Downregulation of Human DAB2IP Gene Expression in Renal Cell Carcinoma Results in Resistance to Ionizing Radiation

Ionizing radiation (IR) can be a very effective regimen for targeting localized tumors or areas of invasion after surgical resection (1). Recent advances in external beam radiotherapy have significantly improved the therapeutic index by combining both imaging-guided precision targeting with the delivery of high doses using three-dimensional fractionation. The major trigger of the cellular response to IR is its destructive impact on genome integrity. Cells can mount a coordinated response to IR by activating a network of interacting signaling pathways, collectively known as the DNA damage response (DDR; ref. 2). There are two main pathways to repair DNA double-strand breaks (DSB): nonhomologous end joining (NHEJ) and homologous recombination (HR; refs. 3, 4). In response to DSB formation, it has been shown that phosphorylation of H2AX occurs in the position of Ser139, then many DDR proteins such as RAD50, MDC1, and BRCA-1 attracted to γH2AX foci (5, 6). Subsequently, DNA-binding enzyme PARP-1 is recruited to modulate the activity of the DNA repair systems (7, 8) and has a primary role in the process of poly(ADP-ribosyl)ation, which is responsible for the major poly(ADP-ribosyl)ation activity observed during DDR. PARP-1 binds to the damaged DNA sites and initiates the formation of a poly-ADP scaffold that recruits other members of DDR pathway, indicating its pivotal role in DNA repair after DNA-damaging agents (9). Because the ability of cells to effectively execute DDR signaling is essential for restoring genomic stability and for promoting survival following DNA damage, PARP-1 is a fundamentally important member in response to DNA-damaging agent and overexpression of PARP-1 is often found in many cancers and considered to contribute to progression of cancer such as BRCA-mutated ovarian and breast cancer (10–12).

Renal cell carcinoma (RCC) accounts for 90% of renal cancer and its incident rate has risen during previous decade (13). Primary treatment for localized RCC is surgical resection; however, 30% of patients still continue to develop metastatic disease after surgical resection (14, 15). In metastatic cancer, radiotherapy has been used for palliation routinely for brain and other extracranial lesions with respectable response rates, but a significant proportion of RCC tumors are highly radioresistant with conventional radiation (14, 15). The mechanism associated with IR-resistance of RCC is not fully understood yet, and deeper understanding of the responsible mechanisms would provide highly appealing targets for clinical therapy.

DOC-2/DAB2 interactive protein (DAB2IP), a potent tumor suppressor, is frequently lost in RCC (16, 17). DAB2IP is known to regulate various biological process including cell survival, apoptosis, and epithelial-to-mesenchymal transition (EMT) through the inhibition of several pathways (18). We have further demonstrated the comprehensive inhibitory mechanisms of DAB2IP on cancer stem cell regulation (19, 20). Particularly, we first demonstrated the nuclear localization of DAB2IP that can impact on gene transcription (19). In this study, we observed that loss of DAB2IP in RCC cells exhibit IR-resistance and that restoration of DAB2IP expression resensitizes them to IR. We further identified that DAB2IP can directly interact with PARP-1 protein and affects PARP-1 protein turnover by recruiting E3-ligases (e.g., RanBP2, TRIP12, and RNF40). Indeed, PARP-1 protein expression was inversely correlated with DAB2IP expression.

As mentioned, the biological consequences of radiation leading to cell death are influenced by the activation of DDR in target cells (2). Our results clearly show that elevated PARP-1 in RCC cells may underlie IR-resistance by accelerating DDR. In contrast, knocking-down PARP-1 expression in IR-resistant RCC cells significantly increases their response to IR. Considering PARP-1 as a druggable target, we evaluated a PARP-1 inhibitor (PARPi; Olaparib) in an in vivo RCC xenograft model and a patient-derived xenograft (PDX) model in combination with radiotherapy and demonstrate a significant improvement in the therapeutic efficacy of radiotherapy. Overall, this study not only unveils a mechanism of radioresistance in RCC but also provides a rational therapeutic strategy to resensitize radioresistant RCC.

