Deacetylase Activity Is Required for STAT5-Dependent GM-CSF Functional Activity in Macrophages and Differentiation to Dendritic Cells¹

Transcription is a highly regulated process that requires the orchestration and coordination of many elements involved in the production of transcription factors and in their posttranslational modifications, which alter their activity, as well as in chromatin remodeling. The binding of transcription factors to the promoter of a target gene and the recruitment of other regulators lead to localized changes in chromatin structure, which, depending on the gene, could have a positive or negative effect on transcription. In several genes, the recruitment of proteins with histone acetyl transferase (HAT)³ activity promotes acetylation of histone proteins (particularly H3 and H4), which leads to relaxed chromatin structure, the binding of core transcription machinery to DNA, and the initiation of transcription. In contrast, the recruitment of histone deacetylases (HDACs) causes deacetylation of histones, chromosomal condensation, and gene repression (1, 2). However, depending on the genes, deacetylation is also associated with the activation of transcription as demonstrated by the observation that histone hyperacetylation inhibits the expression of many genes (3). Moreover, some transcription factors must be deacetylated to induce transcription, such as C/EBPβ, which activates Id-1 expression only when it is deacetylated (4). In addition to histone acetylation, other histone and nonhistone protein modifications (acetylation, phosphorylation, and methylation) are required to regulate transcription (5, 6, 7).

Macrophages are part of innate immunity and play a critical role in host defense mechanisms. These cells originate from undifferentiated stem cells and require specific growth factors (M-CSF, IL-3, and GM-CSF) for their generation (8). In the presence of growth factors and cytokines, macrophages may proliferate, differentiate to specific cell types depending on the tissue, or become activated and develop their functional activities. GM-CSF has a profound influence on macrophage biology. It promotes the differentiation, proliferation, and survival of these cells in addition to many other functions, including Ag presentation, chemotaxis, phagocytosis, and release of reactive oxygen intermediates (9, 10, 11). GM-CSF is of particular relevance for dendritic cell (DC) development and maturation (12). In addition, this growth factor regulates several biological functions of other immune cells such as neutrophils, eosinophils, basophils, and lymphocytes (13).

GM-CSF exerts its functions after interaction with the corresponding receptor, which phosphorylates at least two distinct domains in the β-chain of the receptor (14). One of these domains induces the activation of mitogen activated protein kinases and the PI-3K/Akt/p21^waf-1 pathway. The other mediates the activation of the JAK2-STAT5 signaling pathway. STAT proteins are present in a latent form in the cytoplasm and migrate to the nucleus following cytokine-induced phosphorylation and dimerization. Once in the nucleus, STAT dimers bind to specific DNA-binding sites and activate the transcription of target genes (15). In addition, STAT activity is modulated through interaction with a variety of proteins, including cofactors and other transcription factors (16). Moreover, STAT family members interact with proteins with HAT activity (17, 18, 19) as well as with members of the HDAC family (4, 20). These observations indicate that STAT proteins are involved in chromatin remodeling. In particular, in T and B lymphocytes IL-2- or IL-3-induced transcriptional activity through STAT5 requires the activity of a HDAC (4, 21). On the basis of this finding, it has been proposed that administration of the deacetylase activity inhibitor trichostatin A (TSA) would be beneficial in STAT5-associated cancers (22). However, the effects of TSA on macrophages treated with GM-CSF are unknown. To address this question, we treated bone marrow-derived macrophages with TSA. GM-CSF-dependent proliferation and MHC class II gene expression were reduced under this treatment, as was the differentiation of DCs from bone marrow cells. The expression of genes involved in proliferation and differentiation was also impaired in macrophages and DCs treated with TSA. Taken together, these results indicate that deacetylase activity is essential for the proper function of macrophages and for the generation of DCs.

Recombinant murine cytokines and the JAK2 inhibitor AG490 were purchased from Sigma-Aldrich and TSA was from Tocris Bioscience. The Abs used were anti-STAT5a, anti-STAT5b (R&D Systems), anti-phospho-STAT5a/b Y694/Y699 (Upstate Biotechnology), anti-phospho-Akt (Cell Signaling), anti-β-actin (Sigma-Aldrich), anti-RNA Pol II (N20) (Santa Cruz Biotechnology), anti-MHC class II (I-A) FITC conjugate (Chemicon), anti-Cd11c PE conjugate, and anti-CD16/CD32 (BD Pharmingen). Peroxidase-conjugated anti-rabbit (Jackson ImmunoResearch Laboratories) or anti-mouse (Sigma-Aldrich) was used as a secondary Ab. All other chemicals were of the highest purity grade available and were purchased from Sigma-Aldrich. Deionized water further purified with Millipore Milli-Q System A10 was used.

