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.     

In vitro traits of adenovirus-transformed cell lines and their relevance to tumorigenicity in nude mice

Six independently isolated adenovirus 2-transformed rat cell lines and one adenovirus 5-transformed human cell line have been examined in vitro for serum growth requirements, saturation density, anchorage-independent growth, proteolytic enzyme activity and the presence of LETS glycoprotein and T antigen. This series of adenovirus-transformed cell lines exhibits an oncogenic spectrum ranging from being tumorigenic in immunocompetent rats through to nontumorigenic in adult nude mice. The relevance of the in vitro findings to growth potential in vivo is discussed.     

Tumor formation by SV40-transformed human cells in nude mice: the role of SV40 T antigens

Human cells transformed in vitro by SV40 rarely form tumors in nude mice. We examined whether these cells as a group are inherently nontumorigenic or whether they are potentially tumorigenic but rejected by the athymic host, possibly by nonspecific immune mechanisms. SV80 and NG8 are SV40-transformed human cell lines that express all of the transformed properties, including anchorage-independent growth, but do not form tumors in adult nude mice after injection of as many as 10(8) cells. Both the SV80 and NG8 cell lines have SV40-specific transplantation antigens which crossreact with those present on SV40-transformed (but tumorigenic) rodent cells. We found that SV80 cells, though not NG8 cells, induced progressively growing lethal tumors if the cells are injected repeatedly into neonatal nude mice. Somatic cell hybrids between SV80 or NG8 cells and a highly tumorigenic cell line derived from a human tumor continue to express the virus-induced antigens and fail to form tumors in adult nude mice. These results strongly suggest that at least for some SV40-transformed human cells, the failure to form tumors in nude mice may be due to their expression of virus-induced transplantation antigens rather than the absence of tumorigenic potential.     

Nuclear control of tumorigenicity in cells reconstructed by PEG-induced fusion of cell fragments

The techniques of somatic cell hybridization have provided a valuable means of studying mechanisms of regulation of mammalian cell differentiation and transformation. Most previous studies have indicated that fusions between tumorigenic and nontumorigenic cells result in hybrid cells that are usually tumorigenic. In recent years it has been demonstrated that the phenotypic expression of tumorigenicity is at least partially due to the extensive chromosome loss that occurs in most interspecific and some intraspecific hybrid cells. In the present study we have utilized enucleation techniques that permit cells to be divided into nuclear (karyoplast) and cytoplasmic (cytoplast) cell fragments. Even though these nuclear and cytoplasmic fragments are metabolically stable for short periods of time, in our hands they ultimately degenerate. Viable cells can be reconstructed by PEG-induced fusion of karyoplasts to cytoplasts. Since reconstructed cells apparently do not segregate chromosomes, they may provide a clearer understanding of the interactions between the nucleus and the cytoplasm in the control of the expression of tumorigenicity. We have reconstructed cells using karyoplasts from the tumorigenic Y-1 cell line and cytoplasts from a nontumorigenic cell line, A-MT-BU-A1. In addition we have reconstructed cells containing Y-1 cytoplasts and A-MT-BU-A1 karyoplasts. The reconstructed cells porduced were assayed for tumorigenicity by their ability to grow in soft agar and in nude mice. The results of these experiments indicate that the reconstructed cells containing a tumorigenic nucleus and a nontumorigenic cytoplasm ultimately are tumorigenic and conversely the reconstructed cells containing a nontumorigenic nucleus and a tumorigenic cytoplasm are nontumorigenic. These experiments support the concept that with these cell lines the nucleus (karyoplast) is sufficient to control the phenotypic expression of tumorigenicity.     

