Mutant Forms of PTEN Associated with Autism

Background: The phosphatase and tensin homolog deleted on chromosome 10 (PTEN) was implicated for a subset of autistic patients with macrocephaly. Six independent studies have identified patients in this subclass that have one normal and one mutated PTEN gene. Because loss of both PTEN genes is embryonic lethal, we do not anticipate any patients with both PTEN genes mutated. The PTEN protein is a phosphatidylinositol phosphate phosphatase specific for the 3-position of the inositol ring. PTEN and phosphoinositide 3-kinase have opposing effects on PI(3,4,5)P3 and, consequently, on cell proliferation and survival. PTEN's structure consists of a short N-terminal PI(4,5)P2 –binding domain, a phosphatase domain rich in alpha-helix, a C2 domain dominated by beta-sheet and a C-terminal tail with several phosphorylation sites. Recently, we and others have proposed a model for regulation of PTEN by PI(4,5)P2. Efficient action of PTEN requires three steps. First, PTEN binds to PI(4,5)P2 in the membrane. Second, PTEN undergoes a conformational change that increases alpha-helical content and activates the phosphatase domain. Third, PTEN diffuses on the surface of the membrane and hydrolyzes PI(3,4,5)P3. Objectives: The autism-associated PTEN mutations might have a number of different effects, including null mutations with complete loss of PTEN phosphatase activity hypomorph mutations with reduced activity and gain of function in which the mutation results in a novel function. A first step in determining the effects of these mutations is to express the mutated PTEN proteins and determine if they are active as a lipid phosphatase. Methods: Recombinant PTEN was expressed in bacteria. Binding of PTEN protein to lipid vesicles was measured by doping the vesicles with a dansyl-lipid. Fluorescence from the PTEN tryptophans is quenched by dansyl. As a result, the fluorescence from vesicle-bound PTEN is quenched. PTEN protein is titrated with the dansyl-doped vesicles, allowing calculation of binding constants. Conformational changes of PTEN were detected by infrared spectroscopy. The spectra shift depending on the content of alpha helix and beta sheet. Phosphatase activity was measured by a colorimetric assay in which free phosphate results in a green color. Results: The laboratories that discovered the autism-associated PTEN mutations did not characterize the resulting PTEN proteins. In this study, we characterized the histidine 93 to arginine (H93R) mutation, which is near the phosphatase active site. We found that the H93R mutation affects multiple steps of PTEN action. First, the mutation affects binding of PTEN to membranes bearing negatively charged lipids. Binding of H93R PTEN to phosphatidylserine-bearing membranes was more avid than that of wild type PTEN. Second, the phosphatase activity was decreased for H93R PTEN. Third, H93R PTEN showed increased localization in the nuclei of U87MG glioblastoma cells. Conclusions: The autism-associated H93R mutation of PTEN reduces but does not eliminate phosphatase activity. Furthermore, the mutation increases localization of PTEN in the nucleus. Further studies are needed to determine the consesquences of this increased nuclear localization.