Background: Castration-resistant prostate cancer (PCa) represents a serious health challenge. Based on mechanistically-supported rationale we explored new thera- peutic options based on clinically available drugs with anticancer effects, including inhibitors of PARP1 enzyme (PARPi), and histone deacetylases (vorinostat), respectively, and disulfiram (DSF, known as alcohol-abuse drug Antabuse) and its copper-chelating metabolite CuET that inhibit protein turnover.Methods: Drugs and their combination with ionizing radiation (IR) were tested in various cytotoxicity assays in three human PCa cell lines including radio-resistant stem-cell like derived cells. Mechanistically, DNA damage repair, heat shock and unfolded protein response (UPR) pathways were assessed by immunofluorescence and immunoblotting.Results: We observed enhanced sensitivity to PARPi/IR in PC3 cells consistent with lower homologous recombination (HR) repair. Vorinostat sensitized DU145 cells to PARPi/IR and decreased mutant p53. Vorinostat also impaired HR-mediated DNA repair, as determined by Rad51 foci formation and downregulation of TOPBP1 protein, and overcame radio-resistance of stem-cell like DU145-derived cells. All PCa models responded well to CuET or DSF combined with copper. We demonstrated that DSF interacts with copper in the culture media and forms adequate levels of CuET indicating that DSF/copper and CuET may be considered as comparable treatments. Both DSF/copper and CuET evoked hallmarks of UPR in PCa cells, documented by upregulation of ATF4, CHOP and phospho-eIF2α, with ensuing heat shock response encompassing activation of HSF1 and HSP70. Further enhancing the cytotoxicity of CuET, combination with an inhibitor of the anti-apoptotic protein survivin (YM155, currently undergoing clinical trials) promoted the UPR-induced toxicity, yielding synergistic effects of CuET and YM155.Conclusions: We propose that targeting genotoxic and proteotoxic stress responsesby combinations of available drugs could inspire innovative strategies to treat castration-resistant PCa.

1| INTRODUCTION
Prostate cancer (PCa) is the most frequently diagnosed malignancy in men and one of the major causes of cancer-related death in developed countries.1 PCa is initially androgen-dependent and responds to androgen deprivation therapies, however, the disease ultimately progresses into a hormone-independent and largely incurable stage with metastases to the bones, lung, brain, or liver.Aberrations in the DNA damage response (DDR) machinery are common in cancer and represent potential targets for therapeutic intervention.2 PARP1 activity is important in sensing and signaling DNA damage that arises both endogenously, for example, through generation of oxidative DNA lesions and DNA single-strand breaks (SSBs), or exogenously, such as due to ionizing radiation (IR) exposure or treatment with various chemotherapeutics. Exposure of cycling cells to inhibitors of PARP1 (PARPi) causes excessive unrepaired SSBs and acceleration of DNA replication3 leading to replication stress and formation of DNA double-strand breaks (DSBs), toxic lesions preferentially repaired by homologous recombination (HR). HR defects due to mutations or silencing of factors such as BRCA1/2 sensitize cells to PARPi, as shown for ovarian, breast and also metastatic prostate cancer.4Defects in DNA damage sensors, signaling kinases or nucleotide excision repair also sensitize to PARPi4 suggesting that the therapeutic potential of PARPi might extend beyond the BRCA1/2-defective tumors. There is also an urgent need to identify and validate potential biomarkers to predict sensitivity of individual tumors to PARPi, exemplified for PCa by the fusion oncogene TMPRSS2-ERG or loss of the PTEN tumor suppressor.

