Histone deacetylase inhibitors reverse gene silencing in Friedreich's ataxia
426
Citation
43
Reference
10
Related Paper
Citation Trend
Keywords:
Frataxin
Histone deacetylase 5
Histone deacetylase 2
HDAC11
David Kadosh and Kevin Struhl Department of Biological Chemistry and Molecular Pharmacology, Harvard Medical School, Boston, Massachusetts 02115 USA
Histone deacetylase 5
HDAC11
HDAC4
Histone deacetylase 2
SAP30
HDAC10
Cite
Citations (235)
Histone deacetylase 2
HDAC11
Histone deacetylase 5
HDAC4
SAP30
HDAC10
Histone code
Histone octamer
Histone H4
Cite
Citations (288)
Histone acetyltransferase and histone deacetylase activities determine the acetylation status of histones, and have the ability to regulate gene expression through chromatin remodeling. A controlled balance between histone acetylation and histone deacetylation appears to be essential for normal cell growth. In cancer cells, some genes are repressed by inappropriate recruitment of histone deacetylases. The histone deacetylase inhibitors (HDACI) belong to the class of anticancer drugs that are effective in killing proliferating and non-proliferating tumor cells. In this review we discuss molecular mechanisms involved in the induction of cell cycle arrest, differentiation and induction of apoptosis in tumor cells by HDACI.
HDAC11
Histone deacetylase 5
HDAC4
Histone acetyltransferase
Histone deacetylase 2
HDAC10
SAP30
Histone H4
Cite
Citations (7)
SDS3 (suppressor of defective silencing 3) was originally identified in a screen for mutations that cause increased silencing of a crippled HMR silencer in arap1 mutant background. In addition, sds3mutants have phenotypes very similar to those seen in sin3and rpd3 mutants, suggesting that it functions in the same genetic pathway. In this manuscript we demonstrate that Sds3p is an integral subunit of a previously identified high molecular weight Rpd3p·Sin3p containing yeast histone deacetylase complex. By analyzing an sds3Δ strain we show that, in the absence of Sds3p, Sin3p can be chromatographically separated from Rpd3p, indicating that Sds3p promotes the integrity of the complex. Moreover, the remaining Rpd3p complex in the sds3Δ strain had little or no histone deacetylase activity. Thus, Sds3p plays important roles in the integrity and catalytic activity of the Rpd3p·Sin3p complex.
Histone deacetylase 5
HDAC11
HDAC10
Histone deacetylase 2
HDAC4
Histone H4
SAP30
Cite
Citations (97)
Recently we identified a new family of histone deacetylases in higher eukaryotes related to yeast HDA1 and showed their differentiation-dependent expression. Data presented here indicate that HDAC5 (previously named mHDA1), one member of this family, might be a potent regulator of cell differentiation by interacting specifically with determinant transcription factors. We found that HDAC5 was able to interact in vivo and in vitro with MEF2A, a MADS box transcription factor, and to strongly inhibit its transcriptional activity. Surprisingly, this repression was independent of HDAC5 deacetylase domain. The N-terminal non-deacetylase domain of HDAC5 was able to ensure an efficient repression of MEF2A-dependent transcription. We then mapped protein domains involved in the HDAC5-MEF2A interaction and showed that MADS box/MEF2-domain region of MEF2A interacts specifically with a limited region in the N-terminal part of HDAC5 which also possesses a distinct repressor domain. These data show that two independent class II histone deacetylases HDAC4 and HDAC5 are able to interact with members of the MEF2 transcription factor family and regulate their transcriptional activity, thus suggesting a critical role for these deacetylases in the control of cell proliferation/differentiation. Recently we identified a new family of histone deacetylases in higher eukaryotes related to yeast HDA1 and showed their differentiation-dependent expression. Data presented here indicate that HDAC5 (previously named mHDA1), one member of this family, might be a potent regulator of cell differentiation by interacting specifically with determinant transcription factors. We found that HDAC5 was able to interact in vivo and in vitro with MEF2A, a MADS box transcription factor, and to strongly inhibit its transcriptional activity. Surprisingly, this repression was independent of HDAC5 deacetylase domain. The N-terminal non-deacetylase domain of HDAC5 was able to ensure an efficient repression of MEF2A-dependent transcription. We then mapped protein domains involved in the HDAC5-MEF2A interaction and showed that MADS box/MEF2-domain region of MEF2A interacts specifically with a limited region in the N-terminal part of HDAC5 which also possesses a distinct repressor domain. These data show that two independent class II histone deacetylases HDAC4 and HDAC5 are able to interact with members of the MEF2 transcription factor family and regulate their transcriptional activity, thus suggesting a critical role for these deacetylases in the control of cell proliferation/differentiation. chloramphenicol acetyltransferase glutathione S-transferase Acetylation of chromatin proteins and transcription factors is part of a complex signaling system that is largely involved in the control of gene expression (1.Turner B.M. Cell Dev. Biol. 1999; 10: 165-167Google Scholar). Thus far, the specific involvement of histone acetyltransferases and deacetylases in the control of individual gene expression has been clearly established (2.Kuo M.H. Allis D. BioEssays. 