This article is mentioned in:

Introduction

Head and neck squamous cell carcinoma (HNSCC) is the sixth most common cancer in the world with approximately 800,000 cases being diagnosed every year (1). Advanced HNSCC shows poor prognosis and resistance to standard therapy consisting of cisplatin. Because 60-80% of HNSCC cases carry a TP53 mutation (2-4) and human papillomavirus (HPV)-positive HNSCC shows impaired p53 function (5), therapeutic targeting of the TP53 mutation has become a focus of research. Additionally, as p53 is functional in normal somatic cells, targeted therapy against TP53 mutations can relieve adverse effects. Thus, owing to its selective cytotoxicity against TP53-mutated carcinoma cells, adavosertib (Adv), a WEE1 G2 checkpoint kinase (WEE1) inhibitor, has been the recent focus of study (6,7).

WEE1 is a kinase that regulates replication stress and cell cycle arrest at the G2/M checkpoint. DNA damage during the S-phase activates WEE1, which in turn phosphorylates cyclin-dependent kinase 1 (CDK1) at Tyr15 to inactivate its kinase activity (8). Because p53 regulates the G1/S checkpoint (9), TP53-mutated cells rely on the G2/M checkpoint to arrest the cell cycle to repair their DNA damage. Therefore, Adv treatment of TP53-mutated cells leads to the cells entering M-phase without DNA repair, which in turn causes cell death named mitotic catastrophe. Mitotic catastrophe is a type of cell death characterized by DNA damage, abnormalities in the mitotic apparatus, dysfunction of the mitotic checkpoint, and failure in the occurrence of normal mitosis (10,11), but there is no clear definition. As Adv abrogates cell cycle arrest, it was suggested that the co-administration of DNA-damaging drugs with Adv enhances cytotoxicity (7,12,13). In the present study, we sought to identify drug combinations to enhance Adv-induced cytotoxicity without losing selective cell death in TP53-mutated cells.

It has been reported that histone deacetylase (HDAC) inhibitors induce growth arrest, differentiation, and apoptosis in vitro, and suppress tumor growth in xenograft mouse models with various cancer cell lines including HNSCC (14-16). Additionally, HDAC inhibitors have been shown to enhance cytotoxicity in combination with DNA-damaging anticancer drugs (17,18) or Adv (19-21). We recently reported that ricolinostat (RCS), a selective histone deacetylase 6 (HDAC6) inhibitor, exhibits potent cell growth inhibition in HNSCC cell lines in vitro in combination with the proteasome inhibitor bortezomib (22). Of note, HDAC6 differs from other HDACs by deacetylating cytoplasmic proteins, including α-tubulin. HDAC6 has also been reported to be involved in the cell cycle via deacetylation of α-tubulin (23). Additionally, inhibition of HDAC6 has been reported to sensitize cancer cells to DNA-damaging drugs (24-26). As the disruption of the cell cycle or DNA damage enhances mitotic catastrophe, in the present study, we assessed the effect of the combined treatment of Adv and RCS on HNSCC cell lines with TP53 mutation (CAL27, SAS, HSC-3, Detroit562, and OSC-19) or impaired p53 function by HPV-infection (UPCI-SCC154).

Materials and methods

Reagents

Adavosertib (AZD1775, MK-1775, hereafter referred to as Adv) was purchased from MedChemExpress. Ricolinostat (ACY-1215, hereafter referred to as RCS) was purchased from Selleck Chemicals. Adv and RCS were dissolved in dimethyl sulfoxide (DMSO) (Wako Pure Chemical Industries) to make the stock solution at a concentration of 10 mM for Adv and 5 mM for RCS. Doxorubicin (DOX) was purchased from Cayman Chemical Co. z-VAD-fmk was purchased from Peptide Institute (Japan).