PCa is a heterogeneous disease reflecting both genetic and epigenetic alterations.7 Six epigenetics-modulating drugs targeting DNA methylation or histone deacetylation have been approved for cancer treatment.7 Since epigenetic regulation is complex, preclinical studies are required to generate patient stratification hypotheses and identify predictive biomarkers. Epigenetic reprogramming after loss of Rb and p53 tumor suppressors diminishes androgen receptor expression and is associated with resistence to antiandrogen therapy.8,9 From this point of view, both PC3 and DU145 cells,lacking AR and possessing mutations in tumor suppressors, represent relevant models of a subgroup of aggressive prostate cancer.Another approach to PCa treatment could be drug repurposing, with potentially multifaceted benefits for clinical implementation of new treatment options. DSF is among such possible candidates, showing anti-tumor activity in multiple studies. Recently, we discov- ered the molecular target and mode of action of DSF, thereby strengthening DSF’s potential as an anticancer drug.10 Many cancers become resistant to monotherapy through diverse mechanisms, posing a major challenge in contemporary oncology. Drug combination could overcome resistance to single compounds, thus it is vital to find the drugs that act synergistically and are well tolerated.11Here, we describe differential responses to PARPi and IR in cellular models of aggressive PCa: PC3 (typical for loss of p53 and PTEN), DU145 (mutated p53 and Rb) and radioresistant stem-like PCa cells.12 Moreover, we show that HDAC inhibition alters expression of HR proteins and potentiates cytotoxicity of IR, and that DSF’s active metabolite, diethyldithiocarbamate-copper complex (CuET), activates heat shock response and UPR, showing synergistic toxic effect in combination with a survivin inhibitor-YM155 in human PCa models.

2| MATERIAL AND METHODS
DU145 and PC3 cell lines were cultured in DMEM medium, LNCaP in RPMI medium and LAPC4 in IMDM. DMEM, RPMI, and IMDM media were supplemented with 10% fetal bovine serum and penicillin/ streptomycin. IMDM medium was additionally supplemented with 1 nM R1881. RWPE-1 cells were cultured in a keratinocyte serum-free medium supplemented with the bovine pituitary extract and human recombinant epidermal growth factor (Thermo Scientific, Waltham, MA). EP156T cells were cultured as described previously.13 All cell cultures were maintained in humidified 5% CO2 atmosphere at 37°C. LAPC4, EP156T and RWPE1 cells were kindly provided by Prof. Zoran Culig and Prof. Helmut Klocker (Innsbruck Medical University). Other cell lines were purchased from the European Collection of CellCultures (ECACC) and authenticated by AmpfISTR™ Identifiler PCR Amplification Kit (Applied Biosystems, Foster City, CA).For clonogenic cell survival assay, cells were plated in 6-well plates at 200-500 cells per plate. Next day the cells received appropriate treatment and kept in culture for 7-14 days. Colonies of approximately 50 cells were visualized by 1% crystal violet in 96% ethanol, and their number and total area were counted. Results were confirmed in three independent experiments. For XTT assay, cells were plated at a density of 10 000 per well in a 96-well plate. The next day, cells were treated as indicated. After 48 h, an XTT assay was performed according to the manufacturer’s instructions (Applichem, Darmstadt, Germany).

XTT solution was added to the medium and incubated for 30-60 min, and then the dye intensity was measured at the 475 nm wavelength using a spectrometer (TECAN, Infinite M200PRO, Mannedorf, Switzerland).The KU58948 inhibitor was obtained from AstraZeneca (London, UK). Vorinostat, MK132, nutlin 3, DSF, tunicamycin, thapsigargin and CuCl2 were purchased from Sigma-Aldrich, YM155 from Selleckchem and copper diethyldithiocarbamate (CuET) from TCI Chemicals. Ionizing radiation was delivered using Xstrahl RS research cabinet gamma irradiator.Equal amounts of cell lysates were separated by SDS-PAGE on handcast or precast gel (Invitrogen, Carlsbad, CA), and then transferred onto nitrocellulose membrane. The membrane was blocked with 5% milk in Tris-buffered saline containing 0.1% Tween 20 for 1 h at room temperature, and then incubated overnight at 4°C or 1 h at room temperature with one of the following primary antibodies against: p53 (FL-393, Cell Signaling, Danvers, MA), Rad51 (ab63801, Abcam, Cambridge, UK), GAPDH (GTX30666, GeneTex), alpha-tubulin (H- 300, Santa Cruz, Dallas, TX), BRCA1 (D-9, Santa Cruz), KU70 (N3H10, Santa Cruz), KU80 (ab3107, Abcam), DNA-PKcs (clone 18-2 Thermo Scientific), lamin B (M-20, Santa Cruz), TopBP1 (A300-111A, Bethyl, Montgomery, TX), BRCA2 (A300-005A, Bethyl), ATF4 (ABE387,Merck-Millipore), CHOP (L63F7, Cell Signaling), p-eIF2a (S51, Cell signalling), HSP70(C92FBA-5, Enzo), followed by detection by secondary antibodies: goat-anti mouse and goat-anti rabbit (GE Healthcare). HRP conjugated secondary antibodies were visualized by ECL detection reagent (Thermo Scientific).

