Adenovirus-mediated increase of HNF-3 levels stimulates expression of transthyretin and sonic hedgehog, which is associated with F9 cell differentiation toward the visceral endoderm lineage.

Retinoic acid-induced differentiation of mouse F9 embryonal carcinoma cells toward the visceral endoderm lineage is accompanied by increased expression of the Forkhead Box (Fox) transcription factors hepatocyte nuclear factor 3a (HNF-3alpha) and HNF-3beta, suggesting that they play a crucial role in visceral endoderm development. Retinoic acid stimulation results in a cascade of HNF-3 induction in which HNF-3alpha is a primary target for retinoic acid action and its increase is required for subsequent induction of HNF-3beta expression. Increased expression of HNF-3beta precedes activation of its known target genes, including transthyretin (TTR), Sonic hedgehog (Shh), HNF-1alpha, HNF-1beta, and HNF-4alpha. In order to examine whether increased HNF-3 expression is sufficient to induce expression of its downstream target genes without retinoic acid stimulation, we have used adenovirus-based expression vectors to increase HNF-3 protein levels in F9 cells. We demonstrate that adenovirus-mediated increase of HNF-3alpha levels in F9 cells is sufficient to induce activation of endogenous HNF-3beta levels followed by increased TTR and Shh expression. Furthermore, we show that elevated HNF-3beta levels stimulate expression of endogenous TTR and Shh without retinoic acid stimulation. Moreover, ectopic HNF-3 levels in undifferentiated F9 cells are insufficient to induce HNF-3alpha, HNF-1alpha, HNF-1beta, and HNF-4alpha expression, suggesting that their transcriptional activation required other regulatory proteins induced by the retinoic acid differentiation program. Finally, our studies demonstrate the utility of cell infections with adenovirus expressing distinct transcription factors to identify endogenous target genes, which are assembled with the appropriate nucleosome structure.

The HNF-3alpha transcription factor is a primary target for retinoic acid action.

We have previously demonstrated that gene expression of the hepatocyte nuclear factor 3alpha (HNF-3alpha) transcription factor is activated during retinoic-acid-induced differentiation of F9 embryonal carcinoma cells (A. Jacob et al. (1994). Nucleic Acids Res. 22, 2126-2133). We have extended these studies and now show that HNF-3alpha mRNA is induced approximately 6 h after addition of retinoic acid to the cells, peaks at 1 day postdifferentiation, and then declines to undetectable levels. Furthermore, HNF-3alpha induction occurs in the absence of de novo protein synthesis, suggesting that it is a primary target for retinoic acid action. In order to corroborate this hypothesis, we have mapped the cis-acting HNF-3alpha promoter site that mediates the retinoic acid response. DNA sequence analysis indicates that the HNF-3alpha promoter contains an authentic retinoic acid response element (RARE) of the DR5 class. As expected, this element is able to confer retinoic acid responsiveness to a heterologous promoter. In addition, the HNF-3alpha-specific RARE is able to interact with various retinoic acid receptor heterodimers of the RAR/RXR type. Since HNF-3alpha is induced early during mammalian neurogenesis, our data shed new light on the connection between retinoic-acid-mediated HNF-3alpha activation and establishment of the neuronal phenotype.

Differential induction of HNF-3 transcription factors during neuronal differentiation.

We have investigated the regulation of transcription factors HNF-3alpha and HNF-3beta during the retinoic acid-mediated differentiation of mouse P19 cells. Retinoic acid treatment converts P19 stem cells into neurons and astrocytes and we have clearly shown that gene expression of both HNF-3alpha and HNF-3beta is activated during this process. HNF-3alpha transcription was detected 2 h after addition of retinoic acid and took place in the absence of de novo protein synthesis. This suggests that HNF-3alpha is a primary target for retinoic acid action. HNF-3alpha induction displays a biphasic profile and HNF-3alpha mRNA reaches maximal levels at 2 and 6 days postdifferentiation. Additional experiments strongly suggest that the second peak is due to HNF-3alpha induction in postmitotic neurons. P19 stem cells, on the other hand, do not contain any detectable HNF-3alpha mRNA. According to our studies, the retinoic acid-mediated induction of HNF-3alpha occurs at the level of transcriptional initiation and is conferred by distal promoter sequences. In comparison to HNF-3alpha, HNF-3beta induction is a subsequent event and detectable levels of HNF-3beta mRNA materialize approximately 1 day after addition of retinoic acid to P19 stem cells. Time course studies firmly demonstrate that HNF-3beta mRNA peaks at about 2 days postdifferentiation and then declines to virtually unreadable levels. This temporal pattern is consistent with HNF-3beta being a secondary target for retinoic acid. In analogy to HNF-3alpha, HNF-3beta activation also takes place at the level of transcriptional initiation. Recent studies implicate HNF-3alpha and HNF-3beta in early mammalian neurogenesis. The detection of HNF-3alpha/beta activation during P19 cell differentiation provides us with a convenient cell culture system to elucidate the induction mechanism and the precise role of both transcriptional regulators in the formation of neuronal cells.

