Distinct mechanisms contribute to acquired cisplatin resistance of urothelial carcinoma cells

Cisplatin (CisPt) is frequently used in the therapy of urothelial carcinoma (UC). Its therapeutic efficacy is limited by inherent or acquired drug resistance. Here, we comparatively investigated the CisPt-induced response of two different parental urothelial carcinoma cell lines (RT-112, J-82) with that of respective drug resistant variants (RT-112R, J-82R) obtained upon month-long CisPt selection. Parental RT-112 cells were ~2.5 fold more resistant to CisPt than J-82 cells and showed a different expression pattern of CisPt-related resistance factors. CisPt resistant RT-112R and J-82R variants revealed a 2–3-fold increased CisPt resistance as compared to their corresponding parental counterparts. Acquired CisPt resistance was accompanied by morphological alterations resembling epithelial mesenchymal transition (EMT). RT-112R cells revealed lower apoptotic frequency and more pronounced G2/M arrest following CisPt exposure than RT-112 cells, whereas no differences in death induction were observed between J-82 and J-82R cells. CisPt resistant J-82R cells however were characterized by a reduced formation of CisPt-induced DNA damage and related DNA damage response (DDR) as compared to J-82 cells. Such difference was not observed between RT-112R and RT-112 cells. J-82R cells showed an enhanced sensitivity to pharmacological inhibition of checkpoint kinase 1 (Chk1) and, moreover, could be re-sensitized to CisPt upon Chk1 inhibition. Based on the data we suggest that mechanisms of acquired CisPt resistance of individual UC cells are substantially different, with apoptosis- and DDR-related mechanisms being of particular relevance. Moreover, the findings indicate that targeting of Chk1 might be useful to overcome acquired CisPt resistance of certain subtypes of UC.


INTRODUCTION
Bladder cancer is a frequent type of cancer worldwide. In most countries, the majority of bladder cancers are urothelial carcinoma (UC) [1]. Cisplatin (CisPt)based therapeutic regimen are commonly used both in the perioperative (neoadjuvant, adjuvant) setting for muscleinvasive UC as well as in the palliative setting for recurrent or metastatic UC [2]. CisPt enters cells by passive diffusion as well as by help of transporters [3,4]. Upon replacement of its chloride ligands by water, DNA adducts are formed by S N 2 mechanism (nucleophilic substitution) [5]. The vast majority (60-80%) of DNA adducts generated by CisPt are DNA intrastrand crosslinks (GpG and ApG). Only 1-2% of the DNA crosslinks formed by CisPt are DNA interstrand crosslinks [6,7]. Platinuminduced DNA crosslinks cause a substantial distortion of the DNA double helix, resulting in transcription and replication blockage [8,9]. In consequence of stalled replication forks, DNA double-strand breaks (DSBs) can arise as secondary lesions [10]. DSBs are potent triggers of cell death [11] and can be repaired by DNA doublestrand break repair (homologous recombination (HR) or non-homologous end joining (NHEJ)). The removal of CisPt-induced DNA crosslinks involves nucleotide excision repair (NER), including transcription-coupled Research Paper www.impactjournals.com/oncotarget NER (TC-NER) [9]. The relevance of NER mechanisms for the tumor cell response to CisPt is highlighted by the fact that the expression of the NER factor ERCC1 predicts the therapeutic efficacy of CisPt in lung tumors [12,13] and also seems to be of relevance for UC [14]. The efficacy of platinum-based therapy is limited by intrinsic or acquired drug resistance [15]. Factors that contribute to CisPt resistance are manifold and are poorly characterized for UC [16,17]. Recently, mechanisms affecting resistance to CisPt have been classified according to their site of action as pre-, on-, post-and off-target [17], with drug transport, DNA repair, apoptosis and signal transduction at membranes, respectively, being representatives of these mechanisms.
Upon induction of DNA damage a highly complex cellular stress response program, known as the DNA damage response (DDR), is activated. It tightly controls cell cycle progression by activation of cell cycle checkpoints and fine-tunes mechanisms of DNA repair and cell death [18,19]. Activation of the DDR is considered as an inducible barrier against early tumorigenesis [20,21] and, moreover, to precede genomic instability in bladder cancer [22]. In case of error prone repair of DSBs, genomic instability of bladder carcinomas is favoured [23]. DSBs as well as replication-and transcriptionblocking DNA lesions are particular efficient activators of the DDR. The PI3-like kinases ATM and ATR play key roles in the regulation of the DDR [24,25]. These kinases phosphorylate numerous substrates, among others checkpoint kinases (e.g. Chk1, Chk2) and p53, which eventually affect survival or death of the damaged cell. The relevance of DDR mechanisms for the CisPt sensitivity of UC cells and, most importantly, for CisPt resistant variants, is largely unknown.
Urothelial cancer cells segregate into epithelial and mesenchymal subsets [26]. Therefore, we included both RT-112 and J-82 cells, which are described as UC cells of mainly epithelial-and mesenchymal-like phenotype, respectively [27,28], in our study. To select CisPt resistant variants we took into account that the therapeutic regimen commonly used in CisPt-based anticancer therapy comprises repetitive treatment cycles, where CisPt is administered by infusion, intermitted by treatment free periods. Therefore, parental RT-112 and J-82 UC cells were selected for resistance by multiple pulse-treatments with CisPt followed by extended recovery periods. The aim of the study was to comparatively analyze CisPt resistant UC cell variants (RT-112 R and J-82 R ) with their respective parental cell types regarding (i) DDR capacity, (ii) the expression of putative CisPt resistance factors as suggested by Galluzzi et al. [17] and (iii) the response to a subset of pharmacological modifiers of the DDR and DNA repair, including inhibitors of checkpoint kinases, which are promising novel anticancer compounds acting by selectively increasing replicative stress and cell death in malignant cells [29]. In doing so, we aimed to identify mechanisms that are of relevance for acquired CisPt resistance of UC cells and, moreover, to figure out therapeutic options to overcome their CisPt resistant phenotype.