Evaluation of mitochondrial function and metabolic reprogramming during tumor progression in a cell model of skin carcinogenesis

Metabolic reprogramming from mitochondrial aerobic respiration to aerobic glycolysis is a hallmark of cancer. However, whether it is caused by a dysfunction in the oxidative phosphorylation pathway is still under debate. In this work, we have analyzed the bioenergetic cellular (BEC) index and the relative cell ability to grow in the presence of either galactose or glucose as sources of sugar (Gal/Glu index) of a system formed by four epidermal cell lines with increasing tumorigenic potentials, ranging from nontumorigenic to highly malignant. We find that the BEC index gradually decreases whereas the Gal/Glu index increases with tumorigenicity, indicating that a progressive metabolic adaptation to aerobic glycolysis occurs in tumor cells associated with malignancy. Interestingly, this metabolic adaptation does not appear to be caused by damaged respiration, since the expression and activity of components of the respiratory chain complexes were unchanged in the cell lines. Moreover, the corresponding mitochondrial ATP synthetic abilities of the cell lines were found similar. The production of reactive oxygen species was also measured. A shift in ROS generation was found when compared nontumorigenic with tumorigenic cell lines, the latter exhibiting about threefold higher ROS levels than nontumorigenic cells. This result indicates that oxidative stress is an early event during tumor progression.     

The use of agarose in the determination of anchorage-independent growth.

At the present time, growth in agar suspension is one of the best in vitro correlates of tumorigenicity. Growth in agarose, however, has not been evaluated extensively as an in vitro criterion for tumorigenicity. In the present study we have tested 19 cell lines, including six mouse-human hybrids, for growth in agarose and agar in the presence and absence of exogenous hypoxanthine. None of the six nontumorigenic cell lines grew in agar or agarose. Ten of the 13 tumorigenic cell lines grew in both agar and agarose with about equal efficiency. The remaining three tumorigenic cell lines grew well in agarose but poorly or not at all in agar. Hypoxanthine did not stimulate the growth in agar or agarose of any of the cell lines except BHK. We conclude that growth in agarose may be a more sensitive marker for tumorigenicity than growth in agar and that BHK is exceptional in its sensitivity to supplemental purines.     

The use of agarose in the determination of anchorage-independent growth

Summary At the present time, growth in agar suspension is one of the best in vitro correlates of tumorigenicity. Growth in agarose, however, has not been evaluated extensively as an in vitro criterion for tumorigenicity. In the present study we have tested 19 cell lines, including six mouse-human hybrids, for growth in agarose and agar in the presence and absence of exogenous hypoxanthine. None of the six nontumorigenic cell lines grew in agar or agarose. Ten of the 13 tumorigenic cell lines grew in both agar and agarose with about equal efficiency. The remaining three tumorigenic cell lines grew well in agarose but poorly or not at all in agar. Hypoxanthine did not stimulate the growth in agar or agarose of any of the cell lines except BHK. We conclude that growth in agarose may be a more sensitive marker for tumorigenicity than growth in agar and that BHK is exceptional in its sensitivity to supplemental purines. This research was supported by DHEW NCI Contract No. NO-CP-2-3234 and by USPHS Medical Scientist Training Grant GM 02042-07.     

The conversion of mouse skin squamous cell carcinomas to spindle cell carcinomas is a recessive event

Squamous carcinomas of both human and rodent origin can undergo a transition to a more invasive, metastatic phenotype involving reorganization of the cytoskeleton, loss of cell adhesion molecules such as E-cadherin and acquisition of a fibroblastoid or spindle cell morphology. We have developed a series of cell lines from mouse skin tumors which represent different stages of carcinogenesis, including benign papillomas, and clonally related squamous and spindle carcinomas derived from the same primary tumor. Some spindle cells continue to express keratins, but with a poorly organized keratin filament network, whereas in others no keratin expression is detectable. All of the spindle cells lack expression of the cell adhesion molecule E-cadherin and the desmosomal component desmoplakin. Loss of these cell surface proteins therefore appears to precede the destabilization of the keratin network. At the genetic level, it is not known whether such changes involve activation of dominantly acting oncogenes or loss of a suppressor function which controls epithelial differentiation. To examine this question, we have carried out a series of fusion experiments between a highly malignant mouse skin spindle cell carcinoma and cell lines derived from premalignant or malignant mouse skin tumors, including both squamous and spindle carcinoma variants. The results show that the spindle cell phenotype as determined by cell morphology and lack of expression of keratin, E-cadherin, and desmoplakin proteins, is recessive in all hybrids with squamous cells. The hybrids expressed all of these differentiation markers, and showed suppression of tumorigenicity to a variable level dependent upon the tumorigenic properties of the less malignant fusion partner. Our results suggest that acquisition of the spindle cell phenotype involves functional loss of a gene(s) which controls epithelial differentiation.     

