REPORTS ON SEMINARS

(1) Seminar on "Xenobiotic Metabolizing Enzymes and Their Roles in Cancer Susceptibility"

This workshop was held on February 20-22, 1994, at the Princeville Hotel, Kauai, Hawaii. The organizers were Dr. Ryuichi Kato, Keio University, Tokyo, and Dr. Frank J. Gonzalez, National Cancer Institute, Bethesda, Maryland. There were seven participants from the United States, one from Great Britain, seven from Japan and two observers from Japan.
The purpose of the seminar was to discuss and exchange information on the xenobiotic enzymes responsible for metabolic activation of chemical carcinogens. It has been recognized for many years that these enzymes are required for the process of chemical carcinogenesis. They convert inert procarcinogens to their proximate DNA-binding metabolite. Mutations at the cellular level can then begin to drive the process of cancer initiation and progression. Since humans have varying susceptibilities to cancer, it is suspected that levels of the activating enzymes could be a determinant in cancer risk. Indeed, these enzymes exhibit a large degree of inter-individual variability in their expression and some of this variability is genetically determined. Published studies over the past few years have indicated that levels of the polymorphic P450s and certain transferase enzymes are associated with cancer risk. The U.S. participants were largely studying human xenobiotic metabolism and cancer association, while the majority of Japanese participants were concerned with animal models. Thus, a balanced program of fundamental basic research and applied and developmental human studies was achieved. A number of relationships were formed in an effort to expand collaborative efforts.
Dr. Jeffrey Idle opened the workshop with a discussion of the history of pharmacogenetics. He noted earlier pivotal studies by the late Professor R.T. Williams, who was among the first to recognize the marked species differences in xenobiotic metabolizing enzymes. Dr. Idle suggested that the field of drug/xenobiotic metabolism is not as active as it once was with respect to carcinogen activation/inactivation and that we still have a lot to learn in this area. During the mid-1970s, Dr. Idle's group discovered a polymorphism in a cytochrome P450 CYP2D6 and a later pivotal study was the first to establish a possible link of a P450 polymorphism with cancer. Other polymorphisms in N-acetyltransferase-1 and a form of glutathione S-transferase are also associated with cancer risk. During the 1980s and early 1990s, the field of molecular epidemiology has exploded with new findings, indicating the association of drug oxidation polymorphisms with cancer risk; but many of the studies are highly flawed due to the low involvement of knowledgeable epidemiologists. Dr. Idle said that the time is ripe to launch prospective studies to determine association of polymorphisms with cancer risk.
Dr. Stephanie J. London discussed the intricacies of designing a proper study to address disease associations in case-control studies. Frequently, these studies analyze cancer cases as compared to a group of controls without careful consideration of the composition of the control group. The design and recruitment of a proper control population requires considerable effort that is often not appreciated by those not trained in epidemiology. Dr. London emphasized that study design is critical and that a great deal of effort is required to recruit cases and proper controls. She stressed that ethnic variation, age, sex, history of possible carcinogen exposures, and other factors need to be considered. By analyzing the numerous studies published to date attempting to establish a relationship between the CYP2D6 polymorphism and cancer risk but yielding different results, she felt that perhaps a single study would pass the scrutiny of an expert epidemiologist. Dr. London discussed her efforts in designing a large study in collaboration with Dr. Idle, in which over 800 lung cancer cases and an equal or greater number of matched controls are enrolled. Several polymorphic enzymes will be monitored.
Dr. Robert Branch is studying the association of bladder cancer with drug oxidation polymorphisms. Case-control studies were carried out analyzing levels of CYP2C19 (S-mephenytoin 4-hydroxylase), CYP3A4 (dapsone N-hydroxylation), and CYP2D6 (debrisoquine 4-hydroxylase) using cocktails of the three drugs. Patients with aggressive bladder cancer (G3 histopathology) were found to have an under-representation of poor metabolizers. These patients also had a low dapsone recovery ratio and a history of excess alcohol intake. No differences were found between controls and patients in CYP2C19 activity. In patients with non-aggressive bladder cancer, tumor recurrence occurred significantly less frequently in patients in the lowest tertile of debrisoquine recovery ratio. Together, these data suggest an involvement of P450 substrates/procarcinogens in the etiology of bladder cancer and that measurements of activity of selected P450s might predict subjects at high risk of developing bladder cancer.
Dr. Minroe Watanabe discussed his studies on the association of CYP2E1 genotypes with lung cancer. A restriction fragment length polymorphism using the enzyme Dra I was found and one of the alleles was over-represented in cancer patients as compared to matched controls. An association was also found between extent of smoking and genotype. Dr. Watanabe also presented data on the mechanism of regulation of CYP1A1 in the experimental hamster model. In contrast to the rodent systems, the CYP1A1 gene was not significantly induced by 3-methylcholanthrene in liver but was markedly induced in lung. Studies are in progress analyzing the regulatory elements of the hamster gene.
Dr. Douglas Bell spoke on his studies of smoking-associated bladder cancer. Homozygous deletion of a glutathione S-transferase GSTM1 confers an 80 percent increased risk for this cancer. No differences were found between cases and controls in the frequency of the N-acetyltransferase polymorphism of NAT2. Studies are underway to determine whether various alleles of CYP1A1, CYP2E1 and CYP2D6 cause increased risk for bladder cancer.
Dr. Mary Ann Butler reported on her studies of CYP1A2 and NAT2 phenotyping using caffeine. In three populations, the expected bimodal distribution of NAT2 phenotype was found to be indicative of the common NAT2 polymorphism. The distribution of CYP1A2 metabolic activity was trimodal. Smoking was found to be associated with increased activity only in Caucasians and not in Chinese or African Americans. In a case-control study, rapid CYP1A2 and NAT2 phenotypes were more prevalent in colorectal/polyp cases than in controls. The rapid-rapid phenotype, age and consumption of well-done meat yielded odds ratios of 3.35, 5.96 and 2.18, respectively. These studies suggest that phenotyping using caffeine may be of use in determining susceptible individuals.
Dr. Ryuichi Kato discussed his studies on acetyltransferase activities in hamster liver. This species is an ideal model, having highest activity for arylamine carcinogens in experimental animals and a clear genetic polymorphism. Two forms of enzyme designated AT-I and AT-II were purified and their cDNAs cloned. AT-I is capable of N-acetylation of arylamines and O-acetylation of amino fluorene and food mutagens such as Glu-P-1. A point mutation in AT-II is responsible for an amino acid change to a premature termination codon, resulting in the polymorphism. Dr. Kato also presented data on a new bacterial tester strain for promutagens and procarcinogens that is capable of detecting a large variety of carcinogenic arylamines and amides.
Dr. Tetsuya Kamataki presented data on a new eukaryotic cell system for detecting promutagens and procarcinogens, especially the heterocyclic amines. The key components of this cell are the human CYP1A2 and NAT2. Only the polymorphic NAT2 and not the NAT1 was capable of activating these compounds. Dr. Kamataki also presented his studies on the phenotypic analysis of CYP1A2 and NAT2 in Japanese subjects. He reported that the frequency of rapid metabolizers of these enzymes was 86 percent and 91 percent, respectively. No differences were found in any nucleotide sequence, either the exons or upstream regulator regions, that could account for the differences in CYP1A2 phenotype. The genotypes of NAT2 were in agreement with previously published studies.
Dr. Curtis Omiecinski reported studies on the presence of xenobiotic metabolizing enzymes in cervical and oral epithelial cells immortalized by human papillomavirus Type l6E6/E7. The hypothesis that viral infection and exposure to environmental chemicals (e.g., smoking) enhances carcinogenesis was examined. A number of P450s and epoxide hydratase (mEH) were analyzed in these cells using RT-PCR. Among the findings were the presence of mEH, CYP1A1 and CYP1A2 transcripts in cell cultures, with the latter two being inducible by protypical agents. CYP2El was present in a number of cell lines except one, while CYP2D6 transcripts were not detected. This P450 was found in lung and other tissues. These studies are suggestive of target cell activation of procarcinogens by P450s and mEH.
Dr. Yoshiaki Fujii-Kuriyama discussed studies on CYP1A1 gene regulation and the role of the Ah receptor and AhR nuclear translocator or Arnt. He showed data demonstrating that both proteins bind to the xenobiotic regulatory element XRE. Co-transfection experiments demonstrated the function of these proteins. Dr. Fujii-Kuriyama also presented the sequence of the Ah receptor (ligand binding subunit) from responsive B6 and non-responsive D2 mice. A number of amino acid differences were detected, only two of which might be responsible for the low affinity of the D2 receptor. The human AhR was also cloned and sequenced from HepG2 cells and liver. Two apparently allelic variants were found and on the basis of comparisons with the mouse receptors, they may have different ligand affinities.
Dr. Yoshiaki Funae presented data on the activation of 3-methoxy-4-aminoazobenzene (3-MeO-AAB) by mouse kidney microsomes. Antibodies to a large number of rodent P450s failed to inhibit this reaction but lauric acid and!!!-naphthoflavone were inhibitory. Three P450s were purified from mouse kidney; P450 designated Mk-3 had highest activity toward 3-MeO-AAB. Mk-1 and Mk-2 had lower activity. This P450 may be the mouse ortholog of CYP4B1 that has been characterized in rabbits and humans.
Dr. Kenneth Korzekwa presented data on the mechanism of P450 catalysis. Studies on CYP19 (aromatase) indicate that the rate limiting steps in oxidation is reduction through electron flux. Dr. Korzekwa also presented data on the mechanism of activation of CYP3A4 by 7,8-benzoflavone. This inhibitor was also found to be a substrate for the enzyme and it caused an increase in Vmax for phenanthrene without affecting Km. Phenanthrene, in turn, caused a decrease in Vmax for 7,8-benzoflavone metabolism without altering Km. The presence of inhibitors or activators alters the regio-selective hydroxylations of CYP3A4 substrates.
Dr. Yasushi Yamazoe reported on a new rat sulfotransferase cDNA, designated ST1C1. Expression of ST1C1 was restricted to males and the cDNA encoded enzyme was a hydroxylamine sulfotransferase. This enzyme was also able to catalyze the O-sulfation and DNA binding of N-hydroxy-2-acetylaminofluorene. It might be of importance in the metabolic activation of this and related chemicals to the proximate DNA-binding metabolites in vivo.
Dr. Tadashi Watanabe described studies on the bioactivation of various polycyclic arylmethanols such as 7-hydroxymethyl-12-methylbenz[a]anthracene and S-hydroxymethyl chrysene by rat liver cytosolic sulphotransferases. The activated derivatives were found to bind to the exocyclic amino groups of adenine and guanine in DNA. Four forms of sulfotransferases were isolated and cDNA cloning suggested the presence of at least five subunits. Cloning of two related (possibly allelic) forms designated STa were described as having a single amino acid difference. Dr. Watabe also described cDNA encoding a new glutathione S-transferase form belonging to the theta class designated Yrs-Yrs. This form is capable of reacting with reactive sulfate esters of arylmethanols.
Dr. Frank Gonzalez described studies on aflatoxin B1 activation by human P450s. cDNA expression demonstrated that five out of the 12 P450s could convert this procarcinogen to a DNA-binding metabolite. Using a eukaryote cell activation system, the CYP1A2 had the highest activity followed by CYP3A4. These studies were suggested by kinetic measurements of aflatoxin B1 epoxide formation using human liver microsomes and expressed enzyme. These data suggest that CYP1A2 may be the dominant P450 in humans exposed to low levels of this toxin and hepatocarcinogen.

