Open Full Page

Home , DDB1, ERCC2, ERCC4, ERCC5, ERCC6, RAD51

Published OnlineFirst January 29, 2015; DOI: 10.1158/0008-5472.CAN-14-2593

Cancer Priority Report Research

A Unique Subset of Epithelial Ovarian Cancers with Platinum Sensitivity and PARP Inhibitor Resistance Raphael Ceccaldi1, Kevin W. O'Connor1, Kent W. Mouw1,2, Adam Y. Li1, Ursula A. Matulonis3, Alan D. D'Andrea1, and Panagiotis A. Konstantinopoulos3

Abstract

Platinum and PARP inhibitor (PARPi) sensitivity commonly tions. Furthermore, patients with tumors with NER alterations coexist in epithelial ovarian cancer (EOC) due to the high had similar OS and PFS as BRCA1/2-mutated patients, suggest- prevalence of alterations in the homologous recombination ingthatNERpathwayinactivation in EOC conferred enhanced (HR) DNA repair pathway that confer sensitivity to both drugs. platinum sensitivity, similar to BRCA1/2-mutated tumors. In this report, we describe a unique subset of EOC with Moreover, two NER mutations (ERCC6-Q524 and ERCC4- alterations in another DNA repair pathway, the nucleotide A583T), identified in the two most platinum-sensitive tumors, excision repair (NER) pathway, which may exhibit a discor- were functionally associated with platinum sensitivity in vitro. dance in sensitivities to these drugs. Specifically, 8% of high- Importantly, neither NER alteration affected HR or conferred grade serous EOC from The Cancer Genome Atlas dataset sensitivity to PARPi or other double-strand break–inducing exhibited NER alterations, including nonsynonymous or splice agents. Overall, our findings reveal a new mechanism of plat- site mutations and homozygous deletions of NER genes. inum sensitivity in EOC that, unlike defective HR, may lead to a Tumors with NER alterations were associated with improved discordance in sensitivity to platinum and PARPi, with poten- overall survival (OS) and progression-free survival (PFS), com- tial implications for previously reported and ongoing PARPi pared with patients without NER alterations or BRCA1/2 muta- trials in this disease. Cancer Res; 75(4); 628–34. 2015 AACR.

Introduction inhibitors of PARP inhibitor (PARPi), a novel class of anticancer agents, which exhibit synthetic lethality in tumors with a defective Platinum analogues (cisplatin and carboplatin) are active HR pathway (4, 5). agents against epithelial ovarian cancer (EOC) and constitute the Although defective HR is an important mediator of platinum backbone of first-line chemotherapy used in this disease (1, 2). sensitivity in EOC, repair of platinum-induced DNA damage does The enhanced platinum sensitivity of EOC tumors is thought to be not involve only the HR pathway. The nucleotide excision repair related to an underlying defect in homologous repair (HR)– (NER) pathway is a highly conserved and remarkably versatile mediated DNA repair, particularly in those tumors with high- DNA repair pathway that functions to identify and repair bulky grade serous histology. Specifically, approximately 50% of the DNA cross-links generated by a variety of genotoxic agents includ- high-grade serous EOCs in The Cancer Genome Atlas (TCGA) ing platinum (6). Indeed, more than 90% of platinum–DNA dataset harbored genetic or epigenetic alterations involving the lesions are intrastrand cross-links, which are repaired by the NER HR pathway, with germline and somatic BRCA1/2 mutations pathway (6–8). Furthermore, biallelic germline mutations of NER occurring in 15% and 6% to 7%, respectively (3). Tumors with pathway genes lead to extreme platinum sensitivity observed in defective HR have also been shown to be exquisitely sensitive to patients with Xeroderma pigmentosum or Cockayne syndrome (9, 10). Defective HR contributes to sensitivity to both platinum and 1Department of Radiation Oncology, Dana-Farber Cancer Institute, PARPi, explaining why responsiveness to PARPi is closely asso- 2 Harvard Medical School, Boston, Massachusetts. Harvard Radiation ciated with platinum sensitivity in EOC (11). In fact, completed Oncology Program, Harvard Medical School, Boston, Massachusetts. 3Department of Medical Oncology, Medical Gynecologic Oncology and on-going clinical trials of PARPi routinely enroll patients with Program, Dana-Farber Cancer Institute, Harvard Medical School, platinum-sensitive disease in an attempt to enrich for HR-defec- Boston, Massachusetts. tive tumors that are likely to respond to PARPi (12). However, Note: Supplementary data for this article are available at Cancer Research platinum and PARPi sensitivity do not always coexist in EOC. Online (http://cancerres.aacrjournals.org/). Specifically, not all platinum-sensitive tumors respond to PARPi Corresponding Authors: Panagiotis A. Konstantinopoulos, Dana-Farber Cancer (11), and EOC tumors that become PARPi-resistant retain the Institute, YC-1424, 450 Brookline Avenue, Boston, MA 02215. Phone: 617-632- potential to respond to subsequent platinum-based chemother- 6232; Fax: 617-632-3479; E-mail: apy (13). Understanding the mechanisms of discordance between [email protected]; and Alan D. D'Andrea, E-mail: platinum and PARPi sensitivity may help optimize the adminis- [email protected] tration of these agents in the clinical management of EOC and doi: 10.1158/0008-5472.CAN-14-2593 may also aid in the development of novel therapies to overcome 2015 American Association for Cancer Research. resistance.

