During inflammatory events, the induction of immediate-early genes, such as TNF-α, is regulated by signaling cascades including the JAK/STAT, NF-κB, and the p38 MAPK pathways, which result in phosphorylation-dependent activation of transcription factors. We observed the direct interaction of histone deacetylase (HDAC) 3, a class I histone deacetylase, with MAPK11 (p38 β isoform) by West-Western-based screening analysis, pull-down assay, and two-hybrid system analysis. Results further indicated that HDAC3 decreases the MAPK11 phosphorylation state and inhibits the activity of the MAPK11-dependent transcription factor, activating transcription factor-2 (ATF-2). LPS-mediated activation of ATF-2 was inhibited by HDAC3 in a time- and dose-dependent manner. Inhibition of HDAC3 expression by RNA interference resulted in increased ATF-2 activation in response to LPS stimulation. In agreement with decreased ATF-2 transcriptional activity by HDAC3, HDAC3-repressed TNF gene expression, and TNF protein production observed in response to LPS stimulation. Therefore, our results indicate that HDAC3 interacts directly and selectively with MAPK11, represses ATF-2 transcriptional activity, and acts as a regulator of TNF gene expression in LPS-stimulated cells, especially in mononuclear phagocytes.

Mechanisms of transcriptional activation have received considerable attention in recent years; however, it is becoming evident that transcriptional repression, which may be mediated by a number of different mechanisms, is just as important (1). Relaxation of the chromatin fiber facilitates transcription and is regulated by two competing enzymatic activities, histone acetyltransferases (HATs)3 and histone deacetylases (HDACs), which modify the acetylation state of histone proteins and other promoter-bound transcription factors (2). A still growing number of HDACs has been identified in various systems and model organisms (3). On the basis of their similarity to corresponding yeast ancestor proteins, these enzymes have been classified into three distinct families of which class I HDACs (HDAC1, 2, 3, and 8) are closely related to the yeast transcriptional regulator, RPD3 (4). Class II HDACs (HDAC4, 5, 6, 7, 9, and 10) are similar to the yeast deacetylase, Hda1p (5). Finally, a third class of HDACs, which contains the NAD-dependent sirtuin proteins (SIRT1-SIRT7), which are homologous to yeast SIR2, has recently been identified (6). The human histone deacetylase, HDAC3, is a 49 kDa protein that is expressed by many different cell lines and multiple normal human tissues and is located both in the cell nucleus and in the cytoplasm (4, 7, 8). Although the C-terminal proportion of HDAC3 contributes to the protein’s localization in the nucleus via a nuclear localization sequence, which is sufficient for the transport of a reporter fusion protein into the nucleus and is required for both its deacetylase and transcriptional repression activities, the central portion of the HDAC3 protein possesses a nuclear export signal (8). HDAC3 preferentially deacetylates histone lysines 5 and 12 of H4 and lysine 5 of H2A (9). HDAC3 forms oligomers in vitro and in vivo and gains its enzymatic activity in association with multiprotein repressor complexes that contain silencing mediator for retinoid and thyroid hormone receptors and nuclear receptor corepressors (4, 10, 11). Apart from deacetylating histone proteins and its capability to alter chromosome structure, which affects transcription factor access to DNA, HDAC3 has also been described to directly associate with nonhistone proteins, such as the ATP-dependent chaperone protein, heat shock protein 70 (9) and with class II HDACs (7). In addition, HDAC3 has been linked with the direct suppression of the transcriptional potential of the transcription factors GATA-2, YY1, and TFII-I (12, 13, 14) and with the deacetylation of RelA, a subunit of NF-κB, which has been reported to be exported from the nucleus in an IkBα-dependent fashion (15). HDAC3 has also been demonstrated to inhibit the JNK pathway via G-protein pathway suppressor 2, which is a component of the nuclear receptor corepressor-HDAC3 complex (16).

Four MAPK subgroups have been identified in humans: ERK, JNK/stress-activated protein kinases, the ERK5/big MAPK protein, and p38 MAPK. The p38 group of MAPK contains four isoforms: p38α, p38β, p38γ, and p38δ (17). The p38 MAPK are strongly activated by cytokines, LPS, and environmental stress, but are poorly activated by growth factors and phorbol ester (17). They are widely expressed in many tissues and require dual phosphorylation on threonine and tyrosine residues to become activated. This phosphorylation is mediated by a protein kinase cascade, which consists of MAPK kinases (MKKs) that include MKK3 and MKK6 (17). MAPK11 is known to activate downstream substrates like the transcription factors serum response factor accessory protein-1a and activating transcription factor-2 (ATF-2) through phosphorylation (18). ATF-2 is a member of the ATF-CREB family of transcription factors, which have been implicated in cell cycle progression, cell differentiation, transformation, and immune response (19). Being a leucine zipper transcription factor, ATF-2 binds an 8-bp response element (5′-TGACGTCA-3′), while as a homo-/heterodimer it associates with other members of the ATF family or with members of the Jun/Fos family of transcription factors (20). Most common is the ATF-2/c-Jun heterodimer, which recognizes the AP-1/cAMP response element (CRE) target sequence (20). Upon its phosphorylation on Ser121, ATF-2 associates with p300/CBP, which links it to the basal transcriptional complex (21). Like p300, ATF-2 was also reported to elicit HAT activities that are increased upon its phosphorylation (18).

TNF-α exerts a variety of biological effects, including production of inflammatory cytokines, up-regulation of adhesion molecules, proliferation, differentiation, and cell death (22). Monocytic cells release TNF in response to many stimuli, including the Gram-negative bacterial endotoxin, LPS, TNF itself, phorbol esters, superantigens, and viral agents (23, 24). Regulation is both transcriptional and posttranscriptional, depending on the stimulus, cell type, and possibly differentiation (25, 26). Sequences in the proximal 172 bp and in the −627 to −487 bp region of the human TNF promoter, at least, contribute to transcriptional control in monocytic cells (24, 26, 27). Each of the above stimuli modifies transcription factors interaction with the −116 to −88 bp region. This region includes putative CRE/ATF, NF-κB, CCAAT/enhancer-binding protein, and Ets binding sites (24). Factors binding to the CRE/ATF site (−106 to −99 bp) and the NF-κB site (−97 to −88 bp) cooperate functionally in both monocytic and lymphocytic cells (24, 25).

The purpose of this study was the identification and characterization of nonhistone interaction partners of the human histone deacetylase, HDAC3. In this study, we show that HDAC3, a class I HDAC, interacts with MAPK11, inhibits the transcriptional activity of the MAPK11-dependent transcription factor, ATF-2, and represses TNF gene expression. All together, our data indicate that HDAC3 plays an important role in the regulation of TNF expression in monocytic cells.

The pcDNA3.1-based expression vector for human HDAC3 has been described elsewhere (4). Deletion constructs were generated by standard PCR and cloning procedures (28). HDAC3 constructs were verified by DNA sequencing. MAPK11 cDNAs (GenBank U53442) were kindly provided by Dr. J. Han (The Scripps Research Institute, La Jolla, CA) and the Reference Center of the German Human Genome Project (Deutsches Ressourcenzentrum für Genomforschung (RZPD), Berlin, Germany; GenBank AF031135). The luciferase reporter plasmid driven by the human TNF promoter (pTNF-Luc) was kindly provided by Dr. D. Kwiatkowski (International Child Health Group, University Department of Pediatrics, John Radcliffe Hospital, Oxford, U.K.; Ref. 26). In vivo detection of HDAC3/MAPK11 interactions was done by transient cotransfections of GAL4-HDAC3 and VP16-MAPK11 in the context of a mammalian two-hybrid system (CheckMate; Promega, Mannheim, Germany). In vivo assessment of ATF-2 activation and the upstream p38 signal transduction pathway was done with the PathDetect in vivo signal transduction pathway trans-reporting system from Stratagene (La Jolla, CA). Sequence authenticity was confirmed by cDNA cycle sequencing.

Most of the study was performed with the promonocytic U937 cells and with the Cos7 cells obtained from the American Type Culture Collection (Manassas, VA) and with human THP-1 promonocytes kindly provided by Dr. G. Pancino (Institut Pasteur, Paris, France). U937 and THP-1 cells were cultivated in RPMI 1640 supplemented with 10% FBS. Cos7 cells were cultured in DMEM (Invitrogen Life Technologies, Carlsbad, CA) supplemented with 10% FCS, 1% glutamine, and 1% penicillin/streptomycin. Primary monocyte-derived macrophages (MΦ) were prepared from the peripheral blood of healthy donors and were cultured in RPMI 1640 medium supplemented with 10% (v/v) pooled AB human serum (Sigma-Aldrich, Munich, Germany), as previously described (29).

