• Distinct transcriptional programs stratify ovarian cancer cell lines into the five major histological subtypes

      Barnes, Bethany M.; Nelson, Louisa; Tighe, Anthony; Burghel, George J.; Lin, I-Hsuan; Desai, Sudha; McGrail, Joanne C.; Morgan, Robert D.; Taylor, Stephen S.; orcid: 0000-0003-4621-9326; email: stephen.taylor@manchester.ac.uk (BioMed Central, 2021-09-01)
      Abstract: Background: Epithelial ovarian cancer (OC) is a heterogenous disease consisting of five major histologically distinct subtypes: high-grade serous (HGSOC), low-grade serous (LGSOC), endometrioid (ENOC), clear cell (CCOC) and mucinous (MOC). Although HGSOC is the most prevalent subtype, representing 70–80% of cases, a 2013 landmark study by Domcke et al. found that the most frequently used OC cell lines are not molecularly representative of this subtype. This raises the question, if not HGSOC, from which subtype do these cell lines derive? Indeed, non-HGSOC subtypes often respond poorly to chemotherapy; therefore, representative models are imperative for developing new targeted therapeutics. Methods: Non-negative matrix factorisation (NMF) was applied to transcriptomic data from 44 OC cell lines in the Cancer Cell Line Encyclopedia, assessing the quality of clustering into 2–10 groups. Epithelial OC subtypes were assigned to cell lines optimally clustered into five transcriptionally distinct classes, confirmed by integration with subtype-specific mutations. A transcriptional subtype classifier was then developed by trialling three machine learning algorithms using subtype-specific metagenes defined by NMF. The ability of classifiers to predict subtype was tested using RNA sequencing of a living biobank of patient-derived OC models. Results: Application of NMF optimally clustered the 44 cell lines into five transcriptionally distinct groups. Close inspection of orthogonal datasets revealed this five-cluster delineation corresponds to the five major OC subtypes. This NMF-based classification validates the Domcke et al. analysis, in identifying lines most representative of HGSOC, and additionally identifies models representing the four other subtypes. However, NMF of the cell lines into two clusters did not align with the dualistic model of OC and suggests this classification is an oversimplification. Subtype designation of patient-derived models by a random forest transcriptional classifier aligned with prior diagnosis in 76% of unambiguous cases. In cases where there was disagreement, this often indicated potential alternative diagnosis, supported by a review of histological, molecular and clinical features. Conclusions: This robust classification informs the selection of the most appropriate models for all five histotypes. Following further refinement on larger training cohorts, the transcriptional classification may represent a useful tool to support the classification of new model systems of OC subtypes.
    • Replication catastrophe is responsible for intrinsic PAR glycohydrolase inhibitor-sensitivity in patient-derived ovarian cancer models

      Coulson-Gilmer, Camilla; Morgan, Robert D.; Nelson, Louisa; Barnes, Bethany M.; Tighe, Anthony; Wardenaar, René; Spierings, Diana C. J.; Schlecht, Helene; Burghel, George J.; Foijer, Floris; et al. (BioMed Central, 2021-10-16)
      Abstract: Background: Patients with ovarian cancer often present at advanced stage and, following initial treatment success, develop recurrent drug-resistant disease. PARP inhibitors (PARPi) are yielding unprecedented survival benefits for women with BRCA-deficient disease. However, options remain limited for disease that is platinum-resistant and/or has inherent or acquired PARPi-resistance. PARG, the PAR glycohydrolase that counterbalances PARP activity, is an emerging target with potential to selectively kill tumour cells harbouring oncogene-induced DNA replication and metabolic vulnerabilities. Clinical development of PARG inhibitors (PARGi) will however require predictive biomarkers, in turn requiring an understanding of their mode of action. Furthermore, differential sensitivity to PARPi is key for expanding treatment options available for patients. Methods: A panel of 10 ovarian cancer cell lines and a living biobank of patient-derived ovarian cancer models (OCMs) were screened for PARGi-sensitivity using short- and long-term growth assays. PARGi-sensitivity was characterized using established markers for DNA replication stress, namely replication fibre asymmetry, RPA foci, KAP1 and Chk1 phosphorylation, and pan-nuclear γH2AX, indicating DNA replication catastrophe. Finally, gene expression in sensitive and resistant cells was also examined using NanoString or RNAseq. Results: PARGi sensitivity was identified in both ovarian cancer cell lines and patient-derived OCMs, with sensitivity accompanied by markers of persistent replication stress, and a pre-mitotic cell cycle block. Moreover, DNA replication genes are down-regulated in PARGi-sensitive cell lines consistent with an inherent DNA replication vulnerability. However, DNA replication gene expression did not predict PARGi-sensitivity in OCMs. The subset of patient-derived OCMs that are sensitive to single-agent PARG inhibition, includes models that are PARPi- and/or platinum-resistant, indicating that PARG inhibitors may represent an alternative treatment strategy for women with otherwise limited therapeutic options. Conclusions: We discover that a subset of ovarian cancers are intrinsically sensitive to pharmacological PARG blockade, including drug-resistant disease, underpinned by a common mechanism of replication catastrophe. We explore the use of a transcript-based biomarker, and provide insight into the design of future clinical trials of PARGi in patients with ovarian cancer. However, our results highlight the complexity of developing a predictive biomarker for PARGi sensitivity.