Characterization of the CisPt response of the urothelial carcinoma cells RT-112 and J-82
In light of the fact that urothelial cancer cells can segregate into epithelial and mesenchymal subsets [26], RT-112 und J-82 cells, which are representative of epithelial-and mesenchymal-like UC cells, respectively [27,28], were used in the present study. RT-112 cells differ from J-82 regarding morphology ( Figure 1A) and a higher mRNA expression of the epithelial marker E-cadherin as well as a lower expression of the mesenchymal marker vimentin ( Figure 1B) as expected. Proliferation rate was higher in RT-112 as compared to J-82 cells ( Figure 1C). Analyzing the influence of CisPt on cell viability 24-72 h after CisPt pulse-treatment, we observed that RT-112 cells are 2-3-fold more resistant to moderate doses of CisPt than J-82 cells ( Figure 1D-1F). This is reflected by IC 50 /IC 80 values of 10.7 µM / 44.3 µM and 3.9 µM / 13.5 µM for RT-112 and J-82, respectively, as determined after a post-incubation period of 72 h by the Alamar blue assay ( Figure 1F). This difference in drug sensitivity is not detectable anymore at very high CisPt doses of ≥ 80 µM ( Figure 1D-1G). Measuring cell viability via an alternative method, i.e. the Neutral red assay, similar results were obtained ( Figure 1G). Based on a recent report of Galluzzi et al. [17], who has classified putative CisPt resistance factors of tumor cells, we assembled a 96 well-based quantitative real-time (qRT) PCR array to comparatively analyze the mRNA expression of these factors in RT-112 and J-82 cells. The results of this analysis revealed large cell type-specific differences in the basal mRNA expression of both pre-, on-, post-as well as off-target factors [17]. In more detail, we observed a significantly stronger mRNA expression of ATP7A, BRCA1, VDAC, Calpain, p53, Caspase 6 and ERBB2 in RT-112 cells as compared to J-82 cells. By contrast, J-82 cells revealed an enhanced expression of MT1A, XAF1, BCL2, DYRK1VB, HMOX1, GPX1 and HSPA1B as compared to RT-112 cells (Figure 2A, 2B). Analysing gene expression 72 h after treatment with the IC 50 of CisPt, we found upregulation of GPX1 and XAF1 concommitantly in both RT-112 and J-82 cells ( Figure 2C, 2D). Notably, J-82 cells responded to CisPt treatment with the upregulation of various DNA repair-related factors (i.e. BRCA1, BRCA2, MSH2, XRCC3) ( Figure 2D). This response was not found in RT-112 cells ( Figure 2C). Taken together, the data show that both basal and CisPt-stimulated mRNA expression of factors affecting CisPt sensitivity [17] considerably vary between the two examined UC cell lines, indicating that the basal defence capacity of epithelial-and mesenchymallike UC cells against CisPt-induced injury might be different. This hypothesis needs future confirmation by analyzing the CisPt response of additional UC cell lines of epithelial or mesenchymal origin both in vitro and in vivo.