To carry out transient transfections, the DNA concentration was kept constant in the different samples by using the corresponding empty vector. A total of 5 × 106 cells were transfected with 5–10 μg of total plasmid DNA, which included the firefly luciferase-expressing vector using the DEAE-dextran procedure (30). At 48 h posttransfection, luciferase activity was measured in cell lysates using a luminometer (TD-20/20; Promega) as previously described (30). Luciferase expression was normalized with respect to protein concentrations using the Detergent-Compatible Protein Assay (Bio-Rad, Munich, Germany). For HDAC3 RNA interference experiments, 10 μg of pBS/U6/siHDAC3 (Upstate Biotechnology, Lake Placid, NY) were incubated with 106 cells in 300 μl of medium, 10% FCS, and electroporated at 200 V. Cell lysates were prepared 3 days postelectroporation and analyzed by Western blotting using anti-HDAC3 (Upstate Biotechnology) and anti-tubulin Abs (Sigma-Aldrich). For reporter gene assays in U937 cells, 2 μg of pATF-2 were cotransfected with 2 μg of pBS/U6/SiHDAC3 and luciferase activity was measured 3 days after transfection. To examine TNF promoter-driven gene expression, 107 U937 cells were cotransfected with 4 μg of pTNF-Luc and 2 μg of pHDAC3 vector or an empty control vector, using the DEAE-dextran procedure (30). Twenty-four hours later, the cells were stimulated with LPS at 100 ng/ml (Sigma-Aldrich) and, at 48 h posttransfection, luciferase activity was measured. All transfections were done in triplicate.

Nuclear extracts were prepared from U937 and THP-1 cells transfected with a HDAC3 expression vector or the empty control vector and left unstimulated or stimulated with 100 or 1000 ng/ml LPS during different periods of time. To measure AP-1 and NF-κB activation, EMSA were conducted as previously described (30, 31). To determine ATF-2 activation, we used the Nufshift ATF-2 kit (Active Motif, Rixensart, Belgium). Nuclear extracts were incubated with 32P-end-labeled 23-mer double-stranded ATF-2 oligonucleotide, 5′-GATTCAATGACATCACGGCTGTG-3′ (bold indicates ATF-2 binding sites). The specificity of the binding was examined by competition with a mutated unlabeled oligonucleotide 5′-GATTCAAGAACATAGCGGCTGTG-3′. The DNA-protein complex formed was analyzed on a 6% native polyacrylamide gel. The dried gels were visualized and radioactive bands quantified by a PhosphorImager (Molecular Dynamics, Sunnyvale, CA) using ImageQuant software (Amersham Biosciences, Piscataway, NJ). The double-stranded oligonucleotides with AP-1 or NF-κB-binding motifs were obtained from Eurobio (Paris, France).

U937 cells were transfected with a HDAC3 expressing vector (pHDAC3) or the empty vector (mock), and treated 24 h later with 100 ng/ml LPS. After 0, 7, and 28 min of treatment, cells were treated with 1% paraformaldehyde at room temperature. After 10 min, the cells were harvested, and chromatin immunoprecipitation assays were performed in accordance with the manufacturer’s instructions (Upstate Biotechnology). Immunoprecipitations were done overnight at 4°C with 10 μg of anti-ATF-2 Ab or 10 μg of anti-cFos Ab used as an irrelevant Ab (Active Motifs). Immune complexes were collected with protein A-agarose preblocked with sonicated salmon sperm DNA and BSA (Upstate Biotechnology). Immunoprecipitates were washed three times in washing buffer (immunoprecipitation buffer supplemented with 0.1 mg/ml yeast tRNA) and three times in washing buffer containing 500 mM NaCl. Immune complexes were eluted with 1% SDS and 100 mM NaHCO3. Formaldehyde cross-links were reversed by incubating the samples with 200 mM NaCl at 65°C for 4 h. The immunoprecipitated DNA was purified by proteinase K treatment, phenol-chloroform extractions, and ethanol precipitation. Immunoprecipitated DNA and input (nonimmunoprecipitated) chromatin were analyzed by PCR using the following primer pair for human TNF promoter detection (sense, 5′-GTC CCC AAC TTT CCA AAT CC-3′; and antisense, 5′-CAA CCA GCG GAA AAC TTC CTT-3′). PCR was performed using the following primer pair for human β-globin gene detection (sense, 5′-GAA GAG CCA AGG ACA GGT AC-3′; and antisense, 5′-CAA CTT CAT CCA CGT TCA CC-3′). Amplification products were run on a 2% agarose gel. Quantitative real-time PCR was performed according to manufacturer’s instructions using the ABI Prism 7000 (Applied Biosystems, Foster City, CA).

Protein-protein interaction screening was performed on hEx1 high-density protein expression cDNA libraries with recombinant HDAC3 protein which was expressed from the prokaryotic expression vector pET-32 (Novagen, Madison, WI) according to previously published protocols (32).

Cos7 cells, U937 cells, or primary MΦ were left untreated or were treated with 100 ng/ml LPS for different periods of time. Cell lysates were precleared by adding 50 μl of Protein G Plus/Protein A-Agarose (Calbiochem-Novabiochem, Bad Soden, Germany) during 1 h at 4°C. The cleared supernatants were removed, combined with 10 μg/ml anti-HDAC3 or anti-MAPK11 (Santa Cruz Biotechnology, Santa Cruz, CA) Abs and incubated overnight at 4°C. Immune complexes were washed in the presence of protease inhibitors and the bound proteins were eluted with sample buffer and run on a 10% SDS-PAGE gels. SDS-PAGE and Western blot analysis were performed according to standard procedures (28). Western blots were developed with the ECL detection kit (Amersham Biosciences, Piscataway, NJ).

Fragments of cDNA encoding human HDAC3 were produced using convenient restriction enzymes and PCR methods and then cloned into the GST fusion vector pGEX-4T-3 (Amersham Biosciences). The GST constructs were transformed into the Escherichia coli strain, BL21, (Stratagene) and the GST fusion proteins were purified according to manufacturer’s instructions. MAPK11 was transcribed and translated in vitro in the presence of [35S]methionine (Amersham Biosciences) with the SP6 or T7-coupled reticulocyte lysate systems for 90 min, 30°C in accordance with the manufacturer’s instructions (Promega). For pull-down assays, 20 μg of the GST fusion proteins were incubated overnight at 4°C with 5 μl of the MAPK11 in vitro-transcribed and translated reaction mixture in PBS. The suspension was then washed three times in PBS, denaturized, and subsequently separated by SDS-PAGE before autoradiography.

Kinase assays were then conducted with the TNT T7 (or SP6) Quick Coupled Transcription/Translation (Promega) in accordance with the manufacturer’s instructions. pcDNA3.1-HDAC3 and the pcDNA3.1 empty vectors were transcribed and translated in the presence of 1 mM methionine without biotinylated tRNA, while MAPK11 was transcribed and translated in vitro under the same conditions but in the presence of 2 μl of biotinylated tRNA (Promega) with the T7-coupled reticulocyte lysate system for 90 min at 30°C. After incubation for 5 min at room temperature, [32P]MAPK11 samples were washed three times (wash solution; Calbiochem-Novabiochem) from the free [γ-32P]ATP and collected on a streptavidin membrane, which was followed by the addition of a liquid scintillation mixture (Roth, Karlsruhe, Germany) and quantification on a beta counter (PerkinElmer, Dreieich, Germany).

