Butyrate inhibits the retinoic acid-induced differentiation of F9 teratocarcinoma stem cells

Mechanisms Regulating Stemness and Differentiation in Embryonal Carcinoma Cells

Just over ten years have passed since the seminal Takahashi-Yamanaka paper, and while most attention nowadays is on induced, embryonic, and cancer stem cells, much of the pioneering work arose from studies with embryonal carcinoma cells (ECCs) derived from teratocarcinomas. This original work was broad in scope, but eventually led the way for us to focus on the components involved in the gene regulation of stemness and differentiation. As the name implies, ECCs are malignant in nature, yet maintain the ability to differentiate into the 3 germ layers and extraembryonic tissues, as well as behave normally when reintroduced into a healthy blastocyst. Retinoic acid signaling has been thoroughly interrogated in ECCs, especially in the F9 and P19 murine cell models, and while we have touched on this aspect, this review purposely highlights how some key transcription factors regulate pluripotency and cell stemness prior to this signaling. Another major focus is on the epigenetic regulation of ECCs and stem cells, and, towards that end, this review closes on what we see as a new frontier in combating aging and human disease, namely, how cellular metabolism shapes the epigenetic landscape and hence the pluripotency of all stem cells.

Dissecting the retinoid-induced differentiation of F9 embryonal stem cells by integrative genomics

Retinoic acid (RA) triggers physiological processes by activating heterodimeric transcription factors (TFs) comprising retinoic acid receptor (RARα, β, γ) and retinoid X receptor (RXRα, β, γ). How a single signal induces highly complex temporally controlled networks that ultimately orchestrate physiological processes is unclear. Using an RA-inducible differentiation model, we defined the temporal changes in the genome-wide binding patterns of RARγ and RXRα and correlated them with transcription regulation. Unexpectedly, both receptors displayed a highly dynamic binding, with different RXRα heterodimers targeting identical loci. Comparison of RARγ and RXRα co-binding at RA-regulated genes identified putative RXRα-RARγ target genes that were validated with subtype-selective agonists. Gene-regulatory decisions during differentiation were inferred from TF-target gene information and temporal gene expression. This analysis revealed six distinct co-expression paths of which RXRα-RARγ is associated with transcription activation, while Sox2 and Egr1 were predicted to regulate repression. Finally, RXRα-RARγ regulatory networks were reconstructed through integration of functional co-citations. Our analysis provides a dynamic view of RA signalling during cell differentiation, reveals RAR heterodimer dynamics and promiscuity, and predicts decisions that diversify the RA signal into distinct gene-regulatory programs.

Role of retinoids in differentiation and growth of embryonal carcinoma cells

To study how retinoids promote differentiation and inhibit proliferation of embryonal carcinoma (EC) cells, we have followed their intracellular fate. Retinoic acid (RA) is effectively metabolized to more polar compounds by many EC lines. Unlike RA, retinol is slowly metabolized. Our inability to detect conversion of retinol to RA might indicate that the two retinoids elicit their effects on EC cells in different ways. Retinol added to cultures quickly appears in the nuclear fraction; the proportion associated with nuclei after detergent extraction is initially very low but increases with time. Retinol and RA might be translocated to nuclei by their respective binding proteins [cellular retinol-binding protein (CRBP) and cellular retinoic acid-binding protein (CRABP)]: isolated EC nuclei have specific, independent binding sites for both holoproteins but not their ligands. CRABP cannot be detected in the nucleoplasm of untreated EC cells, but activity is measurable after cells are exposed to RA. Interestingly, incubation with retinol promotes movement of both CRBP and CRABP into the nucleoplasmic fraction. Finally, we have demonstrated that brief exposure to RA dramatically reduces the cloning efficiency of EC cells. Since some cells are unaffected even by lengthy exposures to RA whereas the growth of their progeny is inhibited, we suggest that EC cells can become epigenetically refractory to RA.

