(1) Seminar on “Heterocyclic Amines: New Insights into Risk for Humans”
This workshop was held on March 15-16, 1995, at the Marriott Hotel, Maui Hawaii. The organizers were Dr. Keiji Wakabayashi, National Cancer Center Research Institute. Tokyo, and Dr. Snorri S Thorgeirsson, National Cancer Institute, Bethesda, Maryland. There were nine participants from the United States and seven from Japan
Dr. Keiji Wakabayashi welcomed the attendees and gave the introductory remarks. He provided interesting historical perspective on the discovery of the food-derived heterocyclic amines and reviewed the substantial progress made toward understanding the formation, chemistry, and the biological effects of these compounds. He also pointed out the important role that the US-Japan program has played in the research on food-derived heterocyclic amines.
The first report by Dr. Herman A.J. Schut addressed the subject of repair and inhibition of DNA adducts of heterocyclic amines. It has been shown that both the heterocyclic amines 2-amino-3-methylimidazo[4,5-f]quinoline (IQ) and 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) are carcinogenic in CDF1 mice and Fischer-344 rats. After administration in vivo, both IQ and PhIP form DNA adducts in target organs, as well as in nontarget organs of both species. Using 32P-postlabeling analysis Dr. Schut examined whether adduct persistence can be related to target organ susceptibility and whether adduct formation can be prevented by adding inhibitors to the diet. He found that, with a couple of exceptions, neither the extent of adduct formation, nor its rate of removal is related to target organ susceptibility. It is possible, however, to use adducts in white blood cells as surrogate markers for overall exposure to IQ and PhIP. Dr. Schut also showed that adduct removal depends on the way IQ and PhIP are administered, i.e., as a single oral dose, or at continuous low doses in the diet. Adding CLA (conjugated linoleic acid) to the diet inhibits adduct formation in both target and nontarget organs. Similarly, feeding a diet containing 20% (w/w) menhaden oil (a fish oil, rich in 2-3 fatty acids) inhibits adduct formation in various organs. Thus, adduct formation is not always related to target organ sensitivity to heterocyclic amines. Adduct formation, however, can be prevented, at least in part, by changes in the type and quantity of fatty acids in the diet.
Dr. Yasushi Yamazoe presented studies on activation and inhibition of carcinogenic N-substituted arylamines. Dr. Yamazoe pointed out that organic chemicals in our environment may be important etiological factors in human cancer. Although the amounts are low, N-substituted aryl compounds such as arylamines and nitroarenes are contained in our daily foods or present as air pollutants. These chemicals are biologically activated first via cytochrome P450 oxidation into N-hydroxy derivatives, and further activated to their ultimate derivatives by O-esterifications involving O-acetylation and sulfation. Dr. Yamazoe has recently isolated five different cDNAs from liver cDNA libraries of rats and humans, which code for sulfotransferases involved in the biotransformation of N-hydroxyarylamines. Rat cDNAs expressed in COS-1 cells (ST1A1, ST1B1 and ST1C1) showed 28-30 kDa in SDS-PAGE, and catalyzed sulfations of N-hydroxy-arylamines as well as phenols. The different sulfotransferases, however, show distinct substrate specificities. A typical hepatocarcinogen, N-hydroxy-2-acetylaminofluorene, is known to produce liver tumors at higher rates in males rats than in female rats. Consistent with this observation, the activation of N-hydroxy-2-acetyl-aminofluorene was more than 70 times higher in the presence of a male-specific form, ST1C1 (HAST1), than ST1A1 (ST1V). No hybridizable band with ST1C1 cDNA was detected in human livers. Thus, human livers show low levels of PAPS-dependent DNA binding of N-hydroxy-2-acetylaminofluorene. However, PAPS-dependent activation of N-hydroxy-PhIP was higher in liver cytosols of humans than rats. Human sulfotransferases, ST1A2 and ST1A3, mediated the activation of N-hydroxy-PhIP at higher rates than the rat counterpart, ST1A1. Dr. Yamazoe’s work has demonstrated a wide diversity of sulfotransferase-mediated activation of N-hydroxyarylamines in human livers.
Dr. Nicholas P, Lang next described studies on human metabolic polymorphisms and cancer risk related to the heterocyclic amines. Dr. Lang indicated that variations in cancer incidence rates among human populations has generally been ascribed to variations in exposure to putative carcinogens. However, animal data has clearly demonstrated that exposure alone is insufficient to explain observed variations in tumor incidence among strains and species
Dr. Lang then discussed comparative studies of the heterocyclic amine PhIP, in mouse, rat and human tissue that have help elucidate the mechanisms of inter-species and organ-specific differences in genotoxicity and carcinogenesis Incubation of PhIP with hepatic microsomes demonstrated the conversion to the genotoxic metabolite, N-hydroxy-PhIP by humans at a rate 1.8- and 1.4-fold higher than rats and mice respectively. Rodent microsomes also converted PhIP to its non-genotoxic 4’-hydroxy derivative. This activity was negligible in human hepatic microsomes.
The further activation of N-hydroxy-PhIP by cytosolic O-acetyltransferase was examined using PhIP-DNA binding as a measure of bioactivation. Acetyl CoA-dependent DNA binding of N-hydroxy-PhIP was detected with both human and rodent hepatic cytosols and showed a significant inter-species difference. The enzyme activity in rats (65.6 ± 9.7 pmol bound/mg DNA/30 min, N=3) was similar to that in human rapid acetylators slow acetylators (59.9 ± 6.2, N=3), but was 2.6-fold higher than that in human slow acetylators (25.5 ± 5.7, N=7); p < 0.01). Mice had the lowest O-acetyl-transferase activity for the activation of N-hydroxy-PhIP, which was only 11% of that in rats and 12% and 29% of that in human rapid and slow acetylators, respectively. Human and rat hepatic cytosols exhibited higher levels of glutathione S-transferase (GST) activity that significantly inhibited the O-acetyltransferase-mediated DNA binding of N-hydroxy-PhIP; no such effect was observed by mouse hepatic cytosols. Moreover, the GSTS had no inhibitory effect on sulfotransferase-mediated or ATP-dependent DNA binding of N-hydroxy-PhIP.
Dr. Lang’s results suggest that the differential metabolic activation and detoxification of PhIP in rats and mice are likely to be critical factors in determining the species differences and organ specificity to PhIP genotoxicity and carcinogenesis.
By evaluating naturally occurring variations in exposure and in metabolism in a group of human subjects, Dr. Lang has identified an association between variations in exposure and in metabolism and increased risk for the development of colon cancer or polyps. This study demonstrated that the lowest risk phenotype (slow NAT2 and rapid CYP1A2) combined with low exposure to heterocyclic amines (through the consumption of meat that is rare to medium well done), provided a relative odds ratio of 1 while the highest risk phenotype (rapid NAT2 and rapid CyP1A2) combined with high exposure (through the consumption of meat well done) had a relative odds ratio of 6.45. Dr. Lang indicated that studies in progress will extend this research to include the roles of sulfotransferase and GSTµ in addition to NAT2 and CYP1A2 in human cancer risk following exposure to heterocyclic amines.
It is clear that continued study of the activation as well as the detoxification pathways for heterocyclic amines in humans has the potential to yield information that will allow reduction of cancer risk through modification in dietary habits and cooking practices
Dr. Tetsuya Kamataki discussed the role of CYP3A7 and N-hydroxyacetyl-transferase in the activation of heterocyclic amines in human fetuses. Heterocyclic amines such as IQ and MeIQ show genotoxicity only after undergoing metabolic activation mainly by CYP1A2 and polymorphic N-acetyltransferase in adult human livers. Dr. Kamataki described recent studies that show that these promutagen/carcinogens are activated by CYP3A7 rather than CYP1A in human fetal livers.
Dr. Kamataki has purified four forms of cytochrome P-450, namely P450HFLa-d, and showed that P450HFLa was one of the major forms of cytochrome P450 in human fetal livers. Also, results of analysis of nucleotide sequence of a cDNA clone obtained by using the antibodies to P450HFLa as a probe indicated that this form of cytochrome P450 belonged to the CYP3A family and named as CYP3A7
To further examine the toxicological significance of CYP3A7, Dr. Kamataki established a cell line expressing CYP3A7. CHL cells derived from Chinese hamster lung was transformed with a plasmid carrying cDNA encoding CYP3A7. The expression of CYP3A7 in the transformants was confirmed by Northern and Western blot analyses and by their ability to activate aflatoxin B1. Addition of!!
!-naphthoflavone, a known activator of CYP3A, enhanced the sensitivity of cells to aflatoxin B1. Triacetyloleandomycin a suicidal inhibitor of CYP3A, reduced the sensitivity. These results support that CYP3A7 enzyme expressed in the cells activated the mycotoxin. Interestingly, cell lines expressing CYP3A7 together with NADPH-cytochrome P450 reductase showed high sensitivity to aflatoxin B1, while they did not show any significant sensitivity to heterocyclic amines. However, introduction of cDNA coding for polymorphic N-acetyltransferase (NAT2) but not monomorphic N-acetyltransferase (NAT1) to the cells expressing CYP3A7 resulted in the induction of cytotoxicity upon addition of heterocyclic amines to a culture medium. Dr. Kamataki’s results indicate that in the established cell lines expressing both CYP3A7 and polymorphic NAT2 can activate heterocyclic amines and cause both cytotoxicity and genotoxicity. Dr. Rashmi Sinha described interdisciplinary epidemiologic investigations into the role of heterocyclic amines in human cancer. Heterocyclic aromatic amines (HAs), byproducts of meats cooked at high temperature, are known mutagens and animal carcinogens, although evidence from population studies is limited. In order to perform the required interdisciplinary epidemiologic investigations, Dr. Sinha has embarked on a series of methodologic studies to create the tools necessary to elucidate the carcinogenic effects of HAs in humans. First, she developed a set of questions on methods used to cook meats. This is a modification of a widely used food frequency instrument, the Health Habits and History Questionnaire. Questions on methods of cooking meats, internal doneness and external browning were embedded within the frequency questions are a meat cooking module. Photographs of the meats were taken to help subjects recognize the way they prefer their meat prepared. This module is being used by investigators at the National Cancer Institute and other research centers to determine whether well-done meat consumption is a risk factor for colon polyps, and cancers of the colon, breast, lung, pancreas, and prostate. Second, to quantify HAs consumption in individuals, Dr. Sinha has established a database for 5 HAs, similar to nutrient content of foods. The database was designed to link the type of meat with the cooking method and degree of doneness and yield HA values. For each method of cooking and level of doneness, meat was cooked at five separate times and then pooled to make one composite sample which was analyzed. During meat preparation, careful measurements of internal temperature of meats, temperature of the cooking surface, and amount of loss of fat and moisture were recorded. The values from this database will be linked to the questionnaire responses to permit a more specific determination of the respondents exposure to HAs. Third, Dr. Sinha described a metabolic study carried out in 66 healthy non-smokers (33 males and 33 females) to examine biomarkers of HAs which could be used in cancer etiological studies. The roles of genetic variation in cytochrome P4501A2 (CKYP1A2) and acetyltransferase (NAT2) were also investigated. NAT2 activity remained unchanged through the study but CYP1A2 activity increased in most of the subjects after consuming meat cooked at high temperature, suggesting induction by some compound(s) formed during high temperature cooking (not polycyclic aromatic hydrocarbons as these were equal in both periods). There was a high within-person correlation for CYP1A2 such that subjects with low activity after eating low temperature cooked meat tended to stay low even after consuming high temperature cooked meat, and the subjects who were high tended to stay high. The results from Dr. Sinha’s study indicate both a fixed and an inducible component in the regulation of CYP1A2. She also found increasing CYP1A2 activity associated with decreasing levels of 2-amino-3,8-dimethylimidazo[4,5f]quinoxaline (MeIQx) in the urine when adjusted for amount of meat eaten, suggesting CYP1A2 is metabolizing MelQx to the N-hydroxy metabolite. Thus, if HAs are human carcinogens, then meat cooked at high temperatures may pose an increased cancer risk as it contains both inducers of CYP1A2 and procarcinogens IQx, DiMeIQx, and PhIP) known to be activated by this enzyme. Dr. Sinha suggested that these types of investigations may provide the basis for further etiologic studies that will clarify the role of HAS in human carcinogenesis.
Dr. Hikoya Hayatsu presented recent results on direct interactions of reagents with heterocyclic amines. Recent interest in the inhibitory effect of chlorophyllin against genotoxic actions of HAs has prompted Dr. Hayatsu to study further the mechanisms underlying the inhibitions. In 1980, Dr. Hayatsu reported the inhibition of the HA-mediated Salmonella-mutagenesis by hemin and chlorophyllin, and speculated that the inhibition was a result of complexing between HAs and the pigments. The complex formation between chlorophyllin and HAs was then given ample experimental evidence, and the inhibitions found with the Salmonella and the Drosophila assays were ascribed to the complexing.
Dr. Hayatsu has now found that Fe-chlorin e6-Na3 can mediate a chemical conversion of metabolically activated forms of Has. Using samples of metallo-chlorins with high purity, he explored their suppressive effects against the directly acting mutagenicity of N-hydroxylated Trp-P-2 (Trp-P-2(NHOH)). Fe-chlorin e6-Na3 showed strong inhibition: at 3 x l0-8 M, with an equimoiar Trp-P-2(NHOH) in the prelncubation mixture, it showed 50% suppression of the mutagenicity (TA98, - S9). Dr. Hayatsu analyzed the reaction mixture with the use of HPLC and ultraviolet spectroscopy, and found that the Trp-P-2(NHOH) was converted within a few minutes into Trp-P-2(NO), a nitroso derivative, the mutagenicity of which is 1/4 of that of Trp-P-2(NHOH). He had previously shown that superoxide dismutase (SOD) can convert hydroxylamino-HAs into nitroso-HAs under aerobic conditions. With an assay for an SOD activity using the formazan formation from nitroblue tetrazolium, Dr. Hayatsu has now detected an SOD activity in Fe-chlorin e6-Na3. No activity was found in Cu-chlorin e6-Na3 or FeCl3. Dr. Hayatsu has also found that the mutagenicity of Trp-P-2(NO) itself was subject to inhibition by Fe-chlorin e6-Na3.
Dr. Hayatsu concluded that the strong suppressive effects of Fe-chlorin e6-Na3 against hydroxylamino-HAs are ascribable to the SOD-like function of the pigment, coupled with its complexing ability with the planar surface of the HA molecules.