Complex expression of the genes coding for plasminogen activators and their inhibitors in HeLa-smooth muscle cell hybrids.

As cells progress through the multistep process of neoplastic transformation, they eventually acquire the property of invasive behavior. Although both plasminogen activators (PA) and their inhibitors (PAI) contribute to this process, their regulation in normal and transformed cells remains poorly defined. Because somatic cell hybrids provide useful tools for examining the transformation pathway, tumorigenic and invasive HeLa cells were fused with human normal vascular smooth muscle cells and tested for invasion-related parameters, including the expression of PA and PAI genes, and matrix degradation. Both parental cell lines produced large amounts of PAI activities with no detectable PA in either cellular or secreted form. Opposite findings were obtained with the hybrid cell lines, which demonstrated the presence of receptor-bound and secreted PA but absence of enzymatically measurable PAI activities. Both urokinase-type and tissue-type PA were found in cell-associated and secreted form in the hybrid cells. In addition, expression of the urokinase-type PA receptor gene was found in the three hybrid cells and the vascular smooth muscle cells but not in the HeLa cells. Expression of active, receptor-bound and secreted PA provided the nontumorigenic hybrid cells with the enzymatic tools to degrade extracellular proteins in a plasminogen-dependent manner. Thus, the hybrid cells lost tumorigenicity while retaining the tissue-degrading capability of HeLa cells. These hybrid cell lines should prove to be important reagents for investigating the complex regulatory control of PA and PAI gene expression.     

Clonal sublines of rat neurotumor RT4 and cell differentiation: II. A conversion coupling of tumorigenicity and a glial property

RT4 is a neurotumor induced by ethylnitrosourea injection of a newborn BDIX rat. We demonstrated previously that heterogeneity in early cultures of RT4 tumor cells can be regularly reproduced in cultures of clonal stem cells (cell type conversion). Our previous studies included morphology, differentiation of neural properties, and chromosome number of “tumor-derived” and “stem cell-derived” differentiated cells. In this paper, these two sets of differentiated cells were examined further for three additional parameters, all of which are related to malignancy. The stem cell (AC) and one type of differentiated cell (D) cause tumors when subcutaneously injected into syngeneic animals, while the other two types (B and E) do not. The amounts of a 250,000 molecular weight cell surface protein, which is probably equivalent to LETS protein (large external transformation-sensitive protein) of hamster and mouse, and the levels of plasminogen activator were examined as possible markers of malignancy. As anticipated, nontumorigenic cells generally have a large amount of the 250,000 molecular weight cell surface protein and are low in plasminogen activator activities, whereas the reverse is true for tumorigenic cells. This supports the idea that B and E cells are nontumorigenic revertants. The cell type conversion phenomenon of RT4 neurotumor and the differentiation of mouse teratoma and myeloid leukemic cells share a number of similarities, but differ in that differentiated RT4 cells can propagate in vitro even after loss of tumorigenicity. The concomitant expression of tumorigenicity and the S100 protein production of the D cell, or of nontumorigenicity and B and E cell differentiation upon the conversion of the stem cell, may suggest a regulational coupling between the tumorigenicity and the expression of a glial protein (S100 protein) in D cells.     

Effects of reduced calcium on proliferation and cell viability in tumorigenic and nontumorigenic rat tracheal epithelial cell lines

We have compared the in vitro growth and viability of tumorigenic and nontumorigenic rat tracheal epithelial cell lines over a range of calcium concentrations from 0.003 to 0.85 mM. A greater dependence on calcium for proliferation was seen in the nontumorigenic line as compared to the tumorigenic line at both the colony formation level and in mass cultures. In the latter culture condition, a marked differential effect on cell survival was also demonstrated. These differences in calcium dependence were seen in media containing fetal bovine serum or low concentrations of newborn calf serum and in a serum-free medium developed for these cells. The effect was also independent of the method used for calcium removal i.e., either by chelex treatment or the inclusion of EGTA. Therefore, loss of calcium dependence may be associated with tumorigenicity in rat tracheal epithelial cells offering a selectable marker for neoplastic cells in carcinogen-exposed preneoplastic cell populations.     