To identify potential protein/protein interactions of human HDAC3, we screened high-density protein expression filter membranes that contained 55,296 clones from a human fetal brain library by Far-Western analysis with recombinant HDAC3 as a bait (32). We identified MAPK11 as a potential binding candidate (Fig. 1,A). To further test whether this interaction was direct or indirect, we expressed HDAC3 as a GST fusion protein in E. coli and tested its ability to interact with in vitro translated MAPK11. MAPK11 bound to HDAC3 (Fig. 1,B). The two bands detected on the gel correspond to phosphorylated and unphosphorylated MAPK11, as demonstrated using an anti-phospho-MAPK11 Ab in parallel (data not shown). Endogenous MAPK11 coimmunoprecipitated with HDAC3 in Cos7 cells, promonocytic U937 cells, and primary MΦ, whereas the isotype control immunoprecipitation showed no associated MAPK11 protein (Fig. 1,C). The interaction of HDAC3 and MAPK11 was also demonstrated by transient transfection of Cos7 cells (data not shown). These data indicate that HDAC3 interacts with MAPK11, not only in vitro, but also within several cell types, including primary MΦ. Lysates of Cos7 cells were immunoprecipitated with Abs directed against other members of class I HDACs (HDAC1, HDAC2, and HDAC8) and then detected by Western blot with the anti-MAPK11 Ab. MAPK11 did not interact endogenously with other members of class I HDACs (data not shown), indicating that HDAC3/MAPK11 interaction was specific. We further investigated the interaction between HDAC3 and MAPK11 using a mammalian two-hybrid assay. We used HDAC3, which was fused to the GAL4 DNA-binding domain (pBIND) as the bait vector and MAPK11, which was fused to the VP16-activation domain (pACT), as the prey vector (Fig. 2,A). These constructs were transfected into mammalian cells along with the pG5Luc vector, which contains five GAL4 binding sites upstream of a minimal TATA box upstream of the luciferase gene. The interaction between the GAL4-HDAC3 and VP16-MAPK11 fusion constructs resulted in a 4-fold increase of relative firefly luciferase expression over the negative control (20,345 vs 5,259 relative light units (rlu); Fig. 2,B). Expression of pGAL4-HDAC3 alone or pVP16-MAPK11 alone did not result in increased luciferase activity vs negative controls (5305 and 378 rlu, respectively; Fig. 2 B).

FIGURE 1.

HDAC3 and MAPK11 interact both in vitro and in vivo. A, Image analysis of a high-density protein filter array, which has been screened by nonradioactive West-Western blotting using human HDAC3 protein as a probe. A magnified section of the blot is shown, indicating the interacting double-spotted MAPK11 clones (black). Guide dots (white) on the high-density filter membrane facilitate the identification of xy-coordinates of positive clones. B, Binding of MAPK11 to HDAC3 was measured in pull-down assays with in vitro translated 35S-labeled MAPK11 (left lane, input; middle lane, pGEX-4T-3 empty vector; right lane, 428 aa full-length HDAC3). The double bands represent phosphorylated and unphosphorylated MAPK11. Input corresponds to 10% of the material used for pull-down. Results are representative of three independent experiments. C, Total cellular extracts from Cos7 cells, U937 cells, or primary MΦ, were immunoprecipitated with antisera specific for HDAC3 (upper panel) or MAPK11 (lower panel), or isotype controls. Immunoprecipitated material was analyzed by Western blotting with antisera against MAPK11 (upper panel) or HDAC3 (lower panel).

FIGURE 1.

HDAC3 and MAPK11 interact both in vitro and in vivo. A, Image analysis of a high-density protein filter array, which has been screened by nonradioactive West-Western blotting using human HDAC3 protein as a probe. A magnified section of the blot is shown, indicating the interacting double-spotted MAPK11 clones (black). Guide dots (white) on the high-density filter membrane facilitate the identification of xy-coordinates of positive clones. B, Binding of MAPK11 to HDAC3 was measured in pull-down assays with in vitro translated 35S-labeled MAPK11 (left lane, input; middle lane, pGEX-4T-3 empty vector; right lane, 428 aa full-length HDAC3). The double bands represent phosphorylated and unphosphorylated MAPK11. Input corresponds to 10% of the material used for pull-down. Results are representative of three independent experiments. C, Total cellular extracts from Cos7 cells, U937 cells, or primary MΦ, were immunoprecipitated with antisera specific for HDAC3 (upper panel) or MAPK11 (lower panel), or isotype controls. Immunoprecipitated material was analyzed by Western blotting with antisera against MAPK11 (upper panel) or HDAC3 (lower panel).

Close modal
FIGURE 2.

HDAC3 and MAPK11 interact in vitro in a mammalian two hybrid assay. A, Schematic representation of expression constructs, which have been used for cotransfection experiments in this mammalian two hybrid setting. B, Mammalian two hybrid analysis with HDAC3 being fused to the GAL4 DNA-binding domain and MAPK11 being fused to the VP16 activator domain. Luciferase assays were conducted on total extracts from Cos 7 cells that have been transfected with the luciferase expressing construct pG5-Luc, pBIND-HDAC3, pACTMAPK11, or control plasmids. Results represent the average of a triplicate experiment where luciferase activity was normalized to protein expression. As a positive control, two plasmids, pACT-MyoD and pBIND-Id, were cotransfected while cotransfection of their empty vectors was used as a negative control. Results represent mean values of three independent experiments.

FIGURE 2.

HDAC3 and MAPK11 interact in vitro in a mammalian two hybrid assay. A, Schematic representation of expression constructs, which have been used for cotransfection experiments in this mammalian two hybrid setting. B, Mammalian two hybrid analysis with HDAC3 being fused to the GAL4 DNA-binding domain and MAPK11 being fused to the VP16 activator domain. Luciferase assays were conducted on total extracts from Cos 7 cells that have been transfected with the luciferase expressing construct pG5-Luc, pBIND-HDAC3, pACTMAPK11, or control plasmids. Results represent the average of a triplicate experiment where luciferase activity was normalized to protein expression. As a positive control, two plasmids, pACT-MyoD and pBIND-Id, were cotransfected while cotransfection of their empty vectors was used as a negative control. Results represent mean values of three independent experiments.

Close modal

To determine the MAPK11 binding domain(s), we generated a series of plasmids expressing sequential C-terminal truncations of the GST-tagged HDAC3 protein at amino acid positions 3, 31, 49, 59, 66, 137, 198, 266, and 331 (Fig. 3). We then analyzed their MAPK11 binding capacity by GST pull-down assays (Fig. 3). MAPK11 binding capacity was totally annihilated in GST pull-down assays using GST-tagged HDAC3 constructs with truncations upstream of aa 49 (Fig. 3). These results are consistent with the hypothesis that HDAC3 interacts with MAPK11 via its N terminus.

FIGURE 3.

HDAC3 interacts withMAPK11 via its N terminus. A, MAPK11 binds to HDAC3 via its N terminus as demonstrated using wild-type or truncated HDAC3 constructs. Binding of MAPK11 to the C or N terminus of HDAC3 was measured in pull-down assays with in vitro translated 35S-labeled MAPK11. Input corresponds to 10% of the material used for pulldown. Results are representative of two independent experiments. B, Schematic representation of wild-type (428 aa) and truncated HDAC3 constructs. HDAC3 functional domains contain the MAPK11 binding site (upstream aa 49), the oligomerization domain (aa 1–122), the nuclear export signal (aa 180–313), the nuclear localization signal (aa 313–428), and the HDAC domain (aa 401–428; Ref. 8 ).

FIGURE 3.

HDAC3 interacts withMAPK11 via its N terminus. A, MAPK11 binds to HDAC3 via its N terminus as demonstrated using wild-type or truncated HDAC3 constructs. Binding of MAPK11 to the C or N terminus of HDAC3 was measured in pull-down assays with in vitro translated 35S-labeled MAPK11. Input corresponds to 10% of the material used for pulldown. Results are representative of two independent experiments. B, Schematic representation of wild-type (428 aa) and truncated HDAC3 constructs. HDAC3 functional domains contain the MAPK11 binding site (upstream aa 49), the oligomerization domain (aa 1–122), the nuclear export signal (aa 180–313), the nuclear localization signal (aa 313–428), and the HDAC domain (aa 401–428; Ref. 8 ).

Close modal

Using in vitro translated and transcribed MAPK11, human HDAC3 decreased significantly the MAPK11 phosphorylation state, as measured by an in vitro kinase assay (Fig. 4,A). Because HDAC3 reduced the MAPK11 phosphorylation state in vitro, we further assessed HDAC3 effect on the MAPK11 phosphorylation state in vivo. Because LPS has been reported to increase the MAPK11 phosphorylation state (33), we assessed the effect of HDAC3 on the LPS-induced MAPK11 phosphorylation state using an anti-phospho-p38 (Thr180/Tyr182) polyclonal Ab. LPS treatment increased the MAPK11 phosphorylation state which was significantly inhibited in U937 cells transiently transfected with an HDAC3 expressing vector construct, but not with an empty control vector (Fig. 4,B). Because activation of MAPK11 results in increased ATF-2 activation (34), we assessed the state of ATF-2 phosphorylation following LPS stimulation, in the presence or absence of HDAC3 overexpression. Increased ATF-2 phosphorylation was observed following LPS treatment of U937 cells and was significantly inhibited upon transient transfection of cells with a HDAC3-expressing vector construct but not with the empty vector (Fig. 4,C). All together, these data indicate that both MAPK11 and ATF-2 phosphorylation states are negatively regulated by HDAC3 in LPS-treated cells. Following LPS treatment, the amount of endogenous MAPK11/HDAC3 interacting complexes decreased transiently in primary MΦ (Fig. 4 D) and in U937 cells (data not shown), indicating that LPS relieves HDAC3/MAPK11 interaction.

