**6. Ubiquitin and small ubiquitin-like modifier in exocytosis**

#### **6.1 Ubiquitination**

Post-translational modification with ubiquitin has also been recognised as an important sorting signal on cargo transported by the endosomal network (specifically as a signal for internalisation), particularly at the late endosome (LE)/multivesicular body (MVB) and at the trans-Golgi apparatus. For example, at the LE/MVB, the proteins that make up the endosomal sorting complex required for transport (ESCRT) machinery (ESCRT I and ESCRT II) are known to contain ubiquitin-binding domains that enable them to recognise ubiquitinated cargo proteins and sort them into internalised vesicles destined for lysosomal degradation or for secretion events (reviewed in (Hurley, 2010)).

Ubiquitin has an established role in regulating protein relocation and targeted destruction at the proteasome (see (Hershko & Ciechanover, 1998; Hershko, 2005) for some excellent reviews). The three Rho GTPases, Rho1/A, Rac1, and Cdc42 have now been demonstrated to be ubiquitinated and degraded under certain stimuli (de la Vega, *et al*., 2011) (Figure 1&2). Furthermore, both Rac1 and Cdc42 ubiquitination and protein levels are increased when cells are treated with protease inhibitors (Doye, *et al*., 2006). Inactive Rho1/A was shown to be ubiquitinated by the E3 ubiquitin ligase Smurf1 and degraded in migrating Mv1Lu epithelial cells (H. R. Wang, *et al*., 2003). Ubiquitination of Rho1/A may be required to prevent the Rho1/A mediated formation of actin stress fibres at the leading edge of migrating cells, and to allow the Cdc42 and Rac1 mediated dynamic actin rearrangement necessary for anterograde delivery of membranes to the leading edge of migrating cells (Y. Wang, *et al*., 2003). This site

Molecular Machinery Regulating Exocytosis 85

of GRAIL-mediated ubiquitination, RhoGDI inhibition is restricted to Rho1/A, but not Rac1

Sumo is a small ubiquitin-like modifier that, like ubiquitin, can be covalently attached to a protein via an internal lysine residue on the target and can serve to modify its function (for recent reviews see (Wang & Dasso, 2009; Wilkinson & Henley, 2010)). Sumoylation is emerging an additional level of control over the proteins that regulate exocytosis (Figure 2). At least two SNARE proteins are believed to be sumoylated. Sumoylation of the SNARE accessory protein, Tomosyn, relieves its inhibitory effect on SNARE complex assembly, and thereby on exocytosis (Williams, *et al*., 2011). In response to Ca2+ signaling, sumoylation is known to inhibit exocytosis of insulin granules following their docking at the plasma membrane, and this is most likely to occur through SynaptotagminVII (Dai, *et al*., 2011). In addition, the Rho GTPase Rac1 was found to be sumoylated in response to hepatocyte growth factor stimulation of a number of cell lines (HEK293T, MDCKII, HeLa, and Cos7 cells (Castillo-Lluva, *et al*., 2010)). Sumoylation of Rac1 resulted in sustained activation of Rac1 which promoted the formation of lamellipodia and cell motility (Castillo-Lluva, *et al*., 2010). Sumoylation as a post-translational modification of the proteins in the exocytic pathway is a new field of research and will undoubtedly be found to regulate many more of these proteins.

Here we have illustrated that there is a complex array of specialist molecular machinery that is used to control each step in the process of exocytosis. Emerging evidence suggests that there is a highly organised regulatory network required to achieve control of exocytosis. This involves the post-translational modification of the vesicular machinery and membrane associated proteins that orchestrate exocytosis; including the addition of lipid moieties, phosphorylation, and ubiquitination and sumoylation. These post-translational modifications are responsible for mediating protein intracellular localization, proteinprotein interactions, complex assembly, and ultimately protein function. The dynamics and precision of exocytosis often require multiple modifications of a single protein in order to tightly control temporal/spatial function. Moreover, to ensure the harmonious reaction of the cell to a specific stimulation, these post-translational modifications respond to a variety of cell-type specific signaling events. The challenge facing researchers in this field is to investigate the cross-talk between different modifications in the context of a specific signal, and to determine how these are coordinated with other cellular functions. Thus, it is tempting to speculate about an even higher point of control in the regulation of exocytosis, involving proteins that recognise post-translational modifications and facilitate appropriate

Adamson, P., Marshall, C. J., Hall, A. & Tilbrook, P. A. (1992). Post-translational

(October 1992), pp.20033-20038, 0021-9258

modifications of p21rho proteins. *The Journal of biological chemistry*. Vol. 267, No.28

or Cdc42 (Su, *et al*., 2006).

**7. Concluding remarks** 

functional interaction.

