Actin-Binding Proteins in Nerve Cell Growth Cones

Transcription

1 J Pharmacol Sci 105, 6 11 (2007) Journal of Pharmacological Sciences 2007 The Japanese Pharmacological Society Current Perspective Actin-Binding Proteins in Nerve Cell Growth Cones Ryoki Ishikawa 1, * and Kazuhiro Kohama 1 1 Department of Molecular and Cellular Pharmacology, Gunma University Graduate School of Medicine, Maebashi, Gunma , Japan Received June 25, 2007 Abstract. The motility of the growth cone, an intracellular apparatus located at the tip of the axon in developing neurons, is thought to govern axonal path-finding and the construction of neuronal networks. Growth cones contain an actin-rich cytoskeleton, and their dynamics are regulated by a wide variety of actin-binding proteins and motor proteins. In this review, we will focus on the principal functions of these proteins, their mutual interactions in vitro, and their possible roles in the dynamics of nerve cell growth cones. Keywords: actin, myosin, actin-binding protein, growth cone dynamics, neurite elongation Actin organization in neuronal growth cones At the periphery of growth cones (P-region), there exist highly dynamic actin organizations, the lamellipodia and filopodia (Fig. 1). The former are composed of a flat-sheet network of actin filaments (F-actin), while the latter are composed of packed bundles of F-actin. Both structures have polarity, with the barbed end (or fastgrowing end) of F-actin oriented toward the tip and the pointed end (or slow-glowing end) oriented toward the basal region. Globular actins (G-actin) are continuously incorporated into the barbed end and depolymerize and/or disassemble from the pointed end, resulting in the translocation of actin molecules in lamellipodia /filopodia from the tip to the basal region, a process known as the retrograde flow of actin (1). The balance between actin polymerization at the tip and loss at the basal region has been suggested to govern the forward/retractive movement of growth cones. The center of a growth cone (C-region) is thicker than its P-region and is thought to function as a terminal for vesicles and macromolecules transported along axonal microtubules. Adhesion plaques and some stress-fiber-like structures are observed in this area, which are more static than the peripheral actin cytoskeleton. *Corresponding author. ryoki1@med.gunma-u.ac.jp Published online in J-STAGE: September 8, 2007 doi: /jphs.CP Invited article Roles of structural proteins in the formation of lamellipodia and filopodia Arp2/3 is a 230-kDa macromolecular complex, which contains two actin-homologous subunits (actin-related protein 2: Arp2 and actin-related protein 3: Arp3) and five different subunits, is located in the peripheral portion of lamellipodia and is thought to be responsible for the branching formation of F-actin (2). When Arp2/3 is incorporated into F-actin, Arp2 and Arp3 molecules can function as seeds for actin polymerization, that is, G-actin can bind and start to polymerize from Arp2 and from Arp3, resulting in branching of F-actin. As a result of the position of Arp2 and Arp3 in the complex, both newly generated ends are barbed ends, with an angle between mother and daughter filaments of 70 degrees (3). Because Arp2/3 increases the number of barbed ends in a concentration-dependent fashion, it greatly enhances actin polymerization in lamellipodia. Fascin is a 55-kDa actin bundling protein localized in filopodia (4). The in vitro reconstituted structure of actin/fascin bundles is quite similar to the in vivo structure of actin bundles in filopodia. It bundles F-actin with a distance between cross-linked filaments in a bundle of 8 9 nm in vitro (5). Furthermore, actin bundles formed by fascin have polarity, that is, all the filaments in the bundle face the same direction. Fascin is thought to eliminate Arp2/3 from lamellipodia and fasten F-actin to form filopodia (6), although the mechanism of initiation of filopodia formation is still unclear. 6