Bone marrow-derived macrophages were isolated from 6-wk-old BALB/c mice (Harlan Ibérica) as described (23). To differentiate to DCs, bone marrow cells were cultured in DMEM, 10% FCS, and 5 ng/ml GM-CSF (12). At days 2 and 4, the plates were shaken and the culture supernatant was collected and replaced by fresh medium with GM-CSF. At days 6 and 8, plates were fed aspirating supernatants (without shaking) and medium with GM-CSF was added. At day 8, cells were stimulated with 1 μg/ml LPS from Escherichia coli (Sigma-Aldrich) for 48 h. At day 10 the plates were shaken, the supernatant was collected, and DCs were separated from adherent macrophages.

RNA was extracted with the EZ-RNA Kit (Biological Industries) and treated with DNase (Roche). One microgram was retrotranscribed using Moloney murine leukemia virus reverse transcriptase RNase H Minus (Promega) and real-time PCR was performed as described (24). Data were expressed as relative to the β-actin expression in each sample. The primers used are available on request. Primers and conditions for HDACs were as described (25).

Macrophage proliferation was measured by [³H]thymidine incorporation as described (26). Cell death was assessed by FACS analysis using the rAnnexin-V-FITC kit (Bender MedSystems).

This assay was conducted using specific Abs and cytofluorimetric analysis as described (27). Stained cell suspensions were analyzed using an Epics XL flow cytometer (Beckman Coulter).

Cells were lysed as described (28). Immunoprecipitation assays were performed as described (28) using 150 μg of cell lysates and 2 μl of anti-STAT5a or anti-STAT5b Abs.

Nuclear extracts were prepared from macrophages (29) and EMSAs were performed (30). For supershift experiments, 2 μl of anti-STAT5a and anti-STAT5b Abs were added and incubated for 30 min. The probe corresponds to a γ-IFN activated sequence (GAS)-like element (underlined) from the promoter of β-casein (5′- AGATTTCTAGGAATTCAAATC-3′).

Cells (20 × 10⁶) were cross-linked with 1% paraformaldehyde for 20 min at room temperature. After two washes with ice-cold PBS, cells were collected in 3 ml of 0.1M Tris-HCl (pH 9.4) and 10 mM DTT and incubated for 15 min at 30°C. Cells were centrifuged for 5 min at 2000 × g at 4°C and the pellets were washed sequentially by pipetting with ice-cold PBS, buffer I (0.25% Triton X-100, 10 mM EDTA, 0.5 mM EGTA, and 10 mM HEPES (pH 6.5)) and buffer II (200 mM NaCl, 1 mM EDTA, 0.5 mM EGTA, and 10 mM HEPES (pH 6.5)). Next, 300 μl of lysis buffer (1% SDS, 10 mM EDTA, 50 mM Tris-HCl (pH 8.1), 1 mM DTT, 1 mM PMSF, 0.01 mg/ml aprotinin, 0.01 mg/ml leupeptin, 0.086 mg/ml iodoacetamide, and 1 mM sodium orthovanadate) was added and lysates were sonicated on ice using the Ikasonic U200S Control (Ika Labortechnik) (15 pulses of 10 s, 30% cycle and 30% amplitude). Size of fragments obtained (between 200 and 1200 bp) was confirmed by electrophoresis. Soluble chromatin was collected after centrifugation at 14,000 rpm at 4°C for 10 min and diluted to 1/10 in dilution buffer (1% Triton X-100, 2 mM EDTA, 150 mM NaCl, 20 mM Tris-HCl (pH 8.1), 1 mM DTT, 1 mM PMSF, 0.01 mg/ml aprotinin, 0.01 mg/ml leupeptin, 0.086 mg/ml iodoacetamide, and 1 mM sodium orthovanadate). Soluble chromatin (1–5%) was kept as input control.