Cell lines and culture conditions

The human oral squamous cell carcinoma cell line CAL27, the human pharyngeal squamous carcinoma cell line Detroit562, human tongue squamous carcinoma cell line UPCI-SCC154, the human breast mammary gland adenocarcinoma cell lines MCF7 and MDA-MB-231, and the human lung adenocarcinoma cell line A549 were purchased from the American Type Culture Collection (ATCC). The human tongue squamous carcinoma cell lines SAS, HSC-3, and OSC-19 cells were obtained from the JCRB Cell Bank (Osaka, Japan). A549 and MCF7 cells possess wild-type TP53. MDA-MB-231, CAL27, HSC-3, SAS, Detroit562, and OSC-19 cells carry mutant TP53. UPCI-SCC-154 cells are HPV-positive and show impaired p53 function. CAL27, Detroit562, MCF7, MDA-MB-231, and A549 cells were cultured in Dulbecco's modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS; Biosera) and 1% penicillin/streptomycin solution (Wako Pure Chemical Industries). UPCI-SCC154 cells were cultured in MEM supplemented with 10% FBS and 1% penicillin/streptomycin solution. SAS and OSC-19 cells were cultured in DMEM/F12 medium supplemented with 10% FBS and 1% penicillin/streptomycin solution. HSC-3 cells were cultured in Eagle's minimum essential medium (EMEM) supplemented with 10% FBS and 1% penicillin/streptomycin solution. All cell lines were cultured at 37°C in a humidified incubator containing 5% CO2 and 95% air. In all experiments, as a control, cells were treated with control medium containing less than 1% DMSO to match the amount of solvent brought in by the drug. All cell line experiments were conducted within 10 passages after thawing. Mycoplasma contamination was tested routinely using the e-Myco™ Mycoplasma PCR Detection kit ver.2.0 (iNtRON Biotechnology, Inc.).

Establishment of TP53-KO A549 and TP53-KO MCF7 cells

TP53-KO A549 cells were established as previously described (27). Briefly, A549 cells were transfected with pSpCas9 (BB)-2A-Puro (PX459) V2.0 plasmid vector (a gift from Dr Feng Zhang; plasmid cat. no. 48139; Addgene) with the following sequence (5′-CAC CGT CCA TTG CTT GGG ACG GCA A-3′), selected with puromycin for 2 days, and grown without puromycin to select single colonies. TP53-KO MCF7 cells were also established in the same way.

Establishment of CAL27/H2B-mCherry/AcGFP-α-tubulin cells

To establish CAL27 cells expressing H2B-mCherry and AcGFP-α-tubulin, CAL27 cells were infected with lentiviruses and positive clones were selected using puromycin and blasticidin. Lentiviruses were produced in 293T cells (ATCC) by transfection of the following plasmids: pMD2.G (gift from Dr Didier Trono: Addgene #12259), psPAX2 (gift from Dr Didier Trono: Addgene #12260), pLenti6-H2B-mCherry (gift from Dr Torsten Wittmann: Addgene plasmid #89766) (28), and pLenti-AcGFP-α-tubulin. For pLenti-AcGFP-α-tubulin construction, TUBA1A cDNA was cloned into pAcGFP1-Hyg-C1 vector (Takara Bio Inc.), and then, the AcGFP-α-tubulin fused gene was inserted into the pLentiN vector (gift from Dr Karl Munger: Addgene #37444) (29).

RNA interference

For the gene silencing of HDAC6 in CAL 27 cells, HDAC6 siRNA and control siRNA were synthesized as follows (Japan Bio Services Co., Ltd.): siHDAC6#2: sense GCU UAU UUA AGU GUU AAU AdT dT and antisense UAU UAA CAC UUA AAU AAG CdA dC; siHDAC6#3: sense GGU UUU UGC UUU UUC AAC UdT dT and antisense AGU UGA AAA AGC AAA AAC CdG dC; siHDAC6#4: sense GCA UAU GUA AUA AAG UAC AdT dT and antisense UGU ACU UUA UUA CAU AUG CdA dA; control siLuc: sense CUU ACG CUG AGU ACU UCG AdT dT and antisense UCG AAG UAC UCA GCG UAA GdT dT. siRNAs were diluted to 200 nM in Opti-MEM I (Thermo Fisher Scientific, Inc.). Transfection was performed using Lipofectamine RNAiMAX transfection reagent (Thermo Fisher Scientific, Inc.) according to the manufacturer's instructions. Knockdown (KD) efficiency was assessed using western blotting.

Assessment of cell death

The dead cell counts were assessed by staining with propidium iodide (PI) (FUJIFILM Wako Chemical Corp.), and the number of red fluorescent signals was counted using the IncuCyte ZOOM (Sartorius) automated live cell imaging system. The cells were treated with Adv with or without RCS in the presence of PI (2.5 µg/ml) for up to 48 h in 96-well plates in tetraplicate (30).