Regulation of ribosomal RNA gene transcription during retinoic acid-induced differentiation of mouse teratocarcinoma cells.

We have examined the mechanism of regulation of rRNA synthesis in mouse F9 teratocarcinoma cells that were induced to differentiate by retinoic acid and dibutyryl cAMP. Ribosomal RNA (rRNA) synthesis was significantly reduced during differentiation of F9 cells into parietal endoderm cells. Nuclear run-on assay revealed that the rRNA gene transcription rates were reduced in differentiated cells, and this phenomenon could be mimicked by in vitro transcription assay using nuclear extracts prepared from F9 stem and F9 parietal endoderm cells. Analysis of the DNA-binding activities of two RNA polymerase I (pol I) transcription factors E1BF/Ku and UBF revealed decreased affinity for their cognate recognition sequences. Immunoblot analysis showed a marked reduction in the amounts of E1BF/Ku and UBF in the differentiated cells. Analysis of the steady-state RNA levels for the smaller subunit of E1BF/Ku and for UBF in differentiating F9 cells revealed decreased mRNA synthesis and increase in message level for the differentiation-specific marker laminin B1 with progression of the differentiated status of the cells. This study has demonstrated that differentiation of mouse F9 teratocarcinoma cells into parietal endoderm cells leads to diminished rRNA synthesis, which may be mediated by reduced DNA-binding activities and amounts of at least two pol I transcription factors.

Cell cycle regulation of a novel DNA binding complex in Saccharomyces cerevisiae with E2F-like properties.

Using a biochemical approach, we have detected an activity in Saccharomyces cerevisiae extract that displays the same DNA binding specificity as the mammalian E2F transcription factor and interacts with TTTCGCGC promoter elements. Additional studies revealed that this factor, termed SCELA (S. cerevisiae E2F-like activity), also binds to the closely related SCB promoter sequences. SCB sites (consensus: TTTCGTG) are involved in the cell cycle regulation of several S. cerevisiae cyclin genes and have been shown to interact with the heterodimeric yeast Swi4-Swi6 complex. However, genetic studies clearly demonstrate that SCELA is not related to Swi4 or Swi6. These experiments imply that SCB sites are able to interact with at least two activities: Swi4-Swi6 and SCELA. Because SCB sites are critical for the periodic activation of cell cycle genes, we asked whether SCELA is regulated during yeast cell cycle. Employing a temperature-sensitive strain, we were able to demonstrate that the DNA binding activity of SCELA oscillates during the cell cycle and reaches its maximum at the transition between the G1 and S phases. Preliminary studies suggest that this fluctuation is mediated by phosphorylation/dephosphorylation events. Further characterization of SCELA by UV cross-linking experiments indicate a molecular mass of 47 kDa for this activity. In addition, we present evidence strongly suggesting that SCELA is actually the DNA binding moiety of a large 300-kDa protein complex. Together, these studies firmly indicate that SCELA (as part of a larger complex) plays a critical role in cell cycle regulation of SCB-containing genes, such as CLN cyclins and HO endonuclease. This hypothesis is consistent with other studies that conclude that the SCB-mediated cell cycle oscillation of CLN cyclins and HO requires activities that are distinct from Swi4-Swi6. Finally, it is worth mentioning that the similarities between SCELA and E2F, which is a crucial component in mammalian cell cycle regulation, extend well beyond the DNA binding specificity. In analogy to E2F, SCELA oscillates during the cell cycle, interacts with other cellular activities, and binds to promoter elements that are known mediators of cell cycle control. We will discuss possible functions for SCELA in yeast cell cycle regulation and its relationship to E2F.

Delayed activation of HNF-3 beta upon retinoic acid-induced teratocarcinoma cell differentiation.

