PhIP Exposure Research Articles

Detection of 2-amino-1-methyl-6-phenylimidazo [4,5-b]-pyridine (PhIP)–DNA adducts in human pancreatic tissues

Recent epidemiological investigations have observed an association between the consumption of grilled or barbecued meat and an increased risk of pancreatic cancer, suggesting that dietary exposure to heterocyclic aromatic amines (HCA) may contribute to the development of this disease. 2-Amino-1-methyl-6-phenylimidazo [4,5-b]-pyridine (PhIP) is the most abundant HCA found in well-done and grilled meats. To determine whether HCA-induced DNA damage is present in the human pancreas, immunohistochemistry and computer-assisted image analysis were used to measure PhIP–DNA adducts in 54 normal pancreatic tissues (N) from persons without pancreatic cancer and in 38 normal adjacent pancreatic tissues (A) and in 39 cancer tissues (T) from 68 patients with pancreatic adenocarcinoma. PhIP–DNA adducts were detected in 53 N, 34 A and 39 T samples. Mean values (±SD) of the absorbency for PhIP staining were 0.22±0.04, 0.24±0.04, and 0.24±0.03 for N, A, and T samples, respectively (p=0.004). Using the median absorbency (0.21) of the samples from normal controls as the cut-off, 71% of A and 77% of T tissues, compared with 48% of N tissues, were distributed in the higher range (p=0.009). The odds ratio of pancreatic cancer was 3.4 (95% confidence interval 1.5–7.5, p=0.002) for individuals with a higher level of PhIP–DNA adducts. This is the first report of the detection of PhIP–DNA adducts in human pancreatic tissue samples obtained from patients with unknown exposure to HCA. Although limited by the small sample size, these preliminary results suggest that PhIP exposure may contribute to human pancreatic cancer development.

Impact of genetic background on spontaneous or 2-amino-1-methyl-6-phenylimidazo[4,5- b]pyridine (PhIP)-induced intestinal tumorigenesis in Min/+ mice

We have studied the impact of genetic background on susceptibility to spontaneous or 2-amino-1-methyl-6-phenylimidazo[4,5- b]pyridine (PhIP)-induced intestinal tumorigenesis. The increase in small intestinal tumor number after PhIP exposure was 3.8- and 3.7-fold above the spontaneous levels in multiple intestinal neoplasia ( Min)/+ F1 mice with AKR/J and A/J backgrounds, respectively, compared with only 3-fold in C57BL/6J mice. In the colon, PhIP increased the number of tumors slightly more in C57BL/6J mice (3.3-fold) than in A/J mice (3.0-fold). AKR/J mice had no colonic tumors. Most of the tumors were located in the distal two-thirds of the small intestine in all three strains.

Mlh1-dependent suppression of specific mutations induced in vivo by the food-borne carcinogen 2-amino-1-methyl-6-phenylimidazo [4,5- b] pyridine (PhIP)

Disruption of the DNA mismatch repair (MMR) pathway results in elevated mutation rates, inappropriate survival of cells bearing DNA damage, and increased cancer risk. Relatively little is known about the potential impact of environmentally relevant carcinogens on cancer risk in individuals with MMR-deficiency. We determined the effect of MMR status ( Mlh1 +/+ versus Mlh1 −/−) on mutagenesis induced by the cooked-meat mutagen, 2-amino-1-methyl-6-phenylimidazo [4,5- b] pyridine (PhIP) within cII and supFG1 transgene reporters. Despite being a lymphomagen in mice, PhIP was not mutagenic in thymus. In colon, PhIP exposure induced 3-fold more mutations in Mlh1-deficient mice compared to their Mlh1 +/+ littermates. Similar induction was seen in Mlh1 −/− small intestine. Analysis of mutational spectra revealed that G/C to T/A transversions, the “signature PhIP mutation”, were induced to similar levels regardless of Mlh1 status. In contrast, Mlh1 −/− mice exhibited hypermutability to frameshifts, G/C to A/T transitions, and G/C to C/G transversions. Thus, both the level and types of mutation induced by PhIP are influenced by the activity of the MMR system. MMR may suppress PhIP-induced mutation through recognition and processing of specific mispairs (PhIP-G/T, PhIP-G/G, and PhIP-G/loop mispairs). In contrast, the PhIP-G/A mispair is unlikely to be a MMR substrate. In addition, the similar induction of both transversions and transitions in Mlh1 −/− mice suggests that mutagenic bypass of PhIP-G is similarly efficient with dATP, dTTP, and dGTP, in contrast to previously published conclusions. Our data suggests that MMR-deficiency would increase the likelihood of PhIP-induced carcinogenic mutations. Further evaluation of the risk that consumption of heterocyclic amines may impart to MMR-deficient individuals therefore is warranted.