Dr. James Felton discussed recent studies on genetic toxicology and human dosimetry of heterocyclic amines from cooked foods. Dr. Felton indicated that mutagenic/carcinogenic heterocyclic amines are generated in food when cooked over 150°C. These compounds are present from 0.1 to 200 ppb depending on the food and the cooking conditions. Dr.. Felton also showed that heterocyclic amines are not only present in cooked red meat, fish, and chicken, but are also present at lower levels in baked and fried grain products. Mutagenic activity of fried ground beef cooked at 230°C for 10 minutes per side has 800 ± 37 TA98 revertants per gram cooked weight. Barbecued beef patties had 9130 ± 1890 TA98 revertants per gram but the heterocyclic amine content of 3.0 ± 1.8 ng/g MeIQx, 3.1 ± 1.3 ng/g DiMeIQx, 30 ± 9.7 ng/g PhIP, and 1.6 ± 1.0 ng/g IQ accounted for only 28% of the mutagenic activity. PhIP in beef fried at temperatures from 170 to 250°C is formed at a rate of about 0.1 nanograms per °C. In contrast, IQ only increases 0.006 ng per °C, suggesting that PhIP formation is more dependent on temperature.
Dr. Felton described studies using Chinese hamster ovary cells to assess the role of specific metabolic enzymes and DNA repair on cytotoxicity and mutation frequency and spectra. He showed that PhIP produces very specific mutations in exons 2 and 3 of the aprt gene in these cells. Seventy percent of the mutations seen (all transversions) are at 3 base-sites in this gene. However, spontaneous mutations showed none of these lesions. This particular cell line expressed cytochrome P450IA1 but lacked excision repair. Addition of a human NAT-2 expressed gene to these cells had no effect on the cytotoxicity of PhIP, but it caused a marked increase (1000-fold) in the cytotoxicity with IQ, a much weaker mutagen in cells without NAT-2.
Dr. Felton discussed data on the average consumption of foods containing heterocyclic amines determined by statistically analyzing the intakes of 3563 individuals who provided 3 day dietary records in a USDA sponsored random survey of the US population during 1989. Dietary intakes were determined to be in the following descending order PhIP > AC > MeIQx > DiMeIQx > IQ. An upper-bound estimate of the incremental cancer risk is 1.1 x 10-4, based on body surface areas. Dr. Felton concluded that consumption of meat and fish products contributed the most risk.
Dr. Keiji Wakabayashi presented studies on human exposure to mutagenic and carcinogenic heterocyclic amines. Twenty-three heterocyclic amines (HAs) have been identified as mutagens to S. typhimurium strains in a variety of heated materials. Among these, ten have been proven to be carcinogenic in rodents. Dr. Wakabayashi has estimated human exposure levels to carcinogenic HAs, by measuring their amounts in various kinds of cooked foods and in urine samples of healthy volunteers. Based on these data, the daily exposure to 2-amino-3,8-dimethylimidazo[4,5f]quinoxaline (MeIQx) was estimated to be 0.2 µg-2.6 µg per person. He also analyzed MelQx-DNA adducts in human surgical and autopsy tissue samples by 32P-postlabeling method under adduct intensification conditions modified to include additional digestion with nuclease P1 and phosphodiesterase I after kination with 5’-hydroxyl termini. DNA samples from colon, rectum an kidney were found to contain an adduct spot corresponding to the authentic 5’-pdG-C8-MeIQx by TLC and HPLC, and their levels were 18, 14 and 1.8 per 1010 nucleotides, respectively. In contrast, no PhIP-DNA adducts were detected in human tissues.
Dr. Wakabayashi discussed studies on!!
!-carboline compound, norharman, that is produced by heating L-tryptophan and is known to be present in cooked foods and in cigarette smoke at levels even higher than the carcinogenic Has. Norharman is not mutagenic to Salmonella strains, but becomes mutagenic to S, typhimurium TA98 with S9 mix only in the presence of non-mutagenic aromatic amines including aniline and o-toluidine. Dr. Wakabayashi examined whether DNA adducts were formed in DNA of S. typhimurium TA98 by treatment with norharman and aromatic amines using 32P-postlabeling analysis. When norharman was incubated with aniline for 6 hr at 37°C, three adduct spots were detected at a total level of 10.8/108 nucleotides. In the case of norharman with o-toluidine, three adduct spots were detected and their levels totaled 3.7/108 nucleotides. However, norharman, aniline and o-toluidlne did not themselves produce any DNA adduct spots. Dr. Wakabayashi concluded that DNA adduct formation by norharman with aromatic amines is related to co-mutagenic action of norharman in S. ty:phimurium TA98. Dr. W. G. Stillwell described recent studies on the analysis of heterocyclic amines and their conjugates in human urine. The heterocyclic aromatic amines formed in cooked meat represent an important class of mutagenic/carcinogenic substances. Exposure to these compounds in humans may be estimated by measuring the excretion of the unchanged amine and/or the conjugates in urine. The heterocyclic aromatic amine MeIQx and related food-borne chemicals are bio-transformed in animals to give a variety of detoxified as well as genotoxic products, including ring- and amine-hydroxylated metabolites, as well as sulfamate, sulfate, and glucuronide conjugates. In determining exposure of humans to food-borne mutagens GC-MS procedures have primarily been used for the analysis of the heterocyclic aromatic amines in urine. Many of the conjugates, however, due to their polarity and thermal instability are not amenable to analysis by conventional mass spectrometry. Electrospray ionization (ESI) LC-MS affords a useful technique for the direct analysis of the aromatic amine and its conjugates. Selected ion monitoring procedures and the use of narrowbore or microbore HPLC columns provide very good detection limits in the picogram range for the standard compounds. Dr. Stillwell has analyzed by ESI-LC-MS urine of MeIQx-treated rats for the presence of polar conjugates and have begun analytical studies of human urine. She described a new method for the analysis of the food mutagen PhIP in human urine using ESI-LC-MS. Dr. Stillwell discussed preliminary results of a comparative study of urinary excretion levels of MeIQx and PhIP in ethnically-diverse subpopulations of Los Angelos County. The excretion of MeIQx was highest in blacks followed by Asians and lowest in whites.
Dr. Roderick Dashwood discussed approaches and blocking agents employed to inhibit colon carcinogenesis induced by heterocyclic amines. Aberrant crypt foci (ACF) are preneoplastic lesions that have been detected in human colon resections, and in animals treated with such carcinogens as azoxymethane and dimethylhydrazine. Dr. Dashwood used the ACF assay as a screen for potential chemopreventive agents against the heterocyclic amine colon carcinogens IQ and (PhIP). Typically, IQ or PhIP are given to male F-344 rats by gavage on alternating days in weeks 3-4, and ACF are read after week 16. Using this carcinogen exposure protocol, Dr. Dashwood identified mutations in the Ki-ras gene of IQ-induced ACF. In chemoprevention studies, inhibitors have been given before and during carcinogen treatment to examine “blocking agent” activity, or starting one week after the carcinogen exposure (i.e. during the post-initiation phase) to detect “suppressing agents.” Conjugated linoleic acids (CLA), indole-3-carbinol (I3C), and chlorophyllin (CHL) have been identified as promising inhibitors in these assays. Dr. Dashwood has confirmed the inhibitory activity of CHL in the ACF assay in a one year carcinogenicity bioassay, in which the tumor incidence was reduced not only in the colon but also in the small intestine, liver and Zymbal’s gland. I3C offered potent inhibitory activity against PhIP-induced ACF, DNA adducts, and metabolic activation in the male and female F-344 rat. The mechanism by which CLA inhibits IQ-induced ACF and DNA adduct formation in the colon is under investigation. Trapping/interception of the procarcinogen, scavenging of electrophiles, and induction of detoxification pathways appear to abe minimal. However, Dr. Dashwood has identified the inhibition of enzyme that activate the heterocyclic amines (cytochrome P4501A2, prostaglandin H synthase) as potential mechanisms.
Dr. Katsumi Imaida presented recent data on chemoprevention of heterocyclic amine-induced carcinogenesis in rats by antioxidants. Chemoprevention of carcinogenesis has been extensively investigated in a variety of organs in animal models. Dr. Imaida has investigated the effects of several chemopreventive agents, most of them antioxidants, on organ carcinogenesis induced by heterocyclic amines.
The liver medium-term bioassay for carcinogens is suitable for prediction of carcinogenic, promoting and inhibiting effects of test compounds in a relatively short period (8 weeks). In the study described by Dr. Imaida, F344 rats were injected diethylnitrosamine (DEN) i.p. at 200 mg/kg, b.w.; two weeks later animals were placed on a basal diet containing 2-amino-6-methyldipyridol[1,2-a:3’,2’-d]imidazole (Glu-P-1). Concomitant treatments of Glu-P-1 with several antioxidants, such as 1-O-hexyl-2,3,5-trimethylhydroquinone (HTHQ), green tee catechins (GTC),!!
!-tocopherol,!!!-carotene, d-limonene were also performed. At week 3, all animals underwent a 2/3 partial hepatectomy, and at week 8, all animals were sacrificed and glutathione S-transferase placental form (GST-P) positive foci in the liver were measured. Glu-P-1 itself significantly increased the numbers and areas of GST-P positive foci in the liver. However, significant inhibitory effects were shown by the concomitant administration of HTHQ, GTC and!!!-carotene. Furthermore, inhibitory effects by HTHQ were dose dependent. PhIP is carcinogenic for the mammary gland of female F344 rats. Dr. Imaida investigated the chemopreventive effects of several antioxidants on the mammary gland carcinogenesis induced by PhIP in female F344 rats. Rats were treated with PhIP in the diet at a dose of 200 ppm in diet for 52 weeks. Antioxidants, such as HTHQ, GTC,!!
!-tocopherol, chlorophillin, ellagic acid were concomitantly administered with PhIP in the diet at doses of 0.5 or 1.0%. The incidence of palpable tumors of the mammary gland was reduced with treatments of HTHQ,!!!-tocopherol, and chlorophyllin. Dr. Imaida concluded that some antioxidants have inhibitory activity on heterocyclic amine induced liver and mammary gland carcinogenesis in rats. HTHQ showed the most potent chemopreventive activity of the compounds.
Dr. Elizabeth Snyderwine gave an update on recent studies on heterocyclic amines in the Laboratory of Experimental Carcinogenesis (LEC) at NCI. Studies in the LEC, most notably with nonhuman primates, supported that metabolic processing is an important factor in the carcinogenicity of food-derived heterocyclic amine (HAs). In cynomolgus monkeys, IQ and MeIQx differ significantly in extent of metabolic activation and in carcinogenic potency. In nonhuman primates and other models metabolic activation of the HAs proceeds by N-hydroxylation followed by phase II esterification. The reactive metabolites form DNA adducts that are responsible for mutagenicity and genotoxicity, and are associated with the cardiotoxic effects of these compounds. As a means of assessing the role of HA-DNA adducts in the initiation of carcinogenesis. Dr. Snyderwine has investigated the mutagenic specificity of IQ- and PhIP-DNA adducts in the supF shuttle vector system. The findings support a direct involvement of PhIP-guanine adducts in G to T transversion mutations, that might, in critical genes in vivo, be responsible for initiation of carcinogenesis. Most recently Dr. Snyderwine has been studying the metabolic processing, DNA adduct formation, and mammary gland carcinogenicity of PhIP in female Sprague-Dawley rats and the influence of dietary fat on tumorigenicity. These studies have shown that a 10-dose regimen of PhIP induced mammary gland cancer within 27 weeks and that dietary fat enhanced the incidence and severity of the tumors. In rats treated with PhIP, PhIP-DNA adducts were detected in mammary epithelial cells, the putative target site for mammary gland cancer. The mammary gland has a high phase II O-acetyltransferase activity but a very low capacity toward N-hydroxylation. This high O-acetyltransferase activity might account for the formation of PhIP-DNA adducts in the mammary gland from circulating N-hydroxy-PhIP generated by hepatic metabolism. In human models including cultured human mammary epithelial cells and mammary gland samples from reduction mammaplasty, O-acetyltransferase activation of N-hydroxy-PhIP has been observed. Dr. Snyderwine concluded that to better evaluate the risk of human mamrnary gland cancer from dietary HAs, studies clarifying the HA metabolites distributed to the mammary gland are still needed.
Dr. Tomoyuki Shirai described studies on accelerated mammary carcinogenesis by transplacental, neonatal, and subsequent dietary exposure to PhIP in rats. A food-derived carcinogen, PhIP has been shown to induce mammary and colon carcinomas in rats when given orally. Dr. Shirai indicated that from environmental viewpoint, the potential for continuous exposure to PhIP from prenatal stages is very important. The present experiment was thus carried out in order to assess the carcinogenic effects of PhIP when given maternally and post-neonatally. Female Sprague-Dawley rats were administered PhIP in the diet at a dose of 100 ppm for a total of 15 weeks from 7 weeks of age. They were mated with non-treated males when they were 14-weeks old and delivery and weaning of offspring were completed during the administration period. As controls, additional females were given the basal diet without PhIP and delivered of offspring similarly to the treated animals. The offspring from each group were nursed by their own dams and subdivided into three dietary groups for each sex at weaning, receiving 100 ppm or 25 ppm of PhIP or basal diet. The incidence of mammary carcinomas in females at 47 weeks of age was highest in the progeny from the PhIP-treated dams thereafter given 100 ppm PhIP (70.0%: 14/20 rats). The value was significantly higher than the 17.1% (6/35 rats) for the basal diet group from PhIP-treated dams, but not different from the 61.9% (13/2 1 rats) of the 100 ppm diet group from non-treated dams. In the 25 ppm PhIP treated offspring, however, the mammary carcinoma incidence was significantly higher (P<0.01) in the progeny of PhIP-treated dams (45.0%: 9/20 rats) than those from the non-treated dams (4.8%: 1/21 rats). In the basal diet groups, the mammary carcinoma incidence in the progeny of PhIP-treated dams (17.1%) was higher than in the non-treated dams (3.3%: 1/30 rats), but the difference was not statistically significant. Similar trends were observed for tumor multiplicity. Histologically, most palpable tumors were adenocarcinomas. In male offspring, while a few carcinomas were induced in the 100 ppm groups, no difference was observed between progeny from the PhIP-treated and non-treated mothers. Dr. Shirai concluded that the present results demonstrate a clear increase in mammary tumor induction in female rat after transplacental and neonatal exposure to PhIP. He also confirmed excretion of PhIP into the breast milk as reported earlier. Dr. Shirai noted that no macroscopic intestinal tumors were found in either dams or offspring in the present experiment.
