The Globular Tail Domain of Myosin Va Functions as an Inhibitor of the Myosin Va Motor*

The actin-activated ATPase activity of full-length mammalian myosin Va is well regulated by Ca2+, whereas that of truncated myosin Va without the C-terminal globular tail domain (GTD) is not. Here, we have found that exogenous GTD is capable of inhibiting the actin-activated ATPase activity of GTD-deleted myosin Va. A series of truncated constructs of myosin Va further showed that the entire length of the first coiled-coil (coil-1) of the tail domain is critical for GTD-dependent regulation of myosin Va and that deletion of 58 residues from the C-terminal end of coil-1 markedly hampered regulation. Negative staining electron microscopy revealed that GTD-deleted myosin Va formed a “Y”-shaped structure, which was converted to a triangular shape, similar to the structure of full-length myosin Va in the inhibited state, by addition of exogenous GTD. In contrast, the triangular shape was not observed when the C-terminal 58 residues of coil-1 were deleted, even in the presence of exogenous GTD. Based on these results, we propose a model for the formation of the inhibited state of myosin Va. GTD binds to the C-terminal end of coil-1. The neck-tail junction of myosin Va is flexible, and the long neck enables the head domain to reach the GTD associated with the end of coil-1. Once the head interacts with the GTD, the triangular inhibited conformation is stabilized. Consistent with this model, we found that shortening of the neck of myosin Va by two IQ motifs abolished the regulation by GTD, whereas regulation was partially restored by shortening of coil-1 by an amount comparable to that of the two IQ motifs.