Transcription Factor IIA tau is associated with undifferentiated cells and its gene expression is repressed in primary neurons at the chromatin level in vivo

The levels of General Transcription Factor (TF) IIA were examined during mammalian brain development and in rat embryo fibroblasts and transformed cell lines. The large TFIIA subunit paralogues alphabeta and tau are largely produced in unsynchronized cell lines, yet only TFIIA alphabeta is observed in a number of differentiated tissue extracts. Steady-state protein levels of the TFIIA tau, alphabeta, and gamma subunits were significantly reduced when human embryonal (ec) and hepatic carcinoma cell lines were stimulated to differentiate with either all-trans-retinoic acid (ATRA) or sodium butyrate. ATRA-treated NT2-ec cells required replating to induce a neuronal phenotype and loss of detectable TFIIA tau and gamma proteins. High levels of TFIIA tau, alphabeta, and gamma and Sp factors were identified in extracts from human fetal and rat embryonic day-18 brains, but not in human and rat adult brain extracts. A high histone H3 Lys9/Lys4 methylation ratio was observed in the TFIIA tau promoter of primary hippocampal neurons from day-18 rat embryos, suggesting that repressive epigenetic marks of chromatin prevent TFIIA tau from being transcribed in neurons. We conclude that TFIIA tau is associated with undifferentiated cells during development, yet is down-regulated at the chromatin level upon cellular differentiation.

A parallel association between differentiation and induction of galectin-1, and inhibition of galectin-3 by retinoic acid in mouse embryonal carcinoma F9 cells

Soluble endogenous lactoside-binding lectins, galectins, have been implicated in cell adhesion, growth, differentiation, neoplastic transformation, and metastasis. Two major classes of these lectins, galectin-1 and galectin-3, are developmentally regulated. To explore the mechanisms by which the expression of the galectins is regulated and to examine their association with the differentiation processes induced by all-trans retinoic acid (RA), dibutyryl cyclic AMP (Bt2cAMP) and their combination, we used the murine embryonal carcinoma (EC) cell line F9 and its RA-resistant mutant, RA-3-10. RA induced endodermal differentiation and a concurrent induction of galectin-1 and its complementary glycoconjugates (laminin and lysosomal-associated membrane protein, LAMP) in the F9 wild-type (wt) line, but failed to induce differentiation and had no effects on or even reduced the expression of galectin-1, laminin, and LAMP in the RA-3-10 line. On the other hand, RA inhibited expression of galectin-3 in the wild-type line but had no effect on the RA-3-10 line. The galectin-1 gene is at least partially regulated at the transcriptional level. These results demonstrate a parallel association between differentiation and induction of galectin-1, and inhibition of galectin-3 in F9 cells by RA. The study suggests that a regulated expression of galectins and their complementary glycoconjugates is involved in the differentiation pathway induced by RA in F9 cells.

RB phosphorylation in sodium butyrate-resistant HL-60 cells: cross-resistance to retinoic acid but not vitamin D3

To examine the potential coupling between inducible cellular changes in RB (retinoblastoma) tumor suppressor protein phosphorylation and ability to G0 growth arrest and differentiate, HL-60 promyelocytic leukemia cells were cultured in incremental sodium butyrate (NaB) concentrations and thereby made resistant to the growth inhibitory effects of sodium butyrate, which normally induces G0 arrest and monocytic differentiation in wild type HL-60 cells. The resistant cells were also unable to differentiate in response to NaB, indicating that a regulatory function controlling both G0 growth arrest and differentiation had been affected. The induced resistance was not genetic in origin since the cells regained the ability to G0 arrest and differentiate after being recultured in medium free of sodium butyrate for only three days. The resistant cells had similar cell cycle phase durations as the original wild type cells. The resistant cells retained the ability to both G0 arrest and differentiate in response to 1,25-dihydroxy vitamin D3 (VD3), normally an inducer of G0 arrest and monocytic differentiation in wild type cells. However, they were cross-resistant to retinoic acid (RA), another ligand for the same steroid thyroid hormone receptor family, which induces G0 arrest and myeloid differentiation in wild type cells. The ability to G0 arrest and phenotypically differentiate in response to RA were both grossly impaired. Unlike wild type cells which undergo early down-regulation and then hypophosphorylation of the RB protein when induced to differentiate, in resistant cells, hypophosphorylation of RB in response to NaB was grossly retarded. These changes in RB protein occurred faster when the cells were treated with VD3. In contrast, the changes in RB phosphorylation occurred significantly slower when the cells were treated with RA. The results suggest a coupling between the ability to G0 growth arrest and phenotypically convert and the ability to hypophosphorylate RB.

