Genetic studies in Caenorhabditis elegans identified an evolutionarily conserved CED-2 (CrkII), CED-5 (DOCK180), CED-12 (ELMO), CED-10 (Rac1) module important for cell migration and phagocytosis of apoptotic cells. Previous studies have shown that DOCK180 and ELMO comprise an unconventional bipartite Dbl homology domain-independent Rac guanine nucleotide exchange factor (Rac-GEF); but it is still unclear how CrkII functions in Rac-GEF activity. In this study, we have characterized a unique function of CrkII in phagocytosis and Rac activation mediated by the C-terminal SH3 domain, a region of CrkII that has no clear cellular or biochemical function. We found that mutations that disrupt the C-terminal SH3 domain of CrkII (CrkII-SH3-C) abrogate engulfment of apoptotic cells and impair cell spreading on extracellular matrix. Surprisingly, despite the effects on engulfment, W276K CrkII strongly potentiated Rac-GTP loading when ectopically expressed in HEK 293T cells. Contrary to the effects of the true dominant negative SH2 domain mutants (R38K CrkII) and SH3-N domain mutants (W170K CrkII) that prevent macromolecular assembly of signaling proteins, W276K CrkII increases association between DOCK180 and CrkII as well as constitutive tethering of the Crk/DOCK180/ELMO protein complex that interacted with RhoG. Our results indicate that while N-terminal SH3 of CrkII promotes assembly between CrkII and DOCK180, the C-terminal SH3 of CrkII regulates the stability and turnover of the DOCK180/ELMO complex. Studies with W276K CrkII may offer a unique opportunity to study the structure and function of the DOCK180/ELMO Rac-GEF.

Crk II and Crk L have both cytosolic and nuclear functions. While Crk L is a bona fide nuclear signaling protein because of its ability to bind tyrosine-phosphorylated STAT5 and act as a transcriptional coactivator, the function of nuclear Crk II is less well understood. The present study was undertaken to investigate whether Crk II is in the nucleus, how Crk II translocates into the nucleus, whether it possesses a functional NES, and to determine if nuclear Crk II affects cell cycle checkpoints and promotes apoptosis. Toward this goal, we used several independent techniques to show that a significant percentage of the total endogenous Crk II partitions in the nucleus in mammalian cells, where it forms distinct complexes with DOCK180, Wee1, and Abl. We found no evidence that Crk II bound to Crm1 nor that the localization of GFP-Crk II was sensitive to LMB, an inhibitor of Crm1. To better define the significance of nuclear Crk II localization, we generated a GFP-Crk II protein (GFP-Crk-nuc) fused to three tandem nuclear localization signals derived from the SV40 large T-antigen. GFP-Crk-nuc exhibited exclusive nuclear localization, and in contrast to wild-type Crk, GFP-Crk-nuc expressing cells could not be propagated upon selection in G418-containing media, suggesting nuclear accumulation of Crk II caused either growth arrest or apoptosis. When transiently transfected cells were FACS sorted, GFP-expressing cells showed defective cell adhesion on tissue culture surfaces and showed an increased level of apoptosis assessed by pycnotic nuclei, annexin V staining, and PARP cleavage. Although we found that Crk II bound to the cell cycle protein Wee1, expression of GFP-Crk-nuc did not induce a G2/M cell cycle block or cause increased Cdc2 Tyr15 phosphorylation. Finally, upon UV stimulation, we found that endogenous Crk II translocated to the nucleus and potentiated the extent of UV-inducible apoptosis after 4 h. These data suggest that nuclear compartmentalization of Crk II antagonizes its cytoskeletal functions and assign a proapoptotic role to the nuclear pool of Crk II.

No abstract