628 Cancer Res; 75(4) February 15, 2015

Discordance in Platinum and PARPi Response in Ovarian Cancer

In this study, we show that NER alterations are present in EOC after transfection, and protein expression was concurrently ana- and that these alterations are associated with a phenotype of lyzed by immunoblot. clinical platinum sensitivity that is similar to that of BRCA1/2- mutated tumors with improved overall survival (OS) and pro- HR efficiency gression-free survival (PFS). Moreover, we show that NER altera- HR efficiency was measured using the DR-GFP reporter assay. tions are associated with increased cellular platinum sensitivity Briefly, 48 hours before SceI transfection, U20S-DR-GFP cells were and that, unlike defective HR, lead to discordance between transfected with Control siRNA or siRNA targeting BRCA2, platinum and PARPi sensitivity, a finding that may have impor- ERCC4, or ERCC6. The HR activity was determined by FACS tant clinical and translational implications. quantification of viable GFP-positive cells 96 hours after SceI cDNA transfection. For RAD51 immunofluorescence, cells were fi Materials and Methods xed with 4% paraformaldehyde for 10 minutes at room tem- perature, followed by extraction with 0.3% Triton X-100 for 10 Evaluation of NER alterations in TCGA dataset minutes on ice. Incubation with the primary antibody (anti- We accessed data for 316 patients with high-grade serous EOCs RAD51; Santa Cruz Biotechnology) was performed at 37C. in the TCGA dataset for which DNA copy number, promoter methylation, and whole-exome sequencing information was Cell survival assays available (3, 14, 15). We evaluated for nonsynonymous or splice Cells were seeded in 96-well plates and treated with increasing site mutations, promoter hypermethylation, and homozygous concentrations of cisplatin (Sigma), Doxorubicin (Sigma), and deletions in the following NER genes: XPA, XPC, DDB1, ERCC4, Rucaparib (AG-014699; Selleckchem). After 72 hours, viability ERCC5, ERCC2, ERCC3, ERCC1, ERCC6, PCNA, ERCC8, LIG1, was assessed 72 hours by adding CellTiter-Glo (Promega) and RAD23B, MNAT1, MMS19, RFC1, and XAB2. For each NER measuring luminescence using a luminescence plate reader. Sur- mutation, we also evaluated the copy-number status to assess vival at each drug concentration was plotted as a percentage of the whether it was accompanied by heterozygous loss. Homozygous survival in drug-free media. deletion status was determined by GISTIC methodology (as applied in the ovarian TCGA dataset), and we focused only on Statistical analysis those with concurrent low mRNA expression levels of the corre- The t test and the Fisher exact test were used to analyze sponding NER gene (i.e., mRNA expression z score of less than the clinical and experimental data. Significance was defined as 3). Promoter hypermethylation was assessed using the same a P < 0.05; all reported P values are two sided. OS and PFS curves criteria described in the ovarian TCGA dataset publication (3). were generated by the Kaplan–Meier method, and statistical significance was assessed using the log-rank test. Tumors that harbored Cell culture both NER and BRCA1/2 mutations (n ¼ 4) were not included in the fi ERCC6 (GM16095)- and ERCC4 (GM08437)-de cient PFS or OS survival analysis. Inclusion of these tumors did not fi immortalized broblast cell lines (Coriell Cell Repository) were significantly change the results of the PFS and OS analyses cultured in DMEM (Invitrogen), supplemented with 10% FBS, 1% Penicillin Streptomycin and L-glutamine. ERCC6-deficient line was complemented with either wild-type or mutant N-terminal Results and Discussion Flag-tagged ERCC6, cloned in a pOZ vector. The ERCC4-deficient NER alterations are present in EOC and are associated with line was complemented with C-terminal Myc-tagged wild-type or clinical platinum sensitivity mutant ERCC4, cloned in a pLenti vector (Origene). We curated the EOC TCGA dataset to assess potential inactivat- For Western blotting, whole cell extracts were prepared by ing events of the NER pathway, including mutations, homozy- lysing cells in RIPA with complete protease inhibitor. Lysates gous deletions, and promoter hypermethylation of NER genes were resolved on a polyacrylamide gel, transferred to a PVDF (3, 14, 15). We found that a total of 24 (8%) of 316 EOCs membrane, and incubated with primary antibodies [BRCA2 harbored either NER mutations or homozygous deletions of NER (ab-1; Calbiochem), BRCA1 (ab-1; Calbiochem), ERCC4 genes. Specifically, we identified 19 cases with nonsynonymous or (D3G8C; Cell Signaling Technology), ERCC6 (D-7; Santa Cruz splice site NER gene mutations (all somatic) and 6 cases with Biotechnology), Flag (M2; Sigma), and b-actin (Cell Signaling homozygous deletions of NER genes among the 316 sequenced Technology)]. Signal was detected using an ECL kit (Pierce) EOCs of the TCGA dataset (Fig. 1A). None of the NER genes were and visualized with a Fuji LAS-3000 luminescent image ana- found to harbor promoter hypermethylation. All NER mutations lyzer system. were mutually exclusive, i.e., no individual tumor harbored mutations in more than one NER gene. Furthermore, NER muta- Transfection tions were mutually exclusive with homozygous deletions of the Targeted genes were depleted by transient transfection of siRNA NER genes with the exception of one case that harbored both an 0 directed against BRCA2 (5 -GAAGAAUGCAGGUUUAAUAdTdT- ERCC5 mutation and homozygous deletion of ERCC2. Of the 19 0 3 ), ERCC4 (ERCC4-5), or ERCC6 (ERCC6-2). All siRNA duplexes cases with NER mutations, 7 (36.8%) were accompanied by were from Qiagen. siRNA duplexes were transfected at a final heterozygous loss of the respective NER gene (Fig. 1A), indicating concentration of 100 nmol/L using the Lipofectamine Reagent that in these cases both wild-type alleles had been lost. 0 (Life Technologies) according to the manufacturer s recommen- Importantly, patients with tumors with NER alterations exhib- dations. In all experiments, a fraction of transfected cells was ited statistically significantly higher median OS (63.5 vs. 41.5 analyzed by immunoblot to assess knockdown efficiency. Trans- months, respectively; log-rank P ¼ 0.048) and a trend toward fection of plasmid cDNA was achieved using Lipofectamine LTX statistically significantly higher median PFS (30.4 vs. 14.7 (Life Technologies). Functional assays were performed 48 hours months, respectively; log-rank P ¼ 0.069) compared with patients

