Simon CM et al. (2013), Pattern formation of Rho GTPases in single cell...

XB-ART-46448

Mol Biol Cell 2013 Feb 01;243:421-32. doi: 10.1091/mbc.E12-08-0634.

Show Gene links Show Anatomy links

Pattern formation of Rho GTPases in single cell wound healing.

Simon CM , Vaughan EM , Bement WM , Edelstein-Keshet L .

???displayArticle.abstract???
The Rho GTPases-Rho, Rac, and Cdc42-control an enormous variety of processes, many of which reflect activation of these GTPases in spatially confined and mutually exclusive zones. By using mathematical models and experimental results to establish model parameters, we analyze the formation and segregation of Rho and Cdc42 zones during Xenopus oocyte wound repair and the role played by Abr, a dual guanine nucleotide exchange factor-GTPase-activating protein, in this process. The Rho and Cdc42 zones are found to be best represented as manifestations of spatially modulated bistability, and local positive feedback between Abr and Rho can account for the maintenance and dynamic properties of the Rho zone. In contrast, the invocation of an Abr-independent positive feedback loop is required to account for Cdc42 spatial bistability. In addition, the model replicates the results of previous in vivo experiments in which Abr activity is manipulated. Further, simulating the model with two closely spaced wounds made nonintuitive predictions about the Rho and Cdc42 patterns; these predictions were confirmed by experiment. We conclude that the model is a useful tool for analysis of Rho GTPase signaling and that the Rho GTPases can be fruitfully considered as components of intracellular pattern formation systems.

???displayArticle.pubmedLink??? 23264464
???displayArticle.pmcLink??? PMC3564532
???displayArticle.link??? Mol Biol Cell
???displayArticle.grants??? [+]

Species referenced: Xenopus laevis
Genes referenced: abr akt1 cdc42 fubp1 rac1 rho rho.2 rhoa

???attribute.lit??? ???displayArticles.show???

	FIGURE 1:. Basic features of GTPase response to wounding. (A) Time course showing individual frames generated from a 4D movie of active Rho (green) and active Cdc42 (red) during oocyte wound response. Each frame represents a projection of six optical sections. Time after wounding in minutes:seconds. (B) Plot of fluorescence intensity of GFP-rGBD (RhoA activity) within the zone at different times after wounding and intensity at regions 10 μm outside the zone. Each point represents an average of eight measurements taken at different positions within or outside the zone. The positions of the zone (and thus the measurements) were determined from line intensity scans and thus represent moving frames of reference. (C) Plot of fluorescence intensity of RFP-wGBD (Cdc42 activity) within the zone at different times after wounding and intensity at regions 10 μm outside the zone. Each point represents an average of eight measurements taken at different positions within or outside the zone as in B.
	FIGURE 2:. Segregation of the Rho and Cdc42 zones. (A) Data kymograph of 10-pixel-wide slices from a movie used to generate Figure 1A, showing active Rho (green) and active Cdc42 (red). Here t is time in seconds, r is distance from the wound center (spans 40 μm), and the wound is toward the left of the kymograph. The first, second, and third tick marks indicate the time of wounding, the initial condition (time zero) for the model, and the final simulation time, respectively. (B) Plots of intensity profiles (data) of probe for active Rho (green) or active Cdc42 (red) corresponding to slices presented in A. Dashed line represents the wound edge.
	FIGURE 3:. Models of GTPase pattern formation. (A) Schematic showing the basic scheme of Rho or Cdc42 association with the plasma membrane. (B) Schematic showing the three basic models generated and tested in this study. Model 2 contains proposed interactions based on experiments, and Model 3 contains a positive feedback loop for Cdc42 postulated by our model.
	FIGURE 4:. Model 1 simulations of RhoA activity and Abr vs. distance from the center of the wound. (A) Linear kinetics for the Abr-mediated activation of Rho result in any stimulus of RhoA or Abr decaying to the background level for all parameter regimes. (B) Michaelis–Menten kinetics fail to sustain the pattern as well. (C) A Hill function for the Abr-mediated Rho activation results in the sustenance of the Rho pattern, as observed in vivo.
	FIGURE 5:. Model 2 and 3 simulations of Cdc42 activity vs. distance from the center of the wound. We use the RhoA–Abr Model 1 described earlier. (A) Simulation of Model 2. For every parameter regime and any stimulus of Cdc42, any Cdc42 activity outside of the RhoA zone will decay to the background level without positive feedback for Cdc42. (B) Model 3 with a linear form of Cdc42 autocatalysis fails to capture the maintenance of the Cdc42 zone. (C) Model 3 with a Hill-function representation of Cdc42 autocatalysis results in the sustenance of the Cdc42 peak, as observed in vivo. The experimental intensity profiles from Figure 2B are plotted as points with the final Model 3 simulation. The simulated (solid) and experimental (dashed) wound edge is indicated with the black, vertical line.
	FIGURE 6:. In silico experiments to recapitulate in vivo results. (A) To model WT Abr overexpression, [A] is increased by 20%. The RhoA zone broadens to overtake the Cdc42 zone but does not significantly increase in intensity. (B) On overexpression of GEF-dead Abr, the RhoA and Cdc42 zones cannot be sustained. (C) On overexpression of GAP-dead Abr, the zones overlap and broaden in comparison to controls. (D) On C3 inactivation of RhoA, the Cdc42 system retains its bistable property and its pattern still forms. The Cdc42 activity zone brightens due to an elevated high steady state and broadens since the Rho zone is not present to suppress it in the region closer to the wound.
	FIGURE 7:. Model phase portraits reveal the effect of in silico experiments on the number and values of steady states. Left, Abr–Rho phase plane showing active Rho (green) and Abr (black) nullclines. Steady states (SS) are at points of intersection. For comparison, SS Rho activity levels for the controls are indicated with horizontal lines. Red diagrams, dC/dt vs. C (where C* = active Cdc42), showing rate of activation (solid) and inactivation (dotted) (SS Cdc42 activity levels for controls are indicated by the vertical lines.) Top, “control” (WT) simulations. Second row, WT Abr overexpression (as in Figure 6A). Third row, GEF-dead Abr (as in Figure 6B). Fourth row, GAP-dead Abr (as in Figure 6C). Last row, C3 Rho (as in Figure 6D). Note the disappearance and/or shifting steady-state values in the distinct treatments. Also note the higher turnover rates inside the Rho zone by comparing the ordinates in columns 1 and 2.
	FIGURE 8:. Two wound simulations (bottom) and experiments (top). (A) When the wounds are far apart (35 μm in silico), they behave independently. (B) When the wounds are very close (1 μm in silico), they behave as a single wound surrounded by single oval-shaped zones of Rho and Cdc42. (C) When wounds are close enough for the initial Rho gradients to overlap (20 μm in silico), the Rho zones between the wounds fuse together to create a zone wider than twice the width of that in a single control wound. The Rho recruits enough Abr to suppress the Cdc42 in this region with its GAP domain. (D) When wounds are close enough to have overlapping Cdc42 gradients but not significantly overlapping Rho gradients (26 μm in silico), the Cdc42 zones between the wounds fuse together to create a zone more than twice the width of that in a single control wound.

References [+] :

Artavanis-Tsakonas, Notch signaling: cell fate control and signal integration in development. 1999, Pubmed