Like our bodies, individual cells have a “skin” that can repair itself after an injury. The skin of a cell is comprised of two layers, the outer cell membrane and the inner cell cortex. While the membrane is lipid-based, the cortex is largely composed of the protein actin, which forms a filamentous network. When a cell experiences physical or chemical trauma, both layers must be repaired. First, the membranes of intracellular vesicles fuse to plug the hole. Next, an actin-based ring forms around the wound site that, with help from the motor protein myosin, translocates inward to close the wound. While the basics of single cell wound repair are understood, much work remains to identify the molecular players that regulate this process.
Toward this goal, the Parkhurst Laboratory (Basic Sciences Division) “recently developed the first genetically amenable model system for studying cell wound repair using the fruit fly” says Dr. Mitsutoshi Nakamura, a post-doctoral fellow with Dr. Susan Parkhurst. Their system uses Drosophila embryos, which have the useful property of existing as a giant single cell at early stages due to a delay in cytokinesis. Wounds are generated by focusing a laser on the cortical surface of an embryo, and wound repair can be observed with the use of fluorescently tagged reporters and time-lapse microscopy (Figure 1).
Repair of the cell cortex requires a family of proteins known as the Rho GTPases, which organize into structured arrays at wound sites. In work published this fall in the Journal of Cell Biology, Dr. Nakamura and his colleagues in the Parkhurst lab set out to understand how these GTPase arrays form. They focused on three key GTPases—Rho1, Rac1 and Cdc42—that are conserved from yeast to humans.
GTPases function by cycling between GTP- and GDP-bound forms. This process often requires other proteins, such as GTPase activating proteins (GAPs) and guanine nucleotide exchange factors (GEFs). The researchers examined the localization patterns of fluorescently tagged RhoGAPs and RhoGEFs at wound sites and were surprised to observe that all three tested GEF proteins—Pebble, RhoGEF2 and RhoGEF3—formed distinct spatial arrays around wound sites within 30 seconds. Each GEF was shown to adopt a slightly different pattern; for example, Pebble localizes to clearly different areas than RhoGEF2 but shows some overlap with RhoGEF3. By contrast, the GAP protein Tumbleweed did not form a particular pattern.
Next, Dr. Nakamura asked whether actin localization and dynamics at wound sites are influenced by particular GAP or GEF gene(s). While knockdown of tumbleweed via RNA interference did not have strong effects, knocking down any of the three RhoGEFs delayed actin accumulation at the wound edge and slowed down wound closure. In the case of pebble and RhoGEF3 knockdowns, actin mis-localized to within the wound site, rather than around it. Based on these results, the authors concluded that the RhoGEFs are essential in spatially organizing repair machinery at wound sites.
To determine whether patterning of the RhoGEFs controls spatial organization of the GTPase(s), the researchers tested whether knockdown of the GEFs disrupts GTPase array formation. Indeed, GEF knockdown led to mis-localization of all three GTPases, indicating that the GEFs are required for GTPase patterning. Strikingly, the GTPase Rho1 is not recruited to wound sites at all in the absence of RhoGEF2, indicating a strong dependence of Rho1 localization on RhoGEF2.
Because it was clear that RhoGEF2 pre-patterns Rho1, the authors next asked what controls patterning of RhoGEF2 itself. They had noticed that the localization of RhoGEF2 tended to be similar to that of actin structures, so they tested whether the drug latrunculin B (LatB), which de-polymerizes actin filaments, affected RhoGEF2 localization at early time points. Indeed, RhoGEF2 could not localize to the wound site at all in the presence of LatB, suggesting that an actin network is required for the formation of RhoGEF2 arrays.
Actin network formation is a highly regulated process governed by cytoskeleton regulatory proteins. Of particular relevance to actin dynamics during wound healing are the Annexins (Anx), which are known to form structured arrays and accumulate at sites of injury. In particular, the authors observed that (i) AnxB9 is recruited to the wound location extremely rapidly (Figure 1), even when the actin network is disrupted by LatB and (ii) mutation of AnxB9 delayed actin accumulation at wound sites and completely prevented RhoGEF2 recruitment. These results support a model in which AnxB9 is responsible for stabilization of actin, which subsequently recruits the RhoGEFs and then the Rho GTPases (Figure 2). Consistent with this order of events, stabilization of polymerized actin by the drug phalloidin partially restored RhoGEF2 recruitment in the AnxB9 mutant.
Single cell wound healing requires both membrane sealing and repair of the underlying cell cortex. Central to this process are the Rho GTPase signaling proteins, which must be accurately localized to function properly. The Parkhurst lab has now implicated RhoGEFs, the annexin AnxB9 and actin itself in patterning Rho GTPases at sites of cortical injury. In the future, Dr. Nakamura plans to continue “identifying new molecules involved in the repair process and mechanisms leading to the earliest steps”.
Nakamura M, Verboon JM and Parkhurst SM. “Prepatterning by RhoGEFs governs Rho GTPase spatiotemporal dynamics during wound repair.” Journal of Cell Biology. 2017 Dec 4;216(12):3959–3969. doi: 10.1083/jcb.201704145
This research was supported by the National Institutes of Health.