We performed this study to determine whether human mesothelial cells are capable of undergoing neoplastic change in vitro and to observe their interaction with the activated c-Ha-ras (HRAS1) oncogene EJ-ras, which has a role in the development of many malignant human tumors. Mesothelial cells are presumed to be the progenitor cells of malignant mesothelioma, a cancer strongly correlated with asbestos exposure. Previously, we established a non-tumorigenic cell line, MeT-5A, from normal human mesothelial cells after transfection with a plasmid containing the simian virus 40 (SV40) early-region genes. In the present study, we performed transfection of a plasmid containing the EJ-ras gene and the neomycin-resistance gene into these cells and selected a population resistant to G418, a neomycin analogue. Cells from this cell line formed rapidly growing sc tumors in NIH Swiss athymic nude mice, but untransfected with the vector DNA and selected for G418 resistance formed no tumors. The tumors formed by EJ-ras-transfected cells were established in vitro, and cells from these tumor cell lines exhibited a characteristic altered morphology. The cells had the same isoenzyme phenotype as the parent cells, and they expressed the mutant EJ-ras p21 protein. This first demonstration of malignant transformation of human mesothelial cells in vitro may permit molecular analysis of mesothelial carcinogenesis.

Overexpression of c-raf-1 and the myc family of protooncogenes is primarily associated with small cell carcinoma, which accounts for approximately 25% of human lung cancer. To determine the functional significance of the c-raf-1 and/or c-myc gene expression in lung carcinogenesis and to delineate the relationship between protooncogene expression and tumor phenotype, we introduced both protooncogenes, alone or in combination, into human bronchial epithelial cells. Two retroviral recombinants, pZip-raf and pZip-myc, containing the complete coding sequences of the human c-raf-1 and murine c-myc genes, respectively, were constructed and transfected into simian virus 40 large tumor antigen-immortalized bronchial epithelial cells (BEAS-2B); this was followed by selection for G418 resistance. BEAS-2B cells expressing both the transfected c-raf-1 and c-myc sequences formed large cell carcinomas in athymic nude mice with a latency of 4-21 weeks, whereas either pZip-raf- or pZip-myc-transfected cells were nontumorigenic after 12 months. Cell lines established from tumors (designated RMT) revealed the presence of the cotransfected c-raf-1 and c-myc sequences and expressed morphological, chromosomal, and isoenzyme markers, which identified BEAS-2B cells as the progenitor line of the tumors. A significant increase in the mRNA levels of neuron-specific enolase was detected in BEAS-2B cells containing both the c-raf-1 and c-myc genes and derived tumor cell lines. The data demonstrate that the concomitant expression of the c-raf and c-myc protooncogenes causes neoplastic transformation of human bronchial epithelial cells resulting in large cell carcinomas with certain neuroendocrine markers. The presented model system should be useful in studies of molecular events involved in multistage lung carcinogenesis.

Chromosomal changes in liver epithelial cells undergoing AFB1* induced transformation were studied in two well-characterized epithelial cell cultures derived from rat liver. Diploid liver cells, which did not possess any in vitro transformation marker or the ability to induce tumors in vivo, were treated with the hepatocarcinogen AFB1 and the karyotypes were analyzed sequentially at regular intervals. Although a high percentage of aneuploid cells was seen after AFB1 treatment, no “stem-line” was formed during the period studied. Marker chromosomes appeared prior to the acquisition of malignant potential in liver cells treated repeatedly with AFB1. Chromosomes #1 and #2, were found to be preferentially involved in the formation of marker chromosomes induced by AFB1. Neither marker chromosomes nor malignant potential were noted for 22 weeks after a single treatment with AFB1, to an asynchronous population of rat liver cells. However, when a semi-synchronous population of liver cells were treated once with AFB1, marker chromosomes were observed, but only in a small percentage of the cells. Although the direct implication of the chromosomal alterations in the initiation stage of the carcinogenic process is unclear, procedures known to favor the selective growth of transformed cells in vitro and in vivo resulted in a specific enrichment of the cells with marker chromosomes, suggesting that such cells may be important at least in the progression of the carcinogenic process.

Activated ras oncogenes have previously been implicated in the pathogenesis of human lung carcinomas. A v-Ha-ras-containing retrovirus, Zip-ras, was generated by inserting the coding region of the v-Ha-ras oncogene into the Zip-NeoSV(X) [Cepko et al., Cell 37:1053-1062, 1984] retroviral vector. Amphotrophic Zip-ras retrovirus was used to infect an SV40 large T antigen-positive immortalized cell line, BEAS-2B, derived from normal bronchial epithelial cells, the predominant progenitor cells of human lung carcinomas. Zip-ras-infected BEAS-2B cells selected for G418 resistance formed anaplastic carcinomas in 12 of 15 athymic nude mice (latency 3 wk), whereas Zip-NeoSV(X)-infected BEAS-2B control cultures inoculated into 12 nude mice formed no tumors after a minimum of 7 mo. Tumor cell lines were established and demonstrated to be of human epithelial origin and to express v-Ha-ras p21 protein. A common feature of the tumor cell lines was an increase in ploidy. The increased efficiency of neoplastic transformation by v-Ha-ras of cell lines as compared with our previous results with normal bronchial epithelial cells [Yoakum et al., Science 227:1174-1179, 1985] is consistent with the hypothesis that the "immortalization" step is rate-limiting in in vitro human epithelial cell carcinogenesis.