Sodium butyrate inhibits myogenesis by interfering with the transcriptional activation function of MyoD and myogenin

Sodium butyrate reversibly inhibits muscle differentiation and blocks the expression of many muscle-specific genes in both proliferating myoblasts and differentiated myotubes. We investigated the role of the basic helix-loop-helix (bHLH) myogenic determinator proteins MyoD and myogenin in this inhibition. Our data suggest that both MyoD and myogenin are not able to function as transcriptional activators in the presence of butyrate, although both apparently retain the ability to bind DNA. Transcription of MyoD itself is extinguished in butyrate-treated myoblasts and myotubes, an effect that may be due to the inability of MyoD to autoactivate its own transcription. We present evidence that the HLH region of MyoD is essential for butyrate inhibition of MyoD. In contrast to MyoD and myogenin, butyrate does not inhibit the ubiquitous basic HLH protein E2-5 from functioning as a transcriptional activator.

Sodium butyrate inhibits myogenesis by interfering with the transcriptional activation function of MyoD and myogenin

Sodium butyrate reversibly inhibits muscle differentiation and blocks the expression of many muscle-specific genes in both proliferating myoblasts and differentiated myotubes. We investigated the role of the basic helix-loop-helix (bHLH) myogenic determinator proteins MyoD and myogenin in this inhibition. Our data suggest that both MyoD and myogenin are not able to function as transcriptional activators in the presence of butyrate, although both apparently retain the ability to bind DNA. Transcription of MyoD itself is extinguished in butyrate-treated myoblasts and myotubes, an effect that may be due to the inability of MyoD to autoactivate its own transcription. We present evidence that the HLH region of MyoD is essential for butyrate inhibition of MyoD. In contrast to MyoD and myogenin, butyrate does not inhibit the ubiquitous basic HLH protein E2-5 from functioning as a transcriptional activator.

Effects of growth medium and cyclic AMP analogues on the cAMP-induced differentiation of F9 teratocarcinoma cells

F9 teratocarcinoma cells differentiate into parietal endodermlike cells when treated with retinoic acid (RA) and cyclic AMP (cAMP). We have previously found that F9 cells can be induced to differentiate by treatment with cAMP in the absence of RA. In the course of determining why other investigators had failed to observe cAMP-induced differentiation, we found that the growth medium played an important role in determining the response of F9 cells to differentiating agents. When F9 cells were grown in minimal essential medium (MEM) and treated with either 8-bromo-cAMP (8BrcA) + 1-methyl, 3-isobutylxanthine (MIX), or dibutyryl cAMP (DBcA) + theophylline (T), they differentiated to parietal endodermlike cells without any requirement for exogenous RA. However, when F9 cells were grown in Dulbecco's modified Eagle's medium (DME), DBcA/T failed to induce differentiation alone and required exogenous RA to induce formation of parietal endoderm-like cells. 8BrcA/MIX alone was still able to induce some differentiation, although the extent was not as great as those cells grown in MEM. These results could not be explained by the different growth-promoting properties of the two culture media because there was no difference in the doubling time of F9 cells grown in either medium. Likewise, RA and cAMP both inhibited growth to the same extent in either medium. Inasmuch as almost all published reports on F9 cell differentiation have used DME as a growth medium, and RA with or without DBcA/T as the differentiating agents, these studies would not have had the appropriate conditions to detect the cAMP-induced differentiation of F9 cells.

