Cloning and Characterization of the Mouse Histone Deacetylase-2 Gene*

Histone deacetylase-2 (HDAC2) is a component of a complex that mediates transcriptional repression in mammalian cells. A mouse HDAC2 cDNA was used to identify several recombinant clones containing the entire mouse HDAC2 gene. The mouse HDAC2 gene spans over 36 kilobase pairs and is composed of 14 exons (ranging from 58 to 362 nucleotides in length) and 13 introns (ranging from 75 base pairs to 19 kilobase pairs in length). Primer extension analysis with total RNA from NIH3T3 cells revealed a major transcriptional start site at 221 base pairs 5′ of the ATG translational start codon. Upstream of the transcriptional start site, no canonical TATA box was found, but binding sites for several known transcription factors were identified. Transient transfection studies with 5′ deletion mutants localized the promoter to no more than 76 base pairs upstream from the major transcriptional start site. Fluorescence in situhybridization mapped mouse HDAC2 to chromosomal location 10B1, which is in close proximity to the growth factor-inducible genefisp-12. Information concerning the genomic organization and promoter of HDAC2 will be useful in studies of the regulation of histone deacetylase activities, which in turn are important in studies of the regulation of transcriptional repression in mammalian cells.

A key event in the regulation of eukaryotic gene expression is the posttranslational modification of nucleosomal histones, which converts regions of chromosomes into transcriptionally active or inactive chromatin. The most common posttranslational modification of histones is the acetylation of ⑀-amino groups on conserved lysine residues in the amino-terminal tail domains of the histones. Hyperacetylation of histones generally correlates with transcriptionally active chromatin, perhaps by increasing the accessibility of transcription factors to nucleosomal DNA, whereas hypoacetylation of histones correlates with transcriptional silencing.
An acetyltransferase enzyme responsible for histone acetylation, HAT A, 1 was first identified in Tetrahymena (1). HAT A possesses a high degree of amino acid sequence similarity to the yeast protein GCN5, which also catalyzes histone acetylation. Subsequently, several very well characterized transcription factors, p300/CBP, TAF II 230/250, PCAF, SRC-1, and ACTR, were found to contain histone acetylating activities (2)(3)(4)(5)(6)(7). In addition, several newly discovered histone acetyltransferases (e.g. MOF, ESA1, and TIP60) were recently found in yeast, Drosophila, and humans (8,9). Each of these histone acetyltransferases may have a particular substrate specificity. For example, whereas p300/CBP can acetylate all four core histones, GCN5 acetylates only histones H3 and H4. Further, different acetylases are specific with regard to which histone amino acids they will acetylate. Most interestingly, MOF appears to acetylate a particular lysine on histone H4 along the X chromosome but does not affect histones on other chromosomes.
Because histone acetylation and deacetylation play equally crucial roles in gene regulation, a complete picture of how transcriptional regulation is achieved requires that deacetylase enzymes be identified and their mechanisms clearly understood. Similar to acetyltransferases, genes that encode histone deacetylases have eluded identification until recent years. In yeast, the HDA1 protein, which shares sequence similarity with RPD3, is a subunit of a large histone deacetylase complex, HDA (10). RPD3 is also associated with another yeast histone deacetylase complex, HDB. Using a trapoxin (an inhibitor of histone deacetylase) affinity matrix, Taunton et al. (11) purified and cloned a human 55-kDa protein related to the yeast protein RPD3. Immunoprecipitation of this 55-kDa protein, HD1 (later renamed HDAC1 (12)), showed that it contains histone deacetylase activity. Pure recombinant HDAC1 also deacetylates histones in the absence of protein cofactors (13). HDAC1 binds to the retinoblastoma gene product and to MeCP2 to repress transcription (14 -18). It is also important in repressing transactivation mediated by the progesterone receptor (19) and plays a crucial role in human acute promyelocytic leukemia (20,21). A second mammalian histone deacetylase protein, mRPD3 (renamed HDAC2 (22)), with high homology to yeast RPD3 was identified in our laboratory based on a yeast two-hybrid experiment with the YY1 transcription factor as a bait (23). YY1 negatively regulates transcription by tethering HDAC2 to DNA as a corepressor. Both HDAC1 and HDAC2 exist in a complex with the corepressor mSIN3 and mediate Mad transcriptional repression (12,22,24). In addition, HDAC1 and HDAC2 are essential components of two thyroid hormone receptor corepressors, N-CoR and SMRT (25)(26)(27). Finally, a third human RPD3-related protein, HDAC3, which possesses histone deacetylase activity, was recently cloned (28,29). Like HDAC1 and HDAC2, HDAC3 represses transcription, binds transcription factor YY1, and is ubiquitously expressed in many different cell types (28).
