Suppression of tumorigenicity in somatic cell hybrids. IV. Chromosomes of normal human cells associated with suppression of tumorigenicity in hybrids with D98AH2 carcinoma cells

An analysis for cosegregation of chromosomes and tumorigenicity in 52 hybrids of human diploid X D98AH2 human carcinoma-derived cells reveals the consistent presence of four copies of chromosome 11 in all nontumorigenic hybrids (two from each of the parental cells) and a consistent loss of one or two copies of the 11 in all tumor cells derived from tumorigenic hybrids that grow in nude mice. In our earlier study, assays with restriction fragment length polymorphic (RFLP) markers for the cell parent origin of the chromosomes 11 in the hybrids indicated that at least one of the Nos. 11 lost in the tumor cells is from the diploid. Thus both Nos. 11 of the diploid seem to be required for complete and stable suppression of the tumorigenic phenotype. The results of the present study suggest that chromosome 2 may also carry suppressor information, but this causes only partial suppression of the tumorigenic phenotype in the absence of both Nos. 11. On the other hand, when the hybrids contain full complements of the 2 and the 11, suppression is very stable. All other chromosomes except for Nos. 1, 16, 17, 19, and 21 are clearly discordant with suppression. The latter chromosomes are not discordant often enough to allow their exclusion as possible carriers of suppressor information, particularly in the absence of RFLP evaluations. It is clear, however, that if they do carry such information it is not adequate for maintaining a stably suppressed phenotype in the absence of both Nos. 11 of the diploid.     

Patterns of chromosome segregation in chinese hamster × mouse cell hybrids between permanent cell lines and thymus cells

Hybrids between a tumorigenic Chinese hamster cell line (DC3F-aza) and normal mouse thymus cells very rapidly lost most of their mouse chromosomes, whereas hybrids between tumorigenic mouse cell lines (either Cl.1D of L cell line origin, or PCC4-aza1 teratocarcinoma cells) and normal Chinese hamster thymus cells lost most of their hamster chromosomes. From three such fusion experiments, 20 cell lines were developed which all followed the same evolution, namely, the elimination of the majority of the chromosomes contributed by the normal thymus cell. In some hybrids, the elimination process resulted in the total absence of intact chromosomes contributed by the thymus cell parent. Such hybrids were distinguished from revertant parental cells growing in the selective hybrids were distinguished from revertant parental cells growing in the selective medium by the presence of at least one enzyme in their cell extracts which displayed the electrophoretic mobility of the enzyme of the thymus cell parent. These observations, together with data from other reports, suggest that, as a rule, interspecific cell hybrids which develop upon fusion between normal diploid cells and tumorigenic cell lines maintain the chromosomes of the latter and eliminate preferentially many or most of the chromosomes contributed by the normal cell parents, independent of the respective species of the parental cells.     

DNA polymorphisms indicate loss of heterozygosity for chromosome 11 of D98AH2 cells

Studies with cell hybrids of normal diploid cells fused with tumorigenic D98AH2 (D98) cells had implicated human chromosome 11 of a normal cell as carrying tumorigenicity suppressing information. The cervical carcinoma-derived D98 (HeLa) cells contain two copies of chromosome 11. In this study, analysis of restriction fragment length polymorphism of DNA from D98 cells digested with one of nine restriction endonucleases and hybridized with five DNA probes for highly polymorphic regions on the short arm of chromosome 11 detected no heterozygosity at the insulin (INS), Harvey murine sarcoma virus 1 (HRAS1), and the beta-globin cluster (HBBC) regions. The low probability of an individual being homozygous at all these loci suggests that the Nos. 11 of the D98 cells are both copies of only one of the original homologs, or at least of the short arm segment examined. This indicates that the D98 cells could express altered or lost genes associated with tumorigenicity, even if such changes were recessive. In tumorigenically suppressed hybrids the Nos. 11 of the normal cell could then be complementing this genetic defect of the D98 cells.     

