Class I histone deacetylase inhibitor entinostat suppresses regulatory T cells and enhances immunotherapies in renal and prostate cancer models

Abstract

Background: Immunosuppressive factors such as regulatory T cells (Tregs) limit the efficacy of immunotherapies. Histone deacetylase (HDAC) inhibitors have been reported to have antitumor activity in different malignancies and immunomodulatory effects. Herein, we report the Tregs-targeting and immune-promoting effect of a class I specific HDAC inhibitor, entinostat, in combination with either IL-2 in a murine renal cell carcinoma (RENCA) model or a survivin-based vaccine therapy (SurVaxM) in a castration resistant prostate cancer (CR Myc-CaP) model.

Methods and results: RENCA or CR Myc-CaP tumors were implanted orthotopically or subcutaneously, respectively. Inoculated mice were randomized into four treatment groups: vehicle, entinostat, cytokine or vaccine, and combination. Tregs in the blood were assessed by FACS analysis. Real time quantitative PCR and Western blot analysis of isolated T cell subpopulations from spleen were performed to determine Foxp3 gene and protein expression. The suppressive function of Tregs was tested by T cell proliferation assay. Low dose (5 mg/kg) entinostat reduced Foxp3 levels in Tregs and this was associated with enhanced tumor growth inhibition in combination with either IL-2 or a SurVaxM vaccine. Entinostat down-regulated Foxp3 expression transcriptionally and blocked Tregs suppressive function without affecting T effector cells (Teffs). In vitro low dose entinostat (0.5 µM) induced STAT3 acetylation and a specific inhibitor of STAT3 partially rescued entinostat-induced down-regulation of Foxp3, suggesting that STAT3 signaling is involved in Foxp3 down-regulation by entinostat.

Conclusions: These results demonstrate a novel immunomodulatory effect of class I HDAC inhibition and provide a rationale for the clinical testing of entinostat to enhance cancer immunotherapy.

Figure 1. Enhancement of IL-2 therapy by entinostat is associated with inhibition of Tregs.

Mice with orthotopic inoculation of RENCA cells were treated for 5 days. Blood was drawn from mice on the fifth day and stained for antibodies specific for CD4, CD25, and Foxp3. Tumor weights were measured at the end of two weeks of treatment. A and B, Effect of entinostat on Tregs in tumor bearing mice. A, Effects of vehicle, IL-2, entinostat, or combination treatment on Tregs Foxp3 expression. Cells were stained and subjected to flow cytometry analysis. The dot plots were gated for CD4⁺ cells. The rectangular area encloses the CD4⁺Foxp3⁺ cells, the numbers on the top represent the percentage of Foxp3⁺ cells. The numbers on the bottom in the area represent the mean fluorescence intensity (MFI) of Foxp3 PE staining of CD4⁺Foxp3⁺ cells. B, Quantification of Tregs percentage in CD4 population (left panel) and Foxp3 levels (MFI) in Tregs (right panel) by FACS analysis. Values are means and error bars represent S.D. for 5–7 samples per group. In right panel, p = 0.0011 for IL-2 vs. vehicle; p = 0.000009 for entinostat vs. vehicle. Results are representative of three separate experiments. C, Tumor weight measurements. Columns, mean grams of tumor; Bars, S.D. n = 5–7. p = 0.0209 for entinostat vs. vehicle; p = 0.0077 for combination vs. vehicle; p = 0.0272 for combination vs. entinostat. Results are representative of three separate experiments. D, Entinostat enhanced IFN-γ type immune response induced by IL-2 treatment. Splenocytes (1×10⁶ cells) were harvested and stimulated with PMA (20 ng/ml) and Ionomycin (1 µg/ml) for 5 hours in the presence of Brefeldin A. Cells were then stained for surface markers and intracellular IFN-γ. Histograms show percentage of IFN-γ expressing cells in CD8 population. p = 0.01 for combination vs. IL-2. E, Entinostat reduced tumor infiltration of Tregs. Tumor sections were stained with anti-Foxp3 antibody to show infiltration of Tregs. Histogram shows average numbers of stained Tregs in random 20× resolution bright fields (Tregs number in each field was obtained by blinded count). F, Entinostat treatment induced H3 histone acetylation in splenocytes. BALB/c mice were treated with vehicle (0.5% methocel) or 5 mg/kg/day entinostat by gavage for 5 days. Cells were harvested from spleens and subjected to Western blot analysis for acetylated-H3 histone.

