Analysis of steroid resistance in lymphoid cell hybrids

Evidence for trans regulation of apoptosis in intertypic somatic cell hybrids

The genetic components required for glucocorticoid induction of apoptosis were studied by using somatic cell hybridization. Intertypic whole-cell hybrids were generated by crossing the glucocorticoid-resistant rat liver cell line Fado-2 with the glucocorticoid-sensitive mouse thymoma cell line BW5147.3. Morphological and biochemical criteria were used to assess sensitivity or resistance to glucocorticoid-induced cell death. Both phenotypes were observed, and all of the hybrids retained a functional glucocorticoid receptor as judged by their abilities to induce the metallothionein gene in response to dexamethasone (Dex). Sensitivity to apoptosis did not correlate with morphological phenotype in that not all suspension cells were sensitive. The effect of glucocorticoids on the expression of apoptosis-linked genes was analyzed in a subset of Dex-sensitive and Dex-resistant hybrids. p53 and c-myc mRNAs were present in parental cells as well as sensitive and resistant hybrid cells, and their levels were not affected by glucocorticoid treatment. bcl-2 expression was restricted to the thymoma cell line and was also not affected by glucocorticoids. We did not detect any bcl-2 mRNA in the hepatoma cell line and the hybrids, suggesting that, as with most tissue-specific genes, bcl-2 is regulated in trans. Furthermore, while the majority of hybrids analyzed retained a full complement of mouse chromosomes, sensitive hybrids were missing some rat chromosomes (preferentially chromosomes 16 and 19), indicating that apoptosis is subject to trans repression. Resistant cells thus appear to repress the activity or synthesis of a nuclear factor that interacts with a glucocorticoid-dependent gene(s) to activate the cell death pathway.

Evidence for trans regulation of apoptosis in intertypic somatic cell hybrids

The genetic components required for glucocorticoid induction of apoptosis were studied by using somatic cell hybridization. Intertypic whole-cell hybrids were generated by crossing the glucocorticoid-resistant rat liver cell line Fado-2 with the glucocorticoid-sensitive mouse thymoma cell line BW5147.3. Morphological and biochemical criteria were used to assess sensitivity or resistance to glucocorticoid-induced cell death. Both phenotypes were observed, and all of the hybrids retained a functional glucocorticoid receptor as judged by their abilities to induce the metallothionein gene in response to dexamethasone (Dex). Sensitivity to apoptosis did not correlate with morphological phenotype in that not all suspension cells were sensitive. The effect of glucocorticoids on the expression of apoptosis-linked genes was analyzed in a subset of Dex-sensitive and Dex-resistant hybrids. p53 and c-myc mRNAs were present in parental cells as well as sensitive and resistant hybrid cells, and their levels were not affected by glucocorticoid treatment. bcl-2 expression was restricted to the thymoma cell line and was also not affected by glucocorticoids. We did not detect any bcl-2 mRNA in the hepatoma cell line and the hybrids, suggesting that, as with most tissue-specific genes, bcl-2 is regulated in trans. Furthermore, while the majority of hybrids analyzed retained a full complement of mouse chromosomes, sensitive hybrids were missing some rat chromosomes (preferentially chromosomes 16 and 19), indicating that apoptosis is subject to trans repression. Resistant cells thus appear to repress the activity or synthesis of a nuclear factor that interacts with a glucocorticoid-dependent gene(s) to activate the cell death pathway.