Human RCC cell lines (786O, 769P, and ACHN) were obtained from ATCC and cultured in RPMI1640 (Life Technologies) containing 10% FBS. Immortalized human proximal tubular cell line, HK cell, was obtained from the China Center for Type Culture Collection and cultured in RPMI1640 containing 10% FBS, and transformed human embryonic kidney, HEK293T cells were from ATCC and cultured in DMEM (Life Technologies) containing 5% FBS. All these cell lines were used within 15 passages and authenticated with the short tandem repeat (STR) profiling by Genomic Core in UT Southwestern (UTSW) periodically and Mycoplasma testing was performed by MycoAlert Kit (Lonza Walkersville, Inc.) every quarterly to ensure mycoplasma-free.

The DAB2IP expression plasmid was prepared as described previously (21), and expression plasmid for human PARP-1 was obtained from Dr. Binhua Zhou (University of Kentucky College of Medicine, KY; ref. 22). Gene knockdown was performed using pLKO.1 plasmid expressing gene-specific shRNA and purchased from the National RNAi Core Facility of Taiwan (Academia Sinica, Taipei, Taiwan). Cells were plated with 70% confluence and transfection was carried out by using Xfect (Clontech) according to the manufacturer's instructions.

Primary antibodies used were as follows: rabbit polyclonal anti-RAD51, mouse polyclonal anti-PARP-1, and monoclonal anti-BRCA1 (Santa Cruz Biotechnology), rabbit polyclonal anti-ubiquitine (Cell Signaling Technology), anti-phospho-γ-H2AX (Ser139; Millipore), mouse monoclonal anti-Flag and anti-actin (Sigma-Aldrich), and mouse monoclonal anti-Ku70 and anti-Ku80 were homemade antibodies.

For in vitro experiment, cells were irradiated in ambient air using a JL Shepherd Mark 1-68 ¹³⁷Cs irradiator at a dose rate of 3.47 Gy/min (23). For in vivo irradiation, mice were anesthetized and subcutaneous tumors were focally irradiated using XRad 320 (Precision X-Ray, 250 kV/s, 15 mA, 20.8 Gy/min).

Exponentially growing cells were trypsinized and counted. Cells were diluted serially to appropriate concentrations and plated onto 60-mm dish for 4 hours. Then, cells in triplicates were treated with increasing doses of IR (0, 2, 4, 6, and 8 Gy). After 6 days of incubation, the colonies were fixed and stained with 4% formaldehyde in PBS containing 0.05% crystal violet. Colonies containing more than 50 cells were counted. Surviving fraction was calculated as (mean colony counts)/[(cells inoculated for irradiation) × (plating efficiency)], in which plating efficiency was defined as (mean colony counts)/(cells inoculated for control). The data are presented as the mean ± SEM of at least three independent experiments. The curve S = e – (αD + βD2) was fitted to the experimental data using a least square fit algorithm using the program Sigma Plot 11.0 (Systat Software).

A DSB repair assay was done by counting phospho-γH2AX foci after radiation. Cells seeded on glass coverslip were allowed to attach, and were exposed to a total dose of 2 Gy radiation. Cells were fixed with 4% paraformaldehyde and permeabilized with 0.5% Triton X-100 for 20 minutes at room temperature. Then the samples were blocked with 5% BSA and 1% normal goat serum for 1 hour. Then, cells were incubated with primary antibody, anti-phospho-Histone γH2AX (Ser139; 1:2,000) overnight at 4°C. Samples were washed three times for 5 minutes each in PBS, and then incubated with Alexa Fluor 488–conjugated secondary antibody for 1 hour. Nuclei were counterstained with DAPI and stained cells were analyzed under a fluorescence microscope (BZ-X710; Keyence).

HEK293T cells transfected with flag-tagged DAB2IP were lysed using NE-PER Nuclear and Cytoplasmic Extraction Reagents (Thermo Fisher Scientific) and nuclear fraction was immunoprecipitated with anti-flag antibody. The protein complex mixtures were run into the SDS-PAGE gel and stained with Coomassie-Blue. Identification of proteins in bands cut from Coomassie-stained gel was performed by Orbitrap Elite mass spectrometry platforms (Thermo Fisher Scientific), using short reverse-phase LC/MS-MS methods. Proteins were identified from samples using our in-house data analysis pipeline (CPFP) of proteomics core at The University of Texas Southwestern Medical Center at Dallas.