1998; 20: 615-626Crossref PubMed Scopus (1073) Google Scholar, 3.Struhl K. Genes Dev. 1998; 12: 599-606Crossref PubMed Scopus (1557) Google Scholar). One major role of these enzymes is the control of cell differentiation in response to specific signals. Evidence exists for the participation of the RPD3-related members in this process (4.Johnson C.A. Turner B.M. Cell Dev. Biol. 1999; 10: 179-188Crossref PubMed Scopus (93) Google Scholar). Recently, however, a new family of higher eukaryotic histone deacetylases, distinct from the already characterized RPD3-related members, has been identified (5.Verdel A. Khochbin S. J. Biol. Chem. 1999; 274: 2440-2445Abstract Full Text Full Text PDF PubMed Scopus (214) Google Scholar, 6.Grozinger C.M. Hassig C.A. Schreiber S.L. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 4868-4873Crossref PubMed Scopus (661) Google Scholar, 7.Fischle W. Emiliani S. Hendzel M.J. Nagase T. Nomura N. Voelter W. Verdin E.A. J. Biol. Chem. 1999; 274: 11713-11720Abstract Full Text Full Text PDF PubMed Scopus (212) Google Scholar). These enzymes are related to yeast HDA1 histone deacetylase and within the cloned members, two show sequence homology and the same domain organization and are called HDAC4 and HDAC5 (5.Verdel A. Khochbin S. J. Biol. Chem. 1999; 274: 2440-2445Abstract Full Text Full Text PDF PubMed Scopus (214) Google Scholar, 6.Grozinger C.M. Hassig C.A. Schreiber S.L. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 4868-4873Crossref PubMed Scopus (661) Google Scholar). Despite this homology, HDAC4 and HDAC5 are probably capable of exerting distinct functions, since immunoprecipitation experiments showed that in cells they can be associated with different partners (6.Grozinger C.M. Hassig C.A. Schreiber S.L. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 4868-4873Crossref PubMed Scopus (661) Google Scholar). A member, named mHDA2/HDAC6, shows unique features within deacetylases, in that it possesses two HDA1 homology domains (5.Verdel A. Khochbin S. J. Biol. Chem. 1999; 274: 2440-2445Abstract Full Text Full Text PDF PubMed Scopus (214) Google Scholar, 6.Grozinger C.M. Hassig C.A. Schreiber S.L. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 4868-4873Crossref PubMed Scopus (661) Google Scholar). In contrast to the RPD3-related members, the expression of these genes is not ubiquitous. HDAC5 and mHDA2/HDAC6 expression is activated upon cell differentiation (5.Verdel A. Khochbin S. J. Biol. Chem. 1999; 274: 2440-2445Abstract Full Text Full Text PDF PubMed Scopus (214) Google Scholar). These observations suggest that members of the class II histone deacetylases may play a specific role in the regulation of cell differentiation. Looking for potential partners of HDAC5, we obtained evidence of interaction between HDAC5 and MEF2 transcription factors. We therefore focused our efforts on investigating this issue. The MEF2 family of transcription factors belongs to the large family of MADS box transcriptional regulators present from yeast to humans (8.Theissen G. Kim J.T. Saedler H. J. Mol. Evol. 1996; 43: 484-516Crossref PubMed Scopus (409) Google Scholar). Besides their established role in myogenesis (9.Black B.L. Olson E.N. Annu. Rev. Cell Dev. Biol. 1998; 14: 167-196Crossref PubMed Scopus (856) Google Scholar), MEF2 family members have been implicated in gene activation in response to mitogenic signaling (10, 11; for review, see Ref. 9.Black B.L. Olson E.N. Annu. Rev. Cell Dev. Biol. 1998; 14: 167-196Crossref PubMed Scopus (856) Google Scholar). Very recently, it has been shown that one member of the class II histone deacetylase, HDAC4, can interact with two different members of MEF2 family, MEF2A (12.Miska E.A. Karlsson C. Langley E. Nielsen S.J. Pines J. Kouzarides T. EMBO J. 1999; 18: 5099-5107Crossref PubMed Scopus (472) Google Scholar) and MEF2C (13.Wang H.A. Bertos N.R. Vezmar M. Pelletier N. Crosato M. Heng H.H. Th'ng J. Han J. Yang X.J. Mol. Cell. Biol. 1999; 19: 7816-7827Crossref PubMed Scopus (264) Google Scholar). Here we show that HDAC5, a distinct member of class II deacetylase, is also able to interact specifically with the MEF2A transcription factor and to repress its transcriptional activator capacity. Data presented here and the fact that the induction of various differentiation programs was found to be associated with an up-regulation of HDAC5 gene expression (5.Verdel A. Khochbin S. J. Biol. Chem. 1999; 274: 2440-2445Abstract Full Text Full Text PDF PubMed Scopus (214) Google Scholar), suggested that HDAC5 might be a general regulator of cell differentiation. We discuss here the possibility that HDAC5 might control the early stages of various differentiation programs by inhibiting the precocious activity of determinant transcription factors. MEF2-Luc, L8G5-Luc, and L8-Luc reporter plasmids were generated from MEF2-CAT1 (14.Yu Y-T. Breitbart R.E. Smoot L.B. Lee Y. Mahdavi V. Nadal-Ginard B. Genes Dev. 1992; 6: 1783-1798Crossref PubMed Scopus (385) Google Scholar), L8G5-CAT, and L8-CAT (15.Hollenberg S.M. Sternglanz