The biochemical mechanism of Rho GTPase membrane binding, activation and retention in activity patterning

Abstract

Rho GTPases form plasma membrane-associated patterns that control the cytoskeleton during cell division, morphogenesis, migration, and wound repair. Their patterning involves transitions between inactive cytosolic and active membrane-bound states, regulated by guanine nucleotide exchange factors (GEFs), GTPase-activating proteins (GAPs), and guanine nucleotide dissociation inhibitors (GDIs). However, the relationships between these transitions and role of different regulators remain unclear. We developed a novel reconstitution approach to study Rho GTPase patterning with all major GTPase regulators in a biochemically defined system. We show that Rho GTPase dissociation from RhoGDI is rate-limiting for its membrane association. Rho GTPase activation occurs after membrane insertion, which is unaffected by GEF activity. Once activated, Rho GTPases are retained at the membrane through effector interactions, essential for their enrichment at activation sites. Thus, high cytosolic levels of RhoGDI-bound GTPases ensure a constant supply of inactive GTPases for the membrane, where GEF-mediated activation and effector binding stabilize them. These results delineate the route by which Rho GTPase patterns are established and define stage-dependent roles of its regulators.

Keywords: Cell Polarity; Membrane Signaling; Rho GTPases; Self-organization; Small G Proteins.

Figure 1. Quantifying patterning of total and active Cdc42 in vivo.

(A) Scheme of laser-induced wounding of Xenopus laevis oocyte and enrichment of wGBD and IT-Cdc42. (B) Representative micrograph of wGBD (mCh-wGBD) and IT-Cdc42 (IT-GFP-Cdc42) at the plasma membrane before wounding (top, Pre-wound), after wounding prior to zone formation (middle, Pre-zone), and after wounding when the zones are fully formed (bottom, Zone). Scale bar 20 µm. (C) Kymograph of the micrograph from (B) generated by radially averaging signal intensity around the wound over time (Moe et al, ; see “Methods”). The arrow denotes when the wound occurred, and W denotes the wound location. The yellow line “P” indicates the Pre-zone time point and the yellow line “Z” indicates the Zone time point. (D) Line scan generated from the kymograph in (C) to show IT-Cdc42 and wGBD fluorescence intensity at the Pre-zone time point as a function of distance from the wound center. (E) Line scan generated from the kymograph in (C) to show IT-Cdc42 and wGBD fluorescence intensity at the Zone time point as a function of distance from the wound center. (F) Patterning indices of IT-Cdc42 and wGBD at the Pre-zone time point compared to the Zone time point. Two-sample t tests were used to determine statistical significance between Pre-zone and Zone time points; (mean ± SD). wGDB P value = 4.20128E-09, IT-Cdc42 P value = 1.0572E-08. wGBD and IT-Cdc42 imaged in the same cells, n = 12 individual cells in N = 1 experiment. (G) Timing of IT-Cdc42 and wGBD enrichment around the wound. Mean (Shaded area = SD), n = 9 individual cells in N = 1 experiment per condition.

Figure 2. Rho GTPase activation from soluble RhoGDI complexes by membrane-bound GEFs in vitro kinetically coincides with GDI dissociation.

