Combination of ACY-241 and JQ1 Synergistically Suppresses Metastasis of HNSCC via Regulation of MMP-2 and MMP-9

Overexpression of histone deacetylase 6 (HDAC6) and bromodomain-containing protein 4 (BRD4) is related to aggressiveness of head and neck squamous carcinoma (HNSCC). Based on studies that HDAC6 and BRD4 are potential therapeutic targets of HNSCC, we hypothesized that the combination treatment of BET inhibitor JQ1 and HDAC6-selective inhibitor ACY-241 could exhibit synergistic anticancer effects in human papillomavirus (HPV)-positive and HPV-negative HNSCC cells. In this study, HNSCC cell growth and viability were measured by CCK-8 assay, apoptosis was analyzed by flow cytometry, and metastasis was studied by wound healing and transwell assays. Furthermore, immunoblotting is conducted to investigate proteins that modulate apoptosis or metastasis. Here, we report that the combination of ACY-241 and JQ1 shows synergistic cell growth inhibition, viability reduction, and apoptosis induction in HNSCC cells through inactivation of AKT and NF-κB signaling. Importantly, we demonstrate that combined treatment of ACY-241 and JQ1 synergistically suppresses TNF-α-induced migration and invasion via dysregulating matrix metalloproteinase (MMP)-2, MMP-9, and MT1-MMP. Overall, the combination of ACY-241 and JQ1 significantly suppresses proliferation and metastasis in HPV-positive and HPV-negative HNSCC. Collectively, these findings suggest that the co-inhibition of BET and HDAC6 can be a new therapeutic strategy in HNSCC.


Introduction
Head and neck cancer is the sixth most common cancer worldwide and occurs in the head and neck area, including the larynx, nasal cavity, oral cavity, paranasal sinuses, pharynx, and salivary glands. Despite various primary tumor sites, more than 95% of these epithelial tumors are head and neck squamous cell carcinoma (HNSCC) [1]. Although human papillomavirus (HPV) infection is not the major risk factor of all HNSCCs [2,3], HPV-positivity relates to improved response to treatments and survival rates in a subset of HNSCCs [4]. In particular, HPV-16 is strongly associated with the development of oropharyngeal squamous cell carcinoma (OPSCC), and 71% of node metastatic OPSCC are detected with HPV-16 [5]. Interestingly, HPV-positive HNSCCs are more metastatic to cervical lymph nodes compared to HPV-negative tumors [6]. Distant metastasis is rare in HNSCC compared to other cancers such as stomach, pancreas, lung, breast, or kidney [7], but increasing incidence of  Values represent mean ± SD from three independent experiments (n = 3). * p < 0.05, ** p < 0.01, or *** p < 0.001 vs. DMSO control. Next, we treated ACY-241 and JQ1 in combination to HPV-positive and HPV-negative HNSCC cells and analyzed the cell viability using CCK-8 assays. A substantial decrease in viability was observed following combined treatment compared with single agents. Synergism was evaluated using the Chou and Talalay method [24]. The combination of ACY-241 and JQ1 showed synergistic cytotoxicity with a combination index (CI) of <1.0 (Figure 2A,B). These data confirmed the robust anti-proliferative effect when ACY-241 and JQ1 were combined in both HPV-positive and HPV-negative HNSCC cells.  Next, we treated ACY-241 and JQ1 in combination to HPV-positive and HPV-negative HNSCC cells and analyzed the cell viability using CCK-8 assays. A substantial decrease in viability was observed following combined treatment compared with single agents. Synergism was evaluated using the Chou and Talalay method [24]. The combination of ACY-241 and JQ1 showed synergistic cytotoxicity with a combination index (CI) of < 1.0 (Figure 2A,B). These data confirmed the robust anti-proliferative effect when ACY-241 and JQ1 were combined in both HPV-positive and HPVnegative HNSCC cells. Cell viability was measured by CCK-8 assay (48 h). Synergism of ACY-241 and JQ1 was determined by combination index (CI) using Chou-Talalay method. (C,D) Inhibitory enzymatic effects of ACY-241 and JQ1 were confirmed by immunoblot analysis of acetyl α-tubulin and c-Myc, respectively. α-tubulin and GAPDH were used as loading controls. Protein levels were quantified relative to the loading control. Total protein was extracted after 24 h of ACY-241 (4 µM) or JQ1 (2 µM) treatment alone or in combination. Values represent mean ± SD (n = 3). * p < 0.05, ** p < 0.01, or *** p < 0.001 vs. DMSO control, $ p < 0.05, $$ p < 0.01, or $$$ p < 0.001 vs. ACY-241-treated group, # p < 0.05, ## p < 0.01, or ### p < 0.001 vs. JQ1-treated group. ns = not significant.