The immunocomplexes were precipitated with Dynabeads Protein G (Life Technologies) and subjected to Western blot analysis. For Western blot analysis, cells were lysed and subjected to electrophoresis on 4% to 12% Bolt gels (Life Technologies). Separated proteins were electroblotted onto nitrocellulose membranes, and membranes were incubated with 5% nonfat dry milk (w/v) for 1 hour and then washed in PBS containing 0.1% Tween 20. Membranes were then incubated with primary antibody, and antibody binding was detected using the appropriate secondary antibody coupled with horseradish peroxidase.

Formalin-fixed, paraffin-embedded sections were de-paraffinized, rehydrated, and subjected to heat-induced antigens retrieval (citrate buffer, pH 6.0). Sections were blocked with goat serum, and incubated with appropriate primary antibody and developed with 3,3′-diaminobenzidine chromogen followed by counterstaining with hematoxylin and eosin.

Cells were transfected with pcDNA3.1 His-Ubiquitin and flag-DAB2IP. Approximately 36 hours after transfection, cells were treated with 10 μmol/L MG132 for 6 hours. Then, cells were harvested and the input fraction was prepared using RIPA buffer. His-tagged protein was pulled down using Dynabeads His-Tag Isolation & Pulldown (Life Technologies) according to manufacturer's instruction. The eluted samples were collected and both of the input and pull-down fraction were subjected to SDS-PAGE and Western blot analyses.

All animal work was approved by the Institutional Animal Care and Use Committee. ACHN (3 × 10⁶ cells/site) cells mixed with 50% Matrigel (BD Biosciences) in 0.1 mL were injected subcutaneously as well. Animals were randomly divided into 4 groups including untreated control, radiation alone (2 Gy), Olaparib alone (15 mg/kg), and combination treatment of radiation and Olaparib. When tumors developed to a measurable size, mice were treated with a total dose of 2 Gy using the XRad 320 (Precision X-Ray). Then, Olaparib was given orally every day for 2 weeks. Tumor volume (cubic millimeters) was measured and calculated by using the ellipsoid formula (π/6 × length × width × depth).

For PDX model, frozen RCC–PDX tissue samples (20–30 mm³) from UT Southwestern Kidney Cancer and SPORE Program were implanted subcutaneously in the right posterior flanks of 6-week-old male NOD/SCID mice (24). When tumors reached 150 to 200 mm³, tumor-bearing mice were randomized, anesthetized with isoflurane, and subjected to a single fraction of X-rays. Animal body weights and tumor volumes were measured twice weekly until tumor volume reached a maximum size (2,000 mm³) or up to 120 days after radiation treatment. X-rays were delivered to tumor area using a 10-mm-diameter collimator attached to the Xrad-320 small animal irradiator to minimized radiation exposure to the rest animal body.

All error bars in graphical data represent mean ± SD. Student two-tailed t test was used for the determination of statistical relevance between groups, and P < 0.01 was considered statistically significant. All statistical analyses were performed with GraphPad Prism software.

DAB2IP is a tumor-suppressor protein capable of modulating the cytoplasmic steps of various oncogenic pathways (18). In RCC, we have found that expression of DAB2IP is lost in high-grade tumors from different subtypes of RCC, which is correlated with poor survival and resistance to therapy. Also, loss of DAB2IP can promote stem cell phenotypes in RCC cells (20). Thus, we further investigated whether decreased DAB2IP expression enhanced radioresistance in RCC. Indeed, DAB2IP-negative RCC cell lines such as ACHN and 769P were more resistant to IR than their counterparts with ectopic expression of a DAB2IP cDNA; the clonal survival results indicated that DAB2IP-positive cells (D) were significantly sensitive to IR compared with those transfected with vector control (Neo; Fig. 1A). For example, the survival fraction at 2 Gy (SF₂) for ACHN and 769P was reduced from 0.50 ± 0.005 and 0.47 ± 0.019 to 0.34 ± 0.027 and 0.33 ± 0.009, respectively.