(A) Scheme of the FRET-based nucleotide exchange assays from Cy5-GDP:Cy3-Cdc42:RhoGDI complexes in the presence of ITSN_cat-coated LUVs. (B) Decrease of the acceptor/donor fluorescence intensity ratio as a function of time. At t = 0 s, Cy5-GDP:Cy3-Cdc42:RhoGDI1 or Cy5-GDP:Cy3-Cdc42 complexes (100 nM) were mixed with either ITSN_cat-His₁₀ (10 nM), either in solution or pre-bound to PM LUVs (125 µM total lipids containing 0.2% Ni²⁺-NTA-DGS, or PM LUVs alone as indicated in the presence of excess unlabeled GDP. (C) At t = 0 s, Cy5-GDP:Cy3-Cdc42:RhoGDI1 (100 nM) were mixed in the presence of excess unlabeled GDP with either ITSN_cat-His₁₀ at concentrations as indicated, which was pre-bound to PM LUVs (125 µM) or PM LUVs alone. (D) Scheme of the FRET-based nucleotide exchange assays from Cy5-GDP:Cy3-Rac1:RhoGDI complexes in the presence of DOCK1:ELMO1-coated LUVs. (E) Time traces of the acceptor/donor fluorescence intensity ratio as a function of time. At t = 0 s, Cy5-GDP:Cy3-Rac1:RhoGDI1 or Cy5-GDP:Cy3-Rac1 complexes (100 nM) were mixed with either DOCK1:ELMO1 (10 nM), either in solution or pre-bound to PM LUVs (125 µM total lipids, 4% PtdIns(3,4,5)P₃), or PM LUVs alone as indicated in the presence of excess unlabeled GDP. (F) At t = 0 s, Cy5-GDP:Cy3-Cdc42:RhoGDI1 or Cy5-GDP:Cy3-Cdc42 complexes (100 nM) were mixed in the presence of excess unlabeled GDP with either DOCK1:ELMO1 complexes at concentrations as indicated that were pre-bound to PM LUVs (125 µM total lipids, 4% PtdIns(3,4,5)P₃), or PM LUVs alone. (G) Scheme of the FRET-based GDI dissociation assay. (H) Decrease of the normalized acceptor/donor fluorescence intensity ratio as a function of time (brown). At t = 0 s Cy3-Cdc42:A674-RhoGDI1 complexes (80 nM) are mixed with excess unlabeled RhoGDI1 (5 µM). Time courses of nucleotide exchange at saturating GEF concentrations either in the presence of ITSN_cat-His₁₀ (violet) or DOCK1:ELMO1 (green) as observed in (C) or (F) are shown for comparison. (I) Maximal rates of nucleotide exchange by ITSN_cat-His₁₀ (violet) or DOCK1:ELMO1 (green) compared to dissociation of RhoGDI1 (brown) obtained from mono-exponential fits to the data in (H). All traces shown represent the mean from three independent experiments (N = 3).

Figure 3. RhoGDI dissociation precedes Rho GTPase membrane binding in vitro.

(A) Scheme of neutravidin-mediated recruitment of biotinylated, dual-labeled Cdc42:RhoGDI1 complexes to functionalized surfaces. (B) TIRFM images of Biotin-Cy3-Cdc42 (left) A647-RhoGDI1 (center) complexes (100 pM, ≤2 min after dilution) on neutravidin-coated biotin-PEG surfaces. Scale bar = 10 μm, N = 4 independent experiments, n = 968 observations. (C) Normalized fluorescence intensities of co-localized Cy3-Cdc42 (green) and RhoGDI1 (magenta) molecules imaged at maximal frame rate (22 ms) during recruitment at t = 0 s. Co-localization during the first three frames after recruitment was considered (grey box, see “Methods”). (D) Dual-color TIRFM images of a Cy3-Cdc42:A647-RhoGDI1 complex (upper) compared to a Cdc42 only (lower) recruitment event under conditions as (B). Scale bar = 1 μm. (E) Mean fraction ± SD of Cy3-Cdc42 molecules co-recruited with A647-RhoGDI1 (n = 968, N = 4, SD). (F) Scheme of two potential Rho GTPase recruitment mechanisms; either in complex with RhoGDI or free. (G) TIRFM images (Cy3-Cdc42, A647-GDI1, merge, and trajectories of Cdc42, left-to-right) of a PM SLB exposed to freshly-diluted Cy3-Cdc42:A647-GDI1 complexes (100 pM) in solution. Scale bar = 10 μm. (Image from (G) also used in Appendix Fig. S2B). (H) Dual-color TIRFM images at indicated times of a Cy3-Cdc42 molecule upon recruitment (upper, t = 0 s at recruitment) and dissociation (lower, t = 0 s at time of dissociation) on a PM SLB. Notice the absence of co-localized A647-RhoGDI1. Scale bar = 1 μm. (I) Mean fraction ± SD of Cy3-Cdc42 molecules co-recruited with A647-RhoGDI1 to PM SLBs (n = 7851, N = 3) compared to the positive control (n = 968, N = 4) in (E). T test was used to determine significance. (J) Probability (dots) ± SD (area) of Cy3-Cdc42 and A647-GDI1 co-localization as a function of Cdc42 membrane lifetime with t = 0 s as the moment of recruitment. Vertical grey box demarks the frames evaluated for co-recuitment. n = 7851, N = 3.

