Learn more: PMC Disclaimer | PMC Copyright Notice
Platinum and PARP inhibitor resistance due to over-expression of microRNA-622 in BRCA1-mutant ovarian cancer
Associated Data
Abstract
High-grade serous ovarian carcinomas (HGSOCs) with BRCA1/2 mutations exhibit improved outcome and sensitivity to double-strand DNA break (DSB)-inducing agents [i.e. platinum and Poly(ADP-ribose) polymerase inhibitors (PARPis)] due to an underlying defect in homologous recombination (HR). However, resistance to platinum and PARPis represents a significant barrier to the long-term survival of these patients. Although, BRCA1/2-reversion mutations are a clinically validated resistance mechanism, they account for less than half of platinum resistant BRCA1/2-mutated HGSOCs. We uncover a resistance mechanism by which a microRNA, miR-622 induces resistance to PARPis and platinum in BRCA1-mutant HGSOCs by targeting the Ku complex and restoring HR-mediated DSB repair., Physiologically, miR-622 inversely correlates with Ku expression during the cell cycle, suppressing non-homologous end joining and facilitating HR-mediated DSB repair in S-phase. Importantly, high expression of miR-622 in BRCA1-deficient HGSOCs is associated with worse outcome after platinum chemotherapy, indicating microRNA-mediated resistance through HR rescue.
INTRODUCTION
Approximately 15-20% of patients with epithelial ovarian cancer (EOC) harbor germline (10-15%) or somatic (6-7%) BRCA1 or BRCA2 mutations(TCGA, 2011). Furthermore, epigenetic silencing of BRCA1 via promoter hypermethylation occurs in approximately 10-20% of EOCs. Due to the underlying defect in DNA repair via homologous recombination (HR), patients with BRCA1/2-inactivated EOCs exhibit enhanced sensitivity to platinum analogues and other cytotoxic drugs that induce double strand DNA breaks (DSBs) such as the poly-ADP ribose polymerase inhibitors (PARPis)(Fong et al., 2009). Of these drugs, olaparib was granted accelerated approval by the U.S. FDA for use in EOC patients with germline BRCA1/2 mutations (Fong et al., 2009). However, a substantial fraction of these patients do not respond or eventually develop resistance to these agents suggesting that de novo and acquired platinum and PARPi resistance is a significant clinical problem in HR-defective EOCs. The most common mechanism of resistance to these agents in BRCA1/2-mutated tumors is secondary intragenic mutations restoring BRCA1 or BRCA2 protein functionality; 46% of platinum resistant BRCA-mutated EOCs exhibit tumor-specific secondary mutations that restore the ORF of either BRCA1 or BRCA2(Norquist et al., 2011).
The interplay of the two major mechanistically distinct DSB repair pathways, HR and non-homologous end joining (NHEJ) (Chapman et al., 2012b; Ciccia and Elledge, 2010) is also critical for resistance to platinum and PARPis. Surprisingly, the sensitivity of BRCA1-mutant tumors to PARP inhibitors is almost completely abolished by loss of the NHEJ factor 53BP1 (Bouwman et al., 2010; Bunting et al., 2010; Chapman et al., 2012a), which also correlates with the restoration of competent HR. Furthermore a recent small hairpin (sh) RNA screen for hairpins promoting survival of BRCA1-deficient mouse mammary tumors to PARPi identified 53BP1 and REV7, a factor implicated in NHEJ, as the top hits (Boersma et al., 2015; Xu et al., 2015). However, unlike BRCA1/2 reversion mutations, these resistance mechanisms have not been shown to be clinically relevant for patients with BRCA1/2-inactivated EOCs. However, it is feasible that the NHEJ pathway may be relevant for PARPi resistance in EOCs, and other NHEJ factors may contribute to the resistant phenotype.
Here, we uncover mechanism of resistance to PARPi and platinum in BRCA1-mutated EOCs that involves miRNA-mediated regulation of NHEJ. Specifically, we have identified a miRNA, miR-622 that regulates the expression of the Ku-complex and specifically suppresses NHEJ during S-phase. Consistent with this effect, overexpression of miR-622 rescues the HR-deficiency of BRCA1-mutant ovarian tumor lines and induces resistance to PARPi and platinum-based drugs. Furthermore, expression of miR-622 in two cohorts of patients with BRCA1-inactivated EOCs correlates with reduced disease-free survival after platinum-based therapy, suggesting direct clinical relevance in patients with EOC
RESULTS
miR-622 ‘desensitizes’ BRCA1 mutant cells to PARP inhibitors/platinum-based therapy
Recently, we used PARPi sensitivity as a marker for HR-deficiency to conduct a functional screen for identifying miRNAs that down-regulate HR in a breast cancer line, MDA-MB231(Choi et al., 2014). We characterized the miRNAs (miR-1255b, miR-193b* and miR-148b*) that suppress HR by down-regulating the expression of BRCA1, BRCA2 and RAD51. Strikingly, in that screen, six miRNAs (miR-644, miR-492, miR-613, miR-577, miR-622 and miR-126*)(Choi et al., 2014) demonstrated a surprising trend of inducing PARPi resistance. Our original screen was conducted in a BRCA-proficient breast tumor line MDA-MB231 and we assessed the impact of these miRNAs on PARPi sensitivity in MDA-MB231. Considering the BRCA mutant cells are responsiveness to PARPi, and therefore we also examined the impact of these miRNAs in a BRCA1-mutant breast line, MDA-MB436. There was no significant impact of miR-644, miR-492, miR-613, miR-577 and miR-126* on PARPi sensitivity in MDA-MB231 and MDA-MB436 cells (Supp Fig. 1A), however miR-622 significantly induced resistance to the clinical grade PARP inhibitors, olaparib and veliparib, specifically in the MDA-MB436 cells (Supp Fig. 1B). Furthermore, we tested the impact of miR-622 on PARPi sensitivity on the BRCA1-mutant EOC line, UWB1.289 and found that overexpression of miR-622 caused resistance to both PARPis, olaparib and veliparib (ABT-888) (Fig. 1A). Interestingly, miR-622 expression also caused resistance to the platinum-based chemotherapeutic agents, carboplatin and cisplatin in the BRCA1-mutated UWB1.289 cells (Fig. 1A). Importantly, restoring BRCA1 expression in UWB1.289 cells completely negates the impact of miR-622 on PARPi sensitivity and also sensitivity to platinum drugs (Supp Fig. 1C). In order to exclude the possibility that the Brca1-mutant lines MDA-MB436 and UWB1.289 have acquired other unaccounted mutations which may contribute to the phenotype induced by miR-622, we expressed miR-622 in BRCA1-null mouse embryonic fibroblasts (MEF) and assessed sensitivity to olaparib and cisplatin. Consistent with our previous results, miR-622 significantly ‘desensitized’BRCA1−/−MEFs to both drugs (Fig. 1B) but did not impact the sensitivity of their wild type counterparts (Supp Fig. 1D). Together, these data suggest that the impact of miR-622 on PARPi and platinum-based therapy is specific to the loss of BRCA1.
