Rho GTPases are essential regulators of cytoskeletal reorganization, but how they do so during neuronal morphogenesis in vivo is poorly understood. The actin-depolymerizing and actin-severing protein factor cofilin, encoded by twinstar, is essential for axon growth in Drosophila neurons. Cofilin function in axon growth is inhibited by LIM kinase and activated by Slingshot phosphatase. Dephosphorylating cofilin appears to be the major function of Slingshot in regulating axon growth in vivo. Genetic data provide evidence that Rho or Rac/Cdc42, via effector kinases Rho-associated kinase (Rok, also named Rho kinase or ROCK), or p21-activated kinase (Pak), respectively, activate LIM kinase to inhibit axon growth. Importantly, Rac also activates a Pak-independent pathway that promotes axon growth, and different RacGEFs regulate these distinct pathways. These genetic analyses reveal convergent and divergent pathways from Rho GTPases to the cytoskeleton during axon growth in vivo and suggest that different developmental outcomes could be achieved by biases in pathway selection (Ng, 2004).

From yeast to mammals, cofilin plays an essential morphogenetic role by promoting the rapid turnover of actin filaments through severing filamentous actin (F-actin) and depolymerizing actin filaments from the pointed ends (Bamburg, 1999). Mutations in the twinstar (tsr) gene, which encodes the unique Drosophila homolog of cofilin, result in neuroblast proliferation, spermatogenesis, and defects in epithelial morphogenesis (Gunsalus, 1995; Chen, 2001). Drosophila cofilin was first identified in a screen for genes that induce aberrant cell shapes in fission yeast (Edwards, 1994). In mammalian cells, cofilin activity is inhibited by phosphorylation at serine 3, which is mediated by LIM kinase (LIMK). LIMK is activated through phosphorylation by Pak or Rok effector kinases for Rac/Cdc42 or Rho, respectively. In Drosophila, one LIMK gene has been found that can also phosphorylate cofilin at serine 3 (LIMK1-Flybase) (Ohashi, 2000), but how LIMK1 is regulated is unknown. Cofilin is dephosphorylated by Slingshotphosphatase (Ssh) (Niwa, 2002). Drosophila ssh mutants exhibit defects in epithelial morphogenesis that are characterized by high levels of F-actin and cofilin phosphorylation, suggesting that Ssh regulates actin dynamics through cofilin dephosphorylation (Ng, 2004 and references therein).

The function and regulation of cofilin in neuronal morphogenesis in vivo has not been reported. Loss of cofilin is shown to result in severe axon growth defects in Drosophila neurons. Cofilin function is positively regulated by Ssh phosphatase and negatively regulated by LIMK during axon growth. Cofilin dephosphorylation appears to be the major function of Ssh, since axon growth defects in ssh mutant neurons are suppressed by expressing active forms of cofilin. Genetic evidence is provided that both the Rho-Rok and the Rac/Cdc42-Pak pathways positively regulate Drosophila LIMK. Importantly, while the LIMK pathway acts to inhibit axon growth, Rac also signals through a Pak-independent pathway that acts antagonistically to LIMK to promote axon growth. Furthermore, two distinct RacGEFs appear to be selectively engaged to regulate these different pathways. These genetic results indicate that multiple Rho GTPase signaling pathways converge on a common downstream target, cofilin, to regulate axon growth. At the same time, Rho GTPases also regulate divergent downstream pathways that act in an antagonistic fashion to coordinate growth cone motility (Ng, 2004).

