Background
Bladder cancer is the seventh most common cancer type worldwide with about 300,000 newly diagnosed cases per year [[
1]]. One-third of the patients are diagnosed with a muscle invasive carcinoma and up to 50% of patients already present with or developed metastases within the first two years. While patients with a non-muscle invasive papillary urothelial carcinoma expect a rather good prognosis, long term survival of patients suffering from metastatic disease does not exceed 20% [[
2]]. Although significant responses rates are observed after treatment with cisplatin based combination chemotherapy, the majority of patients will develop disease recurrence presenting with cisplatin resistance [[
3]-[
5]].
Epigenetic alterations have been proposed as a driving force of malignancy [[
6]-[
8]]. In particular, histone deacetylases (HDACs) are associated with the development and progression of several cancer types [[
9],[
10]]. The human HDAC family is composed of 18 genes and is classified based on the sequence homology to their yeast orthologues Rpd3, HdaI and Sir2 and their domain organization: HDAC1, HDAC2, HDAC3 and HDAC8 (class I); HDAC4, HDAC5, HDAC7 and HDAC9 (class IIa); HDAC6 and HDAC10 (class II b); HDAC11 (class IV) and seven sirtuins (class III) [[
11]-[
13]]. The classical HDACs catalyze the Zn
2+ dependent deacetylation of acetyl-lysine residues [[
11]]. Expression profiles of class I HDACs are prognostic in various malignancies e.g. gastric, prostate and ovarian cancer [[
14]-[
16]]. In general, HDACs are considered to act as transcriptional co-repressors because high HDAC activity is associated with transcriptionally inactive chromatin [[
17],[
18]]. Although many HDACs deacetylate histones the analysis of the human acetylome indicates that the deacetylation of non-histone proteins represents a considerable part of their action [[
19],[
20]]. Substrates include p53 [[
21]], cohesion subunit SMC3 [[
22]] and α-tubulin [[
23]]. HDAC inhibitors are useful in the therapy of several hematological malignancies and are currently also investigated in the treatment of solid cancers [[
24],[
25]].
The expression of HDAC8 has been described in a variety of cancer entities e.g. colon, breast lung, pancreas and ovary cancer [[
26]]. HDAC8 is the most recently identified class I HDAC. It is a protein of 377 amino acids and contains a NLS in the center of the catalytic domain [[
27]-[
29]]. HDAC8 has a conserved motif for phosphorylation by protein kinase A (PKA), which negatively impacts its catalytic activity [[
30],[
31]]. While class I HDAC family members form nuclear multiprotein complexes that interact with other chromatin modifiers and transcription factors, HDAC8 has not been found to do so [[
17]]. Its intracellular localization seems to depend on the cell type. In gastric adenocarcinoma, esophageal squamous carcinoma, prostate adenocarcinoma and breast carcinoma HDAC8 localization has been described as both nuclear and cytosolic but as exclusively nuclear for non-cancerous gastric, esophageal or prostate epithelium [[
26]]. A predominantly cytosolic distribution of HDAC8 was described for prostate cancer cells [[
32]] and for differentiating smooth muscle cells [[
33]].
In the highly malignant childhood cancer neuroblastoma high HDAC8 expression significantly correlates with poor prognostic markers and poor overall and event-free survival. In cultured neuroblastoma cells knockdown and pharmacological inhibition of HDAC8 resulted in inhibition of proliferation, reduced clonogenic growth, cell cycle arrest and differentiation [[
34]]. Furthermore, HDAC8 specific inhibition selectively induces apoptosis in T-cell derived lymphoma and leukemic cells [[
35]]. In hepatocellular carcinoma overexpression of HDAC8 promotes proliferation and inhibits apoptosis. HDAC8 knockdown inhibits proliferation and enhances apoptosis in hepatocellular carcinoma cells via up-regulation of p53 [[
36]]. In human breast cancer cell lines overexpression of HDAC1, HDAC6 or HDAC8 contributes to increased invasion and metalloproteinase-9 (MMP-9) expression [[
37]]. Furthermore, HDAC8 promotes lung, colon and cervical cancer cell proliferation [[
31]] and may regulate telomerase activity [[
38]].
A recently published analysis of HDAC expression patterns in urothelial carcinoma cell lines and tissues showed a deregulation of several HDACs in urothelial cancer. These findings include up-regulation of HDAC2 and HDAC8 and down-regulation of HDAC4, HDAC5, and HDAC7 [[
39]]. Given the promising results in neuroblastoma [[
35]], we sought to determine whether the selective targeting of HDAC8 might serve as an appropriate therapy for urothelial carcinoma.
