The Ca2+-activated Cl− channel transmembrane proteins with unknown function 16 A (TMEM16A; also known as anoctamin 1 or discovered on gastrointestinal stromal tumor 1) plays an important role in facilitating the cell growth and metastasis of TMEM16A-expressing cancer cells. Histone deacetylase (HDAC) inhibitors (HDACi) are useful agents for cancer therapy, but it remains unclear whether ion channels are epigenetically regulated by them. Using real-time polymerase chain reaction, Western blot analysis, and whole-cell patch-clamp assays, we found a significant decrease in TMEM16A expression and its functional activity was induced by the vorinostat, a pan-HDACi in TMEM16A-expressing human cancer cell lines, the prostatic cancer cell line PC-3, and the breast cancer cell line YMB-1. TMEM16A downregulation was not induced by the chemotherapy drug paclitaxel in either cell type. Pharmacologic blockade of HDAC3 by 1 μM T247 [N-(2-aminophenyl)-4-[1-(2-thiophen-3-ylethyl)-1H-,,triazol-4-yl]benzamide], a HDAC3-selective HDACi, elicited a large decrease in TMEM16A expression and functional activity in both cell types, and pharmacologic blockade of HDAC2 by AATB [4-(acetylamino)-N-[2-amino-5-(2-thienyl)phenyl]-benzamide; 300 nM] elicited partial inhibition of TMEM16A expression (∼40%) in both. Pharmacologic blockade of HDAC1 or HDAC6 did not elicit any significant change in TMEM16A expression, respectively. In addition, inhibition of HDAC3 induced by small interfering RNA elicited a large decrease in TMEM16A transcripts in both cell types. Taken together, in malignancies with a frequent gene amplification of TMEM16A, HDAC3 inhibition may exert suppressive effects on cancer cell viability via downregulation of TMEM16A.
The balance of histone acetyltransferase (HAT) and HDAC activity is important in the maintenance of gene regulation (Narlikar et al., 2002). In various disease states, including cancer, this balance is altered, generally leading to a decrease in gene transcription. HDAC inhibitors (HDACis) are involved in the progression of mitosis through alterations in chromatin acetylation and heat-shock protein expression (Li et al., 2013). There are currently four classes of HDACis, and some of them have been shown to possess potent anticancer activity. A small-molecule inhibitor of class I and II HDACs, vorinostat (also known as suberanilohydroxamic acid), is approved for the treatment of cutaneous T-cell lymphoma in the United States and is also being developed for other solid tumors (Marks, 2007; Iwamoto et al., 2013). However, pan-HDACis, such as vorinostat, have poor selectivity and high toxicity, so a selective HDACi with improved efficacy and reduced toxicity would be expected to have therapeutic activity and clinical utility in a variety of cancers.
Ion channels contribute to a variety of cancer processes, such as proliferation, apoptosis, migration, and invasion, via regulation of the resting membrane potential and intracellular Ca2+ signaling (Yang and Brackenbury, 2013; Lang and Stournaras, 2014; Pardo and Stühmer, 2014). Pharmacologic blockade of ion channels is an attractive target to both suppress tumor growth and prevent tumor metastasis. Transmembrane protein with unknown function 16 A (TMEM16A; also known as anoctamin 1 or discovered on gastrointestinal stromal tumor 1) was identified as a pore-forming subunit of the Ca2+-activated Cl− channel, with eight putative transmembrane domains and cytosolic N- and C-termini (Caputo et al., 2008; Schroeder et al., 2008; Yang et al., 2008), and functions as a dimer (Sheridan et al., 2011; Ohshiro et al., 2014). TMEM16A is located on chromosome 11q13, which is frequently amplified in many types of human cancer (Akervall et al., 1995; Huang et al., 2002), and plays an important role in driving the amplification of 11q13 (Komatsu et al., 2006). Overexpression of TMEM16A in human cancer is associated with the facilitation of tumor growth and migration, and TMEM16A especially contributes to the tumorigenesis that occurs in head-and-neck squamous cell carcinoma (Ayoub et al., 2010; Duvvuri et al., 2012; Ruiz et al., 2012; Jacobsen et al., 2013). Small interfering RNA (siRNA)–based downregulation and pharmacologic blockade of TMEM16A decrease the proliferation of breast and prostate cancer cells, thus TMEM16A is an attractive therapeutic target and novel biomarker for both of these cancers (Liu et al., 2012; Britschgi et al., 2013). HDACs are also considered to be therapeutic targets for both cancers (Ellis et al., 2009; Munster et al., 2009), but it remains elusive whether TMEM16A is epigenetically-regulated by HDACis in TMEM16A-expressing cancer cells.
