Cryo-EM structure and biochemical analysis reveal the basis of the functional difference between human PI3KC3-C1 and -C2.

Phosphatidylinositol 3-phosphate (PI3P) plays essential roles in vesicular trafficking, organelle biogenesis and autophagy. Two class III phosphatidylinositol 3-kinase (PI3KC3) complexes have been identified in mammals, the ATG14L complex (PI3KC3-C1) and the UVRAG complex (PI3KC3-C2). PI3KC3-C1 is crucial for autophagosome biogenesis, and PI3KC3-C2 is involved in various membrane trafficking events. Here we report the cryo-EM structures of human PI3KC3-C1 and PI3KC3-C2 at sub-nanometer resolution. The two structures share a common L-shaped overall architecture with distinct features. EM examination revealed that PI3KC3-C1 "stands up" on lipid monolayers, with the ATG14L BATs domain and the VPS34 C-terminal domain (CTD) directly contacting the membrane. Biochemical dissection indicated that the ATG14L BATs domain is responsible for membrane anchoring, whereas the CTD of VPS34 determines the orientation. Furthermore, PI3KC3-C2 binds much more weakly than PI3KC3-C1 to both PI-containing liposomes and purified endoplasmic reticulum (ER) vesicles, a property that is specifically determined by the ATG14L BATs domain. The in vivo ER localization analysis indicated that the BATs domain was required for ER localization of PI3KC3. We propose that the different lipid binding capacity is the key factor that differentiates the functions of PI3KC3-C1 and PI3KC3-C2 in autophagy.

Baculovirus production of fully-active phosphoinositide 3-kinase alpha as a p85alpha-p110alpha fusion for X-ray crystallographic analysis with ATP competitive enzyme inhibitors.

Phosphoinositide 3-kinases have been targeted for therapeutic research because they are key components of a cell signaling cascade controlling proliferation, growth, and survival. Direct activation of the PI3Kalpha pathway contributes to the development and progression of solid tumors in breast, endometrial, colon, ovarian, and gastric cancers. In the context of a drug discovery effort, the availability of a robust crystallographic system is a means to understand the subtle differences between ATP competitive inhibitor interactions with the active site and their selectivity against other PI3Kinase enzymes. To generate a suitable recombinant design for this purpose, a p85alpha-p110alpha fusion system was developed which enabled the expression and purification of a stoichiometrically homogeneous, constitutively active enzyme for structure determination with potent ATP competitive inhibitors (Raha et al., in preparation) [56]. This approach has yielded preparations with activity and inhibition characteristics comparable to those of the full-length PI3Kalpha from which X-ray diffracting crystals were grown with inhibitors bound in the active site.