2 Actin Cytoskeleton in Growth Cone 7 Fig. 1. Possible mechanisms of actin dynamics in nerve cell growth cone. Mechanism and accelerators of actin treadmilling As observed in retrograde flow in vivo, a similar flow of actin molecules in F-actin, termed treadmilling, is observed in vitro. The affinity of G-actin for the barbed end of F-actin is about ten-fold that for the pointed end of F-actin at physiological salt concentration, reflecting the difference in speed of polymerization between ends. When actin polymerization reaches equilibrium, the concentration of G-actin in solution is higher than the dissociation constant (K d ) of the barbed end but lower than the Kd of the pointed end. Therefore, G-actin is continuously incorporated into the barbed end and dissociates from the pointed end, although the net length of the F-actin filament is unchanged. This condition is termed treadmilling. The actin molecule has a nucleotide-binding pocket and can associate with ATP or ADP. The affinity of ATP-bound G-actin for F-actin

3 8 R Ishikawa and K Kohama filament is much higher than that of ADP-bound G- actin. Furthermore, ATP which has been incorporated into F-actin as a form of ATP-bound G-actin is hydrolyzed to ADP inside the filament. ATP-bound actins thus concentrate on the barbed-end side while ADP-bound actins concentrate on the pointed-end side when treadmilling occurs. It should be noted that even when actin polymerization has reached equilibrium, the chemical energy of ATP is consumed and used for actin dynamics in treadmilling. Treadmilling appears to be a fundamental model of actin dynamics in retrograde flow, although the speed of treadmilling is much slower than that of retrograde flow. How is the speed of treadmilling enhanced? N-WASP, in combination with Arp2/ 3, greatly enhances actin polymerization into the barbed end of the filament (2, 7). N-WASP is a 65-kDa neural protein that has two G-actin binding sites, one Arp2/3 binding site, and one profilin binding site. In the inactivated state, actin-binding and Arp2/3-binding sites are covered by an internal inhibitory domain. When cdc42 and PIP2 bind to N-WASP, both sites are uncovered, resulting in acceleration of actin polymerization (2, 7). WAVE, a kda protein whose C-terminal domain is homologous to that of N-WASP, has one G-actin binding site, one Arp2/3 binding site, and one profilin binding site; and it promotes actin polymerization (7). Because N-WASP is present throughout the lamellipodia while WAVE is concentrated in the leading edge of migrating lamellipodia, WAVE may play a more important role in actin polymerization in lamellipodia in growth cones (7). The mammalian homologue of formin homology protein family (mdir/formin) is located at the tip of filopodia and accelerates actin nucleation and remains at the barbed end of the filament (8). It has weak capping activity, but unlike gelsolin and capping protein, the capped-filament can still elongate (8). The mdir/formin attached to the barbed end of the filament moves processively along the F-actin filament and remains at the barbed end (9). The Ena/VASP family protein, a G-actin and F-actin binding protein, is located at the tip of filopodia and the leading edge of lamellipodia and greatly enhances actin polymerization in the presence of capping proteins, perhaps because it inhibits the activity of capping protein (10). Other factors promoting actin polymerization are profilin and ADF/ cofilin. Profilin is a G-actin binding protein with a molecular weight of kda. It is localized in the leading edge of growth cones. Profilin enhances the release of ADP from G-actin, resulting in acceleration of binding of ATP to G-actin. As noted above, ATP-bound G-actin has ten-fold greater affinity for the barbed end than ADP-bound G-actin, and profilin greatly enhances actin polymerization (11). However, at high concentrations, profilin functions as a G-actin sequestering protein and promotes the depolymerization of F-actin. ADF/cofilin is an F-actin/G-actin binding protein, and it decreases the stability of F-actin by twisting the helix of the actin filament (12). It has higher affinity for ADP-bound actin than for ATP-bound actin. ADF/cofilin thus greatly enhances the depolymerization of F-actin from the pointed end of the filament since ADP-bound actin molecules concentrate on the pointed end side of F-actin (12). ADF/cofilin loses activity when ser-3 