We previously isolated a novel human transforming sequence from a primary stomach cancer and identified the gene as an activated version of the c-raf-1 gene which is the human homologue of v-raf, a viral oncogene encoding a serine/threonine-specific protein kinase. Of 57 kbp of the human sequence isolated, a region of 40 kbp was found to be the minimum functional unit for the transforming activity, because a cosmid clone harboring this region was capable of inducing foci upon transfection. The size of the transcript of the transforming c-raf-1 gene was estimated to be about 2.8kb. Analyses of cDNA clones of this gene revealed that the gene was generated by substitution of the 5'-sequence (exons 1-5) of the normal c-raf-1 gene with an unrelated human sequence. We identified a region in the genomic clone where the rearrangement had occurred. The rearranged EcoRI fragment was detected in all the primary transformants obtained from two independent transfections, suggesting that the recombination had occurred in the primary cancer. The substituted cDNA sequence is composed of an open reading frame, which joins to exon 6 of the c-raf-1 gene in an in-phase manner. The substituted open reading frame encodes an extremely hydrophobic polypeptide. Thus, the putative product of the transforming gene seems to have a hydrophobic stretch ahead of the ser/thr-protein kinase domain of the c-raf-1 gene product. These results suggest that the truncation or replacement of the amino-terminal domain of the c-raf-1 protein leads to constitutive activation of the protein kinase residing in the downstream domain.

Primary human hepatocellular carcinomas (HCC) were surveyed for oncogenes by transformation assays using NIH3T3 cells and the calcium phosphate coprecipitation method. High-molecular-weight DNA obtained from the HCC tissues was employed for this purpose. One new transforming DNA, named lca (for liver cancer) was obtained, and its properties were studied in detail. This transforming DNA had a linkage to the Alu sequence and was cloned in lambda phage. Restriction enzyme analyses showed that the minimum size of the lca DNA is about 10.5 kilobase pairs and that its nucleotide sequence is different from that of any known human gene or retroviral oncogene. Southern blot analyses of DNAs from flow-sorted human chromosomes and human-mouse somatic cell hybrids indicated that the lca DNA is located on human chromosome 2. It showed identical restriction enzyme cleavage profiles with its counterpart DNA obtained from normal tissue, indicating that there is no extensive DNA rearrangement associated with the activation process of the lca DNA. The normal counterpart DNA shows no transforming activity. An independently obtained transforming DNA from another HCC exhibited identical restriction enzyme cleavage profiles. Thus, lca DNA is likely to represent a commonly encountered transforming DNA in HCC.

In hepatitis B virus (HBV) related hepatoma samples there is a high frequency of HBV-DNA integration into the chromosome, but the frequency of integration in the early-stage of carcinogenesis is expected to be even higher. Structural an analysis of integrated HBV-DNA and chromosomal flanking DNAs disclosed that the marked rearrangement of chromosomal DNA is not directly linked to carcinogenesis. HBV-DNA integration and chromosomal DNA depletion occur even in chronic hepatitis tissues. Integration is, therefore, thought to trigger a series of reactions which lead to carcinogenesis. However, the oncogene is not detected within the HBV genome, and the HBV integration site in the chromosome is variable. As one approach to clarifying the cause and effect relationship between HBV-DNA integration and liver carcinogenesis DNA transfection experiments were conducted using mouse NIH3T3 cells to detect hepatoma-related oncogenes. As a result, the well-known N-ras gene and an unknown oncogene were independently isolated from different hepatoma tissues.

One of the major causes of failure of cancer chemotherapy is the proliferation of specific drug-resistant tumor cells during treatment. Drug-resistant tumor cells, however, usually bear biochemical changes which are related to the resistance mechanisms. New modalities against resistant cells could be possible if we were able to characterize these biochemical changes. Vincristine (VCR)- and adriamycin (ADM)-resistant tumor sublines show cross-resistance (pleiotropic drug resistance) to other unrelated drugs. VCR- and ADM-resistant sublines possess an enhanced outward transport of antitumor agents, which results in a low accumulation of antitumor agents in the cells. The cells express unique glycoproteins in the plasma membrane, and possess a higher calcium content in the cells. They also have double-minute chromosomes and homogeneously staining regions in chromosomes. By targeting for these biochemical changes, we have established new modalities against drug-resistant tumor cells. Calcium channel blockers inhibited the enhanced outward transport of VCR and ADM from resistant tumor cells, and thus overcame resistance to these agents. This approach showed potential usefulness in clinical trials. Another possible approach against drug-resistant tumor cells could be the utilization of monoclonal antibodies against unique glycoproteins in the plasma membrane of resistant tumor cells. We have developed monoclonal antibodies against adriamycin-resistant human myelogenous leukemia K562. Our recent progress with work on calcium channel blockers and monoclonal antibodies are discussed in this paper.