PARTICIPANTS

UNITED STATES

Dr. Douglas A. Bell
Senior Staff Fellow
National Institute of Environmental Health Sciences
Research Triangle Park, NC 27709

Dr. Robert A. Branch
Director, Center for Clinical Pharmacology
University of Pittsburgh Medical Center
Pittsburgh, PA 15261

Dr. Mary Ann Butler
National Institute for Occupational Safety and Health
Division of Biomedical and Behavioral Sciences
Cincinnati, OH 45226

Dr. Frank J. Gonzalez
Chief, Nucleic Acid Section
National Cancer Institute
National Institutes of Health
Bethesda, MD 20892

Dr. Jeffrey R. Idle
Chairman, Department of Pharmacological Sciences
The Medical School
Newcastle upon Tyne NE2 4HH
United Kingdom

Dr. Kenneth R. Korzekwa
Senior Staff Fellow
National Cancer Institute
National Institutes of Health
Bethesda, MD 20892

Dr. Stephanie J. London
Department of Preventive Medicine
University of Southern California School of Medicine
Los Angeles, CA 90033
Dr. Curtis J. Omiecinski
Department of Environmental Health
University of Washington
Seattle, WA 98195

JAPAN

Dr. Yoshiaki Fujii-Kuriyama
Department of Chemistry
Faculty of Sciences
Tohoku University
Sendai 980

Dr. Yoshihiko Funae
Laboratory of Chemistry
Osaka City University and Medical School
Osaka

Dr. Tetsuya Kamataki
Director, Division of Drug Metabolism
Faculty of Sciences
Hokkaidi University
Sapporo 060

Dr. Ryuichi Kato
Chairman, Department of Pharmacology
Keio University School of Medicine
Tokyo 160

Dr. Yasushi Yamazoe
Department of Pharmacology
Keio University School of Medicine
Tokyo 160

Dr. Tadashi Watabe
Laboratory of Drug Metabolism and Toxicology
Department of Hygienic Chemistry
Tokyo College of Pharmacy
Tokyo 192-03

Dr. Minroe Watanabe
Director, Department of Molecular Genetics
Institute of Development, Aging and Cancer
Tohoku University
Sendai 980



(2) Seminar on "Oncogenes and Tumor Suppressor Genes in Cancer"