After appropriate treatment cells were fixed with 4% formaldehyde for 15 min at room temperature, washed with PBS and permeabilized with 0.5% Triton X-100 in PBS for 5 min. The samples on the plastic inserts cutted directly from cultivation plates using CNC machine were then immunostained with primary antibodies against Rad51 (ab63201,Abcam), cyclin A (6E6, Leica), BRCA1 (D-9, Santa Cruz), p53 (FL-393, Santa Cruz), HSF1 (4356S Cell Signaling), followed by a fluorochrome- conjugated secondary antibodies: Alexa Fluor-488 or Alexa Fluor-568 (Invitrogen). Nuclei were visualized by Hoechst 33342 at room temperature for 5 min before mounting. Images were automatically recorded using an inverted fluorescence microscope BX71 (Olympus) and ScanR Acquisition software (Olympus), analyzed with ScanR Analysis software (Olympus), and evaluated with Statistica software (StatSoft).DU145 cells were transfected with anti-p53 siRNA (Eurofins Geno- mics-GUC CAG AUG AAG CUC CCA GAA) and NT siRNA (Eurofins Genomics-UAA UGU AUU GGA ACG CAU A) using Lipofectamine RNAiMAX transfection reagent (Invitrogen) according to manufac- turer’s recommended protocol. After 24 h, cells were either collected for Western blot analysis or used for immunofluorescence analysis.

Cells were harvested at indicated times after treatment (both adherent and detached cells were collected) and fixed in cold 70% ethanol. After treatment with RNaseA, samples were stained with propidium iodide (PI). Cellular DNA content was analyzed using flow cytometer BD FACSVerse (BD Biosciences), and collected data were processed using BD FACSuite (BD Biosciences). At least 10 000 cells per sample were analyzed.Activity of caspase-3 and -7 was quantified by cleavage of fluorogenic substrate CellEvent™ Caspase-3/7 Green Detection Reagent (Ther- moFisher Scientific). Briefly, samples were prepared in staining buffer (140 mM NaCl, 4 mM KCl, 0.75 mM MgCl2, 10 mM HEPES) supple- mented with 2% FBS, 0.5 µM CellEvent™ Caspase-3/7 Green Detection Reagent and incubated for 45 min at room temperature in the dark.Subsequently, 0.5 µg/mL DAPI was added and samples were analyzed by flow cytometry using BD FACSVerse (BD Biosciences), at least 10 000 events were acquired per sample. Collected data were processed by BD FACSSuite (BD Biosciences).To measure the formation of diethyldithiocarbamate-copper complex (CuET) in vitro a complete cell culture medium (DMEM, 10% FBS, 1% penicillin/streptomycin) was incubated with 1 µM disulfiram or 1 µM disulfiram plus 1 µM copper (ii) chloride, and 1 µM CuET as a control. After 3 h of incubation in 37 °C, 5% CO2, the samples were vortexed and mixed with acetone in a ratio 1:250. The mixture was centrifuged 18 000g for 2 min at 4°C. The CuET complex in supernatant was analyzed by HPLC-MS method as described previously.10 The quantification of CuET complex was calculated according to the calibration curve.

