J Biol Chem, Vol. 275, Issue 20, 15254-15264, May 19, 2000

Cloning and Characterization of a Novel Human Class I Histone Deacetylase That Functions as a Transcription Repressor^*

Erding Hu^§, Zunxuan Chen, Todd Fredrickson, Yuan Zhu^¶, Robert Kirkpatrick, Gui-Feng Zhang, Kyung Johanson, Chiu-Mei Sung^**, Ronggang Liu, and James Winkler^**

From the Departments of  Renal Pharmacology, ^** Oncology, ^¶ Molecular Biology,  Medicinal Chemistry, and  Gene Expression Sciences, SmithKline Beecham Pharmaceuticals, King of Prussia, Pennsylvania 19403

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Histone acetylation alters chromatin state by modifying lysines on histone and plays an important role in modulating gene transcription. A dynamic balance of histone acetylation/deacetylation is maintained by histone acetyltransferases and histone deacetylases. Emerging evidence suggests that a family of histone deacetylases may exist to regulate diverse cellular functions, including chromatin structure, gene expression, cell cycle progression, and oncogenesis. We describe here a novel human histone deacetylase, named HDAC8, cloned from human kidney. HDAC8 encodes 377 amino acid residues and shares extensive homology to several known HDACs, in particular a histone deacetylase from Arabidopsis thaliana. Northern blot analyses revealed that HDAC8 expression pattern for HDAC8 is distinct from that for HDAC1 and HDAC3, and expression of HDAC8 mRNA occurs in multiple organs including heart, lung, kidney, and pancreas. HDAC8 mRNA was also observed in several cell lines derived from cancerous tissues. When expressed in HEK293 cells, HDAC8 exhibited deacetylase activity toward acetylated histone, indicating that this protein is a bona fide histone deacetylase. Its histone deacetylase activity was inhibited by trichostatin and other known histone deacetylase inhibitors. Furthermore, active recombinant HDAC8 was expressed and purified from Escherichia coli. When ectopically expressed in cells, HDAC8 was found to be localized to the nucleus. Co-transfection experiments demonstrated that expression of HDAC8 repressed a viral SV40 early promoter activity. These results indicate that HDAC8 is a novel member of the histone deacetylase family, which may play a role in the development of a broad range of tissues and potentially in the etiology of cancer.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Histone acetylation involves the modification of lysines on histones at specific residues (1, 2). The addition of an acetyl group on lysines is catalyzed by histone acetyltransferase (HAT),¹ while histone deacetylases (HDAC) act to remove the acetyl group. Steady state histone acetylation is controlled by the balance of enzymatic activity of both HAT and HDACs (3). Hypo-acetylated histones increase the positive charge of histone and condense the chromatin. Conversely, hyperacetylated histones neutralize the electrostatic charge and de-condense the chromatin (3, 4). Transcription activators (SRC-1 (Ref. 5), p300 (Ref. 6), ACTR (Ref. 7), and PCAF (Ref. 8)) and repressors (Sin3 (Ref. 9), pRB (Refs. 10 and 11), YY1 (Ref. 12), and NcoR (Ref. 13)) have been shown to be associated with HAT and HDACs, respectively. Several transcriptional activators even contain intrinsic HAT activities (SRC-1, p300, and PCAF) (14-16). Furthermore, the enzymatic activity of HAT and HDAC are important for transcriptional activation or repression in vivo (17-19). Several oncogenes (pRB (Ref. 10), BRCA-1 and -2 (Refs. 20 and 21), PML-RAR (Ref. 22), and a zinc finger protein mutated in leukemia (Ref. 23)) have been shown to be associated with HATs or HDACs. In particular, the viral oncogene E1A has been shown to modulate histone acetyltransferase activity of p300 and PCAF, thus changing the dynamics of chromatin structure (24, 25) and triggering gene transcription associated with oncogenic transformation.

Recent studies have revealed multiple forms of HDAC in mammalian cells (26-28) with at least six human HDAC genes identified: HDAC1 (29, 30), HDAC2 (also named RPD3-like protein) (12), HDAC3 (also named HD3-c, RPD3-2A, or 2B) (31-34), h-HDAC4 (also named HDAC-A) (26, 27), h-HDAC5 (26), h-HDAC6 (26), and h-HDAC7 (35, 36). Some HDAC (e.g. HDAC3) mRNAs have differential splice forms that have distinct N termini. HDACs are now classified into two general classes: 1) class I (HDAC1, HDAC2, and HDAC3) enzymes have approximately 400-500 amino acids and are primarily located in the nucleus; 2) class II (HDAC4, HDAC5, HDAC6, and HDAC7) enzymes are much larger proteins with around 1000 amino acids, and HDAC6 has duplications of its catalytic domain (26-28). Although the precise function of these two classes of HDACs is unknown, evidence suggests that they are involved in diverse biological activities, ranging from hormone-induced gene regulation (37-40) to apoptosis (41, 42). Importantly, different HDAC form distinct complexes with different sets of transcription regulators and presumably control gene expression from different promoters (26, 27). Several HDAC inhibitors, including sodium butyrate and trichostatin (TSA), have been shown to have a wide range of effects in cell culture and in animals. These effects include specific gene activation (38, 40, 43), cell proliferation and cell cycle arrest (44), as well as induction of differentiation (22, 41, 45). In addition, two HDAC inhibitors (MS-27-275) (46) and FR901228 (47) have demonstrated tumor growth suppression in vivo. These results suggest complex regulatory roles for HDACs in many vital cellular processes and potential therapeutic utility of HDAC inhibitors.