Figure 4. RhoGEFs act on membrane-bound Rho GTPases and not their soluble forms.

(A) Scheme of the FRET-based single molecule nucleotide exchange assay on PM SLBs. (B) Cy5-GDP:Cy3-Cdc42 complexes (50 pM) recruited to PM SLBs (containing 2% Ni²⁺-NTA-DGS) coated or not with high densities (350 molecules/µm²) of ITSN_cat imaged by sensitized Cy5 emission by TIRFM. Scale bar = 10 μm. (C) Survival fraction of Cy5-GDP as a function of time after membrane recruitment in the absence (pink) and presence (purple) of ITSN_cat. (-ITSN_cat n = 3348, N = 3, +ITSN_cat n = 3023, N = 3). (D) Scheme of two potential routes to GEF-mediated GTPase recruitment and activation. (E) TIRFM images (Cy3-Cdc42, A647-GDI1, merge, and trajectories of Cdc42, left-to-right) of a PM SLB (containing 2% DGS-NTA(Ni)) coated (bottom) or not (top) with high densities of ITSN_cat exposed to freshly-diluted Cy3-Cdc42:A647-GDI1 complexes (100 pM) in solution. Scale bar = 10 μm. (F) Dual-color TIRFM images at indicated times over the lifetime of a Cy3-Cdc42 molecule on a PM SLB, coated (right) or not (left) with high densities of ITSN_cat (t = 0 s at recruitment). Notice the absence of co-localized A647-RhoGDI1. Scale bar = 1 μm. (G) Mean fraction ±SD of Cy3-Cdc42 molecules co-recruited with A647-RhoGDI1 to PM SLBs in the absence or presence of ITSN_cat compared to the positive control (n = 968, N = 4) (Fig. 3E). (−/+ GEF experiments n = 12911, N = 4 and n = 2890 N = 3 respectively). T test was used to determine significance. (H) Probability (dots) ±SD (area) of Cy3-Cdc42 and A647-GDI1 co-localization as a function of Cdc42 lifetime on SLBs coated with (violet) or without (pink) ITSN_cat. t = 0 s is the moment of recruitment. Vertical grey box demarks the frames evaluated for co-recuitment. -ITSN_cat n = 12911, N = 4, +ITSN_cat n = 2890 N = 3. (I) Scheme of GEF-dependent and independent membrane recruitment of free Rho GTPases (J) Cy3-Cdc42 landing rates (mean ± SD) on PM SLBs in the absence (pink) or presence (purple) of ITSN_cat. -ITSN_cat n = 112911, N = 4, +ITSN_cat n = 2890, N = 3. T test was used to determine significance.

Figure 5. Reconstitution of membrane-templated Rho GTPase activity patterns in vitro.

