The prostate cancer blocking potential of the histone deacetylase inhibitor LBH589 is not enhanced by the multi receptor tyrosine kinase inhibitor TKI258

Summary Pharmacologic options for patients with castration- resistant prostate cancer are limited. It has been suggested that targeting intracellular molecules, which have been altered dur- ing neoplastic development, may slow tumor growth. There- fore, the growth-blocking potential of the histone deacetylase- inhibitor LBH589 and the multiple tyrosine kinase-inhibitor TKI258, applied alone or in combination, was investigated in a panel of prostate cancer cell lines. PC-3, DU-145 or LNCaP cells were treated with various concentrations of LBH589 and/ or TKI258. Tumor cell growth, cell cycle regulating proteins, HDAC3- and HDAC4-expression and histone H3 and H4 acetylation were then evaluated by MTT assay and Western blotting. LBH589 dose-dependently blocked prostate cancer cell growth. In contrast, TKI258 did not down-regulate tumor cell growth up to a 1,000 nM dosage. LBH589 elevated histone H3 and H4 acetylation. The cell cycle regulators cyclin B, cyclin D1, cdk1 and cdk4 were down-regulated in PC-3, where- as the suppressor proteins p21 and p27 were up-regulated in LNCaP by LBH589. TKI258 up-regulated p27 in PC-3 or p21 in LNCaP and additionally elevated cyclin B, cyclin D1, cdk1 and cdk4 in both cell lines. Presumably, the increase in cyclin and cdk caused by TKI258 counteracts the benefit of p21 or p27 up-regulation, resulting in TKI258 non-responsiveness. The LBH589/TKI258-combination was not superior to the LBH589 single-drug use in terms of growth reduction. Obvi- ously, TKI258 did not enhance the sensitivity of prostate cancer cells towards an HDAC based regimen. Therefore, the LBH589/TKI258-combination probably does not provide an optimum strategy in fighting advanced prostate cancer.

Keywords : Prostate cancer . LBH589 . TKI258 . Molecular therapy . Tumor growth

Introduction

Prostate cancer is the third most common cause of death in the industrialized nations [1, 2]. In Europe more than 300,000 men are diagnosed with prostate cancer each year. In the United States 240,000 new cases and 33,000 deaths were recorded in 2011 [1]. Carcinoma of the prostate is predominantly found in older men and may be cured when localized, but 30–50 % of the patients develop local or distant recurrence.

Hormone deprivation therapy is still the gold standard for metastatic disease. However, after a median of 18–24 months, nearly all patients progress to androgen independence [3]. Chemotherapy for hormone-refractory prostate cancer (HRPC) is a palliative option and only modestly improves overall survival [4]. At this point there is no cure for metastatic prostate cancer.

Better understanding of the molecular mechanisms lead- ing to neoplastic transformation has led to the development of novel agents, targeting altered proteins found in tumor tissue. Based on clinical and pre-clinical studies, the concept of molecular targeting is becoming increasingly promising as a tool to treat advanced prostate cancer. It has recently been demonstrated that blocking the activity of histone deacetylases (HDAC) by HDAC-inhibitors strongly reduces prostate cancer growth and metastatic dissemination [5–7]. Pharmacologic interference with the HDAC-system might, therefore, be useful in counteracting tumor progression.

However, it has been argued that many cancers might not be sensitive enough to HDAC inhibitors when applied alone, necessitating the introduction of a second drug with a dif- ferent mode of action.Drake et al. have observed elevated tyrosine kinase sig- naling in advanced prostate cancer, making these proteins relevant treatment targets [8]. Several studies have evaluated the anti-tumor potential of tyrosine kinase inhibitors (TKI), including (among others) blockage of vascular endothelial growth factor receptor (VEGFr), epidermal growth factor receptor (EGFr), platelet-derived growth factor receptor (PDGFr), csf-1 or c-KIT [9–12].

Unfortunately, most published clinical trials report only moderate effects in a sub-cohort of patients. Targeting one pathway in the cancer cell therefore does not lead to a sustained response with meaningful tumor shrinkage. The purpose of the present study was to evaluate whether com- bined inhibition of HDAC and tyrosine kinase activity might more effectively inhibit prostate cancer growth than de-activating each pathway separately. The effect of the HDAC-inhibitor LBH589 (panobinostat) and TKI258, given alone or in combination, on a panel of prostate cancer cell lines was investigated. TKI258 (formerly CHIR-258) is a multiple TKI-inhibitor that targets VEGFr 1-3; PDGFr beta, and FGFr 1-3 [13].