(A, B) Viability assays to examine the impact of miR-622 on drug sensitivity. BRCA1-null UWB1.289 cells (A) or BRCA1-deficient MEF cells (B) were transfected with control mimic or miR-622 mimic and treated with vehicle or indicated drug before measurement of viability by luminescence-based ATP quantification. Curves were generated from 3 independent experiments. (C) Association between miR-622 expression levels and DFS and OS in tumors with BRCA1 mutation and BRCA1 promoter hypermethylation in the TCGA dataset based on 50% cut-off. Tumors with BRCA1 mutations and BRCA1 promoter hypermethylation with above median expression levels of miR-622 were associated with worse DFS (left panel, log rank p = 0.03) and OS (right panel, log rank p = 0.03). (D) Association between miR-622 expression levels and DFS and OS in tumors with BRCA1 mutation and BRCA1 promoter hypermethylation in the TCGA dataset based on 20% cut-off. Tumors with BRCA1 mutations and BRCA1 promoter hypermethylation whose expression levels for miR-622 were in the highest quintile were associated with worse DFS (left panel, log rank p = 0.005) and OS (right panel, log rank p = 0.001). (E) DFS and OS in the 10 tumors with the highest mir-622 expression versus the 10 tumors with the lowest miR-622 expression in the TCGA dataset (tumors with BRCA1 mutation and BRCA1 promoter hypermethylation). The 10 tumors with the lowest mir-622 expression were associated with worse DFS (left panel, log rank p = 0.001) and OS (right panel, log rank p = 0.03) compared to the 10 tumors with the highest mir-622 expression.
Expression of miR-622 correlates with response to platinum chemotherapy in BRCA1-inactivated EOCs
To evaluate the association between miR-622 expression and platinum response in EOCs with BRCA1 inactivation, we assessed data from the ovarian TCGA dataset(TCGA, 2011). In that dataset, 89 EOCs (all HGSOCs) exhibited BRCA1-inactivation; 38 EOCs harbored BRCA1-mutations (out of 316 EOCs that underwent whole exome sequencing) while 51 tumors (out of 489 tumors with DNA promoter methylation data) harbored BRCA1 epigenetic silencing via promoter hypermethylation. All patients underwent surgery followed by platinum based chemotherapy. We evaluated the association between miR-622 expression and platinum response using various cut-offs for low versus high miR-622 expression. In all cases, we consistently found that tumors with higher miR-622 expression were associated with inferior response to first line platinum based chemotherapy and worse survival. Specifically, using median miR-622 expression as a threshold to classify BRCA1-inactivated EOCs as exhibiting high versus low miR-622 expression, we found that BRCA1-inactivated tumors with high expression of miR-622 were associated with worse disease-free survival (DFS) (median DFS 14.7 vs 19.8 months respectively, log rank p = 0.03) and overall survival (OS) (median OS 39 vs 49.3 months respectively, log rank p = 0.03) compared with tumors with low miR-622 expression (Fig. 1C). Conversely, there was no association between miR-622 expression and outcome, DFS or OS in the remaining tumors in TCGA dataset, i.e. those without BRCA1 mutations and without BRCA1 promoter hypermethylation (data not shown). This trend was particularly evident in tumors with the highest miR-622 expression, i.e. those whose mir-622 expression was in the highest quintile. Specifically, BRCA1-inactivated tumors whose expression levels for miR-622 were in the highest quintile were associated with worse DFS (median DFS 13.7 vs 18.1 months respectively, log rank p = 0.005) and OS (median OS 35.3 vs 48.3 months respectively, log rank p = 0.001, Fig. 1D).
Furthermore, we compared tumors with the highest miR-622 expression versus those with the lowest miR-622 expression. Specifically, when comparing the top 5, 10 or 15 tumors with the highest miR-622 expression with the lowest 5, 10 or 15 tumors respectively, we consistently found that the tumors with the highest miR-622 expression were associated with inferior response to first line platinum chemotherapy, i.e. worse DFS and OS compared to the tumors with the lowest expression (Fig. 1E and Supp Fig. 1E).
Given the absence of other miRNA expression datasets with sizeable numbers of ovarian tumors with BRCA1-mutations or BRCA1 promoter hypermethylation, we explored the correlation between miR-622 and outcome in tumors with low BRCA1 expression in a different, clinically annotated ovarian cancer dataset (Shih et al., 2011). This dataset included miRNA and mRNA expression data from 60 patients with newly diagnosed FIGO stage III or IV tumors with serous histology, including 3 tumors with BRCA1 mutations. As shown in Supplement Figure 1F, we found similar correlation between high miR-622 expression and inferior outcome to first line platinum based chemotherapy.