(A) Scheme of PI(4)P and PI(4,5)P₂ lipid pattern formation by PIP5K and DrrA-OCRL on PM SLBs. (B) Time-lapse TIRFM images of pattern formation on PM SLBs (initially containing 2% PI4P and PI(4,5)P₂ each) by PIP5K (20 nM) and DrrA-OCRL (here 4 nM, for all following experiments 6 nM were used) visualized by A647-PH (2 nM, yellow) at indicated times before GEF addition. (C) Schematic representation or (D) time-lapse multi-color TIRFM images of PIP lipid patterns on PM SLBs visualized by A647-PH (2 nM, yellow) at indicated times before or after addition of A488-ITSN_cat-PH (1 nM, purple) at t = 0 s. (E) Scheme of the local activation of Cdc42 by ITSN_cat-PH in the PI(4,5)P₂-rich areas of PM SLBs, the resulting binding of the activity probe wCRIB and the diffusion of active Cdc42 into GEF-lacking areas over time. (F) Time-lapse multi-color TIRFM images of Cy3-PH (2 nM, yellow), A488-wCRIB (40 nM, magenta) and A647-Cdc42 (5.5 nM, green) on PIP patterns at indicated times before or after addition of ITSN_cat-PH (1 nM) at t = 0 s. Bottom row shows segmentation based on the lipid pattern. (G) Average intensities inside (filled squares, <I_In > ) and outside (hollow squares, <I_Out > ) of the GEF-containing membrane areas or (H) patterning index (full circle, PI = <I_In > / <I_Out>) of A488-wCRIB (magenta) over time. (I–L) Correspond to (E–H), with the additional presence of OPHN1_cat (20 nM). However, the segmentation is not shown in (J). Additionally, in (K) are the average intensities inside and outside of the GEF-containing membrane areas and in (L) the patterning index of A647-Cdc42 (green) over time. All numeric data represent the mean from three independent experiments (symbols) ±SD (shaded areas) (N = 3). All scale bars are 20 µm. (Image from (J) also used in Appendix Fig. S5A).

Figure 6. Effector proteins can enrich Rho GTPases at membrane sites of their activity.

(A) Scheme or (B) time-lapse multi-color TIRFM images of the effects of RhoGDI1 on templated Rho GTPase activity patterns. Images of Cy3-PH (2 nM, yellow), A488-wCRIB (40 nM, magenta) and A647-Cdc42:RhoGDI1 complexes (600 nM, green) on PIP patterns at indicated times before or after addition of ITSN_cat-PH (1 nM) at t = 0 s in the presence of OPHN1_cat (20 nM). (Image from (B) also used in Fig. EV4D and Appendix Fig. S5A). (C) Patterning indices (PI = <I_In > / <I_Out>) of A488-wCRIB (magenta) or A647-Cdc42 (green) under conditions as in (B) over time. (D) Schematic representation or (E) time-lapse multi-color TIRFM images of the effects of full-length N-WASP on templated Rho GTPase activity patterns. Conditions as in (B) with A647-Cdc42:RhoGDI1 complexes replaced by A647-Cdc42 (5.5 nM, green) and A488-wCRIB replaced by Atto488-N-WASP (40 nM, magenta). (F) corresponds to (C) under conditions as in (E). (G) Schematic representation or (H) time-lapse multi-color TIRFM images of the effects of the synthetic dimerization mimic 2xwCRIB on templated Rho GTPase activity patterns. Conditions as in (E) with Atto488-N-WASP replaced by A488-2xwCRIB (20 nM, magenta). (I) corresponds to (C) under conditions as in (H). Additionally, the grey shaded box highlights the first 100 s after the addition of ITSN_cat-PH. (J) Normalized patterning indices of A488-wCRIB (magenta) or A647-Cdc42 (green) under conditions as in (H) over time with the grey shaded box highlighting the first 100 s after the addition of ITSN_cat-PH. All numeric data represent the mean from three independent experiments (symbols) ±SD (shaded areas) (N = 3). All scale bars are 20 µm.

Figure 7. The mechanism of effector-driven Rho GTPase enrichment at membranes in vitro.

(A) Schematic representation of effector-mediated changes in diffusivity or (B) membrane stability of Cdc42 as possible mechanisms of active Rho GTPase membrane enrichment. (C) Still images (top) and trajectories (bottom) of single A647-Cdc42 molecules (5.5 nM total, 55 pM labeled) on PIP-templated activity patterns either inside (yellow) or outside (black) of GEF-containing regions in the presence of either wCRIB, 2xwCRIB or N-WASP as indicated. (D) Stepsize distribution or (E) MSD over time of Cdc42 trajectories inside (yellow) and outside (black) GEF areas at conditions as indicated. (F) Time-lapse multi-color TIRFM images of PIP-templated Cdc42 activity patterns in the presence of A488-wCRIB or A488-2xwCRIB at indicated times before or after buffer flow-out at t = 0 s. Flow-out buffer contained all proteins at the same concentrations before flow onset, except for A647-Cdc42 that was omitted, and additional free RhoGDI1 (22 nM) to accelerate Cdc42 membrane dissociation. White boxes (4.1 µm × 4.1 µm) indicating ROIs used for analysis. (G) Normalized average intensities of A647-Cdc42 inside (yellow) or outside (black) GEF-containing areas as a function of time after buffer wash-out in the presence of either A488-wCRIB (solid lines) or A488-2xwCRIB (dashed lines). Half-times were obtained by mono-exponential fits to the data. All numeric data represent the mean from three independent experiments (symbols) ±SD (shaded areas) (N = 3). All scale bars are 20 µm.