Prostate and breast cancers are the most common types of cancer in men and women, respectively. In the present study, we examined the epigenetic regulation of TMEM16A using the pan-HDACi vorinostat in the TMEM16A-expressing human prostatic cancer cell line PC-3 and the breast cancer cell line YMB-1, demonstrating that the downregulation of TMEM16A expression/function underlies the inhibition of cell proliferation in TMEM16A-expressing cancer cells. We further identified the HDAC subtype(s) involved in TMEM16A downregulation by pharmacologic HDAC blockade and siRNA-induced HDAC downregulation.
Materials and Methods
The prostate cancer cell lines PC-3 and LNCaP (clone FGC) were supplied from RIKEN BRC (Tsukuba, Japan) through the National Bio-Resource Project of the Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan. The breast cancer cell lines YMB-1 (with the same DNA profile as ZR-75-1), MCF-7, Hs578T-Luc, and BT-549 were supplied from Health Science Research Resources Bank (Osaka, Japan). They were maintained at 37°C in 5% CO2 with RPMI 1640 medium or Dulbecco’s modified Eagle’s medium (Wako Pure Chemical Industries, Osaka, Japan) containing 10% fetal bovine serum (Sigma-Aldrich, St. Louis, MO) and a penicillin (100 units/ml)–streptomycin (0.1 mg/ml) mixture (Wako Pure Chemical).
WST-1 Cell Viability Assay.
A cell-viability assay using WST-1 [2-(4-iodophenyl)-3-(4-nitrophenyl)-5-(2,4-disulfophenyl)-2H-tetrazolium, monosodium salt] was performed according to Dojindo’s suggested protocol (Dojindo, Kumamoto, Japan). Briefly, at a density of 4 × 105 cells/ml, cells were cultured in duplicate in 96-well plates for 0 to 2 days. At 2 hours after the addition of WST-1 reagent into each well, the absorbance was measured in a Bio-Rad microplate reader model 3550 (Bio-Rad Laboratories, Hercules, CA) at a test wavelength of 450 nm, with a reference wavelength of 630 nm. A pair of control and treated samples was prepared from different passage cells, then the same protocol was repeated on another day.
RNA Extraction, Reverse Transcription, and Real-Time Polymerase Chain Reaction.
Total RNA from normal human and tumor tissues was purchased from Clontech Laboratories (Palo, CA), BD Biosciences (San Jose, CA), and BioChain Institute (Hayward, CA). Total RNA extraction from cell lines and reverse transcription were performed as previously reported elsewhere (Ohya et al., 2011). The resulting cDNA products were amplified with gene-specific primers, designated using Primer Express software (v1.5; Applied Biosystems, Foster City, CA).
Quantitative, real-time polymerase chain reaction (PCR) was performed with the use of SYBR Green chemistry on an ABI 7500 sequence detector system (Applied Biosystems) as previously reported elsewhere (Ohya et al., 2013). The transcriptional expression levels were determined 24 hours after compound treatment. The following PCR primers of human origin were used for real-time PCR: TMEM16A (GenBank accession number: NM_018034, 1578–1698), 121 bp; TMEM16B (NM_020373, 2651–2778), 128 bp; histone deacetylase 1 (HDAC1) (NM_004964, 708–824), 117 bp; HDAC2 (NM_001527, 298–405), 108 bp; HDAC3 (NM_003883, 699–819), 121 bp; HDAC4 (NM_006037, 3446–3585), 140 bp; HDAC5 (NM_005474, 3339–3467), 129 bp; HDAC6 (NM_006044, 3517–3637), 121 bp; HDAC7 (NM_015401, 2313–2434, 122 bp); HDAC8 (AF245664, 598–718, 121 bp); HDAC9 (NM_058176, 2252–2374), 123 bp; HDAC10 (CU013303, 1271–1420, 150 bp); HDAC11 (NM_024827, 720–840), 121 bp. β-Actin (ACTB) (NM_001101, 411–511), amplicon = 101 bp. Regression analyses of the mean values of three multiplex real-time PCRs for log10 diluted cDNA were used to generate standard curves. Unknown quantities relative to the standard curve for a particular set of primers were calculated, yielding a transcriptional quantitation of gene products relative to the endogenous standard ACTB. To confirm the nucleotide sequences, the amplified PCR products were sequenced with an ABI PRIZM 3100 genetic analyzer (Applied Biosystems).