Materials and methods

Cell cultures

Human PC cell lines PC-3, DU-145 and LNCaP were obtained from DSMZ (Braunschweig, Germany). Tumor cells were grown and subcultured in RPMI 1640 (Gibco/ Invitrogen; Karlsruhe, Germany), 10 % fetal bovine serum (FBS), 20 mM HEPES-buffer, 2 % glutamine and 1 % penicillin/streptomycin.

Drugs

LBH589 and TKI258 were provided by Novartis Pharma AG, Basel, Switzerland. Both drugs were dissolved in DMSO as 1 mM stock solution and stored in aliquots at −20 °C. Prior to the experiments, the compounds were diluted in cell culture medium to final concentration.

Measurement of tumor cell growth

Prostate carcinoma cells were treated either with LBH589 or with TKI258, or with both compounds. Controls remained untreated. To evaluate whether drug activity depends on the drug exposure time, three sets of experiments were done. 1) Cell growth analysis without LBH589/TKI258 pre-incubation:

Cell cultures were treated with LBH589 and/or TKI258 and immediately subjected to the cell growth assay for 24, 48 and 72h (indicated as “no pre-incubation”). 2) Cell growth analysis with a 3 day LBH589/TKI258 pre-incubation: Cell cultures were pre-treated with LBH589 and/or TKI258 for 3 days. The medium was then replaced by fresh medium (including LBH589 and/or TKI258), and cells were subjected to the cell growth assay for 24, 48 and 72 h (indicated as “3 days pre-incubation”). 3) Cell growth analysis with a 5 day LBH589/TKI258 pre-incubation: Cell cultures were pre-treated with LBH589 and/or TKI258 for 5 days. The medium was then replaced by fresh medium (including LBH589 and/or TKI258), and cells were subjected to the MTT assay for 24, 48 and 72 h (indicated as “5 days pre-incubation”).

Cell growth was assessed using the 3-(4,5-dimethylthia- zol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) dye re- duction assay (Roche Diagnostics, Penzberg, Germany). Treated versus non-treated PC-3, DU-145 or LNCaP cells (0.5 ×104 cells/well) were seeded onto 96-well tissue culture plates. After 24, 48 and 72 h, MTT (0.5 mg/ml) was added for an additional 4 h. Thereafter, cells were lysed in a buffer containing 10 % SDS in 0.01 M HCl. The plates were incubated overnight at 37 °C, 5 % CO2. Absorbance at 570 nm was determined for each well using a microplate ELISA reader. Each experiment was done in triplicate. After subtracting background absorbance, results were expressed as mean cell number.

Apoptosis and necrosis

To exclude toxic effects of the compounds, cell viability was determined by trypan blue (Gibco/Invitrogen). For apoptosis detection the expression of Annexin V/propidium iodide (PI) was evaluated using the Annexin V-FITC Apoptosis Detection kit (BD Pharmingen, Heidelberg, Germany). Tumor cells were washed twice with PBS, and then incubated with 5 μl of Annexin V-FITC and 5 μl of PI in the dark for 15 min at RT. Annexin V binds with high affinity to the membrane phospholipid phosphatidyl- serine which is translocated from the inner to the outer leaflet of the plasma membrane during early apoptosis. Since the membranes are intact in early apoptosis, cells are impermeable to the vital dye PI, i.e. cells are FITC Annexin V positive and PI negative. Late apoptosis is characterized by a loss of membrane integrity, thus cells are both FITC Annexin V and PI positive. Cells were analyzed on a FACScalibur (BD Biosciences, Heidelberg, Germany). Early apoptotic cells were visualized in the lower right quadrant, late apoptotic cells in the upper right quadrant. The percentage of apoptotic cells (early and late) in each quadrant was calculated using Cell- Quest software (BD Biosciences).

Western blot analysis

To explore cell cycle regulating proteins, tumor cell lysates were applied to a 7 % polyacrylamide gel and electrophor- esed for 90 min at 100 V. The protein was then transferred to nitrocellulose membranes. After blocking with non-fat dry milk for 1 h, the membranes were incubated overnight with monoclonal antibodies directed against cell cycle proteins: Cdk1 (IgG1, clone 1), cdk4 (IgG1, clone 97), cyclin B (IgG1, clone 18), cyclin D1 (IgG1, clone G124-326), cyclin E (IgG1, clone HE12), Rb (IgG2a, clone 2), p21 (IgG1, clone 2 G12), p27 (IgG1, clone 57; all: BD Biosciences).