Reversible effects of sodium butyrate on the differentiation of F9 embryonal carcinoma cells

We have studied effects of sodium butyrate on embryonal carcinoma F9 cell differentiation. In the presence of sodium butyrate, F9 cells underwent rapid and drastic morphological changes and expressed marked increases in mRNA levels of various differentiation markers. When sodium butyrate was removed from the cultures, all the examined phenotypes of F9 cell differentiation rapidly reverted to the characteristics of undifferentiated stem cells. However, under the same conditions, when cycloheximide or actinomycin D was added to the cultures, such phenotypic reversion was not observed, but high mRNA levels of the differentiation markers as well as altered cell morphology were retained. These results indicated that the effects of sodium butyrate on induction of teratocarcinoma cell differentiation were reversible and that de novo syntheses of some mRNA(s) and protein(s) were necessary for the reversion of differentiated cells to stem cells.

A2B5-reactive ganglioside expression determines the differentiation stage and capacity of rat insulinoma (RIN) sublines

We have generated rat insulinoma (RIN) sublines AlGh, m5F, A12, A13 and AhGh with increasing surface expression of the A2B5 ganglioside, a marker of differentiation. We asked whether the capacity of the sublines to differentiate was related to their stage of differentiation, as is characteristic of cells within the normal beta-cell lineage. To answer this, we measured the effect of the differentiation inducer sodium butyrate (NaB, 1 mM) on proliferation, insulin content, secretion and biosynthesis, and the expression of A2B5 and 3G5 gangliosides by the sublines. Six days after exposure to NaB, cell numbers/dish ranged from (1-3) x 10(6) compared to (4-6) x 10(6) in control cultures. By day 2, AlGh, m5F, A12, A13 and AhGh cells exposed to NaB contained 1.5-, 1.4-, 1.4-, 1.2- and 1.0-fold higher amounts of insulin, respectively, and by day 6, 3.6-, 2.3- and 1.0-fold higher, and 1.2- and 2.4-fold lower, amounts of insulin, respectively, than control cells. After 2 days, insulin secretion from AlGh, m5F, A12, A13 and AhGh cells was 1.7-, 1.0-, 1.5-, 1.0- and 1.0-fold higher, respectively, and the rate of (pro)insulin biosynthesis 1.7-, 2.3-, 1.3-, 1.0- and 1.0-fold higher, respectively, than control cells. After 6 days, A2B5 ganglioside expression was increased 3-, 1.9- and 2-fold on m5F, A12 and A13 cells, respectively, but was not significantly altered on AlGh and AhGh cells. 3G5 ganglioside expression was increased 1.5- and 8.4-fold, respectively, on AlGh and m5F cells, but was unaltered on A12, A13 and AhGh cells.(ABSTRACT TRUNCATED AT 250 WORDS)

Cyclic adenosine monophosphate-mediated induction of F9 teratocarcinoma differentiation in the absence of retinoic acid

F9 teratocarcinoma stem cells differentiate into parietal endoderm-like cells when given retinoic acid (RA) and dibutyryl cyclic adenosine monophosphate (DB-cAMP). It is generally accepted that the stem cells are resistant to the action of cAMP alone and need to be primed by RA in order to respond to cAMP. In this report, we demonstrate that F9 stem cells differentiate into parietal endoderm-like cells in the absence of exogenous RA when treated with cholera toxin and 1-methyl,3-isobutyl xanthine (CT/MIX) or 8-bromo-cAMP/MIX (8B2-cAMP/MIX). Cells treated with CT/MIX or 8B2-cAMP/MIX were morphologically similar to parietal endoderm-like cells, produced high amounts of plasminogen activator, and synthesized both type IV collagen and laminin mRNA. Conversely, markers made in abundance by stem cells such as stage-specific embryonic antigen (SSEA-1) and an mRNA species of 6.8 kb (pST6-135) were markedly reduced in CT/MIX-treated cells. To prove that cAMP alone could induce differentiation Lipidex-1000, a hydrophobic gel, was used to remove 80-90% of the endogenous serum retinoids. F9 cells grown in this retinoid-depleted serum and treated with 8B2-cAMP/MIX differentiated to parietal endoderm-like cells as shown by both dramatic changes in morphology and induction of type IV collagen mRNA. Our results indicate that the differentiation of F9 to parietal endoderm-like cells can be induced by increased intracellular cAMP and is not strictly dependent on the addition of RA.

