Histone deacetylase inhibitors promote glioma cell death by G2 checkpoint abrogation leading to mitotic catastrophe

Abstract

Glioblastoma multiforme is resistant to conventional anti-tumoral treatments due to its infiltrative nature and capability of relapse; therefore, research efforts focus on characterizing gliomagenesis and identifying molecular targets useful on therapy. New therapeutic strategies are being tested in patients, such as Histone deacetylase inhibitors (HDACi) either alone or in combination with other therapies. Here two HDACi included in clinical trials have been tested, suberanilohydroxamic acid (SAHA) and valproic acid (VPA), to characterize their effects on glioma cell growth in vitro and to determine the molecular changes that promote cancer cell death. We found that both HDACi reduce glioma cell viability, proliferation and clonogenicity. They have multiple effects, such as inducing the production of reactive oxygen species (ROS) and activating the mitochondrial apoptotic pathway, nevertheless cell death is not prevented by the pan-caspase inhibitor Q-VD-OPh. Importantly, we found that HDACi alter cell cycle progression by decreasing the expression of G2 checkpoint kinases Wee1 and checkpoint kinase 1 (Chk1). In addition, HDACi reduce the expression of proteins involved in DNA repair (Rad51), mitotic spindle formation (TPX2) and chromosome segregation (Survivin) in glioma cells and in human glioblastoma multiforme primary cultures. Therefore, HDACi treatment causes glioma cell entry into mitosis before DNA damage could be repaired and to the formation of an aberrant mitotic spindle that results in glioma cell death through mitotic catastrophe-induced apoptosis.

Figure 1

SAHA and VPA reduce glioma cell viability, proliferation and clonogenicity. (a) Dose−response of glioma cell lines U87-MG, U373-MG and LN229 to VPA (upper graph) and SAHA (lower graph) treatment by WST1- cell viability assay. Cells were treated for 48 h with increasing concentrations of HDACi. The LC50 for each cell line was estimated by interpolation in the line-graph. Data shown are mean±S.E.M. from five independent experiments. Statistical comparisons were performed against the most sensitive cell line, U87-MG. (b) Cell viability of glioma cell lines after 48- h treatment with 10 mM VPA, 10 μM SAHA or left untreated. Cells were counted in a hemocytometer by trypan blue exclusion. Data are mean±S.E.M. from five independent experiments. (c) Analysis of cell proliferation by Ki-67-positive nuclei counting after 48 h of HDACi treatment (10 mM VPA or 10 μM SAHA). Data in the graphic are expressed as mean±S.E.M. of four independent experiments. (d) Clonogenic assay of glioma cell lines after 48-h treatment with HDACi (see Materials and Methods for details). Representative images of the stained clones are shown on the left. Bar graph shows results obtained from four independent experiments (Mean±S.E.M.). Statistical analysis were performed by the Student's T-test being *P<0.05; **P<0.01; ***P<0.001

Figure 2

VPA and SAHA induce nuclear condensation and caspase activation in glioma cell lines. (a) Nuclear morphology of U87-MG, U373-MG and LN229 after 24-h treatment with 10 mM VPA, 10 μM SAHA or 100 ng/ml TRAIL. After treatment, cells were fixed in 4% PFA and nuclei were stained with Hoechst 33342. Condensed and abnormal nuclei are indicated with yellow arrowheads. Scale bar= 50 μm. (b) Analysis of caspase-9, caspase-3 and fodrin expression and cleavage on glioma cell lines after 24-h treatment with 10 mM VPA and 10 μM SAHA by western blot. Membrane was reprobed with an anti-α-tubulin antibody to verify equal loading. A representative blot from three independent experiments is shown. (c) Caspase-3 activity in HDACi-treated cells. Glioma cells were treated with 10 mM VPA, 10 μM SAHA alone or in combination with the caspase inhibitor Q-VD-OPh (5 μM) for 24 h. Cell lysates were incubated with Ac-DEVD-AFC fluorogenic substrate up to 8 h, and its cleavage was measured hourly using a fluorometer. Bars depict mean±S.E.M. from four independent experiments. Statistical analysis was performed using the Student's T-test by comparing treated versus non-treated (NT) cells (*P<0.05, **P< 0.01)

Figure 3

HDACi promote DNA fragmentacion in glioma cell lines, which is dependent on caspase activation. (a) DNA fragmentation analysis on glioma cells treated for 48 h with 10 μM SAHA combined or not with the pan-caspase inhibitor Q-VD-OPh (5 μM) by PFGE (HMWF, upper picture) and ladder (LMWF, lower picture) procedures. As a positive control, LN229 cells were treated with 1 μM STS for 48 h. (b) Analysis of the effect of BCL-X overexpression on DNA fragmentation (HMWF and LMWF) induced by 10 mM VPA or 10 μM SAHA on U251-MG glioma cells. Bcl-xL expression was checked by western blot (upper panel). (c) Downregulation of CAD expression with specific shRNA (4 and 5) blocks the formation of 50- Kb fragments and low-molecular weight fragmentation on U251-MG cells treated for 48 h with 10 μM SAHA. (d) Western blot of CAD on U251-MG cells transduced with lentiviral vectors containing CAD shRNA or scrambled sequence (Scr). Equal loading was verified by GAPDH detection of the same membrane (lower panel). Each DNA integrity assay shown in a, b and c was repeated three times and a representative image is shown. (e) Cell viability analysis by cell counting using trypan blue exclusion on glioma cells treated with 10 μM SAHA or 5 μM Q-VD-OPh or both combined for 48 h. Bars depict mean±S.E.M. from four independent experiments. Statistical analysis was performed by the Student's T-test, no significant differences (n.s.) were obtained by comparing SAHA and SAHA+ Q-VD-treated cells