Dr. RJchard H. Adamson discussed studies on the effects of heterocyclic amines in nonhuman primates. Three HAs were selected for studies of metabolic processing, DNA adduct formation, carcinogenesis studies and other toxicologic effects in nonhuman primates, primarily cynomolgus monkeys. The three compounds selected for extensive studies were IQ, 8-MeIQx, and PhIP. Several factors were used in selecting the compounds including structure, their concentrations in cooked food, their mutagenic activity, the carcinogenic spectrum in rodents and the availability of the compound. Thus far, IQ has induced liver tumors in 100% of the monkeys at the 20 mg/kg dose and in 70% of the monkeys at the 10 mg/kg dose. Studies on the metabolic processing and DNA adduction of all three compounds will be presented. Dr. Adamson concluded that based on current data, it is likely that 8-MeIQx will not be carcinogenic in cynomolgus monkeys, but PhIP is likely to be carcinogenic.
Dr. Takashi Sugimura presented recent studies on the genetic alteration induced in rats and mice by food-derived heterocyclic amines, Heterocyclic amines (HAs) are produced by heating fish flesh and meat. Heating creati(ni)ne, sugar and amino acid results in production of highly mutagenic HAs. PhIP is the most abundant HA in cooked meat. and Japanese and Swedish intakes are estimated to be 0.005-0.03 and 0.2-9.5 µg/day, respectively.
Five of ten heterocyclic amines examined for carcinogenicities in experimental animals were proven to induce colon cancer in F344 rats. Although HAs induced no colon tumors in CDF1 mice, MeIQx was found to induce colon tumors in C57B/6N mice. These results suggest that many HA compounds have suffircient potencies to induce colon tumors in rodents.
Dr. Sugimura described studies examining genetic changes in rat colon tumors induced by HAs, Glu-P-1 , IQ and PhIP. While Ki-ras activation was found in only one and p53 mutation in none of 27 colon tumors induced by these HAs, four of eight colon tumors induced by PhIP had five mutations in the APC gene, one of which harbored two mutations. All five mutations were G deletions at GGGA site, and four of these had an even longer consensus surrounding sequence, GTGGGAT. In contrast, APC mutation frequency was low (2/11) in the IQ-induced colon tumors, and these two were base-substitution mutations, C —>T at CpG site and T—>C. Dr. Sugimura and his associates have elucidated that PhIP preferentially forms an adduct of dG-C8-PhIP. IQ is also known to form dG-C8-IQ and dG-N2-IQ adducts, mainly with guanine. The base-substitution mutations in the APC gene of IQ-induced colon tumors may not necessarily be the result of IQ-DNA adducts. Thus, each chemical may have different affinities to different genes. In the rat APC gene, there are 26 GGA sequences and two GTGGGAT sequences, and at two of the latter sequences, two mutations each were induced. Thus PhIP seems to have strong affinity to the GTGGGAT sequence.
Dr. Sugimura also discussed the involvement of mismatch repair enzyme impairment in the colon carcinogenesis by analyzing microsatellite lengths. Various types of 85 simple repeat sequences were analyzed and seven of eight PhIP-induced colon tumors were found to have mutations in one or more loci each. In contrast, no mutations were detected in any of the nine IQ-induced colon tumors. In PhIP-induced colon tumors, mismatch repair enzyme(s) may be impaired. All of the tumors having APC mutations possessed microsatellite mutations. Dr. Sugimura indicated that it remains to be determined whether the APC mutations are the results of impairment of mismatch repair enzyme(s), cooperation of impaired mismatch repair enzyme(s) and adduct formation, or simply due to adduct formation.
Dr. Sugimura described studies analyzing mutation characteristics of a heterocyclic amine, MeIQ, using lac-I transgenic Big Blue® mice in relevance to DNA adduct levels, cell proliferation and carcinogenicity. Hepatocellular carcinomas and colon adenocarcinomas were induced by MeIQ in Big Blue mice. The major DNA adduct of MeIQ was determined to be dG-C8-MeIQ. After feeding with a diet containing 300 ppm MeIQ for 12 weeks, the DNA adduct levels were highest in the liver (28.3/107 ntd), followed by heart, colon forestomach, and bone marrow in that order. Mutation frequencies in these organs were closely correlated with the product of DNA adduct levels and cell proliferation, although colon showed a much higher mutation frequency. Mutation frequencies did not directly correlate with carcinogenicities. Further, Dr. Sugimura showed that mutation spectra in the bone marrow was different from those in the liver and colon.
Dr. Sugimura emphasized that the use of mutation fingerprints of a certain chemical to search for etiology of human cancer, a possibility of different mutational fingerprint in different organs should be taken into consideration.
In his closing remarks, Dr. Snorri Thorgeirsson thanked the participants for their contributions to a meeting characterized by impressive progress and new concepts and insights. He also emphasized the importance for scientists from the U.S and Japan to have this unique opportunity for high quality scientific meetings with detailed private discussions. The material presented provided new insights into both the mechanism by which food derived heterocyclic amines may contribute to human cancer as well as approaches that may be available to inhibit the impact of these compounds on human cancer. Dr. Thorgeirsson also reminded the participants that in spite of the impressive progress that has been made, there were a number of challenges in the field of food-derived heterocyclic amine research still to be conquered. In closing, Dr. Thorgeirsson asked the participants to recognize the tremendous contributions made by Dr. Sugimura and Dr. Adamson to the research on food-derived heterocyclic amines. The participants responded by a standing ovation in honor of Drs Sugimura and Adamson.
PARTICIPANTS
UNITED STATES
Dr. Snorri S. Thorgeirsson
Laboratory of Experimental Carcinogenesis
National Cancer Institute
Bldg. 37, Room 3C28
Bethesda, MD 20892-4255
Dr. Herman A. J. Schut
Department of Pathology
Medical College of Ohio
Toledo, OH 43614
Dr. Nicholas P. Lang
University of Arkansas for Medical Science
College of Medicine
Little Rock, AR 72205
Dr. Rashmi Sinha
Environmental Epidemiology Branch
National Cancer institute
Executive Plaza North, Room 443
Bethesda, MD 20892
Dr. James S. Felton
Lawrence Livermore Laboratories
University of California
P. O. Box 5507
Livermore, CA 94550
Dr. W. G. Stillwell
Division of Toxicology
Massachusetts Institute of Technology
77 Massachusetts Avenue
Cambride, MA 02139
Dr. Roderick H. Dashwood
Department of Environmental Biochemistry
University of Hawaii
1800 East West Road
Honolulu, HI 96822
Dr. Elizabeth G Snyderwine
Head, Chemical Carcinogenesis Section
Laboratory of Experimental Carcinogenesis
National Cancer Institute
Bldg. 37, Room 3C17
Bethesda, MD 20892-4255
Dr. Richard H. Adamson
Vice President for Scientific & Technical Affairs
National Soft Drink Association
1101 Sixteenth Street N W.
Washington, DC 20036
JAPAN
Dr. Takashi Sugimura
President Emeritus
National Cancer Center
1-1 Tsukiji 5-chome, Chuo-ku
Tokyo 104, Japan
Dr. Keiji Wakabayashi
Chief, Biochemistry Division
National Cancer Center
1-1 Tsukiji 5-chome, Chuo-ku
Tokyo 104, Japan
Dr. Yasushi Yamazoe
Faculty of Pharmaceutical Sciences
Tohoku University
Aoba, Aramaki, Adba-ku
Sendai 980, Japan
Dr. Tetsuya Kamataki
Faculty of Pharmaceutical Scieinces
Hokkaido University
Nishi 6-chome, Kita 12-jo
Kita-ku, Sapporo 060, Japan
Dr. Hikoya Hayatsu
First Department of Pathology
Nagoya City University Medical School
1-Kawasumi, Mizuho-ku
Nagoya 467, Japan
Dr. Katsumi Imaida
First Department of Pathology
Nagoya City University Medical School
1-Kawasumi, Mizuho-hu
Nagoya 467, Japan
Dr. Tomoyuki Shirai
First Department of Patholgoy
Nagoya City University Medical School
1-Kawasumi, Mizuho-ku
Nagoya 467, Japan
(2) Seminar on “Coadministration and Multiple Effects of Carcinogenic and Anticarcinogenic Compounds “
This workshop was held on March 13-14, 1995, at the Maui Marriott Hotel, Maui, Hawaii. It was organized by Drs. Tomoyuki Shirai of the Nagoya City University Medical School, and Michael Pariza of the University of Wisconsin-Madison. There were eight participants from the United States, one from Holland, six from Japan, and three Japanese observers.
The goals of the meeting were captured in the terms “coadministration” and “multiple effects.” In the context of this meeting “coadministration” referred to simultaneous exposures to synthetic and naturally-occurring carcinogens and anticarcinogens at concentrations that are typically low. “Multiple effects” referred to the need to take into account for the purposes of risk assessment the full spectrum of effects that each carcinogen or anticarcinogen might induce. Superimposed on these were additional considerations: genetics (between and within species); physiological (individual variations); and the multiple interacting biological responses that a xenobiotic might induce. Discussion and information exchange in these areas is needed to improve risk assessment methodologies for carcinogen exposures, and to increase our understanding of how optimally to use knowledge of anticarcinogens to prevent cancer and improve health.
Dr. Roswell Boutwell opened the workshop with a discussion mechanistic approaches to inhibiting carcinogenesis. Ornithine decarboxylase (ODC), the rate-limiting enzyme in polyamine synthesis, is one of the first biochemical activities to increase during the promotion of epidermal carcinogenesis. For example, 4.5 hours after a single application of 12-O-tetradecanoylphorbol-13-acetate (TPA), ODC in mouse skin is induced 300-fold. Difluoromethylornithine (DFMO), which is a specific reversible inhibitor of ODC, blocks carcinogenesis in every system in which it has been tested. The elucidation of DFMO as an anticarcinogen was based on recognition that it inhibited an essential step in tumor promotion. Inhibition of ODC Induction has been used to screen for potential inhibitors of mouse skin tumor promotion. Inhibition of prostaglandin (PGE2) synthesis may represent another means of mechanistically searching for inhibitors of carcinogenesis.
Dr. Makoto Hayashi followed with a discussion of combination effects of clastogens, which utilizes the mouse peripheral blood micronucleus assay. Model chemicals employed included the simple ethylating agents EMS (ethyl methanesulfonate (EMS), EES (ethyl ethanesulfonate), and ENU (N-ethyl-N-nitrosourea); spindle poisons VICN (vincristine sulfate) and COL (chlchicine); and the oxidizing agent KBrO3. When chemicals with different mechanisms of action were combined, MNRET frequencies were equal to that predicted by simple addition of the effects of the individual chemicals. However, when chemicals with the same mechanism of action were tested together the effect was greater than the sum of the frequencies induced by the individual chemicals .
Dr. Fred Kadlubar discussed genetic polymorphisms in carcinogen-metabolizing enzymes. The distribution for cytochrome P4501A2 (CYP1A2), which catalyzes the metabolism of aromatic and heterocyclic amines found in cigarette smoke and cooked foods, is known to be trimodal. This appears to be due to three phenotypes: homozygous inducible, rapid metabolizers; heterozygous inducible, intermediate metabolizers; and homozygous non-inducible, slow metabolizers. A proposed mechanism for aromatic amine-induced bladder carcinogenesis was presented which involved hepatic N-oxidation by CYP1A2, transport of the unconjugated N-hydroxy metabolite through the blood to the kidney where it is filtered into the bladder lumen. Reabsorption then occurs with conversion of the N-hydroxy metabolite to a reactive N-acetoxy ester that forms DNA adducts and initiates neoplasia.
Dr. Frank Gonzalez pointed out that despite the fact that P450s activate procarcinogens in vitro, there is no direct evidence that they are required for the induction of cancer in intact animals. He presented evidence Indicating that cytochrome P450 enzymes and receptors have physiologically-relevant functions in addition to xenobiotic metabolism. Certain cytochrome P450s appear to be involved with lung development and resistance to infections, regulation of the immune system, and lipid metabolism in species that exhibit peroxisome proliferation. CYP1A2 metabolizes estrogen, bilirubin and uroparphyrinogen; CYP2E1 oxidizes ketone bodies such as acetone and acetal in a pathway of gluconeogenesis believed to be important in extrahepatic tissues.
Dr. Shosuke Kawanishi spoke on oxidative damage induced by interactions between metals and carcinogens. There is considerable interest in the relationship between carcinogenesis and DNA damage induced by oxygen radicals. Various non-mutagenic carcinogens (including benzene-- see next presentation) are capable of damaging DNA through H2O2 formation produced when Cu(II) is added, Peroxisome proliferators may induce an imbalance in peroxisomal enzymes which then increase the production of active oxygen species. DNA damage related to inflammation was also discussed. It was concluded that many carcinogens can collectively cause oxidative damage through direct and indirect mechanisms of radical formation.
Dr. Michele Medinsky discussed benzene toxicity in rodents and humans. Persons exposed to high concentrations of benzene are at increased risk for developing leukemia; what is less clear is the risk to persons exposed to low concentrations of this industrial solvent. Rats and mice appear to metabolize benzene differently than humans. Accordingly, rats and mice may not be good models for predicting human risk. The formation of toxic benzene metabolites is the net result of competing detoxification pathways. Humans are heterogeneous as regards benzene metabolism, a finding that is very like to bear on an individual’s risk.
Dr. Miriam Poirier presented a paper that was co-authored with Dr. Nathaniel Rothman on the formation of DNA adducts in human blood cells following exposure to polycyclic aromatic hydrocarbons (PAHs). Occupational exposure of firefighters to smoke does not appear to increase PAH-DNA adducts in nucleated blood cells. However, frequent consumption of char-broiled beef is related to a transient rise in blood cell DNA adducts, which is enhanced by smoking and alcohol consumption. In another study 6 of 21 non-smoking volunteers who ingested char-broiled hamburgers exhibited increased aromatic adducts in their blood cell DNA. The findings were confirmed in U.S. Army soldiers. It is possible that concomitant dietary administration of antimutagenic agents might reduce the concentration of PAH-DNA adducts and result in chemopreventive effects. This has not yet been tested.
Dr. Harukuni Tokuda discussed a system for developing chemopreventive agents for human use. Chemoprevention was defined as the addition of selected natural and synthetic substances to foods for the purpose of reducing cancer risk. A strategic system was described that consisted of three phases, each with defined criteria and decision points. Studies begin with identification of promising compounds through in vitro screening; followed by assessment as an anticarcinogen in the two-stage mouse epidermal neoplasia model; and finally by assessment against internal cancers in mice. It was proposed that the use of several agents with different modes of action (combination therapy) would be most effective.
