Nonterminally differentiated cells express decreased growth factor responsiveness

Therapeutic Implications of PPARgamma in Human Osteosarcoma

Osteosarcoma (OS) is the most common nonhematologic malignancy of bone in children and adults. Although dysregulation of tumor suppressor genes and oncogenes, such as Rb, p53, and the genes critical to cell cycle control, genetic stability, and apoptosis have been identified in OS, consensus genetic changes that lead to OS development are poorly understood. Disruption of the osteogenic differentiation pathway may be at least in part responsible for OS tumorigenesis. Current OS management involves chemotherapy and surgery. Peroxisome proliferator-activated receptor (PPAR) agonists and/or retinoids can inhibit OS proliferation and induce apoptosis and may inhibit OS growth by promoting osteoblastic terminal differentiation. Thus, safe and effective PPAR agonists and/or retinoid derivatives can be then used as adjuvant therapeutic drugs for OS therapy. Furthermore, these agents have the potential to be used as chemopreventive agents for the OS patients who undergo the resection of the primary bone tumors in order to prevent local recurrence and/or distal pulmonary metastasis.

Senescence represses the nuclear localization of the serum response factor and differentiation regulates its nuclear localization with lineage specificity

The differentiation of cultured 3T3T mesenchymal stem cells into adipocytes represses growth factor responsiveness by limiting the nuclear localization of the serum response factor (SRF) that binds to and activates the promoters of growth control genes that contain the serum response elements (SRE), such as junB and c-fos. The regulation of SRF nuclear localization by adipocyte differentiation is specific, because we show that adipocyte differentiation does not repress the nuclear localization of six other transacting factors. To determine if repression of growth factor responsiveness that occurs during senescence also represses the nuclear localization of SRF, we studied normal human WI-38 fibroblasts at low versus high population doublings. The results show that SRF localizes to the nucleus of proliferative cells whereas in senescent cells SRF can not be detected in the nucleus. This result is apparent in both immunofluorescence assays and in western blot analysis. We next evaluated the cellular distribution of SRF in selected human tissues to determine whether the loss of proliferative potential in vivo could have a different effect on SRF nuclear localization. We found that in cells of the small bowel mucosa, differentiation modulates SRF nuclear localization in an opposite manner. Minimal SRF expression and nuclear localization is evident in undifferentiated cells at the base of crypts whereas increased SRF expression and nuclear localization is evident in differentiated cells at the surface tip of the villus. These results together establish that regulation of SRF expression and nuclear localization is important in senescence and differentiation in a lineage specific manner.

Differentiation, differentiation/gene therapy and cancer

"Differentiation, Differentiation/Gene Therapy and Cancer" is intended to suggest that an understanding of the cell and molecular biology of cell differentiation should advance the development of new cancer therapies. This article, therefore, reviews four general topics and their relationship to each other: (1) the multistep process of cell differentiation in nontransformed and transformed cells, (2) the use of drugs that induce differentiation in vitro as potential clinical differentiation therapy agents for cancer, (3) the evolving emphasis on gene therapy as a new cancer therapy modality, and (4) the concept of differentiation/gene therapy.

Differentiation modulates the balance of positive and negative Jun/AP-1 DNA binding activities to regulate cellular proliferative potential: different effects in nontransformed and transformed cells

Differentiation of 3T3T cells into adipocytes results in the progressive repression of growth factor responsiveness. This is associated with the transcriptional repression of the inducibility of c-jun and junB expression by serum. In contrast, differentiation of SV-40 large T antigen-transformed 3T3T cells (CSV3-1) does not repress growth factor responsiveness nor c-jun or junB inducibility even though CSV3-1 cells can differentiate into adipocytes. To better explain these observations, we have studied compositional changes in AP-1 DNA binding activity attributed to c-Jun, JunB, and JunD during the differentiation process in 3T3T and CSV3-1 cells. The results show that in nontransformed 3T3T cells, differentiation represses AP-1 DNA binding activity via a proportionate downregulation of c-Jun, JunB, and JunD. In contrast, in CSV3-1 cells, AP-1 DNA binding activity increases twofold during differentiation, which is accounted for by an increase in JunD with no change in c-Jun and JunB. If c-Jun and JunB serve as positive regulators and JunD serves as a negative regulator for cell proliferation as suggested by previous studies, the repression of JunD expression in differentiating CSV3-1 cells should be mitogenic because decreasing JunD/AP-1 DNA binding activity would allow c-Jun/AP-1 and JunB/AP-1 DNA binding activities to be dominant. The results confirm this prediction showing that antisense junD oligodeoxyribonucleotides are mitogenic for differentiating CSV3-1 cells whereas antisense c-jun and junB inhibit mitogenesis. These data support the conclusion that differentiation can regulate cellular proliferative potential by modulating the balance of positive and negative Jun/AP-1 DNA binding activities in distinct ways in nontransformed and transformed cells.

