Histone deacetylation contributes to low extracellular superoxide dismutase expression in human idiopathic pulmonary arterial hypertension

Abstract

Epigenetic mechanisms, including DNA methylation and histone acetylation, regulate gene expression in idiopathic pulmonary arterial hypertension (IPAH). These mechanisms can modulate expression of extracellular superoxide dismutase (SOD3 or EC-SOD), a key vascular antioxidant enzyme, and loss of vascular SOD3 worsens outcomes in animal models of pulmonary arterial hypertension. We hypothesized that SOD3 gene expression is decreased in patients with IPAH due to aberrant DNA methylation and/or histone deacetylation. We used lung tissue and pulmonary artery smooth muscle cells (PASMC) from subjects with IPAH at transplantation and from failed donors (FD). Lung SOD3 mRNA expression and activity was decreased in IPAH vs. FD. In contrast, mitochondrial SOD (Mn-SOD or SOD2) protein expression was unchanged and intracellular SOD activity was unchanged. Using bisulfite sequencing in genomic lung or PASMC DNA, we found the methylation status of the SOD3 promoter was similar between FD and IPAH. Furthermore, treatment with 5-aza-2'-deoxycytidine did not increase PASMC SOD3 mRNA, suggesting DNA methylation was not responsible for PASMC SOD3 expression. Though total histone deacetylase (HDAC) activity, histone acetyltransferase (HAT) activity, acetylated histones, and acetylated SP1 were similar between IPAH and FD, treatment with two selective class I HDAC inhibitors increased SOD3 only in IPAH PASMC. Class I HDAC3 siRNA also increased SOD3 expression. Trichostatin A, a pan-HDAC inhibitor, decreased proliferation in IPAH, but not in FD PASMC. These data indicate that histone deacetylation, specifically via class I HDAC3, decreases SOD3 expression in PASMC and HDAC inhibitors may protect IPAH in part by increasing PASMC SOD3 expression.

Keywords: DNA methylation; extracellular superoxide dismutase; histone deacetylation; idiopathic pulmonary arterial hypertension.

Fig. 1.

Decreased lung SOD3 mRNA expression and protein activity in IPAH. A: lung SOD3/β2M mRNA expressed relative to FD (n = 13–16). B: Western blot analysis for lung SOD3 and calnexin, with corresponding densitometry data showing SOD3/calnexin relative to FD (n = 6). C: SOD3 in lung homogenates was separated from the intracellular SODs by using concanavalin A-Sepharose 4B beads to pull down SOD3. SOD3 activity was determined as the difference in activity in supernatant after incubation with concanavalin A-Sepharose 4B (intracellular SOD) or plain Sepharose 4B beads (total SOD) SOD activity was assayed with the SOD assay kit-WST (Dojindo) and expressed as units SOD activity per mg protein (U/mg protein) (n = 6). FD, failed donor; IPAH, idiopathic pulmonary arterial hypertension. *P < 0.05 vs. FD by unpaired t-test.

Fig. 2.

Variable PASMC SOD3 mRNA and protein expression tends to decrease in IPAH. A: PASMC SOD3/β2M mRNA by qPCR expressed relative to FD. B: Western blot for PASMC SOD3 and β-actin, with corresponding densitometry data for SOD3/β-actin relative to FD (n = 6). P > 0.05 vs. FD by unpaired t-test.

Fig. 3.

No change in lung SOD2 expression or intracellular SOD activity in IPAH. A: Western blot for lung SOD2 and calnexin, with corresponding densitometry data for SOD2/calnexin relative to FD (n = 6). P > 0.05 vs. FD by unpaired t-test. SOD2 expression was tested in the same membrane shown in Fig. 1B. B: intracellular SOD activity in lung expressed as units SOD activity per mg protein (U/mg protein) (n = 6). P > 0.05 vs. FD by unpaired t-test. IC-SOD, intracellular SOD.

Fig. 4.

No evidence that the low SOD3 gene expression in IPAH is regulated by DNA methylation of the SOD3 promoter. Lung genomic DNA was subject to bisulfite conversion and sequencing for the 18 CpG sites in the SOD3 promoter. A: percent methylation in the SOD3 promoter region in FD and IPAH lung (n = 4). P > 0.05 vs. FD by unpaired t-test. B: DNA methylation at each of the 18 CpG sites in the SOD3 promoter in FD and IPAH lung. P > 0.05 between FD and IPAH for each CpG site by 2-way ANOVA. C: percent methylation in the SOD3 promoter region in FD and IPAH PASMC (n = 6); *P < 0.05 vs. FD by unpaired t-test. D: DNA methylation at each of the 18 CpG sites in the SOD3 promoter for FD and IPAH PASMC; n = 6. E: PASMC SOD3/β2M mRNA expression following treatment with DNA methyltransferase inhibitor, 1 μM 5-aza-dC, on days 1–4 with harvesting on day 5 in FD vs. IPAH. Data are expressed as change in SOD3/β2M from baseline for each individual (n = 6). P > 0.05 vs. FD by unpaired t-test.