Many human bronchial adenocarcinomas have been shown to contain an activated Ki-ras oncogene (Rodenhuis et al., N. Engl. J. Med. 317 929-935, 1987). To test the hypothesis that activated Ki-ras may be causally related to human bronchial carcinogenesis, v-Ki-ras oncogene was transferred into an established human bronchial epithelial cell line, BEAS-2B, by infection with Kirsten murine sarcoma virus (Ki-MSV) or by transfection with a plasmid containing the transforming region of Ki-MSV. These cells formed poorly differentiated adenocarcinomas in athymic nude mice. Cell lines established from these tumors expressed v-Ki-ras p21 protein and were highly tumorigenic. Whereas serum or transforming growth factor beta 1 induced the BEAS-2B cells at clonal density to undergo growth arrest and squamous differentiation, BEAS-2B cells containing activated ras genes were unaffected by transforming growth factor beta 1 and were mitogenically stimulated by serum.

Transfection of normal human bronchial epithelial (NHBE) cells with a plasmid carrying the ras oncogene of Harvey murine sarcoma virus (v-Ha ras ) changed the growth requirements, terminal differentiation, and tumorigenicity of the recipient cells. One of the cell lines isolated after transfection (TBE-1) was studied extensively and shown to contain v-Ha ras DNA. Total cellular RNA from TBE-1 cells hybridized to v-Ha ras structural gene fragment probes five to eight times more than RNA from parental NHBE cells. The TBE-1 cells expressed phosphorylated v-Ha ras polypeptide p21, showed a reduced requirement for growth-factor supplements, and became aneuploid as an early cellular response to v-Ha ras expression. As the transfectants acquire an indefinite life-span and anchorage independence they became transplantable tumor cells and showed many phenotypic changes suggesting a pleiotropic mechanism for the role of Ha ras in human carcinogenesis.

The subcellular and intralobular distributions of a protein which specifically inhibits the proliferation of normal liver cells were determined in rat liver, using a combination of immunological and biochemical techniques. The IgG fraction from an antiserum raised against the hepatic proliferation inhibitor was isolated by protein A-Sepharose CL-4B chromatography and shown to be highly specific for the antigen using electroimmunodiffusion and affinity chromatography. To determine the intracellular location of the inhibitor, subcellular fractions were prepared from adult rat livers by differential centrifugation. The cytoplasmic fraction contained the biologically active cytostatic inhibitor, whereas the nuclear and mitochondrial fractions were inactive. Cytoplasmic localization of the hepatic proliferation inhibitor was further confirmed by anion exchange high performance liquid chromatography and by double immunodiffusion with the anti-hepatic proliferation inhibitor IgG. When liver sections were subjected to histochemical staining mediated through the immune IgG and an avidin-biotinylated horseradish peroxidase complex, the parenchymal liver cells were stained, but endothelial and connective tissue cells were not. Although some staining was evident throughout the liver parenchyma, the most intensely stained cells were located in the centrilobular region. Moreover, an age-dependent increase in the staining intensity and/or in the number of cells containing the proliferation inhibitor was observed. Preliminary experiments showed that little, if any, staining occurred in hepatocellular carcinoma cells. This highly specific IgG can be used to monitor alterations in the content and location of hepatic proliferation inhibitor in proliferative disorders of the liver.

Clonal growth and serial subculture of diploid liver epithelial cells from neonatal rats were achieved in a serum-free medium (SFM) supplemented with linoleic and oleic acid linked to fatty acid-free bovine serum albumin (fafBSA), epidermal growth factor (EGF), transferrin, insulin, selenous acid, and fetuin. Because it is not known whether factors added to defined media facilitate attachment, support proliferation, or both, a serum-free "attachment medium" was first devised in which cells would attach to the substratum without loss of viability. Then a growth medium that would support cell proliferation was developed. Fetuin enhanced the degree of attachment, and the lipid supplements and EGF induced a marked proliferative response. Serum-free medium supported the formation of colonies equivalent in size, number, and morphology to those obtained in serum-supplemented medium. Cells plated at a higher inoculum density and subcultured regularly for up to 25 wk underwent two to three doublings per week and acquired a flattened epithelial cell morphology. Early passages of rat liver epithelial cells, cultured in SFM may be useful in studies of the regulation of cell proliferation and differentiation.

Experimental conditions known to alter the shape, permeability and organization of cells were used to find out their effects on tyrosine aminotransferase (TAT) in rat liver epithelial cell lines. To produce a more spheroidal morphology for non-malignant cells than that obtainable on plastic, floating collagen and poly(2-hydroxyethylmethacrylate) (poly(HEMA)), were used as substrata. Under these experimental conditions the basal level of TAT activity increased 1.5–2.5-fold. When monolayer cultures were permeabilized by the use of a hypertonic salt solution, the basal activity increased 4–5-fold. TAT activity was also elevated in hepatoma cells cultured in anchorage-independent conditions. The enzyme was not inducible by dexamethasone (DEX) under any of these culture conditions, and the lack of induction was not due to the absence of receptors for this hormone. These studies have shown that the production of TAT, one of the characteristics of adult liver, has persisted in a number of rat epithelial cell lines derived from normal, malignant or regenerating liver and its activity was influenced by the different culture conditions employed.