Carcinogenesis vol.18 no.4 pp.833–841, 1997
`
`Bifidobacterium longum, a lactic acid-producing intestinal
`bacterium inhibits colon cancer and modulates the intermediate
`biomarkers of colon carcinogenesis
`
`carc$$0424
`
`Jagveer Singh, Abraham Rivenson, Mamoru Tomita1,
`Seiichi Shimamura1, Norio Ishibashi1 and
`Bandaru S.Reddy2
`Divisions of Nutritional Carcinogenesis and of Pathology and Toxicology,
`American Health Foundation, Valhalla, NY 10595, USA and 1Nutritional
`Science Laboratory, Research and Development Center, Morinaga Milk
`Industry Co. Ltd, Zama City, Kanagawa 228, Japan
`2To whom correspondence should be addressed
`The human colon can be described as a complex microbial
`ecosystem, comprising several hundred bacterial species.
`Some of these enteric bacteria are beneficial to the host
`and have been shown to exert antimutagenic and anti-
`carcinogenic properties. We have investigated the colon
`tumor inhibitory activity of Bifidobacterium longum, a lactic
`acid-producing enterobacterium. The modifying effects of
`this lactic culture on colonic mucosal and/or tumor cell
`proliferation, ODC activity and ras-p21 oncoprotein expres-
`sion in colon carcinogenesis were also analyzed. Male F344
`rats were fed a modified AIN-76A diet containing 0 or 2%
`lyophilized cultures of B.longum and s.c. administered
`azoxymethane (AOM) dissolved in normal saline at a dose
`of 15 mg/kg body wt, once weekly for 2 weeks. Vehicle
`controls received an equal volume of normal saline s.c.
`Animals were maintained on control or experimental diets
`until termination of the study. Animals intended for analysis
`of cell proliferation were killed 20 weeks after the second
`AOM injection, whereas animals intended for colon tumor
`analysis and measurement of ODC activity and ras-p21
`expression were killed 40 weeks after the last AOM injec-
`tion. The data demonstrate that dietary administration of
`lyophilized cultures of B.longum resulted in significant
`suppression of colon tumor incidence and tumor multipli-
`city and also reduced tumor volume. Results also revealed
`that ingestion of B.longum significantly inhibited AOM-
`induced cell proliferation, ODC activity and expression of
`ras-p21 oncoprotein. Data suggest that oral administration
`of probiotic B.longum exerts strong antitumor activity, as
`indicated by modulation of the intermediate biomarkers
`of colon cancer, and consequently reduced tumor outcome.
`
`Introduction
`Colorectal cancer remains one of the leading causes of cancer
`morbidity and mortality among men and women in Western
`countries, including the USA, where an estimated 133 500
`new cases will lead to ~55 000 deaths in 1996 (1). Despite
`several advances in the treatment of cancer, the therapeutic
`outcome of this malignancy is posing a significant challenge
`to modern medicine. Consequently, primary prevention, early
`detection and secondary prevention are emerging as the most
`
`*Abbreviations: AOM, azoxymethane; IQ, 2-amino-3-methylimidazo[4,5-
`f]quinolone; ODC, ornithine decarboxylase; BrdU, bromodeoxyuridine; DTT,
`dithiothreitol; TBST, Tris-buffered saline containing 0.1% Tween-20.
`
`© Oxford University Press
`
`promising approaches for reducing the morbidity and mortality
`from colorectal cancer (2). Epidemiological and experimental
`studies have demonstrated that high dietary fat intake and lack
`of adequate amounts of dietary fiber increase the risk of colon
`cancer development (3), whereas several micronutrients, trace
`elements present in fruits and vegetables and their synthetic
`analogs reduce the risk of colon cancer (4,5). Of special
`interest in this regard is the beneficial effect of certain lactic
`acid-producing enterobacterial food supplements, the so-called
`probiotics, in the prevention of chronic conditions such as
`cardiovascular disease and cancer (6,7). These lactic cultures,
`which are primarily used for fermentation of milk and other
`dairy products, have been shown to possess antimutagenic and
`anticarcinogenic properties (8–10). In fact,
`the data from
`epidemiological and experimental studies indicate that inges-
`tion of certain lactic cultures, such as lactobacilli and bifido-
`bacteria, or their fermented dairy products reduce the risk of
`certain types of cancer and inhibit tumor growth (11–13). In
`a study in Japan, Kubota found that colon cancer incidence
`was lowest when the colonic population of bifidobacteria was
`highest and that of Clostridium perfringens was lowest (14).