www.aacrjournals.org Cancer Res; 75(4) February 15, 2015 629

Ceccaldi et al.

A B Figure 1. NER alterations in EOC and association Case Amino acid Heterozygous Gene Type of alteration 1.0 NER vs. remaining: number change loss Median OS: with outcome. A, characteristics of 63.5 vs. 41.5 months, P = 0.048 1 ERCC5 Missense mutation G78V Yes NER pathway alterations in EOC 0.8 2 ERCC5 Missense mutation I186T No NER vs. BRCA: tumors of TCGA dataset. B, association 3 ERCC5 Missense mutation D943Y Yes Median OS: 63.5 vs. 59.1 months, P = 0.811 of tumors with NER alterations with ERCC5 Missense mutation S1078F Yes 4 0.6 OS. Tumors with NER alterations ERCC2 Homozygous deletion 5 DDB1 Missense mutation P721L Yes exhibited similar median OS (63.5 vs. 6 DDB1 Splice mutation Q759_splice No 0.4 59.1 months, respectively, P ¼ 0.811) 7 DDB1 Missense mutation E535Q No compared with BRCA1/2-mutated 8 ERCC6 Missense mutation R557G No tumors and statistically significantly 9 ERCC6 Nonsense mutation Q524* No Proportion suviving 0.2 - BRCA mutations 10 ERCC6 Splice mutation T141_splice No - NER alterations higher median OS (63.5 vs. 41.5 11 XPC Missense mutation D121E No months, respectively, P ¼ 0.048) 0.0 - Remaining 12 XPC Missense mutation K387T No compared with the remaining tumors. 13 XPC Missense mutation G757R Yes 0 24 48 72 98 120 144 168 C, association of tumors with NER 14 RFC1 Nonsense mutation Q937* No Months 15 RFC1 Missense mutation L209V No C alterations with PFS. Tumors with NER 16 MNAT1 Missense mutation L171V No 1.0 alterations exhibited similar median 17 RAD23B Missense mutation T195I Yes NER vs. Remaining: Median PFS: PFS (30.4 vs. 19.2 months, respectively, 18 ERCC2 Missense mutation A503G No 30.4 vs. 14.7 months, P = 0.069 P ¼ 0.971) compared with BRCA1/2- 19 ERCC4 Missense mutation A583T Yes 0.8 20 XPA Homozygous deletion NER vs. BRCA: mutated tumors and a trend toward Median PFS: 21 DDB1 Homozygous deletion statistically significantly higher median 0.6 30.4 vs. 19.2 months, P = 0.971 22 ERCC3 Homozygous deletion PFS (30.4 vs. 14.7 months, respectively, 23 RAD23B Homozygous deletion P ¼ 0.069) compared with the 24 RFC1 Homozygous deletion 0.4 remaining tumors. For both B and C, tumors that harbored both NER and 0.2 - BRCA mutations BRCA1/2 mutations (n ¼ 4) were not - NER alterations included in the PFS or OS analysis. Proportion progression free 0.0 - Remaining Inclusion of these tumors did not fi 96 84 72 60 48 36 24 12 0 12 24 36 48 60 72 84 96 signi cantly change the results of the Months PFS and OS analyses.