Localization of HeLa cell tumor-suppressor gene to the long arm of chromosome II.

Cytogenetic and molecular genetic analyses of human intraspecific HeLa x fibroblast hybrids have provided evidence for the presence of a tumor-suppressor gene(s) on chromosome 11 of normal cells. In the present study, we have carried out extensive RFLP analysis of various nontumorigenic and tumorigenic hybrids with at least 50 different chromosome 11-specific probes to determine the precise location of this tumor-suppressor gene(s). Two different hybrid systems, (1) microcell hybrids derived by the transfer of a normal chromosome 11 into a tumorigenic HeLa-derived hybrid cell and (2) somatic cell hybrids derived by the fusion of the HeLa (D98OR) cells to a retinoblastoma (Y79) cell line, were particularly informative. The analysis showed that all but one of the nontumorigenic hybrid cell lines contained a complete copy of the normal chromosome 11. This variant hybrid contained a segment of the long arm but had lost the entire short arm of the chromosome. The tumorigenic microcell and somatic cell hybrids had retained the short arm of the chromosome but had lost at least the q13-23 region of the chromosome. Thus, these results showed a perfect correlation between the presence of the long arm of chromosome 11 and the suppression of the tumorigenic phenotype. We conclude therefore that the gene(s) involved in the suppression of the HeLa cell tumors is localized to the long arm (q arm) of chromosome 11.     

Phenotype and surface antigens of mouse teratocarcinoma x fibroblast cell hybrids

Numerous colonies of hybrids between PCC4-aza 1 teratocarcinoma cells and fibroblasts of the heteroploid Cl.1D cell line were examined. All of the hybrids were fibroblasts showing extinction of the multiple developmental potentialities of the teratocarcinoma cell parent, irrespective of whether the teratocarcinoma parent was diploid or tetraploid. The hybrids did not show loss of any specific chromosome contributed by the PCC4-aza 1 cell parent. In contrast with the PCC4 parental cells which carry F9 antigens and do not express H-2b, the hybrids do not express F9 antigens and carry H-2 alloantigens of both parental specificities. These results suggest that in hybrids whose phenotype is that of the Cl.1D parent, a change may occur in the genetic program of the teratocarcinoma cells.     

Suppression of the transformed phenotype of hepatoma cells after hybridization with normal diploid fibroblasts

Hybrids were generated between mouse hepatoma cells which exhibit a transformed phenotype, and rat normal diploid fibroblasts. Most isolated hybrid clones contain a single set of chromosomes from each parent. Such clones grow to low saturation densities and are unable to grow or to form colonies in soft agar. The transformed phenotype of the parental hepatoma cells is thus suppressed in these hybrids. Suppression is very stable; however, subclones which have regained a transformed phenotype could be selected; these subclones show a significant reduction of their chromosome number. Amongst the hybrid clones isolated after fusion, a few are characterized by an excess of mouse chromosomes and a reduced number of rat chromosomes. Such clones exhibit a transformed phenotype. Our results show that, provided the hybrids contain an almost complete single set of chromosomes of each parent, spontaneous transformation behaves as a recessive trait in hybrids formed with normal diploid cells.     

Genetic control of tumorigenicity in interspecific mammalian cell hybrids.