miR-622 impacts NHEJ mediated repair of DSBs
The NHEJ pathway is composed of at least two branches: the well-studied classical NHEJ (C-NHEJ) and the poorly understood alternative end-joining (A-EJ)(Deriano and Roth, 2013). The molecular details and biological function of A-NHEJ remains largely unclear(Deriano and Roth, 2013). Loss or depletion of factors promoting C-NHEJ (such as 53BP1) or essential for C-NHEJ (such as Ku70) induces PARPi resistance in BRCA1-deficient mouse cells (Bunting et al., 2012; Bunting et al., 2010). To test whether miR-622 indeed impacts NHEJ, we assayed for C-NHEJ and A-NHEJ mediated repair of the yeast endonuclease, I-SceI-induced DSBs using the EJ5-GFP reporter and EJ2-GFP reporter, respectively. These are integrated fluorescence-based reporters (Bennardo et al., 2008) that allow for efficient quantification of the two distinct NHEJ pathways at targeted DSBs. We observed that miR-622 significantly impedes C-NHEJ (Fig. 2A), and enhances A-NHEJ (Fig. 2B). This is consistent with studies showing that depletion of C-NHEJ factors increases the frequency of A-NHEJ (Fattah et al., 2010). Depletion of 53BP1 and Ku70 induces PARPi resistance in BRCA1-mutant cells by restoring HR-mediated repair of DSBs and significantly enhancing genomic stability after PARPi treatment (Bunting et al., 2012; Bunting et al., 2010). Consistent with its impact on NHEJ, we observe that expression of miR-622 in BRCA1−/−MEFs causes a significant decrease in the level of genomic instability (chromosomal aberrations) induced by olaparib treatment (Fig. 2C). To address the mechanism by which miR-622 promotes genome integrity in BRCA1 mutant cells, we tested whether its expression could cause an increase in irradiation-induced Rad51 foci, a measure of the HR-pathway. We found that expression of miR-622 in UWB1.289 cells caused a statistically significant increase in Rad51 foci (Fig. 2D). Importantly, none of these effects are due to alterations in the cell cycle caused by the miR-622 mimics (Supp Fig. 2A).
(A, B) Measurement of C-NHEJ (A) or A-NHEJ (B) mediated repair of I-SceI induced site specific DSBs. Cells carrying a single copy of the recombination substrate with two tandem I-SceI sites were transfected with control mimic, miR-622 mimic, Ku70 siRNA or Ligase4 siRNA before transfection with I-SceI or control vector. In 48 hrs, GFP positive cells were analyzed by flow cytometry. (C) Analysis of genomic instability in metaphases. BRCA1−/− MEF cells were transfected with control miRNA mimic or miR-622, treated with 100nM PARP inhibitor, and measured for abnormal chromosomes in metaphase. (n≥50 metaphases). (D) Analysis of HR-mediated repair by RAD51 focus formation. UWB1.289 cells were transfected with control miRNA mimic or miR-622, stained for RAD51 (green), γH2AX (red) and 4′,6-diamidino-2-phenylindole (DAPI) (blue) 6 hrs after exposure to 10Gy IR. The images were captured by fluorescence microscopy and RAD51 focus-positive cells (with > 20 foci) were quantified by comparing 100 cells
miR-622 regulates expression of the Ku complex
To investigate the mechanism by which miR-622 influences NHEJ and impacts PARP inhibitor sensitivity we used a candidate-based approach whereby all genes implicated in NHEJ were screened for miRNA recognition elements (MREs) of miR-622 using the PITA algorithm. This algorithm is unique in allowing G:U wobbles or seed mismatches, and identifies base pairing beyond the 5'end of the miRNA, predicts the sites not restricted to the 3'UTR of mRNA and identifies non-canonical MREs for specific miRNA/mRNA combinations(Lal et al., 2009). Using this algorithm, miR-622 was predicted to target the transcripts of 53BP1, Ku70, Ku80, APTX and APLF (Supp Fig. 3). We assessed the impact of over-expressing miR-622 in UWB1.289 cells on the mRNA level of these genes and observed a significant reduction in the transcripts of 53BP1, Ku70 and Ku80 (Fig. 3A). Subsequently, we determined the impact of these miRNAs on the protein level of their putative targets. Over-expressing miR-622 reduces the protein levels of Ku70 and Ku80 in UWB1.289 cells. The basal expression of the Ku proteins is lower in MEFs, and the impact of miR-622 on Ku70 and Ku80 in BRCA1−/−MEFs is even more pronounced (Fig. 3B). On the contrary, there was no detectable impact of miR-622 on 53BP1 in the UWB1.289 cells. To test for association of miR-622 with the Ku70 and Ku80 transcripts we captured miRNA-mRNA complexes using streptavidin-coated beads from cells transfected with biotinylated forms of the miRNA mimics (Lal et al., 2011; Orom and Lund, 2007). The amount of Ku70, Ku80 and 53BP1 transcripts was measured in the pull-downs, and the enrichment was assessed relative to pull-down with biotinylated control mimic and also with GAPDH. Consistent with our previous results, miR-622 selectively pulled-down Ku70 and Ku80 transcripts but not the 53BP1 transcript (Fig. 3C). To verify further that Ku70 and Ku80 are targets of miR-622 and confirm that the interaction is mediated by the predicted MREs we used luciferase reporter assays. The predicted MREs (Fig. 3D) were cloned in the 3'UTR of the luciferase gene, and expression monitored in cells transfected with the miR-622 mimic (Fig. 3E). As anticipated, there was significant decrease in luciferase activity, and this was ‘rescued’ by point mutations that disrupt base pairing between miR-622 and their corresponding MREs in Ku70 and Ku80 (Fig. 3F). Together these results suggest that miR-622 regulates the expression of the Ku complex by direct interaction with Ku70 and Ku80 transcripts.