At the genomic level, IR can elicit both single- and double-strand DNA breaks that evoke a multifaceted DDR. Therefore, enhanced DNA repair is a mechanism whereby cells may become resistant to IR (2). An early step in DSB repair involves the rapid phosphorylation of γH2AX at damaged sites (5, 25). To measure the induction and repair of IR-induced DSBs, cells exposed to a total dose of 2 Gy were collected at the indicated times and subjected to immunofluorescence staining for phospho-γH2AX. DSB repair kinetics can be determined by counting the phospho-γH2AX remaining foci over time. The results showed that the rate of DSB repair was significantly increased in DAB2IP-negative cells compared with DAB2IP-positive cells (Fig. 1B). The initial expression levels of phospho-γH2AX remaining foci were not different in DAB2IP-negative and -positive cell line. Rapid induction of DNA damage by IR was detected within 0.5 hour in all of the cell lines similarly, but more than 90% repair was completed at 8 hours following IR in DAB2IP-negative (i.e., Neo) cells, whereas DAB2IP-positive cells still retained nearly 50% of foci even after 24 hours. These results suggest that loss of DAB2IP in RCC enhances IR-resistance by accelerating DSB repair kinetics.

In addition to GTPase activating protein activity, several studies have demonstrated that DAB2IP can function as a scaffold protein to interact with different enzymes or transcriptional factor involved in epigenetic regulation (19, 20). Knowing the nuclear localization of DAB2IP, we decided to explore whether DAB2IP can associate with DDR machinery. Nuclear proteins were immunoprecipitated with DAB2IP antibody and subjected to mass spectrometry; 538 protein candidates were identified (spectral counts > 3, ratio against IgG > 2). We further analyzed these candidate proteins based on the highly interconnected networks using Reactome (26). These proteins were involved in various biological processes including metabolism, ubiquitination, and cell cycle. Noticeably, several proteins involved in DNA repair were ranked highly in gene ontology (GO) enrichment analyses (Fig. 2A). Among them, particularly, PARP-1 was consistently and strongly associated with endogenous DAB2IP in several RCC cell lines determined by reciprocal IP (Fig. 2B). PARP-1 is a ubiquitous enzyme that transfers ADP-ribose units from donor β-NAD onto various substrate proteins (7), and commonly involved in many cellular functions such as transcriptional modulation or DDR which can facilitate repair of radiation-induced DNA breaks (8). Also, DAB2IP status does not affect the expression of key repair factors in NHEJ (Ku70, Ku80) and HR (RAD51, BRCA1) in ACHN and 769P cells (Supplementary Fig. S1A). Thus, we further investigated the impact of DAB2IP on PARP-1 and observed an inverse correlation between DAB2IP and PARP-1 protein expression (Fig. 2C). PARP-1 protein expression was dramatically increased after knocking down endogenous DAB2IP in HK and 786O cells, whereas an ectopic expression of DAB2IP in 769P or 786O KD cells could suppress the expression PARP-1 protein. Consistently, the enzymatic activity of PARP-1 determined by in vitro assay was significantly reduced in cells with increased DAB2IP (Fig. 2D).

Importantly, in addition to in vitro models, we also discovered that PARP-1 protein was highly elevated in kidney tissues derived from DAB2IP-knockout mice (KO) compared with that from wild-type mice (Fig. 2E and F). In contrast to the significant suppression of PARP-1 protein levels by DAB2IP, PARP-1 mRNA expression levels remained unchanged in RCC cell lines or specimens (Supplementary Fig. S1B and S1C), suggesting that DAB2IP regulates PARP-1 protein expression in posttranscription level. As we had observed that both DAB2IP and PARP-1 can be found together in a complex on immunoprecipitation (IP) experiments, we conjecture whether this interaction may regulate PARP-1 levels.