Although increasing evidence suggests that the mammalian RPD3 proteins are closely linked to histone deacetylation and transcriptional repression, little is known concerning how deacetylase levels are regulated. The HDAC mRNAs and proteins are present in many different tissues and cell types, and thus HDACs can be generally regarded as ubiquitous enzymes (28). 2 Nevertheless, there are clearly situations in which mammalian deacetylase is regulated. For example, expression of the mouse HDAC1 mRNA can be induced by interleukin-2 in murine T cells (30). In peripheral blood mononuclear cells, HDAC3 mRNA increased with activation of the cells by PHA, phorbol 12-myristate 13-acetate, and ␣-CD3 but was down-regulated in the presence of granulocyte-macrophage colony-stimulating factor (31). In addition, the mouse HDAC2 mRNA is expressed in lower levels in the mouse spleen compared with other tissues and might be induced by phorbol ester treatment. 2 Most interestingly, HDAC2 is expressed in very low levels in Jurkat T cells (13). As a first step toward understanding how histone deacetylase activity may be regulated in mammalian cells, we report here the isolation, the chromosomal localization, and a detailed promoter analysis of the mouse HDAC2 gene.
Phage DNA and bacterial plasmid DNA were purified by standard methods (32), digested with different restriction enzymes, separated on agarose gels, and analyzed by Southern hybridization. Selected genomic DNA fragments were subcloned into pGEM7Zf or pGEM5Zf vectors (Promega), and DNA sequences were determined by the dideoxy chaintermination method (34).
Chromosomal Localization of the Mouse HDAC2 Gene-Phage DNA purified from clone DASHII8 was labeled with digoxigenin dUTP by nick translation. The labeled probe was combined with sheared mouse DNA and hybridized to metaphase chromosomes derived from mouse embryonic fibroblasts in a solution containing 50% formamide, 10% dextran sulfate, and 2ϫ SSC (0.3 M sodium chloride, 0.03 M sodium citrate). Specific hybridization signals were detected by incubating the hybridized slides in fluoresceinated antidigoxigenin antibodies followed by counterstaining with 4Ј,6-diamindino-2-phenyl-indole. A second experiment was conducted in which a probe derived from mouse microsatellites that is specific for the centromeric region of chromosome 10 was cohybridized with the DASHII8 clone. This positive control probe was derived from a set of oligodeoxynucleotide primers from the D10MIT181 map pair. The clone identification number is 5797 from the ES mouse 129/OLA library at Genome Systems, Inc. A total of 80 metaphase cells were analyzed, and only chromosomes with HDAC2 and centromeric probe spots on both chromosome arms were scored as specific.