Increased proteoglycan synthesis following the differentiation of F9 embryonal carcinoma cells: formation of a differentiation-specific proteoheparan sulfate

We have examined changes in proteoglycan synthesis by F9 embryonal carcinoma cells after the cells have been treated with retinoic acid or retinoic acid plus cholera toxin. Retinoic acid is known to stimulate the differentiation of this cell type to a primitive endoderm-like cell characterized by the production of basement membrane components such as type IV collagen, laminin and proteoglycans. We have now demonstrated that proteoglycan synthesis and secretion were further stimulated when cholera toxin was added in addition to retinoic acid. Moreover, media of these fully differentiated cells was found to contain a different species of proteoheparan sulfate not produced by stem cells or retinoic acid-treated cells. This proteoheparan sulfate had a high density upon CsCl gradient centrifugation. The protein core of this proteoheparan sulfate was estimated by SDS gel electrophoresis to be approximately 15,000 daltons.

Sodium butyrate selectively antagonizes the inhibitory effect of retinoids on cornified envelope formation in cultured human keratinocytes

Sodium butyrate affects cell differentiation in confluent epidermal keratinocyte cultures by considerably increasing the spontaneous formation of cross-linked envelopes in normal human keratinocytes (NHK). It also favors the development of envelope competence in the Simian virus-40 (SV-40)-transformed human foreskin keratinocyte line SV-K14. It completely abolishes the inhibitory effect of serum and retinoic acid on the expression of plasma membrane-associated transglutaminase. However, other markers of epidermal differentiation that are also under the control of retinoids such as keratins or the enzyme cholesterol sulfotransferase are not affected by butyrate. The level of the cellular retinoic acid binding protein (CRABP) is considerably increased in its presence. Butyrate does not interfere with the binding of retinoids to their cellular binding proteins. Our observations suggest that sodium butyrate stimulates cornified envelope formation via the induction of the plasma membrane-associated transglutaminase required for cornified envelope synthesis and, additionally, by abolishing the inhibitory effect of retinoids on the expression of this enzyme.

Stimulation of choline acetyltransferase activity by retinoic acid and sodium butyrate in a cultured human neuroblastoma

Choline acetyltransferase (Acetyl-CoA:choline O-acetyltransferase, EC 2.3.1.6, abbreviated ChAT), the biosynthetic enzyme for acetylcholine and acetylcholinesterase (EC 3.1.1.7, abbreviated AChE) are expressed in a human cholinergic neuroblastoma cell line, MC-IXC. We have shown that ChAT activity can be regulated in culture by retinoic acid, an active metabolite of vitamin A, and by sodium butyrate, an organic fatty acid. Optimal concentrations of these agents produce 4.3-fold and 1.6-fold increases in ChAT activity, respectively. The effects of retinoic acid are statistically significant after 24 h, whereas for sodium butyrate significant differences are seen only after 48 h. Since retinoic acid stimulation of ChAT activity was reversed only by trypsin treatment and not by removal of retinoic acid from the medium, this suggests that this agent may be acting at the level of the cell surface. Other differentiating conditions, such as culture in serum-free medium or addition of 1-2% dimethylsulfoxide did not increase ChAT activity. Acetylcholinesterase activity was shown to increase only in the presence of sodium butyrate, suggesting that retinoic acid and sodium butyrate may be acting via different pathways. Retinoic acid and sodium butyrate both seem to be permissive rather than instructive in regulating ChAT activity in that they are unable to induce ChAT expression de novo in cell lines which do not already express ChAT activity.