We have identified a novel class I human histone deacetylase, HDAC8. Our data suggest HDAC8 is a unique HDAC expressed in a wide range of human tissues as well as cancer cells. Our results also indicate that HDAC8 is a functional enzyme localized to the nucleus and can act as a transcriptional repressor.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials-- Nylon membrane (Biotran) was purchased from ICN Biotechnologies (Costa Mesa, CA). [-³²P]dATP (3000 Ci/mmol) was purchased from NEN Life Science Products. Oligo(dT)-agarose was obtained from Amersham Pharmacia Biotech. Routine molecular cloning and sequence analyses was performed as described previously (48). Reagents for subcloning and sequencing were purchased from Promega Inc. (Madison, WI). PCR reagents were obtained from Perkin-Elmer Inc. (Norwalk, CT). Random priming labeling kits were obtained from Promega Inc. Human poly(A)⁺ mRNA blot was purchased from CLONTECH (Palo Alto, CA). TSA and butyrate were purchased from Sigma. [³H]Histone was generated by labeling HEK293 cells with [³H]acetate in the presence of 10 mM butyrate as described (49). Computer programs used to analyze the protein sequences includes blastn, blastp, and lasergene-DNA star programs.

Isolation of HDAC8 cDNA and Plasmid Construction-- The GenBank^TM EST data base was searched with human HDAC1 and HDAC3 protein sequences, and several ESTs encoding a putative novel HDAC were identified. One EST sequence (accession no. T99283) was used as a template to isolate a full-length cDNA by reverse transcription-PCR-based 5' and 3' extension using the Marathon cDNA kit from CLONTECH. Full-length cDNA obtained was subcloned into PCR2.1 vector (Invitrogen, Carlsbad, CA) and sequenced. Subsequent subcloning was performed using PCR, and all constructs were sequenced to verify the authenticity. The myc-tagged HDAC8 expression vector was constructed by PCR of the coding sequence of full-length HDAC8 (PCR primers: 5' primer, 5'-GAATTCTTGAGGAGCC-GGAGGAACCG-3' and 5'-GCGGCCGCCAA-CTAGACCACATGCTTCAG-3') and subcloned into pCMV-myc vector (EcoRI and NotI sites) from CLONTECH to make a in-frame fusion protein with myc tag at the N terminus. All expression vectors were sequenced to ensure the constructs were free of PCR errors.

RNA Isolation and Northern Analysis-- RNA isolation and Northern blot analysis were performed as described previously (48).

Immunostaining of HDAC8-- 48 h after transient transfection of myc-HDAC8 construct into either Rat-2 or NIH-3T3 cells, cells were trypsinized and re-seeded onto chamber slides (Lab-Tek II, Nunc, Inc., Naperville, IL). Immunostaining was performed essentially as described (50). Briefly, approximately 10³ cells were seeded into each well of two-chamber slides and incubated overnight. All the staining procedures were performed at room temperature: cells were fixed for 2 min with 20 °C methanol, washed with TBS (0.02 M Tris-HCl, pH 7.4, 0.15 M NaCl), and then permeablized in TBS, 0.5% Triton X-100 for 10 min. After a blocking step (incubation with TBS, 0.1% Triton X-100 containing 2% of BSA for 10 min), cells were washed in TBS, 0.1% Triton X-100 and incubated with c-Myc monoclonal antibody (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) in 1:100 dilution for 2 h. Cells were washed in TBS, 0.1% Triton X-100 thoroughly and followed by incubation of Cy3-conjugated goat anti-mouse IgG (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA) in 1:100 dilution for 45 min. Cells were washed in TBS, 0.1% Triton X-100, and then the chamber walls were removed. The slides were drained, mounted with a mounting medium containing DAPI (Vector Laboratories, Burlingame, CA), and sealed. Fluorescent images were visualized under a fluorescent microscope (Olympus Inc., Tokyo, Japan) using a digital camera and processed using Photoshop (Adobe Inc., San Jose, CA).

Tissue Culture and Transfections-- 293, NIH-3T3, or Rat-2 cells were used for HDAC8 enzymatic assays and immunofluorescent studies. All cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum along with penicillin-streptomycin antibiotics. Transfections were performed when cells reached 80% confluence, and a LipofectAMINE Plus transfection kit was used (Life Technologies, Inc.). Transfections were performed according to manufacturer's protocol.