is phosphorylated by LIM kinase (13). Other actin binding proteins in nerve cell growth cones Tropomyosin, a rod-shaped homomeric actin-binding protein, is located in filopodia and the stress-fiber-like structure in the growth cone (14). It inhibits the spontaneous severing of F-actin in vitro and appears to stabilize actin organization. Caldesmon, a rod-shaped F-actin-binding and calmodulin-binding protein, is localized in the growth cone (15). Caldesmon dissociates from F-actin in the presence of Ca 2+ /calmodulin (15) and may thus function as a Ca 2+ sensor in growth cone dynamics. Drebrin-E is a brain-specific 73-kDa actin-binding protein located in adhesion plaques and the basal region of filopodia in growth cones (16). As discussed below, it appears likely that drebrin-e attenuates actin dynamics by inhibiting the activities of factors promoting actin dynamics in growth cones. Gelsolin, an 82-kDa globular protein, exists in growth cones. It binds to F-actin and severs it, and then it remains and caps the barbed end of F-actin to inhibit actin polymerization (17). Gelsolin also has G-actinbinding activity and enhances the nucleation step of actin polymerization (17). Dendritic filopodia-like structures appear in growth cones and dendritic shafts in cultured neurons from gelsolin knockout mice (18), suggesting that gelsolin may function as an F-actin destroyer in growth cone dynamics. Role of myosins and their regulators in growth cone dynamics The major myosins in neuronal growth cones are myosin-ii and myosin-v. Myosin II, also termed conventional myosin or just myosin, appears to affect growth cone dynamics. Disruption of myosin-ii stalls retrograde flow, resulting in elongation of filopodia (19). This finding suggests that myosin-ii may propel, at least in part, the retrograde flow of filopodial actin, although

4 Actin Cytoskeleton in Growth Cone 9 the principal mechanism of retrograde flow involves treadmilling of actin polymerization. However, myosin- II exists as two subtypes in nonmuscle cells, myosin-iia and myosin-iib, which appear likely to have different functions in growth cones. Myosin-IIA is located in adhesion plaques in C-regions while myosin-iib is distributed in patches along actin filaments in both P- and C-regions in growth cones. Disruption of myosin- IIA expression by antisense RNA weakened the adhesion of growth cones but had no effects on neurite outgrowth in a neuroblastoma cell line (20). On the other hand, knockdown of myosin-iib inhibits neurite outgrowth but has no effects on cell attachment (20, 21). Myosin-IIA may thus govern cell attachment, while myosin-iib may govern growth cone dynamics. Myosin-V is a two-headed myosin located in filopodia, membrane vesicles, and plasma membrane in growth cones; and it appears to be responsible for vesicle and macromolecule transport. It has a vesicle-binding site in its tail and has a higher affinity for F-actin than myosin-ii in the presence of ATP (22). Thus, a single myosin-v molecule can slide on F-actin without detaching from it during the ATPase cycle (23). Myosin- V can thus transport vesicles along the actin cytoskeleton in growth cones. Myosin-V also appears to affect the dynamics of filopodia. Disruption of myosin- V decreases the length of filopodia in cultured DRG neurons (24), although the finding that cultured SCG neurons from myosin-va knockout mice exhibit normal growth cone morphology remains controversial. Another minor myosin, myosin-x, which can bind to β-integrin through its tail, is concentrated in the tips of filopodia. Overexpression of myosin-x increases the number and length of filopodia, suggesting that myosin- X governs filopodia dynamics in nerve cell growth cones (25). Mutual interaction of cytoskeletal proteins One of the principal purposes of this review is to focus on mutual interaction of cytoskeletal proteins in nerve cell growth cones, which is summarized in Table 1. Among these proteins, tropomyosin and caldesmon affect the activities of a wide variety of other actinbinding proteins. Tropomyosin inhibits the actin binding of fascin, ADF/cofilin, filamin, and α-actinin in vitro. It also partially inhibits the actin-severing activity of