This seminar was held on March 15-17, 1994, at the Maui Prince Hotel, Maui, Hawaii. The organizers were Dr. Masauki Terasha, National Cancer Research Institute, Tokyo, and Dr. Stuart A. Aaronson, Mount Sinai School of Medicine, New York. There were 14 participants and one observer-eight (8) from Japan and seven (7) from the United States.
Dr. Takashi Sugimura welcomed the attendees with opening remarks. He provided attendees with interesting historical perspectives concerning the US-Japan program and highlighted some of the important contemporary issues in cancer research.
The first report by Dr. George Vande Woude addressed itself to met (HGF/SF'): interactions affecting morphogenesis, tumorigenesis, and metastasis. The met proto-oncogene!!!product, Met, is the tyrosine kinase growth factor receptor for hepatocyte growth factor/scatter factor (HGF/SF) . NIH/3T3 cells express HGF/SF endogenously and become tumorigenic in nude mice via an autocrine mechanism when murine Met is expressed ectopically (Metmu cells). Likewise, NIH/3T3 cells co-expressing human Met and human HGF/SF (HMH cells) are highly tumorigenic in nude mice. Paracrine Met-HGF/SF signal transduction produces mitogenic, motogenic, and morphogenic responses in a cell type dependent fashion, properties that contribute to tumor invasion and metastasis. Metmu and HMH cells are invasive in vitro in Boyden chamber basement membrane matrigel assays. These cells secrete type I and IV collagenase, showing that autocrine Met-HGF/SF signal transduction enhances fibroblast cell motility and induces the protease activity necessary for matrigel invasion. In experimental and spontaneous metastasis assays, Metmu or HMH cells metastasize to the lung. However, lower numbers of subcutaneously injected Metmu and HMH cells produced invasive tumors in the heart, diaphragm, salivary gland, and retroperitoneum. With tumor passage in nude mice, Met expression increases, enhancing the activity of Metmu and HMH cells in in vitro and in vivo metastasis assays. Autocrine mediated Met-HGF/SF signal transduction in NIH/3T3 mesenchymal cells may be an important system for understanding the biological process of metastasis.
Dr. Masabumi Shibuya presented studies on the role of the vascular endothelial growth factor-fems-like tyrosine kinase (VEGF-Flt) 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 basic FGFs, TGF!!!and TNF have been reported, the endothelial cell-specific growth factors and signal transduction system are still poorly understood.
Recently a new endothelial cell-regulatory factor, VEGF/VPF (vascular permeability factor), has been molecularly cloned and characterized. Further, Flt receptor, which we originally isolated from human placenta cDNA library (now designated as Flt-1), has recently been shown to bind specifically with VEGF/VPF at high affinity. Accumulating evidence strongly suggests that the VEGF-Flt receptor system is deeply involved in endothelial cell-specific growth and differentiation as well as in regulation of vascular permeability. Dr. Shibuya described aspects of the unique structure of the Flt receptor. It contains seven immunoglobulin-like domains in the extracellular domain, whereas Fms/Kit/PDGFR bear five Ig-like domains in this region. Further, a tyrosine residue conserved within the kinase insert among the Fms/Kit/PDGFR is not present in the kinase insert of Flt-1 receptor. These characteristics suggest that signal transduction pathways from Flt may be different from those generated by Fms/Kit/PDGFR.
In human Graves' disease (hyperthyroidism), in addition to enlargement of thyroid tissue, hypervascularity is commonly clinically observed. Although a continuous stimulation of thyroid follicular cells by autoimmune anti-TSH receptor antibody is well characterized in this disease, the mechanism of hypervascularity is not clearly understood. Dr. Shibuya presented evidence supporting the involvement of VEGF/Flt in this vascular abnormality.
Paracrine stimulation of tumor angiogenesis appears to occur through the VEGF-Flt system. Using primary and transplantable human glioblastomas as samples, Dr. Shibuya found that most of the tumor cells expressed VEGF mRNA but not flt-1 mRNA. flt-1 mRNA was detectable as being of "host" origin. Only the mouse-typeflt-1 transcript was expressed in glioblastomas maintained in nude mice. These results suggest that at least some human tumors may utilize the VEGF-Flt receptor system for induction of angiogenesis in a paracrine manner.
Dr. Stuart A. Aaronson next described new approaches to the identification of oncogenes and other growth regulatory molecules. Genes that act at rate-limiting steps in mitogenic signaling pathways are important determinants of malignancy. Amplification and/or overexpression of genes encoding growth factors or their receptors contributes to a wide variety of human tumors. Other genes that act early in the intracellular transduction of growth factor signals are commonly implicated as oncogenes as well.
Dr. Aaronson described efforts to dissect the biochemical cascade initiated by growth factor triggering as well as new approaches aimed at identifying other critical genes in these pathways utilizing cDNA expression cloning. He further discussed some recently cloned human growth factors which appear to be major paracrine effectors of normal epithelial cell proliferation. Mechanisms by which such factors may influence tumor progression include neoangiogenesis and invasion as mediated by HGF/scatter factor. KGF is an epithelial cell specific growth factor which is released by stromal fibroblasts of many epithelial tissues.
Recent studies have shown that KGF release by fibroblasts of sex hormone responsive tissues, is hormone dependent. Thus, KGF may contribute to the progression of sex hormone responsive malignancies.
Dr. Shin-ichi Aizawa discussed functional analyses of non-receptor tyrosine kinases with mutant mice. An increasing number of non-receptor protein tyrosine kinases (NR-PTKS) has recently been identified with the aid of PCR technique and others. Their general structure is characterized not only by the tyrosine kinase catalytic domain, but also by the presence of src homology (SH) domains. NR-PTKs can be classified into several families: the src family, the abl family, the fps family, the syk family, the jak family and the tec family. In addition, there exist several NR-PTKs such as csk and fak that appear not to constitute a family. Some NR-PTKs are expressed ubiquitously, while others are more restricted. Src family kinases are considered to locate at the inner surface of the plasma membrane through myristyl residue and to associate with cell surface receptors thereby transducing signals brought by specific ligand binding. On the other hand, some NR-PTKS function as the essential regulators of src family kinases or as the second transducers of signals via src family kinases. For example, csk uniquely phosphorylates the tyrosine at the C-terminal end of src family kinases thereby suppressing their activity. Fak is activated by src family kinases and play a role in focal adhesions, possibly regulating the cytoskeletal architecture. Fak is a unique NR-PTK that has no SH2, SH2' or SH3 domain. Dr. Aizawa also identified a novel non-receptor tyrosine kinase named srm (src-related kinase lacking c-terminal regulatory tyrosine and N-terminal myristylation site) that may constitute a new family of NR-PTKs.
Mutant mice were generated by gene targeting in ES cells of fyn, csk, fak and srm loci. Taking Dr. Aizawa's and others' experience on these genes together, generally, the mutation is lethal to mouse development when the genes do not constitute a family, while the phenotypes are more limited than those expected from expression and the in vitro observations when the genes constitute a family.
Dr. Tadatsugu Taniguchi described cytokine signaling and target genes. Cytokines regulate cell growth in a positive or negative manner by inducing their respective target genes. Dr. Taniguchi has been focusing on two cytokine systems, the IL-2 and IFN systems. IL-2, one of the best studied cytokines, is a potent growth stimulatory factor for lymphocytes, whereas the IFNs are well known "negative growth factors," which inhibit the growth of various cell types including hematopoietic cells.
Recently, a number of cytokine receptors, including IL-2 receptor (IL-2R), have been analyzed. The IL-2R is unique in that it is made up of at least three distinct membrane components: the!!!chain (IL-2R!!!), the!!!chain (IL-2R!!!), and the!!!chain (IL-2R!!!).
Dr. Taniguchi and others have provided evidence for the requirement of the cytoplasmic regions of IL-2R!!!and IL-2!!!, but not IL-2R!!!, in IL-2 induced signaling. The IL-2R!!!chain has been shown to interact, both physically and functionally, with the non-receptor tyrosine kinase src family members (p56lck, p59fyn, p53/56lyn). Likewise, other members of the cytokine receptor family seem to utilize another non-receptor tyrosine kinase member. He and his coworkers have identified several target genes which mediate IL-2-induced cell proliferation of a hematopoietic cell line. Evidence suggests that the IL-2R!!!-src kinase pathway is linked to p21ras and subsequently to c-fos, c-jun activation. On the other hand, another signaling pathway(s) emanating from IL-2R!!!, IL-2R!!!heterodimer, but not linked to the src kinase pathway, leads to activation of genes such as c-myc, bcl-2 and others, followed by the progression of the cell cycle.
Dr. Taniguchi presented recent results on the role of these target genes in cell growth control. During the study of the IFN system regulation, his group identified two transcription factors, IRF-1 (activator) and IRF-2 (repressor). IRF-1 is IFN-inducible, and evidence suggests that IRF-1 is in fact, a target gene critical for the IFN functioning as negative regulator of cell growth. More recently, Dr. Taniguchi provided evidence that IRF- I and IRF-2 manifest anti-oncogenic and oncogenic potentials, respectively. In order to gain further insights on the role of the IRFs, mice deficient in either IRF- 1 or IRF-2 or both were generated. The role of these two IRFS in cell growth regulation and host defense against infections was discussed.
The next speaker was Dr. Carlo Croce, who described the molecular genetics of acute leukemias. The ALL-1 gene, located on chromosome 11q23, is fused to a variety of other genes by reciprocal chromosomal translocations. Dr. Croce's group recently reported the detection by Southern blot of ALL-1 gene rearrangements in adult patients with acute myeloid leukemia lacking cytogenetic evidence of 11q23 translocations. These include patients with normal karyotypes as well as patients with trisomy 11. To characterize the abnormal ALL-1 genes, they cloned the ALL-1 rearrangements in two patients with trisomy 11. Characterization of the clones demonstrated a direct tandem duplication within a specific region of the ALL-1 gene. The partial ALL-1 duplication was also detected by Southern blot in one patient with a normal karyotype. Partial duplication of ALL- 1, in which a portion of a proto-oncogene is fused with itself, represents a novel genetic mechanism for leukemogenesis.
Dr. Misao Ohki presented studies on gene rearrangements in acute myeloid leukemias. A large number of chromosome aberrations in hematopoietic malignancies have been described. Many of these abnormalities, particularly chromosome translocations, are closely associated with specific histologic or immunologic subtypes. Molecular analyses of translocation breakpoints have led to the identification of many genes involved in the control of cellular growth and/or differentiation. It has become evident that specific chromosome rearrangements result in the altered expression or structure of cellular genes located at breakpoints, leading to neoplastic transformation of hematopoietic cells.
The t(8;21)(q22;q22) translocation is one of the most frequent chromosome abnormalities in acute myeloid leukemia (AML) and is morphologically associated with the M2 subtypes according to the FAB classification. The t(16;21)(p11;q22) translocation has been found in chronic myeloid leukemia (CML) in the blastic crisis and myelodysplastic syndrome (MDS) evolved to AML as well as de novo AML. These breakpoints were analyzed with the help of NotI restriction maps of chromosome 21, constructed using a NotI linking library. The use of the linking clones which arrayed along the chromosome provides a means to systematically survey chromosome aberrations such as translocation. This allowed the identification of the involvement of a new gene AML1 , highly homologous to Drosophila segmentation gene, runt, and a polyomavirus enhancer binding protein PEBP2!!!(also called CBF!!!) in t(8;21) and ERG, a member of the ets-supergene family in t(16;21) on chromosome 21. Both of these translocations result in the formation of the fusion genes.
Sequence analysis of cDNA showed that t(8;21) translocation juxtaposes the AML1 gene with a novel gene, named MTG8, on chromosome 8, resulting in the synthesis of an AML1 -MTG8 fusion transcript. The fusion protein predicted by the AML1-MTG8 transcript consists of the N-terminal region of AML1 and most of MTG8, which contains putative zinc finger DNA binding motifs and proline-rich regions constituting a characteristic feature of transcription factors. In the t(16;21) translocation, the breakpoints were clustered within a single intron in the coding region of the ERG gene, which has recently been reported also to be involved in Ewing's sarcoma. Molecular analysis of the fusion message revealed that the TLS/FUS gene on chromosome 16 fused with the ERG gene to produce the TLS/FUS-ERG chimeric transcript by this translocation. The TLS/FUS gene has been identified as a translocated gene in myeloid liposarcoma by a t(12;16)(q13;p11) translocation and encodes an RNA-binding protein highly homologous to the product of the EWS gene involved in Ewing's sarcoma. Thus, the TLS/FUS-ERG fusion gene in t(16;21) leukemia is predicted to produce a very similar protein to the EWS-ERG chimeric protein responsible for a solid tumor, Ewing's sarcoma.
Dr. Arnold Levine's presentation concerned the P53-MDM-2 regulatory pathway. Mutations in the p53 gene are the single most common genetic alteration in human cancers. These mutations most often produce faulty proteins that persist and are selected for in the tumor as well as a reduction to homozygosity, resulting in the loss of the wild-type p53 allele. The missense mutations that produce altered p53 proteins have three types of phenotypes: (1) loss of function as expected for a tumor suppressor gene, (2) a transdominant activity resulting in the inactivation of the wild-type protein function, and (3) gain of a new function by the altered p53 protein. The function lost is the ability to inhibit cell proliferation and act as a transcription factor on specific DNA templates. The transdominant phenotype permits the mutant p53 protein to complex with the wild-type protein, inhibit its function and transform cells in culture. When a wide variety of human or mouse mutant p53 genes are expressed in the BALB 3T3 cell line with no endogenous p53 genes or protein, the great majority of independent cell lines now produce tumors in nude mice. This is then a "gain of function" related to cell growth and division. In addition, some mutant p53 alleles produce an altered protein that stimulates the transcription of a new gene, the multi-drug resistance gene. The wild-type p53 protein does not promote the transcription of this gene, again demonstrating a gain of function mutation.
A database containing 1,400 different p53 mutant gene sequences has been established and analyzed. The nature of the mutagen involved in causing these mutations may be inferred from the type of mutation detected. Transversion mutations presumably caused by benzpyrene in cigarette smoke are common in lung cancers, while only transition mutations are observed in colon cancers. Mutations associated with ultraviolet photo-products are common in skin cancers, and aflatoxin B1-induced mutations are observed in liver tumors from southern China and Africa. There is a strong selection for the persistence of an altered or faulty protein in these cancers. Inherited p53 mutations occur in families with high cancer incidence.