Immunoprecipitation (IP) and HDAC Assays-- Cells were collected after transfection and lysed with single detergent lysis buffer (50 mM Tris-HCl, pH 8, 150 mM NaCl, 1% Nonidet P-40) supplemented with protease inhibitor complete (Roche Molecular Biochemicals). After spinning at 12,000 × g for 30 min to remove insoluble material, soluble supernatants were collected and used for immunoprecipitation as described (48). Anti-myc monoclonal antibody (1:100) (Calbiochem, Inc., San Diego, CA; Santa Cruz Biotechnology, Inc.) and anti-HDAC1 antibody (Upstate Biotechnology, Inc. Lake Placid, NY) were used for IP. For HDAC enzyme assay, ³H-labeled histone substrate was made using HEK293 cells and labeled with [³H]acetate in the presence of 10 mM butyrate and labeled histone was isolated as described (49). HDAC assays were performed essentially as described (49). Briefly, in a 200-µl reaction containing 10 mM Tris, pH 8, 150 mM NaCl, 1 mM MgCl₂, 5000-10,000 cpm of [³H]histone was added along with protein A/G beads. After 1 h of incubation, 50 µl of stop mixture (1 M HCl and 0.16 M acetic acid) was added to the sample and 1 ml of ethyl acetate was added to extract released tritiated acetate. 800 µl of the ethyl acetate was counted in a liquid scintillation counter (Beckman).

Expression and Purification of HDAC8 in E. coli-- Recombinant HDAC8 was expressed as a C-terminal hexahistadine-tagged fusion protein in E. coli using a T7 Lac promoter-driven vector, pET21b (Novagen), derived from vectors developed by Studier et al. (51, 52). Two rounds of site-directed mutagenesis were performed using the Quickchange mutagenesis kit (Stratagene), introducing silent changes to eliminate two internal NdeI restriction sites. The following primer pairs were used (Set 1, Forward (5'-GCA TTC TTT GAT TGA AGC GTA TGC ACT GCA TAA GCA AAT GAG G-3') and Reverse (5'-CCT CAT TTG CTT ATG CAG TGC ATA CGC TTC AAT CAA AGA ATG C-3'); Set 2, Forward (5'-CCA GAT CAT GAG TTT TTC ACA GCG TAT GGT CCT GAT TAT GTG CTG G-3') and Reverse (5'-CCA GCA CAT AAT CAG GAC CAT ACG CTG TGA AAA ACT CAT GAT CTG G-3')). The HDAC8 coding sequence was PCR-amplified from the resultant mutant clone using the following primer sequences (forward, 5'-TTA TAA ATT CAT ATG GAG GAG CCG GAG GAA CCG GCG-3'; reverse, 5'-TTA TAA ATT AAG CTT ATC AAT GGT GAT GGT GAT GGT GGC TGC CGC GGC CTT CAA TGA CCA CAT GCT TCA GAT TCC CTT TG-3') tailed as indicated on the 5' and 3' ends with NdeI and HindIII restriction endonuclease sites, respectively. The PCR product was digested with NdeI and HindIII and subcloned into the multiple cloning region of pET21b, placing the C-terminal factor Xa, histidine sequence in frame with the HDAC8 coding sequence. The inserted gene sequence was confirmed on both strands by automated dideoxynucleotide sequencing (Applied Biosystems) using T7 promoter and T7 terminator primers. The resultant HDAC8 expression plasmid was then transferred to the lysogenic BL21 DE3 E. coli strain containing a chromosomal copy of T7 RNA polymerase under lacUV5 promoter control (51, 52). Overexpression was performed by induction of mid-log phase cells with 1 mM isopropyl-1-thio--D-galactopyranoside for 3 h at 37 °C. Cell pellets were harvested and frozen at 70 °C prior to cells lysis and purification. For purification of HDAC8, 5 liters of cell paste of HDAC8 with His tag at C terminus were lysed in 100 ml of lysis buffer (25 mM Tris, pH 7.5, 3 mM MgCl₂, 10 mM NaCl, 0.25% Nonidet P-40, and protease inhibitor mixture) and centrifuged at 30,000 × g for 1 h. The cell supernatant was loaded onto a 20-ml nickel-nitrilotriacetic acid column. Column was washed with buffer (25 mM Tris, pH 7.5, 3 mM MgCl₂, 100 mM NaCl) containing 30 mM imidazole and 100 mM imidazole. The His-tagged HDAC8 was eluted with buffer containing 300 mM imidazole. The HDAC8 was dialyzed into buffer (10 mM Tris, pH 7.5, 3 mM MgCl₂, 100 mM NaCl, 10% glycerol) and kept at 70 °C before assaying for HD activity.

HDAC8 Antibody and Western Blot-- Purified HDAC8 as used to generate rabbit polyclonal antiserum against human HDAC8. Western blot as performed as a routine procedure (48) using 1:1000 dilution of unpurified rabbit antiserum.

