The HDAC inhibitor LBH589 induces ERK-dependent prometaphase arrest in prostate cancer via HDAC6 inactivation and down-regulation

Abstract

Histone deacetylase inhibitors (HDACIs) have potent anti-cancer activity in a variety of cancer models. Understanding the molecular mechanisms involved in the therapeutic responsiveness of HDACI is needed before its clinical application. This study aimed to determine if a potent HDACI, LBH589 (Panobinostat), had differential therapeutic responsiveness towards LNCaP and PC-3 prostate cancer (PCa) cells. The former showed prometaphase arrest with subsequent apoptosis upon LBH589 treatment, while the latter was less sensitive and had late G2 arrest. The LBH589 treatment down-regulated HDAC6 and sustained ERK activation, and contributed to prometaphase arrest. Mechanistically, LBH589 inhibited HDAC6 activity, caused its dissociation from protein phosphatase PP1α, and increased 14-3-3ζ acetylation. Acetylated 14-3-3ζ released its mask effect on serine 259 of c-Raf and serine 216 of Cdc25C subsequent to de-phosphorylation by PP1α, which contributed to ERK activation. Enhanced ERK activity by LBH589 further down-regulated HDAC6 protein levels and sustained ERK activation by free-forward regulation. The sustained Cdc25C and ERK activation resulted in early M-phase (prometaphase) arrest and subsequent apoptosis in the most sensitive LNCaP cells but not in PC-3 cells. This study provides pre-clinical evidence that HDAC6 may serve as a sensitive therapeutic target in the treatment of prostate cancer with HDACI LBH589 for clinical translation. This study also posits a novel mechanism of HDAC6 participation in regulating the c-Raf-PP1-ERK signaling pathway and contributing to M phase cell-cycle transition.

Figure 1. HDACI induced G2/M phase cell cycle and growth inhibition in prostate cancer cells via distinct mechanisms.

(A) LBH589 had a dose-dependent effect on growth inhibition in prostate cancer cell lines. The cell lines were treated with 50, 75, 100, and 150 nM of LBH589. Cell survival was measured using MTT assay as described in the Materials and Methods section. Graphs compared the percentage of growth inhibition to vehicle control at 24 (upper) and 48 (lower) hours. (B) LBH589 induced G2/M cell cycle arrest. All cell lines were treated with LBH589 or vehicle control for 24 h. The cell cycles were analyzed by PI-staining and flow cytometry according to DNA content. The increased percentages of G2/M cells were compared to that of vehicle control. (C) LBH589 induced apoptosis in PC-3 and LNCaP cells. Both cell lines were treated with 75nM LBH589 or vehicle control for 24 (upper) and 48 (lower) h. The percentages of apoptotic cells were analyzed by counting the population of DNA fragmentation by flow cytometry. (D–E) HDACIs induced Aurora kinase degradation found in PC-3 but not in LNCaP. PC-3 and LNCaP cells were treated with different HDACIs at indicated concentrations of LBH589, SAHA and SBHA for 24 h and cell lysates were analyzed by western blotting using the indicated antibodies. (F) Profile of proteins regarding progression of G2/M cell cycle. The cells were treated with different dosages of LBH589 for 24 h. The lysate was analyzed by western blotting.

Figure 2. LBH589 induced ERK activation in LNCaP but not PC-3 cells.

(A) LBH589 induced ERK activity is selectively activated in LNCaP cell. PC-3 and LNCaP cells were treated with 75nM LBH589 or vehicle control for 24 h. The levels of indicated proteins were immuno-blotted against individual antibodies. GAPDH was an internal loading control. (B) Dose- and time-dependent correlations of LBH589-mediated ERK activation and Cdc25C hyper-phosphorylation. LNCaP was exposed to various concentration of LBH589 for 24 h (left) or 25 nM LBH589 at different times (right). (C) Dose-dependent correlation of LBH589-induced G2/M arrest. LNCaP was treated with various concentrations of LBH589 for 24 h. Cell cycles were analyzed by flow cytometry. (D) Cdc25C hyper-phosphorylation was confirmed by dephosphorylation assay. LNCaP was treated with 50nM LBH589 for 24 hours. The lysates were incubated with phosphatase (CIP) or combined with phosphatase inhibitor. The hyper-phosphorylated and dephosphorylated Cdc25C were analyzed by immuno-blotting with Cdc25C antibody. (E) Cdc25C hyper-phosphorylation was suppressed by MEK inhibitors. LNCaP were treated with MEK inhibitors, PD (100 µM PD98059) or UO (20 µM UO126) for 30 minutes. LBH589 was then added to the culture after 24 h. ERK activity and Cdc25C hyper-phosphorylation were analyzed by Pi-ERK and Cdc25C antibodies in western blotting.