**8. References** 

**6.2 Sumoylation** 

specific ubiquitination/degradation of Rho1/A appears to be restricted to the lamellipodia and filopodia of the leading edge, where Smurf1 is recruited through atypical protein kinase C zeta (aPKC)-mediated phosphorylation (H. R. Wang, *et al*., 2003). The latter is activated by the Cdc42/Rac1 polarity complex (H. R. Wang, *et al*., 2003). Hence the polarised exocytosis during cell migratory activity is regulated by a hierarchy of post-translational modifications, where ubiquitination of Rho1/A appears to be important for switching between the competing actin modifying activities of Rho1/A and Cdc42/Rac1, which in turn controls phosphorylation dependent ubiquitination activity of Smurf, and thereby Rho degradation. A further level of complexity is added by the down-regulation of Rho1/A in migrating cells by another E3 ligase, the Cul3/BACURD complex (Chen, *et al*., 2009). The Cul3/BACURD complex is a ring finger E3 ubiquitin ligase complex that has been shown to ubiquitinate Rho1/A in a diverse range of organisms from human cell lines (293T and HeLa fibroblasts), insect cells (*Drosophila melanogaster* S2 cells) and amphibians (*Xenopus laevis* embryos). Depletion of the Cul3 and BACURD ligase complex by siRNA results in defective migration of HeLa cells and mouse embryonic fibroblasts, and in embryonic abnormalities resulting from defective cell migration in *Xenopus* embryos (Chen, *et al*., 2009). There is also evidence for the ubiquitination of Rac1 by the ubiquitin E3 ligase POSH2 but the purpose of this ubiquitination is not yet clear (Karkkainen, *et al*., 2010). A proteasomal degradation resistant and thus constitutively active mutant of Rac1 (Rac1b), is found in colorectal and breast cancer tumour cells (Jordan, *et al*., 1999; Schnelzer, *et al*., 2000). Interestingly, RNAi mediated silencing of this mutant results in a failure of cancer cells to undergo an epithelial to mesenchymal transition (Radisky, *et al*., 2005) suggesting a role for ubiquitination of Rac1 in controlling cell motility (Visvikis, *et al*., 2008).

Rho GTPases can also be regulated by the ubiquitination of their GEF activators. Activation of Rho1/A via ubiquitination of PDZ-RhoGEF, was found to be initiated by Cul3/KLHL20 (Lin, *et al*., 2011). Likewise, Cdc42 can be activated via ubiquitination of its GEF, hPEM-2 (Yamaguchi, *et al*., 2008). It is yet to be established whether Smurf–mediated regulation these two Rho GTPases occurs in a coordinated manner. In addition to Smurf, Cdc42 activity could be regulated by ubiquitination and or proteasomal degradation of its GEFs, FGD1 and FGD3 by E3 ligase SCFFWD1/β-TrCP (Hayakawa, *et al*., 2005; Hayakawa, *et al*., 2008). There is an interesting interplay between the regulatory effects of ubiquitination and phosphorylation. At the leading edge of migrating cells, SCFFWD1/β-TrCP ligase recognises only forms of GEFs inactivated by GSK-3 kinase phosphorylation; the latter kinase could in turn be inactivated by aPKC-mediated phosphorylation (Etienne-Manneville & Hall, 2003; H. R. Wang, *et al*., 2003). Therefore, phosphorylation by aPKC appears to be at the nexus of regulation of ubiquitination of small Rho GTPases, promoting degradation in Rho1/A and preventing it in CDC42. Finally, two of the small Rab GTPase GEFs, Rabex5 and Rabring7 (Xu, *et al*., 2010; Yan, *et al*., 2010) are known to have ubiquitin E3 ligase activity (Sakane, *et al*., 2007), and Rabex5 cellular localisation is regulated by its ability to bind a ubiquitin signal (Mattera, *et al*., 2006). As yet there are no known Rab proteins that are themselves ubiquitinated.

Ubiquitination of the negative regulators of small Rho GTPases, RhoGDIs, has also been shown. RhoGDI is ubiquitinated by the E3 ubiquitin ligase GRAIL which, while not resulting in its proteasomal degradation, did appear to increase the stability the RhoGDI protein (Su, *et al*., 2006). This results in sequestration of Rho molecules in the cytosol, blocking their activation and initiation of the Rho signaling pathway, and thereby impairing cytoskeletal polarization or actin polymerization. It is yet to be defined why, in the context of GRAIL-mediated ubiquitination, RhoGDI inhibition is restricted to Rho1/A, but not Rac1 or Cdc42 (Su, *et al*., 2006).