Figure 2. Antitumor activity of Tregs depletion antibody PC61 in RENCA tumor.

A, Plots show effect of vehicle, PC61, entinostat or combination treatment on CD25 and Foxp3 levels. Quantification of Tregs percentage, Foxp3 expression by FACS analysis is shown as histograms. Values are means and error bars represent S.D. for 6–7 samples per group. B, Tumor weight measurements from PC61 depletion experiment. Columns, mean grams of tumor; Bars, S.D.. n = 6–7. * represents p<0.05 for marked point vs. vehicle; ** represents p<0.01 for marked point vs. vehicle.

Figure 3. The HDAC inhibitor, entinostat, enhances peptide vaccine therapy in a castration resistant prostate cancer model.

A and B, Entinostat enhances the antitumor effect of the survivin vaccine, SurVaxM. FVB mice were castrated and subcutaneously inoculated with small castration resistant tumor pieces. Approximately 7 days after inoculation when tumors reached an average size of 50 mm², mice were treated three weeks with vehicle, entinostat (5 mg/kg, 1 dose/day, 5 days/wk), survivin vaccine SurVaxM (100 µg/dose, 1 dose/wk), or combination. A, Single tumor growth graph lines were generated by serial caliper measurements. B, Tumors were harvested at the end of treatment and weighed (combination vs. survivin vaccine, p = 0.003). C, Entinostat treatment reduced Foxp3 levels in Tregs from CR Myc-CaP tumor-bearing mice. After 5 days of treatment, peripheral blood cells were collected from mice, stained for CD4 and Foxp3, and subjected to FACS analysis. Quantitation of Foxp3 mean fluorescence intensity (Foxp3 MFI) (vehicle vs. entinostat, p = 0.0002; combination vs. survivin, p = 0.0004).

Figure 4. Survivin vaccine induces antigen-specific CD8 cells and entinostat enhances IFN-γ induction.

A. Tetramer analysis of splenocytes obtained from mice immunized with SurVaxM. Splenocytes were stained with anti CD8 antibodies and survivin-specific tetramers for flow cytometric analysis. Results are based upon gating of CD8+ T cells and indicate the percent of double labeled cells (CD8+/Tetramer+) with respect to specific tetramer. B, Combination treatment led to CD8⁺ IFN-γ⁺ cells induction. Mice were treated as indicated. Splenocytes were stimulated and intracellular IFN-γ staining was performed as described in Figure 1D. p<0.01 for combination vs. vehicle.

Figure 5. In vivo treatment with entinostat decreases Foxp3 expression in Tregs and inhibits Tregs function.

BALB/c mice were treated with vehicle (0.5% methocel) or 5 mg/kg/day or 20 mg/kg/day of entinostat by gavage for 5 days. Cells were harvested from spleens and lymph nodes. A, In vivo entinostat treatment decreased the expression level of Foxp3 in CD4⁺Foxp3⁺ (Treg) cells. Cells were stained with fluorescence-conjugated antibodies specific to CD4, CD25, and Foxp3 and subjected to flow cytometry analysis. The dot plots were gated for CD4⁺ cells. The rectangular area encloses the CD4⁺Foxp3⁺ cells, the numbers on the top represent the percentage of Foxp3⁺ cells. The numbers on the bottom in the area represent the mean fluorescence intensity of Foxp3 PE staining of CD4⁺Foxp3⁺ cells. The panels on the right represent quantification of Tregs percentage (top) and Foxp3 expression (bottom) in FACS analysis. Graph shows means, error bars represent standard deviations for 6 samples in each group. p = 0.00078 for 5 mg/kg entinostat vs. vehicle; p = 0.000004 for 20 mg/kg entinostat vs. vehicle; p = 0.0038 for 20 mg/kg entinostat vs. 5 mg/kg entinostat. α = 0.05. Results are representative of three separate experiments. B, In vivo entinostat treatment reduced Foxp3 mRNA levels in isolated Tregs. Total RNAs were extracted from isolated CD4⁺CD25⁺ T cells (Tregs) and CD4⁺CD25⁻ T cells (Teffs) and were analyzed for Foxp3 mRNA by real time RT-PCR. Foxp3 mRNA levels were normalized to GAPDH mRNA (or ribosomal RNA RPL13A) and values represent means for three samples (three mice per sample) per group and have relative units. p = 0.0157 for Tregs treated with entinostat vs. Tregs treated with vehicle. Results are representative of three separate experiments. C, In vivo entinostat treatment reduced Foxp3 protein level in Tregs. Total cell protein was extracted from isolated Tregs and Teffs and was analyzed for Foxp3 protein by Western blot. D, The effects of HDAC inhibitor, entinostat, on Tregs suppressive function and Teffs proliferation. Tregs suppression assays: 2×10⁵ CFSE-labeled Teffs were stimulated with 0.5 µg/ml of anti-CD3ε antibody and 4×10⁵ irradiated antigen presenting cells (CD4 cells-depleted splenocytes) and co-cultured with Tregs in different ratios to Teffs for 62–80 hrs. Percentages of divided Teffs were calculated at all ratio points for each treatment. Percentage of suppression of Teffs dividing was calculated for each point by comparing cell dividing at each ratio to the cell dividing in absence of Tregs (0∶1). Left panel: Effect of entinostat treatment on Tregs function. Right panel: Effect of entinostat treatment on proliferation of Teffs. Histogram shows percentage of divided Teffs without Tregs. Data are representative of three separate experiments. Values are means from triplicate measurements. * represents p<0.05 for marked point vs. vehicle; ** represents p<0.01 for marked point vs. vehicle.

