Solution Structure, Backbone Dynamics, and Interaction with Cdc42 of Salmonella Guanine Nucleotide Exchange Factor SopE2 †,‡ Christopher Williams, §,| Edouard E. Galyov, ⊥ and Stefan Bagby* ,§ Department of Biology and Biochemistry, UniVersity of Bath, Bath BA2 7AY, U.K., and DiVision of EnVironmental Microbiology, Institute for Animal Health, Compton Laboratory, Berkshire RG20 7NN, U.K. ReceiVed May 7, 2004; ReVised Manuscript ReceiVed July 13, 2004 ABSTRACT: SopE and SopE2 are delivered by the Salmonella type III secretion system into eukaryotic cells to promote cell invasion. SopE and SopE2 are potent guanine nucleotide exchange factors (GEFs) for Rho GTPases Cdc42 and Rac1 and constitute a novel class of Rho GEFs. Although the sequence of SopE-like GEFs is not at all homologous to those of the Dbl homology domain-containing eukaryotic GEFs, the mechanism of nucleotide release seems to have significant similarities. We have determined the solution structure of the catalytic domain (residues 69-240) of SopE2, showing that SopE2 69-240 comprises two three-helix bundles (R1R4R5 and R2R3R6) arranged in a Λ shape. Compared to the crystal structure of SopE 78-240 in complex with Cdc42, SopE2 69-240 exhibits a less open Λ shape due to movement of SopE 78-240 helices R2 and R5 to accommodate binding to the Cdc42 switch regions. In an NMR titration to investigate the SopE2 69-240 -Cdc42 interaction, the SopE2 69-240 residues affected by binding Cdc42 were very similar to the SopE 78-240 residues that contact Cdc42 in the SopE 78-240 -Cdc42 complex. Analysis of the backbone 15 N dynamics of SopE2 69-240 revealed flexibility in residues that link the two three-helix bundles, including the R3-R4 linker that incorporates a -hairpin and the catalytic loop, and the R5-R6 loop, and flexibility in residues involved in interaction with Cdc42. Together, these observations provide experimental evidence of a previously proposed mechanism of GEF-mediated nucleotide exchange based on the Rac1-Tiam1 complex structure, with SopE/E2 flexibility, particularly in the interbundle loops, enabling conformational rearrangements of the nucleotide binding region of Cdc42 through an induced fit type of binding. Such flexibility in SopE/E2 may also facilitate interaction through adaptive binding with alternative target proteins such as Rab5, allograft inflammatory factor 1, and apolipoprotein A-1. Salmonella enterica are facultative intracellular enteric pathogens that cause a broad range of diseases in a variety of vertebrates. A crucial component in host cell invasion by Salmonella is the specialized type III secretion system (TTSS), 1 a common mechanism by which Gram-negative pathogenic and symbiotic animal and plant bacteria deliver a battery of effector proteins into host cells (1). Among other effects, Salmonella virulence proteins delivered by this system cause pronounced membrane ruffling and actin cytoskeleton rearrangements at the point of contact between the bacterium and host cell, leading to bacterial internaliza- tion. Salmonella invasion of host cells is dependent on host cell Rho GTPases Cdc42 and Rac1 (2). Rho GTPases, a family of monomeric GTP-binding proteins, are key regulators for a wide range of cellular responses, including cytoskeletal reorganization, focal adhe- sions, motility, membrane ruffling, cytokinesis, cell aggrega- tion, cell-cell adhesion, gene expression, mitogenic signal- ing, and malignant transformation (3-6). Rho GTPases act as molecular switches, cycling between GDP-bound (inac- tive) and GTP-bound (active) conformations. All GTPases have a similar core fold consisting of a mixed six-stranded -sheet and five R-helices located on either side. Two highly flexible regions, termed switch I and switch II, exhibit the major conformational differences between GDP-bound and GTP-bound forms and define the on and off states of GTPases. The conserved P-loop (phosphate binding loop) in the binding pocket is involved in stabilizing the active conformation. Mg 2+ is essential for GTPase function and structure, contributing to the tight binding of the nucleotides † This work was supported at the University of Bath by The Wellcome Trust (Grant 060998) and at the Institute for Animal Health by BBSRC. The Wellcome Trust is acknowledged for purchase of the 600 MHz NMR spectrometer (Grant 051902) used in this study. C.W. was supported by a Ph.D. studentship from EPSRC. ‡ Coordinates of the 20 lowest-energy structures and the structure closest to the mean have been deposited in the Protein Data Bank as entries 1R6E and 1R9K, respectively. * To whom correspondence should be addressed: Dr. Stefan Bagby, Department of Biology & Biochemistry, University of Bath, Bath BA2 7AY, UK. Tel: +44 (0)1225 386436. Fax: +44 (0)1225 386779. Email: bsssb@bath.ac.uk. § University of Bath. | Present address: School of Chemistry, University of Bristol, Cantock’s Close, Bristol BS8 1TS, U.K. ⊥ Compton Laboratory. 1 Abbreviations: DH, Dbl homology; DTT, dithiothreitol; EDTA, ethylenediaminetetraacetic acid; GdnHCl, guanidine hydrochloride; GEF, guanine nucleotide exchange factor; GST, glutathione S-trans- ferase; HEPES, 4-(2-hydroxyethyl)piperazine-1-ethanesulfonic acid; HSQC, heteronuclear single-quantum coherence; IPTG, isopropyl -D- thiogalactoside; NMR, nuclear magnetic resonance; NOE, nuclear Overhauser effect; PH, pleckstrin homology; rmsd, root-mean-square deviation; SCV, Salmonella-containing vesicle; Tris, tris(hydroxymeth- yl)aminomethane; TTS, type III secretion; TTSS, type III secretion system. 11998 Biochemistry 2004, 43, 11998-12008 10.1021/bi0490744 CCC: $27.50 © 2004 American Chemical Society Published on Web 08/28/2004