The wild-type, but not the mutant p53 protein, will bind to specific DNA sequences of selected genes and promote the transcription of those genes. A cellular oncogene product, a zinc finger protein called mdm-2, binds to the wild-type p53 protein and blocks its ability to enhance the transcription of a gene with a p53 responsive element. The udm-2 gene is amplified some 5-50 times in about 60 percent of the osteosarcomas and 30 percent of soft tissue sarcomas. In these cells, the mdm-2 protein is overproduced and p53 is usually the wild-type form. In about one-third of the sarcomas with amplified mdm-2 genes, p53 is in the mutant form and these patients have a considerably poorer prognosis than when mdm-2 is not amplified or p53 is wild-type.
Dr. Raymond White spoke on the molecular genetics of the APC gene. A rare inherited syndrome involving multiple adenomatous polyps of the colon (APC), which becomes manifest in early adulthood and confers on carriers a high risk of colon cancer, is caused by mutations in a gene (APC) that was isolated in 1991. Subsequent experiments by Dr. White's group and elsewhere have shown that somatic APC mutations occur as early events in the development of sporadic (non-inherited) colon carcinomas also. Characterization of this gene, therefore, carries a high priority among investigators studying the mechanism of tumorigenesis in common colon cancer. The biochemical function of the protein product of APC is not yet clear; however, direct evidence of its role as a tumor suppressor has been obtained from transfection experiments in which a normal copy of APC was able to restore normal morphology and growth behavior to cells derived from a colon carcinoma.
The gene, which lies on chromosome 5, consists of 15 exons. Its expression in tissues is characterized by alternative splicing, not only within the gene but also involving the initiation region. In experiments designed to yield insight into transcription of APC, researchers in Dr. White's laboratory identified three new coding elements in the genomic region upstream from exon l. At least six distinct mRNA species result from alternative splicing of these 5' elements with exons 1 and 2. Preliminary evidence suggests that splice-forms lacking exon 1 may be expressed in a tissue-specific manner. As exon 1 encodes most of the first heptad repeat in the amino acid sequence of the gene product, a region that earlier experiments had shown to confer the ability of the APC peptide to form homodimers, alternative splicing may have important consequences for functional regulation of APC in certain tissues.
Dr. White's group recently identified the promoter region for APC by cloning the DNA sequence immediately upstream of the most 5' new exon into a CAT expression vector. A construct containing about 1 kilobase of this sequence was able to promote CAT expression in a colon-carcinoma cell line. In collaboration with investigators in France, his group characterized the specific APC mutations present in seven kindreds where inheritance of the mutant allele causes an attenuated phenotype of polyps (AAPC), although risk of colon cancer was still high. Four distinct inactivating mutations were identified at the molecular level (one, a 2-base deletion, was common to four of the families), but all were clustered near the 5' end of the gene. All mutations identified to date in patients with the classical APC phenotype have occurred farther downstream. The fact that mutations in AAPC patients seem to be unique to this 5' region may be relevant to the phenotypic differences between APC and its attenuated variant.
Dr. Mutsuo Sekiguchi addressed his presentation to human genes for enzymes controlling spontaneous and induced mutagenesis. 8-Oxoguanine (8-oxo-7,8-dihydroguanine) is produced in DNA, as well as in nucleotide pools of cells, by active oxygen species normally formed during cellular metabolic processes. 8-Oxoguanine nucleotide can pair with cytosine and adenine nucleotides at almost equal efficiencies, and transversion mutation ensues. The MutT protein of E. coli possesses an enzyme activity to hydrolyze 8-oxo-dGTP to the corresponding nucleoside monophosphate and this activity is responsible for preventing occurrence of A:T to C:G transversion. Human cells contain an enzyme activity to degrade 8-oxo-dGTP in a similar fashion. Dr. Sekiguchi purified this particular human enzyme to physical homogeneity, determined its partial amino acid sequence, then cloned the cDNA for human 8-oxo-dGTPase and examined its nucleotide sequence. The human protein comprises 156 amino acid residues and has some sequence homology with the E. coli MutT protein. When the human cDNA was expressed in mutT-- mutant cells, there was a significant amount of 8-oxo-dGTPase activity. In such cells, the frequency of spontaneous mutation was greatly reduced. Thus, he proposes that the human 8-oxo-dGTPase protects genetic information from the untoward effects of endogenous oxygen radicals.
O6-Methylguanine-DNA methyltransferase plays an important role in cellular defense against mutagens and carcinogens with alkylating activity. Dr. Sekiguchi generated transgenic mice having many copies of the bacterial methyltransferase gene attached to the Chinese hamster metallothionein I gene promoter. He found a statistically significant reduction of tumor formation in transgenic mice that received N-nitroso compounds as compared with non-transgenic mice treated in a similar manner. Certain tumor-derived cell lines, termed Mer-, are defective in the enzyme activity and have an increased sensitivity to alkylating agents. His group cloned the genomic sequence coding for the human O6-methylguanine-DNA methyltransferase and elucidated the structure. The gene consisted of 5 exons and spanned more than 170 kb, while mRNA for the enzyme was 950 nucleotides long. Zero or only low levels of mRNA for the enzyme was detected in Mer- cells, though there was no gross difference in the coding and promoter regions of the gene between Mer+ and Mer- cells.
Dr. Sekiguchi also cloned cDNA for mouse methyltransferase and, using the cDNA as a probe, most of the mouse gene for methyltransferase was isolated. The gene consists of 5 exons and is as large as the human gene. Sequences around the exon/intron junctions for the mouse gene are almost the same as those for the human. These sequences are being used to generate mice without the active methyltransferase.
Dr. Paul Modrich discussed the mismatch repair and genetic stability in E. coli and human cells. E. coli methyl-directed mismatch repair eliminates premutagenic lesions that arise via DNA biosynthetic errors, and components of the repair system also block ectopic recombination between diverged DNA sequences. Dr. Modrich's group has reconstituted an excision repair reaction dependent on ten activities (MutH, MutL, MutS, DNA helicase II, SSB, exonuclease I, RecJ exonuclease, exonuclease VII, DNA polymerase III holoenzyme, and DNA ligase) that accounts for function of the methyl-directed system in replication fidelity. Repair is initiated by the mismatch-provoked, MutHLS- and ATP-dependent incision of the unmodified strand at a hemimethylated d(GATC) sequence. The resulting strand break can occur either 3' or 5' to the mismatch on the unmethylated strand. In the ensuing excision reaction, exonucleolytic hydrolysis initiates at the strand break and removes that portion of the unmodified strand spanning the d(GATC) site and the mismatch. The repair system thus possesses an unusual bi-directional excision capability.
Dr. Modrich's group also identified a possible molecular basis for the role of MutS and MutL in blocking ectopic crossovers: the two proteins modulate RecA-catalyzed heteroduplex formation in response to formation of mismatched base pairs. Although MutS and MutL are without effect on RecA-mediated strand transfer between identical DNA sequences, they abolish heteroduplex formation between sequences that have diverged by 3 percent at the nucleotide level. Inhibition of strand transfer between such "homologous" sequences can be observed in the presence of MutS alone, but the effect is dramatically potentiated by MutL and is due to block of the branch migration stage of the reaction.
Human cell nuclear extracts support strand-specific mismatch correction in a reaction that is similar to bacterial repair with respect to both mismatch specificity and unusual features of mechanism. Like the bacterial system, human mismatch repair also functions in mutation avoidance since several kinds of human mutator cells are deficient in the reaction. The hypermutable MT 1 cell line was isolated by virtue of its ability to survive the cytotoxic effects of MNNG. The phenotypes of this alkylation-tolerant line are similar to those of bacterial mismatch repair mutants; and, in fact, MT 1 cells are defective in mismatch correction. Dr. Modrich also found that a number of RER+ (replication error prone) colorectal tumor cell lines are defective in mismatch repair, consistent with the increased variability of microsatellite repeat sequences that characterize such lines. Preliminary in vitro complementation experiments suggest that distinct defects can lead to the RER + phenotype.
Dr. Masaaki Terada discussed the biological significance of amplification in cancer. For conversion of a normal cell to fully malignant cells with metastatic and invasive capacity, multiple steps of carcinogenesis, including initiation, promotion, and progression, are required. During the course of the multistep carcinogenic process involving the multiple genetic alterations, genomic instability apparently occurs resulting in aneuploidy, which characterizes cancer cells at the late stage. Amplification of proto-oncogenes is one of the characteristics of genetic alterations associated in the late stage of carcinogenesis, apparently caused by genomic instability of the cancer cells.
Dr. Terada and coworkers have isolated and characterized genes frequently amplified in various cancers, including K-sam and HST1. K-sam was originally isolated by the in-gel renaturation method, and is frequently amplified in poorly differentiated stomach cancers. The gene encodes a growth factor receptor belonging to the FGF or HBGF receptor family. Some of the characteristics of the K-sam gene were presented. HST1 was originally isolated as a transforming gene and encodes growth factor belonging to the FGF/HBGF family. The gene is localized on 11ql3 and amplified frequently in esophageal cancer. The amplicon having HST1 on 11ql3 was more than 150 kb. At least four genes on the amplicon have been identified. In addition, by genomic subtraction, Dr. Terada found the presence of several different amplicons in a cancer cell. He has also identified a gene flanking c-erbB-2, which is co-amplified with c-erbB-2 in various cancers. The significance of multiple genetic alterations, including amplification of genes, probably caused by genomic instability, was discussed in relationship to its prognostic importance in cancer.
Dr. David Beach discussed pl6ink4, a specific inhibitor of the cdk4/cyclin kinase. Passage through the cell cycle in mammalian cells requires a group of related kinases known as cyclin-dependent kinases (CDKs) whose activity and substrate specificity depends on their association with a family of positive regulatory subunits known as cyclins. The complexes formed by CDK4 and the D-type cyclins have been implicated in the control of cell proliferation during the G1 phase of the cell cycle. In normal proliferating cells, CDK4 associates with D-type cyclins and with a protein of 16KD molecular weight, p16INK4 (inhibitor of CDK4). In human cells transformed with viral oncoproteins that inactivate the Rb tumor suppressor protein, p16INK4 is overexpressed being the main, if not exclusive, partner of CDK4. Dr. Beach and coworkers isolated a human cDNA clone encoding p16INK4 by using the two-hybrid screening system in yeast with CDK4 as the target protein. They found that p16INK4 specifically associates with CDK4 in vitro and in vivo, and does not associate with other CDKs.
Extracts from insect cells overexpressing CDK4 and D-type cyclins were used to reconstitute active CDK4 kinase. When these extracts are incubated with p16INK4, the kinase activity of CDK4 toward Rb is completely inhibited. Inhibition only occurs when using CDK4, and not CDK2, as a catalytic subunit. In normal cells the balance between p16INK4 and D-type cyclins could determine the level of CDK4 kinase activity. Inactivation of Rb during the mid-G1 phase is thought to be fulfilled by CDK4 associated with D-type cyclins. In transformed cells expressing viral oncoproteins, such as T-antigen, E1A or E7, Rb is constitutively inactivated and the G1/S control is at least partially disrupted. Dr. Beach proposes that the disruption of the G1/S control is sensed by a signaling pathway that upregulates the expression of p16INK4, therefore, inhibiting the CDK4 activity. In the particular case of cells expressing viral oncoproteins, this negative loop is unable to restore the Rb-dependent G1/S control because the Rb function in these cells is constitutively abrogated.
Dr. Hiroto Okayama described the identification of new regulatory genes involved in cell cycle controls in mammals and fission yeast. In virtually all eukaryotes, cdc2 or related kinase-cyclin complexes play a central role in cell cycle control. In the G2 phase of the cell cycle, the onset of mitosis is triggered by the activation of the p34cdc2-cyclin B kinase complex, which is regulated by phosphorylation and dephosphorylation at Tyr15 of p34cdc2 catalyzed by the weel kinase(s) and cdc25 phosphatase(s). This control mechanism is conserved in all eukaryotes so far examined from yeast to mammals. However, another regulatory mechanism(s) has been suggested to operate in some organisms, such as in S. cerevisiae.
Dr. Okayama's group recently isolated a human gene named MIN1 that inhibits mitosis via a mechanism independent of the tyrosine phosphorylation of p34cdc2. This gene encodes p130MIN1, which contains a functional domain in the C-terminal region and a highly inhibitory sequence in the N-terminal region. p130MIN1 itself is, therefore, inactive but is converted to a p50MIN1 nuclear protein by proteolytic removal of the inhibitory sequence. The MIN1 mRNA as well as p50MIN1 is predominantly expressed in G2. A weak homology is detectable in the functional domain with the product of ruml+, a new factor that defines the G1 and G2 phases of the cell cycle. Strikingly, expression of the functional domain effectively rescues the lethality of the weel &Mac198; mikl &Mac198; double disruptant. Thus, mammals appear to contain a novel G2 control system parallel to the weel and cdc25-catalyzed p34cdc2 phosphorylation and dephosphorylation. Drastically reduced expression and abnormal processing of the Min 1 protein and the presence of a nonsense mutation in one MIN 1 allele are found in tumor cells. A possible involvement of this gene in chromosomal aberration associated with malignancy was discussed.
In addition to protein kinases, transcriptional regulatory genes have been known to play an important role in cell cycle control, particularly in G1. The SWI/cdc10+ family is among those genes. cdc10+ is a "start" gene of the fission yeast and required for the onset of premitotic and premeiotic DNA synthesis. Dr. Okayama recently identified two such "start" genes named res1+ and res2+. Both are structurally and functionally related to not only each other but also the SWI/cdc10+ family members. Evidence indicates that the Res1 and Res2 proteins form a complex with the Cdcl0 protein and constitute two parallel "start" systems, the former playing a major role in mitotic cycle and the latter in meiotic cycle. In addition, a gene encoding a one zinc finger protein has been isolated that interacts with res2+ - cdc10+ and plays an essential role in the initiation of premeiotic S. Their functions and mutual interactions were discussed as well.
Dr. Aaronson thanked the participants for their contributions to a meeting characterized by impressive progress and new concepts. Scientists from the U.S. and Japan expressed appreciation for the unique opportunity of having this high-quality scientific meeting with detailed private discussions. The material presented provided new insights into the basic genetic mechanisms of growth regulation, as well as new approaches with immediate relevance to the basic cancer researcher as well as potential application at the clinical level.
Thursday, March 17, 1994, a business meeting was held by Dr. Masaaki Terada and Dr. Stuart A. Aaronson to plan continuing and future scientific collaborations.

