Down-regulation of HDAC3 inhibits growth of cholangiocarcinoma by inducing apoptosis

Class I histone deacetylases (HDACs) inhibit expression of tumor suppressor genes by removing acetyl groups from histone lysine residues, thereby increasing cancer cell survival and proliferation. We evaluated the expression of class I HDACs in cholangiocarcinoma (CCA). HDAC3 expression was specifically increased in CCA tissues and correlated with reduced patient survival. HDAC3 overexpression inhibited apoptosis and promoted CCA cell proliferation. Conversely, HDAC3 knockdown or pharmacological inhibition decreased CCA cell growth and increased caspase-dependent apoptosis. Inhibition of class I HDACs blocked HDAC3-catalyzed deacetylation and increased expression of downstream pro-apoptotic targets in vitro and in vivo. These results demonstrate for the first time that down-regulation of HDAC3 induces apoptosis in human CCA cells, indicating that inhibiting HDAC3 may be an effective therapeutic strategy for treating CCA .


INTRODUCTION
Cholangiocarcinoma (CCA) is a highly malignant adenocarcinoma of mostly unknown etiology, with increasing mortality in many countries [1][2][3]. Most patients are diagnosed at late stages and are not eligible for surgical treatment. As a result, 5-year survival rates of CCA have remained at 10% for the past three decades [4]. Even though chemotherapy has been effective in some CCA patients, many patients are not responsive to conventional chemotherapies, emphasizing the urgent need for novel therapeutic strategies for CCA [1][2][3][4].

Research Paper
HDACs inhibit specific tumor suppressor genes, resulting in an aberrant epigenetic status of cancer cells compared to healthy cells, they may serve as candidate anti-cancer targets [12,13]. Increased expression and activation of HDAC3 play a critical role in epigenetic alterations associated with different types of malignancies [14,15]. However, the role of HDAC3 in CCA has not yet been elucidated.
HDACs inhibitors induce cell cycle arrest and apoptosis in a broad spectrum of cancer cells [16][17][18][19]. Vorinostat (suberoylanilide hydroxamic acid, SAHA) and romidepsin have been approved by the U.S. Food and Drug Administration (FDA) for the treatment of cutaneous T-cell lymphoma [20]. Considering their high anticancer activity in hematological malignancies, it is anticipated that they will find applications for the treatment of solid cancers as well. Novel class I HDACs inhibitors (4SC202, BG45, SBHA) have exhibited an anti-cancer activity both in vitro and in vivo, with marginal toxicity [5,21,22]. However, the effects of class I HDACs inhibitors in CCA have not yet been studied.
Here we show that high protein levels of HDAC3 in CCA tissues are associated with poor survival in patients with CCA. Increased expression of HDAC3 induces proliferation and inhibits apoptosis in CCA cells. Down-regulation of HDAC3 induces apoptosis of CCA cells, resulting in a reduced CCA growth. Together, our findings indicate that HDAC3 induces CCA progression by promoting cell proliferation, and suggest that it may serve as a potential target for therapeutic intervention in the treatment of CCA.

HDAC3 expression is increased in CCA tissues, and associated with reduced patient survival
We used CCA tissues from the Biobank of Nanjing Drum Tower Hospital, which contains clinically annotated data from 60 CCA samples; the clinical characteristics of the study participants are summarized in Table 1 . Using immunohistochemistry (IHC), we found that there was no difference in the expression of HDAC1, HDAC2, or HDAC8 isoenzymes between CCA tissues and their adjacent tissues ( Figure 1A & 1B). Based on the illustrated frequency distribution, there was no difference between the high and low HDAC3 groups with respect to age, sex, histological differentiation grade, tumor size, nodal metastasis, or pathological stage (Table 1). However, when we assessed the expression of HDAC3, we found that it was significantly increased in CCA tissues compared to adjacent tissues ( Figure 1A & 1B). Importantly, the increased HDAC3 expression was associated with a reduced patient survival, whereas other class I HDACs had no correlation with survival ( Figure 1C). These findings indicate that an increased HDAC3 expression in CCA tissues is an independent predictor of a poor prognosis in CCA patients.

HDAC3 promotes growth of CCA cells
High levels of HDAC3 expression and activity play a critical role in cell epigenetic alterations associated with malignancies [14,15]. However, the role of HDAC3 in CCA has not been elucidated. We assessed the expression of HDAC3 in different CCA cell lines and fresh tissues by western blotting. HDAC3 protein levels were increased in CCA tissues compared to normal tissues (Figure 2A), and all six CCA cell lines expressed high protein levels of HDAC3 ( Figure 2B).

HDAC3 inhibition induces CCA cell apoptosis in vitro
Flow cytometry was employed to investigate the mechanism of 4SC202 inhibition of CCA cell proliferation. 4SC202 increased apoptosis in two different CCA cell types ( Figure 4A-4C). Next, we investigated the effect of HDAC3 inhibition on the downstream targets, including acetylated α-Histone3, P53 and Bax [23]. HDAC3 knockdown increased caspase substrate (polyADP ribose polymerase (PARP) and caspase 3) cleavage and P53 expression; this effect was mimicked by 4SC202 treatment (Figure 4D & 4E). These results indicate that the apoptosis of CCA cells is primarily induced by HDAC3 inhibition, and that HDAC3 may be the dominant target of 4SC202.

