2017 Volume 40 Issue 6 Pages 844-851
Epigenetic processes, in addition to genetic abnormalities, play a critical role in refractory malignant diseases and cause the unresponsiveness to various chemotherapeutic regimens and radiotherapy. Herein we demonstrate that histone deacetylase inhibitors (HDACis) can be used to sensitize malignant melanoma B16F10 cells to carbon ion irradiation. The cells were first treated with HDACis (romidepsin [FK228, depsipeptide], trichostatin A [TSA], valproic acid [VPA], and suberanilohydroxamic acid [SAHA, vorinostat]) and were then exposed to two types of radiation (carbon ions and gamma-rays). We found that HDACis enhanced the radiation-induced apoptosis and suppression of clonogenicity that was induced by irradiation, having a greater effect with carbon ion irradiation than with gamma-rays. Carbon ion irradiation and the HDACi treatment induced G2/M and G0/G1 cell cycle arrest, respectively. Thus, it is considered that HDACi treatment enhanced the killing effects of carbon ion irradiation against melanoma cells by inducing the arrest of G1 phase cells, which are sensitive to radiation due to a lack of DNA homologous recombination repair. Based on these findings, we propose that pretreatment with HDACis as radiosensitizers to induce G1 arrest combined with carbon ion irradiation may have clinical efficacy against refractory cancer.
Epigenetic processes, in addition to genetic abnormalities, play a critical role in refractory malignant diseases. With malignant melanoma, as with many malignancies, aberrant transcriptional repression is a hallmark of refractory cancer, which resists not only chemotherapy but also radiotherapy.1,2) To restore gene expression, the use of histone deacetylase inhibitors (HDACis) may hold some promise for treating these resistant cancers, as these molecularly targeted agents that inhibit histone deacetylases have been shown to possess antitumor properties. At present, some HDACis have been approved by the U.S. Food and Drug Administration (FDA) for clinical use as ethical pharmaceuticals.3)
To sensitize the unresponsive cancer cells, the use of HDACis has attracted much attention as epigenetic modifiers. For instance, the use of a HDACi elegantly enhanced immune responsiveness against murine melanoma.4) Moreover, some effects of HDACis in combination with low linear energy transfer (LET) radiation therapy also have recently been reported.5–7) However, findings on HDACi-mediated effects in combination with high-LET radiation therapy remain to be limited8,9) and substantial effects of HDACis in combination with carbon ion irradiation against melanoma cells have not yet been investigated.
We therefore hypothesized that HDACis could be used to sensitize cancer cells to carbon ions as an anticancer therapy. In this study, we investigated a combination of four HDACis (romidepsin [FK228, depsipeptide], trichostatin A [TSA], valproic acid [VPA], and suberanilohydroxamic acid [SAHA, vorinostat]) and two types of radiation (carbon ions and gamma-rays) using an apoptosis analysis, clonogenic survival assay, and cell cycle analysis. Consequently, we demonstrate that HDACis can be used to radiosensitize murine B16F10 melanoma cells to carbon ion irradiation.
Murine B16F10 melanoma cells were purchased from the American Type Culture Collection (Manassas, VA, U.S.A.). The cells were cultured in Dulbecco’s Modified Eagle’s Medium (Thermo Fisher Scientific Inc., Waltham, MA, U.S.A.) supplemented with 10% fetal bovine serum (MP Biomedicals, LLC., Santa Ana, CA, U.S.A.) and 1% antibiotic–antimycotic liquid (Thermo Fisher Scientific Inc.) at 37°C in a humidified atmosphere with 5% CO2.
Romidepsin (FK228, depsipeptide) was obtained from Gloucester Pharmaceuticals, Inc. (Cambridge, MA, U.S.A.); trichostatin A (TSA) was purchased from Wako Pure Chemical Industries, Ltd. (Osaka, Japan); and valproic acid sodium salt (VPA) and suberoylanilide hydroxamic acid (SAHA) were purchased from Sigma-Aldrich Co. LLC (St. Louis, MO, U.S.A.).