PARTICIPANTS

UNITED STATES

Dr. Stuart A. Aaronson
Director, Derald H. Ruttenberg Cancer Center
The Mount Sinai School of Medicine
New York, NY 10029

Dr. David Beach
Senior Staff Scientist
Cold Spring Harbor Laboratory
Cold Spring Harbor, NY 11724

Dr. Carlo M. Croce
Director, Thomas Jefferson Cancer Institute and Cancer Center
Jefferson Medical College of Thomas Jefferson University
Philadelphia, PA 19107

Dr. Arnold J. Levine
Chairman, Department of Microbiology
Princeton University
Princeton, NJ 08544-1014

Dr. Paul Modrich
Department of Biochemistry
Duke University Medical Center
Durham, NC 27710

Dr. Raymond L. White
Department of Human Genetics
University of Utah and Investigator,
Howard Hughes Medical Institute
Salt Lake City, UT 84112

Dr. George Vande Woude
ABL-Basic Research Program
NCI-Frederick Cancer Research and Development Center
Frederick, MD 21702-1201

JAPAN

Dr. Shin-ichi Aizawa
Institute for Medical Genetics
Kumamoto University Medical School
4-24-1 Kuhinji, Kumamoto-shi 862

Dr. Hiroto Okayama
Department of Biochemistry
The University of Tokyo, Faculty of Medicine
7-3-1, Hongo, Bunkyo-ku, Tokyo 113

Dr. Misao Ohki
Chief, Radiobiology Division
National Cancer Center Research Institute
5-1-1, Tsukiji, Chuo-ku, Tokyo 104

Dr. Mutsuo Sekiguchi
Director, Medical Institute of Bioregulation
Kyushu University
3- 1-1, Maidashi, Higashi-ku, Fukuoka-shi 812

Dr. Masabumi Shibuya
Department of Genetics
Institute of Medical Science
University of Tokyo
4-6- I , Shirokanedai, Minato-ku, Tokyo 108

Dr. Tadatsugu Taniguchi
Institute for Molecular and Cellular Biology
Osaka University
1-3, Yamadaoka, Suita-shi, Osaka 565

Dr. Masaaki Terada
Director, National Cancer Center
Research Institute
5-1-1, Tsukiji, Chuo-ku, Tokyo 104

Dr. Takashi Sugimura
President Emeritus
(observer)
National Cancer Center
5-1-1, Tsukiji, Chuo-ku, Tokyo 104



(3) Seminar on "Non-Neoplastic Changes Induced by Carcinogens"

