The coactivator role of histone deacetylase 3 in IL-1-signaling involves deacetylation of p65 NF-κB

Abstract

Histone deacetylase (HDAC) 3, as a cofactor in co-repressor complexes containing silencing mediator for retinoid or thyroid-hormone receptors (SMRT) and nuclear receptor co-repressor (N-CoR), has been shown to repress gene transcription in a variety of contexts. Here, we reveal a novel role for HDAC3 as a positive regulator of IL-1-induced gene expression. Various experimental approaches involving RNAi-mediated knockdown, conditional gene deletion or small molecule inhibitors indicate a positive role of HDAC3 for transcription of the majority of IL-1-induced human or murine genes. This effect was independent from the gene regulatory effects mediated by the broad-spectrum HDAC inhibitor trichostatin A (TSA) and thus suggests IL-1-specific functions for HDAC3. The stimulatory function of HDAC3 for inflammatory gene expression involves a mechanism that uses binding to NF-κB p65 and its deacetylation at various lysines. NF-κB p65-deficient cells stably reconstituted to express acetylation mimicking forms of p65 (p65 K/Q) had largely lost their potential to stimulate IL-1-triggered gene expression, implying that the co-activating property of HDAC3 involves the removal of inhibitory NF-κB p65 acetylations at K122, 123, 314 and 315. These data describe a novel function for HDAC3 as a co-activator in inflammatory signaling pathways and help to explain the anti-inflammatory effects frequently observed for HDAC inhibitors in (pre)clinical use.

Figure 1.

Interference with HDAC3 inhibits IL-8 transcription. (A) HEK293IL-1R cells were stably transfected with an empty shRNA vector (pS-Puro, black bars) or with a plasmid encoding a HDAC3-specific shRNA (pS-HDAC3, white bars) to generate shRNA-mediated HDAC3 knockdown cells. HDAC3 knockdown and control cells were treated for 3 h with IL-1 (10 ng/ml) or/and for 24 h with TSA (300 ng/ml) as indicated. IL-8 mRNA expression was determined by RT-qPCR. Data represent the mean IL-8 expression ± standard errors of the means (SEM) relative to the unstimulated control (n = 2). Black bars: pS-Puro, white bars: pS-HDAC3. The right part shows a control of the knockdown efficiency. Cell lysates were tested by immunoblotting (IB) for the occurrence of HDAC3 and β-actin as a loading control. (B) HDAC3 knockdown (pS-HDAC3, white bars) and control cells (pS-Puro, black bars) were stimulated for the indicated periods with IL-1 and analyzed for IL-8 mRNA expression as described in (A). Data represent the mean IL-8 expression ± SEM relative to the unstimulated control (n = 2). (C) HEK293IL-1R cells were pretreated for 24 h with increasing concentrations of the HDAC3 inhibitor apicidin and stimulated for 3 h with IL-1 as shown. RT-qPCR was used to determine IL-8 mRNA expression, bars show the mean IL-8 expression ± SEM relative to the unstimulated control (n = 2). (D) HDAC3 knockdown (pS-HDAC3, white bars) and control cells (pS-Puro, black bars) were transiently transfected with an IL-8 promoter controlled luciferase reporter gene and SV40-β-gal for normalization. Twenty-four hours later, cells were stimulated for 4 h with IL-1 or were left untreated as indicated. Thereafter, cells were lysed, and luciferase and β-galactosidase activities were determined in cell extracts. Shown are the normalized mean luciferase activities ± SEM from four independent experiments. (E) HEK293IL-1R cells were transiently transfected with an IL-8 promoter controlled luciferase reporter gene as in (D). Twenty-four hours later, cells were treated for 24 h with 1 µg/ml apicidin and thereafter stimulated for 3 h with IL-1 or were left untreated as indicated. Thereafter, cell extracts were prepared, and luciferase and β-galactosidase activities were determined. Shown are the normalized mean luciferase activities ± SEM from two independent experiments.

Figure 2.