Western Blot Analysis.
Proteins from the cell lysates was prepared from PC-3 and YMB-1 cell lines as previously reported elsewhere (Ohya et al., 2013). The protein expression levels were determined 48 hours after compound treatment. Equal amounts of protein (50 µg/lane) were subjected to SDS-PAGE (10%). Blots were incubated with anti-TMEM16A (Abcam, Cambridge, UK) and anti–β-actin (Medical & Biologic Laboratories, Nagoya, Japan) antibodies, then incubated with anti-rabbit and anti-mouse horseradish peroxidase–conjugated IgG (Millipore, Billerica, MA), respectively. An enhanced chemiluminescence detection system (GE Healthcare Japan, Tokyo, Japan) was used for the detection of the bound antibody. The resulting images were analyzed by a LAS-3000mini device (FujiFilm, Tokyo, Japan). A pair of control and treated samples was prepared from different passage cells, and then the same protocol was repeated on another day.
Downregulation of HDACs and TMEM16A by RNA Interference.
Downregulation of HDACs and TMEM16A by siRNA was performed as reported previously elsewhere (Yamazaki et al., 2006). Lipofectamine RNAiMAX reagent (Invitrogen/Life Technologies, Carlsbad, CA) was used for all siRNA transfection procedures. Commercially available siRNA oligonucleotides against human HDAC1 (sc-29343), HDAC2 (sc-29345), HDAC3 (sc-35538), and TMEM16A (sc-76686) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). As a control, control siRNA (type A; Santa Cruz Biotechnology) was used. The amount of siRNA used was optimized by measuring the transfection efficacy of fluorescein-conjugated control siRNA (over 80%, as determined by flow cytometry) and was determined to be a final concentration of 33 nM. At 24 hours after transfection, the expression levels of HDAC and TMEM16A transcripts were assessed by real-time PCR assay. At 48 hours after transfection of TMEM16A siRNA, cell viability was measured by WST-1 assay, as described earlier.
Whole-Cell Patch-Clamp Recordings.
Whole-cell patch-clamp experiments were performed as reported previously elsewhere (Hatano et al., 2013). The resistance of electrodes was 3–5 MΩ when they were filled with the pipette solution [in mM: 110 Cs-aspartate, 30 CsCl, 1 MgCl2, 10 HEPES, 10 EGTA, 2 Na2ATP (pH 7.2 by CsOH)]. The Ca2+ concentration was buffered by adding the appropriate amount of CaCl2 as determined by WINMAXC program (Stanford University, Stanford, CA). Membrane currents and voltage signals were digitized with a computer using an analog-digital converter (PCI6229; National Instruments Japan, Tokyo, Japan). Data acquisition and analysis of whole cell currents were performed using WinWCP4.65, developed by Dr. John Dempster (University of Strathclyde, UK). The liquid junction potential between the pipette and bath solutions (−10 mV) was corrected. Cells were held at −50 mV, and the currents were evoked by pipette solutions containing Ca2+ buffered at 1 μM. The current–voltage relationship of the activated current was determined by stepping the cell from −50 mV to potentials between −50 and +50 mV for 1 second. After depolarization, repolarization to −80 mV for 1 second elicited rapidly declining tail currents. A standard HEPES-buffered bathing solution (HEPES solution) was used, with the following composition (in mM): 137 NaCl, 10 CsCl, 5.9 KCl, 2.2 CaCl2, 1.2 MgCl2, 14 glucose, 10 HEPES (pH 7.4 by NaOH). All experiments were performed at 25 ± 1°C.
Statistical significance among two and multiple groups was evaluated using a paired t test and Tukey’s test after an F test or analysis of variance, respectively. A two-way analysis of variance test was applied to the data shown in Figs. 7 and 8. P < 0.05 and P < 0.01 were considered statistically significant, as indicated in the figures. Data are presented as the mean ± S.E.M.