To investigate histone acetylation, cell lysates were marked with anti-HDAC3 (polyclonal IgG, dilution 1:2.000), anti-HDAC4 (polyclonal IgG; dilution 1:500), anti-histone H3 (IgG, clone Y173), anti-acetylated H3 (IgG, clone Y28, dilution 1:500), anti-histone H4 (polyclonal IgG) and anti-acetylated H4 (Lys8, polyclonal IgG, dilution 1:500; all from Biomol GmbH, Hamburg, Germany).

HRP-conjugated goat-anti-mouse IgG (Upstate Biotech- nology, Lake Placid, NY, USA; dilution 1:5.000) served as the secondary antibody. The membranes were briefly incubat- ed with ECL detection reagent (ECLTM, Amersham/GE Healthcare, München, Germany) to visualize the proteins and then analyzed by the Fusion FX7 system (Peqlab, Erlangen, Germany). β-actin (1:1.000; Sigma, Taufenkirchen, Germany) served as the internal control.

Statistics

All experiments were performed 3–6 times. Statistical signifi- cance was investigated by the Wilcoxon–Mann–Whitney-U- test. Differences were considered statistically significant at a p value less than 0.05.

Results

LBH589 but not TKI258 diminishes tumor cell growth

LBH589 significantly and dose-dependently blocked growth of all investigated prostate cancer cell lines (Fig. 1). There was no correlation between drug exposure time and growth impair- ment. The efficacy of LBH589 was quite similar with the 3 or 5 day “pre-incubation” assay. However, 1 nM LBH589 did not block PC-3 growth when it was added to the cell cultures and the MTT assay was started immediately thereafter (“no pre- incubation”), whereas the same concentration significantly diminished PC-3 growth following 3 or 5 days pre- incubation. Additionally, 1 nM LBH589 did not reduce the number of DU-145 and LNCaP cells in the “no pre- incubation” setting, whereas it did in the “pre-incubation” assays (compared to their respective controls).

TKI258 did not down-regulate tumor cell growth when applied at concentrations of 1–500 nM, independently of whether the cell populations were pre-treated for 3 or 5 days or not pre-treated. 1,000 nM TKI258 also did not alter PC-3 and LNCaP growth but slightly diminished the number of DU-145 cells (Fig. 2a, representative for the “3 day pre- incubation” assay, 500 and 1,000 nM values are shown).

Analysis of apoptotic events was also performed. TKI258 did not trigger early or late apoptosis in PC-3, DU-145 or LNCaP cells. LBH589 did not trigger early or late apoptosis in PC-3 and LNCaP cells. However, both early and late apoptosis were elevated in DU-145 cells, when chronically exposed to 25 or 50 nM LBH589 for 8 days (Fig. 2b).

LBH589-TKI258 combination versus mono drug treatment

The combination studies were based on the 25 nM LBH589 and the 500 and 1,000 nM TKI258 concentration. Figure 3 depicts percentage increase of the tumor cell number calcu- lated 72 h after drug incubation. 25 nM LBH589 signifi- cantly lowered the growth capacity of all evaluated tumor cell lines. 500 nM TKI258 had no effect on PC-3, DU-145 or LNCaP cell number. However, 1,000 nM TKI258 mod- erately suppressed DU-145 (but not PC-3) growth, com- pared to the controls. There was also a trend towards a diminished LNCaP cell number with 1,000 nM TKI258. The LBH589-TKI258 (500 nM) combination was not supe- rior to the LBH589 single drug regimen, whereas the LBH589-TKI258 (1,000 nM) combination reduced DU- 145 and LNCaP cell number more strongly than LBH589 alone.