Combination Treatment of ACY-241 and JQ1 Synergistically Induces Apoptosis in HNSCC Cells
Based on our results, further experiments were conducted with 4 µM of ACY-241 and 2 µM of JQ1, which is the combination of the lowest concentrations that display noticeable synergistic effect. First, enzymatic inhibitory activities of ACY-241 and JQ1 were confirmed by observing their target proteins, acetyl α-tubulin and c-Myc, respectively [25,26]. ACY-241 increased acetylation of α-tubulin, and JQ1 decreased c-Myc in both HPV-positive 2A3 and HPV-negative FaDu HNSCC cells. Furthermore, HDAC6 protein level remained unchanged by ACY-241 ( Figure 2C,D). It has been previously reported that JQ1 did not modify BRD4 protein level [27]. We also confirmed that mRNA levels of HDAC6, BRD2, and BRD4 were unaffected after ACY-241 and JQ1 treatments (Figure S1A-C). As c-Myc oncogene is known to induce proliferation [20], we next performed immunoblotting to determine whether ACY-241 and JQ1 disrupt the apoptotic signaling pathway. PARP and caspase-3 were synergistically cleaved by combination treatment to exhibit pro-apoptotic effects. On the other hand, expression levels of anti-apoptotic proteins XIAP and Bcl-xL were synergistically reduced in both HPV-positive and HPV-negative HNSCC cells ( Figure 3A,B). However, Bcl-2 associated pro-apoptotic proteins, such as Bak, Bax, and Bad, remained unchanged by ACY-241 and JQ1 combination ( Figure S2). To further determine the apoptotic effect of HDAC6 and BET inhibition, flow cytometry analysis was performed to examine apoptosis after annexin V/propidium iodide staining. After 72 hours of combination treatment, early and late apoptosis were synergistically promoted in both HPV-positive and HPV-negative HNSCC cells. The percentage of apoptotic cells was as much as 9-fold higher than the additive effect of single inhibitor treatments ( Figure 3C,D). Collectively, these data show that simultaneous inhibition of HDAC6 and BET is an effective treatment strategy to promote apoptosis in both HPV-positive and HPV-negative HNSCC cells. represent mean ± SD (n = 3). * p < 0.05, ** p < 0.01, or *** p < 0.001 vs. DMSO control, $ p < 0.05, $$ p < 0.01, or $$$ p < 0.001 vs. ACY-241-treated group, # p < 0.05, ## p < 0.01, or ### p < 0.001 vs. JQ1-treated group.