Unique and selective mitogenic effects of vanadate on SV40-transformed cells

Vanadate and insulin both function as unique complete mitogens for SV40-transformed 3T3T cells, designated CSV3-1, but not for nontransformed 3T3T cells. The mitogenic effects induced by vanadate and insulin in CSV3-1 cells are mediated by different signaling mechanisms. For example, vanadate does not stimulate the tyrosine phosphorylation of the insulin receptor beta-subunit nor the 170 kDa insulin receptor substrate-1. Instead, vanadate induces a marked increase in tyrosine phosphorylation of 55 and 64 kDa proteins that is not observed in insulin-stimulated CSV3-1 cells. Perhaps most interestingly, vandate-induced mitogenesis is associated with the selective induction of c-jun and junB expression without significantly inducing c-fos or c-myc. Furthermore, treatment of CSV3-1 cells with genistein abolishes the effects of vanadate on protein tyrosine phosphorylation and c-jun induction. These and related data suggest that modulation of protein tyrosine phosphorylation and c-jun and junB expression may serve the critical roles in mediating vandate-induced mitogenesis in SV40-transformed cells.

Adipocyte differentiation selectively represses the serum inducibility of c-jun and junB by reversible transcription-dependent mechanisms

Nonterminally differentiated 3T3 T adipocytes are resistant to growth stimulation by 10% (vol/vol) fetal bovine serum even though they can be induced to proliferate with extremely high serum concentrations. We now report that in adipocytes 10% fetal bovine serum also fails to typically induce c-jun or junB. Rather, after 10% fetal bovine serum treatment, c-jun and junB expression is markedly repressed after a brief initial slight induction. Gel mobility shift studies confirm that AP-1 DNA binding activity is inhibited in adipocytes. Repression in c-jun and junB inducibility in adipocytes results from transcriptional mechanisms, can be reversed by treatment with protein synthesis inhibitors or higher serum concentrations, and does not affect c-fos or c-myc expression. These data suggest that adipocyte differentiation selectively and transcriptionally represses the inducibility of c-jun and junB so as to decrease the cell's ability to proliferate in response to 10% fetal bovine serum.

Regulation of differentiation and protein kinase C expression in 3T3 T proadipocytes: effects of TGF-beta and transformation

We are studying the mechanisms that regulate proliferation and differentiation of normal 3T3 T proadipocytes and neoplastically transformed clones which have lost the ability to differentiate. The phorbol ester 12-O-tetradecanoylphorbol-13-acetate (TPA) and transforming growth factor beta (TGF-beta) are known inhibitors of the same step of the differentiation process in 3T3 T cells. Here, we examined the expression of the phorbol ester receptor/protein kinase C (PKC) during adipocytic differentiation of 3T3 T cells and its modulation by the differentiation inhibitor TGF-beta. PKC receptor assays were performed using a tritiated analogue of TPA and it was found that PKC receptor levels decreased approximately threefold during differentiation. Northern blot analyses revealed an even greater decrease of PKC transcripts during differentiation. TGF-beta inhibited not only differentiation, but the differentiation-dependent decrease in PKC levels as well. Transformed 3T3 T cells which have lost the ability to differentiate were found to express aberrant levels of PKC. The data suggest that TGF-beta may inhibit differentiation via a PKC-dependent pathway and that disruption of normal PKC levels or its regulation may be involved in the loss of differentiation control in transformed 3T3 T cells.