Nuclear and Cytoplasmic Fractionation-- Cell fractionation was performed as described as Dignam et al. (53) and Osborn et al. (54). Briefly, 10⁷ SW620 cells were collected and washed in phosphate-buffered saline twice, and then cell pellet was resuspended in 100 µl of buffer A (10 mM Hepes, pH 7.9, 10 mM KCl, 1.5 mM MgCl₂, 0.5 mM dithiothreitol, 0.1% Nonidet P-40, 1 mM phenylmethylsulfonyl fluoride). After incubation on ice for 30 min, nuclei was pelleted by microcentrifugation at 3500 × g for 10 min and supernatant collected as cytoplasm. Nuclei pellet was resuspended in 100 µl of buffer C (20 mM Hepes, pH 7.9, 0.8 mM NaCl, 1.5 mM MgCl₂, 25% glycerol, 0.2 mM EDTA, 0.5 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride). Suspension was mixed gently by rocking for 1 h at 4 °C and then centrifuged at 14,000 × g for 30 min. Supernatant was collected as nuclear extract. Protein concentration was determined using a Bio-Rad DC kit, and 50 µl of protein was separated on SDS-PAGE followed by Western blot analysis.

SV40 Promoter Reporter Assays-- 2 µg of SV40 promoter construct (pGL3-promoter luciferase reporter vector, Promega) along with different amount of HDAC8 expression constructs were transiently transfected into Rat-2 cells. The SV40 promoter region in pGL3 (nucleotides 44-244) contains transcription initiation sites as well as multiple Sp1 binding sites (55). This promoter region corresponds to bases 5180-5240 of the SV40 viral genome (GenBank^TM accession no. J02400). Empty vector DNA were used to normalized the amount of transfected DNA to 20 µg/100-mm cell culture dish. Transfections were performed as described. 48 h after transfection, cells were trypsinized and reseeded into a 96-well dish at approximately 10,000 cells/well. After overnight incubation, the luciferase reporter activity were measured on a Top Counter (TopCounter NXT; Packard, Meriden, CT) using a Luclite luciferase assay kit (Packard). When cells were treated with TSA (300 nM), another 24 h were allowed before luciferase activities were measured.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Molecular Cloning of Human HDAC8-- A human EST clone (GenBank^TM accession no. T99283) encoding a novel HDAC was found by searching the human EST data base with conserved domains of human HDAC1 and HDAC3. Subsequently, a full-length cDNA clone for this HDAC was obtained by 5' and 3' rapid amplification of cDNA ends of mRNA from human kidney. Sequence analysis revealed that the cDNA encoded a novel gene that is highly homologous to members of the histone deacetylase family, and we named this gene histone deacetylase 8 (HDAC8).

HDAC8 mRNA encodes 377 amino acids with a predicted molecular mass of 45,240 Da (Fig. 1). There is no apparent hydrophobic leader sequence, and several potential post-translational modification sites are identified (Fig. 1A). They include a potential N-glycosylation site (NWS) at Asn¹³⁶, a cAMP-dependent kinase phosphorylation site (KRAS) at Ser³⁹, and two potential casein kinase II phosphorylation sites (Fig. 1A). A stretch of basic region was also identified (Arg¹⁶⁴ to Lys¹⁶⁸, underlined, Fig. 1A) that could serve as a nuclear localization signal (Fig. 1A).

View larger version (6510274K): [in this window] [in a new window]   Fig. 1.   A, nucleotide and amino acid sequence for human HDAC8. Total amino acids for HDAC8 is 377. The GenBankTM accession number for HDAC8 is AF230097. Several potential functional regions or post translational modification sites are highlighted as follows. A basic region (RLRRK) is underlined as potential nuclear localization signal, a potential N-glycosylation site is underlined (Asn136), a potential site for phosphorylation by cAMP-dependent kinase is underlined (Ser39), and two potential casein kinase II sites are also underlined (Ser63 and Ser83). B, alignment of known human HDACs. The following GenBankTM entries were used for comparisons: U50079 (29), AF005482 (12), U75696 and U75697 (32), AF059650 (33), AF039703 (34), D50405 (30), U31814 (12), and U66914 (31). C, phylogenetic tree analysis of HDAC8 and other human HDACs. Additional GenBankTM entries include: AF132607, AF132608, AF132609 (26), and AB006626 (27). D, HDAC8 contains all nine blocks conserved in other HDACs and bacterial acuC family of hydrolases (56). A histidine (His143) important for catalytic activity of HDAC was labeled with an asterisk (*).