Methods
Cell culture and treatment
The urothelial cancer cell lines (UCCs) VM-CUB1, RT-112, SW-1710, 639-V and UM-UC-3 were cultured in DMEM GlutaMAX-I (Gibco, Life Technologies, Darmstadt, Germany) supplemented with 10% fetal calf serum (GE Healthcare, Piscataway, NJ) at 37°C and 5% CO
2. Cell lines used were provided by Dr. M. A. Knowles (Leeds, UK), Dr. J. Fogh (New York, USA), Dr. Barton Grossmann (Houston, USA) and by the DSMZ (Braunschweig, Germany). Normal urothelial control (NUC) cells were isolated from ureters after nephrectomy and were cultured in keratinocyte serum-free medium (Invitrogen, Life Technologies, Darmstadt, Germany) supplemented with 0.25 ng/ml epidermal growth factor and 12.5 μg/ml bovine pituitary extract [[
40]].
Experiments with inhibitors were performed 24 h after seeding of the cells with a single dose of the selective HDAC8-inhibitors compound 2 (c2; 1-napthohydroxamic acid, (abcr GmbH & Co, Karlsruhe, Germany), compound 5 (c5, δ-naphtyl-trans 2-butenoil hydroxamic acid) and compound 6 (c6, 4-naphtyl-benzoil hydroxamic acid) or the pan HDAC-inhibitor SAHA (suberoylanilide hydroxamic acid; #1009929, Cayman Chemicals, Ann Arbor, MI). C5 and c6 are investigational compounds (described in [[
41]]) and are available on request. Inhibitors were dissolved in DMSO as a stock of 10 mM. Control cells were treated with DMSO only.
UCCs were seeded in 6-well plates and grown for 24 h before transfection. Cells were transfected with 10 nM HDAC8 Silencer® Select validated siRNA (#4390824, s31698, Ambion, Life Technologies, Darmstadt, Germany) or a Silencer® Select Negative Control #2 validated siRNA (#4390846, Ambion, Life Technologies, Darmstadt, Germany) using Lipofectamine RNAi MAX (Invitrogen, Life Technologies, Darmstadt, Germany), according to the manufacturer’s protocol. After transfection cells were incubated for 72 h before use for further experiments.
Determination of mean inhibitory concentrations (IC50) and viability
The mean inhibitory concentrations (IC50) and cell viability were measured trough total cellular ATP as an indicator for viable cells using the CellTiter-Glo® Luminescent Cell Viability Assay (Promega, Mannheim, Germany). UCCs were seeded into 96-well plates and grown for 24 h before inhibitor treatment with the indicated drug concentration or DMSO and further grown for 72 h. In another experiment, cells were plated in 6-well plates and grown for 24 h before siRNA-mediated knockdown of HDAC8. Viability measurements were performed after 72 h by transferring the cells into 96-well plates using CellTiter-Glo Reagent according to manufacturer’s protocol in a Wallac Victor 1420 Multilabel Counter (PerkinElmer, Rodgau, Germany).
The colony forming assay was carried out 72 h after siRNA mediated HDAC8 knockdown and HDAC8 inhibitor treatment. Then, cells were plated in 6-well plates at a density of 500 to 1,000 cells per well. After 10 days, cells were washed with PBS (Biochrom, Merck Millipore, Berlin, Germany), shortly fixed in 50% methanol and incubated for 10 min in 100% methanol. The colonies were stained with Giemsa (Merck, Darmstadt, Germany). Colony number and size was determined with ImageJ software (
http://rsbweb.nih.gov/ij/).
Cell cycle analysis by flow cytometry
UCCs were transfected with HDAC8 siRNA or an irrelevant control siRNA or, in another experiment, cultured with the determined IC
50 concentrations of the HDAC8 selective inhibitors c2, c5 and c6, the pan HDAC-inhibitor SAHA (2.5 μM) or DMSO. Cell cycle analysis was performed after 72 h by staining the attached cells and cells in the supernatant with a PI-buffer containing 50 μg/ml propidium iodide, 0.1% sodium citrate and 0.1% Triton X-100 [[
42]] and flow cytometry using a BD LSR Fortessa cell analyzer system and FACSDiva software 6.2 (Becton Dickinson Biosciences, Heidelberg, Germany).
Migration assay
Cell migration was determined in wound healing assays by means of Ibidi Culture-Insert (Ibidi, Martinsried, Germany). The cell suspension was placed in both compartments allowing growth in the designated area only. The cells were treated with IC50 concentrations of c2, c5, c6 or 2.5 μM SAHA 48 h prior to plating. The Culture-Insert was removed 24 h after cell attachment creating a cell-free gap of approximately 500 μm in a confluent cell layer. The extent of wound closure was examined by phase contrast microscopy with the LuciaG software (Laboratory Imaging s.r.o., Prague, Czech Republic) at time points 0, 3, 6, 9, 12 and 24 h.