Fig. 5.

No change in HDAC activity, HAT activity, or total histone H3 and H4 acetylation between FD and IPAH. Class-specific HDAC activity was determined by incubating tissue or cell extracts with specific synthetic HDAC substrates against class I, class IIa, or class IIb HDACs. Activity levels were measured in FD and IPAH lung (A, C, E) or PASMC (B, D, F). Data are expressed as the fluorescent signal relative to the FD (n = 14–16 for lung; n = 6 for PASMC). P > 0.05 vs. FD by unpaired t-test. G: HAT activity was measured in nuclear extracts isolated from PASMC and expressed as ng/min (n = 6). P > 0.05 vs. FD by unpaired t-test. H: Western blot analysis for acetylated histone H3 (H3ac), acetylated histone H4 (H4ac), and total histone H3 (H3) in histone extracts from FD and IPAH with corresponding densitometry for H3ac (I) or H4ac (J) expressed relative to total H3 (n = 5–6). P > 0.05 vs. FD by unpaired t-test.

Fig. 6.

Treatment with HDAC inhibitors increase SOD3 mRNA expression in PASMC and enhanced cell proliferation in IPAH. A: SOD3 mRNA expression following treatment with the following HDAC inhibitors, expressed as SOD3/β2M. Cells were treated for 24 h with the general HDAC inhibitor trichostatin A (TSA) (200 nM); selective class I HDAC 1, 2, and 3 inhibitor MGCD0103 (MGCD) (1 μM); class I HDAC 1, 2, and 3 inhibitor entinostat (MS275) (1 μM); class I HDAC 1 and 2 inhibitor biaryl-60 (BA-60) (1 μM); class IIB HDAC6 inhibitor tubastatin A (TubA) (1 μM); or dimethyl sulfoxide (DMSO) (1:1,000) (n = 5). B: cell counts at 2 days and 4 days after a 24-h treatment with TSA (200 nM) in FD and IPAH PASMC (n = 4–5). Doubling time was measured by using the xCELLigence Real-Time Cell Analyzer (ACEA Biosciences) to provide a real-time measurement of cell proliferation. Cells treated with either TSA (200 nM) or DMSO (1:1,000) were plated 24 h posttreatment in fresh medium on an E-plate 16 (1,000 cells/well) and monitored continuously over a 48-h period (n = 4–5). *P < 0.05 vs. FD DMSO, #P < 0.05 vs. IPAH DMSO by 2-way ANOVA.

Fig. 7.

siRNA knockdown of class I HDAC3 in PASMC increased SOD3 protein expression. PASMC (Lonza) were transfected with siRNA (Life Technologies) against HDAC1, HDAC2, and/or HDAC3 and combinations of the 3 siRNA molecules. HDAC1-3/β2M mRNA by qPCR expressed relative to siNC (A–C). The data were pooled into 3 groups: 1) HDAC1 alone, 2) HDAC2 and HDAC1&2, and 3) HDAC3, HDAC1&3, and HDAC1,2&3. D: representative Western blot of SOD3 and β-actin for each experimental condition. E: densitometry data for SOD3 expression. Data are expressed as SOD3/β-actin relative to HDAC1. Experiments were repeated at least 3 times; n = 3–4; *P < 0.05 vs. HDAC1 and #P < 0.05 vs. HDAC2 and HDAC1&2 group by 1-way ANOVA. siNC, negative control siRNA.

Fig. 8.

Acetylation of SP1 did not differ between FD and IPAH PASMC. A: representative Western blot data of SP1 and total histone 3 expression in nuclear extracts of FD and IPAH PASMC along with densitometry data. B: SP1 was immunoprecipitated from PASMC nuclear extracts and evaluated for protein acetylation. The Western blot and corresponding densitometry are shown for acetylated lysine (Ac-lysine) and SP1 (n = 3). P > 0.05 by unpaired t-test.

Fig. 9.

Treatment of chronically hypoxic rats with the HDAC inhibitor MGCD0103 increased lung Sod3 mRNA expression. Sod3 mRNA expression in the lungs of 3-wk chronically hypoxic rats treated daily with the HDAC inhibitor MGCD0103 (10 mg/kg) intraperitoneal injections (HX + HDACi) compared with sham treated normoxic (NX) and hypoxic (HX) rats (2) (n = 6). *P < 0.001 vs. NX, #P < 0.001 vs. HX by 1-way ANOVA.