FIGURE 4.

HDAC3 represses both MAPK11 and ATF-2 phosphorylation. A, The level of MAPK11 phosphorylation was assessed, in the absence and presence of human HDAC3 overexpression, using an in vitro kinase assay. Results represent means of three independent experiments. B, LPS-induced MAPK11 phosphorylation is inhibited by HDAC3. Total cellular extracts from U937 cells transiently transfected with an HDAC3-expressing vector construct or with an empty control vector, left untreated or stimulated with 1 μg/ml LPS for different periods of time, were analyzed by Western blotting with an antiserum directed against phospho-MAPK11 or against tubulin as a control (data not shown). Results are representative of two independent experiments. C, LPS-induced ATF-2 phosphorylation is inhibited by HDAC3. Total cellular extracts from U937 cells transiently transfected with an HDAC3-expressing vector construct or with an empty vector, left untreated or stimulated with 1 μg/ml LPS for different periods of time, were analyzed by Western blotting with an antiserum directed against phospho-ATF-2 or against tubulin as a control. Results are representative of two independent experiments. D, Total cellular extracts from primary MΦ left untreated or treated with 100 ng/ml LPS for different periods of time were immunoprecipitated with an antiserum specific for MAPK11 or an isotype control (data not shown). Immunoprecipitated material was analyzed by Western blotting with antisera against HDAC3. Results are representative of two independent experiments.

FIGURE 4.

HDAC3 represses both MAPK11 and ATF-2 phosphorylation. A, The level of MAPK11 phosphorylation was assessed, in the absence and presence of human HDAC3 overexpression, using an in vitro kinase assay. Results represent means of three independent experiments. B, LPS-induced MAPK11 phosphorylation is inhibited by HDAC3. Total cellular extracts from U937 cells transiently transfected with an HDAC3-expressing vector construct or with an empty control vector, left untreated or stimulated with 1 μg/ml LPS for different periods of time, were analyzed by Western blotting with an antiserum directed against phospho-MAPK11 or against tubulin as a control (data not shown). Results are representative of two independent experiments. C, LPS-induced ATF-2 phosphorylation is inhibited by HDAC3. Total cellular extracts from U937 cells transiently transfected with an HDAC3-expressing vector construct or with an empty vector, left untreated or stimulated with 1 μg/ml LPS for different periods of time, were analyzed by Western blotting with an antiserum directed against phospho-ATF-2 or against tubulin as a control. Results are representative of two independent experiments. D, Total cellular extracts from primary MΦ left untreated or treated with 100 ng/ml LPS for different periods of time were immunoprecipitated with an antiserum specific for MAPK11 or an isotype control (data not shown). Immunoprecipitated material was analyzed by Western blotting with antisera against HDAC3. Results are representative of two independent experiments.

Close modal

To determine whether HDAC3 can modulate MAPK11-dependent transcriptional activity, we conducted transient cotransfections of pcDNA3.1-HDAC3 and pcDNA3.1-MAPK11 in the presence of the luciferase reporter gene pFR-Luc and the downstream MAPK11-dependent transcription factor pFA-ATF-2. The activity of ATF-2 is known to be induced by phosphorylated MAPK11 (18). We confirmed that overexpression of MAPK11 up-regulated transcriptional activity of ATF-2 (Fig. 5,A). Furthermore, MAPK11-mediated ATF-2 activation was inhibited by 80% in the presence of HDAC3 overexpression (Fig. 5,A). MKK3-mediated ATF-2 activation was inhibited following HDAC3 expression (Fig. 5,A). The MAPK11-mediated ATF-2 activation was almost completely blocked by the p38 kinase inhibitor, SB202190 (Fig. 5,A). To determine whether HDAC3 had a direct or indirect negative effect on ATF-2-mediated transcription, we cotransfected pcDNA3.1-HDAC3 together with pFA-ATF-2 in the presence of the luciferase reporter gene pFR-Luc. We found that HDAC3 had no direct inhibitory effect on ATF-2 activity (Fig. 5,A), indicating that HDAC3 does not directly repress ATF-2 transcriptional activity, but probably does it through its interaction with MAPK11. To directly examine the contribution of endogenous HDAC3 to ATF-2 repression, we performed loss-of-function studies. To decrease endogenous HDAC3 protein levels, we used a vector directing the expression of a small interference RNA (siRNA) designed to target the cellular HDAC3-encoding mRNA (35). Transient transfection of this vector led to an efficient depletion of HDAC3 levels in Cos7 cells and U937 cells (Fig. 5,B). Concomitantly, expression of HDAC3 siRNA completely abrogated transcriptional repression of ATF-2 following transient expression of HDAC3 in Cos7 cells (Fig. 5,A). These results confirm that HDAC3 is a crucial functional component that represses MAPK11-mediated ATF-2 activation. To assess directly the role of histone acetylation in ATF-2 transcriptional activity, we overexpressed transiently in Cos7 cells a HDAC3 mutant lacking the deacetylase enzymatic activity (HDACΔDEAC; Ref. 4). MAPK11-mediated ATF-2 activation was significantly increased in the presence of HDACΔDEAC overexpression (Fig. 5 C).

FIGURE 5.

HDAC3 represses MAPK11-mediated ATF-2 activation. A, Overexpression of HDAC3 inhibits MAPK11-mediated ATF-2 transcription. Upper panel, Overview of expression constructs, which have been used for cotransfection experiments of Cos7 cells in this experimental setting. Lower left panel, Cotransfection experiments of the luciferase-expressing construct pFR-Luc with pFA-ATF-2, which was fused to the GAL4 DNA-binding domain and pMAPK11, in the presence or absence of HDAC3, indicated that HDAC3 represses ATF-2 transcriptional activity mediated via MAPK11. Transient transfection of pFC-MKK3 in conjunction with ATF-2 was used as positive controls. Results represent the average of triplicate experiments where luciferase activity was normalized to protein expression. Lower right panel, Loss of HDAC3 function releases ATF-2 from repression. Expression constructs coding for ATF-2 were cotransfected with empty (siRNA) or HDAC3 siRNA expression vectors. Reporter activation was measured 2 days after transfection. Results represent the average of triplicate experiments where luciferase activity was normalized to protein expression. B, Expression of HDAC3 siRNA efficiently depletes endogenous HDAC3 protein levels in both U937 cells and Cos7 cells. Lysates of transfected cells were analyzed 3 days after transfection by immunoblotting using an anti-HDAC3 Ab and an anti-tubulin Ab as a control. C, MAPK11-mediated ATF-2 activation is significantly increased in the presence of a HDAC3 mutant lacking its deacetylase enzymatic activity (pHDACΔDEAC). Cotransfection experiments of the luciferase expressing construct pFR-Luc with pFA-ATF-2 and pMAPK11 in the presence or absence of pHDACΔDEAC, indicated that, besides the MAPK11 pathway, HDAC3 represses ATF-2 transcriptional activity via its deacetylase activity. Results represent the average of triplicate experiments where luciferase activity was normalized to protein expression.

FIGURE 5.