Soluble chromatin was precleared with 20 μg of salmon sperm (Amersham Biosciences), 8 μg of unspecific IgGs, and 20 μg of preimmune serum and protein-A-Sepharose at 50% overnight at 4°C in rotation. After centrifugation, supernatants were collected and specific Abs (2 μl of STAT5a and STAT5b; 1 μg of RNA Polymerase II) were added. A control was performed with unspecific IgGs. Mixtures were incubated at 4°C for 6 h in rotation and then incubated overnight at 4°C in rotation with protein-A-Sepharose at 50%. Beads were collected and washed sequentially at 4°C for 10 min with TSE I (150 mM NaCl, 0.1% SDS, 1% Triton X-100, 2 mM EDTA, and 20 mM Tris-HCl (pH 8.1)), TSE II (500 mM NaCl, 0.1% SDS, 1% Triton X-100, 2 mM EDTA, and 20 mM Tris-HCl (pH 8.1)), and buffer III (0.25 LiCl, 1% Nonidet P-40, 1% deoxycholate, 1 mM EDTA, and 10 mM Tris-HCl (pH 8.1)). Beads were washed once with TE buffer by pipetting and immunoprecipitates were eluted three times (20-min incubation) with elution buffer (0.1 M NaHCO₃ and 1% SDS). Reversion of cross-linking was performed overnight by heating samples and input controls at 65°C, and DNA was purified using the QIAquick spin kit (Qiagen). Real-time PCR was performed as described above. The primers used correspond to STAT5 binding sites at Cis promoter and the CAP site (21). Amplification of c-jun coding region was used as a control of nonspecific immunoprecipitation. Data were expressed as relative immunoprecipitation normalized to the c-jun amplification level in each sample.

Cells (6 × 10⁶) were electroporated with 1.5 μM short interfering (si) RNA or no siRNA (mock control) in a final volume of 400 μl using the electroporation system BTX-600 (350 volts, 2300 microfarads, and R1 resistance). Cells were then kept on ice for 5–10 min and plated in complete medium (DMEM, 20% FCS, and 30% L-Cell conditioned medium) for 24 h. siRNA were purchased from MWG Biotech (siMAX siRNA).

To calculate the statistical differences, we used Student’s paired t test.

GM-CSF is a hematopoietic growth factor and immune modulator (13) and plays a critical role in regulating the generation and function of many immune cells. Regarding monocytes and macrophages, GM-CSF promotes the differentiation, proliferation, survival, and activation of these cells (9, 13, 31, 32). To test whether deacetylase activity has any functional implication in GM-CSF-dependent macrophage biology, we first examined the proliferation of macrophages in the presence of the pharmacological HDAC inhibitor TSA. Cells were grown and differentiated in the presence of M-CSF for 7 days. After 18 h of M-CSF-deprivation, macrophages became quiescent. When GM-CSF was added to the culture for 24 h, thymidine incorporation increased in a dose-dependent manner (Fig. 1,A). Treatment with TSA resulted in a dramatic inhibition of GM-CSF-induced proliferation (Fig. 1,A). This experiment was repeated five times and differences between control and TSA-treated cells were significant (p < 0.01). Moreover, cell counting confirmed these results. To test whether this effect was specific for GM-CSF, we analyzed M-CSF-dependent proliferation in the presence of TSA. No significant differences were found (Fig. 1,B). This observation also demonstrates that TSA, at the concentration used in our assays, did not induce cellular toxicity. Stimulation of macrophage proliferation by both M-CSF and GM-CSF requires the activation of extracellular signal-regulated kinases or ERKs (33, 34). The observation that TSA did not affect M-CSF-induced proliferation indicates that this inhibitor/drug does not target ERKs. In some cell types, such as macrophages, STAT5 mediates the GM-CSF-dependent proliferation; we therefore hypothesized that a target of the inhibitory effect of TSA is STAT5 activation. To test this hypothesis, we measured the expression of STAT5 target genes involved in macrophage proliferation by real-time PCR. Treatment with 5 ng/ml GM-CSF increased the expression of Cyclin D1 and c-myc, which are required for cell cycle entry. The induction of these genes was strongly inhibited by the addition of 20 nM TSA (Fig. 1 C), indicating that this agent blocks GM-CSF-dependent proliferation by inhibiting STAT5-dependent gene expression.