Plasmids-pGL2-Basic, which contains a luciferase gene in a promoterless background, was obtained from Promega. pGL2-RE, which contains approximately 2.3 kb of mouse HDAC2 DNA sequence upstream from the transcription start site and 128 bp downstream from the transcription start site, was constructed by digesting clone BAC12214 with EcoRI/EagI, filling in the restriction fragment with Klenow polymerase and using blunt-end ligation to clone the filled-in fragment into the SmaI site of pGL2-Basic. pGL2-REr was constructed identically but with the HDAC2 DNA fragment in the opposite orientation. pGL2-BE (Ϫ1100 to ϩ128), pGL2-SacII (Ϫ373 to ϩ162), and pGL2-XE (Ϫ293 to ϩ128) were similarly constructed with BamHI/EagI, SacII, and XbaI/EagI HDAC2 DNA fragments, respectively, and either Klenow or T4 DNA polymerase. p-959Luc was created by digesting pGL2-BE with NsiI/KpnI and using blunt-end ligation to clone the NsiI/KpnI fragment into pGL2-Basic digested with XbaI. To obtain plasmids containing finely incremental 5Ј progressive deletions of the mouse HDAC2 promoter linked to the luciferase reporter gene, p-959Luc was digested with SmaI/PstI and subjected to exonuclease III digestion. All constructions were verified by dideoxy DNA sequencing. pRL-TK, containing the herpes simplex virus thymidine kinase promoter upstream of the cDNA encoding Renilla luciferase, was obtained from Promega.
Primer Extension Assays-Primer extension reactions were carried out essentially as described previously with minor modifications (32). An antisense oligodeoxynucleotide corresponding to the sequence from position ϩ102 to ϩ118 of the mouse HDAC2 promoter was synthesized and end-labeled with [␥-32 P]ATP and T4 polynucleotide kinase. Total RNA from NIH3T3 cells was isolated using the acid phenol-guanidinium thiocyanate method (35). The oligodeoxynucleotide primer (10 6 CPM) was mixed with 14 g of the NIH3T3 total RNA and ethanolprecipitated. The DNA-RNA mixture was then redissolved in 30 l of hybridization buffer (40 mM PIPES (pH 6.4), 1 mM EDTA (pH 8.0), 0.4 M NaCl, and 80% formamide), denatured at 85°C for 10 min, and annealed at 30°C overnight. The annealed hybridization mixture was then ethanol-precipitated, washed, and redissolved in 20 l of reverse transcriptase buffer (50 mM Tris-Cl (pH 7.6); 60 mM KCl; 10 mM MgCl 2 ; 1 mM each of dATP, dCTP, dGTP, and dTTP; 1 mM dithiothreitol; 1 unit/l RNase inhibitor; and 50 g/ml actinomycin D). Subsequently, 50 units of avian myeloblastosis virus reverse transcriptase was added, and the reactions were incubated for 2 h at 37°C. The reactions were then stopped with EDTA, treated with DNase-free RNase, and phenolchloroform-extracted. Single-stranded DNA was then recovered by ethanol precipitation, washed, and dissolved in 4 l of Tris-EDTA (pH 7.4) and 6 l of formamide loading buffer (80% formamide, 10 mM EDTA (pH 8.0), 1 mg/ml xylene cyanol, and 1 mg/ml bromphenol blue). Samples were heated at 95°C for 5 min and resolved on a 6% polyacrylamide/7 M urea gel. The gel was then dried, and images were obtained by autoradiography.
For transfections, 3 ϫ 10 5 cells were seeded in each 60-mm culture dish for 16 -20 h. One-half g of pRL-TK and 10 g of pGL2 or pGL2derived plasmids were introduced into cells with the calcium phosphate coprecipitation method (36). Thirty-six to forty-eight hours after transfections, cells were harvested, and luciferase activity was determined using the dual-luciferase assay system (Promega).
Electrophoretic Mobility Shift Assays (EMSAs)-Nuclear extract containing Myc/Max/Mad was prepared using a modified Dignam-Roeder protocol (37). 3 Purified recombinant Myc, Max, and Mad proteins were prepared as described (38,39). Recombinant AP2 transcription factor was obtained from Promega. Single-stranded oligodeoxynucleotides were labeled individually with [␥-32 P]ATP and T4 polynucleotide kinase, heated together at 65°C, and allowed to anneal by slow cooling to room temperature. Each reaction contained 20 fmol of labeled DNA, 12 mM HEPES (pH 7.9), 10% glycerol, 5 mM MgCl 2 , 60 mM KCl, 1 mM dithiothreitol, 50 g/ml bovine serum albumin, 0.5 mM EDTA, 0.05% Nonidet P-40, 0.1 or 1 g of poly(dI-dC), and approximately 7-10 g of HeLa cell extract or approximately 10 ng of purified protein in a 12-l total volume. Unlabeled specific and nonspecific competitors were included in some reactions. Reactions were incubated for 10 min at room temperature, separated on a 4% nondenaturing polyacrylamide gel (0.0225 M Tris borate, and 0.0005 M EDTA), dried, and subjected to autoradiography.