`Goldin and Gorbach have shown that dietary supplements
`of Lactobacillus acidophilus suppressed DMH-induced colon
`tumor incidence and enhanced tumor latency in rodents (15).
`Shackelford et al. have described an increased survival rate
`among carcinogen-treated animals fed fermented milk (16).
`Recent data from our laboratory indicate that dietary intake
`of Bifidobacterium longum cultures significantly inhibits the
`development of azoxymethane (AOM*)-induced aberrant crypt
`foci representing putative premalignant lesions (17) and blocks
`the induction of colon and liver tumors by 2-amino-3-methyl-
`imidazo[4,5-f]quinolone (IQ), a food mutagen (18). However,
`the precise mechanism by which these lactic cultures exert
`their antitumorigenic influence is not clear. The antimutagenic
`activity of lactic acid-producing bacteria is suspected to reside
`in the cell wall (19), as lactic acid itself has no reported
`antimutagenic effects (20). Sekine et al. (21) and Okawa et al.
`(22) demonstrated that bifidobacterial as well as lactobacterial
`cell wall preparations induce immunity against tumor induction
`both in vivo and in vitro, with the characteristics of a biological
`response modifier. Studies by Zhang and Ohta (23,24) and by
`Orrhage et al. (25) have indicated that cells of lactic acid-
`producing bacteria bind to food-derived mutagens and decrease
`their absorption in the gut by physically removing them from
`the intestine via feces.
`Polyamines play an essential role in cell proliferation and
`differentiation and participate in macromolecular synthesis.
`Ornithine decarboxylase (ODC, EC 4.1.1.17) is the first and
`rate limiting enzyme of this crucial polyamine biosynthetic
`pathway (26). Increased ODC activity has been observed in
`colon adenomas and carcinomas (27,28) as well as in normal
`appearing colon mucosa adjacent to adenomas (29), reflecting
`the underlying hyperproliferative state of colonic mucosa. We
`and others have found that inhibitors of ODC activity such
`833
`
`Genome Ex. 1004
`Page 1 of 9
`
`

`

`J.Singh et al.
`
`as D,L-a-difluoromethylornithine suppress chemically induced Materials and methods
`colorectal tumorigenesis (30,31). Increasing interest has been
`Animals, diets and carcinogen
`focused on alterations in ODC activity in association with
`Weanling male F344 rats were obtained from Charles River Breeding Laborat-
`clinically premalignant and malignant lesions. This has led to
`ories (Kingston, NY). Lyophilized cultures of B.longum were a generous gift
`studies on the potential use of ODC as an intermediate
`from the Morinaga Milk Industry Co. Ltd (Zama City, Japan). Bifidobacterium
`biomarker of cancer risk (32,33). Another putative biological
`longum was cultured in a fermentor of 30 l capacity in medium containing
`marker that has been extensively analyzed in a variety of
`2% glucose, 1% peptone, 1% yeast extract and 0.5% salt. The incubation time
`was 12–14 h at 37°C until the viable cells reached ~3–43109/ml. The cells
`tissues is the degree of cell proliferation. Enhanced cellular
`were harvested by centrifugation and washed in saline. After mixing with a
`proliferation resulting in anomalous expansion of epithelial
`cryoprotectant solution containing 1% sodium glutamate and 3.5% sucrose,
`cells within the colonic crypts has been found in patients
`the cells were lyophilized at
`the Research and Development Center of
`known to be at high risk for colon cancer (34,35). A similar
`Morinaga Milk Industry Co. Ltd. The viable cells were enumerated by anerobic
`plate count methods as described by Rasic (53). Each gram of lyophilized
`expansion of the proliferative region of actively renewing
`culture contained ~231010 live bacterial cells. All ingredients of the semipur-
`epithelium has been noted in humans with precancerous lesions
`ified AIN-76A diet were obtained from Dyets Inc. (Bethlehem, PA) and stored
`of the stomach (36), esophagus (37) and cervix (38) and in
`at 4°C prior to preparation. The composition of the control diet was as follows
`various organs of carcinogen-treated animals (39). It has also
`(54): 20% casein, 0.3% D,L-methionine, 52% corn starch, 13% dextrose, 5%
`been observed that several antitumor agents and nutritional
`corn oil, 5% Alphacel, 3.5% mineral mix (AIN-76A), 1% vitamin mix (AIN-
`76A) and 0.2% choline bitartrate. Lyophilized cultures of B.longum at the 2%
`factors that have been shown to inhibit colon carcinogenesis
`level, equivalent to 431010 live cells/g diet, were added to the semipurified
`in rodents are also inhibitors of cellular proliferation (40).
