Histone Deacetylase Inhibitor SAHA Is a Promising Treatment of Cushing Disease

Abstract

Context: Remission failure following transsphenoidal surgery in Cushing disease (CD) from pituitary corticotroph tumors (CtTs) remains clinically challenging. Histone deacetylase inhibitors (HDACis) are antitumor drugs approved for clinical use, with the potential to affect adrenocorticotropin hormone (ACTH) hypersecretion by inhibiting pro-opiomelanocortin (POMC) transcription.

Objective: Testing the efficacy of suberoylanilide hydroxamic acid (SAHA) on human and murine ACTH-secreting tumor (AtT-20) cells.

Design: Cell viability, ACTH secretion (enzyme-linked immunosorbent assay), apoptosis, and gene expression profile were investigated on AtT-20 cells. In vivo efficacy was examined in an athymic nude mouse AtT-20 xenograft model. SAHA efficacy against human-derived corticotroph tumor (hCtT) (n = 8) was tested in vitro.

Setting: National Institutes of Health.

Intervention: SAHA (0.5 to 8 µM).

Main outcome measures: AtT-20 and hCtT cell survival, in vitro/invivo ACTH measurements.

Results: SAHA (1 µM) reduced AtT-20 viability to 75% at 24 hours, 43% at 48 hours (analysis of variance; P = 0.002). Apoptosis was confirmed with elevated BAX/Bcl2 ratio and FACS. Intriguingly, early (3-hour) significant decline (70%; P < 0.0001) of secreted ACTH and diminished POMC transcription was observed with SAHA (1 µM). Microarray analysis revealed a direct association between liver X receptor alpha (LXRα) and POMC expression. Accordingly, SAHA reduced LXRα in AtT-20 cells but not in normal murine corticotrophs. Xenografted nude-mice tumor involution (126 ± 33/160 ± 35 vs 337 ± 49 mm3; P = 0.0005) was observed with 5-day intraperitoneal SAHA, with reversal of elevated ACTH (P < 0.0001). SAHA did not affect serum ACTH in nontumor mice. Lastly, we confirmed that SAHA (1 µM/24 h) decreased hCtT survival (78.92%; P = 0.0007) and ACTH secretion (83.64%; P = 0.03).

Conclusion: Our findings demonstrate SAHA's efficacy in reducing survival and ACTH secretion in AtT-20 and hCtT cells, providing a potential intervention for recurrent/unremitting CD.

Trial registration: ClinicalTrials.gov NCT00060541.

Copyright © 2017 Endocrine Society

Figures

Figure 1.

SAHA reduced AtT-20 cell survival by apoptotic induction. Clinically achievable concentrations of SAHA (0.5 to 4 µM) significantly reduced the viability of transformed AtT-20 cells in concentration- and time-dependent manners following 3- to 72- hour treatment. (a) Time course of their survival, as assessed by MTT, was plotted. (b) Representative live microscopy images of AtT-20 cells exposed to various doses of SAHA for 24 hours. (c) FACS analysis revealed increased apoptosis in AtT-20 cells treated with SAHA (1 µM) for 24 hours. (d) Western blot analysis revealed an increase in cleaved PARP, a hallmark of apoptotic cell death, when AtT-20 cells were exposed for 24 hours to SAHA (0.5 to 4 µM). (e) SAHA (0.5 to 4 µM) resulted in an increase in the Bax/Bcl2 ratio at 24 hours in a dose-dependent manner in AtT-20 cells, suggesting that SAHA activates proapoptotic pathways. The Bax/Bcl2 ratio curve was generated from two independent experiments. (f) SAHA (1 µM) reduced cell survival at 24 hours in hCtT cells (n = 7). *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001 compared with corresponding control values. Horizontal bars represent mean ± standard deviation. DMSO, dimethyl sulfoxide.

Figure 2.

Before apoptotic death, SAHA evoked robust and sustained declines in ACTH release and POMC transcription in AtT-20 cells. (a) Early (within 3 hours) exposure to SAHA (4 µM) significantly decreased the survival of AtT-20 cells, as observed in MTT assays. Percentage changes in ACTH secretion by AtT-20 cells upon exposure to increasing concentrations of SAHA for (b) 3 hours and (c) 24 hours; ACTH levels significantly decreased by clinically achievable concentrations of SAHA (0.5 to 4 µM) in both time groups (b, c). Each sample was run in duplicate in every experiment, and each experiment was repeated three times for the ELISA assay at 3 and 24 hours (b, c). (d) Physiologically available concentrations of SAHA led to transcriptional downregulation of POMC in AtT-20 cells after 24-hour exposure. (e) As inferred by the luciferase transfection study, the aforementioned effect was likely caused by suppression of the POMC promoter. (f) Expectedly, SAHA led to a significant decrease in ACTH levels after 24 hours of exposure on hCtT cells. *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001 compared with corresponding control values. Horizontal bars represent mean ± standard deviation.