Deletion of the HDAC3 genes inhibits IL-1-triggered Cxcl2 transcription. (A) Immortalized murine embryonic fibroblast (Mef) lines were generated from mice harboring floxed Hdac3 alleles and the tamoxifen-inducible Cre recombinase gene (33). These cells contain a heterozygous floxed Hdac3 allele in the wild-type background (Hdac3 ^fl/+) or in the knockout background (Hdac3 ^fl/−) and become thus heterozygous for Hdac3 or lack both Hdac3 alleles after Cre-mediated deletion of the floxed gene. These cells were treated for 72 h with increasing concentrations of tamoxifen as shown and analyzed for HDAC3 and β-actin expression by immunoblotting. (B) The cells with the indicated genotypes were treated for 3 days with 10 µM of tamoxifen and then stimulated for 0.5 h with IL-1 as shown. RT-qPCR was used to quantify Cxcl2 mRNA expression. Shown are the mean -fold changes compared with the untreated Hdac3 ^fl/+ control line, error bars represent SEM from three different experiments performed in duplicates.

Figure 3.

Overexpression of HDAC3 enhances mRNA expression of inflammatory genes. (A) HEK293IL-1R cells were transiently transfected to express wild-type HA-tagged p65 along with p300 and HA-tagged HDAC3. Twenty-four hours later, cells were lysed, and mRNA expression of IL-8 and CXCL2 was determined by RT-qPCR. Shown is the mean -fold increase ± SEM relative to the vector-transfected control from two independent experiments. (B) In parallel, expression of HA-tagged p65 and HDAC3 was analyzed by immunoblotting of lysates. One representative out of two experiments is shown. The position of a molecular weight marker is indicated at the left.

Figure 4.

HDAC3 binds to and deacetylates p65 at lysines K310, K314 and K315. (A) HEK293IL-1R cells were transiently transfected to express wild-type His-tagged p65 along with YFP-tagged CBP and HA-tagged HDAC3. Cells were lysed under denaturing conditions, proteins were precipitated by TCA, redissolved and equal amounts were analyzed for general acetylation of p65 with a pan-acetyl-lysine-specific antibody or with antibodies specific for acetylated p65 K310, K314 and K315 as shown. Expression of HDAC3 and YFP-CBP was validated as indicated and equal loading of lanes was confirmed by anti-β-actin antibodies. The position of a molecular weight marker is indicated at the Left (B) HEK293IL-1R cells were transfected to express the indicated combinations of HA-tagged HDAC3 and HA-tagged p65 variants. Following lysis of cells, extracts were used for immunoprecipitation (IP) of the HDAC3 protein, and co-precipitating proteins were revealed with antibodies recognizing p65 NF-κB. IPs with murine (m) IgG were performed in parallel to control the specificity of the interactions. Arrows mark different HDAC3 isoforms.

Figure 5.

Acetylation mimicking mutants of p65 suppress IL-8 gene expression. (A) HEK293T cells were transfected with expression plasmids encoding His-tagged p65 wild-type or a mutated version thereof where K122, K123, K310, K314 and K315 (K5R) were mutated to arginines along with 0.75 and 1.5 µg of a plasmid for YFP-tagged CBP. One fraction of cells was lysed under denaturing conditions to analyze acetylation of p65 with the indicated acetyl-lysine-specific antibodies. The lower part shows the input control ensuring adequate protein expression. (B) HEK293IL-1 R cells were transfected with the indicated p65 wild-type and acetylation-deficient expression vectors plus the IL-8 promoter reporter gene construct and the pSV40-β-galactosidase construct. Twenty-four hours later, cells were lysed, and luciferase reporter gene activity was determined and normalized for β-galactosidase activity. Normalized mean luciferase activities ± SEM relative to the vector control from five independent experiments (upper graph) are shown. In parallel transfections, cells were transfected to express HA-tagged p65 wild-type or the indicated acetylation-mimicking mutants thereof. The next day, cells were lysed, and one portion of the lysates was analyzed for mRNA expression of the endogenous IL-8 gene by RT-qPCR. Bars show mean ± SEM from two independent experiments (lower graph). Another portion was tested by immunoblotting for correct expression of p65. The positions of a molecular weight marker are shown, and the HA-p65 migrates slightly slower than the endogenous p65. (C) HEK293IL-1R cells were transfected with additional p65 mutants and analyzed exactly as described in (B). IL-8 promoter activity represents the mean ± SEM from three independent experiments normalized for protein concentration (upper graph). IL-8 mRNA expression represents the mean fold change ± SEM relative to the vector control from three independent experiments (lower graph). All samples were analyzed for comparable expression of the constructs in parallel, one representative immunoblot is shown.

Figure 6.