Figure 6. STAT3 signaling is involved in down-regulation of Foxp3 by entinostat.

A and B, Entinostat treatment induces STAT3 acetylation. HepG2 cells or splenocytes were treated for 6 hours, harvested, and lyzed for Western immunobloting or immunoprecipitation. A, Entinostat induced STAT3 acetylation in HepG2 cells. Upper panel: Cell lysates were analyzed directly and blotted with anti-STAT3 and actin. Bottom panel, Cell lysates were immunoprecipitated with anti-STAT3 antibody and then blotted with anti-STAT3 antibody and with anti-acetylated lysine antibody. B, Entinostat induced STAT3 acetylation in splenocytes. IP and Western blot were performed as described in A. C and D, Down-regulation of Foxp3 is inhibited by blocking STAT3 signaling. Splenocytes were harvested from BALB/c mice and put in culture. Cultures were treated with vehicle or 0.5 mM specific STAT3 peptide inhibitor for one hour, followed by vehicle or 0.5 µM entinostat treatment for 23 hours. Cells were harvested and analyzed by flow cytometry using fluorescence-conjugated antibodies specific to CD4 and Foxp3. Fixable live/dead dye was used to stain cells and live cells were gated. Cell culture was in absence (panel C) or in presence (panel D) of IL-2. In each condition, histograms show quantification of Foxp3 expression in Tregs by FACS analysis. Graph shows means, error bars represent standard deviations. In C, p = 0.0002 for entinostat vs. STAT3 inhibitor+entinostat. In D, p = 0.00015 for entinostat vs. STAT3 inhibitor+entinostat. Results are representative of three separate experiments.

Figure 7. Class I inhibition, not class II inhibition suppresses Foxp3 expression in Tregs.

Splenocytes were isolated from BALB/c mice and put in culture with different treatment conditions as indicated for 24 hours. Cells were harvested, stained for surface markers and intracellular Foxp3, and subjected to FACS analysis. Plot gating and parameter indication were described in Figure 1. Histograms show quantification of Foxp3 levels in Tregs. A, Both class I HDAC inhibitors, entinostat and MGCD0103, inhibit Foxp3 expression in Tregs. When comparing entinostat with MGCD0103, PE conjugated anti-Foxp3 antibody was used. B, Class I HDAC inhibitor, entinostat, and a pan inhibitor, panobinostat, but not class II HDAC inhibitors, MC1568 and MC1575, inhibit Foxp3 expression in Tregs. When comparing entinostat with class II HDAC inhibitors and the pan inhibitor, panobinostat, the Alexa 700 conjugated anti-Foxp3 antibody was used since other fluorochrome channels are interfered by the autofluorescence of the class II HDAC inhibitors. ** represents p<0.01 for marked treatment vs. vehicle.