Figure 8. Cdc42 activation precedes Cdc42 retention.

(A) Scheme showing photoactivation of IT-PA-Cdc42 at a site of baseline/background Cdc42 activity and a site of increased Cdc42 activity around a wound. (B) Representative image of baseline Cdc42 activity (wGBD) before (left) and immediately after (right) photoactivation of IT-PA-Cdc42. Scale bar 20 µm. (C) Montage of IT-PA-Cdc42 from (B) showing disappearance over time. (D) Plot of IT-PA-Cdc42 fluorescence intensity from (B) over time after photoactivation. (E–G) Micrograph (scale bar 20 µm), montage scale bar 10 µm, and plot of IT-PA-Cdc42 photoactivation within the Cdc42 activity zone around a wound. (H–J) Micrograph (Scale bar 20 µm), montage (scale bar 10 µm), and plot of IT-PA-Cdc42 photoactivation outside the Cdc42 activity zone around a wound. (K) Plot of the t_1/2 to disappearance (mean ± SD) of IT-PA-Cdc42 prior to wounding (PA pre-wound; n = 2,) within the Cdc42 zone for the earliest photoactivation (PA1 at wound; n = 24), within the Cdc42 zone for a later photoactivation (PA2 at wound; n = 24), within the Cdc42 zone for the latest photoactivation (PA3 at wound; n = 20), and outside the Cdc42 zone (PA off wound; n = 22). One way ANOVA with Dunnett comparison was used to determine significance. PA pre-wound vs. PA2 at wound P value = 1.09682E-08, PA pre-wound vs. PA3 at wound P value = 5.60967E-10. (L) Correlation between t_1/2 to disappearance of photoactivation within the Cdc42 zone and time after wounding. A positive correlation was found (P = 0.0043). (M–O) Micrograph (Scale bar 20 µm), montage (scale bar 10 µm), and plot of IT-PA-Cdc42 photoactivation within the Cdc42 activity zone around a wound. (P–R) Micrograph (scale bar 20 µm), montage (scale bar 10 µm), and plot of IT-PA-Cdc42 photoactivation within the Cdc42 activity zone with ITSN-C2 mediated increased Cdc42 activity. (S) Plot comparing the t_1/2 to disappearance (mean ± SD) of IT-PA-Cdc42 in control and ITSN-C2 conditions prior to wounding (PA pre-wound; control n = 10, ITSN-C2 n = 13) within the Cdc42 zone for the earliest photoactivation (PA1 at wound; control n = 7, ITSN-C2 n = 13), within the Cdc42 zone for a later photoactivation (PA2 at wound; control n = 10, ITSN-C2 n = 13), within the Cdc42 zone for the latest photoactivation (PA3 at wound; control n = 10, ITSN-C2 n = 12), and outside the Cdc42 zone in wounded cells (PA off wound; control n = 10, ITSN-C2 n = 13). Two-sample t tests were used to determine significance. (T) Correlation between t_1/2 to disappearance of photoactivation within the Cdc42 zone and time after wounding for control and ITSN-C2 conditions. No correlation was found for control (P = 0.1523), but a positive correlation was found for ITSN-C2 (P = 0.0106).

Figure 9. Model figure.

Model of route to Rho GTPase pattern formation.

Figure EV1. Cy3-Cdc42 follows same patterning as IT-Cdc42 around wounds.

