Valproic acid defines a novel class of HDAC inhibitors inducing differentiation of transformed cells

Abstract

Histone deacetylases (HDACs) play important roles in transcriptional regulation and pathogenesis of cancer. Thus, HDAC inhibitors are candidate drugs for differentiation therapy of cancer. Here, we show that the well-tolerated antiepileptic drug valproic acid is a powerful HDAC inhibitor. Valproic acid relieves HDAC-dependent transcriptional repression and causes hyperacetylation of histones in cultured cells and in vivo. Valproic acid inhibits HDAC activity in vitro, most probably by binding to the catalytic center of HDACs. Most importantly, valproic acid induces differentiation of carcinoma cells, transformed hematopoietic progenitor cells and leukemic blasts from acute myeloid leukemia patients. More over, tumor growth and metastasis formation are significantly reduced in animal experiments. Therefore, valproic acid might serve as an effective drug for cancer therapy.

Fig. 1. HDAC inhibitor-like activation of PPARδ by VPA. (A) A cell line expressing the ligand-binding domain of PPARδ fused to the DNA-binding domain of the glucocorticoid receptor (GR) together with a GR-controlled reporter gene was treated for 40 h with the PPARδ ligand cPGI₂ (5 µM), VPA or the HDAC inhibitors sodium butyrate (But) and TSA (300 nM). Reporter gene activity was monitored by enzymatic assay for alkaline phosphatase. Values were normalized between experiments according to cPGI₂-induced activity. (B) A cell line overexpressing full-length GR was tested as a control. Dexamethasone (1 µM) was used as a GR-specific ligand. Values are means ± SD from duplicate determinations in 2–5 independent experiments.

Fig. 2. VPA relieves HDAC-mediated transcriptional repression. (A) HeLa cells were co-transfected with a UAS-TK-luciferase reporter, an SV40 β-Gal control reporter and vectors expressing either the GAL4 DNA-binding domain (amino acids 1–147) or GAL4 fusions of N-CoR (amino acids 1–2453) or the ligand-binding domains of TRβ (amino acids 165–456) with or without TRIAC, or PPARδ (amino acids 138–440) with or without cPGI₂, respectively. At 24 h after transfection, cells were treated with trapoxin (10 nM), TSA (100 nM) or VPA (1 mM) for 16–20 h. Subsequently, cells were harvested and luciferase and β-galactosidase activities were measured. Results are represented as fold repression relative to GAL4. Values are means ± SD from triplicate determinations in two independent experiments. (B) Phoenix cells were co-transfected with the indicated expression vectors, using a luciferase reporter based on the human RARβ2 promoter (de Thé et al., 1990). VPA (1 or 3 mM) was added 18 h after transfections, and cells were analyzed for reporter activity after an additional 24 h.

Fig. 3. VPA induces accumulation of hyperacetylated histone and inhibits HDAC activity. (A) HDAC inhibitors induce the accumulation of hyperacetylated histones H3 and H4. Both the time course and the required concentration for VPA-induced hyperacetylation were determined by western blot analysis of whole-cell extracts from F9 cells treated with VPA (1 mM if not indicated otherwise) in comparison with TSA (100 nM) and sodium butyrate (NaBu, 5 mM). Treatment was for 12 h or as indicated. Equal loading was confirmed by Coomassie Blue staining. Experiments were performed three times with similar results also in HeLa cells. (B) Histone hyperacetylation in vivo was determined by western blot analysis of histones H3 and H4 from mouse splenocyte nuclear extracts. Three mice each were injected i.p. with 25 ml/kg body weight of 155 mM solutions of NaCl or sodium valproate. Due to the short half-life of VPA in rodents, another dose (50%) was readministered after 5 h. Extracts were prepared 10 h after the initial dose. (C) HDAC activity was determined by the release of [³H]acetate from hyperacetylated radiolabeled histones. Activities were determined in the presence of the indicated compounds in immune precipitates from HEK293T cell extracts with antibodies directed against N-CoR, mSin3 or HDAC2. The HDAC activity which precipitated with a non-related immune serum (NI) was determined for control. Values are presented relative to the activity in the absence of HDAC inhibitors. The 100% values normalized for ∼1 mg of extract in representative experiments correspond to 1000 (N-CoR), 500 (mSin3) and 300 c.p.m. (HDAC2). Data are means ± SD from three independent experiments. (D) HDAC activity was determined in immune precipitates from F9 cell extracts with antibodies directed against N-CoR and in N-CoR-depleted extracts. Efficiency of N-CoR depletion was assessed by western blot for N-CoR in the IP pellet as well as in equivalent amounts of whole-cell extracts before and after depletion (data not shown). (E) IC₅₀ values were calculated as those concentrations required for 50% inhibition of [³H]acetate release. HDAC assays were performed using immune precipitates from F9 cell extracts with antibodies directed against HDACs 2, 5 or 6. HDACs 5 or 6 were precipitated from extracts which had been depleted with antibodies directed against N-CoR, mSin3 and HDACs 1–3.