R. Cheng P.F. Weintraub H. Mol. Cell. Biol. 1995; 15: 3813-3822Crossref PubMed Scopus (585) Google Scholar), by replacing the CAT reporter gene by luciferase (respective promoter regions were cloned in the pGL2 basic vector, Promega). PMT2-MEF2A (14.Yu Y-T. Breitbart R.E. Smoot L.B. Lee Y. Mahdavi V. Nadal-Ginard B. Genes Dev. 1992; 6: 1783-1798Crossref PubMed Scopus (385) Google Scholar), pcDNA-mHDA2/HDAC6 (5.Verdel A. Khochbin S. J. Biol. Chem. 1999; 274: 2440-2445Abstract Full Text Full Text PDF PubMed Scopus (214) Google Scholar), and LexA-VP16 (15.Hollenberg S.M. Sternglanz R. Cheng P.F. Weintraub H. Mol. Cell. Biol. 1995; 15: 3813-3822Crossref PubMed Scopus (585) Google Scholar) expression vectors have been described before. Deletions in HDAC5 were obtained by polymerase chain reaction. The resulting DNA fragments were cloned in pcDNA3.1 (Invitrogen) in-frame with a N-terminal HA-tag. GAL4 DNA-binding domain fusion protein constructs were generated by polymerase chain reaction. DNA fragments coding for HDAC5 amino acids 123–673 and 673–1113 were cloned in pcDNA-GAL4 DB, in-frame with the DNA-binding domain (amino acids 1–147) of GAL4. Mouse HDAC4 and human MITR cDNAs have been obtained by polymerase chain reaction from mouse embryo and human skeletal muscle cDNA libraries, respectively (Marathon ready cDNA,CLONTECH). HeLa cells were seeded in 6-well dishes (105 cells/well) and transfected 2 days later with Exgen 500 (Euromedex) according to the supplier's protocol. In each transfection, DNA amount was kept constant with pcDNA3 empty vector. Cell extracts were prepared 24 h post-transfection. Luciferase and β-galactosidase activities were measured using the "Luciferase Assay System" (Promega) and the Luminescent β-Gal Detection kit (CLONTECH), respectively. Glutathione S-transferase (GST) pull-down assays were performed as described (16.Hagemeier C. Casewell R. Hayhurst G. Sinclair J. Kouzarides T. EMBO J. 1994; 13: 2897-2903Crossref PubMed Scopus (159) Google Scholar). GST fusion proteins were produced in BL21 Escherichia coli transformed with pGEX-5X-3 plasmid (Amersham Pharmacia Biotech), encoding GST alone, or GST fused to various domains of MEF2A protein.35S-Labeled proteins were produced in rabbit reticulocyte lysate from pcDNA plasmids using the TNT transcription/translation kit (Promega) and [35S]methionine (Amersham Pharmacia Biotech). HeLa cells were transfected with Exgen500 as described above using 5 μg of pMT2-Mef2A or pMT2 vector and 5 μg of pcDNA-HDAC5 (HA tagged) or pcDNA empty plasmid per 100-mm plate. 24 h post-transfection, cells were washed in cold phosphate-buffered saline and lysed in 50 mm Tris, pH 8.0, 150 mm NaCl, 5 mm EDTA, 0.5% Nonidet P-40, 1 mm phenylmethylsulfonyl fluoride for 20 min on ice. After centrifugation (12,000 × g, 5 min, 4 °C) extracts were diluted with 1 volume of lysis buffer without Nonidet P-40 and incubated with 4 μg of polyclonal anti-MEF2 antibody for 1 h at 4 °C. 30-μl aliquots of protein G-Sepharose were added and the incubation was continued for 2 h. Precipitates were washed in lysis buffer containing 0.25% Nonidet P-40 and finally re-suspended in SDS-polyacrylamide gel electrophoresis loading buffer. Western blot analysis was performed using standard procedures and immunocomplexes were detected by chemiluminescence (ECL+, Amersham Pharmacia Biotech). Immunofluorescence was performed as described (17.Gorka C. Brocard M.P. Curtet S. Khochbin S. J. Biol. Chem. 1998; 273: 1208-1215Abstract Full Text Full Text PDF PubMed Scopus (11) Google Scholar). The following primary antibodies have been used: anti-MEF2 (C-21, Santa Cruz), anti-HA (3F10, Roche Molecular Biochemicals), and anti-HDAC5 (rabbit polyclonal antibodies raised against a peptide localized at the C-terminal region of the protein). In our previous work we reported the domain organization of HDAC5 histone deacetylase (5.Verdel A. Khochbin S. J. Biol. Chem. 1999; 274: 2440-2445Abstract Full Text Full Text PDF PubMed Scopus (214) Google Scholar). The histone deacetylase domain of this protein was found to be located at the C-terminal half of the protein and the N-terminal non-deacetylase domain did not show any significant homology with sequences present in the data base. In order to have insight into the function of HDAC5, we periodically searched the data bank for new sequences homologous to the N-terminal domain of HDAC5. This approach allowed us to identify a Xenopus cDNA presenting significant sequence homology with the non-deacetylase N-terminal domain of HDAC5 (accession number, Z97214, the putative protein encoded by this cDNA is now known as MITR, Ref. 18.Sparrow D.B. Miska E.A. Langley E. Reynaud-Deonauth S. Kotecha S. Towers N. Spohr G. Kouzarides T. Mohun T.J. EMBO J. 1999; 18: 5085-5098Crossref PubMed Scopus (177) Google Scholar). The GenBank sequence information presented MITR as a potential partner ofXenopus MEF2 homologues. The sequence homology between HDAC5 and MITR suggested that HDAC5 could also be a partner of MEF2 transcription factors. In order to investigate this possibility, we first expressed HDAC5 and MEF2A in HeLa cells and examined the formation of complexes bearing both proteins in vivo by co-immunoprecipitation. Extracts obtained