Accession Number-The nucleotide sequence data reported in this paper will appear in GenBank TM , EMBL, and DDBJ Nucleotide Sequence Data bases under the accession number U93191.

RESULTS
Organization of the Mouse HDAC2 Gene-Using the HDAC2 cDNA as a probe, two positive clones, FIXII11 and DASHII8, were identified out of 2 ϫ 10 6 plaques. In addition, two similar 129/SvJ genomic clones, BAC12213 and BAC12214, were isolated from a bacterial artificial chromosome library with a 277-bp polymerase chain reaction product as the probe. Restriction fragment mapping, Southern blot analysis, and pre-liminary DNA sequencing of the various subclones revealed the organization of the mouse HDAC2 gene (Fig. 1). DASHII8 and FIXII11 contain inserts of about 15.5 kb. Clone FIXII11 contains exons V-XIV, which covers the amino acid residues from position 127 to the 3Ј untranslated region of mouse HDAC2. Clone DASHII8 contains exons II to XII, which represents regions from the last two nucleotides of amino acid residue 18 to the first nucleotide of amino acid residue 460. Restriction enzyme mapping and Southern blot analysis indicated that clones BAC12213 and BAC12214 contain all of the mouse HDAC2 exons (data not shown).
The exon-intron organization of the mouse HDAC2 gene was determined by direct DNA sequencing using oligodeoxynucleotide primers synthesized from both strands at approximately 200-bp intervals in the cDNA sequence. A given exon-intron boundary was indicated when the sequence from genomic clones diverged from that of the cDNA. Subsequent rounds of oligodeoxynucleotide preparation and sequencing completely delineated all exon-introns. The DNA sequences of all splice donor and acceptor sites comply with the invariant GT and AG rule ( Table I) Chromosomal Localization of the Mouse HDAC2 Gene-Purified HDAC2 genomic DNA from clone DASHII8 was used as a probe in fluorescence in situ hybridization. The initial experiment resulted in specific labeling of a medium sized chromosome, which was believed to be chromosome 10 on the basis of 4Ј,6-diamindino-2-phenyl-indole staining ( Fig. 2A). A second experiment was conducted in which a probe specific for the centromeric region of chromosome 10 was cohybridized with a HDAC2-specific probe (Fig. 2B). Measurements of 10 specifically hybridized chromosomes 10 verified that the mouse HDAC2 gene is situated at a position that is 22% of the distance from the heterochromatic-euchromatic boundary to the telomere of chromosome 10, an area that corresponds to band 10B1 (Fig. 2C). Out of 80 metaphase cells analyzed, 54 exhibited specific labeling. No specific hybridization was observed on any other chromosome.
Determination of Mouse HDAC2 Gene Transcriptional Initiation Sites-To determine the transcriptional initiation site of the mouse HDAC2 gene, we used a reverse transcriptase primer extension assay. Primers were designed to span the putative transcriptional start site and then used in extension reactions with total RNA isolated from NIH3T3 cells. Identical reactions were carried out side by side with yeast tRNA as a negative control. The results from one primer revealed two consistently strong signals indicating two major transcriptional start sites, and alignment with a dideoxynucleotide sequence ladder from the same primer revealed that the two strong bands correspond to two Gs within a GC-rich region (Fig. 3). We suspect that transcription is initiated at the preceding A (221 bp 5Ј of the ATG translational start codon) and that the cap structure of the mouse HDAC2 mRNA accounts for the staggered ends.
