Introduction
Malignant pleural mesothelioma (MPM) is a rare and aggressive tumor primarily caused by inhalation of asbestos. Treatment for MPM is most commonly surgery, with chemotherapy/radiotherapy for residual disease [
1,
2]. The prognosis for this disease is dismal, given that complete resection of the tumor is difficult, and the lack of sensitivity to chemotherapy or radiation [
3]. In addition, response to newer agents, such as the anti-angiogenic biological therapies and the kinase targeted small molecules, has been disappointing [
1,
2].
Interestingly, between 23 and 64% of MPMs have somatic inactivating mutations in the BAP1 (BRCA1-associated protein 1) gene [
4‐
7]. BAP1 is also mutated in a fraction of uveal melanomas, renal cell carcinomas, and cholangiocarcinoma [
8]. BAP1 is a member of the BRCA1/BARD1 complex and plays a role in homologous recombination (HR) DNA double strand break (DSB) repair [
9,
10]. BAP1 interacts with the BRCA1 RING finger domain, which mediates BRCA1’s E3 ubiquitin ligase activity [
11]. BAP1 functions as a deubiquitinase and BRCA1 likely recruits it to sites of DSBs to restore histone ubiquitination to native levels after repair is completed [
11].
We hypothesized that BAP1-mutant mesothelioma might be addicted to PARP1-mediated DNA repair pathways, similar to the BRCA1/2-mutant breast and ovarian cancers [
12‐
14]. BAP1-mutant MPM would have defects in the HR DNA repair, and thus may also require PARP1 for proliferation. PARP1 inhibition in BRCA1-deficient cancers results in collapsed replication forks and genomic instability leading to mitotic catastrophe and ultimately cell death [
12,
13]. When a cancer arises from defects in one DNA repair pathway, such as HR in the BRCA1/2-mutant malignancies, they become addicted to other DNA repair pathways for survival during replication [
12‐
14]. A well-known example of this is the inhibition of the DNA repair component PARP1 in the BRCA1-deficient cancers result in replication fork collapse and mitotic catastrophe [
12‐
14]. In this situation, PARP1 inhibition stalls replication forks, which cannot be repaired because of the underlying defect in HR in these cancers [
12]. Sensitivity to PARP1 inhibition is one biomarker of HR deficiency [
12,
14].
Significantly, we discovered that the clinical PARP1 inhibitors, niraparib and olaparib, markedly decreased clonal survival in all MPM cell lines tested, regardless of BAP1 status. This implies that defects in homologous recombination DNA repair are at the origin of most MPM. More importantly, these data suggest a new modality for targeted therapy to be studied in a cancer that desperately needs effective treatment.
Discussion
MPM responds poorly to current chemotherapy or radiation therapy, and many presenting cases are unresectable due to the spread of the disease along pleural planes or chest wall invasion [
1,
2]. The average survival of patients with recurrent or unresectable disease treated with the best chemotherapy regimen (cisplatin or carboplatin plus pemetrexed) is one year [
1‐
3]. Adding the angiogenesis inhibitor bevacizumab to this regimen adds another 6 months of survival [
1,
2]. However, all unresected or recurrent patients eventually die of their disease [
1‐
3]. Many patients with unresectable MPM choose palliative care instead of systemic chemotherapy because current regimens are difficult to tolerate [
2]. Targeted inhibitors against EGFR or VEGFR have been disappointing in their clinical benefit, so there are no approved precision therapies [
2]. Thus, new therapies are needed for this devastating disease.
Between 23 and 64% of malignant pleural mesotheliomas (MPMs) have somatic inactivating mutation in the BAP1 gene [
4‐
7]. Mutation in BAP1 leads to defects in the HR DNA repair pathway, and it is considered a tumor suppressor [
6‐
8]. The HR pathway is required for faithful repair of stalled replication forks, and HR-deficient cells become genomically unstable [
12‐
14]. This genomic instability can result in fusion of the stalled replication forks, generating chromosomal ligations [
12‐
14,
20]. Thus, BAP1 deficiency may be an initiating event in oncogenesis because stalled replication forks are no longer faithfully repaired due to the lack of HR [
12‐
14]. This is similar to the inherited HR deficiencies, such as the BRCA1 and 2 mutations, that lead to breast and ovarian cancers [
12,
13].