HDAC3 is the main target in 4SC202-induced apoptosis in CCA cells
To determine if the pro-apoptotic effect of 4SC202 in CCA cells was due to HDAC3 inhibition, CCA cells were treated with 4SC202 following transfection with HA-tagged HDAC3 vector. HDAC3 protein levels were unchanged by 4SC202 treatment ( Figure 5A & 5B), suggesting that 4SC202 inhibits the activity of HDAC3 directly. Due to the high level of homology between the class I HDACs (HDAC 2 shares 52% identity with HDAC3 [24][25][26]), 4SC202 possibly has a weak inhibitory effect also on HDAC1 and 2 [27]. We therefore sought to determine whether HDAC3 was responsible for the 4SC202-induced apoptosis. To elucidate the direct target of 4SC202, we used an in vitro deacetylation system ( Figure 5C). 4SC202 treatment inhibited HDAC3 deacetylation activity, but only had a marginal inhibitory effect on HDAC1 and 2 ( Figure 5E). We examined the effects of HDACs 1, 2 and 3 on apoptosis related targets and found that only HDAC3 could rescue apoptosis in CCA cell lines ( Figure 5F-5H). These results demonstrate that HDAC3 is the main target of 4SC202 in CCA cell apoptosis.

HDAC3 inhibition induces apoptosis and suppresses cell proliferation in CCA tumor xenografts
In order to evaluate the in vivo anti-cancer effects of HDAC3 inhibition, we employed a CCA tumor xenograft model and found that HDAC3 knockdown cells also showed a low proliferative capacity and tumorigenicity compared to their counterparts ( Figure 6A-6C). 4SC202 administration significantly inhibited tumor growth ( Figure 6D & 6E). The body weights of treated mice were   used as indicators of health [28]. 4SC202 treatment did not affect mouse body weight, which indicated that the mice did not experience evident toxicity in vivo ( Figure 6F).
Histological sections of xenograft tumors were analyzed by TUNEL assay, and stained with antibodies against c-caspase 3 and Ki-67, markers of cell apoptosis and proliferation, respectively [28]. Consistent with the in vitro results, 4SC202 administration increased TUNEL and c-caspase 3 staining and reduced Ki-67 staining in xenograft tissues, confirming the anti-tumor effect of 4SC202 ( Figure 6I). HDAC3 can shuttle in and out of the nucleus, whereas other Class I HDACs are found primarily in the nucleus, as the HDAC3 catalytic domain is positioned much closer to the C-terminus than other Class I HDACs [25]. We found that HDAC3 was mainly localized to the nucleus, but was also observed in the membrane. 4SC202 treatment did not significantly change the location and protein level of HDAC3 in CCA xenograft samples ( Figure 6J). These results demonstrate the pro-apoptotic and anti-proliferative effect of HDAC3 inhibition in vivo.