Morphological assessment

The cells were spread on glass slides using a Cytospin 4 Centrifuge (Thermo Fisher Scientific, Inc.) to prepare glass slides, stained with May-Grünwald-Giemsa and examined under a digital microscope (BZ-X800; KEYENCE Co.).

Western blotting

The cells were lysed using RIPA lysis buffer (Nacalai Tesque) added together with a protease and phosphatase inhibitor cocktail (Nacalai Tesque). Equal amounts of proteins (25 µg) were loaded onto the gels (7.5, 10 and 15% gels were used), separated by SDS-PAGE, and then transferred onto Immobilon-P membranes (Millipore Corp.). These membranes were probed with primary antibodies, such as anti-p53 antibody (Ab) (sc-126, 1/1,000), anti-β-actin Ab (sc-47778, 1/1,000), anti-HDAC6 Ab (sc-11420, 1/1,000), anti-acetylated α-tubulin Ab (sc-23950, 1/1,000), and anti-α-tubulin Ab (sc-5286, 1/1,000) were purchased from Santa Cruz Biotechnology, Inc. Anti-PARP Ab (#9542S, 1/1,000), anti-caspase3 Ab (#9665S, 1/1,000), anti-phospho-p53 Ab (#9286, 1/1,000), anti-p21 Ab (#2947S, 1/1,000), anti-phospho-ATR Ab (#9947, 1/1,000), anti-ATR Ab (#2790, 1/1,000), anti-phospho-Chk1 (Ser345) Ab (#2348, 1/1,000), anti-Chk1 Ab (#2360, 1/1,000), anti-phospho-Chk2 (Thr68) Ab (#2197, 1/1,000), anti-Chk2 Ab (#3440, 1/1,000), anti-phospho-Cdc2 (Tyr15) Ab (#4539, 1/1,000), anti-Cdc2 Ab (#9116, 1/1,000), anti-H2A.X Ab (#7631, 1/1,000), and anti-phospho-histone H2A.X (Ser139) Ab (#9718, 1/1,000) were purchased from Cell Signaling Technology, Inc.

Assessment of mitotic catastrophes

CAL27/H2B-mCherry/AcGFP-α-tubulin cells were seeded on collagen-coated glass-bottom dishes (CELLview #627870, Greiner). The next day, the cells were treated with control medium, Adv, RCS, or Adv + RCS, and time-lapse images were obtained every 10 min using a confocal microscope LSM 700 equipped with CO2, temperature, and humidity controller (Carl Zeiss) or fluorescent microscope BZ-X800 equipped with a time-lapse module (BZ-H4XT) (KEYENCE). The cells were maintained at 37°C and 5% CO2, under humidified conditions. Time-lapse images from fluorescent microscopy were used to analyze the cell fate.

Statistical analysis

All quantitative data are expressed as mean ± standard deviation (SD). Statistical analyses for cell death inhibition with z-VAD-fmk treatment, and cell death in combination with Adv treatment and HDAC6 KD were performed using two-way ANOVA followed by Bonferroni's multiple comparison test. For the analysis of the M-phase duration, the Kruskal-Wallis test followed by Dunn's multiple comparison test was used. Statistical significance was set at P<0.05. All analyses were performed using GraphPad Prism 5 software (GraphPad Software, Inc.). The synergistic effect was assessed as follows. Based on nesting all elapsed hours for each experiment, mixed-effect linear regression analyses were performed by setting the number of cell deaths as the dependent variables and the concentration of the two anticancer drugs as independent variables. To assess the possible synergistic effects of the two drugs, the interaction term was added as one of the independent variables in the model to calculate p-for-interaction. All statistical analyses were performed using Stata version 17.0 (Stata Corp.). When the mono treatment was enough to kill almost all cells, the analysis indicated an 'antagonistic effect', because the dead cell count in the combination treatment group was smaller than the dead cell count in each mono treatment group.