The nature of genetic control of cellular malignancy was investigated by examining the tumorigenicity of a series of interspecific mouse-human cell hybrids in the athymic nude mouse. Two highly malignant but genetically distinct mouse cell lines, A9 and PG19, were hybridized with three normal human diploid fibroblast strains, and 19 independently arising hybrid clones were isolated. Each of these clones was capable of forming progressive lethal tumors in the nude mouse, and thus resembled the malignant parental mouse cells rather than the nonmalignant parental human cells. We failed to obtain any evidence for complete suppression of tumorigenicity in these cell hybrids. The absence of suppression was observed regardless of the extent and composition of the human chromosome complements retained in the hybrid clones; the results of detailed cytological and isoenzyme analyses would make it highly improbable that the observed lack of suppression was due to cellular selection in vivo for a more tumorigenic subpopulation in the injected hybrid cells. These data demonstrate that at least for the parental cell combinations used in this study, no human chromosome, when present singly in the mouse-human cell hybrids, can suppress the tumorigenic phenotype of the mouse cells. Our results are consistent with the view that the suppression of cellular malignancy previously demonstrated in intraspecific (mouse × mouse) somatic cell hybrids does not occur in interspecific (mouse-human) cell hybrids, or alternatively, genetic determinants located on two or more human chromosomes are required simultaneously to suppress the malignancy of the mouse cells in cell hybrids derived from malignant mouse cell and nonmalignant human cells.     

Genetic effects of chromosomal rearrangements in Chinese hamster ovary cells: expression and chromosomal assignment of TK, GALK, ACP1, ADA, and ITPA loci.

Polyethylene glycol-mediated fusion of Chinese hamster ovary (CHO) cells with mouse Cl1D cells produced interspecific somatic cell hybrids which slowly segregated CHO chromosomes. Cytogenetic and isozyme analysis of HAT- and bromodeoxyuridine-selected hybrid subclones and of members of a hybrid clone panel retaining different combinations of CHO chromosomes enabled provisional assignments of the following enzyme loci to CHO chromosomes: TK, GALK, and ACP1 to chromosome 7; TK and GALK to chromosome Z13; ACP1, ADA, and ITPA to chromosome Z8; and ADA and ITPA to chromosome Z9. These genetic markers reflect the origin of each of these Z group chromosomes and indicate the functional activity of alleles located on rearranged chromosomes. Identification of diploid electrophoretic shift mutations for ADA and ITPA was consistent with those observations. Assignment of the functional TK locus in TK+/- CHO-AT3-2 cells indicated that gene deletion may be responsible for TK hemizygosity in this subline.     

The Ha-ras-induced transformed phenotype of rat-1 cells can be suppressed in hybrids with rat embryonic fibroblasts

Somatic cell hybrids were isolated from fusions of diploid embryonic rat fibroblasts with transformed Rat-1 cells which contained 4 to 5 copies of the transforming human Ha-ras 1 gene. In contrast to their transformed parental cells four hybrid clones showed normal morphology, long latency periods of tumorigenicity in newborn rats, anchorage requirement of proliferation, and an eightfold-reduced amount of secreted transforming growth factor activity. Thus these hybrids are called suppressed with regard to expression of the Ha-ras-induced transformed phenotype. Tumorigenic derivatives of the suppressed hybrids that had segregated chromosomes were isolated. Since two of the tumorigenic hybrid clones showed the similar low level of secreted transforming growth factors as the suppressed hybrids, decreased production of transforming growth factor activity is unlikely to be a sufficient criterion for suppression of malignancy. Whereas one of the suppressed hybrids expressed the transforming gene product p21 at a level similar to that of the transformed parental cells, other suppressed hybrids expressed less p21. This suggests that the suppressed phenotype can be regulated at the posttranslational level of p21 but that additional controls of expression of p21 are likely to exist. DNA of the suppressed hybrids transformed Rat-1 cells to proliferation in the presence of semisolid agar. Thus the activated human Ha-ras gene in the suppressed hybrids retained its biological activity even though it did not transform these cells to tumorigenicity.     

Genetic analysis of tumorigenesis: IV. Chromosome reduction and marker segregation in progeny clones from Chinese hamster cell hybrids.

Hybrid cells produced by the fusion of pairs of cells, one a tumorigenic derivative of CHEF/16 and the other a nontumorigenic derivative of CHEF/18, give rise to clones which are largely tetraploid, but rare reduced hybrids with chromosome counts in the diploid range have been recovered from tumors of hybrid origin. This paper describes the recovery in cell culture of reduced hybrids in the diploid range by selection with 5-bromodeoxyuridine (BrdU) or methylcellulose as well as by growth in culture of cells from excised tumors. All selected subclones were tumorigenic and resistant to BrdU, but they segregated for resistance to 6-thioguanine. Unselected subclones were tetraploid, nontumorigenic, and sensitive to both drugs. These data show that chromosome reassortment as well as extensive chromosome reduction both occur in a small fraction of the population during growth of each hybrid clone.     