Age at exposure and Apc status influence the levels of 2-amino-1-methyl-6-phenylimidazo[4,5- b]pyridine (PhIP)-DNA adducts in mouse intestine and liver

We have previously shown that C57BL/6J- Min/+ (multiple intestinal neoplasia) mice, heterozygous for the Min mutation in the adenomatous polyposis coli gene, were more susceptible to intestinal tumorigenesis and had higher intestinal PhIP-DNA adduct levels after exposure to the food mutagen 2-amino-1-methyl-6-phenylimidazo[4,5- b]pyridine (PhIP) on day 12 than on day 36 after birth [I.-L. Steffensen, H.A.J. Schut, J.E. Paulsen, Å. Andreassen, J. Alexander, Intestinal tumorigenesis in multiple intestinal neoplasia mice induced by the food mutagen 2-amino-1-methyl-6-phenylimidazo[4,5- b]pyridine: perinatal susceptibility, regional variation, and correlation with DNA adducts, Cancer Res. 61 (200l) 8689–8696]. In the present study, we have evaluated further whether this difference in susceptibility is related to adduct formation/removal, cell proliferation, apoptosis or expression of the nucleotide excision repair protein Xeroderma pigmentosum group A (XPA) in the intestines. Min/+ and +/+ (wild-type) mice were given a subcutaneous injection of 50 mg/kg bw PhIP on day 12 or 36, and the levels of PhIP-DNA adducts after 8, 12, 24 h, 3 or 7 days were quantified by use of 32P-postlabelling. In Min/+ mice, adduct levels were significantly higher after exposure on day 12 than on day 36 in the middle (1.5- to 8.5-fold) and distal (1.3- to 6.5-fold) small intestine from 8 h to 3 days after administration of PhIP, but not in the colon and proximal small intestine. In the liver – a non-target organ for PhIP – adduct levels were 2.0- to 7.5-fold higher after exposure on day 12 than on day 36 from 8 to 24 h after exposure. Adduct levels were generally higher in the middle (1.1- to 1.8-fold) and distal (1.1- to 2.0-fold) small intestines of Min/+ compared with +/+ mice after PhIP exposure on day 12, i.e. in the area of the intestines previously found also to have the highest number of tumors in Min/ + mice. PhIP increased cell proliferation and the number of apoptotic cells in the intestine and liver. However, the higher susceptibility to intestinal tumorigenesis in Min/+ mice exposed to PhIP at early age, or in Min/+ mice compared with +/+ mice, could not be explained by differences in cell proliferation, apoptosis or expression of the XPA repair protein.

A Mechanistic Basis for the Role of Cycle Arrest in the Genetic Toxicology of the Dietary Carcinogen 2-Amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP)

The heterocyclic amine 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP), formed during the cooking of meat, induces tumors of the prostate, colon, and mammary gland when fed to rats. PhIP is readily absorbed and efficiently metabolized to a genotoxic derivative by CYP1 enzymes. Although metabolism and mutational potential of PhIP have previously been well characterized, the intervening cellular and genomic responses to the chemical are not fully understood. We have examined the cellular response to PhIP exposure in human mammary epithelial MCF10A cells, which retain characteristics of normal breast epithelial cells. Because these cells fail to activate PhIP, they were cocultured with a human lymphoblastoid cell line MCL-5, which constitutively expresses CYP1A1, and have been transfected to express human CYPs1A2, 2A6, 3A4, and 2E1. The MCL-5 cells were irradiated (2,000 rads) prior to coculture, rendering them unable to replicate yet still retaining metabolic competency. MCF10A cells were treated (in the presence of MCL-5 cells) with PhIP (1-100 microM) and harvested at various time-points. Compared to DMSO control, treatment (24 or 48 h) with PhIP resulted in a significant dose-dependent fall in cell number. Cells treated for 48 h then cultured in the absence of PhIP (and MCL-5 cells) for a further 6 days showed a much greater dose-dependent reduction in cell number. Flow cytometric analysis indicated that PhIP treatment (48 h) resulted in a dose-dependent accumulation of cells in the G1 population. Western blotting revealed elevated expression of p53 and the cyclin dependent kinase inhibitor p21WAF1/CIP1 after PhIP treatment. Levels of MDM2, a negative regulator of p53, and the hypophosphorylated form of RB were also elevated, consistent with the triggering of G1 cell cycle checkpoint. These cell cycle effects are critical, as they enable cells to effect genome repair, accept mutation, or eliminate excessively damaged cells.

The effect of UDP-glucuronosyltransferase 1A1 expression on the mutagenicity and metabolism of the cooked-food carcinogen 2-amino-1-methyl-6-phenylimidazo[4,5- b]pyridine in CHO cells