Results

Combined treatment of Adv and RCS synergistically induces cell death in TP53-mutated HNSCC cells

To assess the effect of the combined treatment of Adv and RCS on TP53-mutated HNSCC cells, namely CAL27, HSC-3, SAS, Detroit562, and OSC-19 cells, and HPV-positive UPCI-SCC-154 cells, which showed impaired p53, these cells were treated with Adv and RCS for 48 h and PI-positive dead cell number was measured using a live cell imaging system. In HNSCC cells, treatment with Adv and/or RCS induced cell death in a dose- and time-dependent manner. Notably, co-administration of Adv and RCS synergistically enhanced cell death in four out of five TP53-mutated HNSCC cell lines (except Detroit562 cells) as well as in the HPV-positive UPCI-SCC-154 cell line (Figs. 1 and S1). To address whether this synergistic effect was ubiquitous, breast cancer cell lines (MDA-MB-231 and MCF7) and a lung cancer cell line (A549) were also treated with Adv and/or RCS for 48 h (Figs. 1 and S1). Although co-administration of RCS with Adv led to enhanced cell death in MDA-MB-231 cells with TP53 mutation, little or no enhanced cell death was observed in MCF7 and A549 cells, both carrying wild-type TP53. Because Adv has been reported to induce mitotic catastrophe in TP53-mutated cells, we hypothesized that RCS enhanced Adv-induced mitotic catastrophe in TP53-mutated cells but not in wild-type TP53 cells. To address this issue, we next compared the induction of cell death between TP53 wild-type and TP53 knockout (KO) A549 cells, which were established using the CRISPR-Cas9 system (Fig. 2A). Although TP53-KO A549 cells did not show altered sensitivity to Adv as compared to wild-type A549 cells, co-administration of RCS significantly enhanced cell death in TP53-KO but not TP53-WT A549 cells (Fig. 2B). To further confirm this observation, we established TP53-KO MCF7 cells using the CRISPR-Cas9 system (Fig. S2A). Although wild-type MCF7 cells showed slight synergistic cell death in the combined treatment of Adv and RCS, TP53-KO MCF7 cells showed increased synergistic cell death by the combined treatment as shown in A549 cells (Fig. S2B). This indicated that the combination treatment of Adv and RCS can synergistically induce cell death in cells lacking p53 function.

Figure 1 Ricolinostat enhances adavosertib-induced cytotoxicity in HNSCC cell lines. CAL27, UPCI-SCC-154, HSC-3, SAS, MDA-MB-231, and A549 cells were treated with Adv in combination with RCS for up to 48 h. Cells were monitored using IncuCyte live cell imaging system, and dead cell number was assessed using PI staining. Time-dependent and dose-dependent cell death numbers are shown in the left and the middle panels, respectively. Representative data of three independent experiments are shown. n=3, bar, mean ± SD. Synergistically enhanced cell death in the combination treatment was analyzed as described in Materials and methods and summarized in the right panels. A significant synergistic effect with both Adv and RCS treatment with P<0.05 is shown in red. An antagonistic effect with P<0.05 is shown in blue. HNSCC, head and neck squamous cell carcinoma; Adv, adavosertib; RCS, ricolinostat.

Figure 2 Combined treatment of adavosertib and ricolinostat enhances cell death only in TP53-KO A549 cells. (A) TP53 KO in A549 cells was confirmed by western blotting. Expression of p53, phospho-p53, and p21 was assessed after doxorubicin (DOX) (1 µM) treatment for 24 h. β-actin was used as a loading control. Relative band intensity of p-p53/p53 was calculated and is summarized in the right panel. (B) TP53-WT and TP53-KO A549 cells were treated with Adv in combination with RCS for up to 48 h. Cells were monitored using IncuCyte live cell imaging system, and dead cell number was assessed by PI staining. Dose-dependent and time-dependent cell death numbers are shown. The synergistic effect on cell death in combination treatment was assessed and is summarized at the right. Representative data of three independent experiments are shown. n=3, bar, mean ± SD. Adv, adavosertib; RCS, ricolinostat.

