The goal of this project is to determine the mechanism by which activation of cAMP signaling pathways leads to transcriptional repression of the mouse mammary tumor virus (MMTV) promoter in the physiological context of ordered chromatin. Our previous work has shown that cAMP signaling causes repression of the MMTV promoter in a glucocorticoid-independent fashion only when it is incorporated into ordered, replicating chromatin in cultured cells. In contrast, a transiently-transfected MMTV promoter construct which does not have an organized nucleoprotein structure is activated by cAMP (cyclic AMP) signaling. These results emphasize the importance of studying transcriptional regulation of genes in their native setting of complex chromatin. We have recently determined that the two templates are regulated through different cAMP-dependent mechanisms. Thus the differences in response to cAMP signaling are not due to the same cAMP signaling cascade having two different outcomes due to the distinct structures of the templates. Rather, the structure of the template determines which cascades targets it. This model might be used to explain cell type-specific differences in regulation of the same promoter. If the chromatin structure of the promoter is in distinct configurations in two different cell types, the promoter may respond differently to the same extracellular signal. In the past year we have made significant progress in defining the mechanism by which the MMTV promoter in ordered chromatin is repressed by cAMP signaling. We approached this question through comparison to another agent which represses the promoter, the histone deacetylase inhibitor trichostatin A (TSA). Like cAMP, TSA represses the promoter in a chromatin-specific fashion. However, previous work had indicated that the repression by TSA occurred through a mechanism involving the inhibition of chromatin remodeling machinery required for transcription of the MMTV promoter in ordered chromatin. In contrast, cAMP represses the promoter without affecting the obvious changes in chromatin remodeling. We have shown that the inhibition of the chromatin remodeling machinery is actually a secondary effect of TSA and that the mechanism of TSA repression is very similar to that of cAMP. In both cases, transcription is rapidly inhibited in a fashion independent of changes in chromatin remodeling at the template. Since the primary effect of TSA on cellular metabolism is to inhibit deacetylases and effectively raise the level of histone acetylation, we reasoned that cAMP may have effects on histone acetylation at the MMTV promoter.We used the chromatin immunoprecipitation (ChIP) method to examine the acetylation state of histones at the MMTV promoter. In this method, chromatin from treated cells is fragmented and then exposed to antibodies specific for various acetylated forms of the histones H3 and H4. Chromatin fragments bound to the antibodies are immunoprecipitated. DNA is then isolated from the bound fraction and examined for the presence of particular DNA sequences by PCR. Using this method we determined that different histone acetylation patterns are induced by signaling pathways which either activate or represses the MMTV promoter. When the promoter is activated by glucocorticoids, we observed an increase in H4 acetylation and a decrease in H3 acetylation. In contrast, cAMP signaling caused significant increases in histone H3 acetylation. We were able to determine that all the changes we observed in H3 acetylation were limited to a specific lysine residue (K14) in the amino terminal tail region. Our results indicate that, contrary to the popular notion that increased histone acetylation is invariably associated with transcriptional activation, histone acetylation can also play a role in transcriptional repression. Our work also demonstrates that all histone acetylation events are not functionally equivalent since histone H4 acetylation increased in the activated state while histone H3 acetylation increased in the repressed state. Thus the two histones may have distinct functions at a particular promoter.We were interested in how cAMP signaling caused increased acetylation of histone H3. Since histone H3 is known to be phosphorylated at a residue (serine 10) very close to K14, one possible mechanism would be that a cAMP-dependent kinase cascade results in phosphorylation of histone H3, which then marks the template for the action of a histone acetyltransferase (HAT) specific for histone H3. The phosphorylation of serine 10 may make the histone tail region a better substrate for the HAT. We used the ChIP assay with antibodies specific for phosphoH3 to show that cAMP signaling does cause the phosphorylation of histone H3 at serine 10. In an effort to determine whether this phosphorylation event is carried out directly by the cAMP-activated kinase, PKA, we carried out kinase assays on mononucleosomes and found that PKA efficiently phosphorylates H3 at serine 10 in this context. While we have not proven unequivocally that PKA is the in vivo kinase which carries out this reaction, we have definitively shown that it is highly capable of phosphorylating H3 at the in vivo site.Our work defines a novel mechanism of transcriptional repression in which acetylation and phosphorylation of histones play a causative role. Much of the literature on cAMP signaling and transcription focuses on activation of transcription at promoters containing binding sites for the transcription factor CREB (cAMP response element binding protein). Regulation of transcription by this pathway has been defined by transcription factor interactions. The mechanism by which the MMTV promoter in ordered chromatin is repressed by cAMP signaling is independent of CREB and CBP (CREB binding protein) and involves direct modification of chromatin. We are the first to demonstrate that two different modifications of the same histone occur at a promoter targeted in response to a signal transduction pathway.Our study of cAMP-induced repression of the MMTV promoter has led to another interesting line of investigation. In an effort to determine whether CBP was involved in this mechanism we expressed the adenovirus protein E1A in our cell lines. This protein is thought to bind CBP and inhibit its function in transcription. Unexpectedly we found that E1A has two effects on transcriptional regulation of the MMTV promoter in ordered chromatin. First, it significantly stimulates both basal and glucocorticoid-induced transcription. In addition, it alleviates repression of the promoter by cAMP. We have determined that these effects are mediated by the 13S form of the E1A proteins. Using forms of E1A mutated at various amino acid residues necessary for binding of CBP/p300, Rb (retinoblastoma protein), and the HAT, PCAF, we have determined that CBP binding plays a role in the alleviation of cAMP-induced repression. Our preliminary hypothesis is that E1A is somehow targeted directly to the promoter and thereby brings CBP to the template. The presence of CBP then interferes with the repressive cAMP signal. We plan to use ChIP assays to examine the effects of E1A on the histone acetylation patterns at the promoter. In addition, we will carry out additional mutational analysis to determine how E1A is targeted to the template and how it stimulates transcription in general. Our preliminary results indicate that E1A may inactivate CBP at some promoters as has been reported in the literature but it may re-target it to others where it affects transcriptional regulation.
| Rodriguez-Collazo, Pedro; Snyder, Sara K; Chiffer, Rebecca C et al. (2008) cAMP signaling induces rapid loss of histone H3 phosphorylation in mammary adenocarcinoma-derived cell lines. Exp Cell Res 314:1-10 |
| Mulholland, Niveen M; Snyder, Sara K; Kolla, Sarah S et al. (2003) Chromatin-dependent regulation of the MMTV promoter by cAMP signaling is mediated through distinct pathways. Exp Cell Res 287:361-73 |