Dr. Michael Pariza discussed conjugated linoleic acid (CLA), a new]y recognized diet-derived anticarcinogenic nutrient CLA is an anticarcinogen that induces multiple physiological effects. Its mechanism appears to involve the regulation of fat and protein metabolism and retention. Hence, mice fed CLA-containing diets lose body fat but gain lean body mass. Alterations in hormone balance are involved In cancer reduction due to the calorie effect, which is one of the most powerful means known for reducing cancer risk in rodents. Accordingly, it was proposed that the anticarcinogenic action of CLA may be at least in part a result of similar alterations in hormonal balance caused, in this case, by CLA-induced changes in body com osition and fat/protein metabolism.
Dr. Katsumi Imaida presented a paper that was co-authored by Drs. Ryohei Hasegawa, Tomoyuki Shirai, and Nobuyuki Ito, which discussed the combined effects of mixtures of 20 or 40 pesticides at levels of exposure equivalent to the ADI (acceptable daily intake) or higher. Effects were assessed on carcinogen-induced preneoplastic foci in rat liver and other organs. When fed at the ADI, the pesticide mixtures did not increase neoplastic lesions in any tested organ. However, when fed at 100-times the ADI statistically-significant increases in neoplasms were observed. It was concluded that complex mixtures of pesticides fed at very low doses (e.g., the ADI) were not carcinogenic.
Dr. Thomas Kensler co-authored a paper with Drs. John Groopman and Paul Talalay in which interactions between a chemical carcinogen (aflatoxin B1) and a biological carcinogen (hepatitis B virus) were considered. In a nested case control study conducted in China the risk of hepatocellular carcinoma was null when neither factor was present in biological fluids, about 3 for aflatoxin B1 alone, about 7 for hepatitis B virus alone, and over 50 for those exposed to both aflatoxin and the virus. Hence both factors were highly significant alone but greatly increased risk when present together, a most striking example of “multiple effects” of carcinogens. Various means of reducing risk were discussed. Elimination of exposure is not feasible, but various chemopreventive strategies seem promising. Among the substances under study are the drug oltipraz and various nutrients and phytochemicals.
Dr. Kunitoshi Mitsumori discussed the combined effects of simultaneous administration of goitrogens with different mechanisms of action on rat thyroid carcinogenesis. At low doses the effects were essentially additive, whereas at higher doses (and particularly in combination with a thyroid peroxidase inhibitor and a liver enzyme inducer) the effects were more than additive. The results are important for carcinogenic risk assessment models.
Dr. Tomoyuki Shirai co-authored a paper with Drs. Ryohei Hasegawa, Takashi Sugimura, Masao Hirose, and Nobuyuki Ito, on the combined effects on rat liver carcinogenesis of administering five or ten heterocyclic amine mutagen/carcinogens at their respective carcinogenic doses or 1/5th, 1/10th, 1/25th, or 1/100th of their carcinogenic doses. Synergism was observed in some cases but in general the combined effects were isoadditive.
Dr. Peter J. van Bladeren co-authored a paper with Drs John P. Groten, Victor, J. Feron, Frieke C. Kuper, Eric D. Schoen and Diana Jonker, on a systematic approach to evaluating the toxic potential of combinations of chemicals administered at low doses. To detect interactive effects of only two or three chemicals, isobolograms rather than factorial designs should be utilized. It was shown that the “dose addition” model should be applied only to data on mixtures of similarly acting chemicals with the same target organ; mlsuse of this model can lead to risk overestimation (in the case of dissimilar joint action) or underestimation (in the case of potentiating interactions). It was proposed that safety evaluations of complex mixtures should be determined systematically by concentrating effort on the ten or so most risky chemicals in a mixture.
The workshop was summarized by Dr. Richard H Adamson.
PARTICIPANTS
UNITED STATES
Dr. Richard H Adamson
Vice President for Scientific & Technical Affairs
National Soft Drink Association
1101 Sixteenth Street NW
Washington, DC 20036
Dr. Roswell K. Boutwell
Professor Emeritus, Dept of Oncology
University of. Wisconsin-Madison
1125 McArdle Cancer Research
1400 University Avenue
Madison WI 53706
Dr. Frank J. Gonzalez
Chief, Nucleic Acids Sec.
Lab Molec Carcinogen
National Cancer Institute
Building 37, Room 3E24
9000 Rockville Pike
Bethesda, MD 20892
Dr. Fred F. Kadlubar
N.C.T.R.
HFT-100
Jefferson, AR 72079
Dr. Michele A. Medinsky
Chemical Industry Institute of Toxicology
P.O. Box 12137
Research Triangle Park, NC 27709
Dr. Michael W Pariza
Professor, Chair & Director
Food Microbiology & Toxicology
Food Research Institute
University of Wisconsin-Madison
1925 Willow Drive
Madison, WI 53706-1187
Dr. Miriam C Poirier
National Cancer Institute
Building 37, Room 3B25
9000 Rockville Pike
Bethesda, MD 20892
Dr. Thomas Wells Kensler
Department of Environmental Health Sciences
Johns Hopkins School of Hygiene
615 North Wolfe Street
Baltimore, MD 21205
Prof, Dr. P. J. van Bladeren
Director, TNO Nutrition and Food Research
Utrechtseweg 48
P.O. Box 360
3700 AJ Zeist
The Netherlands
JAPAN
Dr. Makoto Hayashi
Section-Chief, Division of Genetics and Mutagenesis
National Institute of Health Sciences
1-18-1 Kamiyoga, Setagaya-ku, Tokyo 158
Dr. Katsumi Imaida
Assistant Professor, 1st Department of Pathology
Nagoya City University Medical School
1 Kawasumi Mizuho-cho, Mizuho-ku, Nagoya 467
Dr. Shosuke Kawanishi
Assistant Professor, Department of Public Health
Kyoto University
Faculty of Medicine
Yoshida Konoe-cho, Sakyo-ku, Kyoto 606-01
Dr. Kunitoshi Mitsumori
Section-Chief, Division of Pathology
National Institute of Health Sciences
1-18-1 Kamiyoga, Setagaya-ku, Tokyo 158
Dr. Hoyoku Nishino
Chief, Cancer Prevention Division
National Cancer Center Institute
5-1-1 Tsukiji, Chuo-ku, Tokyo 104
Dr. Tomoyuki Shirai
Professor and Chairman, 1st Department of Pathology
Nagoya City University Medical School
1 Kawasumi Mizuho-cho, Mizuho-ku, Nagoya 467
(3) Seminar on “Breast and Prostate Cancer: Molecular Carcinogenesis in Relation to Life Style”
This workshop was held on March 28-29, 1995, at the Maui Prince Hotel, Maui, Hawaii. The organizers were Drs. Ryuichi Yatani of the Mie University, Mie, Japan and J. Carl Barrett of the National Institute of Environmental Health Sciences, Research Triangle Park, North Carolina. There were six participants from the United States, seven from Mie, Japan, and one observer from Japan.
The purpose of the seminar was to discuss and exchange information on the role of lifestyle in breast and prostate cancers and how new information on the molecular mechanisms of these cancers could contribute to this understanding.
Dr. Yoshiyuki Ohno kicked off the workshop with epidemiological remarks on cancer of the breast and the prostate in Japan. He noted that an early step to obtain the etiologic clues of disease is to ascertain the descriptive epidemiology: who are affected and what are demographic features and characteristics of these cancers. This information may be used to generate hypotheses to be tested in analytic research as well as to identify high-risk groups to which primary and/or secondary prevention would be beneficial. He presented descriptive epidemiological features of breast and prostate cancer in Japan and some country-specific risk factors for these cancers. Materials used for descriptive epidemiology are routinely available from national mortality and incidence statistics (Vital Statistics and Cancer Registry Data of Japan). Dr. Ohno stated that age-adjusted rates, computed by direct method using world population as a standard, were used to examine secular trends in mortality and incidence. The major results pertaining to breast cancer are: (1) The age-adjusted mortality rate has been increasing since the 1960’s (4.1/100,000 females in 1950 ro 6.6 in 1991) (2) The age-adjusted incidence rate has also been increasing since 1975 (3) Age-specific rates have been increasing particularly in the age range 40-79 years (4) Multiplicative models have revealed an increasing cohort effect for the mortality and incidence rates among women born after 1900. (5) Decreasing mortality and incidence rates after menopause were ascribed by the models to strong cohort effects. (6) A decreased risk with a large number of live births and an increased risk with regular menstrual cycle were noted among premenopausal women Among postmenopausal women, heavy weight increased the risk, and at later the age at first live birth, the higher the risk. Current smoking elevated breast cancer risk among all women and postmenopausal women.
The results pertaining to prostate cancer are: (1) the age-adjusted mortality rate showed the 25-fold increase over 44 years (0.11/100,000 males in 1947 to 3.96 in 1991). (2) The age-adjusted incidence rate also increased from 4.2/100,000 in 1975 to 7.6 in 1987; about 1.8-times increase over 12 years. (3) Trend in age-specific death and incidence rates were virtually unchanged in 45-59 years old, slightly upward in 60-69 years old, and apparently upward in 70 years old or more. (4) Birth-cohort analysis showed an increasing risk of death with successive birth-cohort between 1871-1875 and 1891-1895, with almost identical risk in birth-cohorts born thereafter. (5) Both age and birth-cohort effects were detected to be quite strong for incidence rate. (6) SMR (standardized mortality ratio: 1989-1991 by prefecture) analysis showed no significant clusters of either high or low SMR (7) Annual number of incident cases was projected to be 26,110 in 2015, with the age-adjusted incidence rate of 13.0/100,000 men in Japan. (8) Dietary evaluation demonstrated the following major findings: (a) The smaller the dietary intake of beta-carotene and vitamin A as well, the higher the risk. (b) This risk reduction by dietary beta-carotene and vitamin A was significant in the older men (70-79 years), but not in the younger men (50-69 years). (c) Beta-carotene and vitamin A contained in green/yellow vegetables were protective, and those in seaweeds and kelp suggestively protective.
Dr. David Hunter discussed diet in the etiology of breast and prostate cancer. He stated that the considerable International differences in cancer rates, together with the increasing incidence rates over tirne In many countries and the results of studies of migrants, suggest that environmental factors are responsible for a major portion of cancer incidence in the United States and other affluent countries. A modest association with alcohol consumption is the best-supported dietary relationship for breast cancer at this time. Dr. Hunter stated that prospective studies suggest that the relationship between dietary fat intake in middle life and breast cancer development is insignificant, although the hypothesis that very low fat intakes may be protective has not been fully tested. Energy intake and growth restriction during growth may explain much of the international variation In breast cancer rates but does not immediately suggest a feasible intervention. Substantial evidence indicates that vitamin A from animal or vegetable sources may reduce risk, but this hypothesis requires further investigation. The influence of adult diet on breast cancer should become clearer in the next five years as a number of large ongoing prospective studies report more data. In a number of case-control and prospective studies red meat consumption and/or animal fat intake increased risk of prostate cancer. A diet low in animal fat and high in vegetables probably does reduce overall cancer risk; it is premature to recommend consumption of specific nutrient supplernents as an alternative.
Dr. Minako Nagao spoke on rat mammary cancer induced by food born carcinogens, heterocyclic amines. Heated meat and fish flesh contains strong mutagens, heterocyclic amines. Nineteen heterocyclic amines which are mutagenic to S. typhimurium TA98 and TA100 were isolated from heated amino acids, proteins or proteinaceous foods, and their structures have been determined. Ail ten compounds examined tbr their carcinogenicities have been proven to be carcinogenic. Three of these, 2-amino-3-methylimidazo[4,5-F]quinoline (IQ), 2-amino-3,4dimethylimidazo[4,5-b]pyridine (PhIP) induced mamrnary adenocarcinomas in rats. PhIP is the most abundant carcinogenic heterocyclic amine in cooked meat, and most people are exposed to this compound almost every day. PhIP is known to be metabolized to NOH-PhIP by CYPIA2, and further to an ultimate form, N-OAc-PhIP by N-acetyltransferase, NAT2. The ultimate form reacts with DNA to yield an adduct, N2-(guanin-8-yl)PhIP. PhIP at a concentration of 100 ppm in the diet, induced mammary cancer in 47% of F344 female rats after 104 wks of feeding. It also induced mammary cancer in 25% of SD female rats after 48 wks of feeding. PhIP mammary carcinogenesis was also enhanced by a high concentration of fat in the diet. In the PhIP induced mammary carcinogenesis of F344 rats, involvement of Ha-ras activation was 18% (3/17), and mutation in p53 was 6%. Microsatellite instability for 65 loci was further analyzed in the mammary tumors of SD x F344 F1 rats, and it was found that eight of the 15 tumors induced by PhIP had mutations. Five of the microsatellite-mutation positive tumors had mutations at multiple loci. In contrast, only two of 12 mammary tumors induced by DMBA had mutations in one microsatellite locus each. Since colon tumors induced by PhIP also but not by IQ had mutations in microsateliite sequences, microsatellite mutations seem to be a characteristic of DNA alterations induced by PhIP. Further, PhIP-induced tumors had a specific LOH on chromosome 10, which is highly homologous to human chromosome 17q. The molecular pathways involved ir PhIP-induced mammary tumors, DMBA-induced mammary tumors and NMU-induced tumors may be different.
Dr. Jeffrey E Green spoke on targeting oncogenesis to the mouse prostate and mammary glands. The study of oncogenesis in the prostate has been limited by the lack of easily useable animal models. In an effort to create a transgenic model for prostate cancer, various genomic regions of the rat C3(1) gene were fused to the SV40 early reg”ion coding for the transforming protein T-antigen. Previous studies have demonstrated that the C3(1) gene can specifically target expression to the prostate in transgenic mice. Multiple hormone response elements are contained within the 5’ flanking region of the C3(1) gene as well as in exon and intron sequences. In the first series of experiments, only the 5’ flanking region of the C3(1) gene was utilized to direct expression of T-antigen. Twelve founder animals were generated of which two lines of transgenic animals could be propagated. One line has been studied extensively and over 50% of the males survrvmg more than 7 months develop adenocarcinoma of the prostate. Dysplastic lesions occur in the prostate as early as three months of age. Interestingly, 100% of the females develop adenocarclnoma of the mammary glands and die by six months of age. The slow and predictable development of adenocarcinoma in the prostate and mammary glands of these transgenic animals provides a model to study the multistep progression of tumorigenesis in these tissues. Cell lines from tumors have been created and used as reagents for PRC differential display which has identified genes which are both up and down-regulated during tumor progression. In situ imnlunocyiochemistry has demonstrated altered cell cycle control and elevated expression of p53 in early lesions which increases with tumor progression. Using the terminal transferase method, levels of apoptosis have been shown to increase in dysplastic lesions and adenomas, but decrease in adenocarcinomas of the prostate and mammary glands. The survival curve of both male and female animals has been extended by oophorectomy and castration, but tumors appear to develop eventually. Hormone supplementation to oophorectomized and castrated animals results in survival curves identical to the intact animals demonstrating that tumor formation in these transgenic animals is hormone-responsive. In addition to the prostate and mammary tumors, these transgenic mice develop other proliferative lesions including unusual mixed tumors of the plantar sweat glands with heterotopic bone formation. These transgenic animals should serve as useful models to study multistep oncogenesis in the prostate and mammary glands as well as being useful for in vivo drug studies.