Evaluation of intracellular signaling and histone acetylation

Molecular analysis was carried out on PC-3 and LNCaP cells. DU-145 cells were excluded, since no difference in the growth behavior compared to PC-3 cells was seen. LBH589 applied alone down-regulated cyclin B and D1 as well as cdk1 and cdk4 in PC-3 cells (Fig. 4). In contrast, cyclin B was not altered and cyclin B, cyclin D1, cdk1 and cdk4 were even slightly enhanced in LNCaP cells, com- pared to the untreated controls. As a further mode of action, LBH589 considerably elevated the protein content of p21 and p27 in LNCaP cells, whereas p27 was not modified by this compound in PC-3 cells. P21 was not detectable either in controls or in drug treated PC-3 cells. Distinct suppression of HDAC3 and 4 was found in both PC-3 and LNCaP cells exposed to LBH589. Additionally, LBH589 profoundly increased histone H3 and H4 acetylation in the tumor cells. Application of TKI258 augmented cyclin B and D1 as well as cdk1 and cdk4 in PC-3 cells and simultaneously forced the enhanced expression of p27. A similar effect was evoked on cyclin B, cyclin D1, cdk1 and cdk4 expression in TKI258 treated LNCaP cells. However, contrary to the PC-3 data, TKI258 triggered the up-regulation of p21 but not of p27 in the LNCaP population.Evaluation of the drug combination on protein synthesis revealed no additive effects in both PC-3 and LNCaP prostate cancer cell lines.

Fig. 1 Dose–response analysis. PC-3, DU-145 or LNCaP cells were treated with various LBH589 concentrations without and with 3 or 5 day pre-incubation. Controls remained untreated. Cells were plated out at 5,000 cells/well and counted after 24, 48 and 72 h using the MTT dye reduction assay. One representative experiment of six is shown. * indicates significant difference to controls.

Discussion

The present investigation deals with the influence of a TKI and/or HDAC-inhibitor on prostate cancer growth. Application of LBH589 as a clinically relevant dose [14] stopped tumor growth whereas TKI258 application did not. Elevation of histone acetylation by the HDAC-inhibitors PXD101 (belinostat) [15], valproic acid [6], suberoylanilide hydroxamic acid (SAHA; vorinostat), trichostatin A [16], FK228 (romidepsin) [17] or MS-275 (entinostat) [18] has been shown to delay cell cycle progression and induce apoptosis in prostate cancer cells. Considering that LBH589 also suppressed tumor growth and that all compounds mentioned previously are characterized by a different chemical structure, it seems likely that deactivating HDAC causes the tumor cell to become less malignant. Whether this holds true with respect to the alteration of further molecular targets is not yet clear. Valproic acid has been shown to diminish the cell cycle regulating proteins cdk1, cdk4 and cyclin B in PC-3 cells in the same manner as LBH589 does on PC-3 cells [6]. Inhibition of the cdk1/cyclin B complex was also induced in PC-3 cells by MS-275 [19]. SAHA treatment effectively down-regulated cyclin D1 and cdk4 in LNCaP [20]. The same action was observed in PC-3 cells, but not in the LNCaP population when exposed to LBH589. Since PC-3 cells are androgen-resistant and LNCaP cells are androgen-sensitive, modification of the intracellular signaling machinery following HDAC-inhibition may, there- fore, depend on the androgen receptor status. In fact, the molecular mode of action of valproic acid on prostate cancer cells varies with the tumor phenotype [21]. A novel report

Fig. 2 a Effect of TKI258 on tumor growth. PC-3, DU-145 or LNCaP cells were treated with various concentrations of TKI258. Controls remained untreated. Cells were plated out at 5,000 cells/well and counted after 24, 48 and 72 h using the MTT dye reduction assay. One represen- tative experiment of six is shown. * indicates significant difference to controls. b Effect of LBH589 on DU-145 cell apoptosis. Early and late points to the effectiveness of PXD101 in altering intracellular signaling in prostate cancer cells, which was dependent on the androgen receptor status [22].

Therapeutic manipulation of the tumor’s epigenome may offer a novel treatment strategy for prostate cancer patients. Clinical phase I and II trials with entinostat [23], vorinostat [24], valproic acid [25] or the novel HDAC-inhibitor SB939 [26] have demonstrated prolonged stable disease in patients with solid tumors. Nevertheless, the overall anti-tumor activity of each HDAC-inhibitor is less compelling, since objective responses have been limited to a small patient cohort and no signs of tumor regression have been observed. It is argued that addition of a further anti-cancer agent is necessary to optimize the HDAC-inhibitor based regimen. However, the way to achieve optimization is controversial.