Combined treatment of Adv and RCS in cancer cells leads to mitotic catastrophe

To identify the mechanism by which the combined treatment of Adv and RCS leads to enhanced cell death, we treated CAL27 cells with Adv and RCS for 24 h and observed their cell morphology. In response to drug treatment, CAL27 cells showed nuclear fragmentation and chromatin condensation (Fig. 3A), which are characteristic features of cells undergoing apoptosis. Western blotting revealed that Adv or RCS treatment induced the cleavage of caspase-3, and co-administration of the two drugs further increased the cleavage of caspase-3 in CAL 27 cells (Fig. 3B). Additionally, Adv-and RCS-induced cell death was abolished in the presence of a pan-caspase inhibitor, z-VAD-fmk (Fig. 3C). These data demonstrated that co-administration of Adv and RCS induced apoptosis. Adv is known to induce mitotic catastrophe, which subsequently leads to apoptotic or non-apoptotic cell death (31). To investigate whether the co-administration of Adv and RCS enhanced mitotic catastrophe, we established CAL27 cells stably expressing AcGFP-α-tubulin and histone-H2B-mCherry to monitor mitosis. Time-lapse imaging showed that most cells treated with Adv could not complete mitosis and underwent cell death, which indicated mitotic catastrophe (Fig. 3D-F). Additionally, Adv treatment resulted in a prolonged duration of mitosis (Fig. 3G). On the contrary, although RCS treatment induced cell death, most of the cell death occurred during interphase, and most RCS-treated cells did not enter metaphase (Fig. 3E and F). This indicated that, unlike Adv, RCS hardly induced mitotic catastrophe. Of note, when CAL27 cells were simultaneously exposed to Adv and RCS, dead cell counts were increased and most of the cell death occurred during mitosis (Fig. 3D-F). However, the duration of the mitotic phase until cell death was not further extended as compared to the cells treated with Adv alone (Fig. 3G). These data showed that RCS enhanced the Adv-induced mitotic catastrophe.

Figure 3 Co-administration of ricolinostat enhances adavosertib-induced mitotic catastrophe in CAL27 cells. (A) CAL27 cells were treated with control, Adv (0.5 µM), RCS (5 µM), or Adv+RCS for 24 h and then stained with May-Grünwald-Giemsa stain. Scale bar, 25 µm. Arrowheads indicate nuclear chromatin condensation. (B) CAL27 cells were treated with control, Adv (0.5 µM), RCS (5 µM), and Adv + RCS for 24 h, and then, cleavage of PARP and caspase-3 were assessed by western blotting. (C) CAL27 cells were treated with control, Adv (0.5 µM), RCS (5 µM), or Adv + RCS in the presence of z-VAD-fmk (0, 25, and 50 µM) for 24 h, and dead cell number was monitored using IncuCyte live cell imaging system by PI staining. n=7, bar, mean ± SD; *P<0.05 vs. 0 µM z-VAD-fmk. (D) Live cell imaging of CAL27 cells expressing AcGFP-α-tubulin and Histone H2B-mCherry. Cells were treated with control, Adv (0.5 µM), or Adv + RCS (5 µM) and monitored using confocal microscopy. Representative images of cells after the nuclear envelope breakdown (NEBD) are shown. Time after the NEBD is shown in the upper right corner of each image. Scale bar, 20 µm. (E) Five representative cell fates in each treated cell with control, Adv (0.5 µM), RCS (5 µM), or Adv + RCS are shown in a tree diagram. (F) The ratio of cell death in mitosis or interphase in each treated cell is summarized in the pie chart. n for each condition is shown at the bottom. Data from three independent experiments are summarized. (G) Time from NEBD to telophase or cell death was assessed and summarized. n=75, 59, 29, 46. Data from three independent experiments are summarized. *P<0.05 vs. the control. PARP, poly(ADP-ribose) polymerase; Adv, adavosertib; RCS, ricolinostat.

RCS promotes Adv-induced mitotic entry along with the increment of γ-H2A.X expression

Next, to address how RCS enhances Adv-induced mitotic catastrophe, we assessed the G2/M checkpoint and DNA damage response (DDR)-related proteins by western blotting in CAL27 cells (Fig. 4A). WEE1, a Ser/Thr protein kinase family member, phosphorylates CDK1 at Tyr15, inhibits its activity, and acts as a negative regulator for entry into mitosis (G2 to M transition) (32). Thus, Adv treatment decreased CDK1 (Tyr15) phosphorylation (Fig. 4A). This indicated that CDK1 was activated, leading to premature mitotic entry. At the same time, we observed increased phosphorylation of Chk1 (Ser345), which indirectly inactivates CDK1. This increased p-Chk1 presumably compensated for the increased CDK1 activity. In contrast, RCS treatment suppressed Chk1 phosphorylation. Thus, co-administration of RCS with Adv suppressed p-Chk1 and p-CDK1. This appeared to further promote forced mitotic entry to induce mitotic catastrophe. In support of these findings, co-administration of the two drugs resulted in further increase in γ-H2A.X expression, a marker of DNA double-strand breaks. These data suggest that RCS enhanced Adv-induced mitotic catastrophe by suppressing p-Chk1 and p-CDK1 (Fig. 4A).