with tumors without NER alterations and BRCA1/2 mutations was complemented with either wild-type ERCC6 or the mutant (Fig. 1B and C). Furthermore, patients with tumors with NER ERCC6-Q524. Expression of wild-type ERCC6 rescued cisplatin alterations exhibited similar outcome (OS and PFS), with tumors sensitivity of ERCC6-deficient cells, whereas complementation harboring BRCA1 or BRCA2 mutations (Fig. 1B and C). with mutant ERCC6-Q524 did not impact cisplatin sensitivity It is widely accepted that the improved OS and PFS observed in (Fig. 2A). To confirm that ERCC6 loss alone is solely sufficient to EOCs with BRCA1/2 mutations are attributed to their enhanced induce cisplatin sensitivity, we assessed cisplatin cytotoxicity sensitivity to platinum chemotherapy due to defective HR-medi- following siRNA knockdown of ERCC6. ERCC6 depletion signif- ated DNA repair (16, 17). In this regard, the improved outcome of icantly increased platinum sensitivity, comparable with BRCA2 EOC tumors with NER alterations, which is similar to that of loss, a major mediator of DNA cross-link repair (Supplementary tumors that harbor BRCA1/2 mutations, supports the notion that Fig. S1A). NER alterations may confer a phenotype of enhanced platinum Given that ERCC6-Q524 was a somatic mutation and not sensitivity. associated with heterozygous loss, we evaluated whether this mutation may exert a dominant-negative effect. We postulated NER alterations are functionally associated with platinum that ERCC6-Q524 may interfere with the function of the wild- sensitivity in vitro type allele and hence increase sensitivity to cisplatin. Indeed, As a proof of principle that NER alterations are functionally introduction of the ERCC6-Q524 variant in ERCC6 wild-type associated with platinum sensitivity, we evaluated two NER 293T cells dramatically increased cisplatin sensitivity compared mutations (ERCC6-Q524 and ERCC4-A583T) that were identi- with cells transfected either with wild-type ERCC6 or control fied in 2 patients (cases 9 and 19, respectively, Fig. 1A) with empty vector, suggesting that this mutation sensitizes cells to advanced-stage high-grade serous EOC that demonstrated the cisplatin by a dominant-negative mechanism (Fig. 2B). This best response to first-line platinum-based chemotherapy. The finding suggests that even in the absence of heterozygous loss, ERCC6-Q524 nonsense mutation was present in a patient with NER mutations may still be functionally relevant in terms of stage IV suboptimally debulked high-grade serous tumor that had platinum sensitivity in EOC. Heterozygous NER mutations asso- complete response to first-line platinum chemotherapy and ciated with platinum sensitivity have been reported in other remained in complete remission for 31.5 months after diagnosis. tumor types (other than EOC) and more recently in bladder The ERCC4-A583T missense mutation was present in a patient cancer (18). How mutation of one NER allele can exert a dom- with stage IIIC optimally debulked high-grade serous tumor who inant-negative phenotype is unclear and several mechanisms are is alive and disease-free 25 months after diagnosis. possible. First, because NER proteins are assembled into large, To determine the functional significance of ERCC6-Q524 on multisubunit complexes, the presence of a mutant protein may platinum sensitivity, we evaluated whether this variant could have a deleterious effect on DNA repair, even when the wild-type rescue platinum sensitivity in an ERCC6-deficient cell line. An protein is present. Alternatively, the mutated protein may bind, ERCC6-deficient immortalized fibroblast cell line (GM16095) but not adequately repair, damaged DNA, thereby preventing

630 Cancer Res; 75(4) February 15, 2015 Cancer Research

Discordance in Platinum and PARPi Response in Ovarian Cancer

A ERCC6−/− (fibroblasts): 125 EV ERCC6 100 ERCC6-Q524* CC6 CC6

125 ERCC6−/− : R R EV E E 75 Q524* 100 (nmol/L) ERCC6 50 50 75 * 50 * ERCC6-Q524*

Percent survival 25 25 β-Actin

0 CDDP IC 0 −/− CDDP (μmol/L): 0 0.05 0.1 0.25 0.5 1 ERCC6 : EV ERCC6 ERCC6 Q524*

B 293T (ERCC6+/+): 125 EV IP Flag ERCC6 6 6 * al 100 ERCC6-Q524* iv CC CC 24

v 2.5 Q5 ER 75 Flag: EV ER 2 μ mol/L) ERCC6 50 ( 1.5 50 1 25 * * Percent sur ERCC6

0.5 Flag M2 Q524* 0

CDDP IC CDDP 0 CDDP (μmol/L): 0 0.05 0.1 0.5 1 2.5 5 10 293T: EV ERCC6 ERCC6 Q524*

C ERCC4−/− (fibroblasts): EV 125 ERCC4 ERCC4-A583T 100 10

4 4 75 8 T CC CC

−/− 83

( μ mol/L) 6 ERCC4 : EV ER ER A5

50 50 4 ERCC4 25 * Percent survival 2 * β-Actin

0 CDDP IC 0 CDDP (μmol/L): 0 0.1 0.5 1 2.5 10 ERCC4−/− : EV ERCC4 ERCC4 A583T

Figure 2. ERCC6-Q524 and ERCC4-A583T sensitize to cisplatin (CDDP). A, ERCC6-deficient fibroblasts [black solid line, ERCC6 / þ empty vector (EV)] are hypersensitive to cisplatin. Introduction of wild-type ERCC6 (black dashed line, ERCC6/ þ WT ERCC6) rescues the sensitivity of the deficient cells, whereas the mutant / protein (red solid line, ERCC6 þ ERCC6-Q524 ) does not rescue survival (left plot). IC50 values and immunoblot showing expression of wild-type (blot against ERCC6) or mutant ERCC6 in ERCC6-deficient fibroblasts (anti-Flag antibody) are shown on the right. B, transient overexpression of the mutant protein ERCC6-Q524 (red solid line) in 293T cells confers sensitivity to cisplatin, compared with cells transfected with either empty vector (black solid line) or wild-type ERCC6 (black dashed line). IC50 values are shown on the right. An immunoblot showing expression of wild-type or mutant ERCC6 in 293T (anti-Flag antibody) is shown on the right. C, ERCC4-deficient fibroblasts (black solid line, ERCC4 / þ EV) are hypersensitive to cisplatin. Introduction of wild-type ERCC4 (black dashed line, ERCC4/ þ WT ERCC4) rescues the sensitivity of the deficient cells, whereas mutant protein (green solid line, ERCC4/ þ ERCC4-A583T) does not significantly rescue survival (left plot). IC50 values and immunoblot showing expression of wild-type or mutant ERCC4 in ERCC4-deficient fibroblasts are shown on the right. Experiments were performed in triplicate and error bars represent 1 SEM. Statistical analyses were performed using one-way ANOVA. Asterisks indicate statistically significant values (for A, P < 104 EV vs. ERCC6 and P < 103 ERCC6 vs. ERCC6-Q524; EV vs. ERCC6-Q524; for B, P < 104 EV vs. ERCC6-Q524 and ERCC6 vs. ERCC6-Q524; EV vs. ERCC6; for C, P < 10 4 EV vs. ERCC4 and P < 10 3 ERCC4 vs. ERCC4-A583T; EV vs. ERCC4-A583T).