Sequence comparisons with other known human HDAC indicated a significant degree of conservation (Fig. 1B), with a 37% similarity to HDAC1, 37% to HDAC3, and around 16% similarity to HDAC4, HDAC5, and HDAC6 over the full-length protein sequences (Fig. 1D). The sequence conservation is, however, much more striking at the domain level. Most HDACs have nine conserved blocks that are presumably important for its catalytic function (26, 56). Indeed, these nine blocks are also found in HDAC8 (Fig. 1D), suggesting that HDAC8 is a functional enzyme. A histidine residue (His¹⁴³) central for the catalytic activity (26) was also found in HDAC8 (Fig. 1D). Phylogenetic tree analysis suggests that evolutionarily HDAC8 is mostly closely related to human h-RPD3-like (GenBank^TM accession no. U31814) and it lies between the evolutionary boundary between class I HDACs and class II HDACs (Fig. 1C). When compared against all known HDACs from different species, HDAC8 is most closely related to a HDAC identified from Arabidopsis thaliana (GenBank^TM accession no. AF014824),² with 38.4% similarity over the full-length protein sequence (data not shown).

Expression of HDAC8 in Normal and Cancerous Human Tissues and Cells-- To examine the expression pattern of HDAC8 in human tissues and cells, we performed Northern blot analyses using HDAC8 cDNA as a probe and compared the expression of HDAC8 with HDAC1 and HDAC3. As shown in Fig. 2, HDAC8 mRNA is expressed at higher levels in normal human brain and pancreas, while lower levels were also detected in wide array of tissues including heart, placenta, liver, and kidney. Compared with HDAC3, the level of HDAC8 mRNA was much lower and the Northern blots required considerably longer exposure time (Fig. 2). Further analysis using quantitative PCR (Taqman) on mRNA isolated from approximately 20 different normal human tissues indicated that the distribution of HDAC8 was indeed widespread and the copy number for HDAC8 mRNA was approximately 1/10 that for HDAC3 mRNA (mRNA copies/ng of genomic DNA) (data not shown).

View larger version (67K): [in this window] [in a new window]   Fig. 2.   Expression of HDAC1, HDAC3, HDAC8, and actin in human tissues and cancer cell lines. Blots containing mRNA from indicated tissues or cells were probed with full-length cDNA HDAC8 probes, stripped, and re-probed with other HDACs. Exposure time is 2 days for HDAC1, 18 h for HDAC3, 3 days for HDAC8, and 8 h for actin. Full coding sequences for HDAC1, HDAC3, and HDAC4 were used as probes for Northern analysis.

Interestingly, two mRNA species were detected for HDAC8 mRNA. The size of the HDAC8 mRNA is approximately 2.0 and 2.4 kb. The same HDAC8 mRNA doublet was also observed in various cancerous cell lines (Fig. 2). One intriguing feature for HDAC8 mRNA was the relative abundance of the longer versus shorter messenger RNA. Apparently, in Burkitt's lymphoma Raji cells and melanoma G361 cells, larger 2.4-kb HDAC8 mRNA species were more abundant than the 2.0-kb species while in other cancer cells (HeLa, lymphoblastic leukemia Molt-4, colon adenocarcinoma SW480, and lung carcinoma A549), the 2.0-kb HDAC8 mRNA dominates.

Enzymatic Activity of HDAC8-- To assess the enzymatic activity, a myc-tagged HDAC8 was transiently transfected into 293 cells and IP experiments were performed to examine histone deacetylase activity. As shown in Fig. 3A, both myc-tagged HDAC3 (a control) and HDAC8 were expressed in the soluble nuclear extract of transiently transfected cells. HDAC8 migrated as a single protein band in a Western blot using anti-myc antibody (Fig. 3A). The observed molecular mass was around 49 kDa. This agrees very well with the predicted molecular mass (44 kDa) from the protein sequence. Nuclear extracts of transfected cells were used for IP analysis and proteins retained on protein A beads were assayed for HDAC activity. As shown in Fig. 3B, myc-HDAC8-transfected cellular extracts clearly exhibited HDAC activity and the activity was inhibited by 300 nM TSA as well as other HDAC inhibitors including butyrate and MS-25-275 (46) (data not shown). As a positive control, a commercial antibody against HDAC1 immunoprecipitated HDAC activity in parent as well as transfected 293 cells (Fig. 3B). These results indicated that HDAC8 encoded an active enzyme when expressed ectopically in 293 cells.

View larger version (23K): [in this window] [in a new window]   Fig. 3.   A, expression of myc-tagged HDAC8 in HEK293 cells. 10 µg of expression plasmid for myc-HDAC8 was transiently transfected into HEK293 cells and soluble fraction was collected. 100 µg of protein was separated on a 4-12% gradient SDS-PAGE and transferred onto a polyvinylidene difluoride membrane for Western blot. Lanes 1 and 2 are cell extracts transfected with two separate clones for myc-HDAC8 expression vector, and lane 3 is extract from cells expressing myc-HDAC3. Lane 4 is untransfected extract. Molecular weight markers are labeled at the right side of the blot. B, enzymatic activity of HDAC8. Soluble extract from transfected 293 cells were prepared as described under "Experimental Procedures," and HDAC activity was measured for extracts from cells transfected with vector or myc-HDAC8. HDAC1 antibody was used as a positive control. HDAC assays were performed in the absence or presence of 300 nM TSA.