DISCUSSION
Transcriptional repression by class I HDACs (1, 2, 3 and 8) plays a physiological role via the maintenance of cell proliferation as well as the inhibition of specific tumor suppressor genes, thereby resulting in aberrant epigenetics in cancer cells [12,13,29,30]. Since genetic alterations and epigenetic changes are reversible, class I HDACs are suitable for pharmacological intervention [7][8][9][10][11]. Among them, the modifications in HDAC8 expression do not affect cancer cell proliferation [31,32], and the expression of other class I HDACs in CCA has not been studied. Therefore, we set out to determine the expression of HDAC1, 2, 3 and 8 in CCA tissues. We found that the expression of HDAC1, 2 and 8 showed no significant differences between CCA tissues and adjacent healthy tissues, while HDAC3 expression was increased in CCA tissues and correlated with clinicopathological factors in CCA.
Determining the impact of overexpressed HDAC3 on CCA cells is also of interest, given the important role of epigenetics in carcinogenesis. Acetylation increases p53 protein stability and upon acetylation of p53 at K120, p53 preferentially activates the expression of proapoptotic genes BAX, PUMA, DR5 and NOXA [33]. CCA cells with high HDAC3 expression were found to be resistant to p53-induced apoptosis. The inability of HDAC1 and 2 to activate p53 in these cells indicates that high HDAC3 expression leads to the packaging of the p53 promoter into a highly repressed state. Suppression of HDAC3 in these cells rendered them responsive to p53-induced apoptosis, growth inhibition, and activation of the downstream proapoptotic gene BAX. Increased HDAC3 expression might therefore be an important mechanism for facilitating cancer cell proliferation in CCA. In fact, selective HDAC3 knockdown induced apoptosis and inhibited proliferation of CCA cells. Consistent with these results, HDAC3 overexpression reversed p53-induced apoptosis, while overexpression of HDAC1 and HDAC2 failed to mimic the HDAC3-like effects.
Novel class I HDACs inhibitors (4SC202, BG45, SBHA) were found to be beneficial for cancer cells with good tolerance [5,21,22]. Our results suggest that class I HDACs inhibitors negate the growth-promoting function of HDAC3 by suppressing HDAC3 activity and its downstream targets. Because of its lower IC50 to CCA cells in vitro, 4SC202 may therefore be an effective agent for preventing CCA carcinogenesis. HDAC 1 and HDAC 2 share 82% identity with each other, while they share 53% and 52% identity with HDAC3, respectively [24][25][26]. Due to the high level of homology among the class I HDACs, it is easy to comprehend why an HDAC3 selective inhibitor would be difficult to identify. Though 4SC202 can show selectivity for HDAC1, 2 and 3 [34], its inhibitory effects on other HDACs besides HDAC3 cannot be ignored. Therefore, we evaluated the inhibitory effect of 4SC202 on HDACs 2 and 3 by employing mass spectrometry, and confirmed 4SC202 could inhibit HDAC3 activity in vitro. At the molecular level, HDAC3 not only partially reversed apoptosis, but also reversed expression of apoptosis-related proteins, whereas HDAC1 and HDAC2 did not have this ability. Consistent with in vitro data, 4SC202 significantly inhibited the in vivo activity of HDAC3, and induced apoptosis in HuCCT1 xenograft tissues. This data suggests that 4SC202 induces CCA cell apoptosis by mainly targeting HDAC3.
As the catalytic domain of HDAC3 is positioned much closer to the C-terminus than other class I HDACs, Xenograft samples were stained with HDAC3 (left) and staining was quantified (right). Data represent the mean ± SEM, n≥3. * p<0.05, ** p<0.01, NS not significant. the structure of HDAC3 is distinct from other class I HDACs [25]. This may explain why the HDAC3 protein can shuttle in and out of the nucleus, whereas other class I HDACs are found primarily in the nucleus [25]. HDAC3 phosphorylation, which is regulated by the kinase c-Src, casein kinase-2, and by the phosphatase PP4, regulates HDAC3 activity and nuclear localization [35,36]. We found that HDACs inhibitor treatment did not significantly change the cellular localization or protein levels of HDAC3 in CCA cells and xenograft samples, indicating that the inhibition of HDAC3 activity, instead of its expression and phosphorylation, contributed to the anti-tumor effect of the inhibitor.
In conclusion, this study shows that HDAC3 is a key factor inducing CCA cell proliferation and inhibiting apoptosis, and that increased HDAC3 expression correlates with a poor prognosis in CCA patients. Class I HDAC inhibitors represent a novel treatment approach for CCA. A better understanding of their function in CCA cells is needed to establish their role in the management of CCA.

Ethics
All experiments were approved by the Ethical Committee of Medical Research, Nanjing Drum Tower Hospital, Affiliated Hospital of Nanjing University Medical School. All animal experiments conformed to protocols approved by animal care and use committees at Nanjing University.

Immunohistochemistry (IHC)
Tumor specimens were fixed in 4% formalin and embedded in paraffin. Two to five human CCA tumor specimens from one patient were used for the IHC study.

Western blotting analysis
Cells were lysed with 0.5% NP40 lysis buffer and western blotting was performed according to the standard protocol. Detection was accomplished with the chemiluminescence ECL plus reagent (Thermo, Grand Island, NY, USA) and the chemiluminescence HRP substrate (Millipore, Billerica, MA, USA), and signals were evaluated by a Tanon 5200Multi scanner (Shanghai, China). Primary antibodies were as follows: HDAC1

Cell viability and clonogenic assay
Cell viability was assessed by the CCK-8 colorimetric assay in 96-well plates (2×10 3 cells/well) (Dijindo, Minato-ku, Tokyo, Japan). The absorbance at 450 nm was recorded using a micro-plate reader. For the clonogenic assay, cells were seeded into 6-well plates at a density of 500 cells/well. Following culture for 10 days, individual colonies were counted after crystal violet staining. www.impactjournals.com/oncotarget

Apoptosis assay
Cells were analyzed for apoptosis with the AnnexinV-FITC/PI Apoptosis Detection Kit (BD, Franklin Lakes, NJ, USA) following the manufacturer's instructions.

Cholangiocarcinoma cancer xenograft model
Nude mice were purchased from the Department of Laboratory Animal Science, Nanjing Drum Tower Hospital. HuCCT1 cells (5×10 6 ) in FBS-free RPMI-1640 were subcutaneously injected into the flanks of mice. When tumors were measurable, mice were treated with intraperitoneal injection (IP) of vehicle control or 4SC202 (50 mg/kg) in 200 μl volume twice a week for three weeks. HDAC3 knockdown HuCCT1 cells and their counterparts were injected into the flanks of the same mice. Tumor volume was calculated using the formula, length (L) x width (W) x height (H) x 0.5236. The Animal Welfare Committee of Nanjing Drum Tower Hospital approved all procedures involving animals.

Statistics
Data are expressed as means ± standard error of the mean (SE). The data were analyzed through one-way ANOVAs followed by post hoc Duncan tests (SPSS 17.0). P<0.05 was considered significant.