(A) Micrograph of Cy3-Cdc42 during Pre-wound, Pre-zone, and Zone time points in the repair process. Scale bar 20 µm. (B) Kymograph of the micrograph from (A) generated by radially averaging signal intensity around the wound over time (Moe et al, ; see “Methods”). The arrow denotes when the wound occurred, and W denotes the wound location. The yellow line “P” indicates the Pre-zone time point and the yellow line “Z” indicates the Zone time point. Vertical and horizontal scale bars are 30 s and 3 μm, respectively. (C) Line scan generated from the kymograph in (B) to show Cy3-Cdc42 fluorescence intensity at the Pre-zone time point as a function of distance from the wound center. (D) Line scan generated from the kymograph in (B) to show Cy3-Cdc42 fluorescence intensity at the Zone time point as a function of distance from the wound center. (E) Patterning index of Cy3-Cdc42 at the Pre-zone time point compared to the Zone time point. A two-sample t-test was used to determine statistical significance; n = 13. P value = 6.48965E-07. (F) Patterning index of Cy3-Cdc42 compared to IT-Cdc42 (IT-Cdc42 data were taken from Fig. 1). A two-sample t-test was used to determine statistical significance. Cy3-Cdc42 n = 13 individual cells, IT-Cdc42 n = 12 individual cells, in N = 1 experiment. (G) Example of wGBD and IT-Cdc42 timing in cells where both accumulate rapidly. (H) Example of wGBD and IT-Cdc42 timing in cells where both accumulate more slowly. (I) Timing of IT-RhoA and rGBD enrichment around the wound (mean, N = 11 per condition, shaded area = SD).

Figure EV2. Achieving bilayer surface coverage of active GEF.

(A) Scheme of the single molecule quantification of RhoGEF density on SLBs. (B) 488-ITSN_cat (left) 488-ITSN_cat(Lac) (center) and 488-ELMO:DOCK (right) recruited to SLBs (containing 2% DGS-NTA(Ni), 22.5% PS and 2% PI(3,4,5)P₃ respectively) Scale bar = 10 μm. (C) Quantification of ITSN_cat density on SLBs (454 molecules/µm², 4.8% surface coverage), left axis. Estimated bilayer surface area coverage of total ITSN_cat right axis (mean ± SD). N = 1 experiment, n = 3–13 fields of view imaged per condition, total spots = 49823. (D) Quantification of ITSN_cat(Lac) density on SLBs. (203 molecules/µm², 2.2% surface coverage), left axis. Estimated bilayer surface area coverage of total ITSN_cat(Lac) right axis (mean ± SD) N = 1 experiment, n = 3–15 fields of view imaged per condition, total spots = 34044. (E) Quantification of ELMO-DOCK density on SLBs. (84 molecules/µm², 2.3% surface coverage), left axis. Estimated bilayer surface area coverage of total ELMO-DOCK right axis (mean ± SD) N = 1 experiment, n = 2–11 fields of view imaged per condition, total spots = 17985. (F) Survival fraction of 400 nM 488-ITSN_cat, 400 nM 488-ITSN_cat(Lac) and 80 nM 488-ELMO-DOCK as a function of time during wash off (shaded area, SD). Flow rate 10 μL/s.

Figure EV3. DOCK family GEF does not provide mechanism for Cdc42:GDI1 recruitment.