`AIN-76A diet at the expense of dextrose. All the control and experimental
`At the molecular level, recent evidence indicates that activa-
`diets were prepared weekly in our laboratory and were stored in a cold room.
`tion of ras proto-oncogenes, coupled with the loss or inactiva-
`Animals had access to food and water at all times and food cups were
`replenished with fresh diet three times per week. AOM (CAS 25843-45-2)
`tion of suppressor genes (anti-oncogenes) induces a malignant
`was obtained from Ash Stevens (Detroit, MI).
`phenotype in colonic cells (41). The ras proto-oncogenes (c-
`Ki-ras, c-Ha-ras and N-ras) constitute a family of highly
`Experimental procedure
`conserved genes encoding a structurally and functionally
`Male F344 rats received at weaning were quarantined for 1 week. Animals
`were assigned to either AOM-treated or vehicle-treated groups and were
`related 21 kd protein, referred to as ras-p21, which is anchored
`housed in plastic cages with filter tops under controlled environmental
`to the cytoplasmic face of the plasma membrane, binds to the
`conditions of 21°C temperature and 50% humidity on a 12 h light/dark cycle.
`guanine nucleotides GTP and GDP and is believed to function
`Beginning at 5 weeks of age, groups of animals were fed the modified AIN
`as a molecular switch in transmembrane signaling events of
`76-A diet containing 0 (for controls) or 2% lyophilized B.longum cultures
`(for experimental groups). Two weeks later, animals intended for carcinogen
`cell growth and differentiation (42). The malignant potential
`treatment were given s.c. injections of AOM dissolved in normal saline at a
`of ras genes is attributed to mutational activation by single
`dose of 15 mg/kg body wt once weekly for 2 weeks. Vehicle controls received
`base substitution, critically in codons 12, 13 or 61, or by
`s.c. injections of a corresponding volume of normal saline. The animals were
`enhanced expression of ras-p21, or both (43,44). More than
`maintained on control or experimental diets until termination of the experiment.
`50% of human colon tumors have been shown to carry
`Body weights were recorded weekly until 16 weeks of age and then every 4
`weeks until termination of the study. Groups of animals intended for cell
`mutations in codon 12 of K-ras (44). An even higher incidence
`proliferation analysis were killed 20 weeks after the second AOM injection,
`of similar mutations has been observed in aberrant crypt
`whereas animals intended for colon tumor evaluation, ODC activity and ras-
`foci representing putative premalignant lesions (45), in colon
`p21 analysis were killed 40 weeks after the second AOM injection.
`adenomas (46) and also in histologically normal appearing
`All animals were killed by CO2 asphyxiation at the scheduled times and
`were carefully necropsied. After laparotomy, the entire stomach and the small
`mucosa adjacent to regions of colon carcinomas (47). Enhanced
`and large intestines were resected and opened longitudinally. The intestinal
`levels of oncogenic forms as well as normal cellular ras-p21
`contents were flushed with ice-cold normal saline. Using a dissection micro-
`have been detected in a variety of human tumors, including
`scope, tumors in the colon and small intestine were grossly recorded for their
`colon cancer (48,49). We and others (50,51) have seen predomi-
`location, number and size. For each tumor, the length (L), width (W) and
`nantly K-ras codon 12 mutations and also enhanced expression
`depth (D) were measured with calipers and estimates of tumor volume (V)
`were made according to the formula V 5 L3W3D3p/6 (53). Colon tumors
`of normal as well as mutated ras-p21 in colon tumors and in
`with a diameter of .0.4 cm were cut into two halves; one portion was used
`uninvolved colonic mucosa of carcinogen-treated rodents,
`for the analysis of ODC and ras-p21 and the other half was processed for
`suggesting an association between expression of activated ras
`histopathological examination of tumor type. In addition, tumors ,0.4 cm in
`and colon tumorigenesis. We and others (51,52) have also
`diameter were used for histopathology. Colonic mucosa that was free of
`tumors from AOM-treated animals and from saline-treated animals was
`provided evidence for inhibition of carcinogen-induced ras
`scraped with a microscope slide, snap-frozen in liquid nitrogen and stored
`mutations and suppression of expression of both normal as
`at –80°C until used for ODC and ras-p21 analysis. For histopathological
`well as mutated ras-p21 by inhibitors of ODC activity and
`evaluation, colonic and small intestinal tumors were fixed in 10% buffered
`other agents known to block colon tumor development. Thus,
`formalin, embedded in paraffin blocks and processed for hematoxylin and
`it is likely that changes in ODC activity and modulation of
`eosin staining. Stained sections were examined for tumor types as described
`(55). Most of the tumors in this study were adenocarcinomas.