UDP-glucuronosyltransferase proteins (UGT) catalyze the glucuronidation of both endogenous and xenobiotic compounds. In previous studies, UGT1A1 has been implicated in the detoxification of certain food-borne carcinogenic-heterocyclic amines. To determine the importance of UDP-glucuronosyltransferase 1A1 (UGT1A1) in the biotransformation of the cooked-food carcinogen 2-amino-1-methyl-6-phenylimidazo[4,5- b]pyridine (PhIP), genetically modified CHO cells that are nucleotide excision repair-deficient, and express cytochrome P4501A2 (UV5P3 cell line) were transfected with a cDNA plasmid of human UGT1A1 to establish the UDP-glucuronosyltransferase 1A1 expressing 5P3hUGT1A1 cell line. Expression of the UGT1A1 gene was verified by screening neo gene expressing clonal isolates (G-418 resistant) for their sensitivity to cell killing from PhIP exposure. Five of 11 clones were chosen for further analysis due to their resistance to cell killing. Western blot analysis was used to confirm the presence of the UGT1A1 and CYP1A2 proteins. All five clones displayed a 52-kDa protein band, which corresponded to a UGT1A1 control protein. Only four of the clones had a protein band that corresponded to the CYP1A2 control protein. Correct fragment size of the cDNAs in the remaining four clones was confirmed by RT-PCR and quantification of the mRNA product was accomplished by real-time RT-PCR. Expression of UGT1A1 in the transfected cells was 10 4–10 5-fold higher relative to the UV5P3 parental cells. One clone (#14) had a 10-fold higher increase in expression at 1.47 × 10 5 over the other three clones. This clone was also the most active in converting N-hydroxy-PhIP to N-hydroxy-PhIP glucuronide conjugates in microsomal metabolism assays. Based on the D 50 values, the cytotoxic effect of PhIP was decreased ∼350-fold in the 5P3hUGT1A1 cells compared to the UV5P3 control cells. In addition, no significant increase in mutation frequency was observed in the transfected cells. These results clearly indicate that UGT1A1 plays a critical role in PhIP biotransformation, providing protection against PhIP-mediated cytotoxicity and mutagenicity.

Reply

We thank Dietrich and Geier for their comments and the opportunity for a deeper discussion of our work. We presented clear evidence that: (1) in human liver microsomes N-OH-PhIP is metabolized into glucuronides predominantly by UGT1A1, and (2) polymorphisms in this gene influence the carcinogen metabolism.1 We agree that no conclusions can be extrapolated regarding the general detoxification of carcinogens, and additional studies evidently are needed to extend our data to the in vivo situation regarding PhIP. Yet, we believe our study presented relevant information for genetic epidemiological studies aimed at identifying genes affecting an individual's susceptibility though dietary exposure, which was the primary aim of our study. We agree that changes in protein expression and activity at different levels of any of the components involved in the biotransformation and transport of PhIP are likely to influence exposure to N-OH-PhIP. In Fig. 1 of their letter, Dietrich and Geier draw our attention to the fact that an altered UGT1A-dependent conjugation of N-OH-PhIP would largely be compensated by other enzymes, such as UGT2 or sulfotransferases. However, this is supported by data obtained in rodents. Indeed, as elegantly demonstrated by our colleagues and others, sulfation is the major N-OH-PhIP–conjugating reaction occurring in rats,2, 3 instead of glucuronidation in humans.4-6 Consequently, alterations of UGT1A in rodents would only affect a small fraction of N-OH-PhIP metabolism, and this loss will be easily compensated by SULT enzymes.3 In addition, whereas UGT2 may play a significant role in N-OH-PhIP glucuronidation in rats,3 we and others1, 7 demonstrated that human UGT2 is not reactive with this substrate, and therefore cannot compensate for a loss of UGT1A. Thus, we emphasize the fact that mechanisms illustrated in Fig. 1 in the letter by Dietrich and Geier are valid for rodents only. After the publication of our work1 Malfatti et al. reported2 that an impaired UGT1A pathway leads to lower DNA adduct formation in the colon of rats. Lower formation of N-OH-PhIP glucuronides in the liver has the potential to decrease the exposure of the colon to N-OH-PhIP-N3-G. In fact, N-OH-PhIP-N3-G can be hydrolysed to N-OH-PhIP by bacterial β-glucuronidases and converted locally to reactive metabolites.2 As a result and consistent with our hypothesis, detoxification by UGT1A in the liver, a predominant site for PhIP metabolism via N3 glucuronidation in addition to N2G formation in humans,4 would be likely to influence the exposure to carcinogens in the intestine and the colon. Finally, we did not intend to confirm the work by Peters et al.8 but rather to concede our hypothesis. We agree that it is complex to compare the level of glucuronidation determined with human liver microsomes to the concentration of glucuronides found in human urine. However, despite the limited statistical power of Peters et al.'s study, it cannot be ignored that UGT1A1 polymorphisms were found to alter the mutagenicity of urinary dietary carcinogens in healthy volunteers. This observation strongly supports the hypothesis that in humans UGT genetic alteration influences the carcinogenic potential of PhIP exposure. Chantal Guillemette*, Hugo Girard*, * Canada Research Chair in Pharmacogenomics, Laval University – Pharmacy, Quebec, Canada.