Figure 4

HDACi promote DNA damage by DSBs formation in glioma cell lines. (a) Analysis of DNA DSB formation by immunofluorescence of γH2AX-positive nuclei in glioma cells treated for 48 h with 10 mM VPA, 10 μM SAHA or 100 μM TMZ. Data are mean±S.E.M. from five independent experiments. (b) WST-1 cell viability assay on glioma cell lines exposed to increasing concentrations of TMZ (from 10  to 500 μM) for 48 h. Data shown are mean±S.E.M. from three independent experiments. (c) Analysis of γH2AX-positive nuclei on glioma cells treated with 5 μM Q-VD-OPh, 10 μM SAHA or both combined for 48 h. Bars depict mean±S.E.M. from four independent experiments. (d) ROS production measured by flow cytometry on glioma cells treated with 10 mM VPA, 10 μM SAHA or with 10 μM SAHA in the presence of 10 mM GSH. Data are mean±S.E.M. from four independent experiments. (e) Analysis of γH2AX-positive nuclei on glioma cells treated with 10 μM SAHA or 100 μM TMZ in the presence or absence of 15 mM N-acetyl-L-cysteine (NAC). Bars depict mean±S.E.M. from four independent experiments. Statistical analysis was performed by the Student's T-Test, *P<0.05, **P< 0.01, ***P<0.001 in comparison with NT cells or ^##P<0.01 in comparison with SAHA-treated cells

Figure 5

HDACi treatment of glioma cell lines alters cell cycle progression, decreasing the percentage of cells in S phase and increasing cell population in G2/M. (a) Cell cycle analysis by flow cytometry of glioma cell lines treated for 24 h with 10 mM VPA, 10 μM SAHA or left untreated. Bar graph shows the mean of the percentage of cell in G0/G1, S and G2/M phases with error bars (S.E.M.) from three independent experiments. Plots on the right are representative results from one experiment. (b) Percentage of apoptotic cells (Annexin V positive and PI negative) and dead cells (PI positive) after 24 h of treatment. Results are mean±S.E.M. from four independent experiments. Statistical analysis was performed by the Student's T-test and significance is shown by *P<0.05; **P<0.01

Figure 6

SAHA and VPA reduce the levels of expression of G2 checkpoint gatekeepers Wee1 and Chk1. (a) Analysis of the effect of HDACi on Wee1 and Chk1 protein expression and on the phosphorylation status of cdc2 (Tyr-15) in glioma cells. Cell cycle regulators, Cyclin B1, p21 and cdc25A, were also analyzed. Cells were treated for 24 h with 10 mM VPA, 10 μM SAHA or 100 μM TMZ, and protein extracts were analyzed by electrophoresis and western blot using specific antibodies against proteins indicated on the left of the panels. Representative blots of three independent experiments are shown. (b) Analysis of Wee1 and Chk1 mRNA after 24 h of HDACi treatment of glioma cell lines by reverse transcription and quantitative real-time PCR. Bars depict the mean±S.E.M. of four independent experiments. (c) Expression of Mus81 endonuclease in glioma cells treated as described above by western blot. Equal loading was verified by α-tubulin detection on the same membrane. (d) Downregulation of the endonuclease Mus81 do not prevent DSBs formation in SAHA-treated U251-MG cells. The efficacy of two specific shRNA against the mRNA of Mus81 (1 and 2) was verified by western blot (upper panel). Percentage of γH2AX-positive cells after 48 h of incubation in the presence of 10 μM SAHA or 100 μM TMZ of lentivirus-transduced cells. Data are mean±S.E.M. from three independent experiments. (e) Analysis of DNA fragmentation after Mus81 downregulation with shRNA. Representative gels of three independent experiments are shown, of high-molecular weight DNA fragmentation (upper panel) and low-molecular weight DNA fragmentation (lower panel) analysis

Figure 7

HDACi promote cell death by mitotic catastrophe. (a) Analysis of U251-MG cells in metaphase after SAHA treatment by immunofluorescence of Survivin (in green, left panel) and TPX2 (in green, middle panel), α-tubulin (in red, left and middle panels), γH2AX (in red, right panel) and nuclear staining with Hoescht 33342 (in blue, all panels). Images were acquired using an Olympus IX-70 confocal microscope and representative pictures from three independent experiments are shown. (b) SAHA and VPA reduce the expression levels of Rad51, Survivin and TPX2 in glioma cell lines. Total cell lysates from glioma cells were analyzed by electrophoresis and western blot using specific antibodies against the proteins indicated on the left. Representative blots from three independent experiments are shown. (c) Analysis of the effect of SAHA on Wee1, Chk1, TPX2 and Survivin protein expression in human GBM primary cultures. Representative blots are shown from three independent experiments. Equal loading was checked by membrane staining with naphtol blue. (d) Caspase 2 protein and mRNA expression in HDACi-treated glioma cell lines. After a 24-h treatment with 10 mM VPA or 10 μM SAHA, cells were processed for western blot analysis of Caspase-2 (upper panel) or RNA was extracted to perform quantitative RT-PCR (lower table). Data are shown as a ratio versus non-treated cells, and the mean±S.E.M. from three independent experiments is summarized. Statistical analysis were performed by the Student's T-test being *P<0.05; **P<0.01. (e) Caspase-2 activity assay on U251-MG cells treated for 24 h with 10 mM VPA, 10 μM SAHA in the presence or absence of 5 μM Q-VD-OPh. Cell lysates were incubated with Ac-VDVAD-AFC fluorogenic substrate up to 8 h and its cleavage was measured hourly using a fluorometer. Bars depict mean±S.E.M. from three independent experiments. Student's T-test was performed by comparing data from treated versus non-treated (NT) cells (*P<0.05, ***P< 0.001)