Postradiation Sensitization of the Histone Deacetylase Inhibitor Valproic Acid

Abstract

Purpose

Preclinical studies evaluating histone deacetylase (HDAC) inhibitor-induced radiosensitization have largely focused on the preirradiation setting based on the assumption that enhanced radiosensitivity was mediated by changes in gene expression. Our previous investigations identified maximal radiosensitization when cells were exposed to HDAC inhibitors in both the preradiation and postradiation setting. We now expand on these studies to determine whether postirradiation exposure alone affects radiosensitivity.

Experimental Design

The effects of the HDAC inhibitor valproic acid (VA) on postirradiation sensitivity in human glioma cell lines were evaluated using a clonogenic assay, exposing cells to VA up to 24 h after irradiation. DNA damage repair was evaluated using γH2AX and 53BP1foci and cell cycle phase distribution was analyzed by flow cytometry. Western blot of acetylated γH2AX was done following histone extraction on AUT gels.

Results

VA enhanced radiosensitivity when delivered up to 24 h after irradiation. Cells accumulated in G₂-M following irradiation, although they returned to baseline at 24 h, mitigating the role of cell cycle redistribution in postirradiation sensitization by VA. At12 h after irradiation, significant γH2AX and 53BP1foci dispersal was shown in the control, although cells exposed to VA after irradiation maintained foci expression. VA alone had no effect on the acetylation or phosphorylation of H2AX, although it did acetylate radiation-induced γH2AX.

Conclusions

These results indicate that VA enhances radiosensitivity at times up to 24 h after irradiation, which has direct clinical application.

Histone proteins organize DNA into regular repeating structures of chromatin. The acetylation status of histones, modulated by histone acetyltransferases and histone deacetylases (HDAC), influences chromatin structure, which in turn regulates gene expression. Alterations in the baseline activity of these enzymes have been shown to play a role in multiple aspects of oncogenesis. Specifically, histone deacetylation has been implicated in such processes as tumor differentiation, proliferation, and metastasis. Based on these findings, targeting these pathways has been an active area of investigation and HDAC inhibitors have rapidly emerged as a novel class of molecularly targeted anticancer agents (1–3). Although HDAC inhibitors have shown antitumor activity as a single agent, what makes these agents particularly attractive is their multitargeted mechanism of action, allowing them to potentially enhance the response of other therapeutic modalities, including chemotherapy, radiation therapy, and other molecularly targeted agents (2, 4).

Numerous reports have shown the capacity of HDAC inhibitors to modulate radiation response (5), although mechanisms contributing toward this favorable interaction have not been defined. Whereas DNA repair has been implicated, transcriptional control of genes involved in DNA response following HDAC inhibition was initially regarded as the underlying mechanism. Based on this rationale, studies testing the capacity of HDAC inhibitors to enhance radiation response involved preirradiation exposure, with treatment times ranging from 1 to 48 h before irradiation (6–15). However, our previous investigations examining the optimal in vitro protocol have shown that preirradiation and postirradiation exposure to HDAC inhibitors results in the most sensitization, suggesting that additional effects, other than gene expression, may be involved (7).

As noted above, HDAC inhibitor effects on gene expression are mediated through changes in chromatin structure, which can independently influence repair of radiation-induced DNA damage (16). Moreover, HDACs have been shown to function late in the DNA repair process, purportedly being involved in chromatin restoration and termination of the repair process (16). The standard radiosensitization protocol of adding the HDAC inhibitor before irradiation does not allow for delineating the potential contribution of gene expression from these later events in the DNA repair process. To better understand the mechanisms involved in HDAC inhibitor-induced radiosensitization and in an attempt to reduce the likely contribution of changes in gene expression, we have investigated the effects of the HDAC inhibitor valproic acid (VA) on glioma cell radiosensitivity when added after irradiation. As described here, VA added at times up to 24 h after irradiation resulted in glioma cell radiosensitization. These findings may have a direct influence on the clinical development of HDAC inhibitors as radiosensitizers. For example, potential limitations for their practical application in the clinical setting, including i.v. formulations and/or short activity requiring coordination of drug delivery 1 to 2 h before daily radiation treatments, may be mitigated.

Materials and Methods

Cell lines and treatment

The U251 and SF539 (human malignant glioma) cell lines were obtained from the Division of Cancer Treatment and Diagnosis Tumor Repository (National Cancer Institute, Frederick, MD) and grown as described (17). VA (sodium salt; Sigma) was dissolved in PBS to a stock concentration of 100 mmol/L and stored at −20°C. Cultures were irradiated using a Pantak X-ray source at a dose rate of 2.8 Gy/min.

Clonogenic assay

To evaluate radiosensitivity, specified numbers of cells were seeded into the individual wells of a six-well tissue culture plate and radiation was delivered 6 h later. At this time, cells had attached but had not yet divided. Cells were exposed to VA at indicated concentrations and time points. Plates were incubated for colony formation for 10 to 14 d. Unless otherwise noted, medium was left unchanged during the duration of the experiment. Colonies were stained with crystal violet, the number of colonies containing at least 50 cells was determined, and surviving fractions were calculated. Survival curves were generated after correcting for cell killing from VA alone.

Cell cycle analysis

Cells were harvested at specified times following irradiation, pelleted by centrifugation, and resuspended in PBS containing 50 μg/mL propidium iodide, 0.1% Triton X-100, and 0.1% sodium citrate. Samples were stored at 4°C for 16 h and vortexed before fluorescence-activated cell sorting analysis (FACScan, FL-3 channel; BD PharMingen). The cell cycle calculations were done with the MultiCycle Program from Phoenix Flow Systems.

Immunofluorescent cytochemistry

Cells were grown and treated with 1.5 mmol/L VA for 24 h. At specified times, medium was aspirated and cells were fixed in 4% paraformaldehyde for 10 min at room temperature. Paraformaldehyde was aspirated, and the cells were rinsed with PBS followed by treatment with 0.1% NP40 in PBS for 15 min on ice. Cells were washed in PBS twice and blocked with 5% bovine serum albumin in PBS for 30 min following which anti-γH2AX or 53BP1 antibody (Upstate Biotechnology) was added at a dilution of 1:300 in 5% bovine serum albumin in PBS and incubated overnight at 4°C. Cells were then washed twice in PBS before incubating in the dark with a FITC-labeled secondary antibody at a dilution of 1:300 in 5% bovine serum albumin in PBS for 30 min. The secondary antibody solution was then aspirated, and the cells were washed four times in PBS. Cells then were incubated in the dark with 4′,6-diamidino-2-phenylindole (1 μg/mL) in PBS for 5 min and coverslips were mounted with an antifade solution (Dako). Slides were then examined on a Leica fluorescent microscope. Images were captured by a charge-coupled device camera. For each treatment condition, γH2AX foci were counted by eye in at least 50 cells from the stored images. To define a γH2AX-positive cell, the number of foci/cell was counted in 100 cells from the control group; a cell was determined to be positive if it had more foci than the average foci/cell in the control group.

Histone extraction

Approximately 5 × 10⁶ cells cultured in the presence or absence of 1.5 mmol/L VA were harvested, collected by centrifugation at 300 × g for 10 min, washed once with PBS, and suspended in ice-cold lysis buffer [20 mmol/L Tris-HCl (pH 7.9), 150 mmol/L NaCl, 1 mmol/L EDTA, 0.1% NP40, 1 mmol/L phenyl-methylsulfonyl fluoride, 1 × protease cocktail Complete]. Cells were Dounce homogenized and histones were collected by centrifugation at 1,000 × g for 10 min and washed thrice with lysis buffer. The pellet was suspended in 5 volumes of ice-cold 0.2 N H₂SO₄. After incubation at 4°C for 2 h, the suspension was centrifuged for 5 min at 15,000 rpm and the supernatant was taken and mixed with 6 volumes of ice-cold acetone, incubated overnight at −20°C, and centrifuged for 5 min at 15,000 rpm and pellet was air dried. The acid-soluble histone fraction was dissolved in original pellet volume of 4 mol/L urea. Protein concentrations were determined by the Bio-Rad detergent-compatible protein assay.

Gel electrophoresis

SDS-PAGE was done as described by Laemmli (18). Electrophoresis with 5% acetic acid, 6 mol/L urea, and 4.6 mmol/L Triton X-100 (AUT) with a final concentration of 12% to 18% acrylamide was carried out according to Bonner et al. (19). Briefly, 100 mg of the extracted histones were loaded into the AUT gel and electrophoresed at 11 mA for 16 h. Before transferring onto nitrocellulose membranes for Western blot analysis, the AUT gels were washed on a shaker for 30 min with 50 mmol/L Tris base, containing 0.5% SDS. The proteins were transferred onto nitrocellulose membrane positioned at the anode side of the gel, with transfer buffer [25 mmol/L Tris, 192 mmol/L glycine, 0.1% SDS (pH 8.3), 20% (v/v) methanol], in a Bio-Rad Mini Transblot apparatus at constant 100 V for 2 h. The membrane was blocked with 5% nonfat dry milk and probed with various antibodies. Enhanced chemiluminescent detection was done using enhanced chemiluminescent reagents according to the vendor’s protocols (Santa Cruz Biotechnology).

Results

The initial aim of these studies was to investigate HDAC inhibitor-induced enhancement of radiation response in the postradiation setting. Toward this end, VA (1.5 mmol/L) was added to the culture medium of the human glioma cell line U251 immediately, 6 h, or 24 h after exposure to graded doses of X-rays and clonogenic survival was determined. In this treatment protocol, VA was not removed from the culture medium during the colony-forming period (10 days); the surviving fraction from VA only was reduced to ~0.53 in U251 cells. Survival curves were generated after normalizing for the cell killing induced by VA alone and are shown in Fig. 1. Treatment with VA immediately after irradiation enhanced U251 cell radiosensitivity with a dose enhancement factor (DEF) at a surviving fraction of 0.1 of 1.46 (Fig. 1A). Delivery of VA at 6 and 24 h after irradiation also resulted in the enhancement of U251 cell radiosensitivity with DEFs of 1.38 and 1.4, respectively (Fig. 1B and C). To compare the sensitization induced by postirradiation VA exposure with that induced when VA was present only before irradiation, U251 cells were exposed for 16 h to VA (1.5 mmol/L) and irradiated followed by replacement with VA-free medium, and clonogenic survival was determined (Fig. 1D). This protocol resulted in a DEF of 1.3, consistent with the degree of radiosensitization reported for other cell lines treated with HDAC inhibitors before irradiation (6–8). When VA was not removed from the medium after irradiation (before and after exposure), the DEF was 1.71, which is also consistent with previous data (7). Thus, whereas maximum radiosensitization seems to be obtained through the exposure to HDAC inhibitor both before and after irradiation, these data suggest that exposure to VA at times out to 24 h after irradiation results in an enhancement in radiosensitivity.

Fig. 1.

The effects of VA on tumor cell radiosensitivity. U251cells were seeded in six-well tissue culture plates, and radiation was delivered 6 h later. Medium was replaced with medium containing VA (1.5 mmol/L) immediately (A), 6 h (B), and 24 h (C) following radiation in U251cells. D, U251cells were exposed to VA (1.5 mmol/L) for 16 h and then irradiated. Following irradiation (IR), cells were either rinsed with PBS and cultures were replaced with fresh growth medium (16 h before) or placed in the incubator without rinse (16 h before and after). Colonies were determined 10 to 14 d later, and survival curves were generated after normalizing for cell killing by VA alone. DEF was calculated at a surviving fraction of 0.1. Columns, mean from three independent experiments; bars, SD.

To determine whether the ability of VA to enhance radiosensitivity when delivered after irradiation was specific for U251 cells, these studies were extended to the human malignant glioma cell line SF539 (Fig. 2). In these experiments, because SF539 is more resistant to VA, a concentration of 3 mmol/L was used, which alone reduced the surviving fraction to 0.47. Survival curves were generated after normalizing for the cell killing induced by VA alone. As shown in U251 cells, the addition of VA at 6 and 24 h after irradiation enhanced SF539 radiosensitivity, with DEFs at a surviving fraction of 0.2 of 1.62 and 1.5, respectively (Fig. 2).

Fig. 2.

The effects of VA on tumor cell radiosensitivity in the postirradiation setting. SF539 cells were seeded in six-well tissue culture plates, and radiation was delivered 6 h later. Cells were then exposed to VA (3 mmol/L) 6 h (A) and 24 h (B) following radiation. Colonies were determined 10 to 14 d later, and survival curves were generated after normalizing for cell killing by VA alone. DEF was calculated at a surviving fraction of 0.2. Columns, mean from three independent experiments; bars, SD.

HDAC inhibitors have been reported to have more cytotoxic activity against cells in the G₂-M phase of the cell cycle (20). Given that radiation can activate cell cycle checkpoints, the putative enhancement in radiosensitivity induced by VA when delivered after irradiation may actually reflect the redistribution of cells into a VA-sensitive phase of the cell cycle. To address this possibility, cell cycle phase distribution was determined for U251 cells following exposure to 2 Gy. As expected, a G₂-M arrest was elicited following radiation, which was maximal at 6 h, and returned to baseline levels by 24 h (Table 1). However, as shown in Fig. 1, greater than additive increases in cell killing were detected when VA was added at 0, 6, and 24 h after irradiation. Thus, these data indicate that redistribution of cells into a VA-sensitive phase of the cell cycle does not account for the greater than additive increase in cell killing.

Table 1.

The influence of radiation on cell cycle phase distribution

	% G1	% S	% G2-M
Control	44.9	48.4	6.7
1 h	40.2	38.9	20.9
6 h	27.6	42.4	30.0
12 h	37.3	36.2	26.5
24 h	57.7	33.7	8.6

NOTE: U251 cells were exposed to 2 Gy radiation and, at specified time points, collected for flow cytometry analysis of cell cycle phase distribution. Data shown are representative from two independent experiments.

To begin to address the mechanism through which VA delivered after irradiation enhances cell killing, analysis of phosphorylated histone H2AX was used as an indicator of DNA damage. At sites of radiation-induced DNA double-strand breaks (DSB), the histone H2AX becomes rapidly phosphorylated (γH2AX), forming readily visible foci. Previous studies using preirradiation treatment protocols showed that VA, as well as other HDAC inhibitors, had no effect on the initial level of irradiation-induced γH2AX foci formation, although it delayed the dispersal of γH2AX foci, suggesting an inhibition of DSB repair (6, 10, 13). To determine whether VA delivered after irradiation affected γH2AX dispersal, VA was added to culture medium at 6 h after irradiation and γH2AX foci were determined 6 h later as an initial evaluation of γH2AX kinetics (Fig. 3A and C). As previously shown for U251 cells, after 2 Gy, nearly all cells expressed γH2AX foci 6 h following irradiation (6). Within 12 h after irradiation, γH2AX foci dispersal was shown in the control group, with nearly a 50% reduction in γH2AX-positive cells. However, cells exposed to VA 6 h after irradiation maintained γH2AX foci expression at 12 h, which was equivalent to levels expressed at 6 h after irradiation. As γH2AX facilitates the accumulation and retention of DNA damage response proteins (16), an initial investigation in determining if persistent γH2AX foci expression in cells exposed to VA was still associated with its repair machinery, immunofluorescent cytochemistry was done evaluating 53BP1 foci expression. The 53BP1 has been shown to participate in the cellular response of DNA damage and form part of the H2AX complex, likely functioning in mitotic checkpoint signaling (21, 22). As shown in Fig. 3B and C, following 2 Gy, 53BP1 showed similar kinetics as γH2AX, suggesting that 53BP1 accumulates and remains within the H2AX foci, which may translate to an incomplete resolution of repair machinery.

Fig. 3.

Exposure to VA after irradiation results in delayed repair kinetics. U251cells were exposed to 2 Gy (t = 0 h) followed by either no further treatment (IR alone) or addition of VA (1.5 mmol/L; t = 6 h). Micrographs were obtained at baseline, 6 h, and 12 h after irradiation and evaluated for γH2AX foci (A) or 53BP1 (B). C, foci evaluated in 200 nuclei per treatment per experiment in two independent experiments. The 6-h point was deemed maximum foci formation (87.2% and 82.8% foci-positive cells for γH2AX and 53BP1, respectively). A significant increase in foci present 12 h after irradiation was observed in cells exposed to VA (P = 0.0053 and 0.0438 for γH2AX and 53BP1, respectively).

Whereas γH2AX is typically evaluated as a marker of DSBs, given that it is a histone, it may also be subjected to hyper-acetylation as a result of HDAC inhibitor exposure. To determine whether VA affects the acetylation status of γH2AX, isolated histones were separated on AUT gels and then subjected to immunoblotting. As shown in Fig. 4, γH2AX is clearly visible at 1 h after exposure to 5 Gy and is substantially reduced by 24 h, which is consistent with the time course of γH2AX foci analyses. The capacity of VA to prolong expression of γH2AX after irradiation is also consistent with foci analysis. Although VA alone had no effect on either the acetylation or phosphorylation of H2AX, an acetylated γH2AX band was clearly visible at 1 h after irradiation and, although reduced in intensity, remained intact at 24 h. Although the significance of γH2AX acetylation following VA exposure remains unclear, with the known chromatin changes associated with the acetylation of core nucleosomal histones, it is tempting to speculate that persistent γH2AX acetylation may have an effect on chromatin remodeling and/or restoration during the DNA repair process.

Fig. 4.

VA acetylates radiation-induced γH2AX. AUT gels were used to resolve acetylated γH2AX (top arrow) from γH2AX (bottom arrow) in U251cells exposed to VA (1.5 mmol/L) for 24 h at specified time points following radiation (5 Gy).

Discussion

The capacity of HDAC inhibitors to enhance radiation response has been recognized by several preclinical investigations, although mechanisms have not been identified (5). Defining specific processes contributing to HDAC inhibitor-induced radiosensitization would aid in their clinical development as radiation sensitizers. For example, the majority of preclinical findings would support administering the HDAC inhibitor 1 to 2 h before daily radiation treatments, which may make daily delivery challenging, especially when testing i.v. formulations. Data presented here indicate that the HDAC inhibitor VA enhanced tumor cell radiosensitivity when delivered immediately and up to at least 24 h after irradiation. These findings may support an alternative, and perhaps more clinically practical, scheduling strategy for future trials designed to evaluate HDAC inhibitors as radiation sensitizers.

Several investigators have used DNA microarray to generate transcriptional profiles to specifically determine the functional consequence of HDAC inhibition on gene expression (23–26). In these studies, only 8% to 10% of genes were modulated by HDAC inhibitors, reiterating the complex nature of transcriptional control. Importantly, Glaser et al. (24) evaluated gene expression profiles in two breast and one bladder cell line. Over the several thousands of genes analyzed whose expression was changed in the individual cell lines following HDAC inhibition, there were only 13 overlapping between the three cell lines. Based on the broad capacity of HDAC inhibitors to enhance radiation response in multiple cell lines across tumor types and what seems to be a cell line–specific control of gene transcription, it seems unlikely that the influence of HDAC inhibitors on transcription per se is the sole mechanism mediating the enhanced radiation response.

Although chromatin remodeling is influenced by HDAC inhibitors, it is also a component of radiation response. For example, higher-order chromatin packaging is a barrier to the detection and repair of DNA damage (16). Studies have shown that DSBs cause chromatin reorganization and induce a local decrease in the density of the chromatin fibers. These remodeling events allow for the area of damage to be more accessible to repair machinery by exposing features of the nucleosome, such as constitutive histone modifications, which are normally concealed in the unperturbed cell (16, 27). Following recruitment of DNA repair proteins, the individual strand breaks are usually repaired rapidly, typically within 2 h (28–30). However, following repair, it is necessary for chromatin to be restored to its original state and signals are required to terminate these repair processes. Interestingly, HDACs have been shown to function late in the DNA repair process, purportedly being involved in chromatin remodeling to its original state and terminating the repair process (16). Based on the presented data, which show continued radiation sensitization up to 24 h after irradiation, it is tempting to speculate that mechanisms underlying HDAC inhibitor sensitization may involve these later events in chromatin remodeling rather than the repair of individual base pairs.

The capacity of VA to acetylate core histone proteins H3 and H4 has been well documented (2), although their influence on H2AX, the histone intimately involved in DNA damage signaling, has not been evaluated. Our studies showed VA exposure results in the acetylation of radiation-induced γH2AX. The particular relevance of acetylation of H2AX and γH2AX is unclear. Most studies have focused on the posttranslational modulation of H2AX during early stages of DNA repair. H2AX has been shown to be acetylated following DNA damage and a required event for its ubiquitination and subsequent release from chromatin in DNA DSB response (31). Persistent acetylation of radiation-induced γH2AX in VA-treated cells may affect repair complex dispersal, leading to persistent recruitment of DNA damage repair proteins. 53BP1 expression displayed similar kinetics as γH2AX when cells were exposed to VA, consistent with the incomplete resolution of repair machinery. This could also reflect the failure to complete chromatin restoration, which would be one of the final steps of DSB repair, consistent with VA-induced radiosensitization at 24 h.