In addition to stimulating cell proliferation, M-CSF and GM-CSF also promote cell survival (9). In the absence of a growth factor, macrophages undergo cell death by apoptosis, whereas treatment of macrophages with M-CSF or GM-CSF rescues these cells from this fate (35). We determined cell death using annexin V staining. The addition of M-CSF after 18 h in growth factor-free medium protected macrophages from apoptosis induced by deprivation of growth factor, and this protection was not inhibited by TSA (Fig. 2,A). Interestingly, GM-CSF also protected the cells from apoptosis even in the presence of TSA (Fig. 2,A). These results indicate that the impaired proliferation observed in the presence of TSA is not due to increased cell death and that TSA selectively affects some of the pathways induced by the binding of GM-CSF with its receptor. After interaction of the ligand (GM-CSF) with the corresponding receptor, several intracytoplasmatic domains become phosphorylated and distinct signal transduction pathways are induced, which correspond to a range of functional activities (36). M-CSF- and GM-CSF-dependent survival of macrophages is mediated by activation of the (PI3K)/Akt/p21^waf-1 pathway. This activation is independent of the signal transduction required for proliferation but is involved in protection from apoptosis (9, 35). As expected, Akt phosphorylation (Fig. 2,B) and p21^waf-1 expression (Fig. 2,C) were not affected by TSA treatment. To further study the STAT5-dependent and -independent activities of GM-CSF-treated macrophages, we attempted to silence the expression of STAT5a and STAT5b using specific siRNAs. We used three distinct sequences but failed to silence the expression of these proteins. Our failure can be explained by the high expression and stability of STAT5 proteins in macrophages. Therefore, we used the specific JAK2 inhibitor tyrphostin AG490 (37). In preliminary experiments we determined that at 25 μM AG490 blocks STAT5A and B phosphorylation without affecting cell viability. Treatment of macrophages with this inhibitor blocked the GM-CSF-dependent STAT5 phosphorylation without affecting the ERK and PI3K pathways (Fig. 3,A). In addition, GM-CSF-dependent proliferation (Fig. 3,B), but not protection from apoptosis (Fig. 3 C), was strongly inhibited in the presence of AG490. These results demonstrate that STAT5 is required for GM-CSF-dependent proliferation of macrophages but not for protection from apoptosis. Thus, in macrophages TSA specifically inhibited the GM-CSF/STAT5 pathway but not ERK- or PI3K-dependent functions.

Macrophages are crucial not only for innate immunity but also as APCs for host defense. Expression of MHC class II molecules at the cell surface is required for this function and therefore for the activation of T cells. The major activator of MHC class II molecules expression in macrophages is IFN-γ (38), a cytokine secreted by Ag-activated Th1 lymphocytes. This cytokine regulates the expression of these molecules not only at a transcriptional but also at a posttranscriptional level, including the regulation of translation and the half-life of the protein (39). In addition to IFN-γ, GM-CSF has a positive effect on Ag presentation capacity (10) and induces the expression of MHC class II molecules (32). These observations led us to evaluate whether GM-CSF induces MHC class II molecules expression in macrophages and whether this induction is impaired in the presence of TSA. GM-CSF induced I-A expression at the cell surface of macrophages as well as at the mRNA level. Pretreatment of these cells with TSA blocked the induction at the cell surface (Fig. 4, A and B) and inhibited the expression of I-A β mRNA (Fig. 4,C). The half-life of I-A β mRNA in macrophages is stable (39), which suggests that TSA inhibits the expression of I-A genes at the transcriptional level. To further investigate this point, we analyzed the expression of the CIITA. The expression of this master regulator is related to both tissue specificity and the quantitative expression of MHC class II genes, and its absence causes severe immunodeficiency (40). In bone marrow-derived macrophages, GM-CSF increased the mRNA levels of CIITA (Fig. 4 D). Strikingly, TSA abolished CIITA expression in GM-CSF-stimulated macrophages, confirming that reduced I-A β expression in TSA-treated macrophages is due to impaired MHC class II transcription. There is an apparent discrepancy between the levels of CIITA mRNA, which was completely inhibited by TSA, and those of I-A β, which was only partially inhibited. This discrepancy can be attributed to the distinct mRNA half-lives; for CIITA mRNA the half-life is short (between 30 and 60 min) (41) whereas I-A β mRNA is stable (>10 h) (42, 43), as is the I-A β protein (39).