3| RESULTS
The standard-of-care therapy for localized PCa is radical prostatectomy followed by fractionated radiotherapy. In patients with disseminated PCa, androgen deprivation is achieved either by surgical or chemical castration. However, tumors often become castration-resistant asdisease progresses.14 Human PC3 and DU145 cell lines both lack androgen receptors and thus represent useful models for PCa patients with androgen-independent tumor growth.8Recent findings showed high response rates to PARPi treatment in patients with PCa defective in DNA repair genes.4 Using colony formation assays that mimic effects of long-lasting therapy, we found PC3 cells more sensitive to the PARPi than DU145 cells (Figure 1A), while normal prostate epithelial RWPE1 and EP156T cells did not respond within the 1000 nM range (Supplementary Figures S1A and S1B). As PARP inhibitors are also candidate radiosensitizers, we testedcombined PARPi and IR to explore potential additive/synergic effects. DU145 and PC3 cells were pre-treated with 100 nM and 1 µM PARPi and irradiated after 24 h. Although PC3 cells were less responsive than DU145 to IR alone, the combination with PARPi was more effective in PC3 than in DU145 cells (Figure 1A). These data suggest that PC3 cells respond well to PARPi monotherapy or combined with IR, while DU145 respond rather poorly, a phenomenon which we decided to study further mechanistically. PARPi is particularly effective in treatment of breast and ovarian cancer with BRCA1/2 mutations.15 BRCA1 along with Rad51 and other factors mediate HR, a high-fidelity DNA repair of DNA DSBs during S and G2 phases of the cell cycle.

As PC3 cells responded well to PARPi and the combination with IR compared to DU145 cells, the functional status of HR- repair was examined using immunofluorescence analysis of RAD51 foci as marker of active HR. These experiments involved pre- treatment of cells with PARPi for 24 h, subsequent IR (4 Gy) and further incubation for 1, 2, 5, or 10 h. Fixed cells were then co-stained for RAD51 and the S/G2 marker cyclin A to focus on the HR-relevant cell- cycle phases (Figures 1B and 1C).16 Quantification showed reduced RAD51 foci in PC3 cells compared to DU145 in cyclin A-positive cells (Figures 1D and 1E) supporting the hypothesis of insufficient HR to explain higher sensitivity of PC3 cells to PARPi. These data are consistent with the notion that HR defects sensitize cancer cells to PARPi, alone or combined with IR17 and extend this concept to PCa.Since DU145 cells display relative resistance to PARPi and the combined PARPi/IR treatment (Figure 1A), we sought to identify a drug able to sensitise this PCa model to PARPi. DU145 harbours p53 mutations (P223L and V274F) thereby providing a model matching PCa patients harbouring p53 mutation with limited treatment options and adverse prognosis.18 We chose the FDA-approved histone deacetylase inhibitor vorinostat (also known as SAHA), reportedly preferentially cytotoxic towards cancer cells with mutated p53.19 Indeed, DU145 cells responded well to vorinostat (Figure 2A) and were more sensitive compared to PC3 (Supplementary Figure S2A). In DU145 cells, vorinostat caused activation of apoptosis markers caspases 3/7 (Supplementary Figure S2E) and G2/M arrest, as determined by flow cytometry (Supplementary Figure S2D) and accumulation of prometaphase cells (Supplementaty Figure S2B).

