CD46 is phosphorylated at tyrosine 354 upon infection of epithelial cells by Neisseria gonorrhoeae

The Neisseria type IV pilus promotes bacterial adhesion to host cells. The pilus binds CD46, a complement-regulatory glycoprotein present on nucleated human cells (Källström et al., 1997). CD46 mutants with truncated cytoplasmic tails fail to support bacterial adhesion (Källström et al., 2001), suggesting that this region of the molecule also plays an important role in infection. Here, we report that infection of human epithelial cells by piliated Neisseria gonorrhoeae (GC) leads to rapid tyrosine phosphorylation of CD46. Studies with wild-type and mutant tail fusion constructs demonstrate that Src kinase phosphorylates tyrosine 354 in the Cyt2 isoform of the CD46 cytoplasmic tail. Consistent with these findings, infection studies show that PP2, a specific Src family kinase inhibitor, but not PP3, an inactive variant of this drug, reduces the ability of epithelial cells to support bacterial adhesion. Several lines of evidence point to the role of c-Yes, a member of the Src family of nonreceptor tyrosine kinases, in CD46 phosphorylation. GC infection causes c-Yes to aggregate in the host cell cortex beneath adherent bacteria, increases binding of c-Yes to CD46, and stimulates c-Yes kinase activity. Finally, c-Yes immunoprecipitated from epithelial cells is able to phosphorylate the wild-type Cyt2 tail but not the mutant derivative in which tyrosine 354 has been substituted with alanine. We conclude that GC infection leads to rapid tyrosine phosphorylation of the CD46 Cyt2 tail and that the Src kinase c-Yes is involved in this reaction. Together, the findings reported here and elsewhere strongly suggest that pilus binding to CD46 is not a simple static process. Rather, they support a model in which pilus interaction with CD46 promotes signaling cascades important for Neisseria infectivity.     

Crystallographic structure of the SH3 domain of the human c-Yes tyrosine kinase: loop flexibility and amyloid aggregation

SH3 domains from the Src family of tyrosine kinases represent an interesting example of the delicate balance between promiscuity and specificity characteristic of proline-rich ligand recognition by SH3 domains. The development of inhibitors of therapeutic potential requires a good understanding of the molecular determinants of binding affinity and specificity and relies on the availability of high quality structural information. Here, we present the first high-resolution crystal structure of the SH3 domain of the c-Yes oncogen. Comparison with other SH3 domains from the Src family revealed significant deviations in the loop regions. In particular, the n-Src loop, highly flexible and partially disordered, is stabilized in an unusual conformation by the establishment of several intramolecular hydrogen bonds. Additionally, we present here the first report of amyloid aggregation by an SH3 domain from the Src family.     

Differential mitotic activation of endogenous c-Src, c-Yes, and Lyn in HeLa cells

Src-family tyrosine kinases (SFKs) play an important role in mitosis. Despite overlapping expression of multiple SFK members, little is known about how individual SFK members are activated in M phase. Here, we examined mitotic activation of endogenous c-Src, c-Yes, and Lyn, which are co-expressed in HeLa cells. c-Src, c-Yes, and Lyn were activated at different levels in M phase, and the activation was inhibited by Cdc2 inactivation. Mitotic c-Src and c-Yes exhibited normal- and retarded-electrophoretic-mobility forms on SDS-polyacrylamide gels, whereas Lyn did not show mobility retardation. Like c-Src, the retardation of electrophoretic mobility of c-Yes was caused by Cdc2-mediated phosphorylation. The normal- and retarded-mobility forms of c-Src were comparably activated, but activation of the retarded-mobility form of c-Yes was higher than that of the normal-mobility form of c-Yes. Thus, these results suggest that endogenous c-Src, c-Yes, and Lyn are differentially activated through Cdc2 activation during M phase.     

Involvement of the SH3 domain in Ca2+-mediated regulation of Src family kinases

When cells are treated with Ca(2+) and Ca(2+)-ionophore, c-Src kinase activity increases, whereas c-Yes kinase activity decreases. This opposite modulation can be reproduced in an in vitro reconstitution assay and is dependent on Ca(2+) and on soluble factors present in cell lysates. Since c-Src and c-Yes share a high degree of homology, with the exception of their N-terminal "unique" domains, their activity was thought to be coordinately regulated. To assess the mechanism of regulation we generated stable cell lines expressing eight different constructs containing wild type c-Src and c-Yes, as well as swaps of the unique domain alone, unique and Src homology 3 (SH3) domains together and the SH3 domain alone. Swapping of the unique domains was not sufficient to reverse the regulation of the chimeric molecules. On the other hand, chimeras containing swaps of the unique plus the SH3 domains displayed reverse regulation, implicating both domains in the regulation of kinase activity by Ca(2+). To rule out the participation of the unique domain, we used chimeric molecules with swapped SH3 domains only and found that the SH3 domain is necessary and sufficient to confer Ca(2+)-mediated regulation of Src and Yes tyrosine kinases.     

Identification of type I and type II inhibitors of c-Yes kinase using in silico and experimental techniques

c-Yes kinase is considered as one of the attractive targets for anti-cancer drug design. The DFG (Asp-Phe-Gly) motif present in most of the kinases will adopt active and inactive conformations, known as DFG-in and DFG-out and their inhibitors are classified into type I and type II, respectively. In the present study, two screening protocols were followed for identification of c-Yes kinase inhibitors. (i) Structure-based virtual screening (SBVS) and (ii) Structure-based (SB) and Pharmacophore-based (PB) tandem screening. In SBVS, the c-Yes kinase structure was obtained from homology modeling and seven ensembles with different active site scaffolds through molecular dynamics (MD) simulations. For SB-PB tandem screening, we modeled ligand bound active and inactive conformations. Physicochemical properties of inhibitors of Src kinase family and c-Yes kinase were used to prepare target focused libraries for screenings. Our screening procedure along with docking showed 520 probable hits in SBVS and tandem screening (120 and 400, respectively). Out of 5000 compounds identified from different computational methods, 2410 were examined using kinase inhibition assays. It includes 266 compounds (5.32%) identified from our method. We observed that 14 compounds (12%) are identified by the present method out of 168 that showed > 30% inhibition. Among them, three compounds are novel, unique, and showed good inhibition. Further, we have studied the binding of these compounds at the DFG-in and DFG-out conformations and reported the probable class (type I or type II). Hence, we suggest that these compounds could be novel drug leads for regulation of colorectal cancer.