The pharmacologic agents niflumic acid, paclitaxel, and bafilomycin-A were obtained from Sigma-Aldrich (Tokyo, Japan) and WST-1 from Dojindo (Kumamoto, Japan). The HDAC inhibitors vorinostat (N-hydroxy-N′-phenyl-octanediamide), AATB [4-(acetylamino)-N-[2-amino-5-(2-thienyl)phenyl]-benzamide], T247 [N-(2-aminophenyl)-4-[1-(2-thiophen-3-ylethyl)-1H-,,triazol-4-yl]benzamide], and NCT-14b [(S)-S-7-(adamant-1-ylamino)-6-(tert-butoxycarbonyl)-7-oxoheptyl-2-methylpropanethioate] were supplied by Professor Y. Suzuki (Kyoto Prefectural University of Medicine). The other agents were obtained from Sigma-Aldrich or Wako Pure Chemical Industries.
Effects of the Pan-HDAC Inhibitor Vorinostat on TMEM16A Expression in PC-3 and YMB-1 Cells.
TMEM16A is an attractive therapeutic target in both prostate and breast cancer (Liu et al., 2012; Britschgi et al., 2013). We first performed quantitative analysis of the TMEM16A transcripts by real-time PCR assay in human prostate and breast tissues (from three different donors for each), human prostate cancer cell lines (PC-3 and LNCaP), human breast cancer cell lines (YMB-1, MCF-7, Hs578T, and BT-549), and human breast tumor tissues (from three different donors) (Figs. 1A and 2A). The TMEM16A transcripts were highly expressed in the PC-3 and YMB-1 cell lines and breast tumor tissues compared with normal tissues and the other cell lines (Figs. 1A and 2A). On the other hand, the expression levels of the TMEM16B/anoctamin 2 transcripts were very low (less than 0.003 relative to ACTB [in arbitrary units]) in all the cell lines and tissues examined (Supplemental Fig. 1, A and B). We next examined the effects of vorinostat (0.1–10 µM) on PC-3 and YMB-1 cell viability. At 48 hours after the treatment with vehicle (dimethylsulfoxide [DMSO]), vorinostat, niflumic acid, control siRNA, and siTMEM16A siRNA (siTMEM16A), the relative cell viability was expressed as a ratio relative to the controls (vehicle and control siRNA). As shown in Figs. 1B and 2B, vorinostat treatment (1 and 10 μM) statistically significantly (P < 0.01) suppressed the cell viability in both types of cells in a concentration-dependent manner. Pharmacologic and siRNA-based blockade of TMEM16A by the Ca2+-activated Cl− channel inhibitor niflumic acid (100 μM) (Figs. 1C and 2C) and TMEM16A siRNA (Figs. 1D and 2D) also statistically significantly (P < 0.01) suppressed cell viability in both types of cells, respectively. Real-time PCR analysis showed that the transfection of TMEM16A siRNA significantly suppressed the expression level of TMEM16A transcripts in PC-3 (∼85% inhibition) and YMB-1 (∼50% inhibition). Quantitative real-time PCR analysis showed that vorinostat treatment markedly suppressed the expression levels of the TMEM16A transcripts in both cell lines in a concentration-dependent manner (Figs. 1E and 2E) without any effect on the expression level of the TMEM16B transcripts (data not shown). The expression levels of TMEM16A in arbitrary units were 0.046 ± 0.003, 0.014 ± 0.001, and 0.005 ± 0.001 in PC-3 cells treated with vehicle (0.1% DMSO), 1 μM vorinostat, and 10 μM vorinostat (n = 4 for each, P < 0.01 versus vehicle control; P < 0.05 versus 1 μM vorinostat treated), respectively, and 0.118 ± 0.006, 0.059 ± 0.004, and 0.021 ± 0.008 in the YMB-1 cells treated with vehicle, 1 μM vorinostat, and 10 μM vorinostat (n = 4 for each, P < 0.01 versus vehicle control; P < 0.05 versus 1 μM vorinostat treated), respectively. Western blot analysis also showed the vorinostat-induced (10 μM) downregulation of the TMEM16A protein in the protein lysate of PC-3 (Fig. 3B) and YMB-1 (Fig. 4B), in correlation with the results obtained by real-time PCR assay. After compensation of the optical density of the TMEM16A protein band signal (115 kDa) with that of the ACTB signal (45 kDa), the TMEM16A density in the vehicle control was expressed as 1.00 (Figs. 3B and 4B). The relative optical density in PC-3 and YMB-1 cells treated with vorinostat (10 μM) was 0.15 ± 0.02 and 0.25 ± 0.03, respectively (n = 4 for each, P < 0.01) (Figs. 3B and 4B).