Combination Treatment of ACY-241 and JQ1 Synergistically Inhibits TNF-α-Induced Effects by Degrading MMP-2 and MMP-9
To investigate the effect of HDAC6 and BET inhibition in metastasis, we tested protein expressions of the MMP family by immunoblotting. The most significantly associated MMPs in metastatic HNSCC are membrane-type 1-matrix metalloproteinase (MT1-MMP), MMP-1, -2, -3, and -9 and tissue inhibitors of metalloproteinase-2 [9,10]. EMT is a crucial cellular program that is observed during cell migration and wound healing, and it is regulated by MMP proteins [28]. Single treatments of ACY-241 and JQ1 substantially downregulated MMP-2, rather than MMP-9, in both HPV-positive 2A3 and HPV-negative FaDu HNSCC cells ( Figure 4A,B). The combination of ACY-241 and JQ1 also synergistically reduced MMP-2 and MMP-9, the type IV collagenase that has a significant function in EMT by altering cell-matrix and cell-cell interactions [29]. Furthermore, this combination suppressed MT1-MMP, the activator of MMP-2 that is correlated with tumorigenesis of HNSCC [10] ( Figure 4A,B). Consistent with protein levels, mRNA levels of MT1-MMP and MMP-2 were synergistically reduced by the combination treatment, but not MMP-9 ( Figure S1D-F). In addition to MT1-MMP, tumor necrosis factor-α (TNF-α), an inducer of MMP-9 secretion in hypopharyngeal and OSCC cell lines [30], was synergistically suppressed by a combination of ACY-241 and JQ1 in both HNSCC cells ( Figure 4C,D and Figure S1G).
Next, we examined EMT factors downstream of TNF-α. Among three EMT transcription factors (Snail, Twist1, and Zeb2), the mRNA level of SNAI1 was decreased by ACY-241 and JQ1 (Figure S1H-J). Moreover, protein expression of Snail was moderately reduced by inhibition of BET and HDAC6 ( Figure 4A,B). TNF-α-induced EMT in HNSCC is known to be involved in the nuclear factor-κB (NF-κB) signaling pathway, and NF-κB activation by TNF-α requires AKT phosphorylation [31]. Thus, we investigated whether ACY-241 and JQ1 affect the AKT/NF-κB signaling pathway in HPV-positive and HPV-negative HNSCC cells. HDAC6 and BET co-inhibition synergistically blocked phosphorylation of AKT and p65, which positively correlates with HNSCC malignancy [32], without changing total ATK and p65 levels ( Figure 4C,D). Together, our results indicate that the combination of ACY-241 and JQ1 suppresses TNF-α-induced responses by dysregulating MT1-MMP, MMP-2, and MMP-9 and by modulating AKT/NF-κB signaling pathways in both HPV-positive and HPV-negative HNSCC cells.

Combination Treatment of ACY-241 and JQ1 Synergistically Impairs Migration and Invasion in HNSCC Cells
Next, we performed wound healing assay and transwell migration assay to examine the migration of tumor cells. After 48 h of combination treatment in 2A3 and FaDu cells, proliferation and wound closure were synergistically diminished, and a significantly small number of cells migrated across the transwell membrane ( Figure 5A-C). Then, we performed transwell invasion assay by coating the insert membrane with Matrigel and providing ECM-like circumstances to study cell invasion. Similar to the results of transwell migration assay, combination treatment decreased the penetration of cells through the membrane ( Figure 5D). To represent migration and invasion capacity of viable cells, the absorbance of cells has been normalized by cell viability from Figure 1. These results suggest that co-targeting HDAC6 and BET weakens the migration and invasion capacity of HPV-positive and HPV-negative HNSCC cells.

Discussion
Combination therapy is an economic therapeutic strategy that combines two or more treatment agents that have already been developed. This approach exhibits more significant therapeutic effects by regulating tumor growth, metastasis, cell cycle, and apoptosis while escaping drug resistance evoked by monotherapies [33]. Synergistic effects of HDAC and BET inhibitors have been observed in multiple cancers [34,35]. In multiple myeloma, sensitivity to JQ1 is increased after cotreatment with HDAC6 inhibitor as confirmed through cell proliferation inhibition and apoptosis promotion [36]. However, the combinational treatment of HDAC6 and BET inhibitors has not yet been studied in HNSCC. The Kaplan-Meier plot showing the overall survival (OS) of HNSCC patients regarding gene expressions of HDAC6, BRD2, and BRD4 demonstrates that patients with high BRD2 or BRD4 display low survival ( Figure S3). In addition, previous studies have demonstrated that HPV-positive HNSCC patients have higher OS than HPV-negative HNSCC patients [37]. Here, we demonstrate that the combination of HDAC6-specific inhibitor ACY-241 and BET inhibitor JQ1 modulates molecular mechanisms of apoptosis and metastasis, displaying more significant synergistic anticancer effects than the additive effect of single inhibitors ( Figure 6). We showed that this combination approach reduces cell growth and viability and activates the caspase-3-dependent apoptosis signaling pathway. Anti-apoptotic protein Bcl-xL is downregulated to promote apoptosis, but other Bcl-2 associated pro-apoptotic proteins remain unchanged by ACY-241 and JQ1 combination ( Figure S2). Moreover, we show by flow cytometry that co-inhibition of HDAC6 and BET causes apoptosis in HNSCC cells.