To determine the mechanism by which deacetylase activity regulates STAT5-dependent gene expression, we studied the signal transduction pathway induced by GM-CSF. Once GM-CSF binds to its receptor, STAT5a and STAT5b become phosphorylated by JAK2 and migrate to the nucleus. TSA treatment of macrophages did not produce any change in STAT5a and STAT5b phosphorylation (Fig. 5,A). To determine the functional activity of the STAT5 proteins, we measured their binding activity to DNA in TSA-treated cells. Using a fragment of the β-casein promoter that contains a GAS box, we found that both STAT5a and STAT5b were present in the complexes, as indicated by the supershift observed when specific Abs were added (Fig. 5,B). The same complexes were present when we used nuclear extracts from TSA-treated cells. Moreover, we analyzed in vivo DNA binding activity of STAT5 by ChIP assays. This activity was not affected by TSA treatment (Fig. 5 C). The specificity of the reaction was checked by using unrelated Abs or a fragment of a promoter that does not contain the GAS box (data not shown). Therefore, these results indicate that TSA does not affect the GM-CSF-JAK2-STAT5 pathway in macrophages.

To identify the HDAC involved in GM-CSF-dependent gene expression, we first analyzed the expression of HDAC members in macrophages over a time course of treatment with GM-CSF. The expression of Hdac-4, -7, and -9 was repressed. In contrast, Hdac-1, -5, and -6 were down-regulated at 1 h, returned to basal levels at 2 h and were then down-regulated again. However, the expression of Hdac-8 and -10 was induced by GM-CSF at 1 h, and thereafter their mRNA levels were reduced. Because Hdac-8 and 10 are the only two HDAC members induced by GM-CSF and their expression peaked at 1 h, correlating with the maximal expression of c-myc and Cyclin D1, we examined whether these HDACs are involved in GM-CSF-dependent gene expression. As a control we included Hdac-2 and -3, which are not expressed by macrophages. We used the siRNA technique to silence the expression of these genes. siRNAs specific for each gene inhibited the expression of the corresponding genes compared with unspecific siRNA (siGL3) and electroporated cells (mock control) (Fig. 6,A). However, expression of c-myc and Cyclin D1 were not altered in macrophages treated with siHDAC8 and siHDAC10 (Fig. 6 B) or with a combination of both (data not shown). The siRNA of all these genes did not affect GM-CSF-dependent proliferation (data not shown). These data indicate that these HDACs are not responsible for the deacetylase activity required for GM-CSF-dependent gene expression.

Although STAT1-, STAT2-, STAT3-, and STAT5-HDAC1 interactions have been described in several cell types (4, 20, 44), we did not detect an interaction between STAT5 and HDAC1 in coimmunoprecipitation assays (data not shown). Furthermore, HDAC1 was not present in the promoter region of Cis, a member of the SOCS (suppressor of cytokine signaling) family of phosphatases that inhibits cytokine signal transduction and whose expression depends on STAT5, as revealed by ChIP experiments (data not shown), thereby confirming the results obtained in T and B lymphocytes treated with IL-2 and IL-3 (21).

Because HATs and HDACs play a crucial role in the formation of transcription preinitiation complexes, we next tested the effect of TSA on the recruitment of RNA polymerase II upon GM-CSF stimulation of macrophages. Using ChIP, we found that RNA polymerase II was recruited to the promoter of a STAT5 target gene like Cis after 15 min of GM-CSF treatment. However, this recruitment did not occur when the cells were pretreated with TSA (Fig. 6 C). This observation indicates that TSA inhibits STAT5-mediated transcription by blocking the recruitment of the basal machinery complex. Taken together, these results suggest that deacetylase activity is required to recruit RNA polymerase II and activate the transcription of the STAT5-dependent genes involved in the proliferation and Ag presentation capacity of macrophages.

DCs are APCs crucial for both innate and adaptive immunity and play a critical role in the induction and control of T cell immunity and in the modulation of the responses of B and NK cells (45). In vitro, DCs originate from MHC class II-negative precursors from bone-marrow cultures in the presence of GM-CSF (45). To study the role of deacetylase activity in DC differentiation, we examined the effect of TSA on bone marrow cells GM-CSF-dependent differentiation to DCs. After 8 days of culture in the presence of 5 ng/ml GM-CSF, we obtained a population of cells containing adherent macrophages and floating DCs. The addition of 1 μg/ml LPS to the culture promoted the maturation of these cells and expression of MHC class II (I-A) molecules on the cell surface. To quantify DC differentiation, FACS analysis was performed using CD11c and I-A as specific markers of these cells in mouse. After 10 days of culture, 70% of the cells obtained were CD11c+ and I-A⁺ (Fig. 7,A). This percentage was dramatically reduced to 25% when TSA was present during differentiation (Fig. 7, A and B). A significant difference was found when we compared four independent experiments (p < 0.01). To exclude an effect of TSA on LPS, we tested the effect of the deacetylase inhibitor by incubating the cells in the presence of GM-CSF without the addition of LPS. Under these conditions we obtained similar results; the presence of TSA decreased the amount of differentiated DCs to 28 ± 6%. This finding suggests that TSA has no effect on LPS stimulation of DCs. The effect of TSA on DC differentiation was specific of growth factor, because M-CSF-dependent differentiation of macrophages was not affected by TSA (Fig. 7 C). These results demonstrate that TSA inhibits bone marrow differentiation to DCs.

