Andrea Miyahira: Hi, I'm Andrea Miyahira here at the Prostate Cancer Foundation. Please welcome doctors Francesca Lorenzin and Giulia Fracassi of University of Trento in Italy. They will be discussing their paper-- CRISPR/Cas9 screens identify LIG1 as a sensitizer of PARP inhibitors in castration-resistant prostate cancer, recently published in the Journal of Clinical Investigation. Thank you both for joining.

Francesca Lorenzin: Thank you for inviting us, Andrea.

Giulia Fracassi: Thank you for having us. So thank you, Dr. Miyahira, for the kind introduction. We are glad to present our recent work about the synthetic lethal interaction between LIG1 and PARP in castration-resistant prostate cancer. We all recognize the significant impact that PARP inhibitors have had in treating castration-resistant prostate cancers.

However, results from clinical trials testing PARP inhibitors have raised questions concerning patient enrollment criteria. In particular, the observation that a subset of biomarker-negative patients benefits from PARP inhibitor-based treatments suggests that there may be defects in other DNA repair genes that lead to sensitivity. In this regard, we have to consider that there are many DNA repair genes beyond those involved in homologous recombinational repair pathway and that some of these genes already show therapeutic potential in preclinical studies testing PARP inhibitors.

So based on these premises, we aimed at the identification of low-frequency DNA repair gene variants as potential biomarkers for castration-resistant prostate cancer. With this purpose, we ran a multiple CRISPR/Cas9 screening in prostate cancer cell lines treated with PARP inhibitors using a custom DNA sgRNA library-- to target a larger set of DNA repair genes. With this approach, we nominated multiple candidates highlighted here in orange, whose knockouts potentially lead to sensitivity to PARP inhibitors.

We then validated the top candidates, including LIG1, EME1, and FAAP24, through survival assays, in which we saw that the knockout of any of these three genes leads to sensitivity to PARP inhibitor in 22Rv1 cells. We next analyzed genomic data from TCGA and Stand Up 2 Cancer PCF data sets. And by focusing on loss of function alterations, we found that the frequency of alteration of our three candidates is comparable with those of FDA-approved biomarkers for olaparib.

Moreover, focusing only on our three hits, we saw that LIG1 is the most frequently aberrant in both primary and advanced prostate cancer patients. And based on these observations, we nominated LIG1 as the most promising candidate to pursue in vitro. By exploring the mechanism underlying the synthetic lethality between LIG1 and PARP, we saw that the combination of LIG1 loss and PARP inhibition increases DNA damage and replication stress, eventually leading to apoptosis in 22Rv1 cell line.

We then explored the therapeutic applicability of the synthetic lethality between LIG1 and PARP. And we found by analyzing TCGA pan-cancer data set that LIG1 is aberrant in multiple tumor types, and that its loss, combined with the olaparib treatment, significantly decreases the viability of multiple cancer cell lines beyond prostate cancer cells. We then tested a combined pharmacological inhibition of LIG1 and PARP, and we saw that by combining LIG1 inhibitors and olaparib treatment, this approach significantly reduces the viability of prostate cancer and breast cancer cell lines, while it has only a mild effect on non-tumorigenic cells.

More interestingly, we saw that when LIG1 inhibitor is combined with the olaparib treatment, it markedly increases the cytotoxic effect of PARP inhibition in BRCA1-mutated cells. Finally, in a collaboration with the laboratory of Dr. Joaquin Mateo, we confirmed that the combination of LIG1 loss and PARP inhibition significantly reduces tumor proliferation in prostate cancer xenograft mouse model. So overall, our findings support the inclusion of LIG1 as a biomarker for PARP inhibitor-based therapy. And they provide also initial evidence for effective PARP inhibitor-based drug combination.

Going forward, our data, our findings highlight the importance of expanding the biomarker discovery beyond the genes involved in homologous recombination repair pathway. And it also underlines the importance of extending the genomic profiling of unselected PARP inhibitor trial cohorts in order to expand the patient population that might benefit from these treatments. With this, I conclude my presentation. I want to thank you all our collaborators and our funding sources, and I want to thank you again for the opportunity to present our work. And now I will be happy to take any questions.