Unfortunately, in the short-term viability assay normal prostate epithelial cells RWPE-1 and EP156T respond similarly, thereby questioning the therapeutic window of vorinostat monotherapy (Supplementary Figures S2F and S2G). Mechanistically, vorinostat treatment should evoke degradation of the accumulated mutant p53 protein reverting its anti-apoptotic effect.19 Indeed, downregulation of p53 by vorinostat (Figures 2C and 2E) was mediated by increased p53 degradation, rescuable by proteasome inhibitor MG132 or nutlin, an inhibitor of MDM2 ubiquitin ligase for p53 (Figure 2D). Importantly, pre-treatment with vorinostat also sensitized the DU145 cells to IR and PARPi (Figures 2B and S2C) suggesting possible impact ofvorinostat on the DDR machinery. This phenomenon was further explored as combinations of IR and/or PARPi with vorinostat could potentially represent feasible treatment strategies.To elucidate how vorinostat potentiates the effects of IR and PARPi, we assessed its impact on the DDR pathways. First, as HDACs regulate gene expression, we examined the levels of multiple HR factors after vorinostat treatment, and observed modest yet noticeable decreases of BRCA1, BRCA2, Rad51, and TopBP1 proteins (Figure 3G). Interestingly, despite the lower total BRCA1 level (Figure 2F), the ability to form IR-induced BRCA1 foci remained unchanged (Figures 2G and 2H). Notably, vorinostat pre-treatment prevented formation of IR-induced Rad51 foci in cyclin A-positive cells (Figures 3A and 3B), suggesting robust impairment of HR explaining the acquired sensitivity to PARPi.

This effect is unlikely attributable to vorinostat-mediated downregulation of mutant p53, because direct downregulation of mut-p53 in DU145 cells by siRNA did not reproduce such phenotype (Figures 3C-E). Interestingly, Ku70, Ku80, and DNA-PK, proteins involved in DSB repair via non- homologous end joining (NHEJ), remained unaffected upon vorinostat treatment (Figure 3F) consistent with differential transcription control of genes involved in distinct DNA repair pathways.20 As radio- resistance in PCa represents a significant issue that lacks suitable cellular models, our team developed a model of radiosurviving PCa cells obtained by exposure of parental DU145 cells to clinically relevant daily fractions of IR to a cumulative dose of 64 Gy (2 Gy applied every 24 h for 32 days). This treatment is not 100% toxic and selects for a radiation-surviving, stem-like cell population.12,21 Importantly, pre-treatment with vorinostat sensitised such cells to IR in colony formation assay (Supplementary Figure S2H) further suggesting vorinostat as an interesting option for combined IR treatment.Prostate, as a mainly secretory organ, is especially dependent on proper function of endoplasmic reticulum (ER) and ER-associated degradation (ERAD). ERAD malfunction or insufficiency leads to ER stress and activation of the unfolded protein response (UPR).22 Several factors of ERAD machinery are upregulated in PCa,23 and UPR activation in PCa has been recently demonstrated, providing a possible vulnerability exploitable therapeutically.

We have recently shown that DSF targets cancer via inhibition of the p97/NPL4 pathway, essential for ERAD.10 DSF’s anticancer activity depends on copper25 and we showed that in vivo, DSF becomes converted into diethyldithiocarbamate, a strong copper chelator forming a stable (CuET) the ultimate anticancer metabolite of DSF.10 CuET accumulates in tumors and paralyzes p97/NPL4-dependent processing of proteins, leading to strong proteotoxic stress, UPR and heat shockresponses (HSR).10 Since this drug is clinically used and well tolerated, it is an ideal candidate for repurposing. Specifically for PCa, DSF might be an interesting therapeutic candidate as it scored highly in PCa cell line models.26First, we treated DU145, PC3 and radiosurviving DU145 cells by DSF, DSF + CuCl2, CuCl2 alone or CuET for 48 h to test for cytotoxicity. All cell lines responded with similar sensitivity within nanomolar range (IC50 around 200 nM) to DSF + CuCl2 and CuET (Figure 4A). To further explore the comparable potencies of DSF + CuCl2 and CuET, we assessed whether CuET forms also in vitro, in media containing DSF and CuCl2. Indeed, we confirmed that CuET complex forms efficiently, indicating that the cell culture effects under DSF + CuCl2 treatment are attributable to CuET (Figure 4B). DSF treatment alone was moderately toxic, likelyreflecting the presence of copper ions in standard growth media, forming some CuET. Notably, unlike treatments with PARPi or HDACi there was an obvious lack of differential responses among the otherwise very heterogeneous cell lines, suggesting a mecha- nisms of action independent of the p53 status or DNA repair defects. To confirm that PCa cells treated by DSF + CuCl2 and CuET are experiencing stress phenotypes similar to other cellular models,10 PCa cells were first examined for activation of HSR.