`cell proliferation and ras-p21 expression that are associated
`with colon tumorigenesis would also be involved in colon
`Cell proliferation assay
`tumor inhibition by B.longum.
`Mucosal cell proliferation was analyzed using the bromodeoxyuridine (BrdU)
`It was therefore of interest to evaluate the colon tumor
`labeling method as described (56). One hour before being killed by CO2
`asphyxiation, animals intended for cell proliferation assay were i.p. injected
`inhibitory properties of dietary B.longum in the established
`with 20 mg/kg body wt BrdU. Their colons were removed, carefully slit open
`colon cancer model. We have analyzed the effect of dietary
`longitudinally from cecal end to rectum, fixed in 80% ethanol and embedded
`B.longum on AOM-induced colon tumorigenesis in male F344
`in paraffin. Sections of 4 mm thickness were cut perpendicular to the mucosal
`rats. We have also examined how this enterobacterial culture
`surface, deparaffinized, rehydrated and incubated for 30 min with 0.3% H2O2
`in methanol to quench the endogenous peroxidase activity and were incubated
`influences ODC activity, cell proliferation and the expression
`again for 20 min in 4 N HCl to denature the DNA. BrdU incorporation in the
`of mutated as well as normal cellular ras-p21 during AOM-
`nucleus was detected using monoclonal anti-BrdU as primary antibody
`induced colon carcinogenesis in order to better understand the
`(Becton-Dickinson, Mountview, CA) and a biotinylated secondary antibody
`underlying mechanisms.
`with a Vectastain ABC kit (Vector Laboratories, Burlingame, CA). Slides
`834
`
`Genome Ex. 1004
`Page 2 of 9
`
`

`

`Inhibition of colon cancer by lactic cultures
`
`Table I. Body weights of male F344 rats treated with AOM or vehicle and fed the control or experimental diets
`
`Experimental group
`
`Body wt (g) at weeks after last AOM/saline injectiona
`
`0
`
`8
`
`16
`
`24
`
`32
`
`40
`
`AOM-treated
`Control diet (42)b
`2% B.longum diet (42)
`Saline-treated
`Control diet (12)
`2% B.longum diet (12)
`
`173 6 10
`175 6 8.6
`
`181 6 8
`181 6 12
`
`292 6 16
`297 6 15
`
`311 6 16
`305 6 18
`
`350 6 21
`358 6 20
`
`370 6 21
`360 6 24
`
`382 6 24
`399 6 23
`
`413 6 26
`395 6 24
`
`412 6 27
`420 6 39
`
`445 6 31
`435 6 27
`
`439 6 36
`451 6 38
`
`476 6 42
`469 6 27
`
`aAnimals were injected s.c. with AOM or saline during the second and third week of being on the control or experimental diet.
`bNumber of animals at the beginning of the study. Twenty weeks after the last AOM injection, 12 animals from each control and experimental dietary goup
`were used for cell proliferation studies.
`
`were then counterstained with hematoxylin, dehydrated, clarified, mounted in
`Permount and examined under a standard light microscope.
