The histone deacetylase inhibitors vorinostat and romidepsin downmodulate IL-10 expression in cutaneous T-cell lymphoma cells

Abstract

Background and purpose: Vorinostat and romidepsin are histone deacetylase inhibitors (HDI), approved for the treatment of cutaneous T-cell lymphoma (CTCL). However, the mechanism(s) by which these drugs exert their anti-cancer effects are not fully understood. Since CTCL is associated with immune dysregulation, we investigated whether these HDI modulated cytokine expression in CTCL cells.

Experimental approach: CTCL cell lines and primary CTCL cells were treated in vitro with vorinostat or romidepsin, or with STAT3 pathway inhibitors. Cell cycle parameters and apoptosis were analysed by propidium iodide and annexin V/propidium iodide staining respectively. Cytokine expression was analysed using QRT-PCR and elisa assays. STAT3 expression/phosphorylation and transcriptional activity were analysed using immunoblotting and transfection/reporter assays respectively.

Key results: Vorinostat and romidepsin strongly down-regulated expression of the immunosuppressive cytokine, interleukin (IL)-10, frequently overexpressed in CTCL, at both the RNA and protein level in CTCL cell lines and at the RNA level in primary CTCL cells. Vorinostat and romidepsin also increased expression of IFNG RNA and decreased expression of IL-2 and IL-4 RNA, although to a lesser extent compared to IL-10. Transient exposure to vorinostat was sufficient to suppress IL-10 secretion but was not sufficient to irreversibly commit cells to undergo cell death. STAT3 pathway inhibitors decreased production of IL-10 and vorinostat/romidepsin partially decreased STAT3-dependent transcription without effects on STAT3 expression or phosphorylation.

Conclusions and implications: These results demonstrate that HDI modulate cytokine expression in CTCL cells, potentially via effects on STAT3. Immunomodulation may contribute to the clinical activity of HDI in this disease.

Figure 1

Histone deacetylase inhibitor-induced histone acetylation. HUT78 cells were treated with the indicated concentrations of vorinostat or romidepsin and the levels of acetylated histone H4 (K12) and total histone H4 were analysed by immunoblotting. Results shown are representative of at least three experiments.

Figure 2

Effect of histone deacetylase inhibitors on cell cycle parameters in HUT78 cells. A. Representative cell cycle profiles showing HUT78 cells treated with DMSO or vorinostat (1 µM or 2.5 µM). B,C. HUT78 cells were treated with the indicated concentrations of vorinostat (B) or romidepsin (C) for 24 h. Cells were treated with DMSO as control. The graphs show the proportion of cells in different phases of the cell cycle determined by flow cytometric analysis of propidium iodide-stained cells. Graphs show mean (±SD) values derived from at least three independent experiments, each performed in duplicate. Statistically significant differences compared to DMSO-treated cells are shown (*P≤ 0.05, **P≤ 0.01, ***P≤ 0.001).

Figure 3

Effect of histone deacetylase inhibitors on apoptosis in HUT78 cells. HUT78 cells were treated with the indicated concentrations of vorinostat (A) or romidepsin (B). After 24 h, induction of apoptosis was analysed by flow cytometric analysis of propidium iodide (PI) and annexin (Ann) V-stained cells. Graphs show mean (±SD) values derived from at least three independent experiments, each performed in duplicate. Statistically significant differences compared to DMSO-treated cells are shown (*P≤ 0.05, **P≤ 0.01, ***P≤ 0.001).

Figure 4

Effect of histone deacetylase inhibitors on cytokine and IL-12RB1/B2 RNA expression in cutaneous T-cell lymphoma cells. A–I. HUT78 or (J) SeAx cells were treated with the indicated concentrations of vorinostat (Vor; µM), romidepsin (Rom; nM) or DMSO (D) as a control. After the indicated times (A) IFNG (B) TNF (C) IL-2 (D) IL-4 (E) IL-5 (F, J) IL-10 (G) IL-13 (H) IL-12RB1 and (I) IL-12RB2 RNA expression was analysed by QRT-PCR. Data are the means (±SD) derived from two to five separate experiments. Statistically significant differences compared to DMSO-treated cells are shown (*P≤ 0.05, **P≤ 0.01, ***P≤ 0.001).

Figure 5

Effect of histone deacetylase inhibitors on IL-10 RNA expression in primary cutaneous T-cell lymphoma (CTCL) cells. A,B. Primary CTCL cells derived from two patients were treated with the indicated concentrations of vorinostat (Vor; µM), romidepsin (Rom; nM) or DMSO (D) as a control. After the indicated times, IL-10 RNA expression was analysed by QRT-PCR. Data are means (±SD) of duplicate determinations.

Figure 6

Effect of histone deacetylase inhibitors (HDI) on IL-10 secretion in cutaneous T-cell lymphoma cells. A. HUT78 cells were pretreated with vorinostat or romidepsin, washed to remove drug and resuspended in fresh HDI-supplemented media for a further 3.5 (vorinostat) or 3 h (romidepsin). IL-10 secretion was analysed by elisa. Data are means (±SD) of at least three separate experiments each performed in duplicate. B. IL-10 secretion from SeAx cells treated with the indicated concentrations of romidepsin or DMSO (D) for 24 h (elisa). Data are mean (±SD) of duplicate determinations. C,D. Washout experiments. HUT78 cells were treated with DMSO, vorinostat (Vor; 3 µM) or left untreated (UT) for 24 h. Cells were treated with vorinostat for 6 h (Vor/6 h), washed thoroughly and analysed at 24 h. C. Sub-G1 DNA content. D. IL-10 elisa. Data are means (±SD) derived from four experiments. Statistically significant differences compared to DMSO-treated cells are shown (*P≤ 0.05, **P≤ 0.01, ***P≤ 0.001).

Figure 7

STAT3-dependent expression of IL-10 expression in HUT78 cells. HUT78 cells were treated with the indicated concentrations of the STAT3 pathway inhibitors [stattic, cucurbitacin I (cucurb) or WP1066] for 8 h. DMSO was used as a solvent control. In (A), expression of total, serine- and tyrosine-phosphorylated STAT3 was analysed by immunoblotting. HSC70 was analysed as a loading control. Results are representative of two independent experiments. Open arrowheads indicate a slower migrating isoform detected in cucurbitacin I-treated cells. In (B), supernatants were collected (D, DMSO) and levels of IL-10 analysed by elisa. Data are means (±SD) from two separate experiments each performed in duplicate. Statistically significant differences compared to DMSO-treated cells are shown (*P≤ 0.05).

Figure 8

Histone deacetylase inhibitors (HDI) decrease STAT3 activity. A. HUT78 cells were treated with the indicated concentrations of vorinostat or romidepsin. After 3 (vorinostat) or 8 h (romidepsin), cells were collected and expression of total, tyrosine- and serine-phosphorylated STAT3 was analysed by immunoblotting. Data are from the same blots but with intervening lanes removed. HSC70, loading control. Results are representative of three separate experiments. B. HUT78 cells were transfected with STAT3-dependent firefly luciferase and control Renilla luciferase plasmids. After 24 h, cells were treated with the indicated concentrations of HDI for 3 (vorinostat) or 8 h (romidepsin) and luciferase assays were performed. The graph shows firefly luciferase activity normalized to Renilla luciferase activity and set to 1.0 for DMSO-treated cells. Data are means (±SD) derived from three separate experiments. Statistically significant differences compared to DMSO-treated cells are shown (**P≤ 0.01, ***P≤ 0.001).