Figure 4 Ricolinostat suppresses phosphorylation of Chk1 and further suppresses p-CDK1 when co-administered with adavosertib. (A) CAL27 cells and (B) TP53-WT and TP53-KO A549 cells were treated with Adv (0.5 µM), RCS (5 µM), and Adv+RCS for 24 h (CAL27 cells) or 48 h (A549 cells), and then the expression of DNA damage response-related proteins (p-Chk2, Chk2, p-ATR, ATR, p-Chk1, Chk1, p-CDK1, CDK1, and γ-H2A.X) was assessed by western blotting. To assess the inhibitory effect of HDAC6 by RCS, the level of acetylated (ac)-α-tubulin was monitored. Expression of β-actin was assessed as the loading control. The relative band intensity of each phosphorylated protein was calculated and summarized at the right. Representative data of three independent experiments are shown. Adv, adavosertib; RCS, ricolinostat; WT, wild-type; Chk, checkpoint kinase; ATR, ATR serine/threonine kinase; CDK1, cyclin-dependent kinase 1.

As shown in Fig. 1, the pronounced cell death induction by combined treatment of Adv and RCS appeared to be dependent on TP53 mutations. We examined whether the co-administration of RCS suppressed p-Chk1 in both WT and TP53-KO A549 cells (Fig. 4B). In both cells, Adv treatment increased p-Chk1 and suppressed p-CDK1, similar to that observed in CAL27 cells. Although co-administration of RCS with Adv suppressed p-CDK1 and p-Chk1 in both cell lines, γ-H2A.X expression was increased only in TP53-KO A549 cells. This may reflect the dependency of TP53-mutated cells on the G2/M checkpoint. These data also suggest that RCS enhances premature mitotic entry by suppressing p-Chk1 in A549 cells. Additionally, we assessed the DDR-related protein expression in Detroit562 cells which carry mutated-TP53 but did not show pronounced cell death by combined treatment of Adv and RCS. Detroit562 cells showed almost the same result as WT A549 cells and failed to further upregulate γ-H2A.X expression as compared to treatment with Adv alone (Fig. S3). This suggests that some of the TP53-mutated cell lines may have gained a compensatory mechanism to p53 and are insensitive to co-administration of RCS with Adv.

RCS enhanced cell death induction via inhibition of HDAC6

It has been reported that although RCS selectively inhibits HDAC6 (IC50=4.7 nM), it also inhibits HDAC1, 2, and 3 (IC50=58, 48, and 51 nM, respectively) at higher concentrations (33). To clarify whether the enhanced cell death mediated by RCS was due to the inhibition of HDAC6, we knocked down HDAC6 using three different siRNAs (siHDAC6 #2, #3, and #4). All three siRNAs efficiently knocked down HDAC6 and increased acetylated α-tubulin, which is a substrate for HDAC6 (Fig. 5A). Although KD of HDAC6 by itself did not change p-Chk1 levels or γ-H2A.X expression levels, Adv treatment in HDAC6-KD cells resulted in increased γ-H2A.X expression, as in the case of CAL27 cells treated with RCS and Adv (Fig. 5B). However, we did not observe the suppression of p-Chk1 or p-CDK1 in Adv-treated HDAC6-KD cells. In terms of cell death induction, treatment with Adv in HDAC6-KD cells significantly and synergistically increased cell death, as in the case of CAL27 cells concomitantly treated with RCS and Adv (Fig. 5C-E). This suggested that increased DNA damage and enhanced cell death by co-administration of RCS with Adv appeared to have been caused by HDAC6 inhibition, but the enhanced mitotic catastrophe caused by RCS addition to Adv treatment may be independent of HDAC6 inhibition.

Figure 5 HDAC6 knockdown (KD) enhances adavosertib-induced cell death in CAL27 cells. (A) Knockdown efficiency of HDAC6 in CAL27 cells with three different HDAC6 siRNAs was assessed by western blotting. (B) CAL27 cells were transfected with siLuc or three different siHDAC6 siRNAs (#2-4) and then treated with or without Adv (1 µM) for 24 h. Expression of DDR-related proteins was detected by western blotting. Expression of HDAC6 and acetylated (ac)-α-tubulin were monitored to assess the KD efficiency. Representative data of three independent experiments are shown. (C and D) CAL27 cells were transfected with siLuc or three different siHDAC6 siRNAs (#2-4) and then treated with various concentrations of Adv up to 24 h. Cells were monitored using IncuCyte live cell imaging system, and dead cell number was assessed using PI staining. Dose-dependent (C) and time-dependent (D) cell death numbers are shown. Representative data of three independent experiments are shown. n=3, bar, mean ± SD. *P<0.05 vs. siLuc. (E) The synergistic effect of cell death in the combination of Adv treatment and HDAC6 KD shown in (C) and (D) was assessed and is summarized. Adv, adavosertib; RCS, ricolinostat; HDAC6, histone deacetylase 6; DDR, DNA damage response.