Suppression of HDAC3 affects p65 recruitment and phosphorylation of RNA polymerase II at the IL-8 promoter. (A) Human KB cells were stably transfected with an empty shRNA vector (pS-Puro, gray bars) or with a plasmid encoding a HDAC3-specific shRNA (pS-HDAC3, white bars) to generate shRNA-mediated HDAC3 knockdown cells. HDAC3 knockdown and control cells were treated for 1 h with IL-1 or were left untreated. IL-8 mRNA expression was determined by RT-qPCR. Data represent the mean IL-8 expression ± SEM relative to the unstimulated control (n = 2). (B) The same cells as in (A) were stimulated for the indicated times with IL-1 or were left untreated. Cytosolic (C), soluble nuclear (N1) and insoluble nuclear (N2) extracts were prepared and analyzed for the presence of the indicated proteins by immunoblotting. The amounts of p65 and HDAC3 in the nuclear fractions were quantitated, and expression levels are shown relative to those of untreated cells. Shown is one out of two experiments. (C) The same cells as in (A) were stimulated for 1 h with IL-1 or were left untreated. Chromatin immunoprecipitation was performed using the indicated antibodies and imunoprecipitated DNA was quantified by qPCR using primers spanning the IL-8 promoter/enhancer region and a non-regulatory upstream region as a negative control. Relative enrichment of immunoprecipitated DNA is shown as percentage of input DNA. Shown are mean values ± SEM from two independent experiments performed in duplicates.

Figure 7.

HDAC3 is a global regulator of the transcriptional IL-1 response. HEK293IL-1 R cells stably transfected with pSuper-Puro (pS-Puro) or with pSuper-Puro encoding a shRNA-directed against HDAC3 (pS-HDAC3), as shown in Figure 1A, were treated for 24 h with vehicle (EtOH, 0.03%) or with TSA (300 ng/ml). These cells remained unstimulated or were treated for 3 h with IL-1. Total RNA was isolated from two independent experimental groups (exp. 1 and exp. 2), labeled cRNA was prepared and was hybridized to whole genome Agilent microarrays. (A) Normalized fluorescence intensity values were used to calculate the total number of genes, which were regulated in the same direction by IL-1 by at least 1.5-fold in both independent experiments compared with the EtOH-treated vehicle control. Data analysis revealed 70 IL-1 regulated genes. Of those, only seven genes were still inducible in the absence of HDAC3, as illustrated by the depicted Venn diagram. (B) The gene set identified in (A) was clustered hierarchically, and ratio values are shown as a colored heatmap. Vertical color-coded bars indicate groups of genes contained in the Venn diagram shown in (A). Brown vertical bars indicate all genes with at least 2-fold regulation by IL-1. The entire data set is shown in Supplementary Table S2.

Figure 8.

Comparison with the pan-HDAC inhibitor TSA reveals differential and restricted effects of HDAC3 and IL-1 on specific sets of genes. (A) The microarray data set from the experiments described in Figure 7 was used to calculate the total number of genes, which were regulated in the same direction by TSA by at least 2-fold in both independent experiments compared with the EtOH-treated vehicle control. This revealed 4241 genes whose overlap between control cells and HDAC3 knockdown cells is illustrated by the depicted Venn diagram. (B) The TSA-regulated set of genes identified in (A) was additionally filtered for HDAC3-dependent genes based on a ratio of pS-HDAC3/pS-Puro (lanes 13, 14) of at least 1.5-fold in two independent experiments revealing 130 genes. Similarly, the set was filtered for genes, which were regulated by IL-1 by at least 1.5-fold revealing 24 genes. Depicted are color-coded ratio values of all 154 genes, which were calculated by dividing fluorescence intensity measurements of vehicle-treated cells or IL-1 or TSA treatments by those obtained for the vehicle control situation of either control cells (lanes 1, 2) or HDAC3 knockdown cells (lanes 7, 8). In lane 13 and 14, ratios were calculated by dividing pS-HDAC3 by pS-Puro values. Both data were sorted according to the mean regulation by TSA in descending order. Vertical green and blue bars indicate HDAC3- and IL-1-dependent genes as described for (A). The ratio values from both independent experiments are shown side by side (exp. 1, 2). Solid black arrows indicate examples of genes, which are elevated by both, TSA and HDAC3 knockdown, whereas dotted black arrows indicate genes with adverse regulation by TSA or HDAC3 knockdown, respectively. (C) Depicted are the color-coded ratio values for the 75 most strongly upregulated genes (red colors) or 75 most strongly downregulated genes (blue colors) in response to TSA from the same data set.