HDAC3 represses MAPK11-mediated ATF-2 activation. A, Overexpression of HDAC3 inhibits MAPK11-mediated ATF-2 transcription. Upper panel, Overview of expression constructs, which have been used for cotransfection experiments of Cos7 cells in this experimental setting. Lower left panel, Cotransfection experiments of the luciferase-expressing construct pFR-Luc with pFA-ATF-2, which was fused to the GAL4 DNA-binding domain and pMAPK11, in the presence or absence of HDAC3, indicated that HDAC3 represses ATF-2 transcriptional activity mediated via MAPK11. Transient transfection of pFC-MKK3 in conjunction with ATF-2 was used as positive controls. Results represent the average of triplicate experiments where luciferase activity was normalized to protein expression. Lower right panel, Loss of HDAC3 function releases ATF-2 from repression. Expression constructs coding for ATF-2 were cotransfected with empty (siRNA) or HDAC3 siRNA expression vectors. Reporter activation was measured 2 days after transfection. Results represent the average of triplicate experiments where luciferase activity was normalized to protein expression. B, Expression of HDAC3 siRNA efficiently depletes endogenous HDAC3 protein levels in both U937 cells and Cos7 cells. Lysates of transfected cells were analyzed 3 days after transfection by immunoblotting using an anti-HDAC3 Ab and an anti-tubulin Ab as a control. C, MAPK11-mediated ATF-2 activation is significantly increased in the presence of a HDAC3 mutant lacking its deacetylase enzymatic activity (pHDACΔDEAC). Cotransfection experiments of the luciferase expressing construct pFR-Luc with pFA-ATF-2 and pMAPK11 in the presence or absence of pHDACΔDEAC, indicated that, besides the MAPK11 pathway, HDAC3 represses ATF-2 transcriptional activity via its deacetylase activity. Results represent the average of triplicate experiments where luciferase activity was normalized to protein expression.

Close modal

In addition to MAPK11 and ATF-2 activation, LPS activates the transcription factors, NF-κB and AP-1 (19, 36, 37). Therefore, we investigated the ability of LPS to activate ATF-2, NF-κB, and AP-1 in monocytic cells U937 and THP-1, in the presence and absence of HDAC3 expression, as measured by EMSA. LPS activated ATF-2 in a time- and dose-dependent manner (Fig. 6,A, and data not shown). ATF-2 activation was time-dependent, with optimal activation occurring at ∼21 min (Fig. 6,A). Disappearance of the ATF-2 band by competition with unlabeled oligonucleotide indicates that the interaction wasspecific (Fig. 6,B). ATF-2 activation observed in response to LPS treatment disappeared following transient transfection with a HDAC3-expressing vector construct but not with the empty control vector (Fig. 6,C). To directly examine the contribution of endogenous HDAC3 to ATF-2 repression, we performed loss-of-function studies. Expression of HDAC3 siRNA resulted in increased ATF-2 DNA binding following LPS stimulation (Fig. 6,D), further indicating HDAC3-mediated transcriptional repression of ATF-2. Following LPS stimulation, activation of NF-κB and AP-1 has been reported in monocytic cells (19, 37). Therefore, we assessed NF-κB and AP-1 activation in LPS-treated U937 cells expressing HDAC3 siRNA. In LPS-treated U937 cells, expression of HDAC3 siRNA resulted in increased NF-κB activation (Fig. 6 E), but had no effect on AP-1 induction (data not shown). These results indicate that HDAC3 inhibits LPS-mediated ATF-2 activation, but also LPS-mediated NF-κB activation, in monocytic cells.

FIGURE 6.

HDAC3 represses LPS-mediated ATF-2 activation. A, Time course of ATF-2 activation following LPS treatment. U937 cells (1 × 106/ml) were treated with 1 μg/ml LPS for different periods of time at 37°C and then ATF-2 activation was measured. B, Specificity of ATF-2 activation by LPS. Nuclear extracts from U937 cells treated with 100 ng/ml LPS for 21 min were incubated 20 min with an unlabeled ATF-2 probe or mutated ATF-2 probe, and then ATF-2 activity was measured as described under Materials and Methods. C, ATF-2 activation following LPS treatment is inhibited by HDAC3 overexpression. THP-1 cells (1 × 106/ml) were transfected with a HDAC3-expressing vector (pHDAC3) or the empty vector (mock), and treated 24 h later with 100 ng/ml LPS for different periods of time at 37°C, and then ATF-2 activation was measured, as described under Materials and Methods. D, ATF-2 activation following LPS treatment is repressed by endogenous HDAC3. U937 cells (1 × 106/ml) were transfected with empty (siRNA) or HDAC3 siRNA expression vectors and treated 24 h later with 100 ng/ml LPS for different periods of time at 37°C, and ATF-2 activation was measured as described under Materials and Methods. E, Following LPS treatment, loss of HDAC3 function releases NF-κB from repression. THP-1 cells (1 × 106/ml) were transfected with empty (siRNA) or HDAC3 siRNA expression vectors and treated 24 h later with 100 ng/ml LPS for different periods of time at 37°C, and NF-κB activation was measured as described under Materials and Methods.

FIGURE 6.

HDAC3 represses LPS-mediated ATF-2 activation. A, Time course of ATF-2 activation following LPS treatment. U937 cells (1 × 106/ml) were treated with 1 μg/ml LPS for different periods of time at 37°C and then ATF-2 activation was measured. B, Specificity of ATF-2 activation by LPS. Nuclear extracts from U937 cells treated with 100 ng/ml LPS for 21 min were incubated 20 min with an unlabeled ATF-2 probe or mutated ATF-2 probe, and then ATF-2 activity was measured as described under Materials and Methods. C, ATF-2 activation following LPS treatment is inhibited by HDAC3 overexpression. THP-1 cells (1 × 106/ml) were transfected with a HDAC3-expressing vector (pHDAC3) or the empty vector (mock), and treated 24 h later with 100 ng/ml LPS for different periods of time at 37°C, and then ATF-2 activation was measured, as described under Materials and Methods. D, ATF-2 activation following LPS treatment is repressed by endogenous HDAC3. U937 cells (1 × 106/ml) were transfected with empty (siRNA) or HDAC3 siRNA expression vectors and treated 24 h later with 100 ng/ml LPS for different periods of time at 37°C, and ATF-2 activation was measured as described under Materials and Methods. E, Following LPS treatment, loss of HDAC3 function releases NF-κB from repression. THP-1 cells (1 × 106/ml) were transfected with empty (siRNA) or HDAC3 siRNA expression vectors and treated 24 h later with 100 ng/ml LPS for different periods of time at 37°C, and NF-κB activation was measured as described under Materials and Methods.

Close modal

Because LPS triggers the production of proinflammatory cytokines, we assessed the role of HDAC3 in the modulation of LPS-mediated TNF production. To determine whether HDAC3 can suppress the transcriptional activity of the human TNF promoter which contains ATF-2-binding domains, we assessed directly the levels of ATF-2 binding in the human TNF promoter during LPS stimulation. Therefore, we performed chromatin immunoprecipitation experiments. U937 cells were stimulated with LPS, in the presence and absence of HDAC3 expression. ATF-2 was immunoprecipitated from the chromatin solution, and the presence of TNF promoter sequences was analyzed by PCR. This analysis revealed that the TNF promoter became relatively enriched in ATF-2 during the first 28 min after LPS stimulation (Fig. 7,A). Enrichment of the TNF promoter in ATF-2 almost totally disappeared following transient transfection with a HDAC3-expressing vector construct but not with the empty control vector (Fig. 7,A). Following LPS stimulation, ATF-2 associates specifically (a 2.5-fold enrichment) with the promoter region of TNF, whereas no increase in binding was observed in the β-globin region used as a negative control (Fig. 7,A). Then, we examined the effect of HDAC3 on the transcriptional activity of the isolated TNF promoter in a reporter assay (26). LPS treatment increased the activity of the transfected TNF promoter construct by ∼2-fold in THP-1 cells (Fig. 7,B). Interestingly, expression of siRNA HDAC3 further increased LPS-induced TNF-promoter activity (Fig. 7,B; p < 0.05), indicating that endogenous HDAC3 represses TNF promoter activity in THP-1 cells. TNF mRNA levels were decreased by 60% in LPS-treated THP-1 cells transfected with a HDAC3-expressing vector vs LPS-treated cells transfected with an empty vector, as measured by RT-PCR assay (Fig. 7,C), further indicating that the inhibition of TNF production by HDAC3 occurred at the transcriptional level. LPS-treated THP-1 cells were treated with an HDAC inhibitor, trichostatin A (TSA), and expression of TNF mRNA was assessed by RT-PCR analysis. LPS-mediated TNF mRNA up-regulation was repressed by HDAC3 and was relieved after the addition of TSA to the culture medium (Fig. 7,D). These data further indicate that the TNF promoter is suppressed by the activity of an HDAC. The production of the TNF protein was almost totally abolished at 8 h after LPS treatment in THP-1 cells expressing transiently pHDAC3 vs LPS-treated control cells transfected with the empty vector (Fig. 7 E). All together, these data indicate that HDAC3 represses TNF expression in LPS-treated monocytic cells.