Dr. Kenji Kamiya discussed the feature of radiation-induced rat mammary cancers. The Copenhagen (COP) rat strain is highly resistant to both spontaneous and chemically induced mammary cancers. The resistant trait is due to the inheritance of autosomal dominant allele termed mammary carcinoma suppressor (mcs) gene. Radiation-induced mammary cancers in COP rats were studied. Female rats of inbred COP, Fischer 344 (F344) and Wistar Furth (WF) strains were exposed to either 50 mg/kg of MNU or 3.0 Gy of 60Co gamma rays at 55-56 days of age. F344 and WF rats were highly susceptible to MNU induction of mammary carcinomas 100%), while COP rats were resistant to the same treatment (11.8%) (P<0.01). On the other hand, when irradiated with 3.0 Gy of 60Cogamma rays, no difference of mammary carcinoma incidences and latencies were observed among the three strains, and the incidence of carcinomas in COP rats (37.0%) was as high as those in F344 (22.6%) and WF rats (26.7%). As benign lesions, adenomas were induced by 60Cogamma rays exposure in the three strains of rats. The incidence of adenomas In COP rats was the highest among the three strains.
Immunohistochemical examination revealed that all of MNU-induced carcinomas lacked expressions of R-MFGM (rat milk fat globule membrane) and gamma-SMA (gamma-smooth muscle actin) which were located in secretory epitheliums of alveoli and myoepitheliums, respectively. In contrast, 68.8% of radiation-induced carcinomas and all of adenomas in three strains were strongly positive for these antigens. These results indicated that both radiation-induced adenomas and carcinomas were derived from the similar portions of differentiated mammary glands, which was different from that of chemical carcinogen-induced mammary carcinomas, Furthermore, all of adenomas examined grew tumors in E2-treated syngenic rats after transplantation. Histological examination of tumors in grafted sites revealed morphological malignant conversion from adenomas to carcinomas. These results may suggest that 60Cogamma-irradiation initially induced adenomas, and some of adenomas eventually progressed to highly differentiated carcinomas in COP rats. Evidence is presented In support of the hypothesis that development of cancer depends on an imbalance between highly frequent initiation and subsequent epigenic cellular interactions that suppress promotion-progression towards cancer from the initiated cells. It has been shown that a subpopulation of rat mammary epithelial cells are clonogenic, i.e., able to give rise to clonal glandular structures when transplanted in syngenic recipients. When the cells are exposed to initiating doses of ionlzing radiation or of nitrosomethylurea (NMU) before transplantation, carcinomas arise within these clonal glandular structures. In the current study, ~13 to ~1,080 surviving 7 Gy gamma irradiated or unirradiated clonogens, or mixtures of both, were transplanted per site. Recipients were then treated to maximize hormonal promotion-progression of initiated grafted clonogens. The cancer incidence per irradiated clonogen in grafts of ~13 clonogens per site indicated that at least one of ~95 clonogens was radiogenically initiated. A similar initiation frequency had been seen in grafts of ~5 NMU-treated clonogens. These initiation frequencies are orders of magnitude greater than the expected specific locus mutation frequencies. In sites grafted with larger cell inocula, cancer incidences per clonogen were suppressed inversely as the numbers of irradiated or NMU-treated clonogens grafted per site were increased. Addition of unirradiated cells to small irradiated cell inocula also suppressed progression in the grafi sites. Radiation and NMU thus produce quantitatively, and perhaps qualitatively, similar carcinogenesis-related sequelae in mammary clonogens.
Dr. Robert Dickson presented data on growth factors and the myc oncogene: A fatal interaction in transgenic mouse mammary tumorigenesis. Previous studies of multiple investigators have shown that estrogen and progesterone can facilitate multiple stages of breast tumor progression. Sex steroids are known to stimulate human breast tumor cell growth at least partially through induction of the growth factor TGF!!
!and the nuclear protooncogene c-myc. It is also known that TGF!!!, other related growth factors of its family, its receptor family members, and c-myc protooncogene are commonly overexpressed and/or amplified at the gene level in human breast cancer. Dr. Dickson stated that a bitransgenic mouse model was used to study the consequences of overexpressed TGF!!
!and myc oncogene. Single transgenic TGF!!!- and myc-expressing strains were utilized which require multiple pregnancies and long latencies (9-12 months) for mammary tumorigenesis. The transgenic MT100, heterozygous for TGF!!!, was mated with MMTV-c-myc, heterozygous for myc. In striking contrast to single transgenic strains, 100% of bitransgenic offspring of both sexes developed multiple mammary tumors at an average age of 66 days with no requirement for female pregnancies. Tumors were locally invasive adenocarcinomas which rapidly grew, requiring sacrifice a few weeks after onset Myc and TGF!!!also synergistically interacted in salivary tumorigenesis in this cross. In situ hybridization analysis of mammary and sallvary tumors revealed a ciose colocalization of TFG and myc mRNA with hyperplastic and malignant areas, further supporting cooperation of the two genes in malignant progression. When bitransgenic tumors were characterized by northern blot analysis, it was observed that cyclin D and myn (the murine heterodimeric partner of myc) were strongly expressed. In addition, a very high S-phase functions of these tumors suggested a cooperative induction of the cell cycle by both transgenes. Cell lines were established from a single transgenic and bitransgenic mammary tumors for further study. Myc-containing lines were selectively sensitive to TG for induction of anchorage Independent cell growth in soft agar Myc-expressing cells from single transgenic tumors uniquely exhibited very high levels of programmed cell death (apoptosis). TGF!!!suppressed apoptosis, thus functioning as a survival factor. In summary, TGF!!!and myc may strongly cooperate in mammary tumorigenesis by stimulating the cell cycle, stimulating anchorage independent growth, and promoting cell survival. Current studies are addressing growth factor specificity of these effects and the possible consequences of this growth factor-oncogene interaction for genetic instability and subsequent mutation of tumor suppressor genes. Dr. Mitsuru Emi reported his studies on molecular genetics of breast cancer. Human breast cancers frequently show loss of heterozygositv. (LOH) and amplification at specific chromosomal regions. To understand the role of these genetic alterations in tumor development and progression, 457 cases of primary breast cancers were studied fbr LOH at chromosomal regions 16q24, 17p13.3, 17q21, and amplification of the erbB2 locus at 17q11.2 and the c-myc locus at 8q24, and compared them with lymph node metastasis, histological type and tumor stage. LOH at 17q21 was more frequently observed in tumors of solid-tubular type (41/75, 55%) than the other types (48/187, 26%) (p<0.0001). LOH at 17p13.3 was more frequent in tumors of scirrhous and solid-tubular types (77/141, 55% and 48/88, 55%) than the other types (29/89, 33%) (p=0.0004). The association of allelic loss at 17p13.3 and 17q21 with specific types of tumor indicates that loss or inactivation of a tumor suppressor genes at these regions may exert their effect in a histological type-specific manner. A comparison of estrogen receptor (ER) and progesterone receptor (PgR) status in tumor cells with data concerning these genetic alterations revealed that LOH at 17q21 was significantly correlated to absence of estrogen receptors (ER)(p<0.0003) or progesterone receptors (PgR)(p<0.0001), and with the absence both (p<0001). Similarly, a significant association was observed between amplification of ERBB2 and the absence of either ER of PgR. LOH at 17p13.3 was associated with the absence of PgR (P<0.01). These data suggest a possible relationship of specific genetic changes on chromosome 17 with hormonal deregulation In the progression of breast cancer.
Predisposing mutations in a BRCA1 gene have been recently identified in germline of 17q-linked breast and ovarian cancer families. Primary breast cancer patients were examined, consisting of 46 early-onset cases (<35 of age), 12 cases from multiply-affected families, and 59 cases with bilateral cancer for mutations in entire coding exons of BRCA1 gene using single strand conformation polymorphism (SSCP) analysis. Four mutations were detected in this panel of 103 patients; a flame-shift due to 2-bp deletion at codon 797, a nonsense mutation at codon 1214, two missense mutations, one at codon 271 leading to Val->Met substitution, and the other at codon 1150 leading to Pro->Ser substitution. All of them were germ-1ine mutations; no somatic mutations were found In these tumors, a finding that support a rather confmed role of BRCA1 in breast carcinogenesis. The mean age of onset in these Japanese patients was 49, in contrast to the mean age of 35 observed among the BRCA1 mutation carriers in a U.S. study. Among the three selection groups, all mutations were found in patients with bilateral tumors.
Dr. Helene Smith reported on the molecular aspects of breast cancer progression. She stated that invasive cancers are unusual in that tumors of similar stage and histology are very heterogenous in clinical behavior. There is much evidence suggesting that breast cancer is many different diseases. Supporting this hypothesis are the observations that molecular lesions causing aberrations of gene expression for oncogenes, tumor suppressor genes, growth factors, proteases, angiogenesis factors, and stromal components all have been seen in a varying, and sometimes small, proportion of breast cancers. It is likely that biologic heterogeneity among breast cancers at least partially results from the combination of genetic lesions present in a given tumor. It is proposed that each step required for malignant growth can be accomplished by more than one genetic lesions, with some of the lesions causing a more aggressive phenotype that others. The biologic heterogeneity of breast cancers also may arise from various orders of acquiring the sum of steps necessary for malignant growth. Supporting the stochastic model are studies characterizing losses of heterozygosity (LOH) associated with ductal carcinoma in situ (DCIS), primary invasive lesions, and metastases. The reasons for proposing this model are that: (1) each lesion can be seen in a proportion of breast cancers and they differ in frequency of occurrence; (2) many lesions can be acquired early or late in progression (after invasive malignancy and thus contributing to the increasing aggressive behavior of metastases); and (3) a particular phenotype (i.e., rapid proliferation at the primary site) correlates with various molecular lesions. With the long-term goal of understanding the earliest stages in breast cancer progression, normal and malignant breast epithelium were characterized for pattern of X-chromosome inactivation. Most carcinomas (DCIS and invasive) were clonal; however, one multicentric DCIS had regions with each X chromosome inactivated suggesting a multiclonal origin. In the normal tissue, entire lobules and large ducts of normal breast tissue had the same inactivated X-chromosome suggesting that the normal breast is made up of mosaic patches derived from a single stem cell. If a normal stem cell undergoes tumor initiation prior to mammary gland differentiation at puberty, a localized region of the mammary tree would be at risk for developing multiple carcinomatous lesions. If so, one would expect to fmd cases where two regions of a localized carcinoma have progressed to malignancy by acquiring different chromosomal lesions. Supporting this hypothesis, a breast cancer case is described in which the Invasive component showed loss of heterozygosity (LOH) at one allele on chromosome 17p, while the associated in situ component displayed loss of the opposite allele. The same X-chromosome was inactivated in both components of the tumor. LOH at this locus must have occurred early in progression since morphologically normal mammary epithelium adjacent to this carcinoma has a partial LOH. The social and ethical implication of these observations are that effective chemoprevention of breast cancer may require targeting prepubescent females.
Dr. Yoshinobu Kubota spoke on tumor suppressor genes in human prostate cancer. The genetic background underlying the growth and the development of human prostate cancer is not yet clear. Here, he searched for mutations in the entire coding regions of the p53, RB, MTS-1, tumor suppressor genes and DNA polymerase B genes among prostate cancers of Japanese patients, using PCR, SSCP and sequencing. Tissue samples were obtained from 31 patients with human primary prostate cancer at Yokohama City University and its affiliated hospitals. Total RNA was extracted from each sample by the AGPC method. cDNA was synthesized from RNA with random hexamers. Short segments of the p53, RB, MTS-1 and DNA polymerase B genes were amplified by PCR with 11, 21, 6 and 7 primer sets covering the respective entire coding regions, and the resulting DNA segments were transcribed with T7 RNA polymerase. The T7 transcripts were electrophoresed and structural aberrations were detected. The PCR products that revealed abnormal patterns in RNA-SSCP were directly ligated to plasmid vectors. At least 10 clones of each ligated PCR product were sequenced. (1) p53 gene; Abnormal electrophoretic patterns were observed in 5 of 31 samples (16%) in RNA-SSCP. Sequence analysis revealed point mutations in exons 5 in two cases and exon 3 in one case, and a single base deletion in exon 5 in one case and exon 7 in one case. Four of these 5 cases were poorly differentiated adenocarcinomas. (2) RB gene; Abnormal electrophoretic patterns were observed in 4 of 25 analyzed samples (16%) in RNA-SSCP. Sequence analysis revealed point mutations in exon 6 in one case and exon 19 in one case. Two of these 4 cases were shown to have mutations in SV-40 T binding regions of this gene and these two cases were reactivated poorly differentiated adenocarcinoma. (3) MTS-l gene; No abnormal electrophoretic patterns in SSCP analysis were deleted among the listed 31 cases. (4) DNA polymerase!!
!gene; 12 of the 31 cases were examined for DNA polymerase!!!mutations. Abnormal patterns were observed in 2 of the 12 cases (17%) in RNA-SSCP. Sequence analysis revealed a single point mutation in one case, and an insertion plus a deletion in another case. Both cases were of advanced grade and stage. Interestingly, one of the cases also contained p53 gene mutation. He also searched for mutations of ras gene (codons 12 and 61 of c-Ha-ras or c-Ki-ras) in these 12 cases. However, no mutations of the ras-genes in these cases were found. These results indicate that p53, RB and DNA polymerase B gene mutations are involved in, at least, a subset of human prostate cancer, suggesting that these mutations have important roles in the progression processes of human prostate carcinogenesis. Dr. W B Isaacs discussed his studies on the genetic alterations in prostate cancer. A number of genetic changes have been documented in prostate cancer, ranging from allelic loss to point mutations and changes in DNA methylation patterns. LOH studies carried out primarily on clinically localized cancers reveal the short arm of chromosome 8 as most frequently showing allelic loss, followed by the long arm of chromosome 16. A recent study of lymph node deposits of metastatic prostate cancer carried out by Dr. G. S. Bova in collaboration with Drs. M. Cher and R. Jensen at the UCSF utilizing comparative genomic hygridization also demonstrated frequent loss of sequences from 8p (often accompanied by gain of sequences on 8q), with LOH on chromosome 13q being even more common than loss of 8p sequences. Deletion mapping on chromosome 8p has focused attention upon a relatively small (<1 megabase) region at 8p22 which has a high likelihood of containing a novel prostate cancer tumor suppressor gene. Evaluation of candidate genes in this region is ongoing in collaboration with Dr. Rob Bookstein in San Diego who has contructed a physical map of this region. Deletion of chromosome 17p is less commonly observed in these tumors and is not tightly linked to mutation of the p53 gene. Alterations in the E-cadherin/!!
