Background and Aims: Osteosarcoma (OS) is the most common primary malignant tumour of the bone in children and adolescents and young adults (AYAs). Currently there are no effective targeted treatments for advanced disease, and tumours rarely possess actionable, genomic alterations. Activated (phosphorylated) protein kinase signalling has been identified as a key driver event in subsets of human osteosarcoma. This study aims to utilise a phosphoproteomic approach to identify aberrantly activated kinases in OS, which novel targeted therapies can be directed against.
Hypothesis: Activated kinase signalling is a key driver event in subsets of human OS. Phosphoproteomic profiling of OS tumours will identify aberrantly activated kinases, which if targeted by matched drugs could indicate novel, effective, patient-specific OS therapies.
Methods: Eighteen unique patient-derived xenograft (PDX) models from high-risk sarcoma patients enrolled in the ZERO Childhood Cancer Precision Medicine Trial (ZERO), including four OS (one with matched disease samples), were subjected to Mass Spectrometry-based phosphoproteomic profiling. Selected activation signatures in OS were validated by Western Blot (WB) and linked to molecular genomics/transcriptomics. Eight drug targets were selected for functional validation in vitro in short and long-term drug viability assays across standard OS (n=3) and newly established OS PDX cell lines from ZERO patients (n=6). Two top-performing novel agents underwent further in vitro combination testing using ZERO’s clinically applicable 150-compound drug library and in vivo exploration. Drug mechanism-of-action was investigated by WB and cell cycle analysis.
Results: In OS, we identified ~3400 phosphosites, corresponding to 1441 unique proteins. Eight novel drug targets with elevated phosphorylation across OS patient models were selected for further validation with ten kinase inhibitors. A pan-CDK inhibitor and dual mTORC1/2 inhibitor demonstrated remarkably low IC50 values of <0.05uM or <0.1uM, respectively, across standard OS and OS PDX cell lines in short-term in vitro assays and were selected for further validation. On-target drug activity and mechanism-of-action was confirmed by WB, where reduced expression of downstream targets was observed. Clonogenic assays further confirmed the remarkable sensitivity of tested cell lines to these drugs. In vitro combination testing revealed synergy of the pan-CDK inhibitor with at least three targeted agents, with one combination demonstrating in vivo efficacy against OS xenograft models.
Conclusion: Our study demonstrates significant anticancer activity of two novel kinase inhibitors in vitro in standard OS and newly established high-risk ZERO OS PDX cell lines. This study demonstrates the potential application of phosphoproteomics for the discovery of actionable targets in high-risk OS, which provides a potential avenue for the discovery of novel, patient-specific OS therapies.