Selection of CisPt resistant UC cell variants
In order to elucidate which mechanisms contribute to acquired CisPt resistance of UC cells and having in mind the therapeutic regimen used in the clinic, RT-112 and J-82 cells were repeatedly pulse-treated twice a week (for each 4 h) with the corresponding IC 50 of CisPt, followed by a recovery period of one week ( Figure 3A). After a total selection time of 10 weeks, CisPt resistant RT-112 R und J-82 R cells were obtained ( Figure 3B-3D). Measuring cell viability by the Alamar blue assay, the resistant variants revealed an about 3-fold increase in the IC 50 as compared to the corresponding parental cells ( Figure 3B-3D). Similar results were obtained using the Neutral red assay (data not shown). Gain of CisPt resistance was accompanied by morphological alterations, in particular cell enlargement and distinct cell protrusions ( Figure 3B-3D). Both RT-112 R and J-82 R cells showed an increased mRNA expression of the intermediate filament vimentin (Figure 3C-3E) as compared to their respective parental cells. As vimentin expression represents a prototypical marker of mesenchymal cells, we hypothesize that the development of an EMT-like phenotype is favoured in epithelial-like RT-112 cells and is further promoted in J-82 cells during the selection of CisPt resistant UC cell variants. A coherence between EMT and acquired drug resistance was reported by others [26,[30][31][32]. Flow cytometry-based analyses performed 72 h after CisPt treatment showed a reduction of apoptotic cell death in RT-112 R cells as compared to RT-112 ( Figure 4A). This effect was only observed in RT-112 R cells ( Figure 4A, upper panel) but not in J-82 R cells ( Figure 4A, lower panel). Both RT-112 R and J-82 R cells were characterized by a more pronounced activation of G2/M checkpoint mechanisms as compared to their corresponding parental counterparts ( Figure 4B). The data show that the mechanisms of acquired CisPt resistance differ between individual UC cell lines with protection from CisPt-induced apoptotic mechanisms and alterations in checkpoint control mechanisms being involved.

Induction of DNA damage and activation of the DNA damage response (DDR) in parental and CisPt resistant UC cell variants
In order to measure the induction of DNA damage following CisPt treatment, ATM/ATR-catalyzed S139 phosphorylation of histone H2AX and the recruitment of 53BP1 to sites of damage were monitored by immunocytochemistry ( Figure 5A-5B). Moreover, the level of CisPt-induced DNA intrastrand crosslinks was monitored by Southwestern analysis (Figure 5C-5D). The formation of nuclear γH2AX foci and 53BP1 foci is part of the DNA damage response (DDR) and is believed to reflect predominantly the formation of DNA double-strand breaks (DSBs) [19]. Following CisPt treatment, DSBs are believed to be mainly generated as secondary lesions from primary DNA platinumadducts that stall replication forks [10]. As observed 4 h and 24 h after CisPt pulse-treatment for 4 h, we found a significant reduction in the number of DSBs in J-82 R cells, but not in RT-112 R cells ( Figure 5A-5B). This finding indicates that CisPt resistance of J-82 R cells, but not of RT-112 R cells, might result from a reduced formation of highly cytotoxic DSBs and/or attenuated DDR following CisPt treatment. Bearing in mind that CisPt-induced DSBs mainly originate from primary Pt-(GpG) DNA adducts, we next monitored the formation of Pt-(GpG) intrastrand crosslinks by Southwestern blot analyses. The data show that DNA intrastrand crosslink formation was significantly lower in the J-82 R subline as compared to J-82 parental cells ( Figure 5D). Based on these observations we suggest that acquired CisPt and CisPt resistant (RT-112 R , J-82 R ) UC cells were treated with the IC 50 or IC 80 of CisPt (according to Figure 1F). After incubation period of 72 h, subG1 fraction (A) and cells present in G2/M phase of the cell cycle (B) were determined by flow cytometry-based analyses. Data shown are the mean ± SD from three independent experiments each performed in duplicate. *statistical significance of parental cells vs. CisPt resistant cells. ***p ≤ 0.001; *p ≤ 0.05. www.impactjournals.com/oncotarget resistance of J-82 cells involves a reduced formation of primary (i.e. Pt-(GpG) adducts) and secondary (i.e. DSBs) DNA damage following CisPt treatment. Mechanistically, it is feasible that pre-target resistance mechanisms such as transport or detoxification mechanisms take part [17]. In this context it is noteworthy that the level of CisPt-induced Pt-(GpG) DNA intrastrand crosslinks is higher in parental J-82 cells as compared to RT-112 cells ( Figure  5C) if the corresponding IC 50 and IC 80 were used. This finding indicates that the level of Pt-(GpG) intrastrand crosslinks does not necessarily predicts the level of cytotoxicity. In a next step we comparatively analyzed the DDR of UC parental cells and corresponding CisPt resistant variants following CisPt treatment by Western blot analysis. The data obtained uncover large variations in the activation of DDR mechanisms already in J-82 versus RT-112 parental cells, as reflected on the levels of γH2AX, p-Chk1, p-p53 and p-Kap1 ( Figure 6A-6B). In general, J-82 revealed a more profound activation of the DDR than RT-112 cells. This is likely related to the higher level of Pt-(GpG) adducts in J-82 cells (see Figure 5C), resulting in stronger activation of DDR mechanisms. Comparative analyses of J-82 cells versus CisPt resistant J-82 R cells showed lower phosphorylation levels of H2AX, Chk1, p53 and Kap1 in the CisPt resistant variants ( Figure 6A). Again, this is in line with the observed decrease in Pt-(GpG) DNA adducts and DSB formation described in J-82 R as compared to J-82 cells (see Figure 5). RT-112 R cells revealed a specifically increased phosphorylation of Chk1 as compared to RT-112 parental cells ( Figure 6B). This is indicative of a selectively increased potency of RT-112 R cells to activate checkpoint control mechanisms that might contribute to protection from CisPt induced apoptotic mechanisms (see Figure 4A).