FIGURE 7.

HDAC3 represses LPS-mediated TNF expression. A, TNF-promoter enrichment in ATF-2 following LPS treatment is inhibited by HDAC3 overexpression. U937 cells (1 × 106/ml) were transfected with a HDAC3-expressing vector (pHDAC3) or the empty control vector (mock), and treated 24 h later with 100 ng/ml LPS for different periods of time at 37°C, and then enrichment of the TNF promoter in ATF-2 was measured using chromatin immunoprecipitation as described under Materials and Methods. Briefly, cross-linked cell extracts from mock or HDAC3-transfected cells expressing chromosomal ATF-2 were immunoprecipitated by anti-ATF-2 Abs and assayed for the presence of the TNF promoter or β-globin regions. Left panel, Amplification products were run on a 2% agarose gel. As a negative control, DNA was amplified from extracts tagged with c-Fos. Input DNA (control) is shown to demonstrate equal DNA amount. Results are representative of two independent experiments. Right panel, Time course of the recruitment of ATF-2 to the TNF promoter. ATF-2 binding to the TNF promoter was measured using quantitative real-time PCR amplification. Results are represented as a histogram of the relative enrichment (immunoprecipitated/input) in DNA associated with ATF-2 relative to untreated cells. Results are representative of two independent experiments. B, Endogenous HDAC3 represses human TNF promoter activity. U937 cells were cotransfected with pTNF-Luc and with empty (siRNA) or HDAC3 siRNA expression vectors, and treated 24 h later with 100 ng/ml LPS for 20 min. Luciferase activity was measured in cell lysates. The results represent means of three independent experiments. C, LPS-mediated TNF gene expression is repressed by HDAC3. THP-1 cells were transiently transfected with an HDAC3-expressing vector construct or with an empty control vector (mock). Twenty-four hours posttransfection, the cells were left untreated or were treated with SB202190 100 μM for 1 h before treatment with 500 ng/ml LPS for 2 h. Total cellular RNA extracts and real-time RT-PCR quantification of human TNF mRNA were performed as described under Materials and Methods. Results represent mean values of three independent experiments. D, LPS-mediated TNF gene expression depends on histone acetylation. U937 cells were transiently transfected with an HDAC3-expressing vector construct or with an empty control vector and treated with 500 ng/ml LPS for 2 h, in the presence or absence of TSA at 1 μM. Total cellular RNA extracts and real-time RT-PCR quantification of human TNF mRNA were performed as described under Materials and Methods. Results are representative of two independent experiments. E, LPS-mediated TNF protein production is repressed by HDAC3. THP-1 cells were transiently transfected with an HDAC3-expressing vector or an empty control vector and treated with 10 ng/ml LPS for different periods of time. TNF protein amounts were quantified in culture supernatants as described under Materials and Methods. Results are representative of two independent experiments.

FIGURE 7.

HDAC3 represses LPS-mediated TNF expression. A, TNF-promoter enrichment in ATF-2 following LPS treatment is inhibited by HDAC3 overexpression. U937 cells (1 × 106/ml) were transfected with a HDAC3-expressing vector (pHDAC3) or the empty control vector (mock), and treated 24 h later with 100 ng/ml LPS for different periods of time at 37°C, and then enrichment of the TNF promoter in ATF-2 was measured using chromatin immunoprecipitation as described under Materials and Methods. Briefly, cross-linked cell extracts from mock or HDAC3-transfected cells expressing chromosomal ATF-2 were immunoprecipitated by anti-ATF-2 Abs and assayed for the presence of the TNF promoter or β-globin regions. Left panel, Amplification products were run on a 2% agarose gel. As a negative control, DNA was amplified from extracts tagged with c-Fos. Input DNA (control) is shown to demonstrate equal DNA amount. Results are representative of two independent experiments. Right panel, Time course of the recruitment of ATF-2 to the TNF promoter. ATF-2 binding to the TNF promoter was measured using quantitative real-time PCR amplification. Results are represented as a histogram of the relative enrichment (immunoprecipitated/input) in DNA associated with ATF-2 relative to untreated cells. Results are representative of two independent experiments. B, Endogenous HDAC3 represses human TNF promoter activity. U937 cells were cotransfected with pTNF-Luc and with empty (siRNA) or HDAC3 siRNA expression vectors, and treated 24 h later with 100 ng/ml LPS for 20 min. Luciferase activity was measured in cell lysates. The results represent means of three independent experiments. C, LPS-mediated TNF gene expression is repressed by HDAC3. THP-1 cells were transiently transfected with an HDAC3-expressing vector construct or with an empty control vector (mock). Twenty-four hours posttransfection, the cells were left untreated or were treated with SB202190 100 μM for 1 h before treatment with 500 ng/ml LPS for 2 h. Total cellular RNA extracts and real-time RT-PCR quantification of human TNF mRNA were performed as described under Materials and Methods. Results represent mean values of three independent experiments. D, LPS-mediated TNF gene expression depends on histone acetylation. U937 cells were transiently transfected with an HDAC3-expressing vector construct or with an empty control vector and treated with 500 ng/ml LPS for 2 h, in the presence or absence of TSA at 1 μM. Total cellular RNA extracts and real-time RT-PCR quantification of human TNF mRNA were performed as described under Materials and Methods. Results are representative of two independent experiments. E, LPS-mediated TNF protein production is repressed by HDAC3. THP-1 cells were transiently transfected with an HDAC3-expressing vector or an empty control vector and treated with 10 ng/ml LPS for different periods of time. TNF protein amounts were quantified in culture supernatants as described under Materials and Methods. Results are representative of two independent experiments.

Close modal

The events leading from MAPK11 phosphorylation to enhanced target gene transcription are poorly understood. Our results indicate that p38 signaling is regulated by the human histone deacetylase, HDAC3, at the level of the MAPK11 protein. Transient transfections of truncated HDAC3 mutants indicated the HDAC3 N terminus to be associated with MAPK11. HDAC3 inhibited the MAPK11-dependent transcriptional activity of ATF-2, a transcription factor binding to the promoter of human TNF gene. In agreement with this observation, HDAC3 inhibited TNF gene expression and TNF protein production by monocytic cells following LPS stimulation. Together our data indicate that HDAC3 participates to the control of proinflammatory cytokine gene expression, such as TNF, via repressing MAPK11/ATF-2 signaling.

Transcription activation by de-repression, as described here for ATF-2, is not uncommon in eukaryotic gene regulation. Similar principles are well described in the cases of NF-κB, nuclear receptors, and numerous other transcription factors. Repression as a ground state might help to avoid inadvertent or spurious activation of target genes, which, in the case of ATF-2, could cause inappropriate TNF production and activation of the immune system. Activation by de-repression has also been described for other transcription factors such as c-Jun (38). Interaction assays, as well as gain- and loss-function studies, identified HDAC3 as a repressor of MAPK11-mediated ATF-2 activation. The function of HDAC3 as a negative transcriptional cofactor has been described in the context of nuclear receptors such as the thyroid hormone or retinoic acid receptors (39). ATF-2 can be relieved of repression either by signal-dependent MAPK11-mediated phosphorylation, such as following LPS stimulation, or signal independently, by titrating out one or more limiting components of the repressor complex through increased ATF-2 protein levels. Both mechanisms are not mutually exclusive and might in principle operate in conjunction to increase transcription activation by ATF-2. Consistently, MAPK11 signaling might not only increase phosphorylation but also the protein levels of ATF-2 thus counteracting repression in two ways.

We observed that a HDAC3 mutant, deleted for its deacetylase domain, strongly activates ATF-2. These data suggest that ATF-2 activation of TNF transcription could be accompanied by local hyperacetylation of histones in agreement with HAT activity elicited by ATF-2 (18). Furthermore, treatment of cells with the histone deacetylase inhibitor, TSA, activates TNF gene expression. These observations suggest that the TNF promoter belongs to a small fraction of cellular promoters that are under the control of a repressive HDAC and can be activated by the inhibition of HDAC activity (40). Because the induction of several immediate early genes is controlled through phosphorylation of transcription factors such as ATF-2 (41), our data suggest that HDAC3 could control ATF-2-mediated transcription via both decreased ATF-2 phosphorylation and decreased acetylation of histone proteins (42).