Downregulation of TMEM16A by Pharmacologic HDAC2 and HDAC3 Inhibition in PC-3 and YMB-1 Cells.
We next examined the molecular characteristics of the HDAC subtypes expressed in PC-3 (Fig. 1F) and YMB-1 (Fig. 2F). In PC-3 cells, the HDAC1, HDAC2, and HDAC3 isoforms were predominantly expressed, and the HDAC1, HDAC2, HDAC3, and HDAC6 isoforms were expressed in YMB-1. In human tumor breast tissues, the HDAC1, HDAC3, and HDAC6 isoforms were highly expressed compared with normal breast tissues (Supplemental Fig. 1, C and D). We therefore examined the effects of selective HDAC inhibition on the TMEM16A expression level using three chemical compounds: an HDAC1 and HDAC2 selective HDACi, AATB (IC50 = 0.007 and 0.049 μM for HDAC1 and HDAC2, respectively, and IC50 >> 10 μM for the other HDAC isoforms) (Methot et al., 2008), an HDAC3 selective HDACi, T247 (IC50 = 0.24 μM for HDAC3, IC50 >> 100 μM for the other HDAC isoforms) (Suzuki et al., 2013), and an HDAC6 selective HDACi, NCT-14b (IC50 = 0.082 μM for HDAC6, IC50 > 3.5μM for the other HDAC isoforms) (Itoh et al., 2007). The chemical structures of these compounds are shown in Supplemental Fig. 3A. Quantitation analyses of the protein expression levels were performed as described earlier.
As shown in Fig. 3, A and B, 1 μM T247 exhibited a similar level of suppression of TMEM16A transcripts and proteins in PC-3 cells. On the other hand, a high concentration (300 nM) of AATB partially suppressed TMEM16A expression, while a low concentration (30 nM) of AATB or 1 μM NCT-14b did not elicit any significant change (Fig. 3, A and B). Similar results were obtained using YMB-1 cells (Fig. 4, A and B). The chemotherapy drugs paclitaxel and bafilomycin-A (a vacuolar-ATPase inhibitor) exhibited marked suppressive effects on cell proliferation in both of these cell lines (not shown), but no suppressive effect on TMEM16A expression was observed (Figs. 3 and 4). Also, niflumic acid (100 μM) did not induce any significant decrease in TMEM16A expression in either cell type (data not shown). Treatment of T247 for 48 hours induced the suppression of cell proliferation in both types of cells, but the suppressive effect was obviously weak compared with vorinostat (Supplemental Fig. 3, B and C). Chou et al. (2008) reported that the hydroxamates, such as vorinostat, have a fast-acting inhibitory mechanism compared with a benzamide-type HDACi, such as AATB, T247, or NCT-14b. In accord with this, vorinostat statistically significantly (P < 0.01) suppressed the expression of TMEM16A transcripts 12 hours after compound supplementation in both cell types, but no significant suppressive effects were observed for the other compounds in either cell type (Supplemental Fig. 2, B and D). Moreover, we examined the effects of vorinostat (1 and 10 μM) and T247 (1 μM) on cell viability in the prostate and breast cancer cell lines expressing TMEM16A less abundantly: LNCaP, MCF-7, and Hs578T. We found that 10 μM vorinostat significantly suppressed the cell viability in LNCaP and Hs578T (Supplemental Fig. 4, A and C), but 1 μM vorinostat and 1 μM T247 showed no significant inhibitory effects on the cell viability in LNCaP, MCF-7, or Hs578T (Supplemental Fig. 4). Taken together, TMEM16A may be mostly downregulated by pharmacologic blockade of HDAC3 in TMEM16A-expressing cancer cells.
Suppressive Effect of HDACis on Ca2+-Activated Cl− Currents in PC-3 and YMB-1 Cells.