repair by an alternative DNA-repair pathway, thus explaining a complete loss of functional ERCC4 in this patient. Furthermore, dominant-negative phenotype as previously described for ERCC2 ERCC4 knockdown alone was able to confer platinum sensitivity in yeast (19). in 293T cells to levels comparable with BRCA2 depletion (Sup- Similarly, we evaluated whether the ERCC4-A583T mutant plementary Fig. S1B). could rescue the platinum sensitivity of ERCC4-deﬁcient Our experiments indicate that the ERCC4-A583T and ERCC6- (GM08437) cells. Reintroduction of wild-type ERCC4, but not Q524 mutations confer platinum sensitivity in vitro and may be the mutant protein, was able to rescue cisplatin sensitivity of responsible for the extreme platinum sensitivity observed in these ERCC4-deﬁcient cells (Fig. 2C). Importantly, ERCC4-A583T was a 2 patients. It is important to underscore that, in both tumors, we somatic mutation associated with heterozygous loss arguing for a could not detect any concurrent HR pathway alterations that

www.aacrjournals.org Cancer Res; 75(4) February 15, 2015 631

Ceccaldi et al.

would otherwise explain the extreme platinum sensitivity of these ERCC4 or ERRC6 did not affect HR efficiency in vitro, as measured patients. Specifically, there were no BRCA1/2 mutations, epige- by direct-repeat GFP recombination (DR-GFP) assay and by netic silencing of BRCA1, mutations in Fanconi Anemia genes, IR-induced RAD51 foci formation, a surrogate for HR efficiency mutations in core HR RAD genes, mutations in DNA damage (Fig. 3B and C). Similarly, reintroduction of wild-type or mutant response HR genes, amplification of EMSY, and homozygous ERCC6 or ERCC4 did not affect sensitivity of ERCC6- and ERCC4- deletion of PTEN detected in these tumors. Taken together, our deficient fibroblasts to the topoisomerase-II inhibitor doxorubi- findings provide the proof of principle that platinum sensitivity cin or the topoisomerase-I inhibitor camptopthecin (Fig. 3D and may occur in EOC as a result of NER alterations. Supplementary Fig. S2), which induce DNA double-strand breaks that are repaired by HR. Together, these results indicate that ERCC6-Q524 and ERCC4-A583T mutations do not affect HR functional loss of ERCC6 or ERCC4 does not impair HR efficiency nor sensitivity to PARPi nor alters sensitivity to PARPi or other double-strand break– We evaluated the association of these NER alterations with inducing agents such as camptothecin and doxorubicin. sensitivity to the PARPi rucaparib. Unlike in the case of cisplatin, Our findings suggest a novel mechanism of discordance expression of wild-type or mutant ERCC6 or ERCC4 did not affect between platinum and PARPi sensitivity in EOC, which may have PARPi sensitivity of ERCC6- and ERCC4-deficient fibroblasts (Fig. important clinical and translational implications. Platinum sen- 3A). Furthermore, because defective HR is a critical mediator of sitivity has been traditionally used as an eligibility criterion for platinum and PARPi sensitivity in EOC, we evaluated whether selection of participation in PARPi clinical trials in an attempt to deficiency in ERCC6 or ERCC4 affected HR in vitro. Inhibition of enrich for HR-defective tumors that may respond to PARPi (12).

A ERCC6−/− (fibroblasts): 125 EV 125 ERCC6 ERCC4−/− (fibroblasts): 100 ERCC6-Q524* 100 EV ERCC4 75 75 ERCC4-A583T 50 50 Percent survival Percent Percent survival Percent 25 25

0 0 PARPi (μmol/L): 0 1 10 20 50 100 PARPi (μmol/L): 0 1 10 20 50 100

BC si Ctrl 2.5 100

2 75 1.5 50 1 si ERCC6 si ERCC4 RAD51 foci HR efficiency 25 0.5 Percent cells with > 4 Percent

0 0 si BRCA2 siRNA: Ctrl BRCA2 ERCC6 ERCC4 U2OS + siRNA: Ctrl ERCC4 ERCC6 BRCA2 D 125 125 ERCC6−/− (fibroblasts): −/− 100 EV 100 ERCC4 (fibroblasts): ERCC6 EV 75 ERCC6-Q524* 75 ERCC4 ERCC4-A583T 50 50 Percent survival Percent

25 survival Percent 25

0 0 Doxorubicin (μmol/L): 0 0.05 0.1 0.5 1 Doxorubicin (μmol/L): 00.050.1 0.5 1

Figure 3. Inactivation of ERCC6 or ERCC4 does not impair HR efficiency and does not confer sensitivity to rucaparib and doxorubicin. A, addition of wild-type (WT) or mutant ERCC6 in ERCC6-deficient fibroblasts does not affect PARPi (rucaparib) sensitivity (left). Addition of wild-type ERCC4 to ERCC4-deficient fibroblasts does not affect PARPi sensitivity (right). B, siRNA-mediated depletion of ERCC6 or ERCC4 does not significantly affect HR as measured by a DR-GFP recombination assay in U2OS cells. BRCA2 depletion is shown as a positive control. Error bars, 1 SEM. Results are representative of three independent experiments. C, siRNA knockdown of ERCC4 or ERCC6 does not affect ionizing radiation-induced RAD51 foci formation in U2OS cells. Cells with 4 RAD51 foci were scored as positive. BRCA2-depleted cells are shown as a positive control. Positive cells were quantified 6 hours after irradiation (5 Gy; left plot). For each condition, 100 cells were counted. Error bars, 1 SEM. Representative micrographs of irradiated cells are shown (right). D, addition of wild-type ERCC6 (black dashes) or mutant protein (red solid line, ERCC6-Q524) in ERCC6-deficient fibroblasts does not impact sensitivity to doxorubicin (left). Addition of wild-type ERCC4 (black dashes) or mutant ERCC4 (red solid line, ERCC4-A583T) in ERCC4-deficient fibroblasts does not impact sensitivity to doxorubicin (right). Error bars, 1 SEM. Results are representative of three independent experiments.