Using Drosophila mushroom body neurons as a genetic model, several signaling pathways through which Rho GTPases regulate axon growth in vivo were examined. Several new insights were obtained from these genetic analyses. Actin polymerization at the cellular leading edge is generally thought to provide the driving force for membrane protrusions such as lamellar extension in migrating cells or filopodia and lamellipodia extensions for neuronal growth cone advance. Cofilin was found to be essential for axon growth in vivo. Since cofilin has both pointed-end depolymerization activity and F-actin severing activity, there are at least two possible explanations, on the basis of biochemical and cell biological studies in other cell types, for its essential role in axon growth (e.g., Carlier, 1997; Svitkina, 1999; Dawe, 2003; Ghosh, 2004). (1) Actin polymerization at the leading edge requires a constant supply of monomeric actin subunits derived from depolymerization at the pointed end. (2) The severing activity of cofilin allows generation of free barbed ends as templates for actin polymerization. While the contributions of either of the above processes to axon growth have not been ruled out, it was found that loss of cofilin does not simply result in a lack of filopodia or lamellipodia. Instead, the overabundance of filopodia- and lamellipodia-like structures retained in cofilin mutant axons suggests a third possibility: growth cone advance is inhibited when filopodia/lamellipodia cannot be disassembled upon the loss of cofilin activity (Ng, 2004).

How is cofilin activity regulated during axon growth? LIM Kinase and Ssh Phosphatase are known to be key regulators of cofilin in axon growth. A number of recent studies have addressed the role of cofilin phosphorylation by LIM-kinase and Ssh phosphatase in cultured neurons. For example, overexpression of active forms of cofilin in rat cortical neurons results in an increase in both the number of filopodia and the degree of neurite extension (Meberg, 2000). Overexpression of active forms of LIMK in chick dorsal root ganglion neurons represses growth cone motility and neurite extension. The LIMK effects are likely to be mediated through cofilin, since cotransfection of either mammalian Ssh or the S3A form of cofilin suppresses the LIMK effects (Endo, 2003). Cofilin phosphorylation by LIMK (Aizawa, 2001) is further implicated in semaphorin-mediated growth cone collapse (Ng, 2004).

In vivo study in Drosophila confirms and extends these in vitro studies in several ways. (1) Using a transgenic rescue assay, it was shown that neither phosphomimetic (S3E) nor nonphosphorylatable (S3A) cofilin or their combination can replace endogenous cofilin function. This suggests that cycles of cofilin phosphorylation ('inactivation') and dephosphorylation ('reactivation') are important during actin turnover to promote axon growth and that in vivo the factors that regulate cofilin phosphorylation must act in a delicate balance to optimize axon growth during development. (2) Loss of Ssh is shown to result in axon growth defects, and these defects can be suppressed by the expression of active cofilin, demonstrating that the major function of Ssh in regulating axon growth is cofilin dephosphorylation. (3) LIMK overexpression is shown to result in axon growth defects analogous to ssh, and this phenotype can be suppressed by the coexpression of Ssh or active cofilin. Taken together with existing biochemical data, these results firmly establish that regulation of cofilin phosphorylation by Ssh phosphatase and LIMK plays a pivotal role in regulating axon growth in vivo (Ng, 2004).

Although cofilin phosphorylation is essential for neuroblast proliferation, neither LIMK nor Ssh appears to be a key regulator. No cell proliferation defects were detected in ssh^-/- neuroblast clones or in LIMK1-overexpressing neurons, in contrast to tsr^-/- clones. It is unlikely that cell proliferation is less sensitive to the reduction of cofilin activity than is axon growth. On the contrary, neuroblast clones homozygous for a hypomorphic allele of tsr (tsr¹) have strong defects in cell proliferation, but no defects in axon growth. Taken together, these data suggest that cofilin phosphorylation during cell proliferation is regulated by a set of kinases/phosphatases different from those that regulate axon growth (Ng, 2004).