Fig. 4. Binding of VPA to corepressor or HDAC2 immunoprecipitates. Antibodies directed against N-CoR, mSin3 and HDAC2 co-precipitate substantial HDAC activity, most probably in the form of multiprotein complexes (Figure 3C). Immunoprecipitates from HEK293T cell extracts were incubated with 2 µCi of ³H-labeled VPA in the presence or absence of TSA (100 nM). Radioactivity retained after washing is shown. Values are means ± SD from three independent experiments carried out in duplicate.

Fig. 5. HDAC inhibition by compounds related to VPA. (A) VPA and the related compounds EHXA, and R- and S-4-yn-VPA at the indicated concentrations were tested for HDAC inhibitory activity. Addition of TSA (100 nM) to the reaction served as a control. The assays were performed with N-CoR immunoprecipitates from HEK293T cells in duplicate (untreated enzyme activity 2205 c.p.m. = 100%). Precipitates of a pre-immune serum served as a negative control. (B) Accumulation of hyperacetylated histones H3 and H4 in F9 cells treated with compounds related to VPA was determined as described in the legend to Figure 3. Cells were treated for 12 h with 1 mM of VPA, R- or S-EHXA, R- or S-4-yn-VPA or VPD. One representative out of two similar experiments is shown.

Fig. 6. VPA induces cell differentiation and apoptosis in carcinoma cell lines and inhibits tumor growth and metastasis formation in the rat. (A) F9 teratocarcinoma cells were treated for 48 h with TSA (30 nM) or VPA (1 mM). The appearance of AP-2 protein as a specific marker of the VPA-induced type of F9 cell differentiation was as described (Werling et al., 2001). One out of two experiments with similar results is shown. (B) Thymidine incorporation into cultures of HT-29 colonic cancer or MT-450 breast carcinoma cells was tested as a parameter for cell proliferation. Cells were cultured for 72 h in the absence or presence of the indicated concentrations of VPA prior to analysis of [³H]thymidine incorporation. The graph shows the means ± SD from quadruple determinations. Similar results were obtained in two additional independent experiments. (C) The appearance of apoptotic cells in VPA-treated cultures of MT-450 cells was analyzed by staining of cell surface-exposed phosphatidylserine by FITC-conjugated annexin V after exclusion of necrotic cells by means of propidium iodide uptake (lower right quadrant of the graphs). Similar results were obtained in a second experiment. (D) Subcutaneous tumor growth and lung metastasis of MT-450 breast cancer cells were analyzed in rats treated with VPA or saline, respectively. Tumor volume was determined at day 21 (onset of VPA treatment) and day 43 (termination of experiment). Lung metastasis visible from the organ surface was scored. Scores: 0, no metastasis; 1, <50 metastases or all metastatic nodules <2 mm in diameter (lower frames); 2, many metastases >2 mm in diameter (upper right frame); 3, most of the lung’s surface consists of metastatic nodules (upper left frame). Values are means ± SD and significance of differences was calculated by Student t-test. (E) Pictures of two preparations each representative for control or VPA-treated rats are shown. The original height of the frames is 25 mm.

Fig. 7. VPA relieves the differentiation block in Kasumi cells and in a murine PML–RAR model. (A) Kasumi-1 cells were treated for 5 days with VPA (1 mM) or TSA (50 nM) in the presence or absence of RA (1 µM). Viable cells were identified by trypan blue dye exclusion and the percentage of differentiated cells was determined by counting NBT-positive cells. Values are means ± SD from three experiments. (B) Lin^– cells were isolated and transduced with the indicated vectors (control, PINCO vector encoding GFP alone; PML–RAR, expression of GFP and PML–RAR). GFP⁺ cells were sorted by FACS and seeded in methylcellulose plates in differentiation medium. Cultures were kept in the absence or presence of VPA (0.2 and 1 mM) for 8–10 days before differentiation was analyzed by expression of the myeloid differentiation marker Mac-1. (C) Lin^– cells were plated in semi-solid medium in the presence of cytokines (IL-3, IL-6, SCF, G-CSF and GM-CSF) and the indicated concentrations of VPA or TSA. After 7 days, colonies were counted. (D and E) Lin^– cells were grown for 36 h in liquid medium in the presence of cytokines, and then analyzed for the presence of apoptotic cells, or for cell cycle status after propidium iodide staining.

Fig. 8. VPA relieves the differentiation block in blast cells from AML patients. (A) Leukemic blasts of AML patients were cultured for 5 days in the absence or presence of RA and VPA (1 mM). Differentiation was evaluated by morphological criteria (+ 10–20%; ++ 20–40%; +++ 50–80%; ++++ 80–100% more mature metamyelocytes and granulocytes than control cultures) and by determination of the percentage of NBT-positive cells. Classification according to the French–American–British (FAB) nomenclature is given. Patients 1 and 2 represent newly diagnosed AML patients, whereas blasts of patients 3 and 4 were analyzed at relapse after chemotherapy. Patient 5 developed an AML after a refractory anemia with excess of blasts (RAEB). NN indicates a normal karyotype in leukemic blasts. ND, not determined. (B) Wright–Giemsa-stained primary blasts from the bone marrow of a newly diagnosed AML patient (patient 1 of A) were treated in culture for 5 days with RA (1 µM), VPA (0.2 or 1 mM), TSA (240 nM) as single agents or with a combination of RA and either VPA or TSA.