Equivocal impact of transplacental and lactational exposure to a food-derived carcinogen, 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine, on prostate and colon lesion development in F344 rats

The carcinogenic potential of maternal dietary exposure to 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) was investigated with the focus on the prostate and mammary glands and colons of the offspring. PhIP-DNA adducts were immunohistochemically detected in all three organs of 3-week-old animals after dams had received 200 ppm PhIP in the diet from 4 weeks before mating to weaning. The sites were essentially the same as in adult rats, suggesting that the distribution of bioactivating enzymes for PhIP do not differ greatly. Ki-67 labeling indices were increased only in the colons of the female offspring at 3-weeks of age. Development of preneoplastic and neoplastic lesions in the prostate and colon was not enhanced when the rats were maintained on PhIP-free diet until 63 weeks of age, except for significant increase in multiplicity of aberrant crypt foci in the females. The present findings indicate that PhIP exposure via placenta and breast milk may be important but that high doses and long periods may be necessary for tumor development.

Human UDP-glucuronosyltransferase 1A1 is the primary enzyme responsible for the N-glucuronidation of N-hydroxy-PhIP in vitro.

UDP-glucuronosyltransferase 1A proteins (UGT1A) catalyze the glucuronidation of many endogenous and xenobiotic compounds including heterocyclic amines and their hydroxylated metabolites. Studies have shown that in humans UGT1A-mediated glucuronidation is an important pathway in the detoxification of food-borne carcinogenic heterocyclic amines. The biotransformation of 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP), the most mass abundant heterocyclic amine found in cooked meats, is highly dependent on cytochrome P4501A2 hydroxylation followed by UGT-catalyzed glucuronidation of the N-hydroxy-PhIP reactive intermediate. To determine which UGT1A proteins are involved in the glucuronidation of N-hydroxy-PhIP, microsomal preparations from baculovirus-infected insect cells that express all of the known functional human UGT1A isozymes (UGT1A1, -1A3, -1A4, -1A6, -1A7, -1A8, -1A9, and -1A10) were exposed to N-hydroxy-PhIP and the reaction products were isolated by HPLC. All UGT1A proteins except UGT1A6 showed some degree of activity toward N-hydroxy-PhIP. The formation of both N-hydroxy-PhIP-N(2)-glucuronide and N-hydroxy-PhIP-N3-glucuronide was both time- and substrate concentration-dependent. UGT1A1 was the most efficient in converting N-hydroxy-PhIP to both conjugates producing five times more of the N(2)-conjugate than UGT1A4, the next most active UGT, and 286 times more than UGT1A7, the least active UGT. With an apparent K(m) of 52 microM and a K(cat) of 114 min(-)(1), UGT1A1 was also the most catalytically efficient in forming N-hydroxy-PhIP-N(2)-glucuronide. The catalytic efficiency for N-hydroxy-PhIP-N3-glucuronide formation was 8, 10, and 6 times lower for UGT1A1, -1A4, and -1A8, respectively, when compared to the K(cat) values for N-hydroxy-PhIP-N(2)-glucuronide formation. These results clearly show that UGT1A1 has the highest specificity for glucuronidating N-hydroxy-PhIP. Polymorphic expression resulting in decreased UGT1A1 activity in humans can cause reduced rates of glucuronidation, which can change the metabolic ratio between bioactivation and detoxification to favor bioactivation. This change will increase the susceptibility to the deleterious effects from PhIP exposure because the capacity to form nontoxic N-hydroxy-PhIP glucuronide conjugates will be diminished.

Characterization of a peptide adduct formed by N-acetoxy-2-amino-1-methyl-6-phenylimidazo[4,5- b]pyridine (PhIP), a reactive intermediate of the food carcinogen PhIP

2-Amino-1-methyl-6-phenylimidazo[4,5- b]pyridine (PhIP) is a member of a class of compounds known as the heterocyclic amines (HCAs) that are formed in meat during cooking. It is a multi-organ carcinogen in rodents forms adducts and with DNA and protein. Although protein adducts are not thought to be involved in cancer development, they may be useful as internal dosimeters of PhIP exposure and bioactivation. Towards the goals of characterizing the adducts formed in humans and the development of an assay for quantitation of adduct levels, we have characterized a peptide adduct formed by the putative genotoxic metabolite, N-acetoxy-PhIP. A model peptide with the internal sequence Leu–Gln–Lys–Cys–Pro–Tyr, which is homologous to a potential target sequence for HCAs in human serum albumin, was reacted with N-acetoxy-PhIP and an adduct was identified and further characterized by LC–ESI-MS/MS. N-acetoxy-PhIP is covalently bound to the peptide via cysteine and the exocyclic amino group of PhIP. Future work is needed to establish if this adduct is formed and is stable in vivo in humans following exposure to PhIP.

Urinary N2-(2'-deoxyguanosin-8-yl)PhIP as a biomarker for PhIP exposure.

