PTEN first debuted as a tumor suppressor essential to abrogate the PKB/Akt pathway by controlling Phosphatidylinositol (3,4,5)-trisphosphate (PI(3,4,5)P3) levels at the plasma membrane (PM). Now PTEN has been shown to possess critical cytosol and nucleus based roles, some of which are phosphatase independent. How then does PTEN triage its cancer-abating function to dephosphorylate PI(3,4,5)P3 with its duties elsewhere in the cell? Single molecule TIRFM studies of PTEN in living cells have revealed a dynamic PTEN- membrane association with membrane dwell times of ~200 ms. Paradoxically, SPR studies indicate that PTEN lingers on anionic bilayers for ~700 s, demonstrating a very high nonspecific electrostatic affinity. A recent computational model developed by Nesbitt et al. reconciles these differences with the prediction that PTEN may establish two membrane interaction orientations (1) a nonspecific electrostatics driven orientation that can lead to accumulation of PI(4,5)P2 at the putative PI(4,5)P2 binding domain (PBD), 2) a substrate driven orientation minimizing the distance between the PTEN active site and the membrane surface, but with a significantly decreased nonspecific electrostatic affinity. This latter orientation should increase the probability of PTEN dissociation post-catalysis, since the specific binding free energy of the 3'-phosphate will be lost. To begin testing the hypotheses set forth by this model, 31P spectra of PI(4,5)P2 and PI(3,4,5)P3 will be acquired to characterize the phosphodiester and monoester resonances and to determine the headgroup orientations by chemical shift tensor analyses. These preliminary lipid-only studies are necessary for analyzing 31P chemical shift correlations with of 13C-15N-labeled PTEN bound to vesicles containing PI(4,5)P2. Next, PTEN expression in E. coli will be optimized to obtain sufficient quantities of labeled protein for NMR analyses. Using a specialized labeling pattern to reduce undesired signals, 2D and 3D correlation spectra will be collected to determine the N-terminal secondary structure by comparing assigned resonances to the TALOS database. The PTEN N-terminus was unresolved in a crystal structure and participates in a PI(4,5)P2 specific binding site found to be essential for PTEN recruitment to the PM. Thus, the PTEN-PI(4,5)P2 interaction will be interrogated by SSNMR to identify the PTEN residues that bind PI(4,5)P2. Each of the above findings will be integrated into a new computational model of PTEN docked at anionic bilayers to improve the previous theoretical characterization of electrostatics driven association. Through the execution of the proposed Research Training Plan, I will gain the expertise and complete the preliminary groundwork required for formally testing the two orientations hypothesis.
Phosphoinositides are specialized lipids that operate analogously to traffic signals: they direct the proteinaceous cellular machinery that drives the cellular processes fundamental to life. Approximately 30 phosphoinositide-modifying proteins, including the tumor suppressor PTEN, maintain the proper balance of phosphoinositides. Failure of these proteins to keep phosphoinositide levels in check is the basis for numerous human diseases and dysfunctions including cancer, bipolar disorder, infertility, diabetes and Down syndrome. The proposal outlined herein seeks to observe and define the atomistic characteristics of phosphoinositides that confer their potent signaling abilities and to define the atomistic interactions of PTEN with phosphoinositides important for PTEN's anti-cancer properties.