RNA isolation and quantitative real-time PCR
Total RNA from cell culture was isolated by the Qiagen RNeasy Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer’s protocol. Synthesis of cDNA was performed using QuantiTect Reverse Transcription Kit (Qiagen, Hilden, Germany) with an extended incubation time of 30 min at 42°C. QRT-PCR was performed using an ABI 7500 Fast PCR instrument (Life Technologies, Darmstadt, Germany) with QuantiTect SYBR Green RT-PCR Kit (Qiagen, Hilden, Germany) according to the manufacture’s protocol. To determine the expression levels of HDAC1 (#QT00015239), HDAC2 (#QT00001890), HDAC3 (#QT00093730) and HDAC8 (#QT00049630) we used QuantiTect Primer assays (Qiagen, Hilden, Germany) at an annealing temperature of 55°C. The expression of the housekeeping gene TATA-box binding protein (TBP) was determined with self-designed primers (forward: 5’-ACAACAGCCTGCCACCTTA-3’; reverse: 5’-GAATAGGCTGTGGGTCAGT-3’). Technical duplicates had less than 10% standard deviation.
Western blot analysis
Western blot analysis of whole-cell extracts were done as described previously [[
39]]. Total protein was extracted by cell lysis in a RIPA-buffer containing 150 mM NaCl, 1% Triton X-100, 0.5% desoxycholate, 1% Nonidet P-40, 0.1% SDS, 1 mM EDTA, 50 mM Tris (pH 7,6) and a protease inhibitor cocktail (10 μl/ml, #P-8340, Sigma-Aldrich) for 30 minutes on ice. Protein concentrations were determined by BCA protein assay (Thermo Scientific, Rockford, IL). After separation in SDS-page gels and transfer to PVDF membranes (Merck Millipore, Berlin, Germany) the membranes were blocked with 5% non-fat milk in TBST (150 mM NaCl, 10 mM Tris, pH 7.4 and 0.1% Tween-20), washed and then incubated with primary antibodies at room temperature for 1 h or at 4°C over night. Primary antibodies were used against HDAC1 (1:1,000, C-19, sc-6298; Santa Cruz Biotechnology, Heidelberg, Germany), HDAC2 (1:5,000, H-54, sc-7899; Santa Cruz Biotechnology, Heidelberg, Germany), HDAC3 (1:1,000, H-99, sc-11417; Santa Cruz Biotechnology, Heidelberg, Germany), HDAC8 (1:400, A-4008; Epigentek, Brooklyn, NY), p21 (1:400, C-19, sc-397; Santa Cruz Biotechnology, Heidelberg, Germany), thymidylate synthase (1:1,000, TS, TS106, Millipore, Temecula, CA), PARP (poly [ADP-ribose] polymerase 1, 1:500, 46D11; Cell Signaling Technology, Inc., Danvers, MA) and acetylated α-tubulin (1:15,000, #T-7451, Sigma Aldrich, St. Louis, Mo). Anti-α-Tubulin B-512 (Sigma Aldrich, St. Louis, MO) was used as loading control in a concentration of 1:50,000. Secondary antibodies were HRP-conjugated goat-anti-mouse antibody (sc-2005; Santa Cruz Biotechnology, Heidelberg, Germany), HRP-conjugated goat-anti-rabbit antibody (sc-2004, Santa Cruz Biotechnology, Heidelberg, Germany) and HRP-conjugated rabbit-anti-goat antibody (P0160; DakoCytomation, Stockholm, Sweden) at a concentration of 1:10,000 to 1:100,000. Bands were visualized by the ECL select chemo luminescence kit (GE Healthcare, Piscataway, NJ) and the WesternBright Quantum kit (Biozym, Hessisch Oldendorf, Germany).
Extraction, purification and analysis of histones
Histones were extracted following a published protocol through sulphuric acid extraction and TCA-precipitation [[
43]]. One μg of each sample was used for western blot analysis with 15% SDS-PAGE gels and PVDF membranes (Merck Millipore, Berlin, Germany) according to the previously-described protocol. The detection of acetylated and non-acetylated histones was performed with primary antibodies against acetylated histone H3 (1:2,000, #39139, Active Motif, La Hulpe, Belgium), total histone H3 (1:1,000, #3638, Cell Signaling Technology, Inc., Danvers, MA), acetylated histone H4 (1:1,000, #39243, Active Motif, La Hulpe, Belgium) and total histone H4 (1:500, #39269, Active Motif, La Hulpe, Belgium).