The seminar was held on March 18-19, 1994, at the Ilikai-Hotel Nikko Waikiki Honolulu, Hawaii. The organizers were Dr. Richard Adamson, National Cancer Institute, Bethesda, Maryland, and Dr. Nobuyuki Ito, Nagoya City University School of Medicine Nagoya, Japan. There were six (6) participants from the United States and nine (9) participants from Japan. The purpose of the meeting was to present current research focused on the noncarcinogenic effects of carcinogens either chemical or viral, at a variety of organ sites.
Participants were initially welcomed by Dr. Richard Adamson, followed by Dr. Takashi Sugimura, who made introductory remarks to the group. In his remarks, he stated that there are epidemiological studies which indicate that the incidence of amyotrophic lateral sclerosis found on Guam may be linked to the consumption of cycad nuts from which the carcinogen cycasin has been isolated. Other examples include the braken fern known to cause hematuria and occasionally bladder cancers in cattle and retinal degeneration in sheep. The active ingredient in braken fern has been identified as aquiloside.
Dr. Cindy Davis discussed the cardiotoxic effects of heterocyclic amines (HCAs) in cultured cardiac myocytes from rats. Monkeys treated with 2-amino-3-methylimidazo[4,5-f]quinoline (IQ) showed both myocyte degeneration as well as mitochondrial changes. Therefore, studies were undertaken in order to develop models to further explore heterocyclic amine cardiotoxicity. Primary cultures of fetal rat cardiac myocytes were exposed to the activated forms of the carcinogens IQ and 2-amino-1-methyl-6-phenylimidazo[4,5,b]pyridine (PhIP). LDH leakage increased in proportion to the carcinogen dose (on average 81 percent with IQ and 30 percent with PhIP). However, pretreatment of the cells with N-acetyl-cysteine 16 hours prior to exposure to either carcinogen significantly reduced the release of both LDH and creatine kinases. Upon examination by electron microscopy, treated cells were found to have swollen and irregular mitochondria and fewer organelles. While ultrastructural damage was found to be more severe in the IQ-treated cells, DNA adducts assayed using 32p-postlabeling techniques were found to be considerably higher in the PhIP-treated cells.
The toxic effects of heterocyclic amines were also evaluated in rats given either IQ or PhIP (100 mg/kg, p.o. 10x). Seven out of eight rats exposed to either carcinogen exhibited light microscopic and ultrastructural cardiac abnormalities. Treated animals had foci of chronic inflammation with myocyte necrosis, myofibrillar dissolution, and disarray and dilation of T-tubules that were not observed in untreated controls. While there were no differences in the level of DNA adducts found in heart mitochondrial versus nuclear DNA between the two compounds, rats fed PhIP showed significantly higher total heart DNA adduct levels than animals fed IQ. These results suggest that in addition to carcinogenicity, these mutagenic heterocyclic amines may also play a role in cardiac degeneration.
Dr. Nobuyuki Ito described ultrastructural and histopathological changes of the myocardium in rats treated with PhIP. PhIP has previously been shown to be carcinogenic for the colon of male F344 rats and the mammary gland of female F344 rats. A dose relationship was established for these tumorigenic effects. In addition to the DNA adduct formation seen in the target organs of the PhIP-treated rats, adduct formation was also observed in the heart, pancreas, and lung. Other heterocyclic amines, such as IQ, were also found to cause adduct formation in the heart, a nontarget organ for their carcinogenicity.
In a long-term study, F344 rats were given PhIP at a dose of 400 ppm in the diet for up to 52 weeks. Histopathological heart lesions, such as vacuolar degeneration, mononuclear cell infiltration, and focal necrosis were observed in the PhIP-treated rats but not in controls. Electron microscopic study revealed mitochondrial and myofibrillar abnormalities. A single application of PhIP of 40 mg/kg, i.g., also induced myocardial changes that were evident at the ultrastructural level.
Other carcinogens, such as IQ, dimethyl aminoazobenzene (DMAB), and diethyl nitrosamine (DEN) were also examined for the induction of heart muscle lesions. IQ and DEN were found to induce similar fine structural changes to PhIP, but not DMAB , which was not associated with any adduct formation in heart tissues.
Electrocardiography of rats was also performed after treatment of F344 male rats with 100 and 200 mg/kg of PhIP, i.p., twice a week for four weeks. After each PhIP administration, heart rate, durations (msec.) of Q, PR, QRS and QT, and voltages (mV) of P, Q, R, S, and T wave were examined, but no remarkable changes were detected in any of these parameters. The results of these studies clearly demonstrate that HCAS such as PhIP, IQ and other carcinogens, e.g. DEN, caused clear morphological changes in heart muscle.
Dr. Unnur Thorgeirsson discussed findings of myocardial lesions induced by IQ, a known hepatocarcinogen in nonhuman primates. A histopathological cardiac study was undertaken in monkeys chronically dosed with IQ, following the discovery of high cardiac IQ-DNA adduct levels. In 80 percent of the hearts examined, abnormalities were detected by light microscopy which included focal myocyte necrosis with or without chronic inflammation; interstitial fibrosis with myocyte hypertrophy or atrophy; and vasculitis. Electron microscopy revealed abnormalities in 90 percent of the hearts, including disruption of the mitochondrial architecture; myofibrillar loss; disorder of sarcomere alignment; and focal myocyte nuclear hypertrophy. While there was no correlation found with the extent of myocardial damage and the length of exposure, some correlation was found with the cumulative dose of IQ. Although the IQ-induced myocardial lesions were focal and not associated with clinical evidence of heart failure in the monkeys, these results further extend previous findings which suggest that chronic exposure to heterocyclic amines in the diet may play a role in the etiology of heart disease in humans.
Dr. Takayuki Ozawa described the detrimental effects of PhIP on mitochondrial respiration among various rat tissues as evaluated by the activity measurement of the electron transfer chain. Wistar rats received 100 mg per kg of PhIP for 2, 4, 8, and 12 weeks. While no effects were observed at two weeks, by four weeks of administration, decreased activities of complex I in the mitochondrial electron transport chain of the heart, diaphragm, and psoas major were seen in a dose-dependent manner. The mode of inhibition suggested that there was a detrimental effect of PhIP on mitochondrial DNA (mtDNA) as opposed to a direct effect of PhIP on complex I.
The presence of deletions and the amount of 8-hydroxy-deoxyguanosine (8-OH-dG), a hydroxyl-radical adduct of deoxyguanosine, in rat heart mtDNA following PhIP administration was quantified. After four weeks of PhIP administration, there was significant accumulation of 8-OH-dG in myocardial mtDNA, and mtDNA fragments were detected with 4.1 and 4.8 kbp deletions using PCR techniques. 8-OH-dG was reported to induce random point mutations as well as local structure alternation such as B to Z transition. These effects could, in turn, trigger a double strand separation, thereby resulting in a large deletion. It was concluded that the accumulation of mtDNA mutations over the course of a lifetime may play an important role in both the aging process and degenerative diseases. The mtDNA mutations induced by PhIP could be another example of somatically acquired mutations triggered by a mutagenic agent that result in myocardial dysfunction.
Dr. Yasushi Yamazoe presented his work on the effect of several HCAS on the benzodiazepine and GABA receptors in rodents. Among amino-!!!-carbolines, Trp-P-1 antagonized the suppressive effect of diazepam on the pentylenetetrazole-induced convulsions and death seen in male mice, whereas Trp-P-2 by itself precipitates seizures and death in these animals. Both Trp-P-1 and Trp-P-2 inhibited the specific binding of [3H]diazepam and [3H]muscimol in rat brain membranes mainly by increasing Kd, indicating that these amino-!!!-carbolines bind to benzodiazepam and GABA receptors. Similar effects on [3H]diazepam binding was also observed with norharman and harmine. However, Lys-P-1 was ineffective and A!!!C and MeA!!!C had only weak potency to displace benzodiazepine binding. Inhibitory effects of Trp-P-1 and Trp-P-2 on specific binding of [3H]flunitrazepam to brain membranes was not affected by the addition of GABA. These results suggest that Trp-P-1 and Trp-P-2 may act as active antagonists at benzodiazepine receptors. The convulsant effect of!!!-carbolines such as Trp-P-1 and Trp-P-2 may be mediated by an action on central benzodiazepine receptors.
Dr. Katsumi Imaida described atrophic changes in the pancreas of rats that were attributable to PhIP or MeA!!!C. The pancreas is one of the organs in which adduct formation can be detected in rats treated with PhIP, despite the fact that PhIP is not carcinogenic for the rat pancreas. Therefore, lesions other than neoplastic changes in the pancreas of rats treated with PhIP and other HCAs, such as MeA!!!C, provide valuable morphological correlates of DNA adducts.
Both male and female F344 rats were treated with doses of PhIP ranging from 50 to 400 ppm in their diet for up to 52 weeks. The resulting pancreatic lesions, especially those found in exocrine acinar cells, were examined histopathologically and were found to exhibit loss of zymogen granules, acinar atrophy, fatty change, and vacuolization. The degranulation and acinar cell atrophy were found to be dose-related. No hepatocyte-like cells were observed and there were no remarkable changes evident in islet cells.
Both male and female F344 rats were exposed to MeA!!!C at dietary levels of 800 ppm for 26 weeks, 400 ppm for 52 weeks, and 200, 100 or 0 ppm for 100 weeks. There was significantly higher induction of pancreatic acinar cell adenomas in the 100-200 ppm groups; MeeA!!!C-treated groups exhibited atrophic changes in acinar cells similar to those lesions observed in PhIP-treated animals. These observations demonstrate that heterocyclic amines such as PhIP and MeeA!!!C can induce atrophic changes in pancreatic acinar cells where adduct formation occurs and that all organs with high adduct levels may exhibit toxic lesions separate from neoplastic changes.
Dr. Minako Nagao discussed atrophic changes in rat salivary glands caused by MeeA!!!C. MeeA!!!C which is present in cooked foods and in cigarette smoke, is known to induce haemangioendothelial sarcomas and hepatocellular carcinomas in CDF1 mice. A preliminary study on the biological effects of MeeA!!!C in rats also demonstrated atrophy of both salivary and pancreatic glands leading to animal deaths. The dose-dependent effects of MeeA!!!C were examined after continuous administration of a 0.01, 0.02, 0.04 or 0.08 percent MeeA!!!C diet to male F344 rats. The 0.01 and 0.02 percent groups were sacrificed at week 100, the 0.04 percent group at week 52 and the 0.08 percent group at week 26. Hepatocellular carcinomas and pancreatic adenomas developed in rats dosed with 0.01 percent and 0.02 percent MeeA!!!C at incidences of 25 percent and 30 percent. Fibromas in the subcutis also developed at a high incidence (70 percent) in the 0.02 percent group. DNA adduct levels after administration of the 0.04 percent MeeA!!!C diet to male F344 rats for four weeks were additionally determined. Adduct formation was pronounced in the liver and was also detected in the salivary glands. Dr. Nagao also stated there was a fourfold increase in cholesterol levels following MeeA!!!C treatment.
The effects of HCAS on poly(ADP-ribose) synthetase were presented by Dr. Kunihiro Ueda. Novel effects, inhibitory and stimulatory, of HCAs On poly(ADP-ribose) synthetase were found using an in vitro assay system. These effects might be related to the activity of the HCAs to induce teratocarcinoma cell differentiation in culture. An extensive survey of inhibitors of poly(ADP-ribose) synthetase found that many aromatic compounds have some inhibitory effects on the synthetase's activity. Among them are the HCAS known to be produced by pyrolysis of amino acids or protein. The strongest effect was observed with Trp-P-1, followed by Trp-P-2 and MeeA!!!C. The IC50 values of Trp-P-1 and Trp-P-2 were 0.22 mM and 2.2 mM, a finding that suggests that Trp-P-1 is as potent as nicotinamide, a well-known inhibitor of the synthetase. Also noteworthy was the fact that Trp-P-2 exhibited not only an inhibitory effect at concentrations above 2 mM but had a remarkably stimulatory effect on the synthetase at around 1 mM.
Future research efforts will focus on whether or not these effects of HCAs on ADPribosyltransferases have any biological significance. Preliminary observations have shown that some of the HCAs, including Trp-P-1(at 0.01mM) and PhIP (at 0.5-1mM), induced multipotent teratocarcinoma cell differentiation in culture which resembles the action known for other inhibitors of poly(ADP-ribose) synthetase.
Dr. Elizabeth Snyderwine described studies on metabolic processing and the noncarcinogenic effects of urethane. Urethane, a well-known animal carcinogen, has been the subject of intensive research for nearly 50 years. Once used as a hypnotic in humans during the 1940s at doses as high as 1 gram per day, urethane has also been used as a chemotherapeutic agent for leukemia patients. In addition to its past medicinal uses, urethane is also found in a wide variety of fermented foods and alcoholic beverages that are consumed in the human diet Bread contains 3-15 ng/g, stone-fruit brandies from 200 to 20,000 ng/g, and about one-third of table wine samples contain more than 10 ng/g.
In alcoholic beverages, urethane is principally formed by reaction of ethanol with urea. Urethane is metabolized by cytochrome P-450IIE1 to vinyl carbamate and subsequently to vinyl carbamate epoxide, the ultimate electrophilic metabolite of urethane. To date 12 cases of urethane-induced hepatitis in humans have been reported. It is theorized that the' hepatic damage is due, in part, to the metabolism of urethane to the electrophilic metabolite and that glutathione may play a role in protecting against this damage.
Following this presentation, Dr. Sugimura noted that some 40 years ago, urethane was used as a vehicle in Japan to dissolve drugs. For example, it was used at concentrations of nearly 20 percent to dissolve salicylate which was administered intravenously to treat fever in patients. Dr. James A. Miller asked him if there were any epidemiological studies on the use of urethane but no data were available at that time.
Dr. Ryohei Hasegawa described research on renal toxicity in male F344 rats treated with MeA!!!C that was found in addition to the liver and pancreatic carcinogenicity by this HCA. The early death of animals in the higher dose groups was attributed to renal failure as well as to liver damage. Serum chemistry data showed increased blood urea nitrogen, decreased protein level and decreased albumin/globulin ratio when compared to controls at week 26, with similar results at week 52. These data paralleled the observed histopathological alterations in the kidney, so-called chronic nephropathy. Immunohistological and histopathological examination revealed abnormal deposition in terms of the shape and amounts of droplets of!!!2u-globulin in the proximal tubules of the treated animals, but the degree of nephropathy was not always correlated with the degree of the deposition.
In addition to!!!2u-globulin-associated renal toxicity, direct action of MeA!!!C on the kidney might have been involved in the early stages of chronic nephropathy. However, no renal cell tumors were found in the 0.02 and 0.01 percent dose groups after 100 weeks, although transitional cell hyperplasia and tumors were induced in the renal pelvis and urinary bladder. Urinary crystal formation was increased in these animals, but obvious stone formation or papillar necrosis were not observed.
Dr. Donald Mattison spoke on the reproductive effects of chemical carcinogens. It was noted that one prominent effect of these chemicals is ovarian toxicity, affecting oocytes, granulosa, theca and/or luteal cells. Using experimental animals, studies have shown that chemical carcinogens such as the polycyclic aromatic hydrocarbons (benzo[a]pyrene, 3-methylcholanthrene and 7,12-dimethylbenzanthracene) and therapeutic agents like cyclophosphamide produce ovarian toxicity in a species, strain, chemical, time, and dose-dependent manner. The human ovary also appears to be sensitive to some of these chemicals. Examples given were smoking and cyclophosphamide. Cellular mechanisms of ovarian toxicity produced by chemical carcinogens has not been fully explored and additional studies are needed in this important area of research.
Dr. Mitsuhiro Osame discussed the association of tropical spastic paraparesis (TSP) and human T-lymphotropic virus type I (HTLV-I). This relationship was first demonstrated in 1985 when a serologic study in Martinique found that 59 percent of patients with TSP had antibodies to HTLV-I. A similar association between HTLV-I and spastic paraparesis was also found in Japan; however, since these patients resided in a temperate rather than tropical climate, the term HTLV-I associated myelopathy (HAM) was proposed for this disease. While initially it was felt that TSP and HAM were distinct from one another, because similar associations between HTLV-I and myelopathy have been found in other areas of the world, the prevailing view is that HAM and TSP are the same disorder.
Two subgroups of HAM cases have been identified which are related either to a previous history of blood transfusion or mother-to-child transmission. The identity of the viruses of HAM and ATL was shown through DNA blotting and DNA sequence analysis from cell lines established from the cerebrospinal fluid of HAM patients. HLA-haplotype-1inked high immune responsiveness against HTLV-I has also been reported.
Dr. Jack Gruber discussed Epstein-Barr virus (EBV), one of the most common human viruses with a worldwide distribution. Over the past 30 years it has come to be associated with several malignant neoplasms such as Burkitt's lymphoma, nasopharyngeal carcinoma, lymphomas in immunodeficient subjects, and Hodgkin's disease, and with several infectious diseases including infectious mononucleosis, fatal mononucleosis, severe chronic mononucleosis, and chronic fatigue syndrome.
EBV is a member of the herpesvirus family. Most people acquire the virus at some point during their lifetime, with infants becoming susceptible to EBV infection as soon as the maternal protection usually present at birth disappears. Most children become infected with EBV and exhibit either no symptoms or experience a very mild illness. However, in the United States and in other developed countries, many persons who were not infected with EBV during their childhood may develop a disease syndrome designated as infectious mononucleosis when infected as adolescents or young adults. Although the symptoms of infectious mononucleosis usually resolve themselves within a few months, EBV remains dormant in cells in the throat and blood throughout the person's lifetime.
EBV may also establish a lifelong latent infection in some cells of the body's immune system—one that may reemerge in a small number of these individuals in conjunction with a variety of neoplastic disease conditions such as Burkitt's lymphoma, nasopharyngeal carcinoma, and B-cell lymphomas. Although EBV appears to play an important role in these diseases, the virus is probably not the sole cause of these malignancies. Ongoing studies seek to explore the role of EBV in disease development as well as develop a safe and efficacious vaccine against EBV and its associated illnesses.
In closing comments, Dr. Adamson briefly summarized the talks, then observed (with reference to the popular Dr. Seuss book Green Eggs and Ham) that the seminar concerned heterocyclic amines and that many additional noncarcinogenic effects of chemical and viral carcinogens remain to be discovered, that discovery favored the careful and thorough observer, and that from strange titles came excellent talks and a unifying seminar.