Cells in the exponential growth phase were seeded at a density of 1×106 cells per 35 mm culture dish. They were then treated with each HDACi for 16 h before irradiation. For detecting acetylated histone (Fig. 1), the following concentrations of HDACis were used: 0–80 nM for FK228, 0–100 nM for TSA, 0–1.6 mM for VPA, and 0–16 µM for SAHA. For later experiments, the following concentrations of HDACis were used, based on the amount that was required to reduce the colony formation rate by 10% when used in isolation: 20 nM FK228, 10 nM TSA, 400 µM VPA, and 4 µM SAHA. Cells were also cultured in dimethyl sulfoxide (DMSO) as a control, at a final concentration of 0.1%.
B16F10 cells were exposed to the indicated concentrations of four HDACis (romidepsin [FK228], trichostatin A [TSA], valproic acid [VPA], and suberanilohydroxamic acid [SAHA]) for 16 h. Cells were lysed and analyzed by Western blotting for anti-acetyl-histone H3 (Lys 9; AcH3-K9), anti-acetyl-histone H3 (Lys 18; AcH3-K18), anti-histone H3, and anti-glyceraldehyde 3-phosphate dehydrogenase (GAPDH; as an internal control).
After being exposed to the HDACis for 16 h, cells were lysed in RIPA Lysis and Extraction Buffer (Thermo Fisher Scientific Inc.), and the protein concentrations were determined using the BCA Protein Assay Kit (Thermo Fisher Scientific Inc.). Proteins were separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), transferred to polyvinylidene difluoride (PVDF) membranes, blocked in 5% bovine serum albumin (BSA) in tris-buffered saline with 1% Tween 20 (TBST), and probed with the following primary antibodies: anti-acetyl-histone H3 (Lys 9), anti-acetyl-histone H3 (Lys 18), and anti-histone H3 (Cell Signaling Technology, Inc., Danvers, MA, U.S.A.). The membranes were then incubated with horseradish peroxidase (HRP)-conjugated secondary antibody (Cell Signaling Technology, Inc.), visualized with a chemiluminescence kit (GE Healthcare, Little Chalfont, Bucks, U.K.) according to the manufacturer’s instructions, and exposed using a LAS-3000 Imaging System (FUJIFILM Corporation, Tokyo, Japan). Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as an internal control.
The cells were exposed to monoenergetic 18.3-MeV/n carbon ions accelerated using an AVF930 cyclotron (Sumitomo Heavy Industries, Tokyo, Japan) at Takasaki Ion Accelerators for Advanced Radiation Application (TIARA) of the Takasaki Advanced Radiation Research Institute (TARRI), National Institutes for Quantum and Radiological Science and Technology (QST). The LET value and dose rates of carbon ions on the cell surface were estimated as previously described,10) giving values of 108 keV/µm and 0.01–0.40 Gy/s, respectively. Gamma-rays were emitted from a cobalt-60 radiation source at TARRI, QST, with an LET value of 0.2 keV/µm11) and a dose rate of 4–48 Gy/h. The culture medium was temporarily removed from the irradiation vessels just before irradiation, and the vessels were capped with a thin sheet of polyimide film (Du Pont–Toray, Tokyo, Japan) to prevent the samples from drying out and becoming contaminated during irradiation at room temperature. Control cells were sham-irradiated and handled in parallel with the test cells. For the clonogenic survival assay (Fig. 2), the irradiation dose was 0–4 Gy for carbon ions and 0–12 Gy for gamma-rays. For later experiments, the irradiation dose was determined based on the dose required to reduce survival to 10% (D10), as estimated from a preliminary study, giving values of 1.8 Gy for carbon ions and 8.1 Gy for gamma-rays.