Localization of the Mouse HDAC2 Promoter-To determine whether DNA sequences upstream of the mouse HDAC2 coding region might functionally direct transcription, various derivatives of the cloned genomic DNA were fused to a luciferase reporter construct and transiently transfected into several different mouse and human cell lines. As shown in Fig. 4, A and  B, in all four cell lines tested, a promoter construct that contains 2.3 kb upstream from the transcription initiation site (pGL2-RE) directed the synthesis of higher activity of luciferase enzyme (as much as 1500-fold in NIH3T3 cells) than of a promoterless reporter plasmid (pGL2-Basic). In contrast, a plasmid containing an identical HDAC2 DNA fragment subcloned in the opposite orientation upstream of the luciferase reporter gene (pGL2-REr) produced only background levels of luciferase enzyme activity similar to pGL2-Basic. These results  indicate the presence of a promoter within Ϫ2300 to ϩ128 of the HDAC2 gene. Deletion of 1.2 kb of 5Ј HDAC2 DNA from pGL2-RE (pGL2-BE) consistently resulted in activation of luciferase activity, suggesting the possible existence of negative cis-acting sequences within Ϫ2.3 and Ϫ1.1 kb. Compared with pGL2-RE, very little change in luciferase activity was observed when the 5Ј HDAC2 sequence was further shortened (pGL2-SacII and pGL2-XE).
To further delineate cis-acting DNA sequences on the HDAC2 promoter that might be important in the transcriptional regulation of the mouse HDAC2 gene, we created 11 additional 5Ј deletion mutants, as well as an internal deletion mutant, transfected them into NIH3T3 cells, and assayed for luciferase activity. As shown in Fig. 4, B and C, there is only slight variation in activity with the different 5Ј deletions from Ϫ1100 to Ϫ293. However, a significant decrease in luciferase activity was seen with an internal deletion from ϩ2 to ϩ114 (compare p-349Luc to p-349[dl2-114]Luc). Moreover, a dramatic decrease in luciferase activity was observed when a segment between Ϫ293 and Ϫ267 was deleted, indicating the presence of a key positive cis-acting regulatory element residing in these 26 bp. A further deletion to Ϫ76 reduced luciferase activity severalfold, but this shortest promoter construct still possesses 27-fold higher luciferase activity compared with pGL2-basic. Thus, our results indicate that the sequence from Ϫ76 to ϩ128 is sufficient to confer promoter activity.
Putative Protein Binding Sites in the Mouse HDAC2 Promoter-The 1.1-kb promoter region of the mouse HDAC2 gene was completely sequenced from both strands (Fig. 5). As is typical of many genes that encode transcription factors, DNA sequences surrounding the HDAC2 gene transcriptional start site are highly GC-rich (79% G or C nucleotide from Ϫ114 to ϩ220) and lack a TATA box. Transcription factor binding site data base searches revealed a number of potential binding sites for ubiquitous and tissue-specific transcription factors. An AP1 binding site is present at Ϫ1039 to Ϫ1029. Three binding sites for the transcription factor upstream stimulatory factor are present between Ϫ771 and Ϫ764, Ϫ465 and Ϫ458, and Ϫ370 and Ϫ361, and overlapping these sites are potential binding sites for Myc/ Max and Sp1. An additional potential Sp1 binding site is located at Ϫ334 to Ϫ322. Overlapping the major transcription initiation site at Ϫ7 to ϩ4 and downstream from the start site at ϩ73 to ϩ84 are two elements that closely resemble binding sites for transcription factor AP2. Potential transcription factor binding sites for C/EBP␤ and AP4 are located at Ϫ721 to Ϫ708 and ϩ193 to ϩ202, respectively.