To determine the impact of DAB2IP on PARP-1 protein expression, we measured the half-life of PARP-1 reflecting its protein stability. We found that PARP-1 half-life in DAB2IP-negative cells was much longer than that in DAB2IP-positive cells, indicating that DAB2IP is involved in PARP-1 protein turnover (Fig. 3A). Indeed, the accumulation of ubiquitinated PARP-1 protein was significantly increased in DAB2IP-positive cells whereas the total PARP-1 was reduced (Fig. 3B). Furthermore, treatment of proteasome inhibitor (MG132) could rescue the expression of PARP-1 protein and enzymatic activity even in DAB2IP-positive cells (Fig. 3C and D). These data indicate that PARP-1 protein turnover requires DAB2IP and involves the ubiquitin–proteasome system (UPS).

Next, we performed an in vivo ubiquitination assay to further confirm the impact of DAB2IP on regulating PARP-1 protein turnover through the UPS. As shown in Fig. 4A, incremental DAB2IP expression in HEK293T or 786O cells resulted in increased ubiquitinated PARP-1 in a dose-dependent manner, suggesting that DAB2IP-mediated UPS plays an important role in regulating PARP-1 protein expression posttranslationally. Although, protein ubiquitination through UPS is known to play a critical role in regulating many biological processes such as cell-cycle progression, DNA repair, apoptosis (27), how PARP-1 is ubiquitinated is not fully characterized. During UPS protein degradation, E3-ligases carry out the final step that catalyzes transfer of ubiquitin from an E2 enzyme to form a covalent bond with a substrate lysine (28). The E3-ligases therefore determine the substrate specificity, which reflects the existence of numerous different E3-ligases compared with few E1 and E2 enzymes (29). To identify the candidate E3-ligases recruited by DAB2IP during PARP-1 degradation, we performed IP followed by mass spectrometry (IP-MS) from MG132-treated cells. Briefly, HEK293T cell transfected with DAB2IP expressing plasmid was treated with MG132 for 6 hours, then cell lysates were immunoprecipitated with PARP-1 or DAB2IP antibody and subjected to MS. Based on IP-MS data, we found three E3-ligases (i.e., RanBP2, TRIP12, and RNF40) consistently associated with both PARP-1 and DAB2IP as potential candidates (Supplementary Fig. S2A). We further performed IP to validate the interaction both DAB2IP and PARP-1 with the 3 different E3-ligases in DAB2IP-negative (293T-Neo) versus DAB2IP-positive (293T-D) cells; the protein–protein interaction was clearly detected in 293T-D cells but not in 293T-Neo cells, indicating that the presence of DAB2IP can accelerate these E3-ligases to degrade PARP-1 protein (Fig. 4B).

To examine the functional role of individual E3-ligases on PARP-1 degradation, DAB2IP-positive RCC cells were transfected with shRNA specifically targeting each E3-ligase. As shown in Fig. 4C, knockdown of the individual E3-ligases remarkably decreased ubiquitinated PARP-1 levels and increased PARP-1 protein expression even in the presence of DAB2IP. These results indicate RanBP2, TRIP12, and RNF40 are critical mediators in PARP-1 protein degradation system, but DAB2IP is required to finish UPS. Noticeably, VHL protein, an E3-ligase commonly associated with RCC development, is not associated with DAB2IP from mass-spectrometry result and do not elicit PARP-1 protein degradation.

To further confirm our findings, TCGA database was analyzed for possible clinical relevance of these 3 E3-ligase in RCC, and RanBP2 was found to be significantly downregulated in tumor tissues compared with the adjacent normal tissues and this correlated with disease progression (Supplementary Fig. S2B) and overall survival of patients with RCC (Supplementary Fig. S2C). Consistently, knocking down of RanBP2 mRNA expression restored PARP-1 enzyme activity in DAB2IP-positive cells, suggesting the critical functional role of RanBP2 in degrading PARP-1 in RCC (Fig. 4D). Furthermore, the stable RanBP2 knockdown RCC cells (Fig. 4E) acquired radioresistance compared with parental DAB2IP-positive cells (Fig. 4F). Altogether, this is a novel mechanism of regulation of PARP-1 protein degradation by specific E3-ligase mediated by a scaffold function of DAB2IP.

We next evaluated the rescue effect of PARP-1 on radioresistance. As shown in Fig. 5A and B, constitutive overexpression of PARP-1 induced radioresistance in DAB2IP-positive cells, whereas knockdown of PARP-1 increased IR-sensitivity in DAB2IP-negative cells.

