, S., Mochayedi, S.:Sources of variability of the Escherichia coli PQ37 geno-toxicity assay (SOS chromotest). Mutat. Res. 252, 51–60(1991).

30. Mersch-Sundermann, V., Schneider, U., Klopman, G.,Rosenkranz H. S.: SOS induction in Escherichia coli andSalmonella mutagenicity: a comparison using 330 com-pounds. Mutagenesis. 9, 205–224 (1994).

31. Morotomi, M., Mutai, M.: In vitro binding of potent mu-tagenic pyrolyzate of intestinal bacteria. J. Natl. CancerInst. 77, 195–201 (1986).

32. Nair, P. P., Davis, K. E., Shami, S., Lagerholm, S.: The in-duction of SOS function in Escherichia coli K-12/PQ37by 4-nitroquinoline oxide (4-NQO) and fecapentaenes-12and -14 is bile salt sensitive: implications for colon car-cinogenesis. Mutat. Res. 447, 179–185 (2000).

33. Orrhage, K., Sillerstrom, E., Gustafson, J. A., Nord, C. E.,Rafter, J.: Binding of mutagenic heterocyclic amines by in-testinal and lactic acid bacteria. Mutat. Res. 311,239–248 (1994).

34. Pool-Zobel, B. L., Neudecker, C., Domizlaff, I.,Schillinger, U., Rumney, C., Moretti, M., Villarini, M.,Scassellati-Sforzolini, G., Rowland, I. R.: Lactobacillus-and Bifidobacterium-mediated antigenotoxicity in thecolon of rats. Nutr. Cancer 26, 365–380 (1996).

35. Quillardet, P., Hofnung, M.: The SOS chromotest, a col-orimetric bacterial assay for genotoxins: procedures:Mutat. Res. 147, 65–67 (1985).

36. Quillardet, P. , Hofnung, M.: The SOS chromotest: a re-view. Mutat. Res. 297, 235–279 (1993).

37. Rolfe, R. D.: The role of probiotic cultures in the controlof gastrointestinal health. J. Nutr. 130, 396S–402S(2000).

38. Rowland, I. R., Grasso, P.: Degradation of N-ni-trosamines by intestinal bacteria. App1. Microbiol. 29,7–12 (1975).

39. Sreekumar, O., Hosono, A.: The antimutagenic propertiesof a polysacchharide produced by Bifidobacterium

Antigenotoxicity of Lactic Acid Bacteria 489

longum and its cultured milk against some heterocyclicamines. Can. J. Microbiol. 44, 1029–1036 (1998).

40. Vaughan, E. E., Mollet, B., deVos, W. M.: Functionality ofprobiotics and intestinal lactobacilli: light in the intestinaltract tunnel. Current Opinion in Biotechnology. 10,505–510 (1999).

41. Wollowski, I., Ji, S., Bakalinsky, A. T., Neudecker, C.,Pool-Zobel, B. L.: Bacteria used for the production of yo-gurt inactivate carcinogens and prevent DNA damage inthe colon of rats. J. Nutr. 129, 77–82 (1999).

42. Wollowski, I., Rechkemmer, G., Pool-Zobel, B.L.: Protec-tive role of probiotics and prebiotics in colon cancer. Am.J. Clin. Nutr. 73/2, 451S–455S (2001).

Corresponding author:Prof. Giovanni Cenci, Dipartimento di Biologia Cellulare eMolecolare, Università di Perugia, Via del Giochetto, I - 06126 Perugia, Italye-mail: gcenci@unipg.it

490 G. Cenci et al.