Levels of fos, ets2, and myb proto-oncogene RNAs correlate with segregation of chromosome 11 of normal cells and with suppression of tumorigenicity in human cell hybrids.

The tumorigenicity in nude mice of human carcinoma-derived D98AH2 (D98) cells is suppressed when cell hybrids are made by fusing these cells with normal human diploid cells. Selection for hybrids that have segregated chromosomes results in the recovery of tumorigenic segregants. These segregants have all lost at least one copy of chromosome 11 of the diploid cell parent. Earlier we found that the parental D98 cells had detectable levels of mRNA specific for 13 of 21 proto-oncogenes examined. To determine if transregulation of proto-oncogenes by genes of the normal cell occurs in such hybrids, the steady-state levels of mRNA specific to 22 proto-oncogenes in the parental cells were compared with those of nontumorigenic D98 X human diploid hybrids as well as with those of their tumorigenic segregants and with the cells of the resulting tumors. The only chromosome consistently segregated in the latter was chromosome 11 of the diploid cell. fos and ets2 RNA levels and the amount of fos protein were consistently elevated in the segregants compared with amounts in the original hybrids. An unexpected finding was the inverse relationship for myb RNA that was barely detected in the parental D98 cells but was at least 10-fold elevated in hybrids that did not have segregated chromosomes compared with those that did. These patterns were evident in RNAs prepared from both subconfluent and confluent cell cultures. The findings suggest that genes of the normal cell parent can affect proto-oncogene expression. Whether the genes affecting fos, ets2, and myb RNA levels are on chromosome 11 and whether these alterations are causally related to the tumorigenic phenotype of the hybrid remain to be determined.     

Angiogenesis-inducing ability of human bladder epithelium cell lines and "spontaneously" transformed murine fibroblasts

Ten human bladder epithelium cell lines were tested for their ability to induce blood vessel formation after intradermal injection into irradiated ST/a mice. Cell lines that were shown to be tumorigenic in nude mice, were able to evoke angiogenesis of a higher intensity than nontumorigenic cell lines. No difference was observed between the angiogeneic ability of tumorigenic cells originating from tumors and from in vitro transformed urothelium of nontumor origin. Similarly the origin of nontumorigenic urothelial cell lines did not show any influence on their angiogeneic abilities, but nontumorigenic cell lines which had undergone "infinite growth transformation" exhibited a higher angiogeneic activity than nontumorigenic cell lines with a finite life. The angiogeneic reaction evoked by human bladder epithelium cell lines showed cell dose- and time-dependence; but it was unrelated to the growth potential of the cultured cells. Two "spontaneously" altered sarcoma-producing murine cell lines showed a higher angiogeneic activity than tumorigenic human bladder epithelial cells. The angiogeneic response to these two murine cell lines was unrelated to morphological signs of transformation and to differences in growth rate, serum requirement, saturation density, anchorage dependence, and isoimmunizing properties.     

Enhanced expression of glucose transporter GLUT3 in tumorigenic HeLa cell hybrids associated with tumor suppressor dysfunction.

Previous studies on human cell hybrids between HeLa and normal human fibroblasts have indicated that the tumorigenicy may be controlled by a putative tumor suppressor gene on chromosome 11. We previously demonstrated a twofold increase in glucose uptake with a reduced Km by tumorigenic HeLa cell hybrids which expressed a highly glycosylated GLUT1. In this study, we reported that a tumorigenic cell hybrid, CGL4, also expressed a glucose transporter isoform, GLUT3, that was undetectable in nontumorigenic CGL1 cells. The expression of GLUT3 together with GLUT1 of 70 kDa was also evident in three gamma-ray-induced tumorigenic clones isolated from CGL1 cells, while control nontumorigenic irradiated cells expressed 50 kDa GLUT1 alone. In accordance with this, GLUT3 mRNA was specifically expressed in tumorigenic cell hybrids. To examine the role of GLUT3, clones which stably overexpress GLUT3 were developed from both CGL1 and CGL4 cells. In these transfectants, the affinity for 2-deoxyglucose markedly increased, in parallel with the amount of expressed GLUT3 irrespective of its N-glycosylation state. These results suggest that the enhanced GLUT3 expression in HeLa cell hybrids associated with the tumorigenic phenotypes may account for the increased affinity for 2-deoxyglucose. Possible roles of the putative tumor suppressor in control of gene expression and glucose uptake is discussed.     