The role of PARP-1 in radioresistance of RCC highlights the potential application of PARPi on radiotherapy in RCC tumors. The efficacy of either a single-agent PARPi or in combination with other cytotoxic drugs has been demonstrated in many types of cancer (30, 31) but not in RCC. Olaparib (AZD2281) is a small molecule PARP-1/2 inhibitor, currently under evaluation in many clinical trials of breast, uterine, colorectal, and ovarian cancers, already FDA-approved for some indications (32, 33). To explore the effect of Olaparib on radioresistant RCC cells (i.e., DAB2IP-negative cells), cells were treated with Olaparib for 6 hours prior to IR. The results showed that Olaparib could sensitize those resistant cells to IR in a dose-dependent manner (Fig. 5C). Therefore, we found that PARP-1 is responsible for radioresistance of RCC cells, and could be a potential target to enhance the efficacy of radiotherapy in RCC treatment.

To further demonstrate the effect of PARPi as a radiosensitizer in RCC tumor, we used a mouse xenograft model carrying ACHN tumors. Radiation was administered as single treatment of 2 Gy at Day 0, and Olaparib was given orally daily for 2 weeks. Treatment with low dose of Olaparib (15 mg/kg, oral gavage daily for 2 weeks) transiently suppressed the tumor growth in the first week after starting treatment (P < 0.05 at Day 7), but tumors regrew shortly thereafter (P = 0.128 and 0.185 at Day 10 and Day 14, respectively; Fig. 6A). Combination therapy, however, shows a robust inhibition of tumor growth compared with tumors treated with IR alone or Olaparib alone (P < 0.05 at 1 week posttreatment) without body weight changes (Fig. 6A; Supplementary Fig. S3A). In comparison to control, the combination treatment showed a remarkable therapeutic efficacy (Fig. 6A; Supplementary Fig. S3A).

Having confirmed the better efficacy of combination therapy, the strategy was extended to the evaluation of PDX models (XP490 and XP334). We selected two PDX models with different levels of DAB2IP and PARP-1, XP490, which had higher DAB2IP levels but lower PARP-1 and XP334 (Fig. 6B). Notably, XP490 was very sensitive to IR (Fig. 6C). In contrast, XP334 showed more resistant to either IR or Olaparib alone, but Olaparib was able to sensitize XP334 cells to IR (Fig. 6D), when used in combination. From genetic profiling of these PDX models, it seems that the IR resistance is independent of the status of VHL, BAP1, and PBRM1 (Supplementary Fig. S3B). On the other hand, XP334 PDX model appeared to be more resistant to IR compare with XP374 which exhibited higher DAB2IP (Supplementary Fig. S3D). These data suggest that DAB2IP levels in company with PARP-1 could provide the prediction of the IR sensitivity. Taken together, we conclude that the downregulation of DAB2IP underlies IR resistance in RCC cells.