Expression of CisPt susceptibility factors in CisPt resistant UC cells
Next, we analyzed the mRNA expression of CisPt-related susceptibility factors reported by Galluzzi et al. [17] in RT-112 R and J-82 R cells as compared to the corresponding parental cells. Regarding RT-112 R cells, we found a significant increase in the mRNA expression of metallothionein (MT1A) and the XIAP-associated factor 1 (XAF1) as compared to RT-112 cells ( Figure 7B). In J-82 R cells we observed an elevated mRNA expression of the antioxidative factors heme oxygenase 1 (HMOX1) and glutathione S-transerase M1 (GSTM1) as well as of XAF1 as compared to the corresponding controls ( Figure 7A). Hence, the two types of CisPt resistant UC cell variants were characterized by an increased mRNA expression of  Figure 1F) for 4 h. After post-incubation periods of 4 h or 24 h cells were harvested for Western blot analyses using phospho-specific antibodies as indicated. For control, cells were irradiated with 10 Gy (IR) and analysis was performed 1 h later. Data shown are representative of two independent experiments. Expression of β-actin was determined as protein loading control. www.impactjournals.com/oncotarget XAF1. In this context we would like to note that selection of CisPt resistant J-82 and RT-112 cells by a selection protocol using continuous treatment with increasing CisPt doses over a time period of 4 month also resulted in increased level of XAF1 mRNA in CisPt resistant J-82 cells but not in RT-112 cells (Supplementary Figure S1). The finding of upregulated XAF1 mRNA expression in drug resistant UC cell variants was unexpected considering that XAF1 is known to inhibit the anti-apoptotic factor XIAP, and hence is anticipated to promote cell death [33]. Correspondingly, high XAF1 level was suggested as predictive marker in pancreatic cancer associated with better overall survival [34]. Therefore, it appears possible that its increased mRNA expression in J-82 R cells accidentially correlates with CisPt resistance but is not causative for acquired CisPt resistance of UC cells. Alternatively, XAF1 might have a so far not yet decribed pro-survival function in CisPt resistant UC cells. In this context it is noteworthy that a cell cycle regulatory function has been suggested for XAF1 in gastrointestinal cancer, which rests on its interaction with Chk1 [35]. Interestingly enough induction of XAF1 mRNA expression was also observed in both J-82 and RT-112 parental cells 72 h after CisPt addition (see Figure   2C-2D). So, forthcoming studies are clearly required to dissect the role of XAF1 in the response of UC cells to CisPt. In addition, the data indicate that the improvement of anti-oxidative capacity, as reflected by the upregulation of HMOX1 and GSTM1, and increased expression of metallothionein MT1A might be of particular relevance for acquired CisPt resistance of some subtypes of UC. Bearing in mind that oxidative stress contributes to the cytotoxicity of CisPt [36,37], upregulation of anti-oxidative mechanisms might be a meaningful cytoprotective strategy of UC cells, as is the upregulation of metallothioneins [38]. Noteworthy, upregulation of the mRNA expression of DNA repair factors (i.e. BRCA1, BRCA2, ERCC1, MLH1, MSH2, XRCC3), which are involved in the repair of CisPt-induced DNA damage, was not observed in the CisPt resistant variants.