from cells transfected with HA-tagged HDAC5, MEF2A, and both expression vectors were immunoprecipitated with anti-MEF2 antibodies. The presence of MEF2A and HA-HDAC5 in immunoprecipitated material was then monitored using anti-HA and anti-MEF2 antibodies. Fig. 1shows that anti-MEF2 antibodies are able to co-immunoprecipitate efficiently HDAC5 when both proteins are expressed in cells (Fig. 1,anti-HA panel, far right). The endogenous MEF2 in HeLa cells can also interact with HDAC5 since ectopically expressed HDAC5 was also found associated with endogenous MEF2, in the absence of exogenous MEF2A expression (Fig. 1). It is interesting to note that HeLa cells transfected with HDAC5-expression vector produced two HA-tagged HDAC5 proteins (Fig. 1, input panel) and within these two proteins only the full-length one could interact with MEF2A. In order to investigate the possibility of direct interaction between HDAC5 and MEF2A, we prepared bacterially expressed fusion proteins containing GST fused to different regions of MEF2A and examined the interaction of these proteins with 35S-labeled full-length HDAC5. Fig.2 A shows that a GST fusion protein harboring the N-terminal half of MEF2A is able to interact efficiently with HDAC5 (Fig. 2 A, GST-MEF2A 1–243), while the C-terminal half and the middle part of the protein did not recognize HDAC5 (Fig. 2 A, GST-MEF2A 244–507 and GST MEF2A 87–243). This observation suggested that the region encompassing MADS box and the so-called MEF2 domain, has the ability to interact with HDAC5. To know whether MADS box or MEF2 domain or both are involved in the interaction with HDAC5, we fused both or each of these domains to GST and analyzed the interaction with HDAC5 as above. These experiments showed that MADS box/MEF domain is sufficient to direct an efficient interaction with HDAC5 (Fig. 2 A, GST MEF2A 1–86). However, the MADS box alone was found to be able to interact weakly with HDAC5 (Fig. 2 A, GST MEF2A 1–57) while the MEF2 domain was not able to do so (Fig. 2 A, GST MEF2A 58–86). The specificity of this interaction was shown by considering the capacity of MEF2A N-terminal region to interact with two other deacetylases, HDAC1 and mHDA2/HDAC6. In this experiment we also used HDAC4 and MITR as known partners of MEF2. Data show in Fig. 2 B confirmed the interaction of these proteins with MEF2A as it has been reported very recently (12.Miska E.A. Karlsson C. Langley E. Nielsen S.J. Pines J. Kouzarides T. EMBO J. 1999; 18: 5099-5107Crossref PubMed Scopus (472) Google Scholar, 13.Wang H.A. Bertos N.R. Vezmar M. Pelletier N. Crosato M. Heng H.H. Th'ng J. Han J. Yang X.J. Mol. Cell. Biol. 1999; 19: 7816-7827Crossref PubMed Scopus (264) Google Scholar, 18.Sparrow D.B. Miska E.A. Langley E. Reynaud-Deonauth S. Kotecha S. Towers N. Spohr G. Kouzarides T. Mohun T.J. EMBO J. 1999; 18: 5085-5098Crossref PubMed Scopus (177) Google Scholar). However, no interaction between MEF2A and HDAC1 or mHDA2/HDAC6 was observed in these conditions (Fig. 2 B), showing the specific interaction between HDAC5 and MEF2A in our assays. To determine regions in HDAC5 involved in the interaction with MEF2A, expression vectors encoding HDAC5 deletion mutants were prepared and used to obtain corresponding 35S-labeled proteins. These proteins were incubated with a GST-MEF2A MADS box/MEF2 domain fusion protein and a pull-down was performed. A truncated HDAC5 molecule lacking the first 123 amino acids was capable of interacting with MEF2A, while the removal of additional 57 amino acids completely abolished the binding (Fig. 2 C, compare HDAC5 123–1113 and 180–1113 proteins). This observation showed that the region of 57 amino acids, located between amino acids 123 and 180 plays a major role in interaction with MEF2A. In order to confirm these findings, we expressed35S-labeled polypeptides corresponding to the 123–292 and 123–673 regions of HDAC5 and found that both interact efficiently with MEF2A (Fig. 2 C, HDAC5 123–292 and HDAC5 123–673, respectively). In these assays the deacetylase domain of HDAC5 could not interact with MEF2A (Fig. 2 C, HDAC5 674–1113). These experiments showed that a limited region of HDAC5 encompassing amino acids 123–180 plays an essential role in the direct interaction with MEF2A. A transient transfection assay was designed to evaluate the role of HDAC5-MEF2A interaction on the transcriptional activity of MEF2A. The reporter system used consisted of the minimal promoter region of myosin heavy chain gene flanked by two MEF2-binding sites (14.Yu Y-T. Breitbart R.E. Smoot L.B. Lee Y. Mahdavi V. Nadal-Ginard B. Genes Dev. 1992; 6: 1783-1798Crossref PubMed Scopus (385) Google Scholar), cloned upstream of a luciferase gene. In HeLa cells, the expression of MEF2A led to a 20-fold activation of reporter gene expression (Fig. 3 A). Co-expression of MEF2A and HDAC5 showed a very efficient inhibition of the activator capacity of MEF2A. Indeed, the use of as little as 1 or 5 ng of HDAC5 expression vector in these co-transfection assays, repressed significantly MEF2A activity (Fig. 3 A). We then showed that the repression of MEF2A activity by HDAC5 was as efficient as that observed by HDAC4 (12.Miska E.A. Karlsson C. Langley E. Nielsen S.J. Pines J. Kouzarides T. EMBO J. 