Isolation of a clone of F9 teratocarcinoma cells "naturally" resistant to G418

Resistance to the neomycin analogue G418 forms the basis of a dominant marker selection system for mammalian (and other) cells transfected with the bacterial neo gene. This system has been particularly effective because of the low incidence of spontaneous conversion to G418 resistance in mammalian cells; no case of resistance to the drug in the absence of the bacterial genes has yet been reported to our knowledge. During the course of transfection experiments, we recently isolated a clone of F9 teratocarcinoma cells which is drug resistant yet has no detectable integrated plasmid sequences, neo RNA transcripts, or aminoglycoside phosphotransferase activity. The G418-resistant clone (F9nr7) did not display enhanced resistance to other cytotoxic drugs tested: colchicine, actinomycin D, cycloheximide, and hygromycin B. Therefore, nr7 cells differ from multidrug-resistant phenotypes previously described. However, this clone is inhibited, relative to control cells, in its response to the differentiation-inducing drugs retinoic acid and dibutyryl cAMP, which suggests that some aspects of general drug metabolism may be altered in these cells.

Effect of sodium butyrate on S-100 protein levels and the cAMP response

Sodium butyrate (NaB), when added to cell cultures, produces a variety of morphological and biochemical changes. We examined its effects, in nM concentrations, on the expression of two glioma cell-associated proteins, glial fibrillary acidic protein (GFAP) and S-100 protein in human glioma-derived cell line (RF), and of S-100 protein in the C6 rat glioma cell line. GFAP levels decreased by about 50% in the RF cell line, and S-100 protein levels decreased protein levels decreased by about 40% after treatment with 1 mM NaB for 48 h. In the C6 rat glioma cell line, isoproterenol with theophylline was found to increase S-100 levels by two-fold over basal levels. NaB was found to inhibit the induction of S-100 protein but exhibited no effect on the basal levels of the protein. Other short chain fatty acids, including sodium propionate and sodium isobutyrate, exhibited partial inhibitory activity. NaB, at an EC50 of 1 mM, was also found to inhibit both the beta-adrenergic and the forskolin-mediated increase in cAMP levels in these cells. This suggests that NaB may inhibit cells from expressing S-100 protein by attenuating cAMP levels.

Butyrate selectively activates the metallothionein gene in teratocarcinoma cells and induces hypersensitivity to metal induction

The expression of metallothionein genes (MT-I and MT-II) was shown to be enhanced within 2 h of addition of 2.5-5 mM sodium butyrate to cultures of teratocarcinoma cells. Both undifferentiated stem cells (F9 and OC15) and differentiated cells (PSA5E and OC15 END) reacted similarly to butyrate by increased accumulation of MT mRNAs. As expected, all of the teratocarcinoma cells that were tested also responded to Zn2+ and Cd2+ by 5- to 10-fold increases in MT mRNA accumulation within 2-24 h of metal addition to the culture media. Surprisingly, MT genes in cells pretreated with butyrate were hypersensitive to metal induction, and this was demonstrated by accumulated transcript levels and by synthesis of MT protein. The maximal metal response was obtained by exposure of cells to butyrate for around 5-8 h together with 10 microM heavy metals. Metal additions to culture media over a range of concentrations and times only induced half the levels of MT mRNA that were achieved by butyrate plus metals. Butyrate enhanced the rate of accumulation of MT mRNA in response to metals, increased the sensitivity of the MT gene to metals, and protected cells from toxic effects of high concentrations of metals. The butyrate and metal ion responses were selective in that no accumulation of c-myc, c-fms, HSP-70, or AFP mRNA was detected. However, c-fos mRNA accumulated in cells exposed to toxic concentrations of metals (50 microM and higher) and this was also potentiated by butyrate treatment. These results suggest that butyrate alters the chromatin conformation of both the MT-I and MT-II genes leading to an accentuated transcriptional response to metals.