PARP-1 is an ADP-ribosylating enzyme, which is activated upon binding to DNA breakpoint induced by IR (7). The critical role of PARP-1 in DNA repair is reflected by its frequent upregulation in cancer cells such as melanomas and breast cancer. Therefore, elevated PARP-1 expression is often considered to be a prognostic feature associated with poor survival rate of cancer patients (10, 34, 35). However, in RCC, the regulation of PARP-1 and its role are not well characterized. In this study, we unveiled a unique mechanism of regulating PARP-1 protein stability by the DAB2IP–E3 ligase complex. Over the past decades, several studies have demonstrated that posttranslational modifications of PARP-1, including ADP-ribosylation, phosphorylation, and acetylation have differential impact on its downstream functions (7, 9, 36) but little is known about PARP-1 ubiquitylation. In an early study, Wang and colleagues suggested that K48 is the major site for PARP-1 polyubiquitination in mouse fibroblast cells leading to protein degradation, and assumed that the polyubiquitination site might be away from the catalytic domain or the automodification site of PARP-1, because NAD+ inhibited PARP-1 ubiquitination (37). Another study in heat shocked HeLa cells indicated that ubiquitinated PARP-1 was bound by E3-ligase RNF4, which led to the clearance and recycling of PARP-1 from the promoter region of target genes (38). In this study, we identify several specific E3-ligases such as RanBP2, TRIP12, and RNF40, responsible for PARP-1 polyubiquitination. More importantly, tumor suppressor DAB2IP is able to facilitate PARP-1 polyubiquitination, which subsequently leads to the reduction of PARP-1 protein levels. A significant clinical correlation of RanBP2 expression is found with overall survival of patients with RCC, supporting the notion that posttranslation modification of PARP-1 plays a critical role in RCC development. The elevated PARP-1 expression facilitates the rapid DSB repair that underlies the radioresistance of many malignant cells. Thus, PARPi have recently been evaluated as a sensitizer for radiotherapy in different cancer types; Olaparib (AZD2281) is currently being examined for breast or ovarian cancers with BRCA1/2 mutation (39). Initial trials have demonstrated significant efficacy for both cancer patients who harbor germline BRCA1/2 mutation; suggesting BRCA1/2 as a selective biomarker for this regimen (40, 41). In addition to breast and ovarian cancer, BRCA1/2 mutations have also increased life-time risk of pancreas, colon, and prostate cancers; Olaparib treatment showed better clinical benefit to those BRCA1/2 mutation-associated patients (33, 42–44). In contrast, BRCA1/2 mutation was not commonly associated with patients with RCC (45). However, increasing studies indicate a potential therapeutic role for PARPi in broader subgroup of solid tumors that have homologous recombination dysfunction (HRD; refs. 46, 47). Patients with HRD, but not BRCA1/2 mutation also showed “BRCA-like” behavior, such as susceptibility to DNA damage agents (11, 48). For example, loss-of-function of PTEN has been shown to yield BRCA-like behavior and increased PARPi susceptibility was shown in tumors harboring PTEN mutation or haploinsufficiency, suggesting PTEN-loss also could be a potential biomarker for predicting PARPi responsiveness (49, 50). Given that efficacy of PARPi in other solid tumors not harboring BRCA1/2 mutations, suggested that PARPi potentially have a broader application in treatment of cancer patients than was previously envisioned, possibly through different mechanisms of action. Indeed, our study demonstrated new evidence for PARPi susceptibility in RCC. The outcome of this study further suggests that either DAB2IP or RanBP2 should be further evaluated as selective biomarker(s) for targeted radiotherapy in patients with RCC. Our understanding of the molecular mechanism underlying radioresistant RCC unveils a potent target PARP-1 and the effect of Olaparib on enhancing the efficacy of RT from both in vitro and in vivo models provides a strong rationale to develop a clinical trial of combining PARPi with IR to overcome radioresistant RCC tumors that have lost DAB2IP expression.

No potential conflicts of interest were disclosed.

Conception and design: E.-J. Yun, D. He, B.P.C. Chen

Development of methodology: E.-J. Yun

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): E.-J. Yun, C.-J. Lin, E. Hernandez, J. Guo, W.-M. Chen, J. Allison, N. Kim, P. Kapur, K. Wu, C.-H. Lai, S.T. Baek, B.P.C. Chen

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): E.-J. Yun, J. Guo, N. Kim, P. Kapur, J. Brugarolas, K. Wu, D. Saha, S.T. Baek, B.P.C. Chen

Writing, review, and/or revision of the manuscript: E.-J. Yun, J. Brugarolas, K. Wu, D. Saha, S.T. Baek, B.P.C. Chen

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): A. Dang, P. Kapur, H. Lin, D. Saha, S.T. Baek

Study supervision: J.-T. Hsieh

We thank Dr. Samarpita Sengupta for editorial assistance. This study was supported in part by the National Research Foundation of Korea (NRF-2018R1D1A1B07040751 to E.-J. Yun, NRF-2018M3C7A1024152 to S.T. Baek), BK21 Plus funded by Ministry of Education of Korea (10Z20130012243 to E.-J. Yun), National Cancer Institute SPORE (P50CA196516 to J. Brugarolas), the Ministry of Science and Technology in Taiwan (MOST103-2911-I-005-507 to H. Lin) and Cancer Prevention Research Institute of Texas (RP160268 to B.P.C. Chen).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.