Discussion

In the present study, the combined WEE1 G2 checkpoint kinase (WEE1) inhibitor adavosertib (Adv) and the histone deacetylase 6 (HDAC6)-selective inhibitor ricolinostat (RCS) treatment exhibited synergistic cytotoxicity in TP53-mutated, impaired p53 function by HPV-infection, or TP53-KO HNSCC, lung cancer, and breast cancer cell lines likely via enhanced induction of Adv-induced mitotic catastrophe (Fig. 6). WEE1 phosphorylates the amino acids Tyr15 and Thr14 of CDK1, which keeps the kinase activity of CDK1 low and prevents entry into mitosis (32). As p53 regulates the G1/S checkpoint, TP53-mutated cells rely on the G2/M checkpoint to arrest the cell cycle to repair their DNA damage. Thus, TP53-mutated cells with increased replication stress appeared to exhibit greater sensitivity to WEE1 inhibitors (34,35). The synergistic effect of the two-drug combination shown in Fig. 1 appeared to be mediated through forced entry of the cell into M phase by CDK1 activation. Indeed, the two-drug combination resulted in enhanced dephosphorylation of CDK1, which is an active state of CDK1, as compared to that in the treatment of Adv alone. This was accompanied by pronounced γ-H2A.X expression, which indicated an increase in DNA double-strand breaks in TP53-mutated cells (Fig. 4A and B). In Adv-treated cells, a compensatory mechanism appeared to be induced, leading to an increased expression of p-Chk1. However, combination treatment with RCS suppressed p-Chk1 upregulation and further activated CDK1 to promote premature mitotic entry (Figs. 4 and 6).

Figure 6 Schema of the effect of adavosertib and ricolinostat on the G2/M checkpoint. In TP53-mutated cells, DNA repair is dependent on the G2/M checkpoint regulated by CDK1. CDK1 is a key regulator inducing cell cycle progression through the G2/M checkpoint. WEE1 phosphorylates and inhibits CDK1. Adavosertib inhibits WEE1 and then CDK1 is dephosphorylated and activated. Chk1 inhibits Cdc25, which dephosphorylates and activates CDK1. Ricolinostat decreases the phosphorylation of Chk1, indicating inhibition of Chk1. Chk, checkpoint kinase; ATR, ATR serine/threonine kinase; ATM, ATM serine/threonine kinase; CDK1, cyclin-dependent kinase 1; HDAC6, histone deacetylase 6; WEE1, WEE1 G2 checkpoint kinase.

Adv is an orally available, first-in-class, reversible WEE1 inhibitor (36). Most clinical trials using WEE1 inhibitors have been conducted as a combination with chemotherapy or radiotherapy (37-40). In the case of WEE1 inhibitor monotherapy against recurrent uterine serous carcinoma, which frequently possesses TP53 mutation, the treatment showed significant elongation of progression-free survival (PFS) and a high response rate (41). Consistent with our results (Figs. 1 and 2), several previous studies have reported that sensitivity to Adv is dependent on TP53 mutation status (6,7,42), whereas a few other reports have shown that TP53 status does not alter sensitivity to Adv (43). This may suggest that TP53 single gene mutation by itself is not sufficient to determine sensitivity to WEE1 inhibitor monotherapy.

HDAC6 is a unique member of the HDAC subfamily possessing two catalytic domains, namely DAC1 with E3-ligase activity and DAC2 with deacetylase activity (44). HDAC6 deacetylates histones; however, cytoplasmic proteins, namely α-tubulin, cortactin, and heat shock protein 90 (HSP90), are also deacetylated by HDAC6 (45-47). The roles of HDAC6 in tumorigenesis and cell-cycle progression have been reported (23,48,49). The overexpression of HDAC6 has been reported in acute myeloid leukemia, oral squamous cell carcinoma, and ovarian cancer (50-52), and higher HDAC6 expression appears to correlate with tumor progression and malignancy in hepatocellular carcinoma and esophageal squamous cell carcinoma (53,54). In our study, HDAC6 inhibition by RCS suppressed cell cycle progression (Fig. 3), which was consistent with previous reports showing that RCS or other HDAC inhibitors caused G2/M phase arrest (53,55). In contrast, in the presence of Adv + RCS, the cells underwent a pronounced mitotic catastrophe, indicating increased entry into the M-phase. The molecular mechanisms underlying this contradictory phenomenon remain unclear.