(A-B) Expression of DDR genes is impacted by miR-622. UWB1.289 cells were transfected with control mimic or miR-622 mimic and mRNA levels of predicted DDR genes were analyzed by qRT-PCR using gene-specific primers and normalized to GAPDH (A). Cell lysates were then analyzed by immunoblot for factors which had statistically significant reduction in mRNA in cells transfected with miR-622 (B). Images were quantified by ImageJ software and the mean ± SD of 3 independent experiments is graphically shown. (C) Interaction of target transcripts with miR-622. UWB1.289 cells were transfected with biotinylated-control mimic or biotinylated miR-622 mimic. The immunoprecipitated RNA was analyzed by qRT-PCR using gene-specific primers and normalized to GAPDH. (D) Predicted MREs were obtained from PITA algorithm (http://genie.weizmann.ac.il/pubs/mir07/mir07_prediction.html) and their mutants were generated by mutating nucleotides providing complementarity to corresponding miRNAs. CDS (coding sequence) means the region in the gene where MRE is located. (I) Luciferase reporter assay to assess direct interaction of miR-622 with target genes. Individual or combinations of predicted miRNA recognition sites (MREs) for each putative target transcript of miR-622 were cloned into the luciferase reporter vector and transfected in UWB1.289 cells along with miR mimics. Renilla luciferase activity of the reporter was measured 48 h after transfection by normalization to an internal firefly luciferase control. (J) Luciferase reporter assay for wild-type or mutant MREs for miRNA-622 targets was performed in the same way as described in Figure 2I. (A-H) Mean ± SD of 3 independent experiments is shown and statistical significance is indicated by * (p<0.05).
miR-622 causes resistance to PARP inhibitor and cisplatin by down-regulating expression of the Ku proteins
We examined the impact of Ku downregulation (using siRNAs) or inhibition (dominant negative Ku(He et al., 2007)) on olaparib and cisplatin sensitivity in parallel with miR-622 over-expression in UWB1.289 cells (Fig. 4A) and in BRCA1−/−MEFs (Fig. 4B). We observe that depletion/inhibition (efficacy of siRNAs shown in Supp. Fig. 4) of the Ku complex and over-expression of miR-622 have a comparable effect on de-sensitizing BRCA1-deficient cells to both olaparib and cisplatin. To determine whether the effect of miR-622 on olaparib and cisplatin sensitivity was indeed mediated via Ku suppression we utilized mouse Ku70 cDNA and rat Ku80 cDNA that lack miR-622 MREs. Next, UWB1.289 cells were co-transfected with miR-622 and mouse Ku70 cDNA or rat Ku80 cDNA. The Ku expression constructs lacking the miR-622 MREs ‘rescued’ the expression of these genes in the presence of miR-622 mimic further validating the predicted MREs (Fig. 4C, right panel). Furthermore, individual expression of the Ku proteins partially ‘rescued’ the impact of miR-622 on olaparib and cisplatin sensitivity (Fig. 4C, left panel).
(A, B) Viability assays to examine the impact of miR-622 on targets. Control mimic, miR-622 mimic, Ku70 siRNA, Ku80 siRNA or dominant negative Ku70 were introduced to UWB1.289 cells (A) or BRCA1-deficient MEF cells (B). Transfected cells were treated with vehicle or indicated drug before viability measurement as explained in Figure 1. (C) Impact of miR target rescue. UWB1.289 cells were transfected with control mimic or miR-622 mimic with or without rat Ku70 cDNA or mouse Ku80 cDNA and treated with vehicle or indicated drug before viability measurement as explained in Figure 1. Expression of introduced genes was examined by immunoblot. (D) Correlation between miR-622 expression levels and Ku80 RNA expression levels and Ku80 protein levels in the TCGA dataset. miR-622 expression levels were statistically significantly inversely correlated with Ku80 RNA expression levels (p = 0.019) and Ku80 protein levels (p=0.029). (E) Correlation between miR-622 expression levels and Ku80 RNA expression levels in a different ovarian cancer miRNA dataset. miR-622 expression levels were statistically significantly inversely correlated with Ku80 RNA expression levels (p = 0.05) in a different ovarian cancer dataset.
Ku80 protein and mRNA expression levels are available in primary EOCs in the ovarian TCGA, and were correlated with miR-622 expression. Consistent with our results, there is statistically significant inverse correlation of miR-622 with both Ku80 protein and mRNA expression in BRCA-inactivated EOCs from the TCGA dataset. Specifically, among the 89 EOCs with either BRCA1 mutations (n=38) or BRCA1 promoter hypermethylation (n=51), miR-622 expression levels were statistically significantly inversely correlated with Ku80 RNA expression levels (p = 0.019) and Ku80 protein levels (p=0.029) as determined by reverse phase protein array (RPPA) in the TCGA dataset (Fig. 4D). This correlation was further confirmed in the independent cohort of EOC patients discussed above; specifically miR-622 expression levels were statistically significantly inversely correlated with Ku80 RNA expression levels (p = 0.05) (Fig. 4E). There was no Ku80 protein expression data in that dataset.
Physiological Relevance of miR-622 mediated suppression of the Ku complex
To explore the physiological relevance of the interactions of miR-622 with Ku70 and Ku80 transcripts we assessed their expression during cell cycle, specifically during the G1 to S transition. Synchronizing, UWB1.289 cells (profiles shown in Supp Fig. 5A) we observe that mRNA levels of Ku70 and Ku80 are reduced in the S-phase relative to the G1 phase (Fig. 5A). Interestingly, miR-622 inversely correlates with Ku70 and Ku80 transcripts, and is significantly up-regulated as cells move into the S-phase. Antagonizing miR-622 induces a specific increase in Ku70 and Ku80 transcripts (Fig. 5B) in the S-phase. To further confirm the cell cycle phase specificity of this phenotype avoiding the artifacts of synchronization, and in a diploid cell line with relatively few genomic abnormalities, we utilized the Fucci system(Sakaue-Sawano et al., 2008) to visualize the G1 phase (mKO2-CDT1-RFP) and S-phase (Geminin-GFP) in hTERT-immortalized retinal pigment epithelial cell line (RPE-1) cells. The G1 cells and S/G2 phases were separated and isolated using fluorescence-activated cell sorting (FACS) selection. Consistent with the previous results miR-622 expression inversely correlated with the Ku70 and Ku80 transcripts (Fig. 5C) and inhibition of miR-622 in RPE-1 caused a significant increase in Ku70 and Ku80 transcripts in the S-phase (Fig. 5D). To further elucidate the cell-cycle based impact of miR-622 on the Ku proteins we utilized the luciferase assays (as in Fig. 3). We confirmed that antagonizing endogenous miR-622 in S-phase significantly increases luciferase activity of constructs with miR-622 recognition elements in the Ku70 and Ku80 transcripts, and this was negated by point mutations that disrupt base pairing between miR-622 and their corresponding binding sites in these transcripts (Supp Fig.5B).