PARTICIPANTS

UNITED STATES

Dr. Richard H. Adamson
Director, Division of Cancer Etiology
National Cancer Institute
Bldg. 31, Rm. 11A03
Bethesda, MD 20892

Dr. Cindy Davis
IRTA Fellow, Laboratory of Experimental Carcinogenesis
National Cancer Institute
Bldg. 3 1, Rm. 3C28
Bethesda, MD 20892

Dr. Jack Gruber
Chief, Biological Carcinogenesis Branch
National Cancer Institute
Executive Plaza North, Rm. 540
Bethesda, MD 20892

Dr. Donald R. Mattison
Dean, Graduate School of Public Health
University of Pittsburgh
Room 111, Parren Hall
130 DeSoto Street
Pittsburgh, Pennsylvania 15261

Dr. Elizabeth Snyderwine
Senior Staff Fellow, Laboratory of Experimental Carcinogenesis
National Cancer Institute
Bldg. 31, Rm 3C28
Bethesda, MD 20892

Dr. Unnur Thorgeirsson
Division of Cancer Etiology
National Cancer Institute
Bldg. 31, Rm. 2D14
Bethesda, MD 20892

JAPAN

Dr. Takashi Sugimura
President Emeritus, National Cancer Center
1-1, Tsukiji 5-chome
Chuo-ku, Tokyo 104

Dr. Ryohei Hasegawa
Lecturer, First Department of Pathology
Nagoya City University Medical School
1 Kawasumi, Mizuho-cho
Mizuho-ku, Nagoya 467

Dr. Katsumi Imaida
Lecturer, First Department of Pathology
Nagoya City University Medical School
1 Kawasumi, Mizuho-cho
Mizuho-ku, Nagoya 467

Dr. Nobuyuki Ito
Professor and Chairman, First Department of Pathology
Nagoya City University School of Medicine
1 Kawasumi, Mizuho-cho
Mizuho-ku, Nagoya 467

Dr. Minako Nagao
Chairman, Carcinogenesis Division
National Cancer Center Research Institute
, Tsukiji 5-chome
Chuo-ku, Tokyo I 04

Dr. Mitsuhiro Osame
Professor and Chairman, Third Department of Internal Medicine
Kagoshima University Faculty of Medicine
8-35-1 Sakuraoka, Kagoshima 890

Dr. Takayuki Ozawa
Professor and Chairman, Department of Biochemistry
Nagoya University School of Medicine
65 Tsurumai
Showa-ku, Nagoya 466

Dr. Kunihiro Ueda
Associate Professor, Kyoto University Faculty of Medicine
Konoe-cho, Yosida
Sacho-ku, Kyoto 606-01

Dr. Yasushi Yamazoe
Associate Professor, Keio University School of Medicine
35 Shinano
Shinjyuku-ku, Tokyo 160



(4) Seminar on "Natural Substances Including Phytochemicals for Primary Prevention of Cancer"

This workshop was held on March 21-22, 1994, at the Ilikai Hotel Nikko Waikiki, Honolulu, Hawaii. The organizers were Dr. Takashi Sugimura of the National Cancer Center, Tokyo, and Dr. Lee Wattenberg of the University of Minnesota Medical School, Minneapolis, Minnesota. There were seven (7) participants from the United States and seven (7) participants from Japan.
Dr. Sugimura provided an overview of cancer prevention research, detailing the progress to date and the problems that remain in this challenging field. He suggested that clarification of the definition of cancer prevention is needed, noting that the terms "primary and secondary cancer prevention," defined as the prevention of cancer development or cancer death, are in widespread use. The category of clinical cancer prevention which seeks to prevent successive primary cancers in patients whose first cancer was successfully treated is also becoming increasingly important. Because cancer patients are at increased risk for developing subsequent cancers, the distinction between the prevention and treatment of cancer for these individuals and for those who are genetically at increased risk for cancer development has become blurred. From an economic standpoint, primary cancer prevention in the general population is clearly more advantageous than either secondary or clinical cancer prevention. Dr. Sugimura concluded by stressing the need to establish new strategies and methods for cancer prevention, characterizing their development as an important challenge for the 21st Century and one of the most urgent problems facing cancer researchers today.
Dr. Wattenberg focused on the importance of phytochemicals in the diet. He noted that plant foodstuffs contain a very large number of chemicals, both major and minor nutrients. There are increasing data to suggest that the minor constituents of plants may play an important role in chemoprevention. Information pertaining to their roles come from three levels:

Dr. Wattenberg concluded by noting that the seminar should provide up-to-date information concerning phytochemicals and their usefulness as chemopreventive agents in humans.
Dr. Isao Tomita presented information about the antimutagenic and antitumorpromoting activity of green tea constituents, including polyphenols and carotenoids. Water extract of green tea and its components was shown to inhibit mutagenesis in bacteria, depending on not only desmutagenic action but also bio-antimutagenic potencies, possibly through the activation of repair enzymes or by direct interaction with DNA. Oral administration of green tea extracts suppressed the MNNG-induced mutagenesis in cells of the gastrointestinal tract in rats. Furthermore, antitumor-promoting activity of green tea extracts and their fractions was also demonstrated. Dr. Tomita also introduced epidemiological evidence which suggested that there may be an inhibitory suppressing effect of green tea on human stomach cancer.
Dr. Allan Conney described research which showed that topical application of a green tea polyphenol fraction inhibits 1 2-0-tetradecanoylphorbol- 1 3-acetate-induced ornithine decarboxylase activity. He then discussed the inhibitory effects of tea on carcinogenesis in mice: administration of green and black tea in the drinking water was found to inhibit ultraviolet B light-induced carcinogenesis in SKH-1 mice previously initiated with 7,12dimethyl-benz[a]anthracene. In other animal models, administration of green tea in the drinking water of mice was shown to inhibit nitrosodiethylamine-induced forestomach and lung tumorigenesis or 4-(methylnitrosamino)- 1-(3-pyridyl)-1-butanone-induced lung tumorigenesis. The data presented by Drs. Tomita and Conney provided persuasive evidence of the possibility of tea as a potential chemopreventive agent.
Dr. Hiroyoki Tsuda discussed the possible chemopreventive properties of naturally occurring materials, such as Aloe powder and plant antioxidants such as Aloe powder (30%, freeze-dried whole leaf powder), É¿-carotene (1%), a-tocopherol (1%), glutathione (5%), diallyl sulfide (1%), vanillin (1%), and quercetin (1%). When administered in the diet in the stage of tumor initiation, these compounds showed significant inhibitory effects on the induction of preneoplastic liver lesions in rats. Aloe, at a dose of 30 percent, also showed an inhibitory effect when given during the tumor promotion stage.
Dr. Mou-Tuan Huang presented data showing that topical application of commercial curcumin inhibited 12-O-tetradecanoylphorbol-13-acetate (TPA)-induced inflammation, ornithine decarboxylase activity, DNA synthesis, hyperplasia, formation of hydrogen peroxide, and tumor promotion in mouse epidermis. Topical application of commercial curcumin also inhibited the formation of benzo[a]pyrene (B[a]P)-DNA adducts and the tumorigenic activity of B[a]P in mouse skin. Commercial curcumin has a potent effect on epidemial cyclooxygenase and lipoxygenase activities in vitro and inhibited arachididonic acid-induced mouse ear edema in vitro. Structure-activity relationship studies revealed that commercial curcumin, pure curcumin (purity greater than 98 percent) and demethoxycurcumin had the same potent inhibitory effect on TPA-induced tumor promotion on mouse skin, and they were more active than bisdemethoxycurcumin and tetrahydrocurcumin. In additional studies, administration of commercial curcumin or pure curcumin in AIN 76A diet had the same strong inhibitory effect on azoxymethane (AOM)-induced colon tumorigenesis in CF-1 mice when given to mice: (a) during the initiation period at two weeks before, during, and for one week after carcinogen administration and (b) during the post-initiation period, one week after carcinogen administration until the end of the experiment or during both the initiation and post-initiation periods. In addition, administration of 0.5-2.0 percent commercial curcumin in AIN 76A diet inhibited B[a]P-induced forestomach tumorigenesis in A/J mice (both initiation and post-initiation), and N-ethyl-N-nitro-N-nitrosoguanidine-induced duodenal tumorigenesis in C57B1/6J mice (post initiation). Since curcumin has little toxic effect, it may be a potentially useful chemopreventive agent in humans.
The possible usefulness of the plant Rosmarius officinalis L. (rosemary), an herb with powerful antioxidant properties, was also discussed. Topical application of an extract of rosemary leaves to the backs of mice inhibited B[a]P-mediated DNA adduct formation and tumor initiation as well as TPA-induced mouse ear edema, ornithine decarboxylase activity, and tumor promotion in the mouse epidermis. Carnesol and ursolic acid, isolated from rosemary, inhibited TPA-induced ear edema and ornithine decarboxylase activity. Topical application of 1,3, or 10µ mol of carnesol together with 5 nmol of TPA twice weekly for 20 weeks to the backs of CD-1 mice previously initiated with DMBA inhibited TPA-induced skin tumors per mouse by 38, 63, or 88 percent, respectively. Topical application of 0.1 or 2µ mol of ursolic acid together with 5 nmol of TPA twice weekly for 20 weeks inhibited skin tumors per mouse by 52 or 6 1 percent, respectively.
Dr. Hideki Mori's talk focused on a number of phytochemicals which suppressed liver carcinogenesis; compounds included chlorogenic acid, ellagic acid, protocatechuic acid, indole-3-carbinol, sinigrin, benzylisothiocyanate, and benzylthiocyanate. Some of these compounds also demonstrated the capacity to suppress carcinogenesis in other organs as well. The inhibitory effect of caffeic acid and ferulic acid on tongue carcinogenesis, and flavoglaucin, gingerol, shikonin, and costunolide on intestinal carcinogenesis, was discussed. Dr. Mori also demonstrated that magnesium as well as calcium had chemopreventive effects on colon carcinogenesis, although the precise mechanism of action is not yet known.
Dr. Michael Pariza discussed the role of phytochemicals as dietary anticarcinogens and suggested that they may be substrates for the production of anticarcinogens by intestinal microflora. He cited the example of linoleic acid, an essential fatty acid found only in plant products. Linoleic acid by itself enhances carcinogenesis in several animal models and is the only fatty acid clearly shown to do so. However, certain bacteria in the rumen of cattle and in the rat colon can convert linoleic acid to its cis-9, trans-11 conjugated dienoic isomer (conjugated linoleic acid or CLA). Unlike linoleic acid, CLA is an extremely potent anticarcinogen. It was hypothesized that its anticarcinogenic action might depend on the modulation of the macrophage-interleukin 1 system.
Dr. Hoyoku Nishino described the cancer preventive activity of various natural carotenoids.!!!-carotene has been suggested to be a possible cancer preventive agent, and anticarcinogenic activity has been demonstrated for other natural carotenoids as well. Some of these compounds, such as!!!-carotene, lutein and fucoxanthin, were shown to be more potent than!!!-carotene in suppressing carcinogenesis in various organs.
Dr. H. Leon Bradlow noted that a role for estrogens in the initiation and prevention of breast cancer has been evident for over a century. Recent studies have shown that estradiol 16!!!-hydroxylation is elevated in women at high risk for breast cancer and in postmenopausal women with breast cancer. Studies on non-involved terminal duct lobular units (TDLU) from tumor-bearing human breasts show elevated rates of 16!!!-hydroxylation and ras P-21 expression relative to TDLU obtained from healthy tissue. Other studies have demonstrated that murine strains with a very high incidence of mammary tumors show elevated levels of 16!!!-hydroxylation. A variety of mammary epithelial cell culture models show similar effects. When the cells are transformed into more tumorigenic lines, 16!!!-hydroxylation was elevated and the alternative pathway 2-hydroxylation was depressed.
Dr. Bradlow noted that the products formed by the metabolism of estradiol is critical: one pathway leads to 2-hydroxyestrone, a compound with minimal estrogenic activity and some level of antiestrogenicity, while the other pathway leads to 16!!!-hydroxyestrone, a fully active estrogen. This compound is a unique estrogen capable of binding covalently to the estrogen receptor and to amino groups on adenosine and guanosine. It can exert more prolonged estrogenic responses than estradiol as well as causing genotoxic responses exemplified by increased unscheduled DNA synthesis, hyperproliferation, and increased anchorageindependent growth. Methods for decreasing 16!!!-hydroxylation have been explored; direct efforts have been unsuccessful. However, indirect decreases have been achieved by inducing an increase in 2-hydroxylation by feeding a variety of dietary indole derivatives. Thus, the addition of indole-3-carbinol to the diet will extend the period of latency and decrease the incidence of mammary tumors in several strains of mice. In addition, this same approach has altered the activity of the human papillomavirus and decreased the rate of growth of laryngeal papillomas in human patients.
Dr. Shoji Shibata described the antitumorigenic activities of terpenoids and flavonoids. Glycyrrhetinic acid, the sapogenin of the licorice terpenoid saponin, glycyrrhizin, showed an antitumor-promoting action in two-stage mouse skin carcinogenesis. On the basis of these findings, some oleanane-type triterpenes that were chemically derived from natural tritepenoids were tested for antitumorigenic activity. The compound 18!!!-olean-12-ene3!!!,23,28-triol was shown to have the highest antitumor-promoting effect among the triterpenoids tested so far. A licorice constituent, licochalcone A, was also found to show remarkable antitumor-promoting activity. Other studies were performed in which a number of simple chalcone derivatives were synthesized and tested for antitumorigenic activity. Of these, the compound 3'-methyl-3-hydroxychalcone was shown to demonstrate the most potent activity in vivo and to suppress the proliferation of cultured human malignant tumor cells, such as HGC-27, PANC-1, HeLa and GOTO. An antitumorigenic mechanism of chalcone derivatives has been suggested which involves type II estrogen binding site.
Dr. Ann R. Kennedy presented data on the anticarcinogenic activity of protease inhibitors. She showed that the Bowman-Birk inhibitor, derived from soybeans, is a particularly effective anticarcinogenic protease inhibitor. BBI has been studied both as a pure protease inhibitor, purified BBI (PBBI), and as an extract of soybeans enriched in BBI, termed BBI concentrate (BBIC). PBBI and/or BBIC have been shown to suppress carcinogenesis in a wide variety of in vivo and in vitro carcinogenesis assay systems. BBI, as BBIC, is currently undergoing testing in clinical trials. In human trials, elevated levels of proteolytic activities known to be affected by BBI are serving as intermediate marker endpoints (IME) in the cells of tissues at higher than normal risk of cancer development or those having premalignant characteristics. In previous animal studies, BBI was shown to be capable of bringing such elevated levels of proteolytic activity back to normal levels. BBIC appears to be a highly promising human cancer chemopreventive agent.
Dr. Kouichi Koshimizu discussed the antitumor promoting phytochemicals found in Thai vegetables. For example, 1'-acetoxychavicol acetate showed potent antitumor promoter activity in in vivo carcinogenesis experiments as well as in in vitro testing. The compound 1,2-di-!!!-1inolenoyl-3-!!!-galactosyl-su-glycerol showed the highest antitumor promoter activity when tested in vitro, and its in vivo activity is currently being evaluated. It was pointed out that many Thai vegetables were shown to contain antitumor-promoting factors in higher concentrations, as compared to vegetables commonly included in the Japanese diet.
Dr. Fung-Lung Chung noted that while the inverse relationship between vegetable consumption and the incidence of various cancers in humans has been well established, the exact nature of the ingredients responsible for the protective effects seen have yet to be determined. His laboratory has been focusing on the identification of the active compounds in cruciferous vegetables for protection against lung tumorigenesis. These studies have shown that phenethyl isothiocyanate (PEITC) and indoles found in cruciferous vegetables inhibited lung tumorigenesis in rodents. Structural activity relationship (SAR) studies showed that the alkyl chain length is critical for the inhibitory activity of arylalkyl isothiocyanates. Thus, the synthetic arylalkyl isothiocyanates with longer alkyl chains, up to 10-carbon, were considerably more potent than the shorter chain analogs. Extended SAR studies have also shown that aryl moiety is not necessary for activity, and those structural features associated with both high lipophilicity and low reactivity are the important determinants for the inhibitory potency of isothiocyanates. Evidence accumulated so far lends support to the theory that the mechanism by which isothiocyanates inhibit NNK lung tumorigenesis is mainly due to the inhibition of the cytochrome P-450 enzymes involved in NNK activation.
Since humans are exposed to PEITC through consumption of cruciferous vegetables, its metabolism in humans has been investigated. Since watercress is rich in gluconastrurtiin, a glucosinolate precursor or PEITC, it was chosen as a source of PEITC in human studies. The N-acetylcystein conjugate of PEITC has been identified as a useful marker for quantifying exposure to PEITC after ingesting watercress. This assay may be useful for epidemiological investigations. Little effect on drug hydroxylation was observed after consumption of watercress, suggesting that another mechanism, as yet unknown, might be important.
The various seminar presentations served to strengthen the view that the study of chemoprevention will continue to be of paramount importance in cancer research. The significance of compounds present in trace amounts in fruits, vegetables, and beverages that are consumed on a daily basis was highlighted throughout the discussions. It was suggested that the preventive effects seen with the consumption of these food constituents may be the result of synergistic interactions among these compounds. It was noted that none of the compounds have been shown to suppress human carcinogenesis in intervention studies and that the mechanism of action for phytochemicals remains to be elucidated. The issue of safety was particularly stressed, since there is evidence to suggest that phytochemicals shown to have anticarcinogenic effects for certain organs may also exhibit tumor promoting effects for other organs.
In conclusion, there was a consensus concerning the clear need to continue to pursue these important studies. The value of maintaining a close and continuous cooperation between the U.S. and Japan was viewed as essential to facilitate this research.

PARTICIPANTS

UNITED STATES

Dr. Lee W. Wattenberg
Department of Laboratory Medicine and Pathology
University of Minnesota Medical School
6-133 Jackson Hall
321 Church Street, S.E.
Minneapolis, MN 55455-0315

Dr. Allan Conney
Chairman, Department of Chemical Biology and Pharmacognosy
Rutgers
The State University of New Jersey
College of Pharmacy
PO Box 789
Piscatway, NJ 08855-0789

Dr. Ann R. Kennedy
Department of Radiation Oncology
University of Pennsylvania School of Medicine
195 John Morgan Building
37th and Hamilton Walk
Philadelphia, PA 19104-6072

Dr. Michael Pariza
Director and Chair
Food Research Institute
Food Microbiology and Toxicology
University of Wisconsin
1925 Willow Drive
Madison, WI 53706

Dr. Fung-Lung Chung
Associate Chief
Division of Chemical Carcinogenesis
American Health Foundation
Naylor Dana Inst. for Disease Prevention
1 Dana Road
Valhalla, NY 10595

Dr. Mou-Tuan Huang
Director, Biochemistry and Research Laboratories
Rutgers
The State University of New Jersey
Piscatway, NJ 08855-0789

Dr. H. Leon Bradlow
Strang Cornell Cancer Research Laboratory
510 East 73rd Street, 2nd Floor
New York, NY 10021-4004

JAPAN

Dr. Takashi Sugimura
President Emeritus
National Cancer Center
5-1-1 Tsukiji
Chuo-ku, Tokyo 104

Dr. Hoyoku Nishino
Chief, Cancer Prevention Division
National Cancer Center Research Institute
5-1-1 Tsukiji
Chuo-ku, Tokyo 104

Dr. Hiroyoki Tsuda
Chief, Chemotherapy Division
National Cancer Center Research Institute
5-1-1 Tsukiji
Chuo-ku, Tokyo 104

Dr. Shoji Shibata
Professor Emeritus
University of Tokyo
Shibata Laboratory for Natural
Medicinal Materials
3-2-7 Yotsuya
Shinjuku-ku, Tokyo 160

Dr. Hideki Mori
1st Department of Pathology
Gifu University School of Medicine
40 Tsukasa-machi
Gifu 500

Dr. Isao Tomita
Laboratory of Health Science
School of Pharmaceutical Science
University of Shizuoka
395 Yada
Shizuoka 422

Dr. Kouichi Koshimizu
Department of Food Science and Technology
Faculty of Agriculture
Kyoto University
Oiwake-cho, Kitashirakawa
Sakyo-ku, Kyoto 606-01