632 Cancer Res; 75(4) February 15, 2015 Cancer Research

Discordance in Platinum and PARPi Response in Ovarian Cancer

A I. NER alterations Platinum sensitivity + BRCA/HR proficiency No PARPi sensitivity

II. PARPi Tx NER alterations Platinum sensitivity Retention of + PARPi resistance Platinum sensitivity PARPi sensitivity NER alterations BRCA/HR deficiency Reversion of HR/BRCA defects

B MDA-MB-436 C MDA-MB-436 (BRCA1−/− fibroblasts): (BRCA1−/− fibroblasts): EV + si ERCC6 MDA-MB-436 EV + si ERCC6 −/− 125 BRCA1 + si Ctrl (BRCA1 ): EV BRCA1 125 BRCA1 + si Ctrl 100 BRCA1 + si ERCC6 100 BRCA1 + si ERCC6 * mol/L) * μ

* 1

1,000 − 300 75 75 + siRNA: si Ctrl si ERCC6 si si ERCC6 si ERCC6 (10 (nmol/L) 750 200 50 50 50 50 ERCC6 500 Percent survival

Percent survival 25 BRCA1 25 100 250 VINCULIN

0 0 PARPi IC CDDP IC CDDP 0 0 μ μ −/− PARPi ( mol/L): 0 1 10 50 100 −/− CDDP ( mol/L): 0 0.05 0.25 0.5 1 MDA-MB-436 : EV BRCA1 MDA-MB-436 : EV BRCA1

+ siRNA: + siRNA: si ERCC6 si ERCC6 si ERCC6 ERCC6 si si Ctrl si Ctrl si Ctrl si ERCC6 si ERCC6 si ERCC6

Figure 4. Discordance between platinum and PARPi due to NER alterations: potential implications. A, potential implications of NER alterations for platinum and PARPi sensitivity in patients with BRCA/HR-proficient (I) and BRCA/HR-deficient (II) tumors. I, if tumors with NER alterations are HR proficient, i.e., they do not harbor concurrent HR alterations that would confer sensitivity to PARPis, they are expected to be platinum sensitive but not PARPi sensitive. II, if tumors with NER alterations are BRCA/HR deficient they will initially be sensitive to both platinum and PARPis. However, if these tumors develop resistance to PARPi via reversion of the HR defect (e.g., via secondary BRCA1/2 mutations), they are predicted to retain some sensitivity to subsequent platinum therapy due to the underlying NER alterations. B, expression of BRCA1 (BRCA1 þ si Ctrl, red solid line) rescues the cisplatin hypersensitivity of BRCA1–/– MDA-MB-436 cells [empty vector (EV) þ si ERCC6, black dashed line]. However, BRCA1-restored cells with ERCC6-depletion (BRCA1 þ si ERCC6, red dashed line) partially retain sensitivity to cisplatin compared with BRCA1-restored line without ERCC6-depletion (BRCA1 þ si Ctrl, red solid line; left plot). Sensitivity is measured by 3-day viability. Error bars, 1 SEM. IC50 values are shown as a measure of the relative cisplatin sensitivity of MDA-MB-436 cell lines (right plot). An immunoblot showing knockdown efficiency of ERCC6 and reexpression of BRCA1 in MDA-MB-436 cells is shown on the right. C, add-back of BRCA1 cDNA (BRCA1 þ si Ctrl, red solid line) restores PARPi sensitivity of BRCA1 / MDA-MB-436 cells (EV þ si ERCC6, black dashed line). Concurrent depletion of ERCC6 (BRCA1 þ si ERCC6) in this cell line does not impact PARPi sensitivity (left plot). Sensitivity is measured by 3-day viability. Error bars, 1 SEM. IC50 values are shown as a measure of the PARPi sensitivity of MDA-MB-436 cell lines (right).

