RPTPα-mediated activation of Src

One of the main signal transduction mechanisms in all eukaryotic organisms is tyrosine phosphorylation. The cellular levels of tyrosine phosphorylation are tightly controlled by the activity of two classes of enzymes with opposing activities: the protein-tyrosine kinases (PTKs) and the protein-tyrosine phosphatases (PTPs). Malfunction of these enzymes results in changes of the delicate tyrosine phosphorylation balance and leads to disease. We are interested in elucidating RPTP regulation and we use RPTPalpha as a proto-typical RPTP. The best characterized substrate of RPTPalpha is Src and dephosphorylation of Src leads to Src activation. We studied mitotic RPTPalpha phosphorylation using phospho-specific antibodies. We discovered that RPTPalpha phospho-Ser204 was almost completely dephosphorylated in mitotic NIH3T3 and HeLa cells, whereas phosphorylation of Ser180 was only slightly affected. Concomitantly, Src pTyr527 and pTyr416 were dephosphorylated, resulting in modest Src activation. Using inhibitors, we found that dephosphorylation of RPTPalpha phospho-Ser204 in mitosis is likely mediated by PP2A. Interestingly, binding of endogenous Src to RPTPalpha was induced in mitosis. Based on these results, we proposed a new model for mitotic activation of Src in which RPTPalpha pSer204 dephosphorylation facilitates Src binding, leading to RPTPalpha-mediated dephosphorylation of Src pTyr527 and pTyr416 and hence activating Src in a tightly controlled manner. Next, we investigated how dephosphorylation of RPTPalpha phospho-Ser204 could lead to increased Src binding. In the vicinity of Ser204, RPTPalpha has two domains that might have a role: a putative SH3 binding site and a helix-turn-helix wedge motif. P210/211L mutations in RPTPalpha abolish the putative PXXP SH3 binding site and concomitantly affect the stability of the wedge. We discovered that Src binding to RPTPalpha-P210/211L is impaired and RPTPalpha-P210/211L mutant has decreased ability to activate Src. Mutations of the Src-SH3 domain (W118A or P133L) increased binding to RPTPalpha indicating that the Src-SH3 domain is not required for binding. Further, different RPTPalpha mutations affecting the wedge structure suggested that the wedge is required for the interaction with Src. In a screen for disease-related RPTPalpha mutations we found a mutation in RPTPalpha consisting of an Arg to His substitution in the phospho-tyrosine recognition loop of the membrane-distal catalytic domain (D2) that turned out not to be associated with the disease. RPTPalpha-R554H mutation depletes the catalytic activity of RPTPalpha-D2 without significantly affecting the total activity of RPTPalpha. In addition, R554H mutation impaired Src binding to RPTPalpha. RPTPalpha lacking the catalytic cysteine (Cys723) from the D2 domain had also diminished ability to bind Src. Decreased Src binding to R554H and C723S RPTPalpha mutants resulted in low Src activation. These results suggested that a catalytically active membrane-distal domain is required for a fully functional RPTPalpha that binds and activates Src. Finally, we used three different approaches to identify new RPTPalpha substrates: (1) study of the tyrosine phosphorylation profiles of wild type and RPTPalpha knock-out mouse embryo fibroblasts, (2) substrate trapping experiments and (3) quantitative mass spectrometry. We discovered a number of interesting candidate substrates that remain to be confirmed. The results described in this thesis led us to propose a new model for activation of Src by RPTPalpha.