PARP1 inhibitors are widely used as drugs that target these BRCA1 and 2-mutant cancers [
12,
13,
18,
19]. PARP1 is essential for three DNA repair pathways—single strand-break repair, base excision repair (BER), and alternative non-homologous end joining (aNHEJ). Inhibition of PARP1 by the clinical PARP1 inhibitor drugs results in the synthetic lethal death of BRCA1 or 2 deficient cells. This cell death has multiple etiologies, all involving chromosomal instability and mitotic catastrophe [
12‐
14,
20].
In the absence of PARP1 activity, SSBs increase in genomic DNA, either from defective SSB repair or arrested BER [
13]. The unrepaired SSBs are converted into DS ends upon replication fork arrival [
12]. In the absence of HR, replication fork collapse can be rescued by the aNHEJ repair pathway, but this pathway also requires PARP1 [
14]. When PARP1 is inhibited in HR-deficient cells, there is an increased possibility that collapsed replication forks can be aberrantly ligated together with classical NHEJ. This leads to chromosome fusions and genomic instability, manifested by micronuclei, and bridging, as seen here [
12,
20]. Another mechanism of replication fork stalling and collapse in these HR-deficient cells is that the clinical PARP1 inhibitors lock PARP1 onto DNA, blocking progression of a replication fork [
13,
18]. In that mechanism, replication forks would also be arrested at the PARP1 inhibitor-DNA complex, and that cell would lack both HR and aNHEJ activities, and thus would have great difficulty in restarting that stalled fork [
12,
13,
18].
Olaparib is an oral PARP1 inhibitor that has been recently approved for the treatment of relapsed ovarian cancer that harbors BRCA1 or 2 mutations [
12,
13]. Niraparib is another approved potent oral PARP1 inhibitor that is well-tolerated [
18,
19]. Results of a large clinical trial using single agent Niraparib in relapsed ovarian cancer demonstrated that it was a safe and highly effective agent in ovarian cancer, regardless of HR or BRCA1/2 status [
19]. The data presented here demonstrate that PARP1 inhibition using niraparib or olaparib was lethal in multiple MPM cell lines, regardless of BAP1 mutational status. The finding that niraparib decreased replication fork restart (Fig.
2a) and increased genomic instability (Fig.
2c, d) is consistent with the molecular mechanism of synthetic lethality in PARP1 inhibition in BRCA1 and 2-deficient breast and ovarian cancers [
12,
13,
18].
Bott et al. stated that they found no difference in the sensitivity to niraparib between BAP1-mutant and wild-type MPM using a cell respiration assay [
7]. There was no data shown and there was no mention of whether niraparib was cytotoxic in any of the cell lines [
7]. Using colony formation assays to measure cytotoxicity, which assesses not just the respiration of a cell population but clonal cell proliferative capacity, we also found that there was little difference in sensitivity to niraparib between BAP1-mutant or wild-type MPM cell lines. However, we found that niraparib was significantly cytotoxic to each MPM cell line, regardless of the BAP1 status (Fig.
1). This is consistent with a recent clinical trial where niraparib demonstrated efficacy in ovarian cancer regardless of HR status [
19]. Olaparib also demonstrated cytotoxicity to all BAP1-mutant MPM cell lines (Fig.
1, supplemental Fig. 1). In H290 BAP1 wild-type MPM cells olaparib had significant cytotoxicity, while in the CRL-2081 BAP1 wild-type MPM cells it has less benefit (Fig.
1e, f, Supplemental Fig. 1).
There are two potential reasons why both wild type and mutant BAP1 were sensitive to niraparib. First, there could be undefined HR defects in the BAP1 wild-type MPM cells. It is possible that all MPM results from HR deficiencies, with BAP1 just being the most common type of deficiency [
4‐
7]. Second, niraparib is such a potent PARP1 inhibitor that unrepaired oxidative DNA damage leads to such massive replication fork collapse that neither an HR-deficient or replete cell can tolerate it. These two mechanisms are not mutually exclusive, and niraparib could be lethal from inhibiting BER just as much from deficient SSB repair or aNHEJ [
13,
14].
PARP1 inhibition at clinically relevant concentrations resulting in significant cytotoxicity in MPM demonstrates that agents such as olaparib and niraparib are promising for use in the treatment of MPM, for which effective treatment is desperately needed. Thus, clinical trials of PARP inhibitors in this difficult to treat malignancy would be warranted. These data are also a further demonstration that other malignancies with acquired defects in HR may be responsive to PARP1 inhibition [
8]. Studying the effects of PARP1 inhibition in the other BAP1-mutant malignancies, such as renal cell carcinoma and cholangiocarcinoma, would be especially important, since cancers also have little beneficial therapy once initial resection has failed [
8].