(A-D) Expression of miR and target transcripts in synchronized cells. (A) UWB1.289 cells were synchronized with mimosine and the relative amount of miR-622 or target mRNA for G1- or S-phase was determined by qRT-PCR (normalized to RNU1). (B) UWB1.289 cells were transfected with control ANT or miR-622 ANT and subsequently synchronized with mimosine. Expression of target mRNA was assessed by qRT-PCR in the G1 and S-phase (normalized to GAPDH). (C) RPE1 Fucci cells were sorted according to cell cycle-based fluorophore expression and the relative amount of miR-622 or target mRNA for G1- or S-phase was quantified by qRT-PC. (D) RPE1 Fucci cells were transfected with control ANT or miR-622 ANT and sorted for cell cycle. Expression of target mRNA was assessed by qRT-PCR in the G1 and S-phase. (A-D) Mean ± SD of 3 independent experiments is shown and statistical significance is indicated by * (p<0.05). (E-G) Impact of miR-622 inhibition on recruitment of DSB proteins. RPE1 Fucci cells were transfected with control ANT or miR-622 ANT and irradiated with 5 Gy (for γH2AX and Mre11, 3 hours after IR) or 10Gy (for RPA2, 4 hours after IR) IR. Cells were stained for Mre11 (red) (E), RPA2 (red) (F) or γH2AX (red) (G) and 4′,6-diamidino-2-phenylindole (blue). The images were captured by fluorescence microscopy and Mre11, RPA2 or γH2AX focus-positive cells (with > 20 foci or >50 foci) at S phase (green) were quantified by comparing 100 cells.
Recruitment of the MRN (Mre11-Rad50-Nbs1) complex is the first step in HR. From the functional standpoint there is a competitive interplay between the Ku complex and MRN complex(Balestrini et al., 2013; Foster et al., 2011). Specifically the over-expression of Ku proteins reduces recruitment of Mre11 to DSBs in the S/G2 phase when HR is the preferred DSB repair pathway(Clerici et al., 2008). Therefore we examined the Mre11 foci in the S-phase of irradiated cells transfected with miR-622 antagomirs. Consistent with increased Ku levels antagonizing miR-622 causes a significant decrease in Mre11 foci (Fig. 5E). Furthermore the subsequent step in HR, which is resection of broken DNA ends and RPA2 foci formation is also reduced by antagonizing miR-622 (Fig. 5F). Importantly, antagonizing miR-622 does not impact the IR induced generation of DSBs (monitored by γ–H2AX, Fig. 5E and 5G). Together, these results strongly suggest that miR-622 plays a role in the optimal expression of the Ku complex during the cell cycle, and potentially facilitates the initiation of HR-mediated DSB repair in the S phase.
DISCUSSION
There is tight regulation of the DSB repair pathways during the cell cycle as HR is restricted to the S/G2 phase and NHEJ is pre-dominant in G1 but has moderate activity throughout the cell cycle. Importantly, the choice of DSB repair pathways during cell cycle is critical for maintaining genomic stability. A decisive factor in this choice is competition between DNA end protection (necessary for NHEJ) and DNA end resection (necessary for HR). Depletion of end protecting factors (such as 53BP1) allows DNA end resection in the G1 phase, thereby impairing NHEJ and causing genomic instability (Helmink et al., 2011),(Escribano-Diaz et al., 2013). Conversely ectopic expression of BRCA1 in the G1 phase via the inhibition/deletion of miRNAs suppressing BRCA1 also allows DSB end resection leading to unrepaired DSBs (Choi et al., 2014; Dimitrov et al., 2013). During the S/G2 phase of the cell cycle the relatively error-free HR pathway is preferred, and NHEJ needs to be restricted. The mechanism via which the NHEJ pathway is restricted in the S-phase remains unknown. Here, we uncover regulation of this step by miR-622. We find that miR-622 plays an important role in maintaining the balance between HR and NHEJ repair pathways during the cell cycle by regulating optimal expression of the Ku complex. The Ku complex is pivotal in pathway choice as it competes with the MRN complex to capture broken DSB ends, and divert it towards the C-NHEJ pathway. MiR-622 suppresses NHEJ through targeting of the Ku-complex during S phase, and enhances initiation of HR-mediated DSB repair in the S phase by facilitating the recruitment of Mre11. Therefore ectopic over-expression of miR-622 can limit NHEJ, and boost the HR pathway.
Another important finding of our study is that this role for miR-622 in maintaining balance between DSB repair pathways may mediate resistance to PARPis and platinum agents in BRCA1-inactivated tumors. Elucidating mechanisms of platinum and PARPi resistance in BRCA-deficient EOCs is critical in order to identify approaches that suppress denovo and emerging resistant clones. Pharmacological effects that alter the cellular response to PARPis including increased expression of ABC transporters, such as the P-glycoprotein (PgP) efflux pump, have been associated with PARPi resistance in BRCA1-mutated breast and ovarian cancer, but their clinical relevance for platinum resistance remains unclear. Furthermore, although a number of resistance mechanisms have been described (Konstantinopoulos et al., 2015), only secondary BRCA1/2 mutations restoring BRCA1/2 protein functionality have been validated in multiple EOC patient cohorts. It is noteworthy that most of these models systems have not investigated ovarian carcinomas thereby undermining their clinical relevance. In this regard, our study highlights a mechanism of PARPi resistance in BRCA1-deficient EOC patients involving miR-622 overexpression, and represents an extension of its physiological role in maintaining the balance of DSB repair pathways.
Importantly, unlike 53BP1 loss which confers only PARPi resistance, this resistance mechanism confers resistance to both platinum and PARPis. Although miRNA expression has been recently implicated in mediating HR deficiency and response to platinum and PARPis (Liu et al., 2015), here we implicate a miRNA in exactly the opposite, i.e. mediating PARPi and platinum resistance by rescuing HR deficiency. Strikingly, the clinical relevance of this resistance mechanism was evident in two different ovarian cancer datasets whereby overexpression of miR-622 was associated with inferior outcome after platinum chemotherapy in BRCA1-inactivated tumors. Of note, the expression of miR-622 was also inversely correlated with protein and mRNA expression levels of Ku80 thereby clinically validating our experimental observations that the association of miR-622 with worse outcome may indeed be related to its targeting of the Ku complex. In conclusion, our work suggests a role for miR-622 in regulating the balance between HR and NHEJ in cell cycle and highlights a potential role of this miRNA as a biomarker of responsiveness to platinum and PARPis in BRCA1-inactivated EOCs. Furthermore, miR-622 may be a promising target for augmenting PARPi and platinum response in BRCA1-inactivated EOCs.
