Investigating PLEKHS1 adaptor function in PI3K signalling and human cancer biology

The Class-IA PI3K signalling pathway regulates key anabolic cellular processes including proliferation, survival, growth and metabolism. Oncogenic hyperactivation of the pathway occurs across tumour types, commonly via genetic alterations in key pathway genes (e.g., *PIK3CA*, *PTEN*). Despite extensive research, the primary drivers of healthy or oncogenic PI3K signalling *in vivo* are largely unknown and PI3K pathway inhibitors show limited clinical benefit in solid tumours. Identifying upstream regulators of oncogenic PI3K activation can reveal novel therapeutic targets and/or biomarkers. Previously generated *in vivo* interactomics data identified the RTK/IRS network as the primary driver of homeostatic PI3K activation in healthy mouse prostate, and PLEKHS1 as the primary driver of oncogenic PI3K activation in cancerous, PTEN-null mouse prostate. PTEN phosphatase is a tumour-suppressor and inactivated commonly in prostate cancers. PLEKHS1 is otherwise a poorly characterised protein. This thesis presents research investigating the relevance and mechanisms of PLEKHS1 adaptor protein function to oncogenic PI3K pathway remodelling in human cancers, with a focus on prostate cancer. TCGA pan-cancer analyses report *IRS1* downregulation and *PLEKHS1* upregulation in several primary carcinomas, compared to healthy tissue. In prostate, endometrial and breast cancers, *PLEKHS1* expression correlates with PI3K pathway activation and alterations in key pathway genes (*PTEN*, *PIK3CA*). In castration-resistant prostate cancer, *PLEKHS1* is under-expressed in neuroendocrine cancers, compared to adenocarcinomas. ChIP-Seq datasets report binding of key prostate cancer transcriptional regulators to the *PLEKHS1* locus (e.g., AR, FOXA1). To investigate PLEKHS1 adaptor protein function *in vitro*, LNCaP prostate cancer cell line is used due to its PTEN-null status and PI3K pathway hyperactivation. Transient over-expression of PLEKHS1 in LNCaP cells generates a robust PIP3 response via YxxM-dependent interactions with endogenous PI3K regulatory subunits (p85, p55γ). PLEKHS1 phosphorylation and PI3K interaction are driven by SFK activity and independent of serum growth factors. In this context and alike *in vivo* findings, the PLEKHS1 PH domain which specifically binds PIP3 and PI(3,4)P2 mediates positive feedback PI3K activation. PLEKHS1 adaptor function is mostly conserved across mouse and human protein isoforms but altered for human isoform 2 which has a truncated PH domain. Endogenous PLEKHS1 function is also investigated in LNCaP cells. CRISPR/Cas9 mediated PLEKHS1 knockout does not significantly affect PIP3 levels, Akt phosphorylation or cell viability in LNCaP cells cultured as monolayers or spheroids. Consistently, no endogenous PLEKHS1-PI3K interaction is detected using co-immunoprecipitation in LNCaP cells. However, other *in vitro* disease models did provide evidence of endogenous SFK-PLEKHS1-PI3K signalling axis. In the PTEN-null breast cancer cell line HCC70, endogenous PLEKHS1 is highly expressed and SFKs mediate YxxM phosphorylation and interaction with PI3K regulatory subunits (p85, p55γ). However, siRNA mediated PLEKHS1 knockdown does not affect PIP3 levels, Akt phosphorylation or cell viability. Mouse dorsolateral prostate derived organoids are used as an alternative *in vitro* model. PTEN-KO organoids show elevated PLEKHS1 protein expression and PIP3 levels compared to WT. PTEN/PLEKHS1-DKO organoids phenocopy *in vivo* data with reduced PIP3 levels and growth compared to PTEN-KO organoids. Collectively, this work substantially advances prior *in vivo* findings from mouse models by (1) demonstrating the relevance of PLEKHS1 adaptor function to PI3K signalling in human cancers using *in vitro* and clinical data, (2) elucidating the molecular mechanisms of the SFK-PLEKHS1-PI3K signalling axis, and (3) revealing highly context-specific utilisation of PLEKHS1 as a primary driver of oncogenic PI3K signalling