Fig. 6. ELMO1^NTD auto-inhibits DOCK2–ELMO1.

a, b GEF activity data. ELMO1 stimulates DOCK2 GEF activity. Deleting ELMO1^NTD stimulates DOCK2–ELMO1 GEF activity 60%. Disrupting the DOCK2 dimerization reduces GEF activity by ~60%. a Experimental data for RAC1 at 0.8 μM. There was no observable spontaneous GDP exchange without the GEF. GEF activity data was monitored by the exchange of fluorescent mant-GTP following an injection of GDP after 1 min (indicated by arrow). b Rates as a function of RAC1 concentration. The initial rates of exchange at increasing substrate (mant-RAC) concentrations were fitted to a Michaelis-Menten equation with the resultant constants shown in Table 1. Data are presented as mean values with error bars of +/− one standard deviation. The experiment shown in (a) and (b) was performed six times. c The crystal structure of the RHOG–ELMO2^RBD complex was superimposed onto the modelled ELMO1^RBD of DOCK2–ELMO1–RAC1 (shown in surface representation with ELMO1^NTD in the open-conformation). d The crystal structure of the RHOG–ELMO2^RBD complex was superimposed onto ELMO1^RBD of DOCK2–ELMO1 (shown in surface representation with ELMO1^NTD in closed-conformation). This shows that RHOG-binding site on ELMO1^RBD is exposed in DOCK2–ELMO1–RAC1, whereas in the binary DOCK2–ELMO complex, with ELMO1 in the down conformation, the RHOG-binding site on ELMO1^RBD is occluded by DOCK2^DHR2. Modelling the ELMO2^NTD–BAI1 complex onto ELMO1^NTD of the DOCK2−ELMO1 binary complex (closed-conformation) shows that BAI1 bound to the ELMO1^EID domain would clash with ELMO1^PH (d), but not in the ternary DOCK2–ELMO1–RAC1 complex (c). Thus, RhoG and BAI1 would only bind to the activated conformation of the binary DOCK2−ELMO1 complex with the ELMO1^NTD in the open-conformation with the DOCK2^DHR2−RAC1 binding site exposed. Source data are provided as a Source Data file.