Figure 3.

SAHA exerts pleiotropic effects on AtT-20 cell gene expression. Following SAHA (1 µM) treatment, the gene expression profile of AtT-20 cells was investigated by whole-genome microarray analysis. (a) Unpaired t test comparison of microarray data identified 2657 DEGs that reached the preselected statistical threshold and are depicted in a volcano plot. (b) From our list of DEGs, 34 KEGG transduction pathways were found to be affected by SAHA. (c) Functional analyses of DEGs with the DAVID resources revealed that SAHA altered the expression of two relevant gene clusters: nuclear hormone receptor and apoptosis; a heat map illustrates the differential expression patterns of these two clusters when AtT-20 cells were exposed to different HDACis. *Bonferroni adjusted. cGMP-PKG, cGMP-dependent protein kinase; HTLV, human T-lymphotropic virus; KEGG, Kyoto Encyclopedia of Genes and Genomes; PPAR, peroxisome proliferator-activated receptors.

Figure 4.

Selective degradation and transcriptional downregulation of LXRα by SAHA occurred in tumors but not in normal corticotrophs. (a) After 3 hours of exposure to increasing doses of SAHA (0.5 to 8 µM), AtT-20 cells displayed decreasing LXRα protein levels in a dose-dependent fashion. (b) RXRα protein levels, however, did not seem to be affected by SAHA. (c) Because of decreased levels of LXRα protein, the effects of SAHA were also assessed on the formation of LXRα/RXRα heterodimers using protein-complex immunoprecipitation methods. LXRα was immunoprecipitated and subjected to immunoblotting with an anti-RXRα antibody. A reduction in heterodimerization was observed with increasing doses of SAHA (0.5 to 8 µM) after 3 hours. (d, e) Furthermore, to evaluate the correlation between protein and messenger levels, increasing doses of SAHA were used on AtT-20 cells, and fold-change message levels were measured by qRT-PCR for (d) LXRα and (e) RXRα. (d) A significant reduction in LXRα messenger RNA (mRNA) fold-change after 3 hours of SAHA treatment (0.5 to 8 µM) was demonstrated. (e) Although message levels for RXRα seem to be reduced with SAHA, statistical significance was not reached. (f) To validate the limited action of SAHA on normal cells, LXRα and RXRα protein levels were also measured in corticotroph cells harvested from healthy nude-mice. Flow-cytometric isolation of anti–corticotropin-releasing hormone receptor 1–positive cells, as shown by the double peak, were subjected to a single dose of SAHA (1 µM) for 3 hours; consequently, no change in LXRα and RXRα protein expression was observed. The results are representative of two independent experiments. ***P ≤ 0.001; ****P ≤ 0.0001 compared with corresponding control values. Horizontal bars represent mean ± standard deviation. x-Axis on densitometric and polymerase chain reaction analyses: Con, control: 0.5 to 0.5 µM, 1 to 1 µM, 2 to 2 µM, 4 to 4 µM, and 8 to 8 µM. IP, immunoprecipitation.

Figure 5.

SAHA arrested tumor growth and reversed ACTH hypersecretion in vivo. (a) SAHA was effective in abrogating tumor growth in a nude-mice AtT-20 xenograft model. (b) Tumor volume was similarly suppressed up to 70% with SAHA (50 mg/kg). (c) In addition, plasma ACTH levels were quantified following 5 days of SAHA treatment; ACTH levels significantly decreased in both treatment groups (25 and 50 mg/kg). (d) Although adrenal gland dimensions were greater in tumor-implanted animals, statistical significance was not reached compared with treatment animals. (e) Similarly, reversal of skin thinning in grafted mice was nonstatistically significant with SAHA treatment. (f) Pictures of skin thickness illustrate cutaneous thinning in tumor groups (scale bars = 10 µm). *P ≤ 0.05, **P ≤ 0.01, ****P ≤ 0.0001, compared with corresponding control values. Horizontal bars represent mean ± standard deviation.

Figure 6.

Proposed mechanism for suppression of POMC transcription by SAHA. Schematic representation of the effects of SAHA on LXRα leading to transcriptional downregulation of POMC. By selectively degrading LXRα in tumor cells, heterodimerization with RXRα at the LXR response element 1 site was decreased; therefore, downstream transcription was reduced.