Our study demonstrated that HDAC6 inhibition by RCS suppressed p-Chk1 expression (Fig. 4A and B), whereas HDAC6 KD slightly increased Chk1 (Fig. 5B). Thus, suppression of p-Chk1 by RCS seemed to be HDAC6-independent. However, as described above, in addition to deacetylase activity, HDAC6 possesses E3-ligase activity, which ubiquitinates and degrades Chk1 and functions in DDR. Therefore, it was reported that HDAC6 deletion resulted in an increase in Chk1, which enhanced radiation-induced cell cycle arrest in non-small cell lung cancer cells (44). In our system, although HDAC6 KD suppressed both activities, RCS interacted only with DAC2 and E3-ligase activity localized in DAC1 appeared to be intact. If the increase in Chk1 due to DAC1 inhibition overcomes the inhibition of p-Chk1 caused by DAC2 inhibition, the HDAC6 KD experiment alone is not sufficient to determine whether RCS suppresses p-Chk1 via HDAC6. To address this, we need to further evaluate the specific disruption/deletion of the DAC2 domain in HDAC6.

In terms of the Chk1-CDK1 axis, Adv treatment of HDAC6-KD cells resulted in increased γ-H2A.X and cell death despite no decrease in p-Chk1 or p-CDK1 (Fig. 5). This indicates that the enhanced cell death by siHDAC6 or RCS is not limited to be mediated by Chk1-CDK1. As previously reported, we also showed that RCS treatment and HDAC6 KD resulted in α-tubulin acetylation (Figs. 4 and 5) (45). It is well-known that post-translational modifications of tubulins contribute to microtubule dynamics, which are crucial for proper spindle organization and cell cycle progression (56). CYLD, a tumor suppressor gene product, has been reported to bind and inactivate HDAC6 along with increased acetylation of tubulin, which negatively regulates cell-cycle progression (23). In addition, tubastatin A, an HDAC6 inhibitor, has been reported to disrupt maturational progression and meiotic apparatus assembly in mouse oocytes (57). Therefore, it is still possible that HDAC6 inhibitors including RCS cause abnormal mitosis by constitutively enhancing the acetylation of tubulins, leading to disruption of microtubule dynamics. This may also play a role in promoting the Adv-induced mitotic catastrophe.

To the best of our knowledge, this is the first report showing that RCS with a higher selectivity against HDAC6 suppresses p-Chk1. Although suppression of p-Chk1 by RCS may be independent of HDAC6, enhanced DNA damage and cell death were observed in both RCS-treated and HDAC6-KD cells in combination with Adv treatment. As prominent cell death in Adv + RCS-treated cells was only observed in TP53-mutated and TP53-KO cells, this drug combination appears to be much less toxic to normal cells with intact TP53 than in TP53-mutated cancer cells. Applying this drug combination in the experiment with the tumor xenograft model appears to have potential for future research. Although in vivo studies remain to be carried out, this drug combination with high specificity for cells lacking p53 function could be a good candidate for the treatment of HNSCC patients carrying TP53 mutations as well as human papillomavirus (HPV)-positive HNSCC patients lacking p53 function.

Supplementary Data

Availability of data and materials

The datasets used during the present study are available from the corresponding authors upon reasonable request.

Authors' contributions

KM (Miyazawa) and NT conceived and designed the present study. KM (Miyake), NT and HK (Kazama) performed the experiments. KM (Miyake), NT and HK (Kikuchi) analyzed, interpretated the data. KM (Miyake), NT and KM (Miyazawa) confirmed the integrity of the data. KM (Miyazawa), NT, MH, KT and KM (Miyake) wrote, reviewed and/or revised the manuscript. All authors read and approved the manuscript and agree to be accountable for all aspects of the research in ensuring that the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Ethics approval and consent to participate

Not applicable.

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Acknowledgments

Not applicable.

Funding

This study was supported by JSPS KAKENHI Grant Number 20K07298 (to NT) and the Cancer Research Grant afforded by the Tokyo Medical University Cancer Research Foundation (to KM).

References