gelsolin (26). Tropomyosin thus appears to maintain a single F-actin filament by protecting it from crosslinking, severing, or destabilizing. However, the affinity of tropomyosin for F-actin is greatly enhanced by caldesmon. When both tropomyosin and caldesmon fully bind to F-actin, gelsolin activity is completely Table 1. cones Mutual interaction of actin-binding proteins in growth Protein Inhibition Activation Tropomyosin Fascin Caldesmon ADF/cofilin α-actinin Filamin Caldesmon Arp2/ 3 Tropomyosin Tropomyosin & Caldesmon a Gelsolin Drebrin-E Myosin-II Drebrin-E Tropomyosin Caldesmon ADF/cofilin Fascin α-actinin Myosin-II Myosin-V Profilin N-WASP WAVE mdir/fromin Ena/VASP a Clear inhibitory effects are observed only in the presence of both tropomyosin and caldesmon. blocked (26). Furthermore, they inhibit the activity of myosin-ii and the actin binding of drebrin-e in vitro. These inhibitions are released in the presence of Ca 2+ /calmodulin. Tropomyosin/caldesmon may thus function as a Ca 2+ sensor of actin dynamics in growth cones through regulating the activities of other actinbinding proteins. Caldesmon has also been reported to inhibit the Arp2/3-mediated nucleation of actin polymerization (27). Drebrin-E also affects the activities of a wide variety of actin-binding proteins. We previously reported that it inhibits the actin-binding and cross-linking activities of fascin and α-actinin (16), the actin binding of tropomyosin (28), and the activities of myosin-ii in vitro. Furthermore, we recently found that drebrin inhibits the activities of myosin-v (29) and cancels ADF/cofilindependent promotion of actin-turnover by inhibiting the binding of ADF/ cofilin to F-actin (30). Since myosins and ADF/cofilin appear to enhance the actin dynamics in growth cones, drebrin-e may shift the state of the actin cytoskeleton from dynamic to static. N-WASP (2, 7), WAVE (7), mdir/forming (8), and Ena/VASP (10) have profilin binding sites. Based on the synergistic promotion of actin polymerization by these proteins and profilin, recruitment of G-actin / profilin complex to sites of polymerization has been

5 10 R Ishikawa and K Kohama suggested to be one of the mechanisms of acceleration of actin polymerization by these proteins. Concluding remark What are the principal factors in actin dynamics? We believe that there are no star players and that instead, many cytoskeletal proteins harmoniously regulate actin organization in growth cones (Fig. 1). At the tip of the growth cone, N-WASP, WAVE, mdir/ formin, and Ena/VASP accelerate actin polymerization. Arp2/3 branches lamellipodial F-actin to form lamellipodia, and fascin fastens F-actin to form filopodia. Myosins pull actin bundles backward to drive retrograde flow. Near the border of the P- and C-regions, ADF/cofilin destabilizes F-actin, tropomyosin and drebrin-e loosen the filopodial bundles, and gelsolin severs F-actin to depolymerize/disassemble filopodia and lamellipodia. Profilin accelerates ADP/ATP exchange of G-actin, and the profilin/g-actin complex is re-incorporated into the actin cytoskeleton at the tip of the growth cone. Myosin- IIA and α-actinin enhance adhesion of the growth cone. Myosin-I and myosin-v transport membrane vesicles and macromolecules. The activity of all of these proteins is coordinated well: they are sometimes activated, sometimes inactivated, sometimes work together, and sometimes compete, resulting in regulation of growth cones dynamics. It is quite tempting to speculate that the spatiotemporal localization and activities of the actin cytoskeleton are governed not only by individual regulation of the activities of proteins, but also by the mutual interactions of each protein. Further characterization of the in vivo roles of modes of interaction of cytoskeletal proteins suggested by in vitro experiments remains to be performed. References 1 Mallavarapu A, Mitchison T. Regulated actin cytoskeleton assembly at filopodium tips controls their extension and retraction. J Cell Biol. 1999;146: Millard TH, Sharp SJ, Machesky LM. Signalling to actin assembly via the WASP (Wiskott-Aldrich syndrome protein)- family proteins and the Arp2/ 3 complex. Biochem J. 2004; 380: Mullins RD, Heuser JA, Pollrd TD. The interaction of Arp2/ 3 complex with actin: Nucleation, high affinity pointed end capping, and formation of branching networks of filaments. Proc Natl Acad Sci U S A. 1998;95: Kureishy N, Sapountzi V, Prag S, Anilkumar N, Adams JC. Fascins, and their roles in cell structure and function. Bioessays. 