`Approximately 35–40 longitudinally oriented crypts were analyzed from
`each animal. Each crypt was scored to determine the crypt height, which is
`defined as total number of cells per crypt column, and the number and position
`of labeled cells within each crypt column. Each crypt column was divided
`into three compartments of equal size, compartment 1 being in the lower third
`of the crypt, compartment 2 in the middle third and compartment 3 in the
`upper third near the luminal surface. Our attention was confined to well-
`oriented crypts in which the base, lumen and top of the crypt could be
`observed. The percentage of labeled cells (labeling index) was determined by
`tabulating the ratio of labeled cells to total number of cells in that region/
`crypt 3100. The total number of cells per crypt column of each animal was
`also determined. The distribution in the proliferative zone was measured as a
`percentage of total labeled cells per crypt column.
`ODC assay
`The colonic mucosal and tumor samples obtained 40 weeks after the last
`AOM or saline injection at the termination of the study were assayed for
`ODC activity using previously described methods (57). Briefly, specimens
`were homogenized in ice-cold buffer [25 mM Tris–HCl, pH 7.5, 2.5 mM
`dithiothreitol (DTT), 0.1 mM EDTA] and centrifuged at 40 000 g for 30 min
`at 4°C. Aliquots of clear suparnatants were added to 50 ml of the reaction
`mixture [15 mM Tris–HCl, pH 7.5, 0.1 mM EDTA, 2.5 mM DTT, 0.04 mM
`pyridoxal phosphate, 0.4 mM L-ornithine and 25 mCi D,L-[14C]ornithine
`hydrochloride (56.6 mCi/mmol; Amersham International, Arlington Heights,
`IL)]. The reaction mixture was incubated at 37°C for 1 h in 163100 mm
`glass tubes sealed with rubber stoppers supporting a center well (Kontes,
`Morton Grove, IL). The released 14CO2 was trapped on microglass fiber discs
`(934-AH, GFA; Whatmann) soaked in saturated barium hydroxide. The
`reaction was stopped by injecting 0.1 ml 2 N sulfuric acid through the rubber
`septum directly into the reaction mixture. The incubation was continued for
`an additional 1 h to completely trap the released 14CO2. Center wells along
`with the filter discs were then transferred to glass scintillation vials and the
`radioactivity was counted in 10 ml scintillation cocktail (Scintisol; ISOLAB
`Inc.). ODC activity was determined by measuring 14CO2 liberated from
`L-[1-14C]ornithine and expressed as pmol 14CO2 released/mg protein/min.
`Measurement of ras-p21 expression
`Differential expression of total as well as mutant ras-p21 was estimated as
`described (58). Briefly, colonic mucosa and tumor samples were washed in
`ice-cold phosphate-buffered saline and suspended in disruption buffer (50
`mM Tris–HCl, pH 7.8, 100 mM NaCl, 10 mM sodium deoxycholate, 1 mM
`DTT, 1 mM phenylmethylsulfonyl fluoride, 1% Triton X-100) containing
`0.1 mM leupeptin and 0.2 mg/ml aprotinin as protease inhibitors. The samples
`were homogenized and left on ice for 30 min. The extracts were clarified by
`centrifugation at 15 000 g for 15 min. at 4°C.
`
`SDS–polyacrylamide gel electrophoresis and Western blotting. SDS–PAGE
`and Western transfers were carried out essentially by the methods of Laemmli
`(59) and Towbin et al. (60) respectively. Clear extracts of colonic mucosa
`and tumors corresponding to 200 or 100 mg total protein respectively were
`solubilized in sample buffer (10% SDS, 600 mM Tris–HCl, pH 6.7, 50%
`glycerol) containing 2-mercaptoethanol and 50 mg/ml bromophenol blue.
`Samples were boiled for 2 min and resolved on 12.5% reduced polyacrylamide
`vertical slab gels with an overlay of 5% polyacrylamide along with low range
`SDS–PAGE pre-stained molecular weight markers (BioRad Laboratories,
`Richmond, CA) and ras-p21 Western blot standards (Oncogene Science,
`Manhasset, NY). Electrophoretically resolved proteins were electrotransferred
`
`onto nitrocellulose membrane (Hybond ECL; Amersham International) in a
`Trans-blot Electrophoretic Transfer Cell (BioRad Laboratories).