MATERIALS AND METHODS
Viability Assay
Viability assays were done as previously described (Choi et al., 2014).
Ovarian Cancer datasets and statistical analysis
Association of miR-622 expression levels with outcome (OS and DFS) was assessed in two clinically annotated ovarian cancer datasets with miRNA expression data. First, we accessed expression data from the ovarian TCGA dataset which included 38 tumors with BRCA1-mutations (out of 316 EOCs that underwent whole exome sequencing) and 51 tumors (out of 489 tumors with DNA promoter methylation data) with BRCA1 epigenetic silencing via promoter hypermethylation. Promoter hypermethylation was assessed using the same criteria described in the ovarian TCGA dataset publication. The second dataset included expression data from 60 patients with newly diagnosed FIGO stage III or IV tumors, all with serous histology (Shih KK et al. Gynecol Oncol 2011). The t test and the Fisher exact test were used to analyze the clinical and experimental data. Correlation between miR-622 and Ku80 expression levels was assessed using the Pearson's correlation coefficient. Significance was defined as a p<0.05; all reported p values are two sided. OS and DFS curves were generated by the Kaplan-Meier method, and statistical significance was assessed using the log-rank test.
Non-homologous End Joining Reporter Assay
NHEJ reporter assays were performed as the HR assays done previously (Choi et al., 2014) by using U2OS cells carrying a single copy of the recombination substrate with two tandem I-SceI sites.
Chromosome Breakage Analysis
Brca1−/− MEF cells were transfected with indicated miRNA mimics for 24 hours followed by treatment with or without the indicated concentrations of PARP inhibitor (Olaparib) for 24, 48 or 72 hours. Cells were exposed to 100 ng/ml colcemid for 2 hours followed by treatment with a hypotonic solution (0.075M KCl) for 20 minutes and fixed with 3:1 methanol/acetic acid solution. Slides were stained with Wright's stain and ≥50 metaphase spreads were scored for aberrations.
Immunofluorescence
Immunofluorescence in UWB1.289 and RPE1 Fucci cells were done as previously described (Lee et al., 2010) using RAD51 (Santa Cruz #sc-8349), γ-H2AX (Cell Signaling #9718S), RPA2 (Abcam #ab2175) and Mre11 (Novus Biologicals #NB100-142)
RNA Isolation and Quantitative Real-Time PCR
Total RNA was prepared and expression was analyzed by qRT-PCR as described previously (Moskwa et al., 2011).
Gene-specific primers used for qRT-PCR are as follows:
53BP1-F-1, GTCATTGAGCAGTTACCTCAG, R-1, GGGAATGTGTAGTATTGCCTG; 53BP1-F-2, ATGGTGGAGACCCATGATCC, R-2, GTCTTCTGGGGACTGGCAAC; KU70-F-1, GTTGATGCCTCCAAGGCTATG, R-2, GCACCTGGATTATCCAGCTC; KU70-F-2, AATTCAGGTGACTCCTCCAG, R-2, TGAAGTGCTGCTGCAGCAC; KU80-F-1, AAGCAAAATCCAACCAGGTTCT, R-1, GAATTGCAGGGAGATGTCACA; KU80-F-2, ACTCTGATCACCAAAGAGGAA, R-2, TGGCAGCTCTCTTAGATTCC; APTX-F, TGGAAGCAGTTGTGATTGGG, R, CACCATGTGGAGAACCTGG; APLF-F, GAAGCCAAATCTATGGTGCTA, R, CTTCATCAAGCACTTGACTGT
Immunoblots
The immunoblots were done as described previously (Lee et al., 2010; Moskwa et al., 2011) with 53BP1 (Cell Signaling Technology #4937), Ku70 (Santa Cruz #sc-1486), Ku80 (Thermo Scientific #PA5-17454) and α-tubulin (Sigma #T5168) antibodies.
Immunoprecipitation of miRNA Targets
Immunoprecipitation of miRNA target with biotinylated miR-622 was done with UWB1.289 cells as previously described (Choi et al. 2014).
Luciferase Assay
The wild type (WT) or mutant (Mt) miRNA recognition elements (MREs) of target genes were synthesized as oligonucleotide sequences, annealed and cloned in psiCHECK2 (Promega) downstream to Renilla luciferase. Luciferase assay in UWB1.289 cells using WT and Mt MRE constructs was done as described previously (Moskwa et al., 2011). The oligonucleotide sequences are as follows:
KU70-MRE1-F, TCGAAAGCAATGAATAAAAGACTGGGAAGAAGCAATGAATAAAAGACTGG, R, GGCCCCAGTCTTTTATTCATTGCTTCTTCCCAGTCTTTTATTCATTGCTT; KU70-MRE2-F, TCGAACCAAGCACTTCCAGGACTGAGAAGACCAAGCACTTCCAGGACTGA, R, GGCCTCAGTCCTGGAAGTGCTTGGTCTTCTCAGTCCTGGAAGTGCTTGGT; KU70-MRE1+2-F, TCGAAAGCAATGAATAAAAGACTGGGAAGACCAAGCACTTCCAGGACTGA, R, GGCCTCAGTCCTGGAAGTGCTTGGTCTTCCCAGTCTTTTATTCATTGCTT; KU80-MRE1-F, TCGAAGCTAAAAAATTAAAGACTGAGAAGAGCTAAAAAATTAAAGACTGA, R, GGCCTCAGTCTTTAATTTTTTAGCTCTTCTCAGTCTTTAATTTTTTAGCT; KU80-MRE2-F, TCGATTTATGAAGAGCATAGACTGCGAAGTTTATGAAGAGCATAGACTGC, R, GGCCGCAGTCTATGCTCTTCATAAACTTCGCAGTCTATGCTCTTCATAAA; KU80-MRE1+2-F, R, GGCCGCAGTCTATGCTCTTCATAAACTTCTCAGTCTTTAATTTTTTAGCT.