Activation of human alpha 1-antitrypsin gene in rat hepatoma x human fetal liver cell hybrids depends on presence of human chromosome 14

In order to study the involvement of human chromosomes in the expression of liver-specific functions, we have produced somatic cell hybrids between a rat hepatoma (7777) cell line and human diploid skin fibroblasts (series XIX) or human fetal liver cells (series XXII). Production of human serum proteins was detected by immunoelectrophoretic analyses of concentrated serum-free hybrid culture supernatants. Human alpha 1-antitrypsin (AAT) was secreted by a subset of hybrids but not by the parental cells. The activated human AAT phenotype segregated concordantly with human chromosome 14 in 18 primarily HAT-selected and five azaguanine back-selected series XXII hybrids. All other chromosomes were excluded as playing a role in AAT expression. Therefore, the AAT gene (PI) is assigned to chromosome 14. This quasi-constitutive expression of a liver-specific function was not observed for the other serum proteins studied, nor was it seen in the skin fibroblast-derived hybrids (series XIX) although AAT was produced by some of them.     

Human intraspecific somatic cell hybrids: a genetic and karyotypic analysis of crosses between lymphocytes and D98/AH-2.

A number of human intraspecific hybrids were produced by fusing the 8-azaguanine-resistant cell line D98/AH-2 with PHA-stimulated lymphocytes from a normal human male, followed by selection in HAT medium. The parent cells differed in zymogram patterns for 4 enzyme systems. Hypoxanthine-guanine phophoribosyltransferase was missing in D98/AH-2 and was determined in the hybrids by the normal gene derived from the lymphocyte donor's X chromosome. The HL-A antigens of the lymphocyte donor as well as the W28 specificity from HeLa were easily recognized by a cytotoxicity assay on the hybrid cells, while D98/AH-2 itself was not killed in the normal way by any HL-4 typing sera. The initial hybrid karyotype in all lines was relatively stable, but slow loss of chromosomes occurred following extended growth in culture. The importance of the culture conditions for the rate of chromosome loss was demonstrated. The behavior of several chromosomes was followed in the hybrids and their derivatives. There was relatively nonspecific loss of small numbers of chromosomes, showing that loss of chromosomes from both the D98/AH-2 and the normal lymphocyte parent can occur. Cell lines resistant to 6-thioguanine were selected from the sensitive hybrids. Most had lost the lymphocyte donor's X chromosome, thereby losing the only active allele for HGPRT present in the initial hybrids. However, one line, DMR41, apparently retained the X chromosome and may have a mutated allele for HGPRT. Two lines that are the products of spontaneous segregation are also described. DM4CS and DM17A.     

Chromosome segregation from cell hybrids. II. Do differences in parental cell growth rates and phase times determine direction of loss?

The hypothesis that the direction of chromosome segregation in cell hybrids is determined by the interaction of parent cell cycles, or S-phase times, predicts that the segregant parent will always be the one with the longer cycle, or the longer S phase, and that late replicating chromosomes will be more frequently lost. We have tested this hypothesis by studying cell cycle parameters of mouse, Chinese hamster, and platypus parent cells and by observing chromosome loss and replication patterns in hybrids between them. Two types of hybrids have been studied: mouse-hamster hybrids showed gradual segregation, in one or other direction, of 10-60% chromosomes, while rodent-platypus hybrids (which could be selected under conditions optimal for either parent cell) showed rapid and extreme segregation of platypus chromosomes. We found no correlation between the direction of segregation and the relative lengths of parental cycle times, or phase times, nor between sequence of replication and frequency with which segregant chromosomes are lost. We therefore conclude that the direction and extent of segregation is not directly determined by the interaction of parental cycle or phase times.     