A genetic analysis of processes regulating cytochrome P4501A1 expression

Cytochrome P4501A1 and its associated aryl hydrocarbon hydroxylase activity are highly inducible in the mouse hepatoma cell line, Hepa-1, by substrates of the enzyme and related compounds, such as 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Mutants of this cell line, deficient in P4501A1 inducibility, were isolated. Some of the mutants show a dominant phenotype. Such mutants may have resulted from a genetic alteration leading to the inappropriate activation of a repressor gene that normally functions to restrict high level inducibility to the liver and certain other organs or to certain developmental stages. One dominant mutant was shown to express a protein that prevents binding of the liganded aryl hydrocarbon (Ah) receptor (which mediates induction of P4501A1) to its recognition sequence in DNA (the xenobiotic responsive element, or XRE). The majority of mutants are recessive, and were assigned to four different complementation groups (which probably correspond to four different genes). Gene A corresponds to the structural gene (Cyp1a-1) for P4501A1. Mutations in genes B, C and D all affect functioning of the Ah receptor. A cDNA for gene C was cloned. The encoded protein (ARNT) is required for ligand-dependent translocation of the Ah receptor to the nucleus and its binding to the XRE. ARNT and the Ah receptor form a heterodimeric complex which binds the XRE in a fashion such that both subunits bind the XRE directly. Both ARNT and the Ah receptor contain basic helix-loop-helix motifs. Such motifs have been identified in several transcription factors that bind DNA as heterodimers or homodimers. The roles of the proteins corresponding to the B and D genes are presently under investigation.

Cloning of a factor required for activity of the Ah (dioxin) receptor

The aryl hydrocarbon (Ah) receptor binds various environmental pollutants, such as polycyclic aromatic hydrocarbons, heterocyclic amines, and polychlorinated aromatic compounds (dioxins, dibenzofurans, and biphenyls), and mediates the carcinogenic effects of these agents. The complementary DNA and part of the gene for an 87-kilodalton human protein that is necessary for Ah receptor function have been cloned. The protein is not the ligand-binding subunit of the receptor but is a factor that is required for the ligand-binding subunit to translocate from the cytosol to the nucleus after binding ligand. The requirement for this factor distinguishes the Ah receptor from the glucocorticoid receptor, to which the Ah receptor has been presumed to be similar. Two portions of the 87-kilodalton protein share sequence similarities with two Drosophila proteins, Per and Sim. Another segment of the protein shows conformity to the consensus sequence for the basic helix-loop-helix motif found in proteins that bind DNA as homodimers or heterodimers.

Genetic and molecular analysis of the Ah receptor and of Cyp1a1 gene expression

The Ah receptor is a soluble protein complex that mediates carcinogenesis by a wide range of environmental pollutants, including polycyclic aromatic hydrocarbons, heterocyclic amines, and polychlorinated aromatic compounds. The best understood activity of the receptor concerns its role in the induction of cytochrome P450IA1. We undertook a somatic cell genetic analysis of P450IA1 induction using the mouse hepatoma cell line, Hepa-1. Clones of Hepa-1 were isolated that are defective in induction of P450IA1. Evidence was obtained that the clones are mutational in origin. Cell fusion experiments demonstrated that a few of the mutants are dominant, while the majority are recessive. The dominant mutants were shown to synthesize a repressor of P450IA1 transcription. The recessive mutants were assigned to 4 complementation groups (probably corresponding to 4 different genes). Complementation group A corresponds to the P450IA1 structural gene. Mutations in the B, C and D genes all affect functioning of the Ah receptor. A 'reverse selection procedure', whereby cells that express P450IA1 inducibility can be selected from a majority population of cells lacking inducibility, was developed. The reverse selection procedure was used to isolate transfectants of representative recessive mutants in which the mutational defects are complemented by exogenously applied genomic DNA. A human DNA-derived transfectant of a C- mutant was used to clone the human C gene. The C gene is not the ligand-binding subunit of the Ah receptor but is a protein that is required for translocation of Ah receptor-ligand complexes from cytoplasm to nucleus. In analogous experiments the dominant gene from one of the dominant mutants was transfected into wild-type Hepa-1 cells. Success in transfecting the dominant gene should provide the means for cloning it.