In PC-3 (Fig. 5) and YMB-1 (Fig. 6) cells, Ca2+-activated Cl− currents were recorded using a whole-cell configuration with a pipette solution containing Ca2+ fixed at 1 µM and 140 mM Cs+. Just after membrane disruption, outwardly rectifying currents were not observed by the application of depolarizing steps (from −50 mV to +50 mV for 1 second) in either cell type treated with vehicle (0.1% DMSO) for 24 to 48 hours (data not shown), but 5 to 10 minutes after membrane disruption, application of the depolarizing steps revealed slowly activating, outwardly rectifying currents and repolarization to −80 mV for 1 second elicited rapidly declining tail currents in the vehicle control (Figs. 5, Aa and Ba, and 6, Aa and Ba), with voltage-dependent characteristics similar to those observed in human TMEM16A-expressing human embryonic kidney 293 (HEK293) cells (Supplemental Fig. 5, Cb and D).
Figures 5, C and D, and 6, C and D, show the current amplitude at the end of the depolarizing pulses. Activation of the outwardly rectifying currents 5 minutes after the membrane disruption was not observed when measured using Ca2+-free pipette solution (Supplemental Fig. 5, A and B). Ca2+-activated Cl− currents were almost completely suppressed by treatment with 10 µM vorinostat (PC-3: Fig. 5, Ab and C; YMB-1: Fig. 6, Ab and C). Furthermore, Ca2+-activated Cl− currents were >90% suppressed by treatment with 1 µM T247 (PC-3: Fig. 5, Bb and D; YMB-1: Fig. 6, Bb and D). The activated currents were sensitive to the selective TMEM16A inhibitor T16inh-A01 (data not shown). These results are consistent with those of the expression profile experiments.
Inhibition of TMEM16 Transcription by siRNA-Induced HDAC3 Downregulation in PC-3 and YMB-1 Cells.
We further examined the effect of selective HDAC1/HDAC2/HDAC3 inhibition using RNA interference. Transfection efficacy was determined using fluorescein isothiocyanate-conjugated control siRNA-A in PC-3 and YMB-1 cells (over 80%, see Materials and Methods). Quantitative real-time PCR analysis showed that HDAC3 downregulation (∼75% inhibition) markedly suppressed the expression level of the TMEM16A transcripts (∼75% inhibition) in PC-3 cells (Fig. 7, C and D). In contrast, HDAC1 or HDAC2 downregulation did not elicit any statistically significant (P > 0.05) decrease in TMEM16A expression (Fig. 7, A, B, and D) in PC-3 cells. Similarly, in YMB-1 cells, HDAC3 (∼65% inhibition) and HDAC2 downregulation (∼60% inhibition) statistically significantly (P < 0.01) suppressed the expression level of the TMEM16A transcripts (∼70% and 40%, respectively) (Fig. 8, B–D). On the other hand, HDAC1 downregulation did not exert any statistically significant (P > 0.05) suppressive effect on TMEM16A expression (Fig. 8, A and D). Control siRNA did not exert any statistically significant effects on HDAC1–3 and TMEM16A expression. The expression levels of TMEM16A (in arbitrary units) were 0.040 ± 0.003, 0.042 ± 0.006, and 0.013 ± 0.003 in HDAC1, HDAC2, and HDAC3 downregulated PC-3 cells, respectively. The expression levels of TMEM16A (in arbitrary units) were 0.095 ± 0.006, 0.062 ± 0.004, and 0.027 ± 0.003 in HDAC1, HDAC2, and HDAC3 downregulated YMB-1 cells, respectively. These data suggest that HDAC3 is the leading candidate for TMEM16A suppression.
In cancer cells, ion channels regulate cell proliferation and apoptosis via the modulation of membrane potential and intracellular Ca2+-mobilization (Yang and Brackenbury, 2013; Lang and Stournaras, 2014; Pardo and Stühmer, 2014) and thus are both potential therapeutic targets and biomarkers for cancer. Histone acetylation and DNA methylation on gene expression play important roles in various cellular processes such as proliferation, differentiation, and apoptosis, and DNA methylation is a key regulator of the transient receptor potential Ca2+ channel TRPC3 and inward-rectifier K+ channel Kir4.1 (Martin-Trujillo et al., 2011; Nwaobi et al., 2014). Epigenetic modification-based therapy with HDACi is one of the novel strategies for cancer treatment. However, the role of HDAC in the ion channel regulatory effect of HDACi on ion channel expression is less clear. Recently, it was reported that the Ca2+-activated Cl− channel TMEM16A is a potential therapeutic target and biomarker for prostate and breast cancers as well as head and neck squamous cell carcinomas (Ayoub et al., 2010; Liu et al., 2012; Britschgi et al., 2013). Britschgi et al. (2013) showed that a reduction in epidermal growth factor receptor signaling by TMEM16A knockdown is strongly associated with a decrease in breast cancer cell viability.