!catenin mediated cell-cell adhesion mechanism appear to be present in almost one third of all prostate cancers, and may be critical to the acquisition of metastatic potential of aggressive prostate cancers. Finally, altered DNA methylation patterns have been found in the majority of prostate cancers examined, suggesting widespread alterations in methylation-modulated gene expression. The presence of multiple changes in these tumors is consistent with the multi-step nature of the transformation process. Finally, efforts to identify prostate cancer susceptibility loci are underway and will hopefully elucidate critical earl events in rosta i carcinogenesis. Dr. Jun Shimazaki spoke on progression of androgen-responsive cells to androgen-independent ones in prostate cancer. Most of prostate cancer responds to androgen ablation, showing androgen-responsive tumor, but gradually loses responsiveness and finally progresses to androgen-independent one. Metastasis occurs from androgen-responsive cancer cells as well as independent cancer cells, thus metastasis is an independent process from loss of andro responsiveness. To clarify mechanism of progression to androgen-independent process, the present study was undertaken to compare frequency of abnormalities in oncogenes, tumor suppressor genes and androgen receptor gene in cancer tissues between androgen-responsive and -independent stages. Genes concerning progression and metastasis in prostate cancer were also studied. Tissues obtained at radical prostatectomy form a total of 30 patients with stages B and C were used as androgen-responsive cancer. For androgen-independent cancer, tissues were obtained from a total of 22 autopsy cases who initially responded to endocrine therapy, then relapsed and died from the progression of cancer. At autopsy, cancer foci were removed from the primary and metastatic tissues including kidney, Iiver and lymph nodes. Specimens from surgery or autopsy were frozen immediately after removal and stored at -80°C. Genomic DNA was extracted and used for PCR. Primers for PCR were as follows: exons 1 and 2 of K, H, N-ras, 5’half of exon 15 of APC, exons 4-9 of p53, exons B-H of androgen receptor. When conformation abnormality was detected on PCR-SSCP, direct sequencing was performed with another aliquot of PCR product. Detection of human papilloma virus (HPV) DNA was performed with PCR using consensus primers on L1 region. To analyze LOH on 8p, genomic DNA was extracted and digested with Msp 1 or Taq 1, then electrophoresed. Newly prepared 12 polymorphic DNA markers (Y. Fujiwara, M. Emi, and Y. Nakamura, Cancer Institute, Tokyo) were used as RFLP markers. To detect metastasis suppressor gene on 8p, highly metastatic Dunning tumor cells of AT6.2 were transferred with human chromosome #8 by microcell-mediated chromosome transfer technique. Changes on metastatic ability on nude mice were measured.
There was no APC gene mutation in prostate cancer although polymorphism was found in some cases without a substitution of amino acid. Four ras point mutations were detected; 1 (13%) and 3 (30%) of stage B+C and autopsy samples, respectively. Sites of mutation were codon 12 or 61 of K-ras and codon 13 or 61 of H-ras, respectively. Cases with ras mutation were poorly differentiated adenocarcinoma. In autopsy samples the same mutaion was detected in both primary and metastatic foci. P53 gene mutation was not found in stage B+C, however, six out of 17 autopsy samples (35%) contained the mutations. Sites of mutation were 3 in exon 5, 2 in exon 7 and 1 in exon 8. Types of mutations in these cases were transversions and single nucleotide substitutions with amino acid change. HPV was detected; 4 (24%) and 3 (18%) of stage B and autopsy samples, respectively. Genotype of HPV was type 16. One case showed HPV integration and p53 gene mutation in different foci. These results suggest that ras mutation is rather early event and p53 mutation correlates with progression. and oncogenic protein of HPV may participate in the development in a subset of prostate cancer cases.
Transfer of human chromosome #8 decreased metastatic ability of AT 6.2 cells without influence on growth rate or tumorigenicity after subcutaneous innoculation to nude mice, suggesting metastasis suppressor gene(s) on chromosome #8. To determine further localization of suppressor gene(s) on chromosome #8, fragmentation of the chromosome by radiation was performed and fragments were transferred in AT 6.2 cells. The cells containing 8p23-q12 were suppressed with metastatic ability.
Allelic losses on 8p were detected in 11 out of 27 informative cases (40%) of stage B+C and 14 out of 17 informative cases (82%) of autopsy cases. In two autopsy cases tissues from normal part and cancer foci in the prostate did not show any abnormality but metastatic foci revealed LOH. Detailed deletion mapping of chromosome 8p defined two distinct, commonly deleted regions; cMSR-32 - C18-2644 at 8p22-p21.3 (1.2 Mb) and C18-1312 - C18-494 at 8p21-8p11.22. Therefore, two different approaches focused the similar localization of chromosome #8 as metastasis suppressor gene(s).
There was no abnormality of androgen receptor in stage B+C, but in 3 out of 22 autopsy cases metastatic foci showed the same point mutation at codon 877 (ACT-GCT, Thr-Ala). This mutation is the same as that reported in LNCaP cells. In addition, one case whose metastatic foci showed codon 877 mutation revealed codon 701 mutation (CTC-CAC, Leu-His). In 3 cases with mutated androgen receptor, two cases showed a PSA decline after withdrawal of antiandrogen (chlormadinone acetate), suggesting codon 877 mutation relates to antiandrogen withdrawal syndrome.
Dr. J. Carl Barrett spoke on BRCA1 AND KAI-1 - new genes involved in breast cancer and prostate cancer. The BRCA1 gene predisposes to familial breast and ovarian caner and has been localized to human chromosome 17q21. By positional cloning techniques, Roger Wiseman and others at NIEHS, in collaboration with scientists at the University of Utah and Myriad Genetics, have identified this gene. A physical map spanning more than 2 Mb of human chromosome 17q21 was constructed by assembling a genomic contlg of YAC, P1, and cosmid cloned In conjunction with pulsed field analysis. Using newly identified microsatellite markers, genetic analysis of recombinant chromosomes in member of large kindreds refmed the localization of BRCA1 to a region of approximately 600 Kb between D17S1321 and D17S1325. More than three dozen candidate expressed sequences were isolated from the critical region by a new method known as solution hybrid capture with normal breast cDNA. DNA sequence analyses of these genes revealed that one of them contained a frameshift mutation in germline DNA from affected individuals of a Utah breast cancer family. Alterations in these gene have been identified in four independent breast/ovarian cancer kindreds, establishing it as the BRCA1 locus. Likewise, mutation screening In a panel of 44 sporadic breast and ovarian tumors with allele loss on chromosome 17q21 revealed four additional BRCA1 alterations including a stop codon in an early-onset cancer patient. Interestingly, each of these alterations were also present in germline DNA from the patients, but not in DNA from more than 100 control chromosomes, suggesting that they were predisposing mutations. The BRCA1 gene product is a ring finger protein that may function as a transcriptional factor. Presently one of every four cancers diagnosed in American males is of prostatic origin. Once prostatic cancer metastasizes, it is a fatal disease for which no therapy presently available is curative. Because of these facts, there is a growing interest in the early detection and screening of men for prostate cancer. Such screening could potentially identify 10 million American men with histological prostatic cancer. It is estimated that approximately 7% of these men will eventually die from progression of their disease if left untreated (Prostate 25: 249, 1994). This raises the critical question of which of the remaining 93% (9,300,000) of men with nonlethal, but potentially life-altering histologically detectable prostatic cancer should receive therapy. There is no diagnostic method presently available which allows men with histologically detectable prostatic cancer to be distinguished between those requiring immediate aggressive therapy from those requiring either delayed or no treatment.
Progression to a metastatic state by such histologically detectable prostatic cancer cells is a defmitive criterion upon which to base such a diagnostic substaging method. Detection of the genetic changes required for metastatic ability by prostatic cancer cells could be used for such a method. Thus, the identification of such genetlc changes are urgently required. Progression of prostatic cancer to a metastatic state involves multiple cellular and genetic changes (Prostate 25: 249, 1994). Using somatic cell hybridization, it was shown that acquisition of metastatic ability requires both the loss of metastasis suppressor function(s) and the activation of oncogenes (Cancer Res 51: 3788, 1991). In further studies using microcell-mediated chromosomal transfer, a gene(s) on human chromosome 8 (Cancer Res 54: 6249, 1994) was mapped, which when introduced into rat prostatic cancer cells are capable of suppressing metastatic ability without affecting tumorigenicity or growth rate in vivo. Attempts to positionally clone the metastasis suppressor gene(s) in the 11p11.2-13 region. Microcell-mediated chromosornal transfer was used to document that this chromosomal 11 location harbors genes that suppress the metastatic ability of rat prostatic cancer cells. Alu PCR was used to amplify human specific sequences encoded by the human 11p11.2-13 region in the metastasis suppressed rat prostatic cancer hybrid cells. Partial sequences of these PCR products were used to generate oligonucleotides probes for screening cDNA libraries. This method identified cDNA clones which were specifically expressed in the metastasis-suppressed cells.. One such clone, termed KAI-1, encodes a 267 amino acid protein with four hydrophobic transmembrane domains and one extracellular hydrophilic domain with three potential N-glycosylation sites (Science 268: 884, 1994). FISH analysis of metaphase spreads of human lymphocytes, using a cDNA probe for KAI-1, demonstrated that it is maps to the p11.2 region of human chromosome 11. The expression of KAI-1 is detected in all normal human tissues tested. The coding region of KAI-1 is evolutionarily conserved from yeast to human. The human KAI-1 was introduced into highly metastatic rat prostatic cancer cells via DNA transfection and resulting KAI-1 high expressing clones isolated. These KAI-1 expressed clones were significantly suppressed in their metastatic ability. KAI-1 expression is high in human normal prostate and benign prostatic hyperplasia but dramatically lower in cancer cell lines derived from metastatic prostate tumors. Presently, matched pairs of normal prostate and primary and metastatic human prostate tumors from the same patient are being examined for deletion, mutation, and loss of heterozygosity and expression of KAI-1 gene.
Dr. Ryuichi Yatani spoke on geographic pathology of the prostatic carcinoma. Prostatic carcinoma is one of the most common malignant neoplasia in Western Europe and in the USA, and its prevalence in Japan still at lower levels. However, the extension of life span and Westernized life-style in Japan have led to increase of prostate cancer among Japanese. Also comparative studies between Japanese and Americans have shown that the characteristics of the prostatic carcinoma among the two populations are different in many aspects. To find special characteristics of prostate cancer for different populations, we have performed international comparison studies about pathological and molecular aspects of latent and clinical prostatic carcinomas. Evaluation of prostates from Japanese, Hawaiian Japanese, white Americans and black Americans over 50 years old demonstrated that the age adjusted overall prevalence of latent prostatic carcinoma was significantly higher in black Americans (36.9%) and white Americans (34.6%) than in Japanese (20.5%). There was no significant difference between Hawaiian Japanese (25.6%) and Japanese in Japan. These differences were mainly for older persons with moderately or poorly differentiated carcinomas (clinical stage A2). No racial differences were observed for the well differentiated cases (clinical stage A1). Comparing the incidence of latent cancer in relation to clinical cancer we observed that the ratio was 138:1 in Japanese, 25:1 in Hawaiian Japanese, 25:1 in white Americans, and 13:1 in black Americans. In addition, the Japanese prostates collected from 1982 to 1993 had a higher rate of latent cancer, showing the rate of 6% and 17% in the fourth and fifth decades of age, than those collected from 1965 to 1979 in each decade. In relation to clinical cancers our comparative study about incidence, histological type, and stage in Japanese, Chinese, Korean, Hawaiian Japanese, white Americans, and black Americans has demonstrated interesting results. For the clinical stage, the incidence of stage B cancers was higher in Hawaiian Japanese and white Americans. Conversely, for Japanese the stage C and D cancers and for Koreans the stage D cancers were more frequent. According to the grade of differentiation we observed that poorly differentiated cancers was rarely seen. Comparing the clinical stage and the histological degree of differentiation we observed that in Japanese, Hawaiian Japanese, and white Americans as much as the clinical stage was advanced the tumor was more undifferentiated. However, for Chinese and Koreans, the clinical stage and the grade of differentiation were not correlated.
About the distribution within the prostate, cancers from the inner portion were more differentiated and had lower progression properties than those from the outer portion of the gland. Cancers derive from the inner portion of the prostate were more frequent in Japanese than in white and black Americans. In Japanese and also in Americans, there was a significant relationship between the volume and the grade of invasion of the tumors. For example, tumors smaller than 0.3 cm3 did not show any capsular invasion, and for tumors smaller than 1.4 cm3 there was no capsular penetration nor distant metastasis. Using polymerase chain reaction technique, racial differences in the frequency and type of mutation of ras oncogene in latent carcinomas were found. Point mutations of the ras family genes were relatively high in Japanese and Columbians: Mutations of K-ras codon 12 and N-ras codon 12 were found in 14% and 9% of Japanese cases, respectively. In contrast, the mutation of K-ras codon 12 alone was found in only 5% of Hawaiian Japanese, H-ras codon 12 in 7% of Caucasian American and H-ras codon 61 in 7% of African-American. The frequency of ras gene mutations and mutation spectrum appear to vary in different ethnic groups. As for p53 tumor suppressor gene, the frequency of p53 gene mutation was low, however was detected in some cases in advanced stages. The p53 gene mutation may play a role in the progression of a subgroup of prostate cancer in Japanese as observed for Americans. Interestingly, a difference was noted between Japanese and Americans in the p53 mutational spectrum, presumably arising from variation In the underlying etiologic factors of specific nutrient supplement as an alternative.