`
`Immunodetection and quantification of ras-p21 using enhanced chemilumines-
`cence (ECL). After transblotting the electrophoretically resolved proteins,
`blots were blocked with a 5% solution of non-fat dried milk in Tris-buffered
`saline containing 0.1% Tween-20 (TBST) and were then incubated with mouse
`monoclonal antibody pan-ras (Ab-2) or rabbit polyclonal antibody pan-rasAsp12
`(Ab-1) (Oncogene Science) diluted in TBST containing 0.5% non-fat dried
`milk. Ab-2 is broadly reactive to p-21 translational products of the H-, K-
`and N-ras genes, whereas Ab-1 specifically reacts only with the mutant ras-
`p21Asp12 and not with ras-p21Gly12 or ras-p21Val12. The rationale for determin-
`ing differential expression of wild-type and Asp12- specific mutant ras-p21
`was that we had, in our previous studies, observed predominantly GfiA
`transitions in codon 12 of K-ras, substituting Gly with Asp, during AOM-
`induced colon carcinogenesis (50). Blots were extensively washed in TBST
`and reincubated with peroxidase-linked secondary antibody (anti-mouse Ig or
`anti-rabbit Ig; Amersham International) diluted in TBST containing 0.5%
`non-fat dried milk. The blots were thoroughly washed in excess TBST and
`probed with an ECL Western blot detection system (Amersham International)
`using Reflection autoradiography films (Du Pont NEN, Boston, MA). The
`autoradiography films were scanned using an Image-master sharp laser scanner
`(PDI, Huntington, NY) and the peak areas representing ras-p21 bands of both
`the standards and samples were integrated.
`Protein determination
`Protein contents in colonic mucosal and tumor homogenates were determined
`by the method of Bradford (61), using bovine serum albumin as standard.
`Statistical analysis
`tumor volume, ODC
`tumor multiplicity,
`Body weights,
`tumor incidence,
`activity, cell proliferation and ras-p21 expression were compared between the
`animals fed control and experimental diets. Tumor incidence, i.e. the percentage
`of animals with tumors, was analyzed by c2 test. Tumor multiplicity, expressed
`as the mean number of tumors per animal, was analyzed by unpaired t-test.
`The data on ODC activity, cell proliferation, ras-p21 expression and body
`weights were analyzed using unpaired t-test and one-way analysis of variance.
`Differences were considered statistically significant at P , 0.05.
`Results
`General observations
`Table I presents the body weights of AOM- and vehicle-treated
`animals fed the control or experimental diets. Analysis of data
`reveals that there were no significant differences in the body
`weights of vehicle-treated animals fed control or B.longum
`diet nor between vehicle-treated and AOM-treated animals fed
`the B.longum diet. However, body weights of AOM-treated
`animals maintained on the control diet were slightly, though
`not significantly, lower in comparison with their counterparts
`on the B.longum diet from 16 weeks after the last AOM
`injection. This may be due to AOM carcinogenicity and
`consequent tumor burden.
`Tumor data
`Table II summarizes the AOM-induced tumors in the colon
`and small intestine in terms of tumor incidence (% animals
`835
`
`Genome Ex. 1004
`Page 3 of 9
`
`

`

`J.Singh et al.
`
`Table II. Effect of dietary lyophilized B.longum on intestinal tumor
`incidence in male F344 rats
`
`Control diet
`
`2% B.longum diet
`
`Colon
`Incidence (% animals with tumors)
`Tumors/animal
`Tumors/tumor-bearing animal
`Tumor volume (mm3)
`Small intestine
`Incidence (% animals with tumors)
`Tumors/animal
`Tumors/tumor-bearing animal
`
`77 (23/30)a,b
`1.8 6 1.27d
`2.3 6 0.9c
`550 6 3099
`
`43 (13/30)
`0.47 6 0.57
`1.08 6 0.28
`
`53 (16/30)
`0.83 6 0.98
`1.56 6 0.80
`32 6 49
`
`30 (9/30)
`0.30 6 0.4
`1.0 6 0.07
`
`aValues in parentheses are number of animals with tumors/effective number
`of animals in that group.
`b,c,dSignificantly different from B.longum group at P , 0.05, 0.01 and 0.001
`respectively.
`
`with tumors), colon tumor multiplicity (number of tumors/
`animal and number of tumors/tumor-bearing animal) and colon
`tumor volume. No tumors were found in vehicle-treated animals
`fed the control or B.longum diets. Dietary administration of
`B.longum cultures significantly inhibited the incidence of colon
`adenocarcinomas (P , 0.05), and colon tumor multiplicity in
`terms of tumors/animal (P , 0.001) and tumors/tumor-bearing
`animal (P , 0.01). Although there was a 91% reduction in
`colon tumor volume in the animals fed the B.longum diet, this
`difference was not statistically significant due to a large
`standard deviation. Animals fed the B.longum diet had fewer
`but a statistically insignificant number of small
`intestinal
`tumors than those fed the control diet.