The main findings in our present study are as follows. 1) Epigenetic modulation occurred of the Ca2+-activated Cl− channel TMEM16A by the clinically available pan-HDACi vorinostat in TMEM16A-expressing cancer cell lines, namely, prostate cancer cell line PC-3 and the breast cancer cell line YMB-1. 2) HDAC subtypes were identified related to epigenetic regulation of TMEM16A by pharmacologic and siRNA-mediated inhibition of HDACs. The chemotherapy drugs paclitaxel and bafilomycin-A exhibited antiproliferative effects (data not shown) without any significant effect on the TMEM16A expression levels in PC-3 and YMB-1 cells (Figs. 3 and 4). Our results suggest that HDAC3 plays a crucial role in the epigenetic regulation of TMEM16A in TMEM16A-expressing cancer cells.
HDAC3 transcripts were highly expressed in PC-3, YMB-1, and human tumor breast tissues (Figs. 1F and 2F; Supplemental Fig. 1D). Pharmacologic and siRNA-mediated HDAC3 inhibition caused marked downregulation of TMEM16A expression in PC-3 and YMB-1 cells (Figs. 3, 4, 7, and 8), and its functional activity essentially disappeared as a result of treatment with the selective HDAC3 inhibitor T247 (1 μM) for 24 to 48 hours in both cell types (Figs. 5 and 6). Correspondingly, cell viability in both cell types was significantly (P < 0.05 or 0.01) inhibited by the treatment with T247 (1 μM), niflumic acid (100 μM), and TMEM16A siRNA, respectively (Figs. 1, C and D, and 2, C and D; Supplemental Fig. 3, B and C). Pharmacologic HDAC2 blockade in both types of cells or siRNA-mediated HDAC2 inhibition in YMB-1 alone caused ∼30% inhibition of the TMEM16A expression (Figs. 4 and 8D) but did not exhibit any significant effect on cell viability (Supplemental Fig. 3, B and C). Significant inhibition of cell viability was observed by the treatment with 100 μM niflumic acid (Figs. 1C and 2C) in both types of cells, but the treatment with 10 μM niflumic acid did not induce any changes (<5% in both) in cell viability (data not shown). These data suggest that significant suppression of the signaling pathways that control cell proliferation (i.e., epidermal growth factor receptor signaling) may not be induced by partial inhibition of TMEM16A activity in either cell type.
Recently, it has been reported that HDAC3 inhibition causes a decrease in proliferation and/or migration in breast cancer cells (Kim et al., 2010; Müller et al., 2013) and that HDAC3 is strongly expressed in breast cancers of a more aggressive tumor type (Müller et al., 2013) . Our previous study showed that human prostate cancer progression may be linked to the changes in the expression levels of Ca2+-activated K+ channel subtypes KCa1.1 and KCa3.1 (Ohya et al., 2009). Britschgi et al. (2013) have shown that amplification of TMEM16A correlates with both disease severity and poor prognosis in human breast cancer. It remains unclear whether amplification of HDAC3 correlates with disease severity and prognosis in breast cancer, but in gliomas a high level of HDAC3 expression correlates with a high tumor grade (Zhu et al., 2013). Taken together, the dysfunction of TMEM16A induced by HDAC3 inhibition is at least in part responsible for the decrease in cell viability in PC-3 and YMB-1 cells, and it may be related to the comparatively poor migration and invasion capacity observed in TMEM16A-overexpressing metastatic cancer cells.