Identification of Myc, Max, Mad, and AP2 Transcription Factors Interacting with the HDAC2 Promoter-Having identified potential transcription factor binding sites on the HDAC2 promoter, it is important to show that these DNA sequences bind their cognate transcription factors. To this end, we chose to study the three potential Myc binding sites as well as the two potential AP2 binding sites. Previously, we and others have found that Mad/Max and mSIN3 repress transcription by recruitment of HDAC1 and HDAC2 (12,22). Our finding of potential Myc binding sites on the HDAC2 promoter suggests the intriguing possibility that HDAC2 might autoregulate its own expression through Mad/Max. AP2 is a critical transcription factor required for vertebrate development and may be involved in cellular transformation. The AP2 consensus recognition sequence is present in regulatory regions of a variety of cellular and viral genes. Most, if not all, AP2 binding sites identified and reported so far are located upstream of transcription start sites of genes. Interestingly, one of the potential AP2 sites in the HDAC2 promoter is located exactly at the transcription start site, and a second AP2 site is located downstream from the transcriptional start. Very few sequence-specific transcription factors have been found to tightly associate with DNA overlapping transcription start sites.
Using EMSAs and nuclear extracts prepared from HeLa cells, three DNA-protein complexes that migrated identically were found with each Myc binding site (Fig. 6A, lanes 2, 9, and  16). Formation of all three complexes were effectively inhibited by specific competitors (lanes 3, 10, and 17), and neither complex I nor complex II was inhibited by nonspecific competitors (lanes 4 -7, 11-14, and 18 -21). However, formation of complex III with the myc-1 probe (Ϫ372 to Ϫ359) was also inhibited by three different nonspecific competitors (lanes 5-7). Similarly, formation of complex III with the myc-3 probe (Ϫ774 to Ϫ761) was inhibited by nonspecific competitors (lanes 19 -21). Taken together, complexes I and II contain proteins that bind specifically to the three Myc binding sites.
Having obtained evidence that there are specific proteins that interact with the three potential Myc binding sites, we wished to determine whether purified recombinant Myc, Max, or Mad proteins bind specifically to these sites. We found that Myc/Max, Mad/Max, and Myc/Max/Mad all formed specific complexes with the three Myc binding sites (Fig. 6B, complexes I, II, and IV (lanes 2, 5, 8, 12, 15, 18, 22, 25, and 28)). These complexes represent sequence-specific protein-DNA interactions because they can be eliminated by the addition of excess specific competitor DNA but not by nonspecific competitors.
To determine whether AP-2 protein binds to the two potential AP2 binding sites located in the HDAC2 promoter, we synthesized oligodeoxynucleotides corresponding to the two AP2 binding sites and performed EMSAs with a HeLa nuclear extract. As shown in Fig. 7A, both AP2 oligodeoxynucleotides formed a specific complex that can be competed away with excess AP2 binding site DNA either from the HDAC2 or the SV40 viral promoter but not by nonspecific DNA (lanes 2-5, 7-10). Using purified human AP2 transcription factor produced in Escherichia coli from a recombinant clone, we tested the ability of the AP2 site overlapping the transcription initiation site (Ϫ7 to ϩ4) to form a specific complex. Similar to the nuclear extract, a specific complex that could be competed away with excess AP2 binding site DNA but not by nonspecific DNA was formed (Fig. 7B). DISCUSSION Histone acetylation and deacetylation play a key role in the regulation of transcription in eukaryotic cells. Recent studies revealed multiple, distinct genes that encode mammalian histone deacetylases. Here, we describe the first cloning of a genomic histone deacetylase gene. The mouse HDAC2 gene has a complex organization and consists of 14 exons scattered over a region of over 36 kb. Most exons of the HDAC2 gene are quite small, reminiscent of many other genes that encode enzymatic activities (e.g. 40 -42). The 5Ј segment surrounding the transcriptional initiation sites is very GC-rich and lacks a TATA box, a feature that is consistently found in many housekeeping and transcription factor genes.
Using fluorescence in situ hybridization, the mouse HDAC2 gene was localized to chromosomal locus 10B1. A deletion,  1-4). The arrow indicates the most likely start site of transcription. Del(10)69H, and a reciprocal translocation between chromosomes 9 and 10, T(9;10)62H, have previously been mapped to this site (43)(44)(45). In addition, a growth factor-inducible gene encoding a secreted cysteine-rich protein, fisp-12, was mapped to the same general area of the murine genome (46). From our fluorescence in situ hybridization and cloning analyses, we concluded that the mouse HDAC2 activity derives from a single gene.