Figure 8.

Comparison with the pan-HDAC inhibitor TSA reveals differential and restricted effects of HDAC3 and IL-1 on specific sets of genes. (A) The microarray data set from the experiments described in Figure 7 was used to calculate the total number of genes, which were regulated in the same direction by TSA by at least 2-fold in both independent experiments compared with the EtOH-treated vehicle control. This revealed 4241 genes whose overlap between control cells and HDAC3 knockdown cells is illustrated by the depicted Venn diagram. (B) The TSA-regulated set of genes identified in (A) was additionally filtered for HDAC3-dependent genes based on a ratio of pS-HDAC3/pS-Puro (lanes 13, 14) of at least 1.5-fold in two independent experiments revealing 130 genes. Similarly, the set was filtered for genes, which were regulated by IL-1 by at least 1.5-fold revealing 24 genes. Depicted are color-coded ratio values of all 154 genes, which were calculated by dividing fluorescence intensity measurements of vehicle-treated cells or IL-1 or TSA treatments by those obtained for the vehicle control situation of either control cells (lanes 1, 2) or HDAC3 knockdown cells (lanes 7, 8). In lane 13 and 14, ratios were calculated by dividing pS-HDAC3 by pS-Puro values. Both data were sorted according to the mean regulation by TSA in descending order. Vertical green and blue bars indicate HDAC3- and IL-1-dependent genes as described for (A). The ratio values from both independent experiments are shown side by side (exp. 1, 2). Solid black arrows indicate examples of genes, which are elevated by both, TSA and HDAC3 knockdown, whereas dotted black arrows indicate genes with adverse regulation by TSA or HDAC3 knockdown, respectively. (C) Depicted are the color-coded ratio values for the 75 most strongly upregulated genes (red colors) or 75 most strongly downregulated genes (blue colors) in response to TSA from the same data set.

Figure 9.

Specific inhibitory effects of p65 NF-κB acetylation-mimicking mutants on IL-1-induced gene expression. (A) NF-κB p65-deficient Mefs (p65^−/−) were reconstituted to express p65 NF-κB wild-type (HA-p65) and its mutant forms as described in (15). A representative western blot ensuring expression of comparable amounts of p65 proteins is shown. (B) Mefs were stimulated for various periods with IL-1 as indicated. Expression of Cxcl2 was analyzed by RT-qPCR. Relative changes of mRNA expression compared with untreated cells reconstituted with p65 NF-κB wild-type were calculated. Shown are the relative mean values ± SEM from three independent experiments compared with cells reconstituted with wild-type p65 (set as 1). (C and D) Two RNA preparations from experiments performed as described in (B) were used to prepare cRNA and to hybridize whole genome Agilent microarrays. A set of 851 p65-dependent or IL-1-induced genes was extracted based on 2-fold regulation in both experiments. This set was further filtered for genes affected by p65 K4Q or p65 K5Q. Color-coded ratio values for the top-ranking genes of these two subsets are shown. The entire sets of data are shown in Supplementary Figure S7A and B. Gray colors indicate genes with low hybridization signals on the microarrays. (C) Expression values were used to identify 851 genes, which were regulated by IL-1 (ratio IL-1 p65 wild-type/p65^−/−) or are dependent on p65 NF-κB (ratio p65 wild-type/p65^−/−). These genes are sorted according to mean regulation by IL-1 in descending order and are indicated by the green vertical bar. Genes regulated by IL-1 and suppressed by p65 K4Q (blue vertical bars) or by p65 K5Q (brown vertical bars) were identified by ratio values of p65 wild-type + IL-1/p65 K4Q + IL-1 or p65 K5Q + IL-1 ≥ 2.0. (D) Expression values were used to identify 167 genes, which were dependent on p65 NF-κB (ratio p65 wild-type/p65^−/−) and regulated by p65 K4Q (yellow vertical bars) or by p65 K5Q (purple vertical bars) based on the ratio p65 K4Q or p65 K5Q/p65 wild-type followed by p65 K4Q/p65 K4R and by p65 K5Q/p65 K5R. These genes are sorted according to mean regulation by p65 K4Q in descending order.

Figure 9.