The oligonucleotides for mutant MREs are as follows:
Mt KU70-MRE1+2-F, TCGAAAGGTTGGAATAAATCTGACGGAAGAGGTAGCTGGAGCATCTGACA, R, GGCCTGTCAGATGCTCCAGCTACCTCTTCCGTCAGATTTATTCCAACCTT; Mt KU80-MRE1+2-F, TCGAACGAAATTAAAGTATCTGACAGAAGTTTATGAAGTCGATTCTGACC, R, GGCCGGTCAGAATCGACTTCATAAACTTCTGTCAGATACTTTAATTTCGT.
Cell Cycle Synchronization and Sorting
Cell synchronization was performed in UWB1.289 cells as previously described in Choi et al. (Choi et al., 2014). Cells transfected with miR-622 antagomir with rat Ku70 or moue Ku80 cDNA (gift from Andre Nussenzweig at National Cancer Institute) were similarly synchronized 48 hrs after transfection. RPE1 Fucci cells were sorted by using BD FACSAria based on fluorophore expression according to cell cycle (RFP-G1 phase, GFP-S/G2/M phase).
miRNA Target Prediction
We used a candidate-based prediction approach using PITA (http://genie.weizmann.ac.il/pubs/mir07/mir07_data.html), to analyze the Human DNA Repair Gene list (http://sciencepark.mdanderson.org/labs/wood/dna_repair_genes.html#Human%20DNA%20Repair%20Genes) which resulted in a list of DDR genes predicted as targets of miRNAs of our interest. Predicted targets are listed in Supplementary Figure 2 and further validated as explained in the manuscript.
ACKNOWLEDGEMENTS
DC is supported by R01 AI101897-01 (NIAID) and R01CA142698-07 (NCI), Basic Scholar Grant (American Cancer Society), Leukemia and Lymphoma Society Scholar Grant, Claudia Adams Barr Program for Innovative Cancer Research, Breast SPORE Pilot Award, First Fund Award and Mary Kay Foundation. PAK is supported by the Susan Smith Center for Women's Cancers, and the Department of Defense Ovarian Cancer Academy Award (W81XWH-10-1-0585). The BRCA1−/− MEFs were a gift from Andre Nussenzweig and NHEJ reporter construct was a gift from Jeremy Stark.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
The authors have no financial conflicts
REFERENCES
- Balestrini A, Ristic D, Dionne I, Liu XZ, Wyman C, Wellinger RJ, Petrini JH. The Ku heterodimer and the metabolism of single-ended DNA double-strand breaks. Cell reports. 2013;3:2033–2045. [PMC free article] [PubMed] [Google Scholar]
- Bennardo N, Cheng A, Huang N, Stark JM. Alternative-NHEJ is a mechanistically distinct pathway of mammalian chromosome break repair. PLoS Genet. 2008;4:e1000110. [PMC free article] [PubMed] [Google Scholar]
- Boersma V, Moatti N, Segura-Bayona S, Peuscher MH, van der Torre J, Wevers BA, Orthwein A, Durocher D, Jacobs JJ. MAD2L2 controls DNA repair at telomeres and DNA breaks by inhibiting 5′ end resection. Nature. 2015 [PMC free article] [PubMed] [Google Scholar]
- Bouwman P, Aly A, Escandell JM, Pieterse M, Bartkova J, van der Gulden H, Hiddingh S, Thanasoula M, Kulkarni A, Yang Q, et al. 53BP1 loss rescues BRCA1 deficiency and is associated with triple-negative and BRCA-mutated breast cancers. Nature structural & molecular biology. 2010;17:688–695. [PMC free article] [PubMed] [Google Scholar]
- Bunting SF, Callen E, Kozak ML, Kim JM, Wong N, Lopez-Contreras AJ, Ludwig T, Baer R, Faryabi RB, Malhowski A, et al. BRCA1 functions independently of homologous recombination in DNA interstrand crosslink repair. Molecular cell. 2012;46:125–135. [PMC free article] [PubMed] [Google Scholar]
- Bunting SF, Callen E, Wong N, Chen HT, Polato F, Gunn A, Bothmer A, Feldhahn N, Fernandez-Capetillo O, Cao L, et al. 53BP1 inhibits homologous recombination in Brca1-deficient cells by blocking resection of DNA breaks. Cell. 2010;141:243–254. [PMC free article] [PubMed] [Google Scholar]
- Chapman JR, Sossick AJ, Boulton SJ, Jackson SP. BRCA1-associated exclusion of 53BP1 from DNA damage sites underlies temporal control of DNA repair. Journal of cell science. 2012a;125:3529–3534. [PMC free article] [PubMed] [Google Scholar]
- Chapman JR, Tayloy GRM, Boulton SJ. Playing the End Game: DNA Double-Strand Break Repair Pathway Choice. Molecular cell. 2012b;47:495–510. [PubMed] [Google Scholar]
- Choi YE, Pan Y, Park E, Konstantinopoulos P, De S, D'Andrea A, Chowdhury D. MicroRNAs down-regulate homologous recombination in the G1 phase of cycling cells to maintain genomic stability. eLife. 2014;3:e02445. [PMC free article] [PubMed] [Google Scholar]
- Ciccia A, Elledge SJ. The DNA damage response: making it safe to play with knives. Molecular cell. 2010;40:179–204. [PMC free article] [PubMed] [Google Scholar]
- Clerici M, Mantiero D, Guerini I, Lucchini G, Longhese MP. The Yku70-Yku80 complex contributes to regulate double-strand break processing and checkpoint activation during the cell cycle. EMBO reports. 2008;9:810–818. [PMC free article] [PubMed] [Google Scholar]
- Deriano L, Roth DB. Modernizing the Nonhomologous End-Joining Repertoire: Alternative and Classical NHEJ Share the Stage. Annu Rev Genet. 2013 [PubMed] [Google Scholar]