Discussion
Combination therapy is an economic therapeutic strategy that combines two or more treatment agents that have already been developed. This approach exhibits more significant therapeutic effects by regulating tumor growth, metastasis, cell cycle, and apoptosis while escaping drug resistance evoked by monotherapies [33]. Synergistic effects of HDAC and BET inhibitors have been observed in multiple cancers [34,35]. In multiple myeloma, sensitivity to JQ1 is increased after cotreatment with HDAC6 inhibitor as confirmed through cell proliferation inhibition and apoptosis promotion [36]. However, the combinational treatment of HDAC6 and BET inhibitors has not yet been studied in HNSCC. The Kaplan-Meier plot showing the overall survival (OS) of HNSCC patients regarding gene expressions of HDAC6, BRD2, and BRD4 demonstrates that patients with high BRD2 or BRD4 display low survival ( Figure S3). In addition, previous studies have demonstrated that HPV-positive HNSCC patients have higher OS than HPV-negative HNSCC patients [37]. Here, we demonstrate that the combination of HDAC6-specific inhibitor ACY-241 and BET inhibitor JQ1 modulates molecular mechanisms of apoptosis and metastasis, displaying more significant synergistic anticancer effects than the additive effect of single inhibitors ( Figure 6). We showed that this combination approach reduces cell growth and viability and activates the caspase-3-dependent apoptosis signaling pathway. Anti-apoptotic protein Bcl-xL is downregulated to promote apoptosis, but other Bcl-2 associated pro-apoptotic proteins remain unchanged by ACY-241 and JQ1 combination ( Figure S2). Moreover, we show by flow cytometry that co-inhibition of HDAC6 and BET causes apoptosis in HNSCC cells.  Previous studies have identified the critical role of MMP-2 and MMP-9 in tumor invasion and lymph node metastasis of HNSCC [10]. In hypopharyngeal cancer tissues, mRNA and protein levels of MMP-9 and MMP-2 are significantly upregulated compared to paracancerous tissues. Moreover, the expression levels are enhanced with the increased lymph node metastasis degree and tumor clinical stages, suggesting that MMP-2 and MMP-9 are involved in the occurrence and aggressiveness of hypopharyngeal cancer [38]. It has been suggested that MMP-2 and MMP-9 are potential prognostic biomarkers of HNSCC [39]. We show that combination therapy of ACY-241 and JQ1 has significant anti-metastatic effects through MMP-2 and MMP-9 modulations. This pathway features the suppression of TNF-α/AKT/NF-κB signaling cascade. TNF-α is a multifunctional cytokine that promotes EMT by activating AKT and NF-κB, resulting in augmented invasion and metastasis in many cancers, including OSCC cells [40]. Correspondingly, metastatic tongue squamous cell carcinoma tissues and cells are detected with highly active NF-κB [41]. NF-κB directly binds to 5'-flanking region of the MT1-MMP gene and MMP-9 promoter, thereby inducing MMP-2 and MMP-9 expressions for dynamic EMT [42,43]. Moreover, active PI3K/AKT increases MMP-9-mediated EMT in laryngeal cancer cells [44]. In consistent with previous findings, we demonstrate that the combination of small molecule inhibitors synergistically suppresses metastasis by modulating MMP-2 and MMP-9 expressions via TNF-α/AKT/NF-κB signaling in HPV-negative and HPV-positive HNSCC.