Growth factor responsiveness: role of MyoD and myogenin

A differentiation-defective variant (DD-1) of the MM14 myoblasts acquired the ability to synthesize DNA in response to treatment with epidermal growth factor (EGF) (R. W. Lim and S. D. Hauschka, 1984, Dev. Biol. 105, 48) and no longer expressed myogenic determinant genes (i.e., MyoD and myogenin) (P.R. Mueller, and B. Wold, 1989, Science 246, 780). To determine the effect of expression of MyoD on EGF responsiveness, DD-1 cells were cotransfected with a MyoD expression vector and with pRSVneo. A clone, MyoDD-1 cells, which was G418 resistant, formed multinuclear syncitia, and also expressed MyoD and myogenin, was further characterized. EGF responsiveness, as assessed by DNA synthesis, was decreased 5- to 10-fold in the MyoDD-1 cells from that in G418-resistant control DD-1 cells, despite similar EGF receptor numbers and binding affinities of the receptors. Responsiveness of MyoDD-1 cells to fibroblast growth factor (FGF) was also diminished although to a lesser extent. To determine the effects of decreased myogenic determinant gene expression on mitogen responsiveness, MM14 myoblasts were grown in medium supplemented with 5 microM 5-bromo-2'-deoxyuridine (BUdR-MM14). BUdR-MM14 cells had decreased expression of MyoD and myogenin, did not fuse, and had an altered morphology, from round to flat. The BUdR effect on fusion and cell shape was reversed by growth in control medium. BUdR-MM14 cells were responsive to EGF and had enhanced responsiveness to FGF. The combined studies support the view that expression of MyoD and/or myogenin contributes to negative regulation of mitogen responsiveness.

Repression of SV40 T oncoprotein expression by DMSO

SV40 large T oncoprotein-transformed murine mesenchymal 3T3 T stem cells (CSV3 cells) can be induced to growth arrest and then differentiate into adipocytes. When differentiation occurs, SV40 T oncoprotein expression is repressed (Estervig et al., J Virol 63:2718, 1989). To determine if repression of T oncoprotein expression can also be induced pharmacologically, the effect of a variety of agents that have been reported to effect differentiation in various cell types but not in 3T3 T or CSV3 cells was tested. This rationale suggests that if any of these agents repress T oncoprotein expression in CSV3 cells, then the results would establish that repression of T oncoprotein expression can be mediated by mechanisms independent of overt differentiation. The results show that dimethylsulfoxide (DMSO) is the only agent tested that represses T oncoprotein expression in CSV3 cells. Repression occurs in a dosage-dependent manner within 24-96 hours after exposure to DMSO. The effect of DMSO on T oncoprotein expression is mediated by posttranslational mechanisms that decrease the stability of the T oncoprotein. DMSO-induced repression of T oncoprotein expression is also associated with reversion of the transformed phenotype in CSV3 cells as demonstrated by the loss of responsiveness to a specific transformation-associated mitogen. These data support the conclusion that the pharmacological repression of T oncoprotein expression represents a form of cancer suppressor activity that can be mediated by a distinct molecular mechanism.

TGF-beta blocks early but not late differentiation-specific gene expression and morphologic differentiation of 3T3 T proadipocytes

Transforming growth factor-beta (TGF-beta) inhibits morphologic differentiation of BALB/c 3T3 T cells as well as other proadipocyte models. Our prior studies suggested that TGF-beta may act only during the early stages of differentiation induction. However, we did not determine whether TGF-beta was differentially effecting expression of any of the various differentiation-specific genes or if it could cause down-regulation of these genes in differentiated cells. Therefore, in the current study we tested the effects of exogenous TGF-beta (0.01-5.0 ng/ml) on morphologic differentiation and on differentiation-dependent gene expression (Northern and slot blot analyses) at various times during differentiation. When induced to differentiate, 3T3 T cells first undergo predifferentiation growth arrest and from this state molecular, biochemical, and morphological differentiation proceeds. Here it was found that when added prior to the onset of differentiation, TGF-beta was a potent inhibitor or morphologic differentiation as well as of the expression of differentiation-specific genes such as lipoprotein lipase (LPL) and glycerol-3-phosphate dehydrogenase (GPD). However, once morphologic differentiation began, TGF-beta was ineffective in blocking differentiation. In addition, exposure of fully differentiated cells to TGF-beta for up to 72 hours caused no decrease of differentiation-specific genes and even a 7-day treatment caused no morphologic dedifferentiation. Tumor necrosis factor also had no detectable effect on fully differentiated cells.

