Class II phosphatidylinositol 3-kinase ß in nutrient signaling, endocytosis and centronuclear myopathy

Abstract

To adapt to the availability of nutrients and growth factors, cells must sense these cues and adapt their metabolism and proliferation accordingly. Defects in these processes have been associated with many diseases ranging from diabetes to cancer. Nutrient signaling is controlled by the master regulator mTOR complex 1 (mTORC1). When nutrients and growth factors are readily available, mTORC1 localizes to lysosomes and phosphorylates target proteins initiating anabolic processes like protein translation and lipid synthesis and inhibiting catabolic processes like autophagy. Amino acids induce the translocation of mTORC1 to the lysosomal surface via the activity of Rag GTPases whereas growth factors lead to the activation of the small GTPase Rheb at the lysosome that interacts with mTORC1 to activate it. When growth factor levels are low, mTORC1 activity is inhibited by lysosomal PI(3,4)P2 through so far unknown mechanisms. PI(3,4)P2 is synthesized from PI(4)P by the class II PI 3- kinase β (PI3KC2β). PI3KC2β itself is inhibited by growth factor signaling. Active growth factor receptors promote the activity of the mTORC2 complex which then activates protein kinase N2 (PKN2) by phosphorylation. PKN2 in turn phosphorylates PI3KC2β in its unstructured N terminal domain, triggering the association of inhibitory 14-3-3 proteins that sequester PI3KC2β in the cytoplasm. How PI3KC2β is actively recruited to the lysosomal surface upon growth factor starvation was previously unknow. In this study I demonstrate that active, GTP bound Rab7A interacts with PI3KC2β and recruits it to the lysosomal surface upon growth factor starvation. Rab7A was required for the interaction of PI3KC2β with the mTORC1 subunit Raptor and PI3KC2β translocation to lysosome. Furthermore, lysosomal PI(3,4)P2 and mTORC1 inhibition depended on Rab7A activity. To further investigate the influence of PI3KC2β on mTORC1 composition, I generated a HEK293T cell line that endogenously expresses eGFP-Raptor using CRISPR/Cas9 mediated genome editing. This cell line allowed for the isolation of mTORC1 by immunoprecipitation. PI3KC2β depletion indeed altered the mTORC1 composition, as the interaction of the inhibitory subunit PRAS40 was reduced by 40 %. The exact mechanisms of this observation remain the subject of further studies. Beyond its role in mTORC1 signaling, PI3KC2β is known as an important regulator in x linked centronuclear myopathy (XLCNM). XLCNM is a severe muscular disorder caused by loss 12 of function mutations in the PI(3)P phosphatase MTM1. The depletion of PI3KC2β in MTM1 knock out mice prevents disease onset by hereto unknown mechanisms. MTM1 is required for the depletion of PI(3)P from recycling endosomes to enable the synthesis of PI(4)P by PI4KIIα which is in turn required for the recruitment of the exocyst complex and exocytosis of cargo destined for recycling. This leads to the endosomal accumulation of β1 integrins and triggers defects in muscle fiber generation through myoblast fusion, a process that depends on proper β1 integrin localization and activation. Using C2C12 myoblast cell lines harboring knockouts (KO) of MTM1, PI3KC2β or both MTM1 and PI3KC2β I could reproduce the myoblast fusion defect induced by MTM1 loss in a cellular model. Furthermore, I could confirm that MTM1 KO C2C12 cells had less active β1 integrin at the cell surface. Contrarily, HeLa cells depleted of PI3KC2β had elevated active β1 integrin surface levels and decreased rates of active β1 integrin endocytosis. Both findings could be rescued with the re-expression wild type PI3KC2β but not with kinase inactive PI3KC2β or a mutant that only synthesizes PI(3)P but not PI(3,4)P2. Thus PI3KC2β is required for active β1 integrin endocytosis trough the synthesis of PI(3,4)P2. In line with that, PI3KC2β localized to clathrin coated pits (CCPs) and co-immunoprecipitated with the integrin endocytosis adaptor Dab2 and the endocytic proteins clathrin heavy chain and intersectin-1. I therefore propose a model where MTM1 is required for active β1 integrin recycling and PI3KC2β is required for active β1 integrin endocytosis. If the recycling process fails due to MTM1 loss, active surface β1 integrin levels can be restored by depleting PI3KC2β and thus reducing the rate of β1 integrin endocytosis.