(A) Scheme of two potential routes to GEF-mediated GTPase recruitment and activation. (B) TIRFM images of single molecule recruitment of Rac1 from complex (100 pM) (Cy3-Rac1, A647-GDI1, merge, and trajectories of Cdc42, left-to-right) on a PM SLB (containing 4% PI(3,4,5)P₃) in the absence (top) or presence (bottom) of 80 nM ELMO-DOCK. Scale bar = 10 μm. (C) Probability (dots) ± SD (area) of Cy3-Rac1 and A647-GDI1 co-localization as a function of Rac1 lifetime on SLBs coated with (dark green) or without (light green) ELMO-DOCK. t = 0 s is the moment of recruitment. Vertical grey box demarks the frames evaluated for co-recruitment. – ELMO-DOCK N = 3, n = 5068, + ELMO-DOCK N = 3, n = 2984. (D) Mean fraction ± SD of Cy3-Rac1 molecules co-recruited with A647-RhoGDI1 to PM SLBs in the absence or presence of ELMO-DOCK compared to the positive control (Fig. 3E). ( ± GEF experiments (– ELMO-DOCK N = 3, n = 5068, + ELMO-DOCK N = 3 n = 2984, control N = 4 n = 968, SD)). Statistical comparisons via T-test. (E) Scheme of GEF-dependent and independent membrane recruitment of free Rho GTPases. (F) Cy3-Rac1 landing rates (mean ± SD) on PM SLBs in in the absence (light green) or presence (dark green) of ELMO-DOCK. – ELMO-DOCK N = 3 independent experiments, n = 5068 observations, + ELMO-DOCK N = 3 independent experiments, n = 2984 observations. Statistical comparisons via t-test.

Figure EV4. RhoGDI1 negligibly affects the reconstitution of membrane-templated Rho GTPase activity patterns in vitro.

(A) Time-lapse multi-color TIRFM images of Cy3-PH (2 nM, yellow), A488-wCRIB (40 nM, magenta) and A647-Cdc42:RhoGDI1 complexes (600 nM, green) on PIP patterns at indicated times before or after addition of ITSN_cat-PH (1 nM) at t = 0 s. Bottom row shows segmentation based on the lipid pattern. (B) Average intensities inside (filled squares, <I_In > ) and outside (hollow squares, <I_Out > ) of the GEF-containing membrane areas or (C) patterning index (full circle, PI = <I_In > / <I_Out>) of A488-wCRIB (magenta) over time. (D–F) correspond to (A–C), with the additional presence of OPHN1_cat (20 nM). However, the segmentation is not shown in (D) (Image from (D) also used in Fig. 6B and Appendix Fig. S5A (lower). Additionally, in (E) are the average intensities inside and outside of the GEF-containing membrane areas and in (F) the patterning index of A647-Cdc42 (green) over time. (G) Time-lapse multi-color TIRFM images of the effects of the synthetic dimerization mimic 2xwCRIB on templated Rho GTPase activity patterns. Conditions as in (D) with A488-wCRIB replaced by A488-2xwCRIB (20 nM, magenta). (H) corresponds to (F) under conditions as in (G). All numeric data represent the mean from three independent experiments (symbols) ± SD (shaded areas) (N = 3). All scale bars are 20 µm as indicated.

Figure EV5. ITSN-C2-BFP is recruited to single cell wounds and increases Cdc42 activity and patterning.

(A) Cell expressing Itsn-C2-BFP (Itsn-C2-BFP) or not (Control). Itsn-C2-BFP Top: Montage of images from confocal time-lapse movie showing accumulation of ITSN-C2-BFP around wounds. Itsn-C2-BFP Bottom: corresponding wGBD signal from images shown in top. Control: example of wGBD signal in cell not expressing ITSN-C2-BFP for comparison. Time in min:sec. Scale bar 20 µm. (B) Plot of Itsn-C2-BFP over time. Data show mean ± SD from 22 cells; relative to pre-wound signal (students t-test, P values 4.71E-11, 1.09-10 and 3.75E-16 for timepoints 20, 30 and 40 s respectively). (C) Plot comparing wGBD accumulation around wounds over time in control cells and cells expressing Itsn-C2-BFP. Data show mean ± SD from 17 cells; Itsn-C2-BFP versus control (students t test, (students t-test, P values 5.61E-10, 4.82-13 and 3.91E-16 for timepoints 20, 30 and 40 s respectively). (D) Plot comparing IT-Cdc42 accumulation around wounds over time in control cells and cells expressing Itsn-C2-BFP. Data show mean ± SD from 12 cells; Itsn-C2-BFP versus control (students t-test, (students t-test, P values 3.63E-4, 6.66E-6 and 2.11E-5 for timepoints 20, 30 and 40 s, respectively).