PARTICIPANTS
UNITED STATES
Dr. J Carl Barrett
Director, Environmental Carcinogenesis Program
National Institute of Environmental Health Sciences
National Institutes of Health
P.O. Box 12233
Research Triangle Park, NC 27709
Dr. Robert B. Dickson
Associate Director for Basic Science
Professor of Cell Biology and Pharmacology
Lombardi Cancer Research Center
Georgetown University Medical Center
3800 Reservoir Road, N.W.
Washington, DC 20007
Dr. JefBrey Green
Laboratory of Molecular Oncology
NCI-FCRDC
P.O. Box 8
Frederick, MD 21701-1201
Dr. David Hunter
Epidemiology
Harvard School of Public-Health
Channing Laboratory
180 Longwood Avenue
Boston, MA 02115
Dr. William Isaacs
James Buchanan Brady Urological Institute
Johns Hopkins Hospital
600 N Wolfe Street
Baltimore, MD 21287
Dr. Helene Smith
Director
Geraldine Brush Cancer Research Institute
2330 Clay Street
San Francisco, CA 94115
JAPAN
Dr. Mitsuru Emi
Nippon Medical School
1-396 Kosugi-cho, Nakahara-ku
Kawasaki 211
Dr. Kenji Kamiya
Research Institute for Radiation Biology
and Medicine
Horoshima University
1-2-3 Kasami, Minami-ku
Horoshima 734
Dr. Hoshinobu Kubota
Yokohama City University
3-9 Fukuura, Kanagawa-ku
Yokohama 236
Dr. Minako Nagao
National Cancer Center Research Institute
101 Tsukiji 5-chome, Chuo-ku
Tokyo 104
Dr. Yohiyuki Ohno
Nagoya University School of Medicine
65 Tsurumai-cho, Showa-ku
Nagoya 466
Dr. Jun Shimazaki
Chiba University School of Medicine
1-8-1, Inobana, Chuo-ku
Chiba 280
Dr. Masatoshi Watanabe
National Cancer Research Institute
1-1 Tsukiji 5-chome, Chuo-ku
Tokyo 104
Dr. Ryuichi Yatani
Dean
Mie University School of Medicine
2-174 Edobashi, Tsu, Mie 514
(4) Seminar on “Experimental Studies on Genetic Intervention for Cancer Prevention”
This seminar was held on March 23-25, 1995, in accordance with the United States - Japan Cooperative Cancer Research Program at the Hapuna Beach Prince Hotel, Kona, Hawaii. There were 12 participants, six from Japan and six from the United States.
Dr. Aaronson welcomed the attendees with opening remarks. He provided attendees with interesting historical perspectives concerning the US - Japan program and highlighting some of the important contemporary issues in cancer research.
Dr. Masabumi Shibuya presented studies on the role of VEGF (Vascular Endothelial Growth Factor)-Flt (fms- like tyrosine kinase) receptor system in normal and tumor angiogenesis. The endothelial cell system is known to be important not only in physiological condition such as vasculogenesis in embryogenesis, but also in many pathological conditions: angiogenesis in wound healing, ocular diseases in diabetes and tumor angiogenesis in vivo. Although several angiogenic factors, acidic and basis FGFs, TGF!!
!and TNF have been reported, endothelial cell-specific growth factors and signal transductions system are still poorly understood. Recently a new endothelial cell-regulatory factor, VEGF/VPF (Vascular Permeability Factor) has been molecularly cloned and characterized. Furthermore, Flt (now designated as Flt-1), which Dr. Shibuya’s group originally isolated from human placenta cDNA library, has recently been shown to bind specifically with VEGF/VPF at high affmity. Dr. Shibuya then discussed the significance of the VEGF-Flt receptor system in endothelial cell-specific growth and differentiation as well as in regulation of vascular permeability. First he showed that VEGF-Flt receptor system has an important role in reciprocal communications between parenchymal cells and endothelial cells in normal tissues. Using rat liver tissue, his group found that two receptor-type tyrpsine kinase-mediated systems, VEGF-Flt receptors and HGF-c-Met receptors are expressed in a cell-type specific manner, and that VEGF has strong growth stimulatory and maintenance activities on sinusoidal endothelial cells.
Next, he presented some data upon paracrine stimulation of tumor angiogenesis through the VEGF-Flt system using primary and transplantable human glioblastomas as examples. In addition to the expression of VEGF, he found expression of Placenta Growth Factor (PlGF) gene, a second gene related to VEGF in these tumors. He purified PlGF using the Baculovirus system and found that PIGF is also able to bind to Flt-1 but not to KDR(Flk-1), another VEGF receptor related to Flt-1.
Although VEGF was originally isolated as a vascular permeability factor, precise mechanisms of the increase in permeability of the vessels induced by VEGF have not been studied. Dr. Shibuya presented evidence that the increase In vascular permeability induced by VEGF is a rapid, transient and probably protein-synthesis-independent process. Unique signal transduction from the Flt receptor family was also discussed.
Finally, he showed some data on the genomic structure of flt-1 gene. To understand the phylogenetical relationship between fms family (fms/kit/flt2 involved in hematopoiesis) and the flt family (involved in vascular system), he isolated all exons of the flt-1 gene from genomic DNA. Based on the exon-intron structure, Dr. Shibuya suggested that the flt gene was evolutionarily generated by insertional mutation of the fms family gene.
Dr. David Cheresh described how integrin aVb3 antagonists promote tumor regression by inducing apoptosis of angiogenic blood vessels. A single intravascular injection of a cyclic peptide or monoclonal antibody antagonist of integrin aVb3 disrupts ongoing angiogenesis on the chick chorioallantoic membrane. This leads to the rapid regression of histologically distinct human tumors transplanted onto this membrane. In fact, aVb3 antagonists also prevent the spontaneous pulmonary metastasis of human melanoma cells. All human tumors examined in this model are aVb3 negative, which suggests that these antagonists have no direct effect on the tumor cells themselves.
Induction of angiogenesis by a tumor or cytokine promotes entry of vascular cells into the cell cycle and expression of integrin aVb3. After angiogenesis is initiated, antagonists of this integrin induce apoptosis of the proliferative angiogenic vascular cells, leaving preexisting quiescent blood vessels unaffected. These studies demonstrate, therefore, that ligation of integrin aVb3 is required for the survival and maturation of newly forming blood vessels, an event essential for the proliferation and metastatic properties of human tumors.
Dr. Masaaki Terada discussed a new approach for genetic intervention. Several methods that enable foreign genes to be transferred into cells and adult animals have been developed. These methods make it possible to understand biological function of genes, and can potentially be useful for genetic interventions.
Dr. Terada’s group developed a new method to introduce a foreign gene efficiently into fetuses; a single intravenous injection of expression plasmid: lipopolyamine complexes into pregnant mice resulted in successful gene transfer in mouse embryos. The transgenes that were introduced were expressed in fetuses and newborn progeny. This method of gene transfer into the embryos facilitates rapid analysis of transgene effects in the fetuses and should be useful for studying gene-deficient animal models to gain transgene functions at desired stages of embryogenesis. The methods also can be used to have fetal gene therapy for prevention of some genetic diseases including hereditary cancers.
Another approach was taken for genetic intervention or for understanding some of the biological function of genes in adult animals. Dr. Terada used an adenovirus vector to understand biological function of hst-1 gene in vivo. When adenovirus-hst-1 recombinant virus was injected into mice, there was a marked increase in the number of peripheral platelets without changes in the other hematological findings. The increase in platelet counts lasted for the following 3 to 4 weeks.
When a mouse was pre-treated with cisplatin or irradiation, there was a marked decrease in number of platelets. When adenovirus containing hst-1 gene was administered prior to administration of cisplatin or treatment of irradiation, the decrease of platelets was not a physiological response. Dr. Terada also presented a new method and data showing that hst-1 protein plays an important physiological role in the development of limb buds. He also presented results showing that in tumors, an amplification unit contains several genes, and that there are multiple amplification units in a single cancer cell.
Dr. Mike Rothe closed the morning session with a presentation of a novel family of signal transducers associated with the cyioplasmic domain of the 75k-D tumor necrosis factor receptor (TNFR). Mutational analysis identified the C-terminal region of 78 amino acids within the cyioplasmic domain of the human 75-kD tumor necrosis factor receptor (TNF-R2) that is required for signal transduction. This region was subsequently shown to mediate the interaction of cyioplasmic factors with TNF-R2. Two of these factors were isolated and molecularly cloned using biochemical purification and the yeast two-hybrid system. TNF receptors associated factor l (TRAF1) and TRAF2 are the first two members of a new protein family containing a novel C-terminal sequence homology motif, the TRAF domain. In addition, TRAF2 contains an N-terminal RING finger motif.. TRAF1 and TRAF2 are associated with the cyioplasmic domain of TNF-R2 in a heterodimeric complex. In this protein complex, TRAF2 contacts the receptor directly, while TRAF1 interacts with TNF-R2 indirectly through heterodime formation with TRAF2. Functional studies indicate that the TRAFs are required for TNF-R2 signal transduction.
Dr. Hirai described the mechanism of leukemogenesis by AML1/EVI-1 chimeric Hisamaru factor in t(3;21) leukemia. Since chimeric proteins generated by chromosomal translocations often play causative roles for leukemogenesis, it is important to investigate both their biological activities and biochemical functions. The t(3;21)(q26;q22) translocation, found in the blastic crisis phase of chronic myelogenous leukemia and in myelodysplastic syndrome-derived leukemia, is thought to cause leukemic cell transformation of hematopoietic stem cells. Dr. Hirai reported the molecular and biochemical analysis of the t(3;21) translocation to reveal the mechanisms for leukemogenesis in t(3;21) leukemias.
The AML1 gene, which is located at the translocation breakpoint of the t(8,21) translocation found in acute myelocytic leukemia, was also rearranged by the t(3,21) translocation. Screening of a cDNA library of the t(3;21)-carrying leukemic cell line resulted in the isolation of AML1/EVI-1 chimeric cDNA. In AMLl/EVI-1 fusion protein, AML1 is disrupted at the end of the runt domain and fused with the entire zinc finger EVI-1 protein. The AMLl/EVI-1 fusion transcript was consistently found among all cases of the t(3;21)-carrying leukemia examined by RNA-based PCR. AML1 is a transcription factor, which is highly homologous with Drosophiia pair-rule gene runt and mouse polyomavirus enhancer binding protein (PEBP) 2!!
!A. AML1 has a runt domain responsible for DNA binding to a PEBP2 consensus sequence and heterodimer formation with a human homologue of PEBP2b EVI-1, which is not normally expressed in normal hemopoietic cells, is a transcription factor with two zinc finger domains. EVI-1 was first identified as a gene existing in a common locus of retroviral integration in murine myeloid leukemias. Recently, Dr. Hirai showed that EVI-1 transactivatesc-fos promoter and raises AP-l activity with dependence on the second zinc finger domain. He investigated the biological role of AMLl/EVI-1 fusion protein in leukemogenesis, employing the 32Dc13 murine IL-3-dependent myeloid cell line, which differentiates to mature granulocyies when treated with granulocyte colony-stimulating factor (G-CSF). He established several 32Dc13 stable cell clones expressing AMLl/EVI-1 fusion protein by transfecting the AML1/EVI-1 expression plasmid. The established clones were tested for their abilities to differentiate and proliferate in the presence of IL-3 or G-CSF. The results demonstrate that the AML1/EVI-1 fusion protein blocks terminal differentiation to mature granulocyies and stimulates cellular proliferation in 32Dc13 cells.
In order to evaluate transforming activity of this protein, AML1/EVI-1 was introduced into Rat1 fibroblasts. The cells expressing the fusion product formed macroscopic colonies in soft agar, indicating that AML1/EVI-1 is a transforming gene.
The transactivation abilities of AML1 through the PEBP2 site were reduced in the presence of AML1/EVI-1 protein, indicating that AML1/EVI-1 dominantly and specifically suppresses the transactivation by AML1. It was revealed that AML1/EVI-1 as well as AML1 specifically binds to a PEBP2 site. The affimity of AML1/EVI-1 for the PEBP2 site was several-fold higher than that of AML1, supporting the hypothesis that AML1/EVI-1 dominantly suppresses AML1 transactivation by competing with AML1 for binding to the PEBP2 site. Transcriptional analysis demonstrated that AML1/EVI-1 expression raises AP-1 activity as does EVI-1. AP-1 activation by AML1/EVI-1 could be correlated with cellular transformation because AML1/EVI-1 is a chimeric transcription factor with dual functions. Experimental results indicate that these dual functions should contribute to leukemic cell transformation. Oncoproteins with dual functions are possibly advantageous to oncogenesis because multistep functional abnormalities of oncogenes are accumulated and finally result in the cell transformation. Thus, AML1/EVI-1 should trigger at least two steps of functional abnormalities required for leukemogenesis.
Dr. Arnold Levine described the regulation of p53 transcription factor activity. The p53 protein is a transcription factor that positively or negatively regulates a set of genes. When activated by p53, these p53-responsive gene products (or their absence) can mediate a G1 arrest in the cell cycle or initiate a program leading to cell death or apoptosis. The p53 protein responds, by increasing its concentration in a cell, to damaged DNA. Thus, p53 mediates the communication between damaged DNA or the proteins involved in its repair and the regulation of the cell cycle or even cell death.
The p53 protein has been divided structurally and functionally into four domains:
1. Amino acid residues 1-42 at the N-termlnus contact two proteins, TAF-31 and TAF-70, in the basal transcription factor, TFIID, which contains the TATA-binding protein located about 25-30 bases prior to the transcription start site. Amino acid residues 14, 19, 22 and 23 in this p53 domain are the critical contact points for TAF-31.
2. Residues 120-290 contain the DNA sequence specific binding domain of p53, located in genes that p53 regulates in a positive fashion.
3. Residues 305-364 contain the protein tetramerization domain, the nuclear localization signal and a cyclin dependent kinase site for phosphorylation.
4. Residues 364-390 contain a flexible protein domain that recognizes DNA or RNA nonspecifically as well as damaged DNA or replication errors. This domain can allosterically or sterically regulate the DNA sequence specific binding domain at residues 120-290 in this protein. In this fashion, the p53 protein can communicate DNA damages or mutations to the transcription factor functional domains which in turn communicate with cell cycle progression.
Dr. Mutsuo Sekiguchi presented the results of his studies on the human mutT homologue gene MTH1 and its possible role in preventing mutations and cancers. Errors in the replication of DNA are a major cause of spontaneous mutation, and organisms are equipped with elaborate mechanisms to avoid such errors and to correct errors after DNA replication. He reported evidence for a novel mechanism that prevents replicational errors by hydrolyzing a mutagenic substrate for DNA synthesis. 8-Oxo-dGTP, an oxidized form of dGTP, can be inserted into opposite adenine and cytosine residues of template DNA with almost equal efficiency. In Escherichia coli, this misincorporation is prevented by MutT protein, which degrades 8-oxo-dGTP to the monophosphate. Lack of the mutT gene Increases the occurrence of A:T to C:G transversions 100- to 10,000-fold over the wild type level.