The food-derived, heterocyclic aromatic amine 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) is genotoxic and is carcinogenic in experimental animals. Studies on the role of PhIP in human diet-related cancer would be aided considerably by the availability of a readily applicable biomarker of the internal dose of the ultimate genotoxic species. PhIP has been shown to adduct primarily at C-8 of deoxyguanosine in DNA and so the DNA repair product N(2)-(2'-deoxyguanosin-8-yl)PhIP is a potential biomarker of DNA adduction and repair after exposure to PhIP. An assay for N(2)-(2'-deoxyguanosin-8-yl)PhIP in urine has been developed based on liquid chromatography mass spectrometry, using a deuterated analogue of the nucleoside as an internal standard and an antibody-mediated extraction procedure. Polyclonal antibodies were raised against the PhIP-nucleotide, PhIP-nucleoside and PhIP-guanine base adducts conjugated to keyhole limpet haemocyanin. Following their evaluation, the immobilized PhIP nucleotide antibody was used for the extraction of N(2)-(2'-deoxyguanosin-8-yl)PhIP from urine. The limit of detection of the assay was 125 pg and the limit of quantification 200 pg for a 50 ml human urine sample. Following oral administration of PhIP (20 mg/kg body wt/day) to rats for 6 days, N(2)-(2'-deoxyguanosin-8-yl) PhIP was readily detected in the urine, reaching steady state over 3 days. This is the first direct demonstration of the urinary elimination of this adduct following exposure to parent amine. The half-life of the adduct with DNA was estimated to be approximately 20 h. The total amount of PhIP recovered in the urine as adduct was <0.5 x 10(-3)% of the dose administered. Levels of the PhIP adduct in urine collections from human subjects ingesting the amine (4.9 micro g) in cooked meat were below the limits of detection, indicating that humans are exposed to a bioactive dose of <3 x 10(-4) of that associated with a non-carcinogenic level in rats.

Sulforaphane and quercetin modulate PhIP-DNA adduct formation in human HepG2 cells and hepatocytes.

The formation of DNA adducts in human HepG2 cells and human hepatocytes exposed to 14C-labelled 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) was examined using Accelerator Mass Spectrometry (AMS). PhIP generated DNA adducts in a linear dose-dependent manner between 100 pM and 20 micro M. Co-treatment with the dietary isothiocyanate, sulforaphane (SFN, 1-10 micro M), or the flavonoid, quercetin (5-20 micro M), significantly reduced the level of PhIP-DNA adducts in a dose-dependent manner. The degree of protection was dependent on PhIP concentration, i.e. after 100 pM PhIP exposure, SFN or quercetin reduced adduct levels to below the limit of detection (0.15 amol PhIP/ micro g DNA) but at higher PhIP exposure (10 nM and 1 micro M), the protection was 60 and 10%, respectively. The involvement of phase I, phase II and DNA repair enzymes in this protection against PhIP-DNA adduct formation was investigated using real-time RT-PCR and enzyme activity assays. In intact HepG2 cells, quercetin inhibited cytochrome P450 (CYP)1A2, the main phase I enzyme responsible for PhIP bioactivation. In contrast, SFN induced phase II detoxification enzymes, UDP-glucuronosyltransferase 1A1 and glutathione S-transferase A1 mRNA expression. SFN and quercetin showed no effect on DNA repair, neither in terms of the level of PhIP-DNA adducts, when cells were treated with phytochemicals after the carcinogen exposure, nor the regulation of mRNA expression of two DNA repair enzymes, apurinic endonuclease and DNA polymerase beta. This study indicates that dietary isothiocyanates and flavonoids modulate phase I and phase II enzyme expression, hence increasing the rate of detoxification of the dietary carcinogen PhIP in human HepG2 cells but do not affect the rate of PhIP-DNA adduct repair. The formation of PhIP-DNA adducts in human hepatocytes was also dose-dependent with PhIP-concentration and the levels of protection by SFN or quercetin were up to 60% after 10 nM PhIP treatment, but showed large inter-individual variation with no observed protection in some individuals.

Factors affecting human heterocyclic amine intake and the metabolism of PhIP

We are working to understand possible human health effects from exposure to heterocyclic amines that are formed in meat during cooking. Laboratory-cooked beef, pork, and chicken are capable of producing tens of nanograms of MeIQx, IFP, and PhIP per gram of meat and smaller amounts of other heteroyclic amines. Well-done restaurant-cooked beef, pork, and chicken may contain PhIP and IFP at concentrations as high as tens of nanograms per gram and MeIQx at levels up to 3 ng/g. Although well-done chicken breast prepared in the laboratory may contain large amounts of PhIP, a survey of flame-grilled meat samples cooked in private homes showed PhIP levels in beef steak and chicken breast are not significantly different ( P=0.36). The extremely high PhIP levels reported in some studies of grilled chicken are not seen in home-cooked samples. Many studies suggest individuals may have varying susceptibility to carcinogens and that diet may influence metabolism, thus affecting cancer susceptibility. To understand the human metabolism of PhIP, we examined urinary metabolites of PhIP in volunteers following a single well-done meat exposure. Using solid-phase extraction and LC/MS/MS, we quantified four major PhIP metabolites in human urine. In addition to investigating individual variation, we examined the interaction of PhIP with a potentially chemopreventive food. In a preliminary study of the effect of broccoli on PhIP metabolism, we fed chicken to six volunteers before and after eating steamed broccoli daily for 3 days. Preliminary results suggest that broccoli, which contains isothiocyanates shown to induce Phases I and II metabolism in vitro, may affect both the rate of metabolite excretion and the metabolic products of a dietary carcinogen. This newly developed methodology will allow us to assess prevention strategies that reduce the possible risks associated with PhIP exposure.