In vivo hybridisation of cultured melanoma cells and isogenic normal mouse cells.

Somatic cell hybrids were derived by fusing tumourigenic and melanogenic melanoma (PAZG) cells with normal diploid male mouse cells in vivo. Their chromosomal composition was equivalent to the sum of both parental genomes and included a Y chromosome lacking in the melanoma parent. Our study showed that in PAZG X C57BL hybrids (MP), tumourigenicity was suppressed but pigmentation was expressed.     

Correlation of the inability to sustain growth in defined serum-free medium with the suppression of tumorigenicity in Wilms' nephroblastoma.

We report the investigation of the growth properties of tumorigenic and reverted nontumorigenic Wilms' nephroblastoma cells when cultured in serum-free medium. Wilms' tumor, a pediatric nephroblastoma, has been associated with deletions encompassing the p13 band of chromosome 11 and an independent loss of heterozygosity at 11p15. Weissman et al. (Science 236:175-180, 1987) transferred a human der(11) chromosome into the G401.6TG.6 Wilms' tumor cell line via the microcell-mediated chromosome transfer technique. The resulting microcell hybrids were nontumorigenic when assayed in nude mice; however these cells retained all of the in vitro growth and morphological characteristics of the tumorigenic parental cells in 10% fetal calf serum (FCS). Segregation of the der(11) chromosome from the nontumorigenic microcell hybrid cells resulted in the reappearance of the tumorigenic phenotype in vivo. In vitro culture of these cell lines in serum-free medium supplemented with 0.1% bovine serum albumin (BSA) and 10 ng/ml Na2O3Se resulted in sustained growth of both the tumorigenic parent and the tumorigenic segregant while the nontumorigenic microcell hybrids were unable to divide. The separate addition of either 10 ng/ml of epidermal growth factor (EGF) or 5 micrograms/ml of insulin did not alter this effect. However, the addition of 5 micrograms/ml of transferrin stimulated the nontumorigenic microcell hybrid cells to grow at a rate comparable to the tumorigenic cells. In addition, conditioned serum-free medium from the tumorigenic parental or tumorigenic segregant cell lines was able to stimulate the growth of the nontumorigenic microcell hybrid cells, whereas the reciprocal experiment had no effect on the growth of the tumorigenic cells. These data suggest that the inability of the microcell hybrid cells to grow in serum-free conditions is correlated with their genetic nontumorigenic phenotype and that a specific growth factor, transferrin, can bypass or alter this negative growth regulatory pathway(s) in vitro.     

Suppression of tumorigenicity,but not invasion,in glioblastoma/HeLa cell hybrids

Somatic cell hybrids between SNB-19 human glioblastoma cells and human D98^OR HeLa parental cells were produced and analyzed for their ability to form tumors in nude mice and to invade reconstituted extracellular matrix (Matrigel). Whereas both the SNB-19 and D98^OR HeLa parental cells form tumors, four of six hybrid lines did not form tumors, even after periods up to six months, suggesting that each cell type can complement the tumorigenicity of the other. SNB-19 cells showed high rates of Matrigel invasion at all cell densities examined, whereas D98^OR HeLa cells showed lower rates of invasion that were further reduced at high cell density. All six hybrid cell lines displayed a combination of these properties: at low cell density, the hybrids showed high rates of invasion, similar to the SNB-19 cells, but the invasion rate diminished at higher cell densities, similar to the D98^OR HeLa cells. Taken together, these results provide new experimental evidence that several distinct genetic changes are involved in generating the tumorigenic and invasive phenotype of glioblastoma cells. © 1995 Wiley-Liss, Inc.     

Identification of a human chromosome 11 gene which is differentially regulated in tumorigenic and nontumorigenic somatic cell hybrids of HeLa cells.