(a) Following HDACi pretreatment and radiation exposure, the cells were incubated for 24 h, and apoptotic cells were detected by analyzing their affinity with annexin V and 7-amino-actinomycin D (7-AAD) using flow cytometric analysis with the Muse™ Cell Analyzer; cells were incubated in DMSO as a control. (b) In carbon ion irradiated cells, the percentage of early apoptotic cells was significantly increased by all four HDACis (romidepsin [FK228], trichostatin A [TSA], valproic acid [VPA], and suberanilohydroxamic acid [SAHA]). (c) In gamma-ray irradiated cells, TSA and VPA did not affect the enhancement of early apoptosis. Data are presented as the mean±S.E. of at least three independent experiments in triplicate and were compared using Student’s t-tests (* p<0.05; ** p<0. 01; n.s., not significant).
Following HDACi pretreatment and radiation exposure, cells were incubated for 24 h, and the rate of apoptosis was measured using the Muse Annexin V and Dead Cell Kit (Merck KGaA, Darmstadt, Germany) according to the manufacturer’s instructions. Data are presented as the mean±standard error (S.E.) of at least three independent experiments in triplicate and were compared using Student’s t-tests at a significance level of p<0.05.
Following HDACi pretreatment and radiation exposure, cells were trypsinized and replated after the appropriate dilution onto 100 mm culture dishes. Between 100 and 10000 cells were seeded per dish to obtain between 30 and 100 well-separated colonies per dish. After 9(±1) d, the colonies were fixed with 2% formalin (Wako Pure Chemical Industries, Ltd.) in phosphate-buffered saline (PBS) (Wako Pure Chemical Industries, Ltd.), stained with 0.002% crystal violet (Wako Pure Chemical Industries, Ltd.), and observed under a stereoscopic microscope (Olympus Corporation, Tokyo, Japan). Any colonies that generated more than 50 cells were considered clonogenic survivors and the number of colonies was normalized based on each unirradiated control. Data are presented as the mean±S.E. of at least three independent experiments in triplicate.
To evaluate the interaction between the HDACis and irradiation, two-way ANOVA was performed, in which the treatment conditions and irradiation dose were included as independent variables, and the number of colonies was the dependent variable. Wherever there was a significant difference between HDACi treated and untreated groups (p<0.05), the significance of the interaction was assessed to determine whether the effects were additive (no significant interaction) or synergistic (significant interaction, p<0.05).
Following HDACi pretreatment and radiation exposure, cells were incubated for 3–24 h and cell cycle analysis was performed using the Muse™ Cell Cycle Kit (Merck KGaA) according to the manufacturer’s instructions. Data are presented as the mean±S.E. of at least three independent experiments in triplicate.
Western blotting was performed to confirm that the four HDACis resulted in histone acetylation in the B16F10 cells. Results showed that treatment with each HDACi acetylated histone H3 in a concentration-dependent manner (Fig. 1). Therefore, in the following experiments, we used the middle concentration of each HDACi: 20 nM for FK228, 10 nM for TSA, 400 µM for VPA, and 4 µM for SAHA. We determined the concentration of each HDACi required to reduce the colony formation rate by 10% when used in isolation (data not shown).
An apoptosis assay was performed to examine whether the HDACis enhanced the killing effects of carbon ion irradiation in B16F10 cells, using an irradiation dose of 1.8 Gy of carbon ions, and the same experiment was conducted using gamma-rays (8.1 Gy) for comparison. Flow cytometry was used to analyze the affinity of these cells with annexin V and 7-amino-actinomycin D (7-AAD). Annexin V binds to phosphatidylserine (PS), which is moved to the surface in early apoptotic cells, while 7-AAD stains dead cells, as it can pass through cell membranes that have been permeabilized by apoptosis. In Fig. 2a, the horizontal axis represents the intensity of annexin V binding to PS in each cell, with cells with a higher intensity of binding being regarded as early apoptotic cells. By contrast, the vertical axis represents the intensity of 7-AAD in each cell, with cells with a higher intensity being regarded as dead cells. We considered annexin V positive but 7-AAD negative cells (plotted in the bottom right of Fig. 2a) to be early apoptotic cells.
As shown in Fig. 2b, each “carbon+HDACi” group showed more increased numbers of apoptotic cells than “carbon+DMSO” group, indicating that all of the examined HDACis enhanced apoptosis induced by carbon ion irradiation. By contrast, when used in combination with gamma-rays, pretreatment with FK228 or SAHA enhanced apoptosis, while pretreatment with TSA or VPA did not at these experimental concentrations (Fig. 2c). None of the HDACis enhanced late apoptosis induced by either type of irradiation (data not shown).
Next, we measured the clonogenic ability of cells that had been pretreated with FK228 or SAHA, which enhanced apoptosis regardless of the type of radiation used. To do this, we counted the number of colonies of pretreated B16F10 cells that formed following exposure to carbon ions or gamma-rays and normalized these values based on the unirradiated control for each treatment (Fig. 3a). We found that pretreatment with FK228 additively enhanced the radiation-induced suppression of clonogenicity caused by carbon ion irradiation (Fig. 3b) because a statistical difference in medication condition (FK228-treated and untreated) was detected (F1, 24=13, p=0.0014) and no significant interaction between FK228 and carbon ion irradiation was detected (F5, 24=1.7, p=0.18). Interestingly, pretreatment with SAHA synergistically enhanced the suppression of clonogenicity induced by carbon ion irradiation and additively enhanced the suppression of clonogenicity induced by gamma-ray irradiation (Fig. 3c). For carbon ion irradiation, a statistical difference in medication condition (SAHA-treated and untreated) was detected (F1, 24=18, p=0.00029), and a significant interaction between SAHA and carbon ion irradiation was detected (F5, 24=4.2, p=0.0069). For gamma-ray irradiation, a statistical difference in medication condition (SAHA-treated and untreated) was detected (F1, 24=28, p<0.001), and no significant interaction between SAHA and gamma-ray irradiation was detected (F5, 24=2.4, p=0.071). Thus, carbon ion irradiation was statistically more effective at suppressing clonogenicity than gamma-ray irradiation when used in combination with both HDACis; and SAHA was more effective in suppressing clonogenicity than FK228 when used in combination with either type of radiation.
(a) Following HDACis pretreatment, cells were irradiated with carbon ions or gamma-rays and then trypsinized and replated after the appropriate dilution; the resulting colonies were fixed and stained after 9(±1) d; control cells were incubated in DMSO. (b) Pretreatment with romidepsin (FK228) reduced the survival fractions, particularly after carbon ion irradiation. (c) Pretreatment with suberanilohydroxamic acid (SAHA) significantly suppressed clonogenicity compared with FK228 after both carbon ion and gamma-ray irradiations. Data are presented as the mean±S.E. of at least three independent experiments in triplicate. To evaluate the interaction between the HDACis and irradiation, two-way ANOVA was carefully performed. Wherever there was a significant difference between the HDACis-treated and -untreated groups (p<0.05), the significance of the interaction was assessed to determine whether the effects were additive (no significant interaction) or synergistic (significant interaction, p<0.05). (* p<0.05; ** p<0.01; n.s., not significant.)
Finally, we further carried out a cell cycle analysis to investigate why FK228 and SAHA have different sensitizing effects. To do this, we used flow cytometry to analyze the amount of DNA in each cell in order to assess the cell cycle phase as a function of incubation time. Three hours after irradiation, SAHA strongly induced G0/G1 arrest (Fig. 4a), while 12 h after irradiation, G2/M arrest was induced by carbon ion irradiation in the absence of any HDACi pretreatment (Fig. 4b). The cell cycle distributions at 3–24 h after irradiation, as a function of incubation time, are showed in Fig. 4c. These results demonstrate that cells were arrested in the G0/G1 phase following pretreatment with HDACis, and the arrest in G2/M was gradually induced by carbon ions post-irradiation (Fig. 4c). SAHA induced G0/G1 arrest more strongly than FK228 at 3 h after sham irradiation at the tested concentrations of HDACis (Fig. 4d). Furthermore, when G0/G1 arrest was induced, G2/M arrest was relatively suppressed (Fig. 4e); and when cells were pretreated with an HDACi, G2/M arrest was also reduced at 12 h after irradiation, which is when there was a peak in irradiation-induced G2/M arrest, instead exhibiting delayed peaks (Fig. 4e). We also obtained similar results using gamma-ray irradiation (data not shown).
(a) Cells were pretreated with the HDACis romidepsin (FK228) and suberanilohydroxamic acid (SAHA), and irradiated; cell cycle analysis was conducted 3 h after irradiation via flow cytometry with the Muse™ Cell Analyzer, showing that HDACis enhanced G0/G1 arrest. Control cells were incubated in DMSO. (b) Twelve hours after irradiation, G2/M arrest was induced by carbon ion irradiation in the absence of HDACi pretreatment. (c) Cell cycle distribution of B16F10 cells pretreated with HDACis at 3–24 h after irradiation. (d, e) The percentage of G0/G1 phase cells (d) and G2/M phase cells (e) plotted as a function of incubation time after irradiation. Data are presented as the mean±S.E. of at least three independent experiments in triplicate.
The treatment of murine B16F10 cells with HDACis led to the acetylation of histones in a concentration-dependent manner (Fig. 1). Since the acetylation and deacetylation of histones are controlled by the balance between histone acetyltransferase (HAT) and HDAC, treatment with HDACis makes HAT predominate, which blocks the deacetylation reaction of HDAC and causes histones to be acetylated.12) It is known that treatment with HDACis can decondense the chromatin structure.13,14) When acetyl groups are added to histones at lysines, the positive charge of the lysine residues may be dampened, weakening their interaction with the negatively charged DNA backbone and resulting in the chromatin conformation becoming more relaxed.15) It has been reported that genomic DNA is vulnerable to radiation when the chromatin structure is relaxed and so, since reactive radicals arising from the radiolysis of water molecules might be a major contributor to DNA damage induced by radiation, this decondensed chromatin may increase the susceptibility to radical attacks.16) Therefore, B16F10 cells that are pretreated with HDACis will be more easily damaged by radiation than untreated cells.
We found that HDACis enhanced radiation-induced apoptosis and suppression of clonogenicity in B16F10 cells (Figs. 2, 3), with these effects being greater in carbon ion irradiated cells than in gamma-ray irradiated cells. It is well known that high-LET radiation holds greater biological effects than low-LET radiation of the same dose.17,18) Our results also showed carbon ions alone had superior cell-killing effects than gamma-rays. The apoptosis analysis showed that all of the examined HDACis effectively sensitized the melanoma cells to carbon ion irradiation (Fig. 2b), whereas pretreatment with TSA or VPA did not enhance apoptosis following gamma-ray irradiation (Fig. 2c). Furthermore, the clonogenic survival assay also showed that carbon ion irradiation was statistically more effective than gamma-ray irradiation when used in combination with each HDACi (Figs. 3b, c).
We expected that the radiosensitizing effects of the HDACis may have been caused by the indirect action of radiation on the chromatin that had been relaxed by pretreatment with the HDACis. However, contrary to this expectation, we found that the effects were greater when the HDACis were used in combination with high-LET radiation, which has a small indirect action, than with low-LET radiation. Therefore, it appears that chromatin decondensation and the indirect action of radiation may not be related to the radiosensitizing effects of HDACis. Similarly, it has been reported that the inhibition of poly(ADP-ribose) polymerase (PARP) 1, which is involved in DNA repair, led to greater sensitization of tumor cells to carbon ion irradiation than to low-LET irradiation.19) Furthermore, it has been reported that the suppression of nonhomologous end-joining repair may result in greater radiosensitivity when used in combination with carbon ions compared to the suppression of homologous recombination repair.20) Therefore, the radiosensitizing mechanisms of HDACis need to consider not only the amount of DNA damage but also the DNA repair responses.
Pretreatment with SAHA synergistically sensitized the cells to carbon ion irradiation and was found to be more effective than pretreatment with FK228 based on the clonogenic survival assay (Fig. 4). Therefore, to address the reason for this difference in the radiosensitizing effects of the examined HDACis, we conducted cell cycle analysis. It is known that carbon ion irradiation and HDACi treatment induce G2/M and G0/G1 arrest, respectively, with HDACis strongly activating the expression of the cyclin-dependent kinase inhibitor p21cip1/waf1, resulting in the induction of G1 arrest.21–24) Although sub-G1 population (Fig. 4) is not in accord with Annexin V-positive apoptotic cells (Fig. 2), it is not surprising because those data represent the difference between mid-late apoptosis (sub-G1) and early apoptosis (Annexin V-positive/7-AAD-negative cells). Our data demonstrated that SAHA significantly induced G0/G1 arrest compared with FK228 at the concentrations of HDACis used in this study. Therefore, since G1 phase cells are sensitive to radiation due to a lack of DNA homologous recombination repair,25) this may explain the greater sensitizing effects of SAHA to carbon ion irradiation.
The difference in G0/G1 arrest induction between SAHA and FK228 may also be related to differences in their target molecules. There are 18 HDACs, which are subdivided into four classes (classes I to IV) based on their homology to yeast HDACs, with class II being further subdivided into classes IIa and b. It is considered that FK228 inhibits class I, VPA inhibits classes I and IIa, and TSA and SAHA inhibit classes I, II, and IV.26) Because each of these HDACs has specific functions, the blocking of different HDACs may have different effects. Indeed, although TSA and VPA had less of an effect on B16F10 melanoma cells in the present study (Fig. 2c), it has previously been reported that TSA sensitizes squamous carcinoma cells to gamma-ray irradiation via G1 arrest,27) and VPA significantly sensitizes retinoblastoma cells to X-ray irradiation.28) Thus, each HDACi has different effects according to the cell line used and acts on different target molecules. This means that although the sensitivity of melanoma cells to carbon ion irradiation was significantly enhanced by pretreatment with SAHA via G1 arrest (Fig. 4), other types of cancer cells may be sensitized to carbon ion irradiation by pretreatment with different HDACis.
In general, activation of the tumor suppressor gene p53 leads to activation of the p21cip1/waf1 gene. The expressed p21 protein then binds to cyclin-dependent kinase (CDK) complexes and inhibits their activity by causing G1 arrest of cell cycle. However, when p53 is inactivated by mutation or deletion, cells with DNA damage proliferate due to progression of the cell cycle. Therefore, aberrant p53 has been attracting attention recently as a target for antitumor strategies. Nevertheless, it has been reported that the transcription activation of p21cip1/waf1 by HDACis is p53-independent.29,30) Consequently, we predict that HDACi treatment can sensitize many types of cancer cells to carbon ion irradiation via G1 arrest regardless of the p53 genotype. Indeed, murine melanoma B16F10 cells have wild-type p53 and showed sensitizing effects to carbon ion irradiation in this study. Thus, we propose that the use of HDACis as radiosensitizers that induce G1 arrest may have p53-independent clinical efficacy in refractory cancer in combination with carbon ion irradiation, although it remains to have to consider various differences between humans and mice.
In conclusion, we investigated the combined effects of HDACis and carbon ion irradiation on the murine melanoma B16F10 cell line. We demonstrated that HDACis enhanced apoptosis and the suppression of clonogenicity induced by irradiation, having a greater effect when used in combination with carbon ions than with gamma-rays. Carbon ion irradiation and HDACi treatment induced G2/M and G0/G1 arrest, respectively, in B16F10 cells. It is considered that HDACis enhanced the killing effects of carbon ion irradiation against melanoma cells by inducing G1 arrest. Thus, the use of HDACis as radiosensitizers may become a promising approach for carbon ion therapy.
We thank Ms. Megumi Harada for technical support about Western blotting. We thank Drs. Tetsuya Sakashita, Michiyo Suzuki, and Masahiro Kikuchi for their useful advice. We are also grateful to the operators of TIARA, and to the operators of the Cobalt-60 irradiation facility at TARRI in QST for their technical cooperation. This work was supported in part by Grants-in-Aid for Challenging Exploratory Research [JSPS KAKENHI Grant Number 26670566] and Scientific Research (C) [JSPS KAKENHI Grant Number 23591627] from the Japan Society for the Promotion of Science.
The authors declare no conflict of interest.