Intestinal tumours induced by the food carcinogen 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine in multiple intestinal neoplasia mice have truncation mutations as well as loss of the wild-type Apc(+) allele.

C57BL/6J-Min/+ (multiple intestinal neoplasia) is a murine model for familial adenomatous polyposis (FAP), where the mice are heterozygous for a nonsense Apc(Min) (adenomatous polyposis coli) mutation, and therefore develop numerous spontaneous adenomas in the small intestine and colon. Neonatal exposure of Min/+ mice to the food carcinogen 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) (eight subcutaneous injections of 25 or 50 mg/kg PhIP to pups or 50 mg/kg PhIP to lactating dams) markedly increased (2--9-fold) the number of intestinal tumours, especially in the small intestine. We examined whether the Apc gene was affected in small intestinal and colonic tumours induced by PhIP. In spontaneous tumours formed in these mice, the main mechanism for tumour induction is loss of the wild-type Apc(+) allele, i.e. loss of heterozygosity (LOH). Also in the PhIP-induced tumours, this is a major mechanism, since large fractions of PhIP-induced tumours had LOH in APC: However, mechanisms other than LOH must also prevail, since a lower frequency of LOH was found in the small intestinal tumours from male mice exposed to PhIP either via breast milk (65%) or by direct injection (68%), compared with the untreated controls (92%). Tumours that had retained the wild-type Apc(+) allele were further analysed for presence of truncated Apc proteins with in vitro synthesized protein (IVSP) assay. Truncated Apc proteins, indicating truncation mutations in exon 15 of the Apc gene, were detected in 20% (8 of 40) of the tumours not showing LOH from the small intestine after PhIP exposure, all in segment 2 (codons 686--1217). Seventeen percent (2 of 12) of the colonic tumours had a truncated Apc protein in segment 3 (codons 1099--1693). Importantly, no truncated proteins were detected in tumours from unexposed mice with apparently retained wild-type Apc(+) allele. These results show that PhIP induces intestinal tumours in the Min/+ mice both by causing LOH and truncation mutations in the wild-type Apc(+) allele.

Age-dependent induction of aberrant crypt foci in rat colon by 2-amino-1-methyl-6-phenylimidazo [4,5-b]pyridine and azoxymethane.

Pups and adult rats received seven oral exposures (three time weekly) of the food mutagen PhIP (50 mg/kg), or two subcutaneous exposures (once weekly) of the experimental carcinogen azoxymethane (3.75 mg/kg). Aberrant crypt foci (ACF) were scored 8 weeks after the first exposure. In addition, lactating dams with suckling pups were orally exposed to 50 mg/kg of PhIP, three times weekly for three weeks. Direct PhIP exposure of pups induced 2.2 times more ACF than similar exposure of adult rats (2.0+/-0.0 versus 0.9+/-0.8, P<0.05). The growth of ACF, expressed as crypt multiplicity AC/ACF, was 3.5 times larger in neonatally exposed rats than in rats exposed in adulthood (8.0+/-7.3 versus 2.3+/-1.6, P<0.05). PhIP exposure via breast milk induced ACF in 3 of 25 animals. However, the difference versus controls, which had no ACF, did not reach statistical significance. Contrary to PhIP, azoxymethane induced more ACF in adult rats than in pups (2.8+/-1.9 versus 4.8+/-1.7, P<0.05). Similarly to PhIP however, azoxymethane induced 3.2 times larger ACF (AC/ACF) in pups than in adult rats (11.9+/-8.4 versus 3.7+/-1.9, P<0.001). Whereas no PhIP-induced ACF (0/15) were observed in the lymphoid follicles, approximately 60% of the azoxymethane-induced ACF (32/56) were located in these structures. This difference was statistically significant (P<0.001). The density of azoxymethane-induced ACF was 80 times larger in the lymphoid follicles than in the surrounding mucosa (P<0.01). Based on the assumption that the formation of ACF with high multiplicity is predicative for the tumour development we conclude that neonatal rats are more susceptible to PhIP and azoxymethane than adult rats.

Mammary gland carcinogenesis by 2-amino-1-methyl-6-phenylimidazo[4,5- b]pyridine in rats: possible mechanisms

2-Amino-1-methyl-6-phenylimidazo[4,5- b]pyridine (PhIP) is a heterocyclic amine derived from cooked meat. Mammary gland cancer can be induced in female Sprague–Dawley rats by administration of several oral doses of PhIP. The mechanism of mammary gland carcinogenesis by PhIP in this rat model is not fully understood but appears to involve several factors. One factor is the formation of PhIP–DNA adducts in the mammary gland after metabolic activation of PhIP. Possible target cell populations include the epithelial cells of the mammary gland terminal end buds (TEBs), putative sites of origin of carcinomas. Another factor involved in the mammary carcinogenicity of PhIP may be an increased proliferation in epithelial cells of the TEBs which occurs after a carcinogenic dose of PhIP is administered. This proliferation would be likely to enhance the fixation of mutations from PhIP–DNA adducts in target cells and facilitate the initiation of carcinogenesis. PhIP exposure also transiently inhibits the development of the mammary gland by retarding the differentiation of TEBs to alveolar buds and lobules. As a consequence, more TEBs are available for neoplastic transformation. Recent studies in rats have also shown that PhIP increases the levels of serum prolactin, a well-recognized promoter of mammary gland cancer, which may further explain the targeting of PhIP to the mammary gland. The results to date indicate that PhIP has multiple effects on the mammary gland and hormone status in rats that could potentially play a role in its ability to induce mammary gland cancer.