J-82 R cells show enhanced sensitivity to a Chk1 inhibitor
In search of pharmacological approaches to overcome acquired CisPt resistance of J-82 R cells, we examined their sensitivity to a selected subset of  pharmacological inhibitors. Unfortunately, these analyses could not be performed with RT-112 R cells because their CisPt resistant phenotype turned out as not stable and got lost upon freezing. For these analyses inhibitors of the DDR-related kinases ATM/ATR (VE-822) as well as of checkpoint (Chk) kinases (AZD-7762 (Chk1 and Chk2 inhibitor) and LY2603618 (Chk1-specific inhibitor)) and Wee1 kinase (MK-1775) were included. Noteworthy, targeting of ATR/Chk1-regulated replicative stress responses of tumor cells has recently been suggested as a novel therapeutic strategy [29]. As additional candidate inhibitors we analyzed the impact of the cyclin-dependent kinase (CDK) inhibitor roscovitine, the multikinase inhibitor sorafenib, which is frequently used as anticancer drug in the clinic, as well as of inhibitors of the DNA repair proteins RAD51 (RI-1) and PARP-1 (olaparib) on the viability of parental J-82 versus resistant J-82 R cells.
As a further candidate inhibitor we employed lovastatin, because statins have been shown to exhibit anticancer activity in various preclinical model systems [39] and are discussed to overcome acquired drug resistance to doxorubicin in neuroblastoma cells [40]. J-82 R cells turned out to be slightly more sensitive to treatment with the pan Chk inhibitor AZD-7762 ( Figure 8A) and showed a significantly enhanced sensitivity to the Chk1-specific inhibitor LY2603618 as compared to parental cells ( Figure 8B). The J-82 R cells also revealed a tendentially enhanced sensitivity to the Wee1 kinase inhibitor MK-1775 ( Figure 8C) but not to the CDK inhibitor roscovitine ( Figure 8D). The pronounced loss of cell viability of J-82 R cells following Chk1 inhibition seems to be specific as it was not observed upon inhibition of ATM/ATR kinase or the DNA repair factors RAD51 and PARP-1 (Table 1). Pre-treatment of J-82 R cells with low non-toxic concentration of Chk inhibitors increased their sensitivity to CisPt ( Figure 8E-8F), indicating that targeting of Chk might be particular useful to overcome acquired CisPt resistance of some subtypes of UC cells. Whether targeting of Chk is equally effective in epithelial and mesenchymal-like UC cells remains to be elucidated in forthcoming studies. Noteworthy, Chk inhibition was reported to overcome CisPt resistance of head and neck cancer cells [41] as well as of clear cell carcinoma of the ovary [42] in vitro, supporting the hypothesis that targeting of Chk might be a useful approach to deal with acquired CisPt resistance of different types of tumor cells. Taken together, our data support the current view that increasing replicative stress in tumor cells might be a promising therapeutic strategy also in UC [29]. In fact, the antitumor potency of Chk1 inhibitors is currently investigated in www.impactjournals.com/oncotarget clinical trials. Whereas AZD-7762 revealed inacceptable cardiotoxicity [43], SCH-900776 seems to be better tolerated in humans [44].
Taken together the data show that different molecular mechanims are involved in acquired resistance of different types of UC cells to CisPt. Apparently, molecularly different sets of CisPt defence programs can become activated in individual UC cells. We hypothesize that acquired CisPt resistance in (epithelial-like) RT-112 cells might be preferentially related to protection from pro-apoptotic mechanisms, whereas gain of CisPt resistance in (mesenchymal-like) J-82 UC cells seems to be characterized by a lower level of CisPt formed DNA damage and attenuated DDR. Mechanisms of transport and DNA repair seem to be of minor relevance for aquired CisPt resistance of UC cells. Hence, therapeutic targeting of apoptosis-and/or DDR-related mechanisms are suggested as preferential to overcome acquired CisPt resistance in UC. Importantly, inhibitors of Chk might be useful to handle CisPt resistance in UC cells. Forthcoming in vivo studies are required to scrutinize the potency of Chk1 specific inhibitors to work against the non-responsiveness of urothelial carcinoma cells to CisPtbased anticancer therapy in a clincally relevant setting.