Specific inhibitory effects of p65 NF-κB acetylation-mimicking mutants on IL-1-induced gene expression. (A) NF-κB p65-deficient Mefs (p65^−/−) were reconstituted to express p65 NF-κB wild-type (HA-p65) and its mutant forms as described in (15). A representative western blot ensuring expression of comparable amounts of p65 proteins is shown. (B) Mefs were stimulated for various periods with IL-1 as indicated. Expression of Cxcl2 was analyzed by RT-qPCR. Relative changes of mRNA expression compared with untreated cells reconstituted with p65 NF-κB wild-type were calculated. Shown are the relative mean values ± SEM from three independent experiments compared with cells reconstituted with wild-type p65 (set as 1). (C and D) Two RNA preparations from experiments performed as described in (B) were used to prepare cRNA and to hybridize whole genome Agilent microarrays. A set of 851 p65-dependent or IL-1-induced genes was extracted based on 2-fold regulation in both experiments. This set was further filtered for genes affected by p65 K4Q or p65 K5Q. Color-coded ratio values for the top-ranking genes of these two subsets are shown. The entire sets of data are shown in Supplementary Figure S7A and B. Gray colors indicate genes with low hybridization signals on the microarrays. (C) Expression values were used to identify 851 genes, which were regulated by IL-1 (ratio IL-1 p65 wild-type/p65^−/−) or are dependent on p65 NF-κB (ratio p65 wild-type/p65^−/−). These genes are sorted according to mean regulation by IL-1 in descending order and are indicated by the green vertical bar. Genes regulated by IL-1 and suppressed by p65 K4Q (blue vertical bars) or by p65 K5Q (brown vertical bars) were identified by ratio values of p65 wild-type + IL-1/p65 K4Q + IL-1 or p65 K5Q + IL-1 ≥ 2.0. (D) Expression values were used to identify 167 genes, which were dependent on p65 NF-κB (ratio p65 wild-type/p65^−/−) and regulated by p65 K4Q (yellow vertical bars) or by p65 K5Q (purple vertical bars) based on the ratio p65 K4Q or p65 K5Q/p65 wild-type followed by p65 K4Q/p65 K4R and by p65 K5Q/p65 K5R. These genes are sorted according to mean regulation by p65 K4Q in descending order.

Figure 9.

Specific inhibitory effects of p65 NF-κB acetylation-mimicking mutants on IL-1-induced gene expression. (A) NF-κB p65-deficient Mefs (p65^−/−) were reconstituted to express p65 NF-κB wild-type (HA-p65) and its mutant forms as described in (15). A representative western blot ensuring expression of comparable amounts of p65 proteins is shown. (B) Mefs were stimulated for various periods with IL-1 as indicated. Expression of Cxcl2 was analyzed by RT-qPCR. Relative changes of mRNA expression compared with untreated cells reconstituted with p65 NF-κB wild-type were calculated. Shown are the relative mean values ± SEM from three independent experiments compared with cells reconstituted with wild-type p65 (set as 1). (C and D) Two RNA preparations from experiments performed as described in (B) were used to prepare cRNA and to hybridize whole genome Agilent microarrays. A set of 851 p65-dependent or IL-1-induced genes was extracted based on 2-fold regulation in both experiments. This set was further filtered for genes affected by p65 K4Q or p65 K5Q. Color-coded ratio values for the top-ranking genes of these two subsets are shown. The entire sets of data are shown in Supplementary Figure S7A and B. Gray colors indicate genes with low hybridization signals on the microarrays. (C) Expression values were used to identify 851 genes, which were regulated by IL-1 (ratio IL-1 p65 wild-type/p65^−/−) or are dependent on p65 NF-κB (ratio p65 wild-type/p65^−/−). These genes are sorted according to mean regulation by IL-1 in descending order and are indicated by the green vertical bar. Genes regulated by IL-1 and suppressed by p65 K4Q (blue vertical bars) or by p65 K5Q (brown vertical bars) were identified by ratio values of p65 wild-type + IL-1/p65 K4Q + IL-1 or p65 K5Q + IL-1 ≥ 2.0. (D) Expression values were used to identify 167 genes, which were dependent on p65 NF-κB (ratio p65 wild-type/p65^−/−) and regulated by p65 K4Q (yellow vertical bars) or by p65 K5Q (purple vertical bars) based on the ratio p65 K4Q or p65 K5Q/p65 wild-type followed by p65 K4Q/p65 K4R and by p65 K5Q/p65 K5R. These genes are sorted according to mean regulation by p65 K4Q in descending order.