A number of regulatory cascades such as JAK/STAT, NF-κB, and MAPK pathways are known to regulate inflammatory gene expression. This process goes along with the phosphorylation-dependent activation of transcription factors while chromatin may selectively be altered by posttranslational acetylation and phosphorylation of histone tails or by direct remodeling of nucleosomes involving ATP-dependent complexes (43), to specifically prepare responsive loci for the binding of determined regulatory factors. Very recently, it has been reported that such changes may be invoked by p38-dependent histone modification, where inflammatory stimuli selectively induce p38 MAPK-dependent phosphorylation and phosphoacetylation of the histone, H3, on the promoters of a subset of cytokine and chemokine genes where p38 activity is required to enhance the accessibility of cryptic NF-κB binding sites (41). In agreement with these data, we observed that HDAC3 represses both LPS-mediated ATF-2 and NF-κB activation in monocytic cells. HDAC3, by interfering with p38 activity, could both decrease ATF-2 transcriptional activity via dephosphorylation and decrease NF-κB activation by diminishing the accessibility of cryptic NF-κB binding sites present in the promoter of TNF gene. In fact, ATF-2 and NF-κB complexes bind sites between −106 and −88 bp of the human TNF promoter (19). Proximate cooperating ATF-2 and NF-κB sites are also present in the promoters of the E-selectin and IFN-β genes (44). After LPS exposure, glucocorticoids suppress TNF transcription in monocytic cells by preventing transactivation of ATF and NF-κB sites between −106 and −88 bp of the promoter and both ATF and NF-κB sites contribute independently but additively to glucocorticoid response (19). In agreement with these data, we observed that HDAC3 inhibits the binding of both ATF-2 and NF-κB complexes to the TNF promoter following LPS stimulation of monocytic cells, suggesting that HDAC3 might mimic the effect of glucocorticoids in regard to the control of proinflammatory cytokines in mononuclear phagocytes. Our results indicate that HDAC3 represses TNF expression primarily at a transcriptional level, although we cannot rule out a posttranscriptional control of TNF expression by HDAC3 (45). We observed that HDAC3 blocks early rather than late LPS-induced TNF secretion, suggesting a temporary and partial effect. In agreement with our data, deacetylation of the RelA subunit of NF-κB by HDAC3 has been shown to act as an intranuclear molecular switch that controls the duration of the NF-κB transcriptional response (15). We also observed that HDAC3 represses the IL-1β gene expression and the IL-1β protein production in LPS-treated THP-1 cells (data not shown), indicating that HDAC3 represses the secretion of proinflammatory cytokines, such as TNF and IL-1β, in activated monocytic cells (46).

Based on our observations, we speculate that HDAC3 plays a critical role in determining the threshold level at which a cell, especially mononuclear phagocytes, produces TNF in response to LPS stimulation. HDAC3 could allow some level of LPS stimulation without a concomitant increase in TNF expression, a situation that might allow for controlled immune activation. When the LPS stimulus is sufficiently strong, the suppressive effect of HDAC3 on TNF expression would be overwhelmed, resulting in TNF expression and enhanced immune response against the pathogen. Therefore, according to this model, HDAC 3 might serve as a rheostat modulating the TNF production in response to LPS stimulation (Fig. 8). Our observations that overexpression of HDAC 3 inhibits TNF production and that inhibition of HDAC 3 expression via siRNA enhances TNF production support this model. Future experiments in transgenic mice will directly test this hypothesis and further define the role of HDAC 3 in the regulation of the immune response.

FIGURE 8.

Model for the regulation of TNF gene expression by HDAC3. This scheme indicates the potential effect of the HDAC3/MAPK11 interaction on ATF-2 activity and TNF gene expression both under basal conditions and following LPS stimulation. In contrast to basal conditions, following LPS stimulation HDAC3 repression is suppressed, MAPK11 which is activated via the p38 signaling cascade, is phosphorylated and subsequently phosphorylates the transcription factor ATF-2. Phosphorylated ATF-2 binds to the human TNF promoter and via the formation of a p300/CREB-binding protein/ATF-2 activator complex stimulates TNF gene expression and subsequently TNF protein production. HDAC3 interaction with NF-κB pathway is not represented.

FIGURE 8.

Model for the regulation of TNF gene expression by HDAC3. This scheme indicates the potential effect of the HDAC3/MAPK11 interaction on ATF-2 activity and TNF gene expression both under basal conditions and following LPS stimulation. In contrast to basal conditions, following LPS stimulation HDAC3 repression is suppressed, MAPK11 which is activated via the p38 signaling cascade, is phosphorylated and subsequently phosphorylates the transcription factor ATF-2. Phosphorylated ATF-2 binds to the human TNF promoter and via the formation of a p300/CREB-binding protein/ATF-2 activator complex stimulates TNF gene expression and subsequently TNF protein production. HDAC3 interaction with NF-κB pathway is not represented.

Close modal

The MAPK11 cDNA has been kindly provided by Dr. J. Han (Department of Immunology, The Scripps Research Institute), and the RZPD. We thank A. Bourgeois for help with the chromatin immunoprecipitation assay.

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

1

This work was supported by the German National Science Foundation (Deutsche Forschungsgemeinschaft, MA 2057/2-3), the Heinrich and Fritz Riese Foundation, the August Scheidel Foundation, and the Franche-Comte University.

3

Abbreviations used in this paper: HAT, histone acetyltransferase; HDAC, histone deacetylase; ATF-2, activating transcription factor-2; MKK, MAPK kinase; CRE, cAMP responsive element; pTNF-Luc, luciferase reporter plasmid driven by the human TNF promoter; MΦ, macrophage; TSA, trichostatin A; siRNA, small interference RNA.