The tumorigenicity of HeLa cells in nude mice can be suppressed by the addition of a normal human chromosome 11 in somatic cell hybrids. We have attempted to identify specific genes involved in this phenomenon by transfecting a complementary DNA expression library into a tumorigenic HeLa-fibroblast hybrid. A cell line designated F2 was isolated which displayed morphological features of the nontumorigenic hybrids, demonstrated reduced tumorigenicity in nude mice, and showed an 85% reduction in alkaline phosphatase, a consistent marker of the tumorigenic phenotype in these cells. F2 contained a single exogenous complementary DNA, which was recovered by polymerase chain reaction and designated HTS1 because of its potential association with "HeLa tumor suppression." Northern blot studies suggested differential regulation of the HTS1 gene dependent on the tumorigenicity of the cell. In nontumorigenic hybrids, RNA species of 2.8, 3.1, and 4.6 kilobases were identified. In two tumorigenic hybrid lines, the 2.8-kilobase species was markedly reduced or absent. Similarly, three nontumorigenic human keratinocyte lines expressed all three RNA species, whereas several tumorigenic cervical carcinoma cell lines lacked the 2.8-kilobase species. Chromosome localization studies mapped the HTS1 gene to chromosome 11p15, a region of chromosome 11 that is believed to contain a tumor suppressor gene. These findings indicate that HTS1 represents a novel chromosome 11 gene which may be a target of the tumor suppressor gene active in this system.     

Introduction of human chromosome 11 via microcell transfer controls tumorigenic expression of HeLa cells.

Both tumorigenic segregant HeLa X human fibroblast hybrids and tumorigenic HeLa (D98/AH-2) cells can be converted to a non-tumorigenic state following introduction of a single copy of a fibroblast t(X;11) chromosome. The translocated chromosome contains approximately 95% of the 11 chromosome and the q26-qter portion of the X chromosome which contains the hypoxanthine guanine phosphoribosyl transferase (HPRT) gene. Introduction of a human X chromosome has no effect on tumorigenic expression. Suppression of tumorigenicity is relieved by selecting cells which have lost the t(X;11) chromosome by growth in medium containing 6-thioguanine (6-TG). Further, reintroduction of the t(X;11) chromosome into tumorigenic 6TGR cells again suppresses tumorigenicity. Thus, the introduction of a single copy of a human chromosome 11 is sufficient to completely suppress the tumorigenic phenotype of HeLa cells and is suggestive of the presence of tumor-suppressor gene(s) on this chromosome.     

A new genetic locus, Bevi, on human chromosome 6 which controls the replication of baboon type C virus in human cells.

Somatic cell hybrids derived from seven independent fusions between mouse X human and hamster X human parental cells were examined for their ability to support the replication of the baboon endogenous type C virus. These hybrids preferentially segregated human chromosomes while retaining rodent chromosomes, as demonstrated by karyotypic and isozyme analysis. A total of 41 primary colonies and 33 secondary subclones were analyzed for viral replication, as well as for the presence of enzyme structural gene markers for 19 of 23 human chromosomes. A syntenic association was seen between the ability of the baboon type C virus to infect and replicate in hybrid cultures and the expression of human malic enzyme-1 (assigned to human chromosome 6). Analysis of 86 highly segregated subclones derived from cells preinfected with baboon type C virus showed that the continued production of baboon type C virus segregated concordantly with the expression of three enzyme genes assigned to human chromosome 6 (malic enzyme-1, phosphoglucomutase-3 and superoxide dismutase-2). Subclones of infected hybrids which lost chromosome 6 and failed to release virus also failed to synthesize the virus-coded major structural protein p30. No syntenic association between baboon virus expression and any of 18 other human chromosomes was observed. These studies define a new gene (designated Bevi) on human chromosome 6 which dominantly controls the replication of baboon type C virus. The data suggest that Bevi may be a preferred integration site for the baboon type C DNA provirus in the human genome.