Blocking of the interaction of HSP90 and MMP-2 and MMP-9 induces metastasis suppression in breast cancer [45] and lung cancer [46], respectively. The dissociation of MMP and HSP90 is caused by the hyperacetylation of HSP90, which is a substrate of HDAC6 [46]. Thus, suppression of migration and invasion is triggered by promoting HDAC6-mediated HSP90/MMP-2 or MMP-9 dissociation and followed by MMP-2 and MMP-9 degradation in breast and lung cancer cells. Consistent with the previous findings, HDAC6 inhibition by ACY-241 blocks migration and invasion of HNSCC cells. In the case of HNSCC, ACY-241 significantly inhibits MMP-2 expression more than it does MMP-9. It has been previously reported that the expression level of MMP-2 is higher than that of MMP-9 in FaDu cells [47]. Moreover, MMP-2 is more correlated with lymph node metastasis than MMP-9 in laryngeal squamous cell carcinoma [48]. These data imply that targeting MMP-2 may be more effective for metastasis suppression than targeting MMP-9 in HNSCC. In addition to ACY-241, although the mechanism is unclear, JQ1 downregulates the expression of MMP2-and MMP-9 and inhibits angiogenesis in glioblastoma tumors [49] and pancreatic ductal adenocarcinoma cells [50]. Here, we demonstrate for the first time that combined treatment of ACY-241 and JQ1 synergistically blocks migration and invasion of HPV-positive and HPV-negative HNSCC cells by degrading MMP2 and MMP-9. These findings suggest that a combination of BET and HDAC6 inhibitors significantly suppresses metastasis in HNSCC irrespective of HPV infection.
Further investigations are required to confirm which member of BET is deregulated by JQ1 to induce anticancer effects. Many studies of JQ1 have focused on BRD4 inhibition because BRD4 is an appealing target in cancer treatment. JQ1 is often called a BRD4 inhibitor due to its strong inhibitory potency for BRD4. However, JQ1 is a pan-BET inhibitor that can moderately inhibit BRD2 by binding to the acetyl-lysine binding pocket of bromodomains [51]. According to our Gene Expression Omnibus (GEO) analysis, BRD2 was significantly overexpressed and positively correlated with HDAC6 in HNSCC patients among four BET proteins ( Figure S4). This suggests that BRD2 inhibition may synergize with HDAC6 inhibition to suppress metastasis in HNSCC. This hypothesis is supported by the study that ZNF281, an EMT transcription factor, is enriched in the BRD2-binding genes that are downregulated by JQ1 [52]. It is also known that BRD2 promotes EMT, while BRD3 and BRD4 negatively regulate this process in triple-negative breast cancer. Thus, BRD2 knockdown or JQ1 treatment reduced the transcription of genes involved in EMT [53]. For these reasons, it is necessary to examine which BET protein binds to the EMT-promoting genes observed in this study (MMPs, TNF-α, AKT, NF-κB) and whether BRD2 overexpression participates in HNSCC metastasis.
In cell growth and viability assays, we obtained greater GI 50 and IC 50 values for HPV-positive HNSCC cells (Table 1). However, this result was not enough to determine that HPV-positive HNSCC cells have higher sensitivity for ACY-241 and JQ1, because a similar degree of synergistic anticancer effects was observed in subsequent experiments. We have shown by GEO analysis that the BRD2 mRNA level is upregulated and displays a stronger positive correlation with HDAC6 mRNA in HPV-positive HNSCC samples compared to HPV-negative samples ( Figure S4). In addition, we interestingly observed that adhesion and migration abilities in HPV-positive 2A3 cells are more impaired compared with those in HPV-negative FaDu cells in the cell culture. This observation can be explained by HPV E6 oncogene expression because E6 disrupts the cell adhesion pathway by TAp63β degradation [54]. Although the significant difference derived from HPV infection is not clearly demonstrated, further investigations are required to uncover the role of HPV infection in metastasis and survival outcome of HNSCC.