Andrea Miyahira: Thank you so much, Dr. Fracassi, for sharing that. So do you know of any patients with alterations in LIG1 or any of these other genes that you identified have been treated with PARP inhibitors, and if so, how they responded? And would you expect differences in PARP inhibitor response between tumors with homozygous versus heterozygous loss?

Francesca Lorenzin: So I can maybe start with responding to the second part of the question and saying that in our hands, full impairment of LIG1 is necessary to have a sensitivity to PARP inhibitors, and this is also consistent to what was reported for other DNA repair genes in patients regarding PARP sensitivity. And so we would have liked to validate our findings in the patient-derived data. And we contacted Professor de Bono to examine the TOPARP-B trial data.

And out of the 98 patients included in the trial, 3 were reported as having homozygous deletion in LIG1, and they were all reported as responders to PARP inhibition. However, as by design, this trial recruits patients that have at least mutation in other-- in a selected panel of DNA repair gene mutations, we could not conclude that LIG1 was the driver of the PARP sensitivity.

However, and interestingly, out of the five [Inaudible]-mutated patients that were in the trial, the only one that was reported as a responder to PARP inhibition was the one with concurrent mutation in LIG1. So this is pointing to the right direction, but is not a conclusive analysis. We also checked for other trials that recruit patients regardless of DNA repair gene mutations. However, this trial profiled patients with a panel of DNA repair genes that did not include LIG1, and also some of the data were not publicly available, therefore precluding any analysis in this direction.

Andrea Miyahira: OK. Thank you. And do you know whether tumors with LIG1 or these other alterations have any unique genomic architectures or specific mutational patterns?

Giulia Fracassi: Maybe I can answer this question. Actually, we didn't perform this analysis but is definitely of interest. I think that the main limitation in this context is that LIG1 is not a frequently aberrant. So we would need to expand the number of tumor samples in order to reach the statistical significance. We already checked for the co-occurrence of defects in LIG1 and other DNA repair genes. But again, because of the low numbers, we didn't reach the statistical significance. But for sure, yes, is something that it might be interesting to analyze.

Andrea Miyahira: OK. Thank you. And based on these data, is LIG1 a promising therapeutic target? And what strategy do you think would be optimal?

Giulia Fracassi: So based on our findings, especially the observation that LIG1 is aberrant both in primary and advanced prostate cancer patients, we believe that LIG1 is a promising therapeutic target. So considering that at the moment there are no LIG1-specific inhibitors suitable for in vivo studies, we believe that the optimal therapeutic strategy might be to-- could be to test PARP inhibitor in a context in which LIG1 is aberrant.

However, considering our data and also recently published results from the Nicholson lab, we think that developing LIG1-specific inhibitors could expand the therapeutic applicability of the synthetic lethality that we identified. In particular, based on preliminary evidence, we think that the combination of LIG1, the use of LIG1 inhibitor, might improve the cytotoxic effect of PARP inhibitor in BRCA-mutated tumors or resensitize BRCA-mutated tumors to that developed PARP inhibitor resistance.

Andrea Miyahira: OK, thank you. And what are your next steps? And are there any translational plans?

Francesca Lorenzin: So ideally, the next step would be to test whether LIG1 aberration leads to sensitivity to PARP in patients. And so design new trials that include LIG1 and also include the other non-homologous repair genes that were previously identified as PARP sensitizers. However, this is not really our expertise.

And so what we can focus on is, for example, the development and testing of a novel LIG1 inhibitor, to test whether the pharmacological inhibition of LIG1 and PARP is really a valuable therapeutic strategy for cancer patients. And also, our screening revealed other hits that are connected to sensitivity to PARP. And so validating and testing these hits could lead to identifying novel biomarkers and novel therapeutic targets. And this might be particularly important or relevant when considering that PARP inhibitors are being also used for early stages of the disease. And so there are being more patients that will be treated with PARP inhibitors.

Andrea Miyahira: OK, thank you so much. Thank you, Dr. Lorenzin and Dr. Fracassi, for sharing this with us today.

Francesca Lorenzin: Thank you, Andrea, for having us.

Giulia Fracassi: Thank you.