Determination of cell viability
Cell viability was determined using the Alamar blue assay [45]. In this assay, viable cells are detectable by their ability to effectively metabolize the non-fluorescent dye resazurin (Sigma, Steinheim (Germany)) to fluorescent resorufin. Cells were incubated for 1.5 h with the resazurin solution (final concentration 40 µM) before fluorescence was measured (excitation: 535 nm, emission: 590 nm, 5 flashes, integration time: 20 µs). Relative viability in the untreated controls was set to 100%. In addition, cell viability was also determined by use of the Neutral red assay [46]. In this assay, viable cells accumulate the red dye 2-methyl-3-amino-7-dimethylaminophenazine in lysosomes, whereas dead cells are unable to do so. Thus, the staining intensity is directly proportional to the number of viable cells. For this assay, cells were incubated for 1.5 h with the neutral red solution (Sigma Aldrich (Steinheim, Germany)) (final concentration 0.1 mg/ml) before fixation with 1% formaldehyde. Afterwards the dye was extracted with 50% ethanol and absorption was measured at 550 nm. Relative viability in the untreated controls was set to 100%.

Flow cytometry-based analysis of cell cycle distribution and cell death
Cell cycle distribution was analyzed by flow cytometry. Adherent cells were trypsinized and combined with the medium that contains floating cells. After centrifugation (800 × g, 5 min, RT), cell pellet was washed and resuspended in PBS. Afterwards, the cells were fixed with ice-cold ethanol (-20°C, ≥ 20 min). After centrifugation (800 × g, 5 min, 4°C) the supernatant was discarded. The cells were resuspended in PBS containing RNase A (Serva Electrophoresis GmbH (Heidelberg, Germany)) (1 µg/µl) and incubated for 1 h at RT. After adding of propidium iodide (Sigma (Steinheim, Germany)) cells were subjected to flow cytometric analysis (Becton Dickinson (Heidelberg, Germany)). The SubG1 fraction was considered as a measure of dead (apoptotic) cells.

Analysis of DNA damage induction
The formation of DNA double-strand breaks (DSBs) was investigated by measuring the level of S139 phosphorylated H2AX (γH2AX), which is a surrogate marker of DNA damage [47,48], by Western blot analysis or by immunocytochemistry-based detection of nuclear γH2AX and 53BP1 foci. For immunocytochemical analysis, the cells were seeded onto coverslips. After treatment the cells were fixed with 4% formaldehyde in phosphate buffered saline (PBS) (MERCK (Darmstadt, Germany)) (15 min, RT) and incubated with ice-cold methanol (over night, -20°C). After blocking (1.5 h, RT; blocking solution: 5% BSA in PBS/0.3% Triton X-100, incubation with γH2AX antibody (mouse) and 53BP1 antibody (rabbit) was performed (1:500, over night, 4°C), followed by further washing with PBS/0.3% Triton X-100 and addition of the secondary fluorescence-labelled antibody (1:500, 1 h, RT, in the dark). After washing, the cells were mounted in Vectashield (Vector Laboratories (Burlingame, CA, USA)) containing DAPI. Nuclear gH2AX and 53BP1 foci were counted by microscopical analysis using an Olympus BX43 fluorescence microscope and the number of co-localized gH2AX and 53BP1 foci was calculated.
The level of Pt-(GpG) DNA intrastrand crosslinks was monitored by Southwestern blot analysis. To this end, genomic DNA was isolated using the "DNeasy Blood and Tissue" kit (Qiagen (Hilden, Germany)). The concentration and purity of the DNA was measured photometrically (NanoVue TM Plus (GE Healthcare, UK)). 0.5 μg of the DNA was diluted in 100 μl of TE buffer, denatured by heating (10 min, 95°C) and cooled on ice. Afterwards, 100 μl ice-cold ammonium acetate (2 M) was added. A nitrocellulose membrane was soaked in 1 M ammonium acetate and fixed into a slot-blot apparatus (Roth (Karlsruhe, Germany)). The DNA was transferred onto the membrane by use of a vacuum pump. After washing with 1 M ammonium acetate and water, the membrane was incubated with 5 × SSC (10 × SSC: 1.5 M NaCl, 150 mM sodium citrate, pH 7.0) for 5 min and baked for 2 h at 80 o C before it was blocked in 5% non-fat milk in TBS/0.1% Tween 20 over night at 4°C. Incubation with the primary antibody directed against Pt-(GpG) intrastrand crosslinks (1:200) [49] was conducted for 1 h at RT, followed by incubation with peroxidase-conjugated anti-rat IgG secondary antibody (1:2000, 2 h, RT). Visualization of the Pt-(GpG) intrastrand crosslinks was done by chemiluminescence and autoradiographies were densitometrically analyzed. Additionally, the membrane was stained with methylene blue (MP Biomedicals (Santa Ana, CA, USA)) to ensure equal DNA loading.