1999; 18: 5099-5107Crossref PubMed Scopus (472) Google Scholar). However, the capacity to repress MEF2A was not shared by all members of class II histone deacetylases, since the expression of the third member of this family (mHDA2/HDAC6) did not alter MEF2A activity (Fig. 3 B). We then tried to show whether HDAC5-MEF2A interaction is necessary to repress MEF2A activity. HeLa cells were co-transfected with the MEF2 responsive reporter, a MEF2A expression vector and vectors expressing different deletion mutants of HDAC5. As shown before, the expression of full-length HDAC5 repressed efficiently MEF2A activity (Fig. 3 D, 1–1113 construct). Interestingly, this repression was found to be independent of the deacetylase domain. Indeed a truncated HDAC5 bearing only the non-deacetylase domain of the protein was almost as efficient in repressing the MEF2A transcriptional activity as the full-length HDAC5 (Fig. 3 D, compare 1–1113 and 123–673 constructs). This repression was found to be dependent on MEF2A-HDAC5 interaction, since a deletion mutant lacking the N-terminal region, defined to be the site of interaction with MEF2A, was absolutely inefficient in repression (Fig. 3 D, 180–1113 construct). However, HDAC5/MEF2A interaction alone was not sufficient to repress transcription since two constructs expressing truncated versions of the N-terminal non-deacetylase domain of HDAC5, able to interact with MEF2Ain vitro, were unable to repress transcription (Fig.3 D, 123–292 and Δ241–292 constructs, respectively). This experiment showed that the N-terminal region of HDAC5 possesses in addition to the MEF2-binding domain, a distinct repression domain. This region is localized C-terminal to the MEF2-binding domain and covers a region that encompasses amino acids 241–292 of HDAC5. Indeed, the deletion of this region abolished HDAC5 repressive activity. However, amino acids located C-terminal to this region are also important for this repressive activity, since the 123–292 version of HDAC5 were unable to repress transcription. The expression of the deacetylase domain alone did not show any repressive activity (Fig. 3 D, 674–1113 construct). Western blots presented in Fig. 3, Cand E, show that the vectors used in these experiments express efficiently all the expected proteins. However, it appeared that most of the constructs used, besides the expected protein (Fig.3 E, arrows), produced at least another protein. We could show that these products are not a result of degradation but are generated after the splicing of a fraction of the messenger transcribed from the plasmid. Indeed the use of different antibodies and the cloning of a mRNA generated after transfection, showed that the full-length mRNA transcribed from the plasmid can undergo splicing and generate variants (not shown and see Fig. 5 B). Since the deacetylase domain of HDAC5 was found to be unnecessary to repress MEF2A transcriptional activity, one could question the ability of this domain to repress transcription. In order to investigate this issue, we targeted the HDAC5 deacetylase domain into a promoter containing GAL4-binding sites. The reporter system used contained eight copies of the binding sites for LexA immediately adjacent to five copies of the binding site for GAL4 (15.Hollenberg S.M. Sternglanz R. Cheng P.F. Weintraub H. Mol. Cell. Biol. 1995; 15: 3813-3822Crossref PubMed Scopus (585) Google Scholar), all cloned upstream of a luciferase gene. In the presence of LexA-VP16 fusion co-activator and the GAL4 DNA-binding domain alone (GAL4-DB), this reporter was activated to high levels of expression (Fig.4, GAL4-DB). An expression vector was prepared expressing a fusion protein containing the deacetylase domain of HDAC5 (amino acids 674–1113) fused to the GAL4 DNA-binding domain. Co-expression of LexA-VP16 and GAL4-HDAC5 showed that the HDAC5 deacetylase domain could inhibit by 10-fold the transcriptional activity of LexA-VP16 (Fig. 4, HDAC5 674–1113 construct). We showed also that the non-deacetylase domain of HDAC5 possesses an independent repression domain capable of repressing MEF2A transcriptional activity (Fig. 3 D). To confirm this finding we also fused HDAC5 N-terminal non-deacetylase domain (amino acids 123–673) to the GAL4 DNA-binding domain. The expression of this fusion protein together with LexA-VP16 led also to an efficient repression of LexA-VP16 transcriptional activity (Fig. 4,HDAC5 123–673 construct). In order to show the specificity of this repressive activity, we prepared an expression vector encoding GAL4 DNA-binding domain fused to the MEF2-binding domain of HDAC5 (amino acid 123–292). This region of HDAC5 possesses the ability to interact with MEF2A (Fig. 2 C) but does not repress its transcriptional activity (Fig. 3 D). Therefore, it should not be able to repress the VP16-dependent transcriptional activation. As expected, the expression of this protein did not affect the ability of LexA-VP16 to direct an efficient transcription of the reporter gene (Fig. 4, HDAC5 123–292). These data showed that the HDAC5 deacetylase domain, although unnecessary to repress MEF2A transcriptional activity, is an active repressor domain. They also confirmed that besides the deacetylase domain, HDAC5 possesses another repression domain located in its N-terminal region. The present report and those published recently indicate that HDAC4 and HDAC5 are able to interact with members of MEF2 transcription factors. These two members of class II deacetylases, although sharing similar domain organization and sequence homology, may be able to ensure different functions in cells, as they have been found to be associated with distinct partners (6.Grozinger C.M. Hassig C.A. Schreiber S.L. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 4868-4873Crossref PubMed Scopus (661) Google Scholar). In order to obtain more evidence of functional differences between these two deacetylases, we transfected HeLa cells with HA-tagged HDAC4 and HDAC5 expression vectors and examined the cellular distribution of these proteins using an anti-HA antibody. In cells transfected with HDAC4 expression vector a significant proportion of cells showed foci of HDAC4 accumulation in the nucleus (Fig.5 A, top left), as it has been previously shown (7.Fischle W. Emiliani S. Hendzel M.J. Nagase T. Nomura N. Voelter W. Verdin E.A. J. Biol. Chem. 1999; 274: 11713-11720Abstract Full Text Full Text PDF PubMed Scopus (212) Google Scholar, 12.Miska E.A. Karlsson C. Langley E. Nielsen S.J. Pines J. Kouzarides T. EMBO J. 1999; 18: 5099-5107Crossref PubMed Scopus (472) Google Scholar) while none of the cells transfected with HA-HDAC5 expression vectors showed this particular pattern of expression (Fig. 5 A, top right). In the majority of these cells, a homogenous distribution of HA-HADC5 was observed in the nucleus (Fig. 5 A, HDAC5 panel). Moreover, Western blot analysis of cells transfected with HA-HDAC5 expression vectors showed the expression of two major forms of the protein, one presenting the expected size and the other was found to have a molecular mass around 83 kDa (Fig. 5 B, HDAC5 panel, α-HA, arrows). The use of an antibody raised against the very C-terminal region of HDAC5 revealed the same bands (Fig. 5 B, HDAC5 panel, α-HDAC5, arrows). This experiment showed that the smaller band is not the result of proteolysis but rather is an altered form of the protein, possessing both the N-terminal HA-tag and the C-terminal epitope. This observation suggests that a fraction of the messenger produced from the vector undergo splicing giving rise to HDAC5 variants. Indeed, we cloned and sequenced a cDNA produced after transfection and confirmed the above hypothesis (not shown). Interestingly, the smaller HDAC5-related protein is not able to interact with MEF2A in vivo (Fig. 1, compare, input and IP α-MEF2). It appeared therefore that transfected cells are capable of generating at least one form of HDAC5 that is unable to interact with MEF2A transcription factor. We are currently analyzing the presence of such HDAC5 variants produced from the endogenous HDAC5 gene. Western blot analysis of cell transfected with HA-HDAC4 showed the presence of a single band at the expected size. These data show that HDAC4 and HDAC5, although both interacting with MEF2A and modulating its activity, are different proteins and are probably involved in distinct regulatory pathways. Within class II histone deacetylases, HDAC4 and HDAC5 encoded by two independent genes, show similar domain organization and sequence homology (6.Grozinger C.M. Hassig C.A. Schreiber S.L. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 4868-4873Crossref PubMed Scopus (661) Google Scholar). However, the analysis of associated proteins showed that in cells these deacetylases might enter functionally distinct complexes. Indeed, HDAC4, but not HDAC5 was found associated with RbAp48 (6.Grozinger C.M. Hassig C.A. Schreiber S.L. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 4868-4873Crossref PubMed Scopus (661) Google Scholar). Data presented here further support this idea and show that when expressed in cells, HDAC4 and HDAC5 present very different patterns of nuclear localization. Moreover, an altered form of HDAC5, unable to interact with MEF2A could be generated in cells after transfection. Differentially spliced forms of HDAC5 have been already observed in different tissues and cell lines (5.Verdel A. Khochbin S. J. Biol. Chem. 1999; 274: 2440-2445Abstract Full Text Full Text PDF PubMed Scopus (214) Google Scholar). 2C. Lemercier, A. Verdel, B. Galloo, S. Curtet, M.-P. Brocard, and S. Khochbin, unpublished observations. All these observations suggest that HDAC4 and HDAC5 exert distinct functions in cells and are probably involved in different regulatory pathways. Our data showed, however, that HDAC5, like HDAC4, is capable of interacting specifically with the MEF2A transcription factor. This interaction completely abolished the transcriptional activity of MEF2A. We were surprised to find that this repression was independent of the deacetylase activity of HDAC5. Indeed, the N-terminal region of HDAC5 was found to interact with MEF2A and repress its transcriptional activity. Moreover, we found that MEF2A interaction and transcriptional repression are ensured by distinct regions of HDAC5 N-terminal non-deacetylase domain. The repression domain in this region of HDAC5 might recruit cellular histone deacetylases to ensure its repressive activity. Indeed, it has been shown that MITR, which shows significant sequence homology with the non-deacetylase domain of HDAC5 can interact with HDAC1 (18.Sparrow D.B. Miska E.A. Langley E. Reynaud-Deonauth S. Kotecha S. Towers N. Spohr G. Kouzarides T. Mohun T.J. EMBO J. 1999; 18: 5085-5098Crossref PubMed Scopus (177) Google Scholar) and moreover, both HDAC4 and HDAC5 were found to be associated with HDAC3 (6.Grozinger C.M. Hassig C.A. Schreiber S.L. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 4868-4873Crossref PubMed Scopus (661) Google Scholar). One may question the role of the histone deacetylase domain of HDAC5 in the activity of this protein. We showed here that this domain is fully functional in repressing transcription when it is targeted to a promoter via the GAL4 DNA-binding domain. We believe therefore that HDAC5 is a multifunctional repressor; it is capable of a repressive activity, independent of or dependent on its own deacetylase domain. However, one should keep in mind that the deacetylase activity of HDAC5 may be required to repress transcriptional activity of endogenous MEF2A-responsive genes. MEF2A, and other transcription factors, may target HDAC5 into chromatin where the deacetylase domain could participate in a chromatin remodeling activity. Moreover, HDAC5 like RPD3-related members, may enter distinct multifunctional complexes via "adaptors." Indeed, evidence does exist that HDAC4 can form a complex when expressed in HeLa cells (7.Fischle W. Emiliani S. Hendzel M.J. Nagase T. Nomura N. Voelter W. Verdin E.A. J. Biol. Chem. 1999; 274: 11713-11720Abstract Full Text Full Text PDF PubMed Scopus (212) Google Scholar). The way HDAC5 and HDAC4 are used in the regulation of gene expression appears therefore to be similar to that of HDAC1 and HDAC2. Indeed, these RPD3-related deacetylases are capable of interacting directly with transcription factors, i.e. YY1, Sp1, etc. (19.Yang W.M. Inouye C. Zeng Y. Bearss D. Seto E. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 12845-12850Crossref PubMed Scopus (485) Google Scholar, 20.Doetzlhofer A. Rotheneder H. Lagger G. Koranda M. Kurtev V. Brosch G. Wintersberger E. Seiser C. Mol. Cell. Biol. 1999; 19: 5504-5511Crossref PubMed Scopus (358) Google Scholar) and/or they may enter various complexes via interaction with molecules such as Sin3. Sin3 seems to play a pivotal role in targeting RPD3-related HDACs into various and distinct complexes (4.Johnson C.A. Turner B.M. Cell Dev. Biol. 1999; 10: 179-188Crossref PubMed Scopus (93) Google Scholar). However, Sin3 was not found associated with HDAC4 and HDAC5 (6.Grozinger C.M. Hassig C.A. Schreiber S.L. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 4868-4873Crossref PubMed Scopus (661) Google Scholar) indicating that other molecules may accomplish a Sin3-like function in targeting HDAC5 and HDAC4 into different complexes. Interestingly, HDAC4, HDAC5, and a new member of the class II histone-deacetylase, HDAC7, were found to interact directly with corepressors N-CoR and SMRT (21.Kao H.Y. Downes M. Ordentlich P. Evans R.M. Genes Dev. 2000; 14: 55-66PubMed Google Scholar, 22.Huang E.Y. Zhang J. Miska E.A. Guenther M.G. Kouzarides T. Lazar M. Genes Dev. 2000; 14: 45-54PubMed Google Scholar), confirming the fact that these new deacetylases can target various complexes in a Sin3-independent fashion. The fact that RbAp48 was found to form a complex specifically with HDAC4 (6.Grozinger C.M. Hassig C.A. Schreiber S.L. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 4868-4873Crossref PubMed Scopus (661) Google Scholar) provides some hints concerning the nature of complexes containing this deacetylase. Indeed, RbAp48, besides its involvement in NRD and Sin3 complexes (23.Tong J.K. Hassig C.A. Schnitzler G.R. Kingston R.E. Schreiber S.L. Nature. 1998; 395: 917-921Crossref PubMed Scopus (552) Google Scholar, 24.Zhang Y. LeRoy G. Seelig H.P. Lane W.S. Reinberg D. Cell. 1998; 95: 279-289Abstract Full Text Full Text PDF PubMed Scopus (688) Google Scholar), where HDAC4 was not found (6.Grozinger C.M. Hassig C.A. Schreiber S.L. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 4868-4873Crossref PubMed Scopus (661) Google Scholar), is associated with chromatin assembly machinery (25.Parthun M.R. Widom J. Gottschling D.E. Cell. 1996; 87: 85-94Abstract Full Text Full Text PDF PubMed Scopus (379) Google Scholar,26.Verreault A. Kaufman P.D. Kobayashi R. Stillman B. Cell. 1996; 87: 95-104Abstract Full Text Full Text PDF PubMed Scopus (528) Google Scholar). A role for HDAC4 in histone metabolism is therefore possible. The inhibition of MEF2A and MEF2C by HDAC4 (12.Miska E.A. Karlsson C. Langley E. Nielsen S.J. Pines J. Kouzarides T. EMBO J. 1999; 18: 5099-5107Crossref PubMed Scopus (472) Google Scholar, 13.Wang H.A. Bertos N.R. Vezmar M. Pelletier N. Crosato M. Heng H.H. Th'ng J. Han J. Yang X.J. Mol. Cell. Biol. 1999; 19: 7816-7827Crossref PubMed Scopus (264) Google Scholar), as well as that of MEF2A by HDAC5 reported here, indicate the possible involvement of these proteins in various regulatory pathways. Since MEF2 transcription factors regulate the activity of numerous muscle-specific genes and play an important role in myogenesis, HDAC4/HDAC5 could be considered as potential regulators of myogenesis. However, a different role for these two proteins in myogenic differentiation could be envisaged since we have observed spliced forms of HDAC5 messenger (but not HDAC4) in myoblasts. 