Immunofluo- rescence analysis confirmed a robust HSR manifested by formation of HSF1 nuclear stress foci27 (Figures 4C and 4D) and increase of heat shock protein 70 (HSP70), the main HSR effector, in all tested cell lines (Figure 4E). The PCa cells also strongly activate UPR manifested by elevated ATF4, CHOP, and phospho-eIF2α, estab- lished UPR markers22 (Figures 5A and 5B).DSF’s toxicity for PCa cell lines26,28 inspired a small pharmacodynamic clinical trial in PCa patients with non-metastatic recurrent prostate cancer.29 The trial failed to show either global demethylation as a presumed pharmacodynamic marker28 or significant changes in PSA levels, consequently concluding that such DSF monotherapy was inefficient in PCa patients. Such failure might reflect, at least in part, the fact that copper was not included into this trial, thus limiting DSF’s anticancer activity that is otherwise apparent from preclinical studies including mouse models.10,30 A new Phase Ib study of intravenous copper loading combined with oral DSF administration in metastatic castration resistant prostate cancer was lunched recently, which should provide more conclusive information about DSF efficacy in patients (ClinicalTrials.gov Identifier: NCT02963051). As DSF alone could be insufficient for eradication of PCa cells in vivo combined therapy could provide a better option. Because UPR, robustly induced by DSF + CuCl2 and CuET treatments, strongly activates cell death, such candidate combinational treatment strategy could exploit inhibition of pro-survival proteins that are known to be overexpressed in cancers, such as survivin.31 Chemical inhibitor of survivin, YM155, showed anticancer activity in preclinical cancer models includingPCa32 and is being evaluated in clinical trials.

Interestingly, synergistic toxicity between YM155 and common UPR inductors thapsigargin and tunicamycin has been recently reported.34 However, these two UPR inducers are very toxic and unsuitable for clinical applications.35 On the other hand, DSF (combined with copper) is relatively well tolerated and thus provides a viable option to potentiate survivin inhibitors. Motivated by this rationale, DSF + CuCl2 and CuET were first compared with thapsigargin and tunicamycin and very good potency in UPR induction was confirmed (Figures 5A and 5B). Next, DU145 and PC3 cells were treated with indicated combinations of DSF (with copper) and YM155. Combination of the drugs led to reduced survival of both DU145 and PC3 cells. (Figures 5C and 5D) revealing moderate synergy as computed using CompuSyn algorithm36 (Figure 5E). Thus combination of two clinically available drugs, YM155 and DSF (supplemented with copper) represents a readily available and potentially efficient treatment option for PCa and also other cancer patients.

4| DISCUSSION
Therapy of advanced PCa still poses a serious challenge in oncology, making any innovative and better alternative treatments highlydesirable. Here we chose two well-characterized cellular models (PC3, DU145) and one experimentally derived model (termed radio- surviving DU145) of castration resistant PCa to explore new therapeutic options. We concentrated on anticancer drugs currently entering or in clinical trials such as PARPi, vorinostat, and DSF, the latter with recently revealed mechanism of action through interference with p97/NPL4- mediated protein turnover. IR was added in some experimental setups as it is known that the standard androgen deprivation treatment may benefit from combination with radiother- apy in locally advanced prostate cancer.37Compared to DU145, PC3 cells showed higher sensitivity to PARPi and IR. Analogous observations were published by others and the differential sensitivity was associated in part with PTEN loss and induction of senescence in PC3 cells.38–40 Here, we add another clue as PC3 cells show low Rad51 foci formation after PARPi and IR suggesting defects in HR-promoted DNA repair. Defects in HR are regarded as a major prerequisite for synthetic lethality in combination with PARPi.41 Based on a phase II clinical trial, combined with next-generation sequencing of DNA repair genes, the PARPi olaparib (Lynparza) received an FDA designation of breakthrough therapy.