- Dimitrov SD, Lu D, Naetar N, Hu Y, Pathania S, Kanellopoulou C, Livingston DM. Physiological modulation of endogenous BRCA1 p220 abundance suppresses DNA damage during the cell cycle. Genes & development. 2013;27:2274–2291. [PMC free article] [PubMed] [Google Scholar]
- Escribano-Diaz C, Orthwein A, Fradet-Turcotte A, Xing M, Young JT, Tkac J, Cook MA, Rosebrock AP, Munro M, Canny MD, et al. A Cell Cycle-Dependent Regulatory Circuit Composed of 53BP1-RIF1 and BRCA1-CtIP Controls DNA Repair Pathway Choice. Molecular cell. 2013 [PubMed] [Google Scholar]
- Fattah F, Lee EH, Weisensel N, Wang Y, Lichter N, Hendrickson EA. Ku regulates the non-homologous end joining pathway choice of DNA double-strand break repair in human somatic cells. PLoS genetics. 2010;6:e1000855. [PMC free article] [PubMed] [Google Scholar]
- Fong PC, Boss DS, Yap TA, Tutt A, Wu P, Mergui-Roelvink M, Mortimer P, Swaisland H, Lau A, O'Connor MJ, et al. Inhibition of poly(ADP-ribose) polymerase in tumors from BRCA mutation carriers. N Engl J Med. 2009;361:123–134. [PubMed] [Google Scholar]
- Foster SS, Balestrini A, Petrini JH. Functional interplay of the Mre11 nuclease and Ku in the response to replication-associated DNA damage. Molecular and cellular biology. 2011;31:4379–4389. [PMC free article] [PubMed] [Google Scholar]
- He F, Li L, Kim D, Wen B, Deng X, Gutin PH, Ling CC, Li GC. Adenovirus-mediated expression of a dominant negative Ku70 fragment radiosensitizes human tumor cells under aerobic and hypoxic conditions. Cancer research. 2007;67:634–642. [PubMed] [Google Scholar]
- Helmink BA, Tubbs AT, Dorsett Y, Bednarski JJ, Walker LM, Feng Z, Sharma GG, McKinnon PJ, Zhang J, Bassing CH, et al. H2AX prevents CtIP-mediated DNA end resection and aberrant repair in G1-phase lymphocytes. Nature. 2011;469:245–249. [PMC free article] [PubMed] [Google Scholar]
- Konstantinopoulos PA, Ceccaldi R, Shapiro GI, D'Andrea AD. Homologous Recombination Deficiency: Exploiting the Fundamental Vulnerability of Ovarian Cancer. Cancer Discov. 2015 [PMC free article] [PubMed] [Google Scholar]
- Lal A, Navarro F, Maher CA, Maliszewski LE, Yan N, O'Day E, Chowdhury D, Dykxhoorn DM, Tsai P, Hofmann O, et al. miR-24 Inhibits cell proliferation by targeting E2F2, MYC, and other cell-cycle genes via binding to “seedless” 3′UTR microRNA recognition elements. Molecular cell. 2009;35:610–625. [PMC free article] [PubMed] [Google Scholar]
- Lal A, Thomas MP, Altschuler G, Navarro F, O'Day E, Li XL, Concepcion C, Han YC, Thiery J, Rajani DK, et al. Capture of microRNA-bound mRNAs identifies the tumor suppressor miR-34a as a regulator of growth factor signaling. PLoS genetics. 2011;7:e1002363. [PMC free article] [PubMed] [Google Scholar]
- Lee DH, Pan Y, Kanner S, Sung P, Borowiec JA, Chowdhury D. A PP4 phosphatase complex dephosphorylates RPA2 to facilitate DNA repair via homologous recombination. Nature structural & molecular biology. 2010;17:365–372. [PMC free article] [PubMed] [Google Scholar]
- Liu G, Yang D, Rupaimoole R, Pecot CV, Sun Y, Mangala LS, Li X, Ji P, Cogdell D, Hu L, et al. Augmentation of Response to Chemotherapy by microRNA-506 Through Regulation of RAD51 in Serous Ovarian Cancers. J Natl Cancer Inst. 2015;107 [PMC free article] [PubMed] [Google Scholar]
- Moskwa P, Buffa FM, Pan Y, Panchakshari R, Gottipati P, Muschel RJ, Beech J, Kulshrestha R, Abdelmohsen K, Weinstock DM, et al. miR-182-mediated downregulation of BRCA1 impacts DNA repair and sensitivity to PARP inhibitors. Molecular cell. 2011;41:210–220. [PMC free article] [PubMed] [Google Scholar]
- Norquist B, Wurz KA, Pennil CC, Garcia R, Gross J, Sakai W, Karlan BY, Taniguchi T, Swisher EM. Secondary somatic mutations restoring BRCA1/2 predict chemotherapy resistance in hereditary ovarian carcinomas. J Clin Oncol. 2011;29:3008–3015. [PMC free article] [PubMed] [Google Scholar]
- Orom UA, Lund AH. Isolation of microRNA targets using biotinylated synthetic microRNAs. Methods. 2007;43:162–165. [PubMed] [Google Scholar]
- Sakaue-Sawano A, Kurokawa H, Morimura T, Hanyu A, Hama H, Osawa H, Kashiwagi S, Fukami K, Miyata T, Miyoshi H, et al. Visualizing spatiotemporal dynamics of multicellular cell-cycle progression. Cell. 2008;132:487–498. [PubMed] [Google Scholar]
- Shih KK, Qin LX, Tanner EJ, Zhou Q, Bisogna M, Dao F, Olvera N, Viale A, Barakat RR, Levine DA. A microRNA survival signature (MiSS) for advanced ovarian cancer. Gynecologic oncology. 2011;121:444–450. [PubMed] [Google Scholar]
- TCGA Integrated genomic analyses of ovarian carcinoma. Nature. 2011;474:609–615. [PMC free article] [PubMed] [Google Scholar]
- Xu G, Chapman JR, Brandsma I, Yuan J, Mistrik M, Bouwman P, Bartkova J, Gogola E, Warmerdam D, Barazas M, et al. REV7 counteracts DNA double-strand break resection and affects PARP inhibition. Nature. 2015 [PMC free article] [PubMed] [Google Scholar]