Human cells contain an enzyme similar to the MutT protein of E. coli, and this enzyme specifically hydrolyzes 8-oxo-dGTP to 8-oxo-dGMP, with a relatively low Km value, as compared with other deoxyribonucleoside triphosphates. The enzyme was purified to physical homogeneity and, based on the partial amino acid sequence, the cDNA was cloned. Dr. Sekiguchi obtained evidence that expression of human cDNA in E. coli mutT-mutant cells suppresses almost completely the occurrence of specific A:T to C:G transversion mutation. Thus, it is most likely that the human 8-oxo-dGTPase has the same antimutagenic capacity as the MutT protein.
To pursue this problem further, it is necessary to elucidate the genomic structure of human 8-oxod-GTPase. Dr. Sekiguchi showed that the MTH1 gene is composed of at least 4 exons and spans about 9 kb. Using as probes appropriate fragments of the genomic sequence, he mapped the gene to human chromosome 7p22. Whether or not 7p22 is related to a certain type of inherited disease or to familial cancer is unknown, but it is likely that a defect in this locus could lead to a high incidence of cancer as well as a high frequency of spontaneous mutation.
Dr. Guo-Min Li demonstrated that h-MutL and h-MutS function as heterodimers in human mismatch repair. Human cell nuclear extracts support strand-specific mismatch correction in a reaction that is similar to E. coli methyl-directed repair with respect to both mismatch specificity and unusual features of mechanism. Like the bacterial system, the human pathway also functions in mutation avoidance since several classes of mutator human cell line are defective in the reaction. These include an alkylation-tolerant line that is resistant to the cytotoxic action of N-methly-N’-nitro-nitrosoguanidine, as well as a number of hypermutable RER+ tumor cell lines. Application of in vitro complementation and traditional fractionation approaches has indicated that the human repair reaction is dependent on at least seven activities. Activities that complement extracts derived from HMLH1- and hMSH2- deficient cells have been obtained in homogeneous form. Both behave as heterodimers, in the former case comprised of hMLH1 and hPMS2 proteins, and in the latter of hMSH2 complexed with a 160 kDA polypeptide Drs. Li and Modrich have found that the MT1 alkylation-tolerant cell line, and the HCT15 and DLD1 colorectal tumor cell lines are selectively defective in correction of base-base mispairs, but are proficient in repair of 2, 3, or 4 nucleotide insertion/deletion mispairs. This differential repair defect distinguishes these three lines from previously characterized hMLH1 or hMSH2-defective cells, which are deficient in repair of both classes of mismatch. Since extracts of MT1 and HCT15 cells are complemented by the (hMSH2)1-(p160)1 heterodimer, and since the latter line appears to be free of hMSH2 mutations, it is likely that the three lines displaying partial repair defects will prove to harbor mutations in the gene encoding p160.
Dr. Jerry Shay addressed his presentation to telomeres, telomerase and tumors. Due to the inability of DNA polymerase to replicate linear DNA at the chromosome (telomere) ends, cells lose 50-200 nucleotides per cell division (end-replication problem). This progressive shortening may be an important molecular mechanism by which cells count their divisions. For example, when telomeric repeats become sufficiently short, a DNA damage signal may be elicited, initiating replicative senescence. In contrast to normal cells, immortal and tumor cells show no net loss of average telomere length with cell division, suggesting that telomere stability may be required for cells to escape from replicative senescence and proliferate indefinitely. Telomerase is a ribonucleoprotein enzyme that synthesizes telomeric DNA in normal germline tissues (such as testes and ovaries) compensating for the end replication problem, thus protecting and stabilizing telomeres.
Dr. Shay developed a PCR-based assay to measure telomerase activity. His recent studies (Science, 266:2100-2015, 1994) demonstrate that telomerase is expressed in almost all immortal human cell lines and primary tumors but is not expressed in normal cells and tissues (except germline tissues). In a survey of primary tumors from 12 tissue types including those of the breast, brain, colon, prostate, lung, uterus, adrenal, head and neck, he demonstrated that 90/101 were positive for telomerase activity, while normal tissues from autopsied patients and mitotically active growths such as uterine leiomyomas (fibroids) were negative. Thus, telomerase is likely to be required for the continued proliferation of almost all human tumors.
In a more comprehensive study (Nature Medicine, 1:249-257, 1995) Dr. Shay’s group demonstrated that 94/100 primary neuroblastoma tumors expressed telomerase activity. Three of the tumors that did not express telomerase activity were in the special 4s category that have distal metastases but often spontaneously regress. The 4s tumors almost always have very short telomeres compared to adjacent normal tissue and he believes that the tumor regression may be due to lack of telomerase activity, resulting in the tumor cells “running out” of proliferation potential. In additional studies yet to be published, Dr. Shay demonstrated that 126/153 (82%) lung carcinomas, 44/55 (80%) renal cell carcinomas, 23/27 (85%) prostate carcinomas, 28/33 (85%) hepatocellular carcinomas, 45/50 (90%) breast carcinomas, and 56/66 (85%) gastric cancers were positive for telomerase activity. Overall these results indicate that reactivation of telomerase may be a critical, and perhaps rate-limiting, step in the progression of most human cancers.
Hybrids between immortal cells that express telomerase and normal cells that lack telomerase have a limited life-span. In another recent study, Dr. Shay found that telomerase activity is repressed in such hybrids. Treatment of lmmortal human cell lines with short oligonucleotides that can serve as substrates for telomerase resulted in telomere elongatlon possibly by altering feedback mechanisms that regulate telomerase activity. He took advantage of this observation to test the hypothesis that elongation of telomeres would extend the life-span of cells in culture. An important human cell line was treated with an oligonucleotide to lengthen its telomeres and then was fused with mortal cells. The life-span of these hybrid cells was no longer than that of the hybrids in which telomeres had not been elongated. These observations provide the first direct evidence for the hypothesis that telomere length determines proliferative capacity of human cells.
Cancer is generally an age-related, multistep process involving numerous genes. Cellular immortalization may be a critical or rate-limited step in these processes. Since telomerase appears to be vital to the continued proliferation of cancer cells, development of anti-cancer agents based on telomerase inhibition may be highly efficacious. Normal somatic cells do not express telomerase, so this type of agent may also possess great specificity, low toxicity and reduced side effects. In addition, tests for telomerase activity in primary tumors may be useful for diagnostic and prognostic purposes.
Dr. Takao Sekiya was the final speaker for the morning session and discussed analysis of genomic DNA including uncharacterized regions: detection of genes and DNA aberrations. To detect aberrations of genomic DNA in human cancers even in uncharacterized regions, DNA fragments were amplified by arbitrarily primed-polymerase chain reaction using a single primer (AP-PCR) and compared semi-quantitatively among template DNAs from primary and metastatic cancers and normal tissue of the colon. Results revealed frequent gain of copy numbers of chromosomes 8 and 13 in advanced carcinomas but not in less-malignant cancers. The use of a single primer in AP-PCR could amplify DNA fragments from average 30 regions of genomic DNA and the sum of lengths of DNA fragments amplified were about 17 kilobases. To bring out the hidden information in such AP-PCR products, the products were further analyzed by single-strand conformation polymorphism (SSCP) method. When 19 samples of cancer DNA from 16 patients with hepatocellular carcinomas (HCCs). Using the 23 primers, nucleotide sequence polymorphisms were detected in about 100 loci. In this study, 289 DNA fragments amplified from these 54 regions showed heterozygosity and cancer specific losses of heterozygosity were detected in 13 of these informative nucleotide sequences in 5 HCCs. By AP-PCR-SSCP analysis, Dr. Sekiya’s group could also find a cancer specific base-substitution in one HCC. AP-PCR-SSCP analysis could be a useful method for detection of genomic alterations in uncharacterized regions in cancer cells.
It is quite a job to detect anonymous genes within a large region of unsequenced DNA.. A method for preferential isolation of DNA fragments associated with CpG islands has been developed. The method includes denaturing gradient gel electrophoresis of cloned DNA fragments digested with four restriction endonucleases recognizing AT-rich sequences. The principle of the method is retention of DNA fragments with long GC-rich sections in the gel after disappearance of most others due to inhibition of strand dissociation after the fragments reach retardation level in the gradient. Nucleotide sequences of the retained fragments show that about half of these fragments appear to be derived from CpG islands. Northern analysis indicated the presence of RNA complementary to most of the retained fragments. This method of segregation of partly melted molecules (SPM) is expected to provide means of convenient and efficient isolation of genes from unsequenced DNA.
Dr. Tadashi Yamamoto presented the results of his studies on the signaling pathway unique to the receptor-type kinase ErbB-2. Arnplification and/or over-expression of the human c-erbB-2 gene often occur in breast cancer and have been correlated with poor clinical prognosis. This suggests that elevated expression of the gene product p185erbB2 results in aggressive tumor growth in vivo. Yet, the precise mechanism of signaling through p 185erbB2 remains to be established.
During search for the ligand specific to ErbB-2, Dr. Yamamoto’s group found that 17b-estradiol (E2) bound to the ErbB-2 protein. Scatchard analysis revealed that the Kd value for the binding was 2.7 nM. E2 treatment of ErbB-2 expressing NIH3T3 derivatives (RC cells) resulted in down-regulation of the protein. E2 also induced morphological transformation of the RC cells but not of the parental NIH3T3 cells. Thus, estrogen is suggested to play an important role in ErbB-2 mediated signaling. Basing on these observations, Dr. Yamamoto prepared nitrogen mustard-conjugated E2 and showed that this compound had killing activity over the ErbB-2 expressing cells.
Several SH2-domaln containing molecules are believed to interact with p185erbB2, tyrosine phosphorylated at the noncatalyiic carboxyl-terminal tail. One of the important pathways mediated by the SH2-containing molecules is the activation of Ras, which is important for stimulation of cell growth. However, Dr. Yamamoto previously demonstrated that p155dCerbB2, having 230 amino acids truncation of the carboxyl-terminus, is mitogenic for NIH3T3 cells. By screening lgt11 expression library with the p155dCerbB2 protem, cDNA clones for a novel protein termed ToB (Transducer of ErbB-2) was obtained. The ToB protein is partly homologous to the BTG-1 gene product that shows anti-proliferative activity. ToB does not contain the SH2 or SH3 motif. Elevated expression of exogenously introduced ToB remarkably suppressed growth of NIH3T3 cells, but growth suppression by ToB was hampered by the presence of the activated c-erbB-2 gene product. Thus, Dr. Yamamoto proposed that the kinase activity of the c-erbB-2 gene product negatively regulates a ToB-mediated anti-proliferative pathway.
Last to speak was Dr. Stuart Aaronson, whose presentation focused on new RAS-related oncogenes of human cancers. Signaling pathways that mediate the normal functions of growth factors are commonly subverted in cancer. Oncogenes identified by a variety of approaches have been shown to function at critical steps in mitogenic signaling. Progression through the cell cycle requires the coordinated actions of members of two complementary classes of growth factors, and oncogenes appear to replace the actions of one set of these growth factors. Growth factors can also influence normal cell differentiation, and constitutive activation of promoting-promoting pathways in cancer cells can modulate the cell phenotype as well. Paracrine actions of growth factors and cytokines may also influence the stepwise series of genetic events that lead to malignancy, and some examples were described.
In the search for novel oncogenes in tumors from which such genes have not yet been identified, Dr. Aaronson’s group applied an expression of cDNA cloning strategy. He generated an expression cDNA library from an ovarian carcinoma line. A potent transforming gene was detected by transfection analysis and identified as TC21, a recently cloned member of the RAS gene superfamily. A single point mutation substituting glutamine for leucine at position 72 was shown to be responsible for activation of transforming properties. While the cDNA clone possessed high transforming activity, the ovarian tumor genomic DNA, which contained the mutated TC21 allele, failed to induce transformed foci. Thus, expression cDNA cloning made it possible to identify and isolate a human oncogene that has made it possible to identify and isolate a human oncogene that has evaded detection by conventional approaches. Approaches aimed at determining the prevalence of this and other ras--related oncogenes in human tumors were discussed.
Dr. Terada thanked the participants for their contributions to a meeting characterized by impressive progress and new concepts. The material presented provided new insights into the basic genetic mechanisms of promoting regulation, and new approaches which had immediate relevance to the basic cancer researcher as well as potential application for genetic intervention for the prevention of cancer at the clinical level.
Saturday, March 25, a business meeting was held mediated by both Dr. Masaaki Terada and Dr. Stuart A. Aaronson.
PARTICIPANTS
UNITED STATES
Dr. Stuart A. Aaronson
The Derald H. Ruttenberg Cancer Center
One Gustave L. Levy Place, Box #1130
New York, NY 10029
Dr. David Cheresh
Scripps Research Institute
10666 North Torrey Pines Road, IMM24
La Jolla, CA 92037
Dr. Arnold Levine
Princeton University
Lewis Thomas Laboratory
Washington Road
Princeton, NJ 08544
Dr. Guo-Min Li
Duke University
Department of Biochemistry
Research Drive
Nanaline Duke Building, Room 150
Durham, NC 27710
Dr. Mike Rothe
Tularik, Inc.
270 East Grand Avenue
South San Francisco, CA 94080
Dr. Jerry W. Shay
University of Texas Southwestern Medical Center
5323 Harry Hines Boulevard
Dallas, TX 75235-9039
JAPAN
Dr. Masaaki Terada
National Cancer Center Research Institute
1-1 , Tsukiji 5-chrome
Chuo-ku, Tokyo 104, Japan
Dr. Hisamaru Hirai
University of Tokyo, Faculty of Medicine
3-1 , Hongo 7-chrome
Bunkyo-ku, Tokyo 113, Japan
Dr. Mutsuo Sekiguchi
Medical Institute of Bioregulation
Kyushu University
1-1 Maidashi 3-chome
Higashi-ku, Fukuoka 812, Japan
Dr. Takao Sekiya
National Cancer Center Research Institute
1-1, Tsukiji 5-chome
Chuo-ku, Tokyo 104, Japan
Dr. Masabumi Shibuya
Institute of Medical Science
University of Tokyo
6-1 Shirokanedai 4-chome
Minato-ku, Tokyo 108, Japan
Dr. Tadashi Yamamoto
Institute of Medical Science
University of Tokyo
6-1 Shirokanedai 4-chome
Minato-ku, Tokyo 108, Japan