Cellular tumorigenicity in nude mice. Test of associations among loss of cell-surface fibronectin, anchorage independence, and tumor-forming ability

Fibronectin (FN; also called large external transformation-sensitive [LETS] protein or cell-surface protein [CSP]) is a large cell-surface glycoprotein that is frequently observed to be either absent or greatly reduced on the surfaces of malignant cells grown in vitro. Because FN may be a useful molecular marker of cellular malignancy, we have carried out an extensive screening to test the specific association among the degree of expression of FN, anchorage-independent growth, and tumorigenicity in the athymic nude mouse. A variety of diploid cell strains and established cell lines were tested for the expression of surface FN by indirect immunofluorescence using rabbit antisera against human cold insoluble globulin, rodent plasma FN, or chicken cell- surface FN. Concomitantly, the cells were assayed for tumor formation in nude mice and for the ability to form colonies in methylcellulose. Tumorigenic cells often showed very low surface fluorescence, confirming earlier reports. However, many highly tumorigenic fibroblast lines from several species stained strongly with all three antisera. In contrast, the anchorage-independent phenotype was nearly always associated with tumorigenicity in approximately 35 cell lines examined in this study. In another series of experiments, FN-positive but anchorage-independent cells were grown as tumors in nude mice and then reintroduced into culture. In five of the six tumor-derived cell lines, cell-surface FN was not significantly reduced; one such cell line showed very little surface FN. Our data thus indicate that the loss of cell-surface FN is not a necessary step in the process of malignant transformation and that the growth of FN-positive cells as tumors does not require a prior selection in vivo for FN-negative subpopulations.     

Decreased transforming growth factor-beta response and binding in insulin-independent teratoma-derived cell lines with increased tumorigenic properties.

The mouse C3H teratoma-derived cell line 1246 is an adipogenic cell line which stringently requires insulin to proliferate and differentiate in defined medium. From this cell line an insulin-independent cell line called 1246-3A was isolated. It was found that, in contrast to 1246 cells, 1246-3A cells had lost the ability to differentiate and became tumorigenic when injected at a density of 10(6) cells/mouse into syngeneic host C3H mice. In addition, they produce in their culture medium transforming growth factor alpha- and beta-like polypeptides which stimulate their proliferation. Highly tumorigenic insulin-independent cell lines able to give rise to tumor when injected at a density of 10(4) cells/mouse were isolated by using an in vitro-in vivo shuttle technique. The highly tumorigenic cell lines have lost the response to TGF-beta 1. The binding of TGF-beta 1 to the nontumorigenic parent cell line or to cells displaying increased tumorigenic properties was investigated. The data presented here indicate that the increased tumorigenicity is accompanied by a progressive decrease of specific binding of TGF-beta 1 to the cells. However, the decreased number of cell surface TGF-beta 1 binding sites in the highly tumorigenic cells did not correlate with an increase in the secretion of TGF-beta protein by the tumorigenic cells, as all of TGF-beta produced by the cells was in a latent form. Affinity cross-linking experiments indicated that the 1246 cell line displayed several TGF-beta cross-linked molecular species (MW 140, 92, and 70 kDa). Increase of tumorigenicity was accompanied by a marked decrease in the intensity of the cross-linked bands, particularly of the 92 and 70 kDa species.     

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.     

Tumorigenicity of revertants from an SV40-transformed line

A syndrome of in vitro properties correlates with the tumorigenicity of SV40-transformed rodent cells. These properties are plasminogen activator production, loss of large actin cables, and anchorage-independent growth. An established rat fibroblast line, its SV40 transformant, several T-antigen negative revertants, and a spontaneous retransformant isolated form one of the revertants were analyzed in vivo for their tumorigenicity and in vitro for the syndrome. The two transformed lines were highly tumorigenic, and had clearly abnormal in vitro properties. The parental rat line was weakly tumorigenic in nude mice and demonstrated a slightly transformed response in the in vitro assays. The revertants were completely nontumorigenic. Expression of the in vitro syndrome was not uniform for all revertants; however, most cell lines maintained the correlation of the syndrome and tumorigenicity.