Genetic analyses show that Rho, Cdc42, and Rac all contribute to activation of the LIMK1 pathway, which leads to axon growth inhibition. However, previous cell biological data in vitro (Kozma, 1997) and loss-of-function mutant analysis in vivo (Lundquist 2001; Hakeda-Suzuki, 2002; Ng, 2002) indicate that Rac GTPases act to promote axon growth. How can one resolve these seemingly opposite effects of Rac GTPases? Several lines of evidence are provided that, in addition to activating LIMK1, Rac GTPases also act via a second pathway to promote axon growth. (1) Reduction of Rac GTPase activity can also enhance the LIMK1 overexpression phenotype, suggesting that Rac can act antagonistically to LIMK1 to promote axon growth. (2) Overexpressing Rac1 Y40C (a mutant with diminished binding to Pak) strongly suppresses the LIMK1 overexpression phenotype. Since Pak activation leads to axon growth inhibition and Pak^-/- neuroblast clones do not have axon growth defects, these data together suggest that the Rac pathway that counteracts the LIMK pathway is Pak independent, which is consistent with previous studies in which transgenically supplied Rac1 Y40C rescued axon growth in the absence of all endogenous Rac (Ng, 2002). (3) Two different RacGEFs have been shown to either enhance or suppress the LIMK pathway. This again suggests that different Rac signaling pathways act antagonistically to regulate axon growth (Ng, 2004).

Given the presence of these two Rac GTPase pathways (LIMK1--->Twinstar dependent and independent pathways), it is likely that, depending on the signaling context, Rac can either inhibit or promote axon growth. Indeed, Rac activation has been shown to either promote or inhibit axon growth in different systems (reviewed in Luo, 2000). In addition, both attractive and repulsive axon guidance cues can signal through Rac GTPases to mediate these opposite effects in vivo. It is proposed that one possible explanation for the above phenomena is that different cues selectively favor either the Pak/LIMK/cofilin phosphorylation pathway or the alternative axon growth-promoting pathway, resulting in different developmental outcomes. The finding that two RacGEFs (Trio and Still life) have opposite effects in modifying LIMK activity suggests that the selection of these pathways could be achieved by selectively engaging different GEFs. How RacGEFs selectively couple to different downstream effector pathways remains to be determined by future experiments (Ng, 2004).

Rac activation stimulates actin polymerization and leads to cell protrusions via lamellipodia formation. However, testing of several major classes of actin polymerization stimulators did not provide evidence that Rac promotes axon growth through the actin polymerization pathway. For instance, the SCAR-Arp2/3 pathway essential for de novo actin polymerization does not appear to contribute to MB axon growth in vivo. This is consistent with a recent study suggesting that the Arp2/3 pathway is also not essential for axon growth in cultured neurons. Although the anticapping protein Enabled is essential for MB axon growth, genetic interaction data argue against its participation in the axon growth-promoting pathway downstream of Rac. This is also consistent with previous genetic analysis in C. elegans, indicating that Rac (Ced-10) and Ena (Unc-34) act in parallel pathways downstream of the netrin receptor to promote axon growth. Another possibility for the axon growth-promoting pathway is that Rac counteracts the LIMK pathway by activating Ssh. Recent in vitro data suggest that Rac can act to dephosphorylate cofilin (Nagata-Ohashi, 2004), thereby promoting actin turnover. However, it is believed that the Rac-dependent axon growth pathway is unlikely to be via Ssh alone, given that, in the absence of ssh, Rac Y40C overexpression can still suppress the ssh growth phenotype. Since axon growth also requires the regulation of microtubule dynamics and vesicle trafficking, both of which are thought to be Rho GTPase dependent, it is proposed that the Rac-mediated axon growth-promoting pathway may involve these processes (Ng, 2004).

In summary, genetic analyses have begun to tease apart the complex signaling networks between Rho GTPases and the actin cytoskeleton in the context of axon growth in vivo. Rho GTPases act through convergent and divergent signaling pathways to regulate axon growth. In addition to cofilin regulation, analyses of other actin polymerization regulators in MB neurons have established the relationship between these signaling pathways and the regulation of axon growth. The pathways identified in this study provide a foundation for future investigations as to how extracellular cues direct growth cone signaling to precisely wire neural circuits in vivo (Ng, 2004).

GENE STRUCTURE

cDNA clone length - 771

Bases in 5' UTR - 168

Exons - 4

Bases in 3' UTR - 156

PROTEIN STRUCTURE

Amino Acids - 148

Structural Domains

See SMART (Simple Modular Architecture Research Tool) for information of Twinstar structure.

date revised: 20 February 2005

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.