Insulin-induced mitogenesis associated with transformation by the SV40 large T antigen

Simian virus 40 (SV40) large T antigen-transformed cells typically show a markedly reduced serum requirement for growth and the inability to growth arrest and differentiate. An SV40 large T antigen-transformed 3T3 T cell line, CSV3-1, that can growth arrest and differentiate into adipocytes with high efficiency has, however, recently been described (Scott et al: Proc. Natl. Acad. Sci. U.S.A. 86:1652-1656, 1989; Estervig et al: J. Virol. 63:2718-2725, 1989; J. Cell. Physiol. 142:552-558, 1990). The results of the current studies using these cells show that whereas quiescent 3T3 T cells show no mitogenic response to insulin, quiescent CSV3-1 cells show a highly significant insulin-induced mitogenic responsiveness in the absence of other added growth factors. Maximum mitogenesis was observed at an insulin concentration of 1 microgram/ml, which induced 40-70% of the cells to undergo DNA synthesis within 48 hours. The half maximum response was achieved with 1-10 ng/ml of insulin. Insulin's mitogenic effect on CSV3-1 cells was evident under several different culture conditions that induce quiescence and was not mediated by any detectable autocrine growth factors that might make CSV3-1 cells competent to respond to insulin. In CSV3-1 cells insulin appears to act on its own receptor rather than on the IGF-1 receptor, because at comparable dosages IGF-1 is 10- to 100-fold less effective than insulin. Insulin also is shown to be a mitogen for another SV40-transformed cell line, CSV3-35, which can be growth arrested; in contrast insulin has no mitogenic effect on two control cell lines that are stably transfected with pSV2neo, a plasmid containing SV40 early promoter/enhancer but lacking large T antigen gene: These results suggest a significant relationship between SV40 T antigen-associated transformation and the expression of mitogenic responsiveness to insulin.

Discrimination of two fibroblast progenitor populations in early explant cultures of hamster gingiva

Numerous metabolic studies have demonstrated heterogeneity of fibroblast populations in culture, yet little is known about the structure of fibroblast populations in adult tissues in vivo. To determine if populations of both cycling and non-cycling cells are present in gingiva, hamsters were labelled with [3H]-thymidine to label cycling cells in vivo, and explanted biopsies were subsequently incubated with bromodeoxyuridine to label cycling cells in vitro. Cycling cells were identified by combined immunohistochemistry and radioautography. Fibroblasts were recognized by the presence of vimentin and the absence of keratin as determined by immunofluorescence. The largest proportion of cells were double-labelled with [3H]-thymidine and bromodeoxyuridine (43.8%) indicating the presence of actively cycling populations that maintained their proliferative status upon explanation. Cultures also exhibited a second population of cells labelled only with bromodeoxyuridine (38.7%) that did not cycle in vivo, but retained the capacity for proliferation in vitro. However, limiting dilution analysis of single-cell suspensions revealed only a single class of progenitors capable of forming large colonies in vitro. Approximately 1 in 190 plated cells was capable of colony-formation, indicating that, upon explanation, a subset of the cycling cells in vitro exhibits extensive proliferative capacity. There was also a small population of cells unlabeled with either [3H]-thymidine or bromodeoxyuridine (9.4%) that appeared to be terminally differentiated. Different substrates, including glass and thin films of gelatin and collagen, did not significantly alter the fraction of cells labelled with [3H]-thymidine. These data demonstrate the existence of 2 separate progenitor-cell populations with different capacities for proliferation in vivo and in vitro.