#### **6.2 Sumoylation**

84 Crosstalk and Integration of Membrane Trafficking Pathways

specific ubiquitination/degradation of Rho1/A appears to be restricted to the lamellipodia and filopodia of the leading edge, where Smurf1 is recruited through atypical protein kinase C zeta (aPKC)-mediated phosphorylation (H. R. Wang, *et al*., 2003). The latter is activated by the Cdc42/Rac1 polarity complex (H. R. Wang, *et al*., 2003). Hence the polarised exocytosis during cell migratory activity is regulated by a hierarchy of post-translational modifications, where ubiquitination of Rho1/A appears to be important for switching between the competing actin modifying activities of Rho1/A and Cdc42/Rac1, which in turn controls phosphorylation dependent ubiquitination activity of Smurf, and thereby Rho degradation. A further level of complexity is added by the down-regulation of Rho1/A in migrating cells by another E3 ligase, the Cul3/BACURD complex (Chen, *et al*., 2009). The Cul3/BACURD complex is a ring finger E3 ubiquitin ligase complex that has been shown to ubiquitinate Rho1/A in a diverse range of organisms from human cell lines (293T and HeLa fibroblasts), insect cells (*Drosophila melanogaster* S2 cells) and amphibians (*Xenopus laevis* embryos). Depletion of the Cul3 and BACURD ligase complex by siRNA results in defective migration of HeLa cells and mouse embryonic fibroblasts, and in embryonic abnormalities resulting from defective cell migration in *Xenopus* embryos (Chen, *et al*., 2009). There is also evidence for the ubiquitination of Rac1 by the ubiquitin E3 ligase POSH2 but the purpose of this ubiquitination is not yet clear (Karkkainen, *et al*., 2010). A proteasomal degradation resistant and thus constitutively active mutant of Rac1 (Rac1b), is found in colorectal and breast cancer tumour cells (Jordan, *et al*., 1999; Schnelzer, *et al*., 2000). Interestingly, RNAi mediated silencing of this mutant results in a failure of cancer cells to undergo an epithelial to mesenchymal transition (Radisky, *et al*., 2005) suggesting a role for ubiquitination of Rac1 in controlling cell motility (Visvikis, *et al*., 2008).

Rho GTPases can also be regulated by the ubiquitination of their GEF activators. Activation of Rho1/A via ubiquitination of PDZ-RhoGEF, was found to be initiated by Cul3/KLHL20 (Lin, *et al*., 2011). Likewise, Cdc42 can be activated via ubiquitination of its GEF, hPEM-2 (Yamaguchi, *et al*., 2008). It is yet to be established whether Smurf–mediated regulation these two Rho GTPases occurs in a coordinated manner. In addition to Smurf, Cdc42 activity could be regulated by ubiquitination and or proteasomal degradation of its GEFs, FGD1 and FGD3 by E3 ligase SCFFWD1/β-TrCP (Hayakawa, *et al*., 2005; Hayakawa, *et al*., 2008). There is an interesting interplay between the regulatory effects of ubiquitination and phosphorylation. At the leading edge of migrating cells, SCFFWD1/β-TrCP ligase recognises only forms of GEFs inactivated by GSK-3 kinase phosphorylation; the latter kinase could in turn be inactivated by aPKC-mediated phosphorylation (Etienne-Manneville & Hall, 2003; H. R. Wang, *et al*., 2003). Therefore, phosphorylation by aPKC appears to be at the nexus of regulation of ubiquitination of small Rho GTPases, promoting degradation in Rho1/A and preventing it in CDC42. Finally, two of the small Rab GTPase GEFs, Rabex5 and Rabring7 (Xu, *et al*., 2010; Yan, *et al*., 2010) are known to have ubiquitin E3 ligase activity (Sakane, *et al*., 2007), and Rabex5 cellular localisation is regulated by its ability to bind a ubiquitin signal (Mattera, *et* 

*al*., 2006). As yet there are no known Rab proteins that are themselves ubiquitinated.

Ubiquitination of the negative regulators of small Rho GTPases, RhoGDIs, has also been shown. RhoGDI is ubiquitinated by the E3 ubiquitin ligase GRAIL which, while not resulting in its proteasomal degradation, did appear to increase the stability the RhoGDI protein (Su, *et al*., 2006). This results in sequestration of Rho molecules in the cytosol, blocking their activation and initiation of the Rho signaling pathway, and thereby impairing cytoskeletal polarization or actin polymerization. It is yet to be defined why, in the context Sumo is a small ubiquitin-like modifier that, like ubiquitin, can be covalently attached to a protein via an internal lysine residue on the target and can serve to modify its function (for recent reviews see (Wang & Dasso, 2009; Wilkinson & Henley, 2010)). Sumoylation is emerging an additional level of control over the proteins that regulate exocytosis (Figure 2). At least two SNARE proteins are believed to be sumoylated. Sumoylation of the SNARE accessory protein, Tomosyn, relieves its inhibitory effect on SNARE complex assembly, and thereby on exocytosis (Williams, *et al*., 2011). In response to Ca2+ signaling, sumoylation is known to inhibit exocytosis of insulin granules following their docking at the plasma membrane, and this is most likely to occur through SynaptotagminVII (Dai, *et al*., 2011). In addition, the Rho GTPase Rac1 was found to be sumoylated in response to hepatocyte growth factor stimulation of a number of cell lines (HEK293T, MDCKII, HeLa, and Cos7 cells (Castillo-Lluva, *et al*., 2010)). Sumoylation of Rac1 resulted in sustained activation of Rac1 which promoted the formation of lamellipodia and cell motility (Castillo-Lluva, *et al*., 2010). Sumoylation as a post-translational modification of the proteins in the exocytic pathway is a new field of research and will undoubtedly be found to regulate many more of these proteins.