Three types of muscle-specific gene expression in fusion-blocked rat skeletal muscle cells: translational control in EGTA-treated cells

When rat skeletal muscle cells were treated with EGTA, an inhibitor of cell fusion, a battery of muscle-specific mRNAs was synthesized but not translated despite the synthesis of many other proteins. Most of the muscle-specific mRNAs were associated with polysomes in fused myotubes, whereas they were found in postpolysomal fractions in EGTA-treated cells. Therefore, in addition to the well-documented transcriptional and posttranscriptional control of muscle-specific genes, translational control of this specific group of genes, presumably involving a Ca2+-dependent process, is also observed in these fusion-blocked cells. These findings and results obtained with other fusion inhibitors demonstrate that three types of muscle-specific gene expression take place in the fusion-blocked cells depending on the inhibitors used: one, neither muscle-specific mRNAs nor proteins are synthesized; two, the mRNAs are synthesized but not translated; and three, both the mRNAs and the proteins are synthesized.

c-myc regulation during retinoic acid-induced differentiation of F9 cells is posttranscriptional and associated with growth arrest

We have shown that c-myc mRNA levels decrease more than 20-fold when F9 teratocarcinoma stem cells are induced to arrest growth and terminally differentiate to parietal endoderm after exposure to retinoic acid and cyclic AMP (Campisi et al., Cell 36:241-247, 1984). Here, we demonstrate that although growth arrest and full expression of the differentiated phenotype required about 3 days, c-myc mRNA declined abruptly between 8 and 16 h after the addition of retinoic acid and cyclic AMP. The decline was independent of cyclic AMP. We found little or no change in the level of c-myc transcription during differentiation, although two other genes showed marked transcriptional regulation. Thus, decreased c-myc mRNA is a consequence of very early posttranscriptional regulation directed by retinoic acid. Differentiation was not fundamental to this regulation. We have shown that sodium butyrate blocks expression of the differentiated phenotype if added within 8 h of retinoic acid and cyclic AMP (Levine et al., Dev. Biol. 105:443-450, 1984). However, butyrate did not inhibit the decrease in c-myc mRNA. Furthermore, F9 cells partially arrested growth without differentiating when grown in isoleucine-deficient medium. Under these conditions, c-myc mRNA levels also declined. Our results suggest that induction of differentiation-specific genes may be under retinoic acid-mediated control dissimilar from that responsible for the decay of c-myc mRNA. In addition, they raise the possibility that growth arrest may be initiated by reduced c-myc expression.

c-myc regulation during retinoic acid-induced differentiation of F9 cells is posttranscriptional and associated with growth arrest

We have shown that c-myc mRNA levels decrease more than 20-fold when F9 teratocarcinoma stem cells are induced to arrest growth and terminally differentiate to parietal endoderm after exposure to retinoic acid and cyclic AMP (Campisi et al., Cell 36:241-247, 1984). Here, we demonstrate that although growth arrest and full expression of the differentiated phenotype required about 3 days, c-myc mRNA declined abruptly between 8 and 16 h after the addition of retinoic acid and cyclic AMP. The decline was independent of cyclic AMP. We found little or no change in the level of c-myc transcription during differentiation, although two other genes showed marked transcriptional regulation. Thus, decreased c-myc mRNA is a consequence of very early posttranscriptional regulation directed by retinoic acid. Differentiation was not fundamental to this regulation. We have shown that sodium butyrate blocks expression of the differentiated phenotype if added within 8 h of retinoic acid and cyclic AMP (Levine et al., Dev. Biol. 105:443-450, 1984). However, butyrate did not inhibit the decrease in c-myc mRNA. Furthermore, F9 cells partially arrested growth without differentiating when grown in isoleucine-deficient medium. Under these conditions, c-myc mRNA levels also declined. Our results suggest that induction of differentiation-specific genes may be under retinoic acid-mediated control dissimilar from that responsible for the decay of c-myc mRNA. In addition, they raise the possibility that growth arrest may be initiated by reduced c-myc expression.