`Cell proliferation
`The data on colonic epithelial cell proliferation were analyzed
`as the number of cells/crypt column and as the rate of cell
`proliferation (labeling index) in the lower, middle and upper
`third compartments of the crypt column and the total crypt
`column (Figure 1 and Table III). There was no significant
`difference in the total number of epithelial cells counted
`between the control and experimental groups (data not shown).
`However, as shown in Figure 1, the number and distribution
`of BrdU-labeled cells per crypt column as well as per crypt
`column compartment significantly differed in the two groups.
`As summarized in Table III, dietary B.longum significantly
`suppressed AOM-induced proliferative indices in the lower,
`middle and upper compartment as well as in the total crypt
`column (P , 0.01–0.001). This inhibitory effect of B.longum
`on AOM-induced cell proliferation was strongly correlated
`with tumor outcome.
`ODC activity
`Table IV summarizes ODC activity in the colonic mucosa
`and in the colon tumors. AOM administration significantly
`enhanced ODC activity in the colonic mucosa of animals fed
`the control as well as B.longum diets as compared with their
`saline-treated counterparts. Dietary intake of B.longum resulted
`in a significant inhibition of AOM-induced ODC activity in
`the colonic mucosa. Ingestion of cultures of B.longum, how-
`ever, did not cause a significant decrease in the steady-state
`levels of colonic mucosal ODC activity as shown in saline-
`treated animals. Colon tumors of AOM-treated animals fed
`the B.longum diet exhibited significantly lower levels of ODC
`activity as compared with the tumors from animals fed the
`control diet. This inhibitory effect of cultures of B.longum on
`836
`
`Fig. 1. Photomicrographs of colonic crypts showing BrdU-labeled cells
`stained immunohistochemically with anti-BrdU monoclonal antibody in
`male F344 rats as described under Materials and methods. (A) AOM-treated
`animals fed the control diet showing expansion of the proliferative
`compartment. (B) AOM-treated animals fed the B.longum diet. (Bar
`represents 10 mm.)
`
`AOM-induced ODC activity is very well correlated with colon
`tumor outcome.
`Western blot analysis and differential expression of ras-p21
`Figure 2A and B demonstrates representative examples of
`Western blot analyses of ras-p21 and ras-p21Asp12. The majority
`of the samples had detectable levels of p21 with pan-reactive
`anti-ras (Ab-1) mouse monoclonal antibody. Ab-1, which does
`not distinguish between wild-type and mutated forms of ras-
`p21, identified a duplet in samples expressing increased levels
`of p21 species that exhibit different electrophoretic mobilities
`(Figure 2A). These samples exhibited a single band when
`probed with anti-pan rasAsp12 (Ab-2) rabbit polyclonal antibody
`representing mutated ras-p21Asp12 (Figure 2B), thereby con-
`
`Genome Ex. 1004
`Page 4 of 9
`
`

`

`Inhibition of colon cancer by lactic cultures
`
`Table III. Effect of dietary lyophilized B.longum on rate of colonic mucosal cell proliferationa during colon carcinogenesis in male F344 rats
`
`Experimental group
`
`Percent labeled cells/total cells
`
`Percent labeled cells/total cells in compartment
`
`Control diet
`2% B.longum diet
`
`Total labeling
`index
`
`18.9 6 1.1b
`12.8 6 1.1d
`
`Lower third
`
`Middle third
`
`Upper third
`
`Lower third
`
`Middle third
`
`Upper third
`
`6.9 6 0.5
`5.2 6 0.5d
`
`8.8 6 0.7
`5.2 6 0.8d
`
`3.2 6 0.5
`2.2 6 0.5d
`
`21.5 6 1.9
`15.7 6 1.7d
`
`26.7 6 2.0
`15.5 6 2.2d
`
`9.9 6 1.4
`7.1 6 1.4c
`
`aCell proliferation is expressed as colonic crypt labeling index (LI): LI 5 (no. labeled cells/total number of cells)3100.
`bMean 6 SD (n 5 12).