To study the mechanism by which TSA inhibits DC differentiation, we examined the expression of several of the genes involved in this process. Transcription factors of the IFN regulatory factor (IRF) family play crucial roles in differentiation as well as in the functional activity of DCs. GM-CSF-dependent development of DCs depends on IRF4 (46), as well as IRF2 (47, 48). In addition, RelB, a member of the NF-κB family, is required for the development of some subsets of DCs (49, 50). We found that IRF4, IRF2, and RelB were expressed in mature DCs. However, TSA inhibited the expression of these genes when differentiation was achieved in the presence of LPS (Fig. 8) or in its absence (data not shown). This effect was specific because IRF1 expression was very low in DCs and was not inhibited by TSA, indicating that IRF1 is not involved in DC development. These results indicate that TSA reduces GM-CSF-dependent DC development by inhibiting the expression of factors involved in differentiation, such as IRF4, IRF2, and RelB.

Chromatin remodeling is an essential mechanism that regulates gene transcription. HATs and HDACs play pivotal roles not only in modifying histone tails but also in modulating the activity of several transcription factors. In this study we provide evidence that deacetylase activity is required for STAT5-dependent GM-CSF induction of the proper function of macrophages and the generation of DCs. We have shown that TSA impairs the proliferation of macrophages under GM-CSF stimulation. The requirement of deacetylase activity is related to GM-CSF and is not a general mechanism of proliferation, because TSA does not inhibit the effect of M-CSF.

To determine the mechanism of deacetylase activity on GM-CSF-dependent functional activities of macrophages, we considered the different signaling transduction pathways activated by this growth factor after interacting with the receptor. In the intracytoplasmic region of the β-chain receptor, at least two domains that regulate distinct pathways have been described (36). The activation of the PI3K/Akt/p21^waf-1 and the Ras-ERK pathways critically depends on the integrity of tyrosines Y577 and Y612 (14). Our experiments show that genes regulated through these pathways are not affected by the inhibition of deacetylase activity. The activation of ERK1/2 is required for GM-CSF- and M-CSF-dependent macrophage proliferation (33, 34). However, TSA did not affect M-CSF-dependent proliferation, which implies that genes that require the ERK pathway are not inhibited. Also, the genes induced through the PI-3K/Akt/p21^waf-1 pathway were not altered, as shown by the cell survival experiments. Although it has been shown that the PI3K-Akt pathway interacts both functionally and physically with STAT5 in hematopoietic cells (51, 52, 53, 54), we have demonstrated that inhibition of JAK2 affects only STAT5 activation in macrophages. The observation that this inhibition impaired the GM-CSF-dependent proliferation of these cells but had no effect on their protection from apoptosis indicates that these two biological functions of GM-CSF depend on distinct pathways in macrophages. This confirms our previous results demonstrating that GM-CSF-dependent protection of macrophages from apoptosis is dependent on the PI3K-Akt pathway, while ERK is required for proliferation (9).

Upon dimerization, two membrane-proximal domains, including a proline-rich domain, become phosphorylated by Jak2, which is constitutively associated in a conserved region designated the Box 1 domain (55). The expression of the genes dependent on this pathway, which involves STAT5, is down-regulated by TSA. GM-CSF, through the JAK2-STAT5 transduction pathway, regulates the expression of Cyclin D1 and c-myc, two genes required for macrophage proliferation (56, 57). As expected, GM-CSF rapidly induced an increase in Cyclin D1 and c-myc mRNA, but TSA blocked this induction. These results demonstrate that TSA impairs the GM-CSF-dependent proliferation of macrophages by inhibiting the expression of STAT5 target genes.