2-Amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) retards mammary gland involution in lactating Sprague-Dawley rats.

2-Amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP), a compound from cooked meat, is an established mammary gland carcinogen in female rats. Four doses of PhIP (150 mg/kg, p.o., once per day) were given to lactating Sprague-Dawley rats separated from their 10-day-old pups to initiate involution of the gland. Twenty-four hours after the last dose, apoptotic index in the mammary gland, as measured by the TUNEL assay, was significantly higher in the gland from control rats than in the PhIP-treated rats (4.757 +/- 1.066 versus 1.905 +/- 0.248%; P < 0.05). In comparison with controls, alveoli in the mammary gland of PhIP-treated rats were also visibly larger and contained more secretory epithelial cells. The expression of Bax, a stimulator of apoptosis, and Bcl-2, an inhibitor of apoptosis, were quantitated by western blotting. Accordingly, Bax expression was 2.7-fold higher in control rats, whereas Bcl-2 expression was 3.1-fold higher in PhIP-treated rats, both changes being statistically different (Student's t-test, P < 0.05). Immunohistochemistry further confirmed a lower expression of Bax and higher expression of Bcl-2 in secretory alveolar epithelial cells of the PhIP-treated mammary gland. The findings are consistent with the notion that exposure to PhIP retarded involution via partial inhibition of programmed cell death. To investigate possible mechanisms for the inhibitory effects of PhIP on mammary gland involution, serum levels of prolactin, an important hormone for the maintenance of lactation, were measured in virgin rats with regular estrous cycles given PhIP (150 mg/kg, p.o.) on the morning of diestrous. After one estrous cycle, on proestrous morning, serum prolactin levels were 1.3-fold higher after PhIP than after control vehicle (one-way ANOVA, Fisher LSD multiple comparison test, P < 0. 05). PhIP exposure during involution was associated with the induction of benign mammary tumors. Seven out of 12 rats developed fibroadenomas, and one developed a tubulopapillary carcinoma within 1 year of receiving PhIP administration during involution (150 mg/kg, p.o., once per day for 5 days), and a high-fat diet (23.5% corn oil). An increase in serum prolactin level and the effects on mammary gland apoptosis seen with PhIP may have implications for the mechanisms of carcinogenic targeting of PhIP to the mammary gland.

The identification of [2-14C]2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine metabolites in humans

[2-(14)C]2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine ([14C]PhIP), a putative human carcinogenic heterocyclic amine found in well-done cooked meat, was administered orally to three colon cancer patients undergoing a partial colonectomy. Forty-eight to seventy-two hours prior to surgery, subjects received a 70-84 microg dose of 14C. Urine and blood were analyzed by HPLC for PhIP and PhIP metabolites. Metabolites were identified based on HPLC co-elution with authentic PhIP metabolite standards, mass spectral analysis and susceptibility to enzymatic cleavage. In two subjects, approximately 90% of the administered [14C]PhIP dose was eliminated in the urine, whereas in the other, only 50% of the dose was found in the urine. One subject excreted three times more radioactivity in the first 4 h than did the others. Twelve radioactive peaks associated with PhIP were detected in the urine samples. The relative amount of each metabolite varied by subject, and the amounts of each metabolite within subjects changed over time. In all three subjects the most abundant urinary metabolite was identified as 2-hydroxyamino-1-methyl-6-phenylimidazo[4,5-b]pyridine-N2-glucuron ide (N-hydroxy-PhIP-N2-glucuronide), accounting for 47-60% of the recovered counts in 24 h. PhIP accounted for <1% of the excreted radiolabel in all three patients. Other metabolites detected in the urine at significant amounts were 4-(2-amino-1-methylimidazo[4,5-b]pyrid-6-yl)phenyl sulfate, N-hydroxy-PhIP-N3-glucuronide and PhIP-N2-glucuronide. In the plasma, N-hydroxy-PhIP-N2-glucuronide accounted for 60, 18 and 20% of the recovered plasma radioactivity at 1 h post PhIP dose in subjects 1, 2 and 3 respectively. Plasma PhIP was 56-17% of the recovered dose at 1 h post exposure. The relatively high concentration of N-hydroxy-PhIP-N2-glucuronide and the fact that it is an indicator of bioactivation make this metabolite a potential biomarker for PhIP exposure and activation. Determining the relative differences in PhIP metabolites among individuals will indicate metabolic differences that may predict individual susceptibility to carcinogenic risk from this suspected dietary carcinogen.