One of the mechanisms underlying HDAC-induced promotion of tumor onset and progression is the repression of tumor suppressor gene transcription (Ropero and Esteller, 2007). Due to the epigenetic modification, HDACi are able to upregulate tumor suppressor gene expression, resulting in inhibited cancer cell viability (Gui et al., 2004). For example, several HDACi upregulate the tumor suppressor p53 (Kim et al., 2001) and the repressor element 1–silencing transcription factor (also called neuron-restrictive silencer factor) (Taylor et al., 2012) in cancer cells. These suppressors play important roles in repressing ion channel expression. Indeed, p53 negatively regulates the voltage-gated K+ channel EAG-1 (ether á go-go-1 K+ channel) (Lin et al., 2011), and repressor element 1-silencing transcription factor negatively regulates the voltage-gated K+ channel KV4.3 and Ca2+-activated K+ channel KCa3.1 in several different types of cells, including cancer cells (Cheong et al., 2005; Uchida et al., 2010; Ohya et al., 2011). HDAC3 selectively binds to many types of transcriptional repressors. For example, HDAC3 represses the following tumor suppressors: DBC-1 (deleted in breast cancer-1) (Chini et al., 2010), CREB-3 (cAMP response element binding protein-3) (Kim et al., 2010), and STAT-3 (signal transducers and activators of transcription-3) (Miao et al., 2014; Minami et al., 2014) in various types of cancer cells, including breast cancer cells. The molecular mechanisms underlying the epigenetic modulation of TMEM16A transcription remain to be elucidated, so further studies (e.g., genetic structure and promoter analysis of TMEM16A and siRNA-mediated knockdown of HDAC3-related tumor suppressors) will be needed to clarify the pathophysiologic significance of epigenetic modulation of TMEM16A in TMEM16A-expressing cancer cells.
In conclusion, this study provides novel evidence that the Ca2+-activated Cl− channel TMEM16A is epigenetically regulated by the inhibition of histone deacetylases in TMEM16A-expressing cancer cells. Of note, TMEM16A transcription in the respective prostate and breast cancer cell lines PC-3 and YMB-1 is mostly regulated by HDAC3. Pharmacologic and siRNA-mediated inhibition of HDAC3 inhibited TMEM16A expression and interfered with cell viability. TMEM16A is known to be overexpressed and to regulate the migration and invasion of metastatic cancer cells. Therefore, these findings constitute important information for an epigenetic modification-based therapeutic strategy by HDACi in metastatic prostate and breast cancers. In addition, TMEM16A is involved in a variety of biologic functions, such as epithelial secretion, neuronal excitability, nociception, and gastrointestinal motility (Duran and Hartzell, 2011). Recent studies have shown that HDAC3 is involved in neurodegenerative, gastrointestinal, and inflammatory diseases (Chen et al., 2012; Alenghat et al., 2013; Venkatraman et al., 2014). Therefore, TMEM16A inhibition via pharmacologic blockade of HDAC3 may have therapeutic potential for a wide range of disorders.
The authors thank Y. Masuno (Kyoto Pharmaceutical University), A. Matsui (Kyoto Pharmaceutical University), and K. Yotsutsuji (Aichi-Gakuin University) for their technical assistance. Pacific Edit (San Francisco, CA) reviewed the manuscript before submission.
Participated in research design: Ohya, Muraki, Suzuki.
Conducted experiments: Matsuba, Niwa, Kanatsuka, Nakazono, Ohya, Muraki, Hatano, Zhan.
Performed data analysis: Matsuba, Niwa, Kanatsuka, Nakazono, Ohya, Muraki, Hatano, Fujii.
Wrote or contributed to the writing of the manuscript: Ohya, Matsuba, Niwa, Muraki, Hatano, Suzuki.
- Received June 5, 2014.
- Accepted September 17, 2014.
This work was supported by a Grant-in-Aid for Scientific Research from the Japan Society for the Promotion of Science (No. 25460111); a research grant from the Promotion and Mutual Aid Cooperation for Private Schools of Japan (Kyoto Pharmaceutical University and Aichi-Gakuin University); Mochida Memorial Foundation for Medical and Pharmaceutical Research; and Uehara Memorial Foundation (to S.O.).
- histone deacetylase
- HDAC inhibitor
- small interfering RNA
- transmembrane proteins with unknown function 16
- polymerase chain reaction
- 2-(4-iodophenyl)-3-(4-nitrophenyl)-5-(2,4-disulfophenyl)-2H-tetrazolium, monosodium salt
- Copyright © 2014 by The American Society for Pharmacology and Experimental Therapeutics