Receptor tyrosine kinase signaling is dysregulated in many human cancers and it has been proposed that deacti- vating this pathway might slow tumor progression. The EGFr-selective TKI gefitinib or erlotinib has been shown to exert antiproliferative and pro-apoptotic effects, particu- larly in PTEN-positive, androgen-sensitive prostate cancer cell lines [27, 28]. Sunitinib and Sorafenib, both multikinase inhibitors of several targets, including VEGFr and PDGFr, apoptosis were evaluated by the Annexin V-FITC Apoptosis Detection assay. The upper right quadrant demonstrates percentage of cells in late apoptosis, the lower right quadrant percentage of cells in early apoptosis. The lower left quadrant shows percentage of vital cells (SDintra-assay< 10 %). Each experiment was done in triplicate and repeated 5 times revealed a similar growth blocking activity, which also correlated to the PTEN expression status [29, 30]. Finally, AEE788, a dual inhibitor of VEGFr and EGFr tyrosine kinases, has been shown to prevent cell growth of a panel of prostate cancer cells in vitro, independent of the PTEN level [31].

The present investigation was designed to evaluate the prostate cancer slowing potential of TKI258, a further tyro- sine kinase inhibitor.
Unfortunately, TKI258 did not influ- ence androgen-sensitive or androgen-resistant prostate cancer cell lines. The lack of influence is puzzling since TKI258 has been shown to inhibit proliferation of breast and urothelial carcinoma cells [32, 33]. Molecular analysis of PC-3 and LNCaP cells demonstrated distinct elevation of the tumor suppressor proteins p27 (PC-3) or p21 (LNCaP) under TKI258. Concurrently, cyclin B and D1 along with cdk1 and 4, which all push the cell cycle forward, were up- regulated by TKI258. Possibly, the increase in cyclin and cdk counteracts the benefit of p21 or p27 up-regulation in prostate cancer, additively leading to drug non-responsiveness. In line with this supposition, it has recently been suggested that simultaneous targeting of FGFr, PDGFr and VEGFr may be associated with increased side effects, leading to a loss in drug

Fig. 3 Growth activity of prostate cancer cells. PC-3, DU-145 or LNCaP cells were treated with 25 nM LBH589, 500 or 1,000 nM TKI258, or the respective drug concentration. Controls remained untreated. 24 h to 72 h percentage increase of the tumor cell number is shown. One representative experiment of six is shown. * indicates significant difference to controls sensitivity [34]. Further studies are required to evaluate whether this mechanism might be restricted to prostate carci- noma or may be valid for other tumor entities.Simultaneous application of TKI258 and LBH589 did not improve the LBH589 single drug regimen. The failure of TKI258 to evoke additive effects on LBH589 treatment may not exclusively be attributed to the inefficacy of TKI258 in reducing tumor cell number when applied alone.

Indeed, low dosed interferon alpha has been shown to augment the anti-tumor potential of HDAC-inhibition on prostate cancer cell growth and invasion, although it was ineffective on its own [6]. Bevacizumab, targeting VEGF, also did not decrease proliferation of prostate cancer cells but still enhanced the anti-proliferative activity of a second compound, docetaxel, in the combinatorial setting [35]. Vice versa, the TKI AEE788 significantly suppressed prostate.

Fig. 4 Western blot analysis of cell cycle proteins, HDAC expression and histone acetylation. PC-3 or LNCaP cells were treated either with 25 nM LBH589, or with 500 or 1,000 nM TKI258, or with both compounds simultaneously.

Histone analysis is restricted to LBH589, since TKI258 evoked no effects. Controls remained untreated. Cell lysates were then subjected to SDS-PAGE and blotted on the membrane incubated with the respective monoclonal antibodies. β-actin served as the internal control. The figure shows one represen- tative of three separate experiments cancer growth. However, the AEE788-valproic acid combi- nation had no advantage over valproic acid mono-treatment [31]. We, therefore, assume that TKI258 lacks the potential to increase the susceptibility of prostate tumor cells to an HDAC-inhibitor based regimen. Whether this might be true for TKI in general is unclear. Evidence has been provided that simultaneous inhibition of HDAC and EGFr may offer greater therapeutic benefits in various cancer types (lung, liver, breast, head and neck, colon, and pancreatic cancer) over single-acting agents [36, 37], and treatment with an HDAC-inhibitor combined with sorafenib demonstrated the highest preclinical efficacy in cholangiocarcinoma and hepatocellular carcinoma models [38, 39]. However, prostate cancer cells were not included in these studies. Interestingly, the AEE788-valproic acid combination, which was not superior to single drug use in prostate carcinoma [31], revealed additive effects in renal cell cancer [40]. We, therefore, assume that simultaneously targeting EGFr, FGFr and/or VEGFr together with HDAC does not provide the optimum strategy to treat prostate cancer patients. Nevertheless,Dovitinib further tests with a broader spectrum of TKI are required to ascertain this statement.