However, our findings suggest that for tumors with NER altera- Furthermore, NER alterations may also, at least partly, explain tions, platinum sensitivity may not always be an accurate predic- another phenomenon, commonly encountered in clinic, whereby tor of PARPi sensitivity. Specifically, if tumors with NER altera- certain patients with BRCA1/2-mutated tumors that become tions are HR proficient, i.e., they do not harbor concurrent HR PARPi-resistant retain the potential to respond to subsequent alterations that would confer sensitivity to PARPis, platinum platinum-based chemotherapy, as has been previously reported sensitivity may not necessarily translate into PARPi sensitivity (13). It is well known that the most common mechanism of (Fig. 4A). Therefore, the presence of NER alterations may explain PARPi resistance in BRCA/HR-deficient tumors is reversion of the why certain patients with extreme platinum sensitivity who are HR defect (i.e., secondary BRCA1 or BRCA2 mutations restoring enrolled in PARPi trials fail to respond to these agents (11, 12). normal BRCA1/2 function; refs. 21–23). In this regard, if BRCA/ Specifically, in the phase II multicentre, open-label, nonrando- HR-deficient tumors also harbor NER alterations, then reversion mized study of olaparib in women with advanced high-grade of the HR defect conferring PARPi resistance (21, 22) will not serous and/or undifferentiated ovarian carcinoma (20), the objec- confer cross-resistance to platinum; therefore, these patients may tive response rate to the PARPi olaparib in patients with platinum- still benefit from subsequent platinum-based therapy (Fig. 4A). sensitive ovarian cancer was 50% in BRCA1/2-negative and Supporting this, restoration of BRCA1 function in BRCA1- and 60% in BRCA1/2-mutated tumors. Furthermore, in the phase I ERCC6-depleted cells completely rescues PARPi sensitivity but dose-escalation and single-stage expansion study of olaparib only partially rescues cisplatin sensitivity (Fig. 4B and C). In this in BRCA1/2-mutated ovarian cancer (11), the clinical benefit regard, in a retrospective review of patients with BRCA1/2-mutat- response rate (radiologic or tumor marker complete or partial ed ovarian cancer who received chemotherapy following disease response or radiologic disease stabilization for 4 months or more) progression on olaparib (i.e., after development of resistance to was 69% in patients with platinum-sensitive disease. Overall, as PARPis), response to platinum-based chemotherapy was 40%, evident by these studies, 30% to 50% of tumors with platinum- thereby suggesting a discordance between platinum and PARPi sensitive ovarian cancer were resistant to PARPi therapy. sensitivity of 40% in this setting (13). Known mechanisms of

www.aacrjournals.org Cancer Res; 75(4) February 15, 2015 633

Ceccaldi et al.

resistance to PARPis in patients with BRCA1/2-mutated EOCs or Authors' Contributions EOCs with BRCAness phenotype (apart from secondary BRCA1/2 Conception and design: R. Ceccaldi, K.W. Mouw, A.D. D'Andrea, mutations) include loss of PARP1 expression, upregulation of P- P.A. Konstantinopoulos glycoprotein efflux Pump transporter, and loss of 53BP1 in Development of methodology: R. Ceccaldi, K.W. Mouw, A.D. D'Andrea, P.A. Konstantinopoulos BRCA1 mutant cells. Of these mechanisms, loss of 53BP1 has Acquisition of data (provided animals, acquired and managed patients, been experimentally proven to confer platinum sensitivity in provided facilities, etc.): R. Ceccaldi, K.W. O'Connor, A.Y. Li, U.A. Matulonis, BRCA1-mutated tumors (24), but its exact clinical relevance in P.A. Konstantinopoulos patients with ovarian cancer is unclear. Analysis and interpretation of data (e.g., statistical analysis, biostatistics, In conclusion, we report for the first time that NER pathway computational analysis): R. Ceccaldi, K.W. Mouw, U.A. Matulonis, alterations (mutations and homozygous deletions) occur in EOC P.A. Konstantinopoulos Writing, review, and/or revision of the manuscript: R. Ceccaldi, K.W. O'Connor, and that these alterations are associated with a phenotype of K.W. Mouw, U.A. Matulonis, A.D. D'Andrea, P.A. Konstantinopoulos clinical platinum sensitivity that is similar to that of BRCA1/2- Administrative, technical, or material support (i.e., reporting or organizing mutated tumors characterized by improved OS and PFS. We data, constructing databases): R. Ceccaldi, P.A. Konstantinopoulos showed that the NER mutations identified in the most plati- Study supervision: A.D. D'Andrea, P.A. Konstantinopoulos num-sensitive tumors (ERCC6-Q524 and ERCC4-A583T) were functionally associated with platinum sensitivity in vitro. Impor- Acknowledgments tantly, these NER alterations did not affect HR and did not confer The authors thank Timur Yusufzai for providing the pOZ-ERCC6 construct sensitivity to PARPi, thus providing a novel mechanism of dis- and Shawn Johnson and Geoffrey Shapiro for providing the MDA-MB-436 ( BRCA1 cDNA) cell line. cordance between platinum and PARPi sensitivity in EOC. Our fi ndings suggest that NER alterations may have a previously Grant Support unrecognized role as biomarkers for selection of patients for R. Ceccaldi is a recipient of the Ovarian Cancer Research Fellowship. P.A. participation in PARPi trials as well as for deciding therapy after Konstantinopoulos is recipient of DOD Ovarian Cancer Research Program development of PARPi resistance. Ovarian Cancer Academy Award W81XWH010-1-0585.

Disclosure of Potential Conﬂicts of Interest Received September 3, 2014; revised October 29, 2014; accepted November No potential conﬂicts of interest were disclosed. 18, 2014; published OnlineFirst January 29, 2015.