Assignment of the human gene for the glucocorticoid receptor to chromosome 5

Human lymphoblastic leukemia cells of line CEM-C7 are glucocroticoid-sensitive and contain glucocorticoid receptors of wild-type characteristics. EL4 mouse lymphoma cells are resistant to lysis by glucocorticoids due to mutant receptors that exhibit abnormal DNA binding. Hybrids between the two cell lines were prepared and analyzed with respect to glucocorticoid responsiveness and to receptor types by DNA-cellulose chromatrography. Sensitive hybrid cell clones contained the CEM-C7-specific receptor in addition to the EL4 type of receptor. Several sensitive hybrid cell clones were used for selection of resistant segregants by growth in the presence of high concentrations of glucocorticoid. These segregants had lost the wild-type CEM-C7 receptor, while the EL4-specific receptor was retained. To identify the human chromosome that was lost concordantly with the CEM-C7 receptor the chromosomes of hybrid cells were studied by alkaline Giemsa (G-11) staining and trypsin/Giemsa banding. All hybrids contained human chromosomes in addition to one to two sets of EL4 chromosomes. Human chromosome 5 was present in all hybrid cell clones that expressed the CEM-C7 receptor and it was absent from those that did not. This absolute correlation was not observed for any other human chromosome. We conclude that the human gene for the glucocorticoid receptor is located on chromosome 5.

Analysis of androgen-sensitivity in rat prostate X mouse kidney cell hybrids

Variant androgen-sensitive cell lines were produced by fusing freshly isolated epithelial cells from the rat ventral prostate with a line of murine renal tumor (RAG) cells. The properties of the cloned lines of the prostate X RAG hybrids can be summarized as follows: (1) the modal chromosome number of the hybrid cell lines ranged from 68 to 176; (2) the cells had doubling times of 7.6-49.5 h; and (3) epitheloid, ameboid and intermediate morphologies were observed among the various lines. The proliferative response of various hybrid lines to treatment with 10 nM 5 alpha-dihydrotestosterone was used to classify the hybrids as either very sensitive (greater than 40% reduction in cell doubling time), sensitive (greater than 10% reduction in doubling time) to androgens, or insensitive (less than 10% reduction in doubling time) to androgens. There was no direct relationship between the androgen-sensitivity of the cells and their androgen receptor content, suggesting that these variant cell lines may be useful for the study of the genetic factors involved in cellular responses to androgens.

Dominant and recessive aryl hydrocarbon hydroxylase-deficient mutants of mouse hepatoma line, Hepa-1, and assignment of recessive mutants to three complementation groups

Fifty-four benzo[a]pyrene (BP)-resistant, aryl hydrocarbon hydroxylase (AHH)-deficient mutants were found to be recessive, while five were dominant. Hybrids between the former mutants and the wild-type were killed by BP, and possessed AHH activities of at least 0.5 (relative to the wild-type). Dominant-mutant--wild-type hybrids were resistant to BP and had activities of about 0.05. Additional experiments assigned the recessive mutants to three complementation groups, designated A through C. Group-B--group-C hybrids were exceptional in possessing a mean AHH activity (0.36), less than the value (0.5) expected from gene dosage. This deficiency was probably due, in part, to instability of AHH activity in these hybrids. However, all hybrids tested retained stable DNA complements, equal to the sum of those of their parents, for 140 days in culture. Previous studies have shown that group B and group C mutations both affect the functioning of a cytosolic receptor required for AHH induction (1).

Evaluation of a mutation test using S49 mouse lymphoma cells and monitoring simultaneously the induction of dexamethasone resistance, 6-thioguanine resistance and ouabain resistance

A test for the detection of chemically induced mutants in S49 mouse lymphoma cells is described. These cells can be plated in parallel in several selective media; the induced frequencies of dexamethasone-resistant, 6-thioguanine-resistant and ouabain-resistant mutants were compared. The first two selection systems permit the detection of all kinds of mutation that result in alteration or partial or complete loss of the gene product concerned, whereas ouabain-resistant mutants can only be induced with strong point mutagens in these cells. Dexamethasone resistance is the marker induced at the highest frequency among these three. Data obtained from this selection system are therefore the most amenable to statistical analysis. Dexamethasone resistance is expressed within a short time after mutagenesis (3 days), and because S49 cells do not display metabolic co-operation, large numbers of cells can be screened. A metabolizing system in vitro with rat-liver homogenate may be included in tests of indirectly acting mutagens. These features make the S49 mutation test system using dexamethasone resistance as the main marker and other markers as internal controls an attractive tool in mutation testing in somatic cells in vitro.