Statistical analysis
Statistical analyses were performed using SPSS 18 (SPSS, Chicago, USA). Significance was measured by the student’s t-test and no-parametric Mann-Whitney U test. P-values of < 0.05 were considered as significant whereas p < 0.01 and p < 0.001 were defined as highly significant. IC50 values and dose-response curves were approximated by non-linear regression analysis using Origin 8.0 (Origin Lab, Northhampton, GB).
Discussion
In this study we present the first systematic analysis of HDAC8 expression and function in urothelial cancer using a set of bladder cancer cell lines representative for the heterogeneity of this tumor. The aim of our study was to evaluate the potential of HDAC8 as a therapeutic target.
Overexpression of HDAC8 has been reported in a considerable number of different cancer entities [[
26],[
34],[
36],[
37]]. In neuroblastoma, in particular, HDAC8 expression was significantly correlated with further poor prognostic markers as well as poor overall and progression-free survival. SiRNA-mediated knockdown and pharmacological inhibition of HDAC8 in neuroblastoma significantly decreased proliferation rate and reduced clonogenic growth, cell cycle arrest, and differentiation [[
34]]. In hepatocellular carcinoma HDAC8 knockdown also suppresses cell proliferation and enhances apoptosis via elevated expression of p53 and acetylation of p53 at Lys382 [[
36]].
As there were indications from our own and other data that HDAC8 is often upregulated in urothelial carcinoma as well [[
39],[
44]], the question arose whether HDAC8 might be a potential target for anticancer treatment in this tumor.
In urothelial cancer cell lines, a variable expression of HDAC8 was observed both at mRNA and protein level [[
39]]. Importantly, mRNA expression levels were comparable to neuroblastoma and breast cancer cells (data not shown). An according variability has also been reported from investigations in further malignomas, e.g. hepatocellular carcinoma cell lines, were also a broad range of HDAC8 expression was observed in cancer cell lines [[
36]]. Differences between mRNA and protein expression indicate that HDAC8 expression and activity in UCCs may be regulated both transcriptionally and on the protein level, e.g. by protein kinase A (PKA) phosphorylation [[
30],[
31]]. In addition, in our UCC panel, a low HDAC8 expression was predominantly observed in UCCs with an epithelial phenotype. Therefore, to cover this range both on protein and mRNA level, we chose to apply a panel of 6 cell lines representing the heterogeneity of the HDAC8 expression instead of focusing on one urothelial cancer cell line.
SiRNA targeting of HDAC8 in UCCs caused a significant reduction of proliferation up to 45% and inhibited clonogenic growth in a cell line-dependent manner. These results were comparable to observations in hepatocellular carcinoma (HCC) and neuroblastoma cells [[
34],[
36]]. Clonogenic growth was most decreased in the mesenchymal cell line SW-1710 which presented the highest HDAC8 protein expression. Treatment with the three different HDAC8 inhibitors c2, c5 and c6 revealed a low sensitivity of UCCs for c2 with a calculated IC
50 value greater than 50 μM. In contrast, neuroblastoma cell lines (BE (2)-C) were more sensitive to treatment with c2, presenting IC
50 values in a range of 10 to 40 μM. In these cells, the HDAC8 inhibitor c2 yielded an similar phenotype at a concentration similar to the
in vitro IC
50 of c2 against HDAC8 [[
41]]. None of the UCCs was inhibited substantially at this concentration by pharmacological treatment with c2.
The inhibitors c5 and c6 significantly reduced the viability of all UCCs, with half inhibitory concentrations between 9 and 20.8 μM. These differences follow the order of the affinity of the inhibitors for HDAC8
in vitro [[
41]]. Though
in vitro affinity of c5 and c6 is 20 - 50 fold higher compared to c2,
in vivo effects on UCC were not as strong as expected.
Focusing on morphological features of UCCs, the data suggested that cells with an epithelial phenotype and low HDAC8 expression are more sensitive towards pharmacological inhibition of HDAC8 with c5 and c6 compared to cells with a mesenchymal phenotype. Specifically, SW-1710 cells (mesenchymal, elevated HDAC8 expression) were least sensitive to the inhibitors c5 and c6 while RT112 cells (epithelial, lowest HDAC8 expression) responded to treatment with c5 and c6 already at low concentrations.