In addition to its role in proliferation and survival, GM-CSF induces the expression of MHC class II molecules in monocytes (32). We have shown that GM-CSF induces I-A expression in macrophages and that this induction is due to an increase in CIITA mRNA levels. TSA abolished the expression of these genes, revealing that deacetylase activity is crucial for CIITA expression and the Ag-presentation capacity of macrophages. CIITA is regulated by several promoters that are tissue- and cytokine-dependent. IFN-γ, through STAT1, affects promoters III and IV, whereas GM-CSF uses I and III (32). It should be noted that promoter III is induced with different kinetics by IFN-γ or GM-CSF. All of these observations indicate that distinct mechanisms are used by these cytokines to regulate CIITA. In both cases, the way by which TSA inhibits the expression of CIITA could be the same, because deacetylase activity is required for transcriptional activity of STAT1 and STAT5 (20, 21, 58).

In B cells, stimulation with IL-3 results in the induction of Id-1, which encodes a dominant negative inhibitor of basic helix-loop-helix transcription factors. The expression of this gene requires the binding of both C/EBPβ and STAT5 to its promoter. STAT5 recruits HDAC1, which results in deacetylation of histones, and also C/EBPβ, whose acetylation diminishes its DNA-binding activity (4). However, in the same cellular model the induction of Cis requires HDAC, but no interaction or direct effect on STAT5 was demonstrated (21). Moreover, the hyperacetylation that occurs concomitantly with STAT5 binding was absent from TSA-treated cells. In OSM (oncostatin M), the acetylation peak observed at the CAP site was increased upon TSA treatment. OSM, a member of the IL-6 family of cytokines, is also a STAT5 target gene closely related structurally, genetically, and functionally to leukemia inhibitory factor. However, OSM-specific biological activities have been reported in hematopoiesis and liver development. This finding indicates that in some STAT5-mediated transcription the inhibitory effect of TSA can occur independently of histone acetylation or chromating remodeling (21). In macrophages treated with GM-CSF, Cis induction was abolished with TSA treatment; but, as described in B cells, we did not observe modifications of STAT5 or interaction with HDAC. In addition, inhibition of Hdac1 expression by siRNA did not affect GM-CSF-dependent proliferation and gene expression in macrophages (data not shown). Although the effect of TSA seems to be independent of histone acetylation, this drug inhibits transcriptional initiation of several STAT5 target genes by preventing the recruitment of the basal transcription machinery to the promoter. Because inhibition of transcription occurs in a few minutes, TSA may act by modifying some of the transcriptional elements at posttranscriptional levels rather than by inhibiting the synthesis of a protein required for transcription.

Another important finding of this study is that TSA inhibits the generation of DCs from bone marrow progenitors, which is reflected by the decreased expression of genes involved in DC development such as IRF2, IRF4, and RelB. The mechanism by which TSA inhibits the expression of these genes remains to be elucidated. It has been proposed that some STATs, as well as members of the NF-κB family, bind to the IRF4 promoter in human monocyte-derived DCs (59). However, although we found that STAT5 is present in the nuclear extracts of DCs and can bind a β-casein probe, we did not detect STAT5 binding to the promoter of IRF4 (data not shown). This observation suggests that STAT5 induces other transcription factors involved in GM-CSF-dependent expression of IRF4. Among these, members of the NF-κB family are reported to regulate the expression not only of IRF4 but also RelB (60). Strikingly, NF-κB activity is inhibited by deacetylase inhibitors via a mechanism involving the suppression of proteasome activity (61, 62). Further studies are required to determine whether this mechanism is responsible for TSA suppression of these genes in bone marrow-derived DCs.

In recent years, several studies have demonstrated the importance of deacetylase activity in the regulation of immune response. Among these, it has been reported that IFN-stimulated transcription and innate antiviral responses are regulated by deacetylase activity (20, 58) and that HDACs act as negative regulators of proinflammatory response in macrophages (25). In addition, inhibition of deacetylase activity alters TLR4-dependent activation and function of macrophages and DCs (63). In this report we show for the first time that HDAC activity is essential for the generation of DCs and for STAT5-dependent functions induced by GM-CSF in macrophages, which are critical for the development of innate and adaptive immunity. Because TSA and other HDAC inhibitors have been proposed for use in tumor treatment (22), their capacity to affect the immune response is of great relevance.

We thank Tanya Yates for editing the manuscript.

The authors have no financial conflict of interest.