Our present results suggest a potential for PARPi in treating PCa, guided by immunohistochemi- cal and/or ex vivo biopsy evaluation of HR biomarkers such as RAD51 foci formation.43 These approaches, while technically challenging, have a potential for clinical implementation aspredictive biomarkers for treatment with PARPi, complementary to genetic analyses of BRCA1/2, ATM or TMPRSS2-ERG.4,6,42 Olaparib has also been recently reported to be effective in combination with, and as maintenance therapy after, first-line endocrine therapy of prostate cancer.44Different therapeutic approaches will be required for castration- resistant PCa cases that are HR repair proficient. Based on our current data, we propose another treatment strategy, involving HDAC inhibitors such as vorinostat. DU145 are among the cell lines with gain-of function p53 mutations,18,45 associated with preferential sensitivity to HDAC inhibition.19 Indeed, these cells responded well to vorinostat, particu- larly when combined with IR, as also noticed by others.46 Consistent with the literature, we observed reduced p53, and modest down- regulation of BRCA1, BRCA2 and Rad51 after vorinostat treatment by immunoblotting. For the first time, we report that vorinostat down- regulates also TOPBP1 which is important for Rad51 loading to chromatin during HR.47 Indeed, pre-treatment with vorinostat resulted in less efficient DNA repair by HR, as documented by lower counts of Rad51 foci in cyclin A-positive cells. Downregulation of mutated p53 by siRNA altered neither Rad51 foci nor the RAD51 protein level, indicating that the effect of HDAC inhibition by vorinostat is more pleiotropic affecting also the HR-promoted DNA repair processes.

Consistently, other HDAC inhibitors, MS-275 and FK228, impaired HR repair.48 The vorinostat-induced DNA repair defect was further corroborated in our experiments of combined treatment with IR and PARPi.As targeting proteotoxic stress pathways represents an emerging promising therapeutic approach for PCa,24 we also tested DSF that impairs protein degradation.10 DSF repurposing for cancer treatment is currently tested in at least eight clinical trials (according to ClinicalTrials.gov) involving various cancers including PCa. Despite DSF monotherapy failed in a clinical trial in PCa patients with non- metastatic recurrent PCa,29 this study did not combine DSF with copper, which is required for DSF’s anticancer activity in vitro25,49 and potentiates its activity in vivo.10,30 Another intriguing option for future treatment is concomitant DSF (ideally supplemented with copper) with other anticancer drugs or IR. Such combinations show promising results in preclinical models50,51 and also in a few clinical trials.52,53 In this study, we demonstrated toxicity of the CuET complex, the main anticancer metabolite of disulfiram in vivo10 as well as potency of DSF in combination with copper. These treatments also induced cellular responses which were reported for other cell lines, including UPR and HSR pathway activation. Such strong activation of UPR prompted us to test the combination of DSF with a survivin inhibitor YM155, reported as being highly potent in combination with UPR inducers thapsigargin and tunicamycin.34 YM155 is a novel anticancer drug undergoing clinical trials and it was already tested as a monotherapy in castration- resistant PCa patients, yet with a rather limited effect.33 The observed synergy between YM155 and CuET/DSF + CuCl2, reported in our present study, provides a further rationale for additional preclinical and/or clinical investigations, with potential implications also for other human malignancies, beyond the treatment of PCa.

5| CONCLUSIONS
Combined IR/PARPi effectively killed HR-impaired PCa cells. Vorino- stat treatment reduced levels of HR factors including TOPBP1, with ensuing enhanced sensitivity to IR and PARPi. DSF/copper was effective against all PCa models, triggering proteotoxic stress, UPR and heat shock pathway activation, highlighting a rationale for combinato- rial therapy blocking anti-apoptotic responses by survivin inhibitors. We propose that targeting genotoxic stress and proteotoxic stress responses by combinations of available drugs could inspire innovative strategies to treat castration-resistant PCa.