Figure 3. LBH589 induced prometaphase arrest in LNCaP but not in PC-3 cells.

(A–B) Immuno-staining of LNCaP and PC-3. Both cells were treated 75 nM LBH589 for 24 h. Immuno-staining (A) with Cdc2 and Cdc25C antibodies and (B) of γ-tubulin (Green) and DAPI (Blue) in vehicle control (left) and LBH589 treatment (right) in LNCaP. (C–D) MEK inhibitor attenuated LBH589-induced prometaphase arrest in LNCaP. The cells were pre-treated 30 with UO126 for 30 minutes and sequentially treated with LBH589 for 24 h. (C) The cell cycles were analyzed by PI-staining and flow cytometry. (D) The metaphase cells were counted by DAPI staining presented as condensate chromatin.

Figure 4. HDAC6 down-regulation correlated with ERK activation in LBH589 treatment.

(A) Screening of HDACs involved in ERK activation. The immuno-blottings of LBH589 treated cell lysates were performed with indicated antibodies. (B) LBH589 induced the down-regulation of HDAC6, which correlated with ERK activation in LNCaP but not in PC-3. (C) Knockdown of HDAC6 increased in ERK activity. 293T was transfected with siRNA of HDAC1, 3, 6, or non-targeting for 72 hours. The lysates were analyzed by immuno-blotting. (D) ERK activity was involved in LBH589-mediated HDAC6 down-regulation. Cell lysates from cells treated with LBH589 or combined with UO126 pre-treatment were immuno-blotted.

Figure 5. LBH589 induced ERK activation by modulating c-Raf activity.

(A) Analysis of ROS production. The indicated cells were treated with 75 nM LBH589 for 24 h. ROS was measured as described in the Materials and Methods section. (B) The pattern of c-Raf signaling pathway on LBH589 treatment. The cells were treated with LBH589 for 24 h and the lysates were immuno-blotted with the indicated antibodies. α-catenin was a loading control.

Figure 6. LBH589 switched the PP1 interacting partner for ERK activation.

(A) LBH589 induced Cdc25C-S216 dephosphorylation in a dose- and time-dependent manner. LNCaP was exposed to 75 nM LBH589 for various times or at indicated LBH589 concentrations for 24 h. Phosphorylation of Cdc25C-S216 was performed by immuno-blotting with Pi-Cdc25C-S216 antibody. (B–C) LBH589 treatment switched the PP1α interacting partner. The immuno-precipitations were conducted with anti-HDAC6 or anti-14-3-3ζ. The precipitated samples were analyzed by immuno-blotting using antibody against PP1α, 14-3-3ζ, Lys-Ac (Pan-acetylated lysine). IgG was a nonspecific pull-down control.

Figure 7. The model of LBH589-induced prometaphase arrest.

LBH589 binding to HDAC6 might cause conformational change on HDAC6, leading to the dissociation of PP1α and enhancement of 14-3-3ζ acetylation. Acetylated 14-3-3ζ increased its interacting affinity with PP1α and interfered in the affinity of PP1α binding with substrate. PP1α then dephosphorylated S259 of c-Raf and S216 of Cdc25C, triggering ERK activation. Constant ERK activation due to the activation of c-Raf might destabilize HDAC6 proteins and hyper-phosphorylate CDC25C, leading to the prometaphase cell cycle arrest of LNCaP cells.