As recently shown in endometrial stroma sarcoma cells, HDAC inhibition may be counteracted by increased activity of the PI3K pathway in PTEN-deficient cells [[
45]]. In our cell line panel, UM-UC-3 are PTEN-deficient, resulting in increased PI3K activity. However, this cell line was not exceptionally resistant either in our previous study using pan-HDAC inhibition [[
39]] or in the present study with HDAC8-specific inhibitors. Accordingly, at least in urothelial cancer, PTEN deficiency does not seem to have a decisive impact on the efficacy of HDAC inhibitors.
Effects of siRNA mediated downregulation and pharmacological inhibition on urothelial cancer cell lines were not thoroughly consistent. Differences might be explained by several factors. For example, knockdown depletes the protein thereby not only affecting enzymatic but also other protein functions for example complex assembly. Inhibitor treatment ideally only suppresses the enzymatic activity while further protein functions should not be affected. Accordingly, also compensatory mechanisms might be different in both conditions. Comparing expression levels of further class I HDACs after knockdown of HDAC8 as well as after pharmacological inhibition, only minor changes were observed. Although upregulation of HDAC1 or HDAC2 was a little more consistently observed after HDAC8 knockdown, they can hardly explain the difference between knockdown and inhibition by c5 or c6. More likely, the stronger effects of the inhibitors may be due to inhibition of other targets in addition to HDAC8.
Neither HDAC8 knockdown nor pharmacological treatment with any compound (except the SAHA control) led to a change in histone H3 or H4 acetylation, a widely used surrogate marker for intracellular HDAC inhibition. This finding suggests that HDAC8, as expected, does not substantially affect overall histone acetylation. In addition, this does also indicate that inhibitor treatment seems to be iso-enzyme specific as other class I HDACs seemed to be not affected. This was also observed in neuroblastoma cell lines after treatment of HDAC8. Global Histone H4 acetylation was not affected by HDAC8 knockdown or by selective inhibitor treatment [[
34]].
In contrast, HDAC8 knockdown in some cell lines and treatment with c5 or c6 resulted in a strong increase of acetylated α-tubulin. The latter finding is in accord with previous findings in HeLa and HEK293 cells [[
45]]. The cytoplasmic protein α-tubulin is especially a substrate of HDAC6 which is predominantly localized in the cytoplasm [[
23]]. HDAC6 influences the cytoskeleton and cell motility via deacetylation of α-tubulin and other cytoskeleton proteins [[
46]].
In vitro, c5 and c6 do not inhibit HDAC6 efficiently. Thus, the best explanation for these observations is that
in vivo HDAC8 directly or indirectly influences α-tubulin acetylation. Similar discrepancies between
in vitro and
in vivo activity of an isoenzyme-selective HDAC inhibitor on tubulin acetylation have been observed by others for valproic acid [[
47]]. These effects on α-tubulin acetylation may relate to the inhibition of cell migration by c5 and c6 we observed in UC cell lines. However, inhibition of HDAC6 as such does not inhibit migration of UCC as efficiently as the HDAC8 inhibitors c5 and c6 [[
48]].
The effects of siRNA mediated knockdown of HDAC8 on cell cycle and apoptosis were limited and few significant effects were seen, such as a decreased S-phase fraction in VM-CUB1 and small changes in thymidylate synthase and p21 expression. In the neuroblastoma cell line BE (2)-C, a G0/G1 arrest has been detected after siRNA-mediated knockdown of HDAC8. This G0/G1 arrest induced by HDAC8 knockdown was associated with p21 mRNA upregulation [[
34]]. In contrast, no effect on the cell cycle was observed in the hepatocellular carcinoma cell lines BEL-7402 and Hep-G2 [[
36]]. This observation fits with our own marginal effects after siRNA-mediated HDAC8 knockdown. The level of apoptosis induction in BEL-7402 and Hep-G2 cells after siRNA-mediated targeting of HDAC8 were comparable to the increase of the subG1-fraction in individual urothelial carcinoma cell lines after targeting of HDAC8 [[
36]].
Concerning the use of inhibitors, effects of pharmacological inhibition on cell cycle distribution by c2 were, as expected, only minor. In contrast, pharmacological inhibition by c5 or c6 resulted in a significant albeit low increase of the sub-G1 fraction in two out of five cell lines and in an apparent G2/M-arrest in four out of five cell lines. Consequently, p21 increased in two cell lines and thymidylate synthase decreased in all but one.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.
The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
ML, MJH, AK, WAS and GN conceived and designed the study. ML and MJH carried out the performed experiments. ML, WAS and GN carried out data analysis and prepared the initial manuscript. SMU provided crucial reagents. MJH, AK, SMU, WAS and GN contributed to the manuscript. WAS and GN supervised the project. All authors read and approved the final manuscript.