Deletion analysis of the HDAC2 promoter suggests the existence of positive and negative regulatory elements residing as far upstream as 2.3 kb from the transcriptional start site, but basal transcription is mediated primarily by elements residing between Ϫ76 and ϩ128. It is conceivable that the actual sequence requirement for the promoter activity is much smaller. In this respect, the HDAC2 gene may be similar to many other housekeeping genes, in which as little as 50 bp is sufficient for directing initiation of transcription (e.g. see Refs. 47 and 48). A series of finer deletion mutants of the HDAC2 promoter coupled with transient transfection analysis should clarify this point.
One noteworthy feature of the HDAC2 promoter is the existence of several important potential transcription factor binding sites, of which many appear to be growth-related. An AP1 binding site, which is induced by treatment with phorbol esters and binds several different proteins, including Jun, Jun B, Fos, and Fos-related antigens (49), is present far upstream from the initiation of transcription. The existence of an AP1 binding site in the mouse HDAC2 gene promoter is consistent with the idea that mouse HDAC2 mRNA may be stimulated by growth factors.
Two AP2 binding sites, one overlapping the transcriptional start site and one located downstream of the transcriptional start, were found in the mouse HDAC2 promoter. Because AP2 mediates gene activation in response to retinoic acid, cyclic AMP, and phorbol esters (50 -52), AP2 may also contribute to growth-induced HDAC2 expression. More interestingly, although cellular factors that bind to the immediate vicinity of a transcription initiation region have previously been identified in several promoters (53)(54)(55)(56)(57)(58)(59)(60)(61)(62)(63)(64)(65), none of them bear similarity to AP2. Our EMSAs and transient transfection assays suggest a novel mechanism of transcriptional regulation by AP2.
Three Myc/Max/Mad binding sites are present in the mouse HDAC2 gene promoter. One of the Myc/Max/Mad binding sites overlaps with the upstream stimulatory factor. Another one extends over the binding site for upstream stimulatory factor. A third site, most proximal to the transcription start, overlaps with an Sp1 site and embraces an upstream stimulatory factor site. Myc protein possesses helix-loop-helix and leucine zipper motifs and can heterodimerize with Max protein via the helixloop-helix motif that is present in both proteins (66). Max may also heterodimerize with another member of the same family, Mad (38). Whereas Myc-Max heterodimers can activate transcription, Mad-Max heterodimers can bind to the same site and repress transcription (67)(68)(69)(70). Interestingly, the Myc protein is expressed at very low levels in resting cells, and its expression is induced with cell growth, whereas Max is expressed at high levels in resting and proliferating cells. Therefore, it was speculated that overexpression of the myc gene, as observed in many cancer cells, could favor production of Myc-Max heterodimers, directing activation of growth-related genes that are repressed in normal cells. The presence of multiple Myc/Max/ Mad binding sites in the promoter of the HDAC2 gene implies that the expression of HDAC2 may be up-regulated by Myc/ Max. Recent studies indicate that HDAC2 associates with the mSIN3 corepressor and mediates Mad-Max transcriptional repression (12,22,24). This suggests the intriguing possibility that HDAC2 expression may be autoregulated through Mad-Max binding to the HDAC2 promoter. Work is now underway to address this issue.
(Ϫ293 to Ϫ267) deletion of the mouse HDAC2 promoter resulted in a dramatic decrease in expression. An inspection of sequences in this area did not reveal a resemblance to any known transcription factor binding sites. Taken together, our results suggest that a key transcription factor that is important for the regulation of HDAC2 is yet to be discovered.
In summary, identification of a genomic deacetylase gene, mouse HDAC2, and knowledge of the organization of this gene and of its promoter sequence should now provide new information that will facilitate a comprehensive study of HDAC2 gene regulation. In addition, with a fully characterized mouse HDAC2 gene in hand, we should be able to begin to address the functional role of HDAC2 by gene replacement experiments.