Expression and Purification of Recombinant HDAC8 in E. coli-- In order to assess the enzymatic activity of HDAC8 further, we sought to express and purify recombinant HDAC8 in E. coli. A T7 RNA polymerase-based E. coli expression system (PET) was used to express HDAC8 (51). A histidine tag was engineered into the C terminus of HDAC8 to assist purification (Fig. 4A). As shown in Fig. 4B, tagged HDAC8 was expressed in the soluble fraction (Fig. 4B, Western panel, lane 3) and HDAC8-His was further purified to near homogeneity (>95%) (Fig. 4B, Western and SDS-PAGE panels, lane 5) using a nickel-nitrilotriacetic acid column. The authenticity of the expressed HDAC8 protein was verified by N-terminal peptide sequencing (data not shown). When assayed for HDAC activity, purified HDAC8 was found to be active toward tritiated histone in a dose-dependent manner (Fig. 4C). To assess if metals such as zinc can modulate the activity of HDAC8, we tested the effects of several mono- and divalent cations on recombinant HDAC8 activity. Unlike bacterial HDAC (58), addition of Zn²⁺ completely inhibited human HDAC8 enzymatic activity (Fig. 4D). Similarly, Cu⁺ and Fe²⁺ also inhibited the HDAC activity. In contrast, K⁺ and Mn²⁺ moderately enhanced the activity while Ba²⁺, Li⁺, and Ca²⁺ had no effects (Fig. 4D). These results clearly indicate that human HDAC8 is an active enzyme and that it may have a different metal requirement that is distinct from bacterial HDAC (58).

View larger version (31K): [in this window] [in a new window]   Fig. 4.   Recombinant expression of HDAC8 in E. coli. A, PET-HDAC8 constructs. A His tag (IEGRGSHHHHHH) was added to the C terminus of HDAC8. B, Coomassie staining of a SDS-PAGE gel and Western blot of HDAC8-His in soluble fraction of E. coli and purified HDAC8-His. Western panel, lane 1, vector-transformed bacterial extract; lanes 2 and 3, HDAC8-transformed extract with (lane 3) and without (lane 2) isopropyl-1-thio--D-galactopyranoside induction; lane 4, blank; lane 5, purified HDAC8 (1 µg). Coomassie stain panel, lane 1, protein molecular weight marker; lane 2, vector-transformed bacterial extract; lanes 3 and 4, HDAC8-transformed extract with (lane 4) and without (lane 3) ITPG induction; lane 5, purified HDAC8 (1 µg). C, activity of recombinant HDAC8. Purified HDAC8-His was used in a routine HDAC assay (see "Experimental Procedures"). Increasing amount of HDAC8-His was used in a 100-µl assay (2.5-20 µg) in the presence or absence of TSA (300 nM). 50,000 cpm of 3H-labeled histone was used as substrate. D, influence of different mono- and divalent cations on recombinant HDAC8 enzymatic activity. 2.5 µg of purified HDAC8-His was incubated with increasing amounts of indicated cations in the presence of 10,000 cpm of 3H-labeled histone for 30 min, and released H3 was counted in a liquid scintillation counter. Inhibition or activation was plotted as percentage of HDAC activity in assaying buffers (see "Experimental Procedures") without any added cations. E, 20,000 cpm of metabolically labeled histone as incubated with the indicated amount of HDAC8 for 1 h at 37 °C. After incubation, the entire reaction mixture as denatured by adding SDS loading buffer and protein was separated on a 4-12% gradient gel (Nu-PAGE) and gel was stained with Coomassie Blue, dried, and exposed to x-ray film.

We also investigated if HDAC8 preferentially deacetylates any given histone in vitro. As shown in Fig. 4E, H3 and H4 were clearly substrates, as the radioactivity corresponding to these histones was diminished with increasing amounts of HDAC8.

Nuclear Localization of HDAC8-- A myc-tagged HDAC8 construct was transiently transfected into either Rat-2 or NIH-3T3 cells, and a Cy3-labeled anti-myc antibody was used to visualized the cellular localization of transfected HDAC8. As shown in Fig. 5, HDAC8 was clearly localized to nucleus. The distribution was throughout the nucleus and coincided with the DAPI (DNA) staining (Fig. 5A). At a higher magnification, certain sporadic regions of the nucleus were devoid of HDAC8 staining, suggesting the possibility of specific exclusion of this HDAC from transcriptionally active areas such as the nucleolus. A similar nuclear localization pattern was also observed for murine class I HDAC1 protein (59) and class II HDAC4 (27).