Clonal growth of lymphoid cells in serum-free media requires elimination of H2O2 toxicity

We have recently described the development of a serum-free medium that contains casein, insulin, testosterone, transferrin, and linoleic acid and that supports the long-term growth of a wide variety of lymphoid cells. A problem of culturing cells in this medium is the difficulty of cloning cells or growing cells at low density. We now describe the formulation of a chemically defined medium that supports the clonal growth of the murine S49 T lymphoma cell line. This medium contains catalase, insulin, transferrin, testosterone, Na2SeO3, and dilinoleoyl phosphatidylcholine and contains less than 50 micrograms/ml total protein. The two novel additions in this medium are catalase, which replaces casein and dilinoleoyl phosphatidylcholine, which substitutes for linoleic acid in this defined medium. In addition to S49 cells, the medium described above supports the long-term growth of other lymphoid cells, including human and murine hybridomas. We propose that catalase functions to degrade H2O2 that is present in the cultures and that casein, bovine serum albumin, and other proteins commonly included in media for cultured cells may also scavenge H2O2. Na2SeO3 also partially protects against the death of cells at clonal density and this protection may, like catalase, be due to removal of H2O2. Our results suggest that H2O2 is an important cytotoxic agent that prevents growth of lymphoid cells during culture in serum-free media and perhaps in serum-containing media as well.

A new determinant of glucocorticoid sensitivity in lymphoid cell lines

The SAK cell line, derived from a spontaneous thymic lymphoma in an AKR mouse, is resistant to lysis by glucocorticoids in spite of the presence of functional glucocorticoid receptor. Receptor function was determined by hormone binding analyses, as well as characterization of hormonal effects on cell growth and on the accumulation of murine leukemia virus and metallothionein mRNAs. SAK cells were fused with a receptor-defective (and therefore resistant) variant of a well-characterized murine thymoma line, W7. The resulting hybrids are glucocorticoid sensitive, demonstrating complementation of the receptor defect in W7 cells by the functional glucocorticoid receptor of SAK. This fusion shows that SAK cells are resistant to the hormone due to the absence of another function designated "I" for lysis. SAK cells were also fused with glucocorticoid-sensitive W7 cells (containing wild-type receptor), generating glucocorticoid-sensitive hybrids, which demonstrate that the dexamethasone-resistant phenotype of the SAK cells is recessive. Resistant derivatives of this hybrid were found which still contain the full amount of receptor. Chromosome analysis revealed that, on the average, the resistant derivatives had lost two chromosomes, suggesting segregation of chromosomes carrying genetic material necessary for the "lysis" function. The drug 5-azacytidine (a known inhibitor of DNA methylation) has been shown to cause heritable changes in gene expression. Treatment of SAK cells with 5-azacytidine generated glucocorticoid-sensitive clones at high frequency, suggesting that the gene(s) involved in the "lysis" function are intact and have been inactivated through a process such as differentiation.

Decreased acquisition of glucocorticoid resistance in tetraploid human lymphoid cells

Resistance to dexamethasone (1 microM) was measured in glucocorticoid-sensitive diploid and tetraploid clones of the human leukaemic cell line CCRF-CEM (clone C7) during continuous culture and after X-ray or chemical mutagenesis. In continuous culture resistant diploid cells accumulated at a rate of about one cell per 10(5) divisions, while the rate for tetraploid cells was less than one per 10(7) divisions. Chemical and X-ray mutagenesis caused a marked increase in the number of resistant diploid cells but had very little effect on tetraploid cells. These results are consistent with a mutational basis for the acquisition of the glucocorticoid-resistant phenotype in human lymphoid cells.