Prostaglandin endoperoxide synthase (cyclooxygenase) mRNA and protein production in mouse myoblasts and a differentiation-defective variant

Northern blot analysis revealed that a differentiation-defective variant (DD-1) of MM14 mouse myoblasts has seven times the prostaglandin endoperoxide synthase mRNA than the parental MM14 myoblasts. There was an even greater increase in the level of prostaglandin endoperoxide synthase protein in the DD-1 cells as compared to that in the MM14 myoblasts. In fact, prostaglandin endoperoxide synthase was not detectable by Western blot analysis of extracts from MM14 myoblasts. Since prostaglandin endoperoxide synthase has been reported to be a gene whose expression is induced transiently, i.e., growth-regulated, upon mitogen stimulation of quiescent cells, the RNA abundance of other growth-regulated genes was examined including: KC, JE, c-myc, 1B6, and vimentin. Northern blot analysis revealed that the mRNA abundance of JE, KC, and c-myc is 12-, 17-, and 2-fold higher, respectively, in growing DD-1 cells than in growing MM14 myoblasts. In contrast, there was little difference in the mRNA abundance of 1B6 and vimentin. These results are consistent with the hypothesis that increases in the levels of expression of prostaglandin endoperoxide synthase and some growth-regulated genes are integral to the expression of the differentiation-defective phenotype and may in fact contribute to this phenotype.

Improved coupling between proliferation-arrest and differentiation-induction in ML-1 human myeloblastic leukemia cells

Proliferation and differentiation are coupled in normal cells and are aberrant in leukemia cells. The studies reported here were aimed at more effectively coupling proliferation-arrest and differentiation-induction in a human myeloblastic leukemia cell line (ML-1). This was accomplished by using reduced serum conditions in conjunction with a differentiation-inducing agent: cells were first incubated in reduced serum [0.3% fetal bovine serum (FBS)] instead of standard conditions (7.5% FBS) and, second, exposed to 12-O-tetradecanoylphorbol-13-acetate (TPA). The effects of this protocol were as follows: first, cell proliferation was slowed and cells accumulated in G0/G1 phase of the cell cycle; this occurred with only a minimal decrease in viability [to approximately 88-92% (0.3% FBS) from greater than or equal to 96% (7.5% FBS)]. Second, the induction of differentiation was accelerated; this allowed the time of exposure to TPA to be decreased. Acceleration of induction was very pronounced when cells were maintained in 0.3% FBS both before and during exposure to TPA, with TPA at concentrations above the minimum sufficient for induction but below those causing significant cytotoxicity; as little as 1 hour of TPA exposure resulted in near-maximal induction (approximately 80%) with this protocol, compared to the greater than or equal to 1 day required with previous standard protocols. In sum, conditions that slow ML-1 cell proliferation (0.3% FBS) enhance TPA-induced differentiation, substantially narrowing the time frame of induction; these conditions should be useful for studying the molecular mechanisms that underlie the induction process.

Three distinct effects of SV40 T-antigen gene transfection on cellular differentiation

SV40 large T-antigen-induced transformation has been reported to block differentiation, but the mechanism(s) of this effect has not been established. The results presented here show that stable transfection of the SV40 T-antigen gene, via the pSV3neo plasmid, has at least three distinct effects on 3T3T adipocyte differentiation. Cells first show a decreased ability to undergo predifferentiation growth arrest, which is a prerequisite for in vitro 3T3T adipocyte differentiation. However, if predifferentiation growth arrest is accomplished by use of stringent differentiation-inducing culture conditions, adipocyte differentiation can occur with high frequency. The pSV3neo-transfected cell clones also show other modifications of the adipocyte differentiation process, including changes in nonterminal (reversible) and terminal (irreversible) steps of adipocyte differentiation. When compared to nontransfected 3T3T cells, the cell clones containing pSV3neo require markedly reduced growth factor concentrations to restimulate proliferation of nonterminally differentiated adipocytes and the terminal step of differentiation is also blocked. These results suggest that integration of the T-antigen gene, through pSV3neo transfection, has multiple effects on the cellular mechanisms of differentiation. It does not block the differentiation process per se; rather it appears to make cells highly sensitive to proliferation signals, thereby making differentiation more difficult.