`cSignificantly different from the control group, P , 0.01.
`dSignificantly different from the control group, P , 0.001.
`
`Fig. 2. Western blot analysis of total ras-p21 (A) and mutated ras-p21Asp12 (B) expression. Extracts of colonic mucosa or tumors were resolved by
`SDS–PAGE, electroblotted onto ECL-Hybond followed by immunodetection using pan-reactive anti-ras-p21 monoclonal antibody (A) or antipan ras-p21Asp12
`polyclonal antibody (B) as described under Materials and methods. (A) Lanes 1–6, ras-p21 Western blotting standards; lanes 6 and 7, colon mucosa of
`AOM-treated animals fed the control and experimental diet respectively; lanes 8 and 9, colon mucosa from saline-treated animals fed the control and
`experimental diets respectively; lanes 10 and 11, colon tumors from AOM-treated animals fed the control and experimental diets respectively. (B) Lanes 12
`and 13, colon tumors; lanes 14 and 15, colon mucosa from AOM-treated animals fed the control and experimental diets respectively.
`
`Table IV. Effect of dietary lyophilized B.longum on colonic mucosal and
`tumor ODC activity in male F344 ratsa
`
`Experimental group
`
`Control diet
`
`2% B.longum diet
`
`AOM-treated
`Mucosa
`Tumor
`Saline-treated
`Mucosa
`
`66 6 10b,c
`456 6 147c
`
`11.5 6 4.3
`
`32 6 6
`101 6 30
`
`10.3 6 3.3
`
`aODC activity is defined as pmol 14CO2 released/mg protein/min.
`bMean 6 SD (n 5 12).
`cSignificantly different from B.longum diet group at P , 0.001.
`
`firming the simultaneous occurrence of both wild-type and
`mutated ras-p21 phenotypes.
`A standard curve of integrated optical density from laser
`densitometric scans representing ras-p21 Western blot stand-
`ards was plotted to quantify immunoreactive ras-p21 protein
`(Figure 3). Table V summarizes the results of Western blot
`analysis for total as well as mutant ras-p21 expression in both
`colonic mucosa and tumors representing AOM-treated animals
`fed the control and B.longum diets. Dietary B.longum signific-
`antly suppressed the expression of total and mutated ras-p21
`
`in colonic mucosa and tumors as compared with the control
`diet (P , 0.01). This inhibitory effect of B.longum cultures on
`AOM-induced ras-p21 expression was again strongly correlated
`with colon tumor outcome.
`
`Discussion
`The main purpose of this study was to evaluate the colon
`tumor inhibitory properties of cultures of B.longum. A previous
`study from our laboratory (17), demonstrating inhibitory effects
`of dietary B.longum on cecal b-glucuronidase activity and the
`development of AOM-induced aberrant crypt foci, provided
`the impetus for studying the influence of this lactic culture on
`colon tumorigenesis in a well-established experimental model.
`Our experiments demonstrate that whereas AOM administra-
`tion induces multiple colon tumors in ~77% of treated animals,
`dietary intake of B.longum significantly suppresses the number
`as well as the size of these tumors. To our knowledge, this is
`the first study providing evidence that ingestion of lyophilized
`cultures of B.longum, a lactic acid-producing bacterium present
`in the human colon, inhibits tumor incidence and multiplicity
`in addition to reducing the overall volume of AOM-induced
`colon tumors.
`Several
`lines of evidence support
`
`the tumor inhibitory
`837
`
`Genome Ex. 1004
`Page 5 of 9
`
`

`

`J.Singh et al.
`
`Fig. 3. ras-p21 standard curve.
`
`properties of lactic acid bacteria. Epidemiological studies
`involving cancer patients and populations at increased risk
`have indicated that consumption of cultured dairy products
`has an inverse correlation with the risk of colon and breast
`cancer (11,62–64). In rodents, Shackelford et al. (16) demon-
`strated a protective effect of orally administered dairy products
`fermented by Streptococcus thermophilus or Lactobacillus
`bulgaricus against chemically induced colon tumors. In another
`experiment, the induction of colon carcinomas by 1,2-dimethyl-
`hydrazine in rats was significantly suppressed by feeding
`animals with viable L.acidophilus cells and it was speculated
`that a similar effect could occur in humans (14). Injections of
`live or dead Bifidobacterium cells into sarcoma