Specific binding of the glucocorticoid-receptor complex to the mouse mammary tumor proviral promoter region

To elucidate the molecular mechanism by which steroid hormones exert their regulatory function, we investigated the interaction of a glucocorticoid-receptor complex with purified DNA fragments from cloned mouse mammary tumor (MMTV) proviral DNA. With a DNA-cellulose binding assay, rat and mouse glucocorticoid receptors were found to interact with a high affinity site or sites in or near the promoter region of the MMTV proviral DNA. The assay allowed the use of unpurified as well as purified receptor, and therefore made it possible to investigate the binding properties of mutant receptors. Two nuclear-transfer-deficient receptors have a decreased affinity for specific as well as unspecific DNA, but are still capable of distinguishing between the two types of DNA. The existence of specific DNA binding sites for steroid-receptor complexes is discussed in the context of a general model of steroid hormone action.

Glucocorticoid-mediated inhibition of ornithine decarboxylyase activity in S49 lymphoma cells

Incubation of wild-type S49 lymphoma cells with glucocorticoids, such as dexamethasone and hydrocortisone, inhibits the activity of ornithine decarboxylase (L-ornithine carboxylyase, EC 4.1.1.17), the rate-limiting enzyme in the pathway of polyamine biosynthesis. The kinetics of this inhibition are more rapid than the glucocorticoid-mediated growth arrest in the G1 phase of the cell cycle or in glucocorticoid-mediated cytolysis of these cells. The inhibition of ornithine decarboxylase activity by corticosteroids is specific for steroids of the glucocorticoid class. Results obtained with variant S49 cells having lesions in the pathways of glucocorticoid or cyclic AMP action indicate that cytoplasmic glucocorticoid receptors, as well as nuclear transfer of steroid--receptor complexes, are required for the inhibition of ornithine decarboxylase activity but that this inhibition does not require hormonal activation of adenylate cyclase or cyclic AMP-dependent protein kinase. Because glucocorticoid-mediated inhibition of ornithine decarboxylase occurs when cellular protein synthesis has decreased less than 20%, this inhibition may represent a specific glucocorticoid-mediated deinduction of ornithine decarboxylase in S49 cells. Inhibition of ornithine decarboxylase activity may offer a useful marker for suppression of growth and cell cycle progression in these and other lymphoma cells.

Specific receptors control steroid sensitivity in lymphoma cell hybrids

Hybrids between different mouse lymphoma cell lines were prepared. 3 sub-lines of S49.1 were used: one of them was glucocorticoid-sensitive (wild-type, wt), the others were resistant with defects either in receptor binding (r-) or in nuclear transfer of the receptor-glucocorticoid complex (nt-). These cells were hybridized with another glucocorticoid-resistant lymphoma, EL4, which is of the nt- phenotype. wild-type glucocorticoid sensitivity was dominant over nt- resistance, and S49.1 wt X EL4 hybrids contained glucocorticoid receptors of both parental cell types. The r- and nt- phenotypes did not complement, suggesting that both defects occurred in the same polypeptide. S49.1 wt X EL4 hybrids produced resistant cells at 3 XC 10(-4) per cell per generation. Sensitive hybrid cell clones were treated with high glucocorticoid doses to select for resistant segregants. The majority of these segregants had lost one chromosome. The concomitant loss of the wild-type S49.1 receptor was observed upon transition from sensitivity to resistance. These results provide direct evidence for the hemizygosity of the glucocorticoid receptor gene in S49.1 cells. Thus, the cell system described here is suitable for assigning the glucocorticoid receptor gene to a specific mouse chromosome.

Glucocorticoid-induced lymphocytolysis: state of the genetic analysis

The glucocorticoid-induced lysis of lymphoid cell lines offers a genetic approach to steroid hormone action because unresponsive variants can easily be selected as resistant to this lytic effect. The present state of analysis of lymphocytolysis in two murine cell lines, the S49 T-lymphoma and the W7 thymoma, is reviewed. All glucocorticoid-resistant variants isolated so far result from various defects in the glucocorticoid receptor. The absence of variants blocked at another step of the lytic mechanism is discussed. The observed hemizygosity of the glucocorticoid receptor locus in the S49 line and the instability of cell hybrids illustrate some of the potential problems encountered in somatic cell genetics.