1
McKenna, N. J., B. W. O’Malley.
2002
. Combinatorial control of gene expression by nuclear receptors and coregulators.
Cell
108
:
465
.
2
Kouzarides, T..
2000
. Acetylation: a regulatory modification to rival phosphorylation?.
EMBO J.
19
:
1176
.
3
Ng, H. H., A. Bird.
2000
. Histone deacetylases: silencers for hire.
Trends Biochem. Sci.
25
:
121
.
4
Emiliani, S., W. Fischle, C. Van Lint, Y. Al-Abed, E. Verdin.
1998
. Characterization of a human RPD3 ortholog, HDAC3.
Proc. Natl. Acad. Sci. USA
95
:
2795
.
5
Grozinger, C. M., C. A. Hassig, S. L. Schreiber.
1999
. Three proteins define a class of human histone deacetylases related to yeast Hda1p.
Proc. Natl. Acad. Sci. USA
96
:
4868
.
6
Grozinger, C. M., E. D. Chao, H. E. Blackwell, D. Moazed, S. L. Schreiber.
2001
. Identification of a class of small molecule inhibitors of the sirtuin family of NAD-dependent deacetylases by phenotypic screening.
J. Biol. Chem.
276
:
38837
.
7
Fischle, W., F. Dequiedt, M. Fillion, M. J. Hendzel, W. Voelter, E. Verdin.
2001
. Human HDAC7 histone deacetylase activity is associated with HDAC3 in vivo.
J. Biol. Chem.
276
:
35826
.
8
Yang, W. M., S. C. Tsai, Y. D. Wen, G. Fejer, E. Seto.
2002
. Functional domains of histone deacetylase-3.
J. Biol. Chem.
277
:
9447
.
9
Johnson, C. A., D. A. White, J. S. Lavender, L. P. O’Neill, B. M. Turner.
2002
. Human class I histone deacetylase complexes show enhanced catalytic activity in the presence of ATP and co-immunoprecipitate with the ATP-dependent chaperone protein Hsp70.
J. Biol. Chem.
277
:
9590
.
10
Guenther, M. G., O. Barak, M. A. Lazar.
2001
. The SMRT and N-CoR corepressors are activating cofactors for histone deacetylase 3.
Mol. Cell. Biol.
21
:
6091
.
11
Li, J., J. Wang, Z. Nawaz, J. M. Liu, J. Qin, J. Wong.
2000
. Both corepressor proteins SMRT and N-CoR exist in large protein complexes containing HDAC3.
EMBO J.
19
:
4342
.
12
Ozawa, Y., M. Towatari, S. Tsuzuki, F. Hayakawa, T. Maeda, Y. Miyata, M. Tanimoto, H. Saito.
2001
. Histone deacetylase 3 associates with and represses the transcription factor GATA-2.
Blood
98
:
2116
.
13
Wen, Y. D., W. D. Cress, A. L. Roy, E. Seto.
2003
. Histone deacetylase 3 binds to and regulates the multifunctional transcription factor TFII-I.
J. Biol. Chem.
278
:
1841
.
14
Yao, Y. L., W. M. Yang, E. Seto.
2001
. Regulation of transcription factor YY1 by acetylation and deacetylation.
Mol. Cell. Biol.
21
:
5979
.
15
Chen, L., W. Fischle, E. Verdin, W. C. Greene.
2001
. Duration of nuclear NF-κB action regulated by reversible acetylation.
Science
293
:
1653
.
16
Zhang, J., M. Kalkum, B. T. Chait, R. G. Roeder.
2002
. The N-CoR-HDAC3 nuclear receptor corepressor complex inhibits the JNK pathway through the integral subunit GPS2.
Mol. Cell.
9
:
611
.
17
Dong, C., R. J. Davis, R. A. Flavell.
2002
. MAP kinases in the immune response.
Annu. Rev. Immunol.
20
:
55
.
18
Kawasaki, H., L. Schiltz, R. Chiu, K. Itakura, K. Taira, Y. Nakatani, K. K. Yokoyama.
2000
. ATF-2 has intrinsic histone acetyltransferase activity which is modulated by phosphorylation.
Nature
405
:
195
.
19
Steer, J. H., K. M. Kroeger, L. J. Abraham, D. A. Joyce.
2000
. Glucocorticoids suppress tumor necrosis factor-α expression by human monocytic THP-1 cells by suppressing transactivation through adjacent NF-κB and c-Jun-activating transcription factor-2 binding sites in the promoter.
J. Biol. Chem.
275
:
18432
.
20
van Dam, H., D. Wilhelm, I. Herr, A. Steffen, P. Herrlich, P. Angel.
1995
. ATF-2 is preferentially activated by stress-activated protein kinases to mediate c-jun induction in response to genotoxic agents.
EMBO J.
14
:
1798
.
21
Kawasaki, H., J. Song, R. Eckner, H. Ugai, R. Chiu, K. Taira, Y. Shi, N. Jones, K. K. Yokoyama.
1998
. p300 and ATF-2 are components of the DRF complex, which regulates retinoic acid- and E1A-mediated transcription of the c-jun gene in F9 cells.
Genes Dev.
12
:
233
.
22
Tracey, K. J., A. Cerami.
1994
. Tumor necrosis factor: a pleiotropic cytokine and therapeutic target.
Annu. Rev. Med.
45
:
491
.
23
Herbein, G., S. Gordon.
1997
. 55- and 75-kilodalton tumor necrosis factor receptors mediate distinct actions in regard to human immunodeficiency virus type 1 replication in primary human macrophages.
J. Virol.
71
:
4150
.
24
Yao, J., N. Mackman, T. S. Edgington, S. T. Fan.
1997
. Lipopolysaccharide induction of the tumor necrosis factor-α promoter in human monocytic cells: regulation by Egr-1, c-Jun, and NF-κB transcription factors.
J. Biol. Chem.
272
:
17795
.
25
Tsai, E. Y., J. Yie, D. Thanos, A. E. Goldfeld.
1996
. Cell-type-specific regulation of the human tumor necrosis factor-α gene in B cells and T cells by NFATp and ATF-2/Jun.
Mol. Cell. Biol.
16
:
5232
.
26
Udalova, I. A., J. C. Knight, V. Vidal, S. A. Nedospasov, D. Kwiatkowski.
1998
. Complex NF-κB interactions at the distal tumor necrosis factor promoter region in human monocytes.
J. Biol. Chem.
273
:
21178
.
27
Tang, X., M. J. Fenton, S. Amar.
2003
. Identification and functional characterization of a novel binding site on TNF-α promoter.
Proc. Natl. Acad. Sci. USA
100
:
4096
.
28
Sambrook, J., E. F. Fritsch, T. Maniatis.
1989
.
Molecular Cloning: A Laboratory Manual
2nd Ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor.
29
Mahlknecht, U., C. Deng, M. C. Lu, T. C. Greenough, J. L. Sullivan, W. A. O’Brien, G. Herbein.
2000
. Resistance to apoptosis in HIV-infected CD4+ T lymphocytes is mediated by macrophages: role for Nef and immune activation in viral persistence.
J. Immunol.
165
:
6437
.
30
Varin, A., S. K. Manna, V. Quivy, A. Z. Decrion, C. Van Lint, G. Herbein, B. B. Aggarwal.
2003
. Exogenous Nef protein activates NF-κB, AP-1, and c-Jun N-terminal kinase and stimulates HIV transcription in promonocytic cells: role in AIDS pathogenesis.
J. Biol. Chem.
278
:
2219
.
31
Manna, S. K., B. B. Aggarwal.
1999
. Immunosuppressive leflunomide metabolite (A77 1726) blocks TNF-dependent nuclear factor-κB activation and gene expression.
J. Immunol.
162
:
2095
.
32
Mahlknecht, U., O. G. Ottmann, D. Hoelzer.
2001
. Far-Western based protein-protein interaction screening of high-density protein filter arrays.
J. Biotechnol.
88
:
89
.
33
Goh, K. C., M. J. deVeer, B. R. Williams.
2000
. The protein kinase PKR is required for p38 MAPK activation and the innate immune response to bacterial endotoxin.
EMBO J.
19
:
4292
.
34
Ouwens, D. M., N. D. de Ruiter, G. C. van der Zon, A. P. Carter, J. Schouten, C. van der Burgt, K. Kooistra, J. L. Bos, J. A. Maassen, H. van Dam.
2002
. Growth factors can activate ATF-2 via a two-step mechanism: phosphorylation of Thr71 through the Ras-MEK-ERK pathway and of Thr69 through RalGDS-Src-p38.
EMBO J.
21
:
3782
.
35
Sui, G., C. Soohoo, B. Affar el, F. Gay, Y. Shi, W. C. Forrester.
2002
. A DNA vector-based RNAi technology to suppress gene expression in mammalian cells.
Proc. Natl. Acad. Sci. USA
99
:
5515
.
36
Han, J., Y. Jiang, Z. Li, V. V. Kravchenko, R. J. Ulevitch.
1997
. Activation of the transcription factor MEF2C by the MAP kinase p38 in inflammation.
Nature
386
:
296
.
37
Karin, M., A. Lin.
2002
. NF-κB at the crossroads of life and death.
Nat. Immunol.
3
:
221
.
38
Weiss, C., S. Schneider, E. F. Wagner, X. Zhang, E. Seto, D. Bohmann.
2003
. JNK phosphorylation relieves HDAC3-dependent suppression of the transcriptional activity of c-Jun.
EMBO J.
22
:
3686
.
39
Glass, C. K., M. G. Rosenfeld.
2000
. The coregulator exchange in transcriptional functions of nuclear receptors.
Genes Dev.
14
:
121
.
40
Van Lint, C., S. Emiliani, E. Verdin.
1996
. The expression of a small fraction of cellular genes is changed in response to histone hyperacetylation.
Gene Expr.
5
:
245
.
41
Saccani, S., S. Pantano, G. Natoli.
2002
. p38-dependent marking of inflammatory genes for increased NF-κB recruitment.
Nat. Immunol.
3
:
69
.
42
Saccani, S., G. Natoli.
2002
. Dynamic changes in histone H3 Lys 9 methylation occurring at tightly regulated inducible inflammatory genes.
Genes Dev.
16
:
2219
.
43
Strahl, B. D., C. D. Allis.
2000
. The language of covalent histone modifications.
Nature
403
:
41
.
44
Whitley, M. Z., D. Thanos, M. A. Read, T. Maniatis, T. Collins.
1994
. A striking similarity in the organization of the E-selectin and β interferon gene promoters.
Mol. Cell. Biol.
14
:
6464
.
45
Dumitru, C. D., J. D. Ceci, C. Tsatsanis, D. Kontoyiannis, K. Stamatakis, J. H. Lin, C. Patriotis, N. A. Jenkins, N. G. Copeland, G. Kollias, P. N. Tsichlis.
2000
. TNF-α induction by LPS is regulated posttranscriptionally via a Tpl2/ERK-dependent pathway.
Cell
103
:
1071
.
46
Hsu, H. Y., M. H. Wen.
2002
. Lipopolysaccharide-mediated reactive oxygen species and signal transduction in the regulation of interleukin-1 gene expression.
J. Biol. Chem.
277
:
22131
.