View larger version (48K): [in this window] [in a new window]   Fig. 5.   A, immunostaining for transfected recombinant HDAC8 in NIH-3T3 cells. Panels A, B, and C are lower magnifications of anti-myc-Cy3, DAPI, and phase contrast images. Panels E and E are higher magnification images of anti-myc-Cy3 and DAPI images. Panels F, G, and H are images for negative control where cells were transfected with empty vector. B, Western blot analysis of total cell protein, the fractionated cytoplasm, and nuclear extract of SW620 cells. 50 µg of protein was separated in SDS-PAGE and analyzed. Lane designations were as follows: lane 1, total cell protein from SW620 cells; lane 2, total cell protein from NIH-3T3 cells; lane 3, total cell protein transfected with myc-HDAC8 expression construct; lanes 4 and 5, cytoplasm and nuclear extract from SW620 cells.

Nuclear localization of HDAC8 was also confirmed in a fractionation experiment (Fig. 5B). The human colon cancer cell line, SW620, was found to express endogenous HDAC8 (Fig. 5B). The endogenous HDAC8 migrated slightly faster than transfected myc-tagged HDAC8 (Fig. 5B, lanes 1-3) since the myc-tagged HDAC8 is 22 amino acids longer. After separation of the cytoplasm and nuclear fractions, HDAC8 was found to be highly enriched in the nuclear fraction in a fashion similar to that for another transcription factor, C/EBP-, while a cytosolic enzyme, glyceraldehyde-3-phosphate dehydrogenase, was localized exclusively to the cytoplasm (Fig. 5B, lanes 4 and 5). These data, together with the immunofluorescence results, strongly suggests that HDAC8 is a nuclear protein.

Transcriptional Repression Mediated by HDAC8-- The effect of HDAC8 expression on viral SV40 promoter activity was examined by co-transfection of a SV40 promoter-driven luciferase and HDAC8 expression constructs. A significant repression of luciferase activity was observed when increasing amounts of HDAC8 expression vector was transfected along with the SV40 promoter-driven luciferase reporter construct (pGL3 promoter) (Fig. 6A). This repression reached a maximal of approximately 70-80% at 10 µg of transfected HDAC8 DNA. In addition, TSA induced a dramatic increase of SV40 promoter-driven luciferase activity in a dose-dependent fashion (Fig. 6B). This increase was repressed up to 50% by the expression of HDAC8 at lower doses of TSA (19-37 nM). However, at higher doses of TSA (300 nM), the induction of luciferase activity was comparable in control and HDAC8-transfected cells (Fig. 6B). Taken together, these results suggest that ectopically expressed HDAC8 behaved as a functional transcriptional repressor toward the viral SV40 early promoter and the repressor effect of HDAC8 can be antagonized by a high dose of TSA. This suggests that the repressor activity of HDAC8 is mediated by HDAC activity.

View larger version (35K): [in this window] [in a new window]   Fig. 6.   A, transcriptional suppression by HDAC8. 2 µg of SV40 early promoter-driven luciferase vector was transfected into Rat-2 cells as described under "Experimental Procedures" An increasing amount of myc-HDAC8 expression vector was transfected along with the reporter, and an empty pCMV vector (Promega, Madison, MI) was used to normalize the total amount of DNA to 20 µg. A expression plasmid (GeneStorm, Invitrogen, Carlsbad, CA) that encodes glucose-6-phosphate dehydrogenase (G6PD) was used as negative control. The reporter activities for each transfection were measured using Luclit kit from Packard Inc. All points were done in triplicate, and transfections were repeated four times. The increasing amount of HDAC8 was visualized by Western blot (lower panel). B, 2 µg of SV40 promoter-driven luciferase and 10 µg of myc-HDAC8 expression vector were co-transfected into Rat-2 cells, and cells were then treated with various indicated concentration of TSA. Luciferase activities were measured 24 h after treatments.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

HDACs belong to a large protein superfamily that encompasses at least two general classes of members. The classification is based on 1) size and sequence characteristics and 2) distinct association with different transfection co-factors. Class I HDACs are smaller (49-60 kDa), whereas class II HDACs are larger than 100 kDa. Both classes of HDAC were initially identified in yeast (60), and plants (61) and more recently in mammalian cells (26, 27, 29). To date, three human class I HDAC have been characterized. In this paper, we describe an additional member of class I HDAC, HDAC8.

HDAC8 is highly homologous to other HDACs found in mammalian cells and in other organisms. In particular, HDAC8 retains all nine conserved domains found in other HDACs and their presumed ancestral gene, bacterial enzyme acuC (56). Phylogenetic tree analysis suggests that HDAC8 diverges from other class I human HDACs early in evolution, and it may, therefore, represent a key point that distinguishes class I and class II HDACs in human. With the identification of HDAC8, there are now four human class I HDACs and three class II HDACs.