Western blot analysis
The activation status of the DNA damage response (DDR) machinery was investigated by Western blot analysis employing a set of phospho-specific antibodies, which detect prototypical factors that become activated by phosphorylation in the course of the DDR. Total cell extracts were obtained by lysing an equal number of cells in Roti ® -Load buffer (Carl Roth GmbH (Karlsruhe, Germany)) (5 min, RT). After sonication (EpiShear™ Probe sonicator, Active Motif (La Hulpe, Belgium)) proteins were denatured by heating (5 min, 95°C) and separated by SDS-PAGE (12.5% gel). Subsequently, proteins were transferred onto a nitrocellulose membrane (GE Healthcare (Little Chalfont, UK)) via the Protean Mini Cell System (BioRad (München, Germany)). After blocking in 5% non-fat milk in TBS/0.1% Tween 20 (MERCK (Darmstadt, Germany)) (2 h, RT), the membrane was incubated with the corresponding primary antibody (1:1000, over night, 4°C). After washing with TBS/0.1% Tween 20 the secondary (peroxidase-conjugated) antibody was added (1:2000, 2 h, RT). For visualization of the bound antibodies the Fusion FX7 imaging system (PeqLab (Erlangen, Germany)) was used.

Quantitative real-time PCR-based mRNA expression analyses
Putative markers of CisPt susceptibility were selected on the basis of a recent review by Galluzzi et al. [17] who has classified putative CisPt resistance factors of tumor cells into mechanisms of pre-, on-, post-and off-target. Based on this report we assembled a 96 wellbased quantitative real-time (qRT) PCR array to analyze the mRNA expression of these factors in RT-112 und J-82 cells. In addition, mRNA expression of the epithelial marker E-cadherin as well as the mesenchymal marker vimentin and the proliferation factors c-Myc and cyclinD1 was analyzed by qRT-PCR. Total RNA was purified using the RNeasy Mini Kit (Qiagen (Hilden, Germany)). The reverse transcriptase (RT) reaction was performed by use of the OmniScript Kit (Qiagen) with 2000 ng of mRNA. For each PCR reaction 40 ng of cDNA and 0.25 µM of the corresponding primers (Eurofins MWG Synthesis GmbH (Ebersberg, Germany)) were used. Quantitative real-time PCR analysis was performed in triplicates employing the QPCR-SYBR Green Fluorescein Mix (Thermo Fisher Scientific (Dreieich, Germany)) and a CFX96 Real-Time System (BioRad (Munich, Germany)) with the Bio-Rad CFX Manager 3.1 software. PCR runs (35-40 cycles) were done as follows: 95°C -10 min; 95°C -15 s; 60°C -30 s; 72°C -40 s; 72°C -10 min. At the end of the runs, melting curves were analyzed to ensure the specificity of the amplification reaction. mRNA levels of β-actin, GAPDH, PPIA, RPL32 and 18S were taken for normalization. If not stated otherwise, relative mRNA expression of untreated control cells was set to 1.0.

Statistical analysis
For statistical analysis the unpaired two-tailed Student's t-test was applied using the GraphPad Prism 5.01 software. p-Values ≤ 0.05 were considered as significant and were marked with an asterisk. www.impactjournals.com/oncotarget CisPt resistant mesenchymal-like J-82 UC cells are characterized by reduced formation of DNA damage and attenuated DDR Acquired CisPt resistance is reversible by pharmacological inhibition of Chk1.