References 1. Bast RC Jr, Hennessy B, Mills GB. The biology of ovarian cancer: new 14. Cerami E, Gao J, Dogrusoz U, Gross BE, Sumer SO, Aksoy BA, et al. opportunities for translation. Nature Rev Cancer 2009;9:415–28. The cBio cancer genomics portal: an open platform for exploring multi- 2. Konstantinopoulos PA, Awtrey CS. Management of ovarian cancer: a 75- dimensional cancer genomics data. Cancer Discov 2012;2:401–4. year-old woman who has completed treatment. JAMA 2012;307:1420–9. 15. Gao J, Aksoy BA, Dogrusoz U, Dresdner G, Gross B, Sumer SO, et al. 3. TCGA. Integrated genomic analyses of ovarian carcinoma. Nature 2011; Integrative analysis of complex cancer genomics and clinical profiles 474:609–15. using the cBioPortal. Sci Signal 2013;6:pl1. 4. Bryant HE, Schultz N, Thomas HD, Parker KM, Flower D, Lopez E, et al. 16. Bolton KL, Chenevix-Trench G, Goh C, Sadetzki S, Ramus SJ, Karlan BY, Specific killing of BRCA2-deficient tumours with inhibitors of poly(ADP- et al. Association between BRCA1 and BRCA2 mutations and survival in ribose) polymerase. Nature 2005;434:913–7. women with invasive epithelial ovarian cancer. JAMA 2012;307:382–90. 5. Farmer H, McCabe N, Lord CJ, Tutt AN, Johnson DA, Richardson TB, et al. 17. Yang D, Khan S, Sun Y, Hess K, Shmulevich I, Sood AK, et al. Association Targeting the DNA repair defect in BRCA mutant cells as a therapeutic of BRCA1 and BRCA2 mutations with survival, chemotherapy sensitivity, strategy. Nature 2005;434:917–21. and gene mutator phenotype in patients with ovarian cancer. JAMA 6. Kennedy RD, D0Andrea AD. DNA repair pathways in clinical practice: 2011;306:1557–65. lessons from pediatric cancer susceptibility syndromes. J Clin Oncol 18. Van Allen EM, Mouw KW, Kim P, Iyer G, Wagle N, Al-Ahmadie H, et al. 2006;24:3799–808. Somatic ERCC2 mutations correlate with cisplatin sensitivity in muscle- 7. D0Andrea AD, Grompe M. The Fanconi anaemia/BRCA pathway. Nat Rev invasive urothelial carcinoma. Cancer Discov 2014;4:1140–53. Cancer 2003;3:23–34. 19. Naumovski L, Friedberg EC. Analysis of the essential and excision repair 8. Kelland L. The resurgence of platinum-based cancer chemotherapy. Nat Rev functions of the RAD3 gene of Saccharomyces cerevisiae by mutagenesis. Cancer 2007;7:573–84. Mol Cell Biol 1986;6:1218–27. 9. Lehmann AR. DNA repair-deficient diseases, xeroderma pigmentosum, 20. Gelmon KA, Tischkowitz M, Mackay H, Swenerton K, Robidoux A, Tonkin K, Cockayne syndrome and trichothiodystrophy. Biochimie 2003;85:1101–11. et al. Olaparib in patients with recurrent high-grade serous or poorly differ- 10. Furuta T, Ueda T, Aune G, Sarasin A, Kraemer KH, Pommier Y. Transcrip- entiated ovarian carcinoma or triple-negative breast cancer: a phase 2, multi- tion-coupled nucleotide excision repair as a determinant of cisplatin centre, open-label, non-randomised study. Lancet Oncol 2011;12:852–61. sensitivity of human cells. Cancer Res 2002;62:4899–902. 21. Norquist B, Wurz KA, Pennil CC, Garcia R, Gross J, Sakai W, et al. Secondary 11. Fong PC, Yap TA, Boss DS, Carden CP, Mergui-Roelvink M, Gourley C, et al. somatic mutations restoring BRCA1/2 predict chemotherapy resistance in Poly(ADP)-ribose polymerase inhibition: frequent durable responses in hereditary ovarian carcinomas. J Clin Oncol 2011;29:3008–15. BRCA carrier ovarian cancer correlating with platinum-free interval. J Clin 22. Swisher EM, Sakai W, Karlan BY, Wurz K, Urban N, Taniguchi T. Secondary Oncol 2010;28:2512–9. BRCA1 mutations in BRCA1-mutated ovarian carcinomas with platinum 12. Ledermann J, Harter P, Gourley C, Friedlander M, Vergote I, Rustin G, et al. resistance. Cancer Res 2008;68:2581–6. Olaparib maintenance therapy in platinum-sensitive relapsed ovarian 23. Sakai W, Swisher EM, Karlan BY, Agarwal MK, Higgins J, Friedman C, et al. cancer. N Engl J Med 2012;366:1382–92. Secondary mutations as a mechanism of cisplatin resistance in BRCA2- 13. Ang JE, Gourley C, Powell CB, High H, Shapira-Frommer R, Castonguay V, mutated cancers. Nature 2008;451:1116–20. et al. Efficacy of chemotherapy in BRCA1/2 mutation carrier ovarian cancer 24. Bunting SF, Callen E, Kozak ML, Kim JM, Wong N, Lopez-Contreras AJ, in the setting of PARP inhibitor resistance: a multi-institutional study. Clin et al. BRCA1 functions independently of homologous recombination in Cancer Res 2013;19:5485–93. DNA interstrand crosslink repair. Mol Cell 2012;46:125–35.

634 Cancer Res; 75(4) February 15, 2015 Cancer Research

A Unique Subset of Epithelial Ovarian Cancers with Platinum Sensitivity and PARP Inhibitor Resistance

Raphael Ceccaldi, Kevin W. O'Connor, Kent W. Mouw, et al.

Cancer Res 2015;75:628-634. Published OnlineFirst January 29, 2015.

Updated version Access the most recent version of this article at: doi:10.1158/0008-5472.CAN-14-2593

Supplementary Access the most recent supplemental material at: Material http://cancerres.aacrjournals.org/content/suppl/2021/03/31/0008-5472.CAN-14-2593.DC1

Cited articles This article cites 24 articles, 10 of which you can access for free at: http://cancerres.aacrjournals.org/content/75/4/628.full#ref-list-1

Citing articles This article has been cited by 10 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/75/4/628.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at Subscriptions [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/75/4/628. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.