The histone deacetylase inhibitor valproic acid inhibits NKG2D expression in natural killer cells through suppression of STAT3 and HDAC3

Abstract

NKG2D is a major activating receptor of NK cells and plays a critical role in tumor immunosurveillance. NKG2D expression in NK cells is inhibited by the histone deacetylase (HDAC) inhibitor valproic acid (VPA) and enhanced by the narrow-spectrum HDAC inhibitor entinostat. We previously demonstrated that entinostat enhanced NKG2D transcription by increasing acetylation of Histones H3 and H4. However, the mechanism by which VPA reduces NKG2D expression in NK cells is not known. We have also shown that NKG2D transcription is regulated by STAT3 phosphorylation. In this study, we investigated regulation of NKG2D expression in NK cells by VPA and entinostat by assessing protein expression, phosphorylation, and interaction of HDACs and STAT3. We find that VPA selectively inhibits STAT3 tyrosine705 phosphorylation, but entinostat does not. STAT3 complexes with HDAC3, and HDAC3 inhibition represses STAT3 phosphorylation and therefore NKG2D expression. NK cells from STAT3 wild-type mice downregulate NKG2D in response to VPA, but not NK cells from STAT3 knockout mice. These results show that VPA is a potent inhibitor of STAT3 phosphorylation and demonstrate that histone acetylation and STAT3 tyrosine705 phosphorylation cooperate in regulating NKG2D expression in NK cells.

Figure 1. Effect of HDAC inhibitors on NKG2D expression after 24 hrs treatment.

NK cells were treated with 0.1 μM of entinostat or 0.5 mM of VPA for 24 hrs, NKG2D expression was accessed by flow cytometry. (A) Representative of 5 independent experiments. (B) Percentage increasing of NKG2D expression was calculated by the formula: (NKG2D Mean in treated NK cells - NKG2D Mean in untreated NK cells)/NKG2D Mean in untreated NK cells *100. The data were pooled from 5 donors. (C) Effect of VPA concentration on NKG2D expression. *p < 0.05; **p < 0.01; ***p < 0.001.

Figure 2. VPA impairs but entinostat enhances NK cell degranulation.

Primary human NK cells were treated with 0.1 μM of entinostat and 0.1 mM of VPA for 24 hrs, respectively; and then co-incubated with COL, HCT-15 and SaOS2 cells (E:T = 4:1) for 4 hrs. NK cell degranulation was evaluated by flow cytometry for CD107a expression. (A) Representative dot plot; (B) Percentage increasing of CD107a expression was calculated by the formula: (percentage CD107a in treated NK cells - percentage CD107a in untreated NK cells)/percentage CD107a in untreated NK cells *100. The data are expressed as mean ± SEM of three independent experiments. *p < 0.05; **p < 0.01; ***p < 0.001.

Figure 3. VPA inactivates STAT3 in NK cells.

(A) Expression and phosphorylation of STAT3 in purified primary NK cells with or without treatment with 1.0 μM entinostat or 0.1 mM VPA. (B) Phosphorylation of STAT1 and STAT5 in purified primary NK cells with or without treatment with 0.1 mM VPA and IL-2. Similar results were obtained in three independent experiments.

Figure 4. NKG2D expression on NK cells from STAT3 wild-type and knockout mice.

Murine NK cells were isolated from spleens of STAT3 wild-type and knockout mice, and then treated with/without 0.1 mM VPA for 24 hrs. NKG2D surface expression on CD3-NKp46+ cells was analyzed by flow cytometry. The data was pooled from 4 mice. *p < 0.05; **p < 0.01; ***p < 0.001; NS, non-significant.

Figure 5. STAT3 interacts with HDAC3 in NK cells.

(A) Proteins were immunoprecipitated from NK cell lysate by HDAC3 monoclonal antibody and isotype IgG, and then detected by STAT3 and HDAC3 antibodies using two separate blots with the same sample to avoid the HDAC3 (49 kDa) to be overlapped with the antibody heavy IgG (55 kDa). (B) Proteins were immunoprecipitated from NK cell lysate by STAT3 monoclonal antibody and isotype IgG, and then detected by HDAC3 and STAT3 antibodies using two separate blots with the same sample. (C) NK cells were treated with Entinostat, VPA and RGFP966. Phosphorylation, acetylation and total protein of STAT3 and acetylated histone H3/H4 were detected by western-blot. Similar results were obtained in three independent experiments.

Figure 6. NKG2D expression on NK cells after treatment with VPA, RGFP966, CAY10603 and PCI34051.

NK cells were treated with indicated concentrations of VPA, RGFP966, CAY10603 and PCI34051 for 24 hrs, and then NKG2D surface expression was analyzed by flow cytometry. The data are pooled from 3–5 donors. *p < 0.05; **p < 0.01; ***p < 0.001.

Figure 7. Model of transcriptional regulation of NKG2D by the phosphorylation of STAT3 and acetylation of Histones-H3 and H4 in NK cells.

HDAC3 is required for STAT3 phospharylation which determines NKG2D transcript in NK cells. VPA inhibits HDAC3 and results in suppression of STAT3 phosphorylation and then reduction of NKG2D expression in NK cells. Histones-H3 and H4 can be acetylated by HAT, and acetylated-H3 and H4 can be deacetylated by HDAC. Entinostat preferentially inhibits HDAC1 and is less active against HDAC3 resulting in enhanced acetylation of H3 and H4 and then increased binding of acetylated H3 and H4 (Ac-H3, Ac-H4) to the promoter of NKG2D, promoting NKG2D transcription.