3C. Lemercier, A. Verdel, B. Galloo, S. Curtet, M.-P. Brocard, and S. Khochbin, unpublished data. Moreover, since MEF2 members are also involved in the regulation of cell proliferation, one might also propose a connection between HDAC4/HDAC5 activity and MEF2 function in this process. Previously, we showed that the expression of HDAC5 was up-regulated during a variety of differentiation programs (5.Verdel A. Khochbin S. J. Biol. Chem. 1999; 274: 2440-2445Abstract Full Text Full Text PDF PubMed Scopus (214) Google Scholar). A careful examination of the timing of HDAC5 expression during the induced differentiation of murine erythroleukemia cells, showed a transient activation of HDAC5 gene expression (5.Verdel A. Khochbin S. J. Biol. Chem. 1999; 274: 2440-2445Abstract Full Text Full Text PDF PubMed Scopus (214) Google Scholar). This accumulation occurs between 6 and 12 h of induction and is followed by decay in mRNA content. This period precedes the commitment and the induction of globin gene expression (17.Gorka C. Brocard M.P. Curtet S. Khochbin S. J. Biol. Chem. 1998; 273: 1208-1215Abstract Full Text Full Text PDF PubMed Scopus (11) Google Scholar). The same transient induction of HDAC5 gene expression was observed during the in vitro differentiation of myoblasts to myotubes.3 It is therefore tempting to think that HDAC5 plays a role in the pre-commitment period of various differentiation programs and functions as an early and transient inhibitor of specific transcription factors. We are grateful to Dr. Jean Jacques Lawrence, the head of INSERM U309 for encouraging this work, to Dr. S. F. Konieczny for MEF2 plasmids and Dr. Mary Callanan for the critical reading of this manuscript. We thank Dr. Duncan Sparrow and Xiang-Jiao Yang for communicating results prior to publication and Amandine Jaulmes for technical assistance.
Histone deacetylase 5
HDAC4
Mef2
Histone deacetylase 2
Histone acetyltransferase
HDAC10
SAP30
Cite
Citations (203)
Regulation of gene expression by mitogen-activated protein kinases (MAPKs) is essential for proper cell adaptation to extracellular stimuli. Exposure of yeast cells to high osmolarity results in rapid activation of the MAPK Hog1, which coordinates the transcriptional programme required for cell survival on osmostress. The mechanisms by which Hog1 and MAPKs in general regulate gene expression are not completely understood, although Hog1 can modify some transcription factors. Here we propose that Hog1 induces gene expression by a mechanism that involves recruiting a specific histone deacetylase complex to the promoters of genes regulated by osmostress. Cells lacking the Rpd3-Sin3 histone deacetylase complex are sensitive to high osmolarity and show compromised expression of osmostress genes. Hog1 interacts physically with Rpd3 in vivo and in vitro and, on stress, targets the deacetylase to specific osmostress-responsive genes. Binding of the Rpd3-Sin3 complex to specific promoters leads to histone deacetylation, entry of RNA polymerase II and induction of gene expression. Together, our data indicate that targeting of the Rpd3 histone deacetylase to osmoresponsive promoters by the MAPK Hog1 is required to induce gene expression on stress. PMID: 14737171
Histone deacetylase 5
HDAC11
HDAC4
HDAC10
SAP30
Histone deacetylase 2
HDAC1
RNA polymerase II
Cite
Citations (0)
Histone deacetylase 5
Histone deacetylase 2
HDAC11
SAP30
HDAC10
HDAC4
Histone acetyltransferase
Histone code
Cite
Citations (30)
HDAC11
Histone deacetylase 5
HDAC10
SAP30
Histone deacetylase 2
HDAC4
Histone acetyltransferase
Cite
Citations (17)
I. Class I Histone Deacetylases Histone Deacetylase 1 Dominique Meunier and Christian Seiser Biochemistry of Multiprotein HDAC Complexes Alejandro Vaquero, Michael Scher, and Danny Reinberg The Biology of HDAC3 Edward Seto The Biology of HDAC8, a Unique Class I Histone Deacetylase David Waltregny and Vincent Castronovo II. Class II Histone Deacetylases Regulation of Muscle Gene Expression by Histone Deacetylases Timothy A. McKinsey and Eric N. Olson The Class IIa Histone Deacetylases: Functions and Regulation Herbert G. Kasler and Eric Verdin Histone Deacetylases in the Response to Misfolded Proteins J. Andrew McKee and Tso-Pang Yao III. Class III Histone Deacetylases Comparison of Sirtuin Sequences Between Archaea and Vertebrates Roy A. Frye Structure of the Sir2 Family of NAD+-Dependent Histone/Protein Deacetylases Kehao Zhao and Ronen Marmorstein The Enzymology of SIR2 Proteins Margie T. Borra and John M. Denu The Class III Protein Deacetylases: Homologs of Yeast Sir2p Bjoern Schwer, Brian J. North, Nidhi Ahuja, Brett Marshall, and Eric Verdin IV. Histone Deacetylase Inhibitors HDAC Inhibitors: Discovery, Development, and Clinical Impacts Akihiro Ito, Norikazu Nishino, and Minoru Yoshida Cell Cycle Targets of Histone Deacetylase Inhibitors Brian Gabrielli HDAC Inhibitors: An Emerging Anticancer Therapeutic Strategy Paul Kwon, Meier Hsu, Dalia Cohen, and Peter Atadja Index
Histone deacetylase 5
HDAC11
Histone deacetylase 2
HDAC10
HDAC4
SAP30
HDAC8
Histone H4
Cite
Citations (20)