Structures reveal a key mechanism of WAVE regulatory complex activation by Rac1 GTPase

Abstract

The Rho-family GTPase Rac1 activates the WAVE regulatory complex (WRC) to drive Arp2/3 complex-mediated actin polymerization in many essential processes. Rac1 binds to WRC at two distinct sites-the A and D sites. Precisely how Rac1 binds and how the binding triggers WRC activation remain unknown. Here we report WRC structures by itself, and when bound to single or double Rac1 molecules, at ~3 Å resolutions by cryogenic-electron microscopy. The structures reveal that Rac1 binds to the two sites by distinct mechanisms, and binding to the A site, but not the D site, drives WRC activation. Activation involves a series of unique conformational changes leading to the release of sequestered WCA (WH2-central-acidic) polypeptide, which stimulates the Arp2/3 complex to polymerize actin. Together with biochemical and cellular analyses, the structures provide a novel mechanistic understanding of how the Rac1-WRC-Arp2/3-actin signaling axis is regulated in diverse biological processes and diseases.

Fig. 1. Cryo-EM structures of the WRC in different Rac1-bound states.

a–c Schematic and cryo-EM density of the indicated WRCs. Black dotted lines indicated by arrowheads are flexible peptide linkers tethering Rac1 to the WRC. The meander sequence is traced by white dotted lines. Other dotted lines and cylinders in the WRC^AD-Rac1 structure refer to sequences of which the densities are not observed in WRC^AD-Rac1, but present in WRC^apo and WRC^D-Rac1. Red dots indicate locations to which Rac1 is tethered. d–f Pyrene-actin polymerization assays measuring activities of the indicated WRCs used in cryo-EM. Reactions use the NMEH20GD buffer (see Methods) and contain 3.5 µM actin (5% pyrene-labeled), 10 nM Arp2/3 complex, 100 nM WRC230WCA or WAVE1 WCA, and/or 6 µM Rac1^QP. A. U. for arbitrary units. Results are representative of at least two independent repeats. Source data for d–f are provided as a Source Data file.

Fig. 2. Interactions mediating Rac1 binding to the D site.

a Side view of the overall structure of Rac1 (cartoon, gold) binding to the D site (semitransparent surface, green). F37 and Y967 side chains are shown as reference points. b Top view and semitransparent surface charge representation of the D site (calculated using APBS in Pymol), showing key interactions between Sra1 and Rac1. Yellow dotted lines indicate polar interactions. White dashed line indicates binding surface boundary. For clarity, the backbones of Switch I and II are shown as loops. c, d Structural comparison of Rac1 and the D site in the bound (dark colors) and unbound (light colors, PDB 3SBD for Rac1) states. Curved arrows indicate side chain flipping upon Rac1 binding. Straight arrow indicates translation of polypeptide backbone. Black dashed lines indicate the interdigitated π-π stacking.

Fig. 3. Interactions mediating Rac1 binding to the A site.

a Side view of the overall structure of Rac1 binding to the A site, using the same color scheme as in Fig. 2. F37 and R190 side chains are shown as reference points. b Top view and semitransparent surface charge representation of the A site, showing key interactions between Sra1 and Rac1. Yellow dotted lines indicate polar interactions. White dashed line indicates binding site boundary. For clarity, the backbone of Rac1 Switch I—β2—β3—Switch II sequence mediating the binding is shown as loops. Dots of different colors indicate residues of which mutations were involved in human disease (red), previously designed and shown to disrupt Rac1 binding (yellow), or newly introduced in this work (blue). c Semitransparent surface representation of the Rac1 surface, showing how Sra1^R190 fits into a deep pocket in Rac1 and how it is supported by interactions surrounding the rim of the pocket. d Pyrene-actin polymerization assays measuring the activities of WRCs carrying indicated mutations at the A site. Reactions use the NMEH20GD buffer (see Methods) and contain 3.5 µM actin (5% pyrene-labeled), 10 nM Arp2/3 complex, 100 nM WRC230WCA or WAVE1 WCA, and/or indicated amounts of Rac1^QP. e Representative fluorescence images and quantification of lamellipodia formation in B16-F1 Sra1/Cyfip2 double KO#3 cells transfected with indicated EGFP-Sra1 variants and stained by phalloidin for F-actin. Statistical significance was assessed from three repeats for differences between cells transfected with WT (wild type) vs. no (-) or indicated mutant constructs concerning cell percentages displaying “no lamellipodia” phenotype (*p < 0.05; ***p < 0.001) and with “lamellipodia” phenotype (+++p < 0.001) using one-way ANOVA with Dunnett’s post hoc test correcting for multiple comparisons. n.s.: not statistically significant. Error bars represent standard errors of means. Cell numbers used for the quantification are shown on top of each column. Exact p-values (WT vs. X) are No lamellipodia: no (-): <0.0001, Y108H: 0.4733, Y108A: 0.8198, N176W: 0.0240, C179R < 0.0001, N183R < 0.0001, S186M < 0.0001, K189M < 0.0001; and Lamellipodia: no (-): <0.0001, Y108H: 0.9706, Y108A: 0.9995, N176W < 0.0001, C179R < 0.0001, N183R < 0.0001, S186M < 0.0001, K189M < 0.0001. Source data for d, e are provided as a Source Data file.

Fig. 4. Rac1 binding to the A site drives a conformational change to release WCA.

a, b Overlay of WRC^AD-Rac1 (dark colors) and WRC^D-Rac1 (light colors) structures showing conformational changes of the A site upon Rac1 binding. Sequences of which the densities are not observed in WRC^AD-Rac1 structure, but are present in WRC^D-Rac1 are indicated by magenta dashed lines and cylinders. The meander region in WAVE1 is traced by the black dotted line. c, d Comparison of the tyrosine lock and “stem” (traced by the black dotted line) region before and after Rac1 binding to the A site. Structures in light colors are from the unbound state in WRC^D-Rac1 and used as reference point for the A-site bound state. Residues critical for stabilizing the tyrosine lock and “stem” components are labeled and shown in sticks. Dots of different colors indicate residues of which mutations were involved in human disease (red), previously designed and shown to disrupt WRC inhibition (yellow), or newly introduced in this work (blue).

Fig. 5. An allosteric competition model explains WRC activation by Rac1 binding to the A site.

a, b Coomassie blue-stained SDS-PAGE gels showing GST-Rac1^P29S (loaded with GDP or GMPPNP) pull-down of WRC^D-Rac1 bearing the indicated mutations: WAVE1^∆PPP in (A) (replacing ¹³¹PPPLNI¹³⁶ with a GSGSGS linker) and WAVE1^Y151E or Sra1^R87C in (B). c Pyrene-actin polymerization assays measuring the activities of WRC^D-Rac1 used in (A-B). Reactions use the NMEH20GD buffer and contain 3.5 µM actin (5% pyrene-labeled), 10 nM Arp2/3 complex, and 100 nM WRC^D-Rac1 carrying indicated mutations. d A “door wedge” model describing the allosteric competition mechanism underlying WRC activation by Rac1 binding to the A site. WRC activation requires Rac1 binding to the A site to swing the door (A site in Sra1 and α4-loop-α5 in WAVE1) against the wedge (Y151 and the “stem” sequence) inserted into the door hinge. The “tug-of-war” between Rac1 binding and the tyrosine lock determines the activity level of the WRC. Phosphorylation (blue dot) of the released Y151 further shifts the equilibrium to provide an additional control of the strength and duration of WRC activation. Note that for clarity, majority of Sra1 and other WRC components not directly involved in activation are not shown in the cartoon. Source data for a–c are provided as a Source Data file.