Choriocarcinoma cells grown in the presence of MTX have developed resistance in two ways. The HCCM derived sublines (relatively high MTX resistant) produced enhanced levels of DHFR and had relatively unimpaired transport of MTX, though altered transport was the primary determinant of response in the CC1 derived sublines (low MTX resistant). Since the selection procedure used was identical, it was assumed that altered MTX transport was insufficient to account entirely for various degrees of resistance. Increased DHFR activity was necessary for the development of high MTX resistance. The overproduction of DHFR was the consequence of amplification of the DHFR gene sequence. The incidence of DMs in metaphases paralleled the degree of resistance. Since DMs were also present in the cells not showing DHFR gene amplification, mechanisms other than DHFR gene multiplication were responsible for the de novo synthesis of DMs.

A hereditary testicular/ovarian teratoma strain (Tera) of rats was developed from the Csk: Wistar-Imamichi strain. As the teratoma consisted of tridermic tissues such as bone, epithelium and neural tissue, it was diagnosed as triphyllomatous teratoma. The frequency of the teratoma was about 25% in either sex, with no sexual difference. Accordingly, the heredity of the teratoma appeared to be an autosomal single recessive trait (symbol, tera).

Based on a result suggesting that the frequency of the Trp-P-2-sensitive persons in patients suffering from cancers at digestive tract is higher than in healthy persons when the sensitive person is defined as the person whose SCE frequency in peripheral lymphocytes induced by in vitro treatment with Trp-P-2 is more than 2 sigma above the mean of the SCEs of healthy population under the identical assay conditions. The possibility to detect the cancer predisposition is discussed.

Case-control analysis of the data from Aichi Cancer Registry revealed that when smoking and family history of lung cancer were combined, relative risk of this cancer increased in an additive manner to the higher level than that for each single factor. This risk increase was larger for squamous cell carcinoma than for adenocarcinoma (relative risk; 4.02 vs 2.20). Other data from three case-control studies were discussed, and a new approach of familial case-control study of cancer was proposed. Finally importance of cancer prevention and detection for families at high risk of cancer was stressed.

Family history of cancer has been regarded as an index of genetic influence. A population-based study revealed some characteristics of dietary habits associated with family history of specific sites (stomach, colo-rectum, uterus and breast) of cancer. Those characteristics resembled those of cancer patients themselves. These results suggested that family history of cancer could also be an indicator of the influence of environmental factors, especially dietary habits.

Six pedigrees of cancer family syndrome, multiple colorectal cancer (89 cases) and colorectal cancer with malignancies in other organs (95 cases), experienced at the National Cancer Center Hospital were reviewed with particular emphasis placed on the familial aggregation and cumulative age incidence. A high association with genetic or familial factors was found in multiple primary colorectal cancer with a family history of colorectal cancer.

Familial breast cancer are clinically characterized by early age of onset, the increased frequency of multiple primary malignant neoplasms including bilateral breast cancer and non-left-side dominance with regard to the site of onset. The genetic consideration of the pedigrees suggests that familial clustering of breast cancer is related to the autosomal dominance of susceptibility, polygenes and incomplete penetrance. The additive effects obtained by combining the genetic with the environmental factors are recognized in familial breast cancer. Clinical surveys and follow-ups of breast-cancer-prone families are required to facilitate the early detection and treatment.

A case-control study of 65 women with ovarian epithelial carcinoma and 130 individually age-matched controls were conducted from 1985 to 1986 in Hokkaido. The following factors were significantly greater in the cases than in the controls: higher prevalence of breast, uterine, or ovarian cancer of their mother or sister; nulliparous; less experience in induced abortion; less frequent ligation of Fallopian tubes; irregular menstruation around at 20 years of age; past history of pathological amenorrhea in the premenopausal group; daily meat or fish consumption in the postmenopausal group. It is inferred that women with ovarian cancer had a genetical predisposition and dysfunctional ovaries from early in the life.

The purpose of this study was to examine whether patients with gastric cancer differ from those with benign gastroduodenal diseases in terms of familial disposition. This series consisted of 1,233 patients with gastric cancer and 2,200 patients with benign gastroduodenal diseases, who had had a gastrectomy between 1972 and 1979 in one hospital. The family information on cancer deaths of first and second relatives of the patients was obtained at the time of operation. The age-adjusted relative risk of positive family history of cancer was higher among the patients with gastric cancer than that among those with benign gastroduodenal diseases. The patients with diffuse type of cancer had a higher risk of positive family history than those with intestinal type of cancer. The non-cancer patients with family history had a slightly higher risk, although not statistical significant, to die from cancer than those without family history. Higher frequency of blood type A was seen in gastric cancer patients with positive family history. These results suggest that certain types of gastric cancer may seem to be associated with genetic factor.

A numerous number of case-control studies have already demonstrated a certain extent of familial clustering for most human cancers. This is the first report of follow-up study with respect to familial predisposition to cancer. About 40 thousands women who participated in a cervical cancer screening program in 1977 were followed up until 1985 by using the population-based cancer registry. Cancer patients who were detected in the cohort population and their controls were compared for the family history of cancer and female reproductive history both of which had been recorded in a computer file beforehand. For large bowel cancer and breast cancer, positive effect of familial predisposition was observed, though not statistically significant.