HNSCC Cell Lines and Culture
The hypopharyngeal cancer cell lines were used to represent HNSCC cell lines. HPV-positive 2A3 cells and HPV-negative FaDu cells were bought from American Type Culture Collection (ATCC). 2A3 and FaDu cells were cultured in Dulbecco's Modified Eagle Medium (Sigma Chemical, St. Louis, MO, USA) supplemented with 10% fetal bovine serum (FBS), 100 U/mL penicillin, and 100 µg/mL streptomycin. Cells were subcultured every 3-4 days and incubated in a humidified atmosphere at 37 • C and 5% CO 2 .

Cell Growth and Viability Assay
Cell growth and viability were determined using CCK-8 assay. Each cell line was seeded at a density of 3 × 10 3 cells in 96-well plates containing 130 µL medium. After overnight incubation, cells were treated with ACY-241 and JQ1 in single and in combination, then incubated for another 24, 48, and 72 h. Thirteen microliters of WST-8 was added to each well, and absorbance was measured at 450 nm using a multimode microplate reader (Tecan Group, Ltd., Mannedorf, Switzerland).

Drug Combination Analysis
Synergism between ACY-241 and JQ1 was evaluated using Chou-Talalay method [24]. Fraction-affected (Fa) versus combination index (CI) plot was drawn using CalcuSyn (Biofosft). The drug combination was considered synergistic when CI was less than 1.

Apoptosis Assay
Flow cytometry analysis (FACS) was used to analyze apoptosis. Each cell line was seeded at a density of 7 × 10 5 -1 × 10 6 cells in 100 mm cell culture dish and treated with 4 µM ACY-241 and 2 µM JQ1 on the next day. After 48 h incubation, cells were washed twice with ice-cold phosphate-buffered saline (PBS) and detached from the plate with trypsin and EDTA. Cell pellets were collected after 3 min centrifugation at 1000 rpm, then resuspended with 4-700 µL of 1× binding buffer. Cells were stained with 5 µL propidium iodide (PI) and 0.5 µL of Annexin V-fluorescein isothiocyanate (FITC) in the dark for 15 min and diluted with 400 µL of 1× binding buffer (Annexin V-FITC Apoptosis Detection Kit; BD556547, BD Pharmingen San Diego, CA, USA). The cells were analyzed using a fluorescence-activated cell sorting flow cytometer.

Wound Healing Assay
2A3 and FaDu cells were seeded in a 6-well plate at a density of 7 × 10 5~1 .5 × 10 6 cells. On the next day, cells were scratched with 200 µL pipette tip and washed twice with serum-free medium, followed by treatment of 4 µM ACY-241 and 2 µM JQ1. The wound was photographed after 0 h and 48 h incubation at 50× magnification.

Transwell Migration and Invasion Assay
Transwell migration and invasion assays were performed using a transwell insert plate (PET membrane, 8 µM pore, SPL 36224, Korea). Transwell invasion assay was conducted after coating the insert membrane with 0.4 mg/mL Matrigel (Corning Matrigel Membrane Matrix, cat# 356234), and transwell migration assay was conducted without coating. Four hundred microliters of cell suspension solution in serum-free medium was added to the upper well at a density of 2 × 10 5 cells, and 500 µL of medium containing 10% FBS was added to the lower well. On the next day, 4 µM ACY-241 and 2 µM JQ1 were added to the lower well. After 48 h incubation, interiors of the inserts were cleaned by wet cotton swabs. Cells were stained with the cell stain solution (0.1% crystal violet, 20% methanol) and visualized under a light microscope at 200× with iSolution Lite (IMT i-Solution Inc., BC, Canada). Then, cells were solubilized in 10% glacial acetic acid, and absorbance was measured at 560 nm using a multimode microplate reader (Tecan Group, Ltd., Mannedorf, Switzerland). To represent migration and invasion capacity of viable cells, the absorbance of cells has been normalized by cell viability.

Western Blot Analysis
2A3 and FaDu cells were seeded at a density of 5 × 10 5 cells and treated with 4 µM ACY-241 and 2 µM JQ1 on the next day. Cells were washed twice with ice-cold PBS and extracted with 100 µL lysis buffer, then lysed by sonication at 20% amplitude. Bradford protein assay was performed to measure protein concentrations. Protein samples were prepared with a 5× sample buffer and loaded to 7.5-12% polyacrylamide gel. After SDS-page, proteins were transferred to nitrocellulose membrane. Membranes were blocked with 5% skim milk at room temperature and incubated with primary antibody against α-tubulin

Statistical Analysis
All results are expressed as means ± standard deviation (SD) of three independent experiments. For the cell growth and viability test of single agents, statistical significance was determined by unpaired two-tailed Student's t-test. Statistical analysis for other data was performed by GraphPad Prism software 7.0 (Graphpad Software, San Diego, CA, USA). One-way or two-way ANOVA followed by post-hoc analysis with Bonferroni's multiple comparison test was used to evaluate statistical significance. p < 0.05 was considered statistically significant for data.

Conclusions
Collectively, we demonstrate that the combined targeting of HDAC6 and BET in HNSCC synergistically promotes apoptosis and suppresses cell growth and viability, migration, and invasion.

Conflicts of Interest:
The authors declare no conflict of interest.