2002;24: Ishikawa R, Sakamoto T, Ando T, Higashi-Fujime S, Kohama K. Polarized actin bundles formed by human fascin-1: their sliding and disassembly on myosin II and myosin V in vitro. J Neurochem. 2003;87: Vignjevic D, Kojima S, Aratyn Y, Danciu O, Svitkina T, Borisy GG. Role of fascin in filopodia protrusion. J Cell Biol. 2006;174: Takenawa T, Miki H. WASP and WAVE family proteins: key molecules for rapid rearrangement of cortical actin filament and cell movement. J Cell Sci. 2001;114: Kover DR. Molecular details of formin-mediated actin assembly. Curr Opin Cell Biol. 2006;18: Higashida C, Miyoshi T, Fujita A, Oceguera-Yanez F, Monypenny J, Andou Y, et al. Actin-polymerization-driven molecular movement of mdir1 in living cells. Science. 2004;303: Kwiatkowski AV, Gertler FG, Loureiro JJ. Function and regulation of Ena/VASP proteins. Trends Cell Biol. 2003;13: Pantaloni D, Carlier MF. How profilin promotes actin filament assembly in the presence of thymosin β4. Cell. 1993;75: Carlier MF, Laurent V, Santolini J, Melki R, Didry D, Xia GX, et al. Actin depolymerizing factor (ADF/ cofilin) enhances the rate of filament turnover: implication in actin-based motility. J Cell Biol. 1997;136: Sarmiere PD, Bamburg JR. Regulation of the neuronal actin cytoskeleton by ADF/cofilin. J Neurobiol. 2004;58: Gunning PW, Schevzov G, Kee AJ, Hardeman EC. Tropomyosin isoforms: divining rods for actin cytoskeleton function. Trends Cell Biol. 2005;15: Sobue K, Sellers JR. Caldesmon, a novel regulator protein in smooth muscle and nonmuscle actomyosin system. J Biol Chem. 1991;266: Sasaki Y, Hayashi K, Shirao T, Ishikawa R, Kohama K. Inhibition by drebrin of the actin-bundling activity of brain fascin, a protein localized in filopodia of growth cones. J Neurochem. 1996;66: Sun HQ, Yamamoto M, Mejillano M, Yin HL. Gelsolin, a multifunctional actin regulatory protein. J Biol Chem. 1999; 274: Lu M, Witke W, Kwiatkowski DJ, Kosik KS. Delayed retraction of filopodia in gelsolin null mice. J Cell Biol. 1997,138; Lin CH, Espreafico EM, Mooseker MS, Forscher P. Myosin drives retrograde F-actin flow in neuronal growth cones. Neuron. 1996;16: Wylie SR, Chantler PD. Separate but linked functions of conventional myosins modulate adhesion and neurite outgrowth. Nat Cell Biol. 2001;3: Bridgman PC, Dave S, Asnes CF, Tullio AN, Adelstein RS. Myosin IIB is required for growth cone motility. J Neurosci. 2001;21: De La Cruz EM, Well AL, Rosenfeld SS, Ostap EM, Sweeney HL. The kinetic mechanism of myosin V. Proc Natl Acad Sci U S A. 1999;96: Mehta AD, Rock RS, Rief M, Spudich JA, Mooseker MS, Cheney RE, Myosin-V is a processive actin-based motor. Nature. 1999;400: Wang FS, Wolenski JS, Cheney RF, Mooseker MS, Jay DG. Function of myosin-v in filopodial extension of neuronal growth cones. Science. 1996;273: Bohil AB, Robertson BW, Cheney RE. Myosin-X is a molecular

6 Actin Cytoskeleton in Growth Cone 11 motor that functions in filopodia frormation. Proc Natl Acad Sci U S A. 2006;103: Ishikawa R, Yamashiro S, Matsumura F. Differential modulation of actin-severing activity of gelsolin by multiple isoforms of cultured rat cell tropomyosin: Potentiation of protective ability of tropomyosin by 83-kDa nonmuscle caldesmon. J Biol Chem. 1989;264: Yamakita Y, Oosawa F, Yamashiro S, Matsumura F. Caldesmon inhibits Arp2/ 3-mediated actin nucleation. J Biol Chem. 2003;278: Ishikawa R, Hyashi K, Shirao T, Xue Y, Takagi T, Sasaki Y, et al. Drebrin, a developmental-associated brain protein from rat embryo, causes the dissociation of tropomyosin from actin filaments. J Biol Chem. 1994;269: Ishikawa R, Katoh K, Takahashi A, Xie C, Oseki K, Watanabe M, et al. Drebrin attenuates the interaction between actin and myosin-v. Biochm Biophys Res Commun. 2007;359: Ishikawa R, Nakamura A, Kohama K. Biochemical aspect of actin-binding proteins in nerve cell growth cones. J Pharmacol Sci. 2007;103 Suppl 1:47P.