Phosphatidylinositol-3-kinase (PI3K) enzymes, producing signaling phosphoinositides at plasma and intracellular membranes, are key in intracellular signaling and vesicular trafficking pathways. PI3K is a family of eight enzymes divided into three classes with various functions in physiology and largely deregulated in cancer. Here, we will review the recent evidence obtained during the last 5 years on the roles of PI3K class I, II and III isoforms in tumor biology and on the anti-tumoral action of PI3K inhibitors in preclinical cancer models. The dependency of tumors to PI3K isoforms is dictated by both genetics and context (e.g., the microenvironment). The understanding of class II/III isoforms in cancer development and progression remains scarce. Nonetheless, the limited available data are consistent and reveal that there is an interdependency between the pathways controlled by all PI3K class members in their role to promote cancer cell proliferation, survival, growth, migration and metabolism. It is unknown whether this feature contributes to partial treatment failure with isoform-selective PI3K inhibitors. Hence, a better understanding of class II/III functions to efficiently inhibit their positive and negative interactions with class I PI3Ks is needed. This research will provide the proof-of-concept to develop combination treatment strategies targeting several PI3K isoforms simultaneously.

Rapid, easy and early pancreatic cancer diagnosis and therapeutic follow up continue to necessitate an increasing attention towards the development of effective treatment strategies for this lethal disease. The non invasive quantitative assessment of pancreatic heterogeneity is limited. Here, we report the development of a preclinical imaging protocol using ultrasonography and shear wave technology in an experimental in situ pancreatic cancer model to measure the evolution of pancreatic rigidity.

For patients with metastatic pancreatic cancer that are not eligible for surgery, signal-targeted therapies have so far failed to significantly improve survival. These therapeutic options have been tested in phase II/III clinical trials mostly in combination with the reference treatment gemcitabine. Innovative therapies aim to annihilate oncogenic dependency, or to normalize the tumoural stroma to allow immune cells to function and/or re-vascularisation to occur. Large scale transcriptomic and genomic analysis revealed that pancreatic cancers display great heterogeneity but failed to clearly delineate specific oncogene dependency, besides oncogenic Kras. Beyond these approaches, proteomics appears to be an appropriate approach to classify signal dependency and to identify specific alterations at the targetable level. However, due to difficulties in sampling, proteomic data for this pathology are scarce. In this review, we will discuss the current state of clinical trials for targeted therapies against pancreatic cancer. We will then highlight the most recent proteomic data for pancreatic tumours and their metastasis, which could help to identify major oncogenic signalling dependencies, as well as provide future leads to explain why pancreatic tumours are intrinsically resistant to signal-targeted therapies. We will finally discuss how studies on phosphatidylinositol-3-kinase (PI3K) signalling, as the paradigmatic pro-tumoural signal downstream of oncogenic Kras in pancreatic cancer, would benefit from exploratory proteomics to increase the efficiency of targeted therapies.

Supplementary Table 2 contains the list of isoform selective peptides in each condition, with quantification and global pathway analysis.

Supplementary Figures file contains merged Figure S1 to S8 including description of cell lines, Human samples, PI3K inhibitor stability, SILAC setup and controls, STRING analysis, quantifications of WB and cellular assays, prediction of kinase motifs found regulated, effects of combination of treatment in vitro, correlation between cell survival and phosphopeptide intensity, effects of combination treatment in Human samples and in vivo.