HDAC8 is an active enzyme both in eukaryotic cells or expressed as a recombinant enzyme in E. coli. The expression and purification of HDAC8 in E. coli suggests that the enzyme is active in the absence of cofactors and without further post-translational modifications that occur in eukaryotic cells. To our knowledge, this is the first report of a mammalian HDAC that can be expressed and purified from E. coli while retaining its enzymatic activity. Our initial biochemical characterization indicates that HDAC8 enzymatic activity can be modulated by mono- and divalent cations. The availability of this expression system should enable us to explore further the structure-functional relationship as well as potential regulatory role of HDAC-associated proteins such as oncoproteins E1A and p53.

Similar to other HDAC, HDAC8 has been localized to the nucleus, suggesting that HDAC8 could play a role in transcriptional regulation. Indeed, our results with the SV40 promoter suggests that HDAC8 functions as a transcriptional repressor. The ability to repress a transiently transfected promoter is consistent with the observation that HDAC1 can also repress transcription from viral promoters (Rous sarcoma virus long terminal repeat) (59). In contrast, transfection of HAT such as p300 and PCAF (62) activates transfected reporter constructs. The SV40 early promoter is known to be regulated by acetylation (47), probably through multiple Sp1 and/or AP1 sites present in the SV40 early promoter. Sp1 sites have been shown to be influenced by TSA and butyrate in the p21 (WAF1/Cip1) promoter (63). It is presently unclear how HDAC influences transfected target gene promoter activity. It is known that transfected DNA rapidly assembles into mini-chromosomes with histone attached (64-66). It is therefore possible that by removing additional acetyl groups from histone H3 or H4 and shifting the balance of histone acetylation, HDAC8 may condense the structure of these mini-chromosomes and restrict the access to transcription activators. It should be noted that histone deacetylase enzyme activity has been shown to be required for gene repression in vivo (18, 19, 67, 68). Alternatively, histone deacetylases may utilize other transcriptional factors as substrates, and the addition and removal of acetyl groups on these transcription factors could be crucial in controlling transcription initiation. The modification of ACTR and subsequent transcriptional activation attenuation induced by estrogen is an example of the importance of acetylation (69). In this instance, acetylation of ACTR by p300 prevented its interaction with estrogen receptor and other co-activators. Additional transcription factors that have been shown to be acetylated include GATA-1 (70), p53 (71), and the high mobility group (HMG) proteins (72). Besides utilizing transcription factors as acetylation substrates, HDAC may also sequester critical co-factors away from the transcriptional initiation complex by protein-protein interaction, thus repressing gene activation. A recent investigation suggesting a direct interaction between sp1 and HDAC1 (73) lends further support for this view. Whether any of these mechanisms play a dominant role in HDAC8-mediated transcription repression on SV40 early promoter warrants further investigation.

HDAC inhibitors have shown potential as anti-cancer agents (74-77). Indeed, two compounds have shown promise in suppressing cancer growth in several animal tumor models (46, 47, 57, 78). It should be noted, however, that these anti-tumor compounds were isolated based on properties other than their ability to inhibit HDACs. Thus, It is unclear whether the anti-tumor efficacy of these compounds is a direct effect of HDAC inhibition. It is possible that specific HDAC inhibitors will provide more potent anti-cancer drugs. The discovery of multiple forms of HDAC, especially those expressed in cancer tissues and cells, suggest that HDACs, including HDAC8, might be good targets for anti-tumor therapeutics.

In summary, we have isolated a novel human class I HDAC, HDAC8, from kidney and this enzyme is expressed in a wide range of human tissues as well as several human cancer cell lines. HDAC8 acetylates histone and is localized in the nucleus. Furthermore, HDAC8 functions as a transcriptional repressor toward a SV40 early promoter. These results provide a new tool to study the physiological role of HDACs in normal and cancer cells.

	ACKNOWLEDGEMENT

We are grateful for the helpful comments and suggestions by Drs. Nick Laping, Richard Edwards, Miklos Gellei, and Laura Fitzgerald. We also thank Drs. David Brooks, Randall Johnson, Alan Shatzman, Derk Bergsma, and Ralph Reverie for their support.

	FOOTNOTES

^* The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The nucleotide sequence(s) reported in this paper has been submitted to the GenBank^TM/EMBL Data Bank with accession number(s) AF230097.

^§ To whom correspondence should be addressed: Dept. of Renal Pharmacology, SmithKline Beecham Pharmaceuticals, 709 Swedeland Rd., King of Prussia, PA 19403. Tel.: 610-270-6087; Fax: 610-270-5681; E-mail: erding_hu-1@sbphrd.com.

Published, JBC Papers in Press, March 9, 2000, DOI 10.1074/jbc.M908988199

² T. Tomihama, K. Shoji, H. Hanyu, and T. Okano, unpublished results.

	ABBREVIATIONS

The abbreviations used are: HAT, histone acetyltransferase; HDAC, histone deacetylase; IP, immunoprecipitation; EST, expressed sequence tag; PAGE, polyacrylamide gel electrophoresis; PCR, polymerase chain reaction; TSA, trichostatin; TBS, Tris-buffered saline; DAPI, 4,6-diamidino-2-phenylindole; kb, kilobase pair(s).

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES