Histone Deacetylase (HDAC) Inhibitor LBH589 Increases Duration of γ-H2AX Foci and Confines HDAC4 to the Cytoplasm in Irradiated Non–Small Cell Lung Cancer

DNA methylation and histone deacetylation are the major epigenetic pathways identified that regulate gene expression (1). Histone modification plays an important regulatory role in many cellular processes, including gene transcription (2, 3) and proliferation (4). Transcriptionally active genes have high levels of core histone acetylation, whereas transcriptionally inactive genes have low levels (5, 6). The balance between these two states is regulated by two enzymes: histone deacetylase (HDAC) and histone acetyltransferase. It is estimated that 4% to 12% of genes are regulated by histone acetylation/deacetylation (7). Abnormal transcriptional repression occurs in many forms of cancer, and HDACs have been implicated in this process (7). For this reason, HDACs have been identified as molecular targets in cancer therapy (7–10). Previous studies have shown HDAC inhibition to result in cell growth arrest, differentiation, apoptosis, and alterations in gene expression in cancer cell lines (11, 12).

In addition to promoting growth arrest by themselves, HDAC inhibitors have been shown to have synergistic effects with cytotoxic cancer therapy (13–16). This synergy is related to the inhibition of DNA repair (17) and synthesis (18). Due to its ability to enhance the effects of anticancer therapy and the potential low side effect profile, HDAC inhibitors are a promising new class of antineoplastic pharmaceuticals.

Non–small cell lung cancer (NSCLC) is often refractory to therapy due to late presentation and relative resistance to nonsurgical treatment modalities. Surgery remains the treatment of choice with 5-year survival rates of 24% to 61% (19). Unresectable lung cancer shows resistance to chemotherapy and radiotherapy, and recurrences are common within the field of irradiation (20). Improved effectiveness of nonsurgical treatments has the potential to improve patient outcomes by decreasing recurrence rates and providing more effective therapy to nonsurgical candidates.

LBH589 is a novel cinnamic hydroxamic acid analogue HDAC inhibitor currently in early clinical trials (21). The compound has been shown to induce acetylation of histones H3 and H4 and hsp90, increase p21 levels, and induce G₁ cell cycle arrest in hematologic malignancies (22, 23). In the present study, the effects of LBH589 on irradiated NSCLC were studied. LBH589 enhanced radiation-induced apoptosis. Furthermore, HDAC inhibition increased the duration of γ-H2AX foci and limited nuclear localization of HDAC4 in irradiated human NSCLC cell lines.

Tumor model and cell culture. NCI-H23 (human lung adenocarcinoma, non–small cell) and NCI-H460 (human large cell lung cancer) cell lines were obtained from the American Type Culture Collection (Manassas, VA) and maintained in RPMI 1640 (Life Technologies, Carlsbad, CA) supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin. Cell lines were maintained in a 37°C, 5% CO₂ incubator. LBH589 was obtained from Novartis Oncology (East Hanover, NJ) and diluted in DMSO to 25 μmol/L stock solutions.

In vitro clonogenic assay. H460 cell lines were cultured, counted, and seeded onto plates at specific cell densities. After treatment with either 50 nmol/L LBH589 or ionizing radiation (IR), cells were returned to 37°C incubation for 10 days. Colonies were fixed for 15 minutes with 70% ethanol and then stained for 15 minutes with 1% methylene blue. After staining, colonies with at least 50 viable cells were counted with a low-power microscope. Surviving fraction was calculated as (mean colony counts) / (cells plated) × [plating efficiency (PE)], where PE was defined as (mean colony counts) / (cells plated for nonirradiated controls). Averages and SEs were calculated.

Apoptosis assays. Apoptosis was measured with Annexin-propidium iodide staining using the protocol from BD PharMingen (San Diego, CA). Briefly, H23 and H460 cell lines were treated with 25 nmol/L LBH589 and 3 Gy irradiation. Twenty-four hours later, cells were suspended, washed twice with PBS, and adjusted to 10⁶ cells/mL in binding buffer. Cell solution (100 μL) was transferred to a 5-mL tube and stained with Annexin V-FITC and propidium iodide for 15 minutes. Binding buffer (400 μL) was added to the samples, and the samples were analyzed immediately by flow cytometry.

In separate experiments, cells were assayed for pyknotic nuclei after staining with 4′,6-diamidino-2-phenylindole (DAPI). Near-confluent H23 and H460 cells were treated with or without 25 nmol/L LBH589, incubated for 18 hours, and then irradiated with 3 Gy. Cells were then fixed and stained with DAPI after 20 hours. One hundred cells from each group were counted by an observer who was blinded to the treatment of each culture. Pyknotic nuclei were recorded, and the percentage of cells undergoing apoptosis were determined from three separate experiments. The mean and SE were calculated for each treatment group.

Cell lysis and immunoblot analysis. H23 and H460 cell lines were treated with or without 25 nmol/L LBH589 and incubated for 18 hours. The cells were then irradiated with 3 Gy and incubated for 6 hours. Cells were washed twice with PBS followed by 180 μL lysis buffer (20 nmol/L Tris, 150 mmol/L NaCl, 1 mmol/L EDTA, 1% Triton X-100, 2.5 mmol/L NaPP_i, 1 mmol/L phenylmethylsulfonyl fluoride, 1 μg/mL leupeptin). Protein concentration was quantified by the Bio-Rad method (Hercules, CA). Total protein (40 μg) was loaded into each well of a 12% SDS-PAGE gel and separated. Protein was transferred onto a polyvinylidene difluoride (PVDF) membrane and probed with antibodies for actin and cleaved caspase-3 (Cell Signaling, Boston, MA) in a 5% milk TBS-Tween 20 (TBST) solution followed by goat anti-rabbit (Sigma, St. Louis, MO) in a 0.5% milk TBST solution. Membranes were exposed to film and later developed.

Tumor volume study. H460 cells (10⁶) were suspended in 0.1 mL of cell medium and then injected into the hind limb of mice. Tumors were allowed to form for 1 week. The mice were then divided into one of four groups: control, IR, 40 mg LBH589, and 40 mg LBH589 plus IR. Groups receiving 40 mg LBH589 were treated twice during the 7-day course of IR. LBH589 and control vehicle were administered via oral gavage. Mice were irradiated 1 hour following drug administration with 2 Gy (X-ray generator) given as five fractions, once daily over a total of 7 days. Tumor volume measurements were taken every 2 to 3 days for the duration of the study, and the average and SE were calculated. The Institutional Animal Care and Use Committee guidelines were followed during all aspects of treatment.

Cell immunostaining. H23 and H460 cell lines were cultured onto slides and then treated with 25 nmol/L LBH589 for 18 hours followed by 3 Gy. For HDAC4 immunostaining, cells were processed 2 hours after IR; for γ-H2AX immunostaining, cells were processed at the indicated times. Cells were washed with cold PBS and fixed with 3.7% paraformaldehyde for 5 to 10 minutes and then with 100% methanol. Slides were washed with PBS and incubated with primary antibody to HDAC4 (Novus Biologicals, Littleton, CO) or γ-H2AX (Novus Biologicals) at a 1:200 dilution overnight at 4°C followed by rhodamine red–labeled secondary or Alexa Fluor 488 secondary (Molecular Probes, Inc., Carlsbad, CA) antibody at a 1:700 dilution. Slides were counterstained with DAPI as described above and analyzed by microscopy. Photographs were taken and scanned into Adobe Photoshop. Positive γ-H2AX staining cells were counted by an observer blinded to treatment conditions, and the mean and SE were determined.

Cytoplasmic and nuclear fractionalization. H460 cells were grown to 90% confluency, washed thrice with PBS, and then transferred to Eppendorf tubes. Hypotonic lysis buffer (200 μL; 10 mmol/L HEPES, 5 mmol/L KCl, 1.5 mmol/L MgCl₂, 0.08% NP40) was added to the cells for 15 minutes. The solution was then centrifuged at 3,000 × g for 5 minutes. The supernatant was collected, and the pellet was washed. Dignin buffer (20 mmol/L HEPES, 1.5 mmol/L MgCl₂, 0.2 mmol/L EDTA, 420 mmol/L NaCl, 50 mmol/L β-glycerophosphate, 25% glycerol) was then added to the pellet for 30 minutes. The solution was centrifuged at 13,000 × g for 10 minutes, and the supernatant was collected. Protein concentration was then calculated as stated above. Western blot analysis of both cytoplasmic and nuclear fractions was done with PAGE and transfer to PVDF membranes. Immunostaining was done using anti-HDAC4, anti-lamin A (Cell Signaling), and anti-actin (Sigma) primary antibodies.

Statistical analysis. The mean and SE were calculated using Microsoft Excel. The Student's t test was used to compare treatment groups. Ps < 0.05 were considered statistically significant.

LBH589 sensitizes NSCLC to the cytotoxic effects of IR. Clonogenic analysis of H460 cell lines was done using LBH589 and 0 to 6 Gy. Figure 1 shows the mean surviving fraction and SE for each treatment group (n = 3). Untreated control cells showed substantial radioresistance, with 6 Gy resulting in only one log reduction in survival. Use of LBH589 1 and 18 hours before IR resulted in a synergistic decrease in colony survival compared with untreated cells as evident by an increase in the negative slope of the dose-response curve. Treatment with LBH589 alone for 18 hours resulted in a significantly reduced PE, whereas treatment for 1 hour alone had no reduced PE compared with the control. These data show that LBH589 enhances the cytotoxic effects of IR in NSCLC cell lines.

LBH589 enhances radiation-induced apoptosis. To study the effect of HDAC inhibition by LBH589 on apoptosis, three in vitro experiments were done. Figure 2 shows results from Annexin V-FITC/propidium iodide flow cytometry analysis of apoptosis. Use of 25 nmol/L LBH589 before 3 Gy significantly increased the number of apoptotic cells from 7% to 30% for H23 cell line (P < 0.001) and from 6% to 25% for H460 cell line (P = 0.003) compared with control. Use of IR alone or LBH589 alone produced only a minimal increase, and the effect of the combined treatment was greater than what would be predicted by an additive effect.

To confirm the ability of LBH589 to sensitize human lung cancer cell lines to radiation-induced apoptosis, nuclear morphology studies were done. Figure 3 shows the mean percentage and SE of pyknotic nuclei determined by DAPI staining. Use of 25 nmol/L LBH589 18 hours before 3 Gy significantly increased the percentage of pyknotic nuclei to >10% for H23 (P < 0.001) and H460 (P = 0.042) cell lines. Untreated H23 and H460 cells had <1% apoptotic nuclei, H23 and H460 cells treated with 3 Gy had 3% and 2% apoptotic nuclei, and H23 and H460 cells treated with LBH589 alone had 4% and 2% apoptotic nuclei, respectively.

Cleavage of caspase-3 was analyzed to verify the role of apoptosis in cells treated with LBH589 and radiation. Western blot analysis was done on H23 and H460 whole-cell lysates. Figure 4 shows the Western immunoblots for cleaved caspase-3 and actin. An increase in caspase-3 cleavage was evident in both H23 and H460 cell lines following treatment with LBH589. Use of LBH589 before IR increased levels of caspase-3 cleavage in H460 cells. This increase, however, was not as prominent in the H23 cell line.

LBH589 enhances tumor growth delay in vivo. H460 cells were injected into the hind limb of mice. After tumor formation, the mice were treated with two oral doses of 40 mg LBH589 and/or five 2 Gy fractions over 7 days. Figure 5 shows the fold increase in tumor volume (Fig. 5A) and the tumor growth delay (Fig. 5B) for each treatment group. Use of LBH589 alone resulted in a modest but significant tumor growth delay of 2 days (P < 0.001). IR alone delayed growth by ∼4 days (P < 0.001). Combined treatment significantly delayed tumor growth by ∼20 days (P < 0.001), indicating that HDAC inhibition enhances the effects of IR on NSCLC tumor growth. In addition, the mice receiving LBH589 showed minimal signs of toxicity during the course of the study as monitored by weight loss and mobility.

LBH589 prolongs the duration of radiation-induced γ-H2AX foci. Immunostaining was done to study γ-H2AX foci present at DNA double-strand breaks (DSB). Figure 6A shows representative photographs of the H23 cell line treated with combinations of LBH589 and IR. The green staining of γ-H2AX foci is shown. Three grays induced γ-H2AX foci as early as 30 minutes following treatment. These foci disappeared by 6 hours in cell lines treated with IR alone. Use of LBH589 alone for 20 hours resulted in a modest increase in γ-H2AX foci. In comparison, LBH589 added 18 hours before IR prolonged the duration of γ-H2AX foci for up to 24 hours after IR (42 hours after LBH589 administration). Furthermore, γ-H2AX foci were seen at 18 and 24 hours after IR in cells undergoing apoptosis. Interestingly, no γ-H2AX foci were present in cells undergoing apoptosis following treatment with radiation without HDAC inhibition. Similar results were seen in the H460 cell line (Supplementary Fig. S1).

We next studied the rate of γ-H2AX foci resolution. Figure 6B shows the number of γ-H2AX foci present 24 hours after IR. Treatment with 3 Gy alone and LBH589 alone resulted in rapid resolution of γ-H2AX (<5% at 24 hours). Use of LBH589 before IR significantly delayed the resolution of γ-H2AX foci with 60% residual foci in both cell lines at 24 hours (P < 0.001). The increased duration of γ-H2AX foci following treatment with LBH589 and IR indicates that HDAC inhibition disrupts the DNA repair process, and this mechanism potentially sensitizes NSCLC to the cytotoxic effects of radiation.

HDAC4 nuclear translocation in irradiated lung cancer cell lines. Immunostaining of HDAC4 was done on H23 and H460 cell lines to identify the effect of LBH589 on HDAC4 compartmentalization. Figure 7 shows representative photographs of the H460 cell line probed with anti-HDAC4 antibodies and rhodamine-labeled secondary antibodies (red) and then counterstained with DAPI (blue). Untreated cells and cells treated with LBH589 alone showed background HDAC4 staining in the cytoplasm and nucleus. When H460 cells were treated with 3 Gy, HDAC4 localized to the nucleus at 2 hours and minimal HDAC4 was present in the cytoplasm. However, LBH589 added before IR markedly limited HDAC4 nuclear localization. A similar effect was seen in the H23 cell line (Supplementary Fig. S2). These results were confirmed in the H460 cell line using anti-HDAC4 antibodies for Western blot analysis of cytoplasmic and nuclear proteins (Fig. 7B).

Previous studies have shown radiosensitization with HDAC inhibitors in several cell lines, including melanoma (17), squamous (18), prostate (24), and colon carcinoma (14–17). In the current study, this radiosensitizing effect was shown in NSCLC cell lines in clonogenic studies. Interestingly, an effect was seen when LBH589 was administered 1 hour before IR. Although this enhancement was less than that of drug administered for 18 hours before IR, it is possible that the effects of LBH589 are in part due to a mechanism that does not require epigenetic changes. The biological effects of HDAC inhibitors were once focused on alteration of gene transcription through histone modification. Recently, many transcription-independent pathways affected by HDAC inhibitors have been identified. These pathways include histone-dependent processes, such as DNA repair (17, 24, 25) and cell cycle regulation (26–28), and histone-independent processes, such as activation of protein kinases, including AKT and extracellular signal-regulated kinase (23).

HDAC inhibition has been shown to induce apoptosis by affecting several pathways. Apoptotic genes activated by HDAC inhibitors include Bak, Bax, p53, caspase-9, caspase-8, caspase-3, Bid, Bad, etc. Antiapoptotic genes repressed by HDAC inhibitors include Bcl-2, C-FLIP, survivin, NF-κB, etc. (7, 13). The essential pathways for apoptosis induction continue to remain under debate; however, p21^WAF1 has been implicated to play an important role in certain tumor cell lines (29–31). In this study, low doses of LBH589 had synergy with IR in inducing apoptosis as evident by Annexin-propidium iodide flow cytometry and nuclear morphology. These data suggest that apoptosis is one mechanism of interaction between LBH589 and IR.

The effect of LBH589 on tumor growth was studied in vivo using human lung cancer xenografts in the nude mouse model. In this study, mice receiving two oral doses of LBH589 and five 2 Gy fractions over 7 days showed significant delay in tumor growth compared with either agent alone. Drug (40 mg/kg) was administered 1 hour before irradiation on days 1 and 4 of the study. The relatively high dose and ∼17 hour half-life (21) allowed for the drug to be theoretically available 1 to 2 days after administration. The mice receiving the drug exhibited few signs of toxicity. In the current experiment, there were no premature deaths and the mice in all groups had similar weights throughout the study. HDAC inhibitors are believed to be effective in enhancing both radiation and chemotherapy treatments while being safe for patients. Phase I clinical data from LBH589 showed minimal and tolerable side effects (21). Because of its potential as an anticancer agent and its low side effect profile, this class of pharmaceutical is promising for future cancer therapy.

Phosphorylation of H2AX (γ-H2AX) is an evolutionarily conserved response to DNA DSBs (32–35). The formation of γ-H2AX is necessary for the recruitment of many factors involved in DNA repair, including NBS1 (MRN complex) and 53BP1, both of which regulate cell cycle checkpoints (36). DNA DSBs induced by IR activate checkpoint pathways that inhibit progression of cells through G₁ and G₂ phases and delay the progression through S phase (36–38). Previous reports have shown the ability of HDAC inhibitors to prolong the duration of γ-H2AX in irradiated melanoma (17) and prostate (24) cell lines. Both of these reports attribute the radiosensitization ability of HDAC inhibitors to the inhibition of DNA repair. The mechanisms behind the inhibition of DNA repair by HDAC inhibitors are not clear. Previous reports have shown that altered gene expression of ATM (39), DNA-PKcs (40), Ku70 (41), and Ku80 (42) leads to radiosensitization. Furthermore, mutations of acetylatable lysine residues on histone H4 have been shown to increase the effect of DNA-damaging agents (43). A recent study by Tamburini and Tyler (44) showed that dynamic changes in histone acetylation accompany homologous recombination and the modulation of histone acetylation is necessary for cell viability. In addition, Kusch et al. (45) have shown that loss of the dTip60 complex increases the duration of γ-H2AX homologues following IR. The mechanism of HDAC inhibitors on DNA repair inhibition is potentially multifactorial, involving both transcriptional regulation of DNA repair genes and direct effects on the acetylation of repair complexes.

The importance of histone acetylation and deacetylation in DNA repair is emerging (46). Kao et al. have shown that IR induces HDAC4 foci formation in the nucleus and conclude that HDAC4 interacts with 53BP1 to mediate the response to DNA damage (47). In the present study, HDAC4 translocated to the nucleus in response to IR. LBH589 added before IR limited this radiation-induced nuclear localization. Inhibiting HDACs blocks the nuclear localization of complexes important in DNA repair, including HDAC4 and possibly 53BP1. DNA repair is also attenuated, and the cell is more susceptible to IR-induced apoptosis. Many proteins in addition to HDAC4 and 53BP1 have been shown to form nuclear foci following DNA damage, including BRCA1 (48), Rad51, and the NSB1/Mre11/Rad50 complex (49, 50). The effect of HDAC inhibitors on the localization and function of these complexes would provide further insight into the role of HDAC in DNA repair.

In this study, LBH589 was shown to sensitize NSCLC to the cytotoxic effects of IR. Furthermore, this compound was shown to be effective with oral administration in animal models. The IC₅₀ in the nanomolar range for LBH589, along with the relatively long half-life and low toxicity, makes this compound promising for future clinical trials. The tumor-selective effect of HDAC inhibitors indicates that these drugs are exciting candidates for cancer therapy.

Grant support: NIH grants RO1-CA112385, RO1-CA70937, RO1-CA88076, and RO1-CA89888; Vanderbilt Lung Cancer Specialized Programs of Research Excellence grant P50-CA90949; and Vanderbilt-Ingram Cancer Center grant CCSG P30-CA6848. K.C. Cuneo is supported by the Vanderbilt University Medical Scholars Program.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.