We have investigated the retinoic acid-mediated activation of the transcriptional regulator HNF-3 beta during differentiation of mouse F9 embryonal carcinoma cells. Using gel shifts, HNF-3 beta DNA binding activity was clearly detected in differentiated cells, while F9 stem cells were devoid of this activity. We also demonstrated that HNF-3 beta mRNA is specific for differentiated cells. Addition of retinoic acid to F9 stem cells results in delayed activation of HNF-3 beta mRNA which can be detected 1-2 days after the initiation of differentiation. HNF-3 beta mRNA concentrations are maximal at approximately 4 days postdifferentiation and stay at elevated levels for at least 4 additional days. Nuclear run-on experiments clearly show that HNF-3 beta is activated at the level of transcriptional initiation, suggesting that the increases of beta-specific DNA binding activity and mRNA concentration are merely a reflection of this activation mechanism. F9 cells can give rise to three distinct differentiated cell types, visceral endoderm, parietal endoderm, and primitive endoderm, and we have observed HNF-3 beta stimulation during the formation of all three tissues. HNF-3 beta stimulation upon visceral endoderm differentiation is accompanied by the activation of HNF-3 target genes such as transthyretin, suggesting that HNF-3 beta is involved in the developmental activation of this gene. In contrast, HNF-3 beta target genes in parietal and primitive endoderm have yet to be identified. However, the stimulation of HNF-3 beta during primitive endoderm formation, which is an extremely early event during murine embryogenesis, points toward a role for the factor in crucial determination processes that occur early during development.

Retinoic acid-mediated activation of HNF-3 alpha during EC stem cell differentiation.

We present evidence demonstrating that the liver-enriched transcription factor HNF-3 alpha is activated upon retinoic acid-induced differentiation of mouse F9 embryonal carcinoma cells. We have detected increases in the DNA binding activity and mRNA level of HNF-3 alpha. Both are reflections of the actual activation mechanism at the level of transcriptional initiation, which we showed with the help of HNF-3 alpha promoter constructs. Time course studies clearly show that HNF-3 alpha activation is a transient event. Employing Northern blots, HNF-3 alpha mRNA can be detected between 16 and 24 hours post-differentiation, reaches its zenith at approximately 1 day, and then declines to virtually undetectable levels. F9 cells can give rise to three distinct differentiated cell types; visceral endoderm, parietal endoderm, and primitive endoderm. We have clearly shown that HNF-3 alpha stimulation occurs upon primitive endoderm formation. In addition, the transcription factor is also activated during the induction of cell lineages that give rise to parietal and visceral endoderm. HNF-3 alpha stimulation upon visceral endoderm differentiation is accompanied by the activation of HNF-3 target genes such as transthyretin, suggesting that HNF-3 alpha is involved in the developmental activation of this gene. In contrast, HNF-3 alpha target genes in parietal and primitive endoderm have yet to be identified. However, the stimulation of HNF-3 alpha during primitive endoderm formation, which is an extremely early event during murine embryogenesis, points towards a role for the factor in crucial determination processes that occur early during development.

Control of gene expression by lipophilic hormones.

Regulation of gene expression in response to steroids, thyroid hormone, and retinoids is mediated by an impressive array of intracellular receptors. Sequence analysis showed that the hormone receptors comprise a large superfamily of ligand-responsive transcription factors. Upon binding to hormones, the receptors interact with specific hormone response elements located in the promoters of numerous genes. Promoter-bound receptors communicate with distinct receptors and/or additional members of the transcriptional machinery, frequently evoking protein-protein interactions. Ultimately, this results in the induction of complex gene systems that control hormone-induced processes such as differentiation, cell growth, and homeostasis. In addition to the genes transcribed by RNA polymerase II, the lipophilic hormones, particularly glucocorticoids, can also modulate RNA polymerase I-directed transcription of the ribosomal gene. For both transcription systems, activation and repression of genes in response to hormones have been reported. Finally, the involvement of hormone receptors in tumorigenesis has been discussed. It is likely that receptor studies will have major implications in the diagnosis and therapy of diseases such as leukemia.

Regulation of E2F/cyclin A-containing complex upon retinoic acid-induced differentiation of teratocarcinoma cells.

Retinoic acid-induced differentiation of mouse P19 teratocarcinoma cells is accompanied by alterations in the level of E2F transcription factor. P19 stem cells contain free, uncomplexed E2F and an E2F complex termed E2F/stem. This stem cell complex is a heterotrimeric protein aggregate consisting of E2F transcription factor, E2F-binding protein (E2F/bp1), and cyclin A. Retinoic acid treatment converts P19 stem cells into differentiated neurons, glial cells, and fibroblasts. The presented experiments clearly show that the level of uncomplexed E2F gradually decreases upon differentiation, and fully differentiated cells do not contain free E2F. In addition, the stem cell-specific E2F aggregate is converted into a smaller complex, termed E2F/diff. This smaller complex, which is specific for differentiated cells, does not contain cyclin A and consists of E2F transcription factor associated with E2F/bp1. Finally, the role of E2F complexes in the cessation of cell proliferation, which accompanies P19 cell differentiation, is discussed.