Proliferation, development and DNA adduct levels in the mammary gland of rats given 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine and a high fat diet.

2-Amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) is a heterocyclic amine derived from cooked meat that is a mammary gland carcinogen in rats. A carcinogenic dose-regimen of PhIP (75 mg/kg, p.o., 10 doses, once per day) was administered to 43-day old female Sprague-Dawley rats, and the rats were then placed on a defined high fat (23.5% corn oil) or low fat (5% corn oil) diet for up to 6 weeks. At various times after carcinogen and diet, and prior to carcinogenesis, we examined the percentage of proliferating cells in terminal end bud (TEB) epithelial structures of the rat mammary gland by proliferating cell nuclear antigen staining, mammary gland architecture by whole mounting, and PhIP-DNA adduct levels in mammary epithelial cells by the 32P-post-labeling assay. Immediately after dosing, the percentage of proliferating epithelial cells in TEBs was significantly higher in PhIP-treated rats than in control rats receiving vehicle only [7.5 +/- 0.9% (n = 99) versus 4.2 +/- 0.6% (n = 127), respectively]. The mammary glands of PhIP-treated rats showed a significantly lower density of alveolar buds (ABs) and a higher density of TEBs than control rats, which suggests that PhIP exposure partially inhibited the normal glandular differentiation of TEBs to ABs. After 6 weeks on the diet, proliferation in TEBs was statistically higher in rats given PhIP plus a high fat diet than in rats given vehicle plus a low fat diet. The mammary glands from rats on a high fat diet also showed a statistically higher density of TEBs when compared with rats on a low fat diet [2.08 +/- 0.34% versus 1.04 +/- 0.20%, respectively (n = 6)]. PhIP-DNA adduct levels were relatively high in mammary epithelial cells of treated rats. At 3 h after the last dose of PhIP, DNA adduct levels [relative adduct labeling (RAL) x 10(7), mean +/- SE] were 10.5 +/- 1.7 (n = 8) and 0.9 +/- 0.2 (n = 7) in epithelial cells isolated from mammary gland and in the liver, respectively. DNA adduct removal rates from the mammary gland were not different between rats on the high fat and low fat diets. Adducts were still detected after 6 weeks on either diet. Thus, events that occurred prior to neoplasia in the mammary glands of PhIP-treated rats include formation of PhIP-DNA adducts at relatively high levels, and enhanced proliferation in TEBs (putative sites of origin of mammary gland carcinomas) and partial inhibition of TEB differentiation. The high fat diet, a promoter of PhIP-induced mammary gland carcinogenesis, appeared to sustain the proliferative effect of PhIP in mammary gland TEBs at a time when PhIP-DNA adducts are still detectable. These early events may contribute to the targeting and carcinogenicity of PhIP to the mammary gland of rats.

DNA adducts of heterocyclic amines: formation, removal and inhibition by dietary components

The dietary mutagens 2-amino-3-methylimidazo[4,5- f]quinoline (IQ) and 2-amino-1-methyl-6-phenylimidazo[4,5- b]pyridine (PhIP) are carcinogenic in rodents. In F344 rats PhIP induces mammary tumors in females and colon tumors in males, while IQ induces tumors principally in the liver, Zymbal gland and intestines. In CDF 1 mice, IQ induces liver, lung and forestomach tumors. We have evaluated the dynamics of formation, removal and inhibition of PhIP– and IQ–DNA adducts in these rodents. After bolus doses (50 mg/kg, by gavage) of IQ or PhIP, both IQ– and PhIP–DNA adducts were removed rapidly from both target and nontarget organs, while after 3–4 weeks of feeding IQ or PhIP (0.01–0.04%) adduct removal was much slower. Gavaging of male F344 rats with PhIP (0.1–1000 μg/kg/day) for 23 days resulted in accumulation of PhIP–DNA adducts in various organs, but adducts were detectable only at 100 or 1000 μg/kg/day. Urinary excretion of unchanged PhIP was a constant proportion (1.6–2.1%) of the daily dose over the entire dose range and was independent of duration of exposure. When weanling female F344 rats were exposed to dietary PhIP (0.01–0.04%) for 1–4 weeks, the presence of either conjugated linoleic acid (CLA; 0.1–1.0%) or indole-3-carbinol (I3C; 0.1%) in the diet inhibited PhIP–DNA adduct formation (58–99%) in various organs, including the mammary gland and the colon. Similarly, the inclusion of 0.075% 4-ipomeanol (IPO) in the diet of male CDF 1 mice exposed for 3 weeks to dietary IQ (0.01%) resulted in inhibition of IQ–DNA adduct formation (30–59%) in the target organs (liver, lungs, stomach) but not in a number of other organs. It is concluded that (1) the rate of PhIP– and IQ–DNA adduct removal depends on the dose and frequency of administration, (2) urinary PhIP may be a good biomarker of recent PhIP exposure and (3) CLA, I3C and IPO are potential chemopreventive agents against PhIP- or IQ-induced tumors in rodents.