The actin-activated ATPase activity of full-length mammalian myosin Va is well regulated by Ca 2؉ , whereas that of truncated myosin Va without the C-terminal globular tail domain (GTD) is not. Here, we have found that exogenous GTD is capable of inhibiting the actin-activated ATPase activity of GTDdeleted myosin Va. A series of truncated constructs of myosin Va further showed that the entire length of the first coiled-coil (coil-1) of the tail domain is critical for GTD-dependent regulation of myosin Va and that deletion of 58 residues from the C-terminal end of coil-1 markedly hampered regulation. Negative staining electron microscopy revealed that GTD-deleted myosin Va formed a "Y"-shaped structure, which was converted to a triangular shape, similar to the structure of full-length myosin Va in the inhibited state, by addition of exogenous GTD. In contrast, the triangular shape was not observed when the C-terminal 58 residues of coil-1 were deleted, even in the presence of exogenous GTD. Based on these results, we propose a model for the formation of the inhibited state of myosin Va. GTD binds to the C-terminal end of coil-1. The neck-tail junction of myosin Va is flexible, and the long neck enables the head domain to reach the GTD associated with the end of coil-1. Once the head interacts with the GTD, the triangular inhibited conformation is stabilized. Consistent with this model, we found that shortening of the neck of myosin Va by two IQ motifs abolished the regulation by GTD, whereas regulation was partially restored by shortening of coil-1 by an amount comparable to that of the two IQ motifs.
Class V myosin is considered one of the oldest classes of myosin, being distributed from low eukaryotes, such as yeast, to vertebrate cells (1,2). In mammalian cells, there are three distinct subclasses of myosin V, named myosin Va, Vb, and Vc (3)(4)(5). Myosin Va is one of the best characterized motor proteins of the myosin superfamily. Myosin Va is a processive motor, which undergoes multiple catalytic cycles coupled to multiple mechanical advances for each diffusional encounter with its actin track (6 -10). This property allows single or a few myosin Va molecules to support movements of an organelle along actin filaments with a large step approximating the 36-nm pseudo repeat of the track.
Myosin Va consists of two identical heavy chains that dimerize through the formation of a coiled-coil structure to form a homodimer. At the N terminus is the motor domain containing the ATP and actin binding sites. The motor domain is followed by a neck that consists of six IQ motifs with the consensus sequence IQXXXRGXXXR, which act as the binding sites for calmodulins or myosin light chains. The next ϳ500 amino acids are predicted to form a series of coiled-coils separated by several flexible regions. The last ϳ400 amino acids form a globular tail domain (3). This C-terminal globular tail domain (GTD), 2 in conjunction with a portion of the coiled-coil region, mediates myosin Va binding to specific membrane-bound organelles such as melanosomes (11).
One of the most important questions is how the motor function of myosin Va is regulated. If myosin Va were constantly active, then the high concentrations of myosin Va found in brain (3) would consume substantial amounts of ATP. Therefore, it is desirable that the mechanoenzymatic activities of myosin Va is regulated in vivo. However, the molecular mechanism of myosin Va regulation is not well understood.
The ATPase activities of tissue-isolated myosin Va (3, 12) and baculovirus-expressed full-length myosin Va (13,14) are well regulated, the actin-activated ATPase activity of myosin Va being significantly increased by micromolar concentrations of Ca 2ϩ . Sedimentation velocity analysis showed that the myosin Va undergoes a Ca 2ϩ -induced conformational transition from 14 S to 11 S (12-14). Electron microscopy revealed that at low ionic strength, myosin Va has an extended conformation in high Ca 2ϩ whereas it forms a folded shape in the presence of EGTA, in which the tail domain is folded back toward the head (12). The conformational transition is closely correlated with activation of the actin-activated ATPase activity of myosin Va (13). On the other hand, truncated myosin Va without the globular tail domain is not regulated by Ca 2ϩ (15,16) and does not undergo a large conformational transition like full-length myosin Va (13,14). These results suggest that the tail domain plays a key role in the Ca 2ϩ -dependent regulation of myosin Va, by inhibiting the ATPase of the head (12)(13)(14). In this model, myosin Va in the inhibited state is in a folded conformation such * This work was supported by an American Heart Association Scientist development grant (to X.-D. L.), National Institutes of Health Grants AR41653, AR048526, and DC006103 (to M. I.), and National Institutes of Health Grant AR34711 (to R. C.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1  that the tail domain interacts with and inhibits the myosin Va motor activity. Cargo binding, high Ca 2ϩ , and/or phosphorylation may reduce the interaction between the head and tail domains, thus activating its activity. Consistent with this model, we recently found that the actinactivated ATPase activity of melanocyte-type myosin Va is significantly activated by the binding of melanophilin (17), an organelle receptor for myosin Va in mouse melanocytes (18 -21). Because myosin Va binds to melanophilin through its tail domain, we proposed that this interaction interferes with the interaction between the tail and head domains of myosin Va, thus relieving the inhibitory effect of the tail domain (17).
In this study, we further investigated the molecular mechanism of the regulation of the motor activity of myosin Va. Our results indicate that the C-terminal globular tail domain of myosin Va directly binds to the head domain, thus inhibiting motor function. Furthermore, our results suggest that the lengths of the neck domain and the first long coiled-coil domain structurally coordinate the binding between the head and the tail domains, and thus are critical for the regulation of myosin Va motor activity.

EXPERIMENTAL PROCEDURES
Materials-Restriction enzymes and modifying enzymes were purchased from New England Biolabs (Beverly, MA), unless indicated otherwise. Pfu Ultra High-Fidelity DNA polymerase was purchased from Stratagene (La Jolla, CA). Oligonucleotides were synthesized by Invitrogen (Carlsbad, CA). Actin was prepared from rabbit skeletal muscle acetone powder according to Spudich and Watt (22). Recombinant calmodulin of Xenopus oocyte (23) was expressed and purified according to a published method (24). Nickel-nitrilotriacetic acid (Ni-NTA)-agarose and mouse monoclonal antibody recognizing penta-His (anti-His tag) were purchased from Qiagen (Hilden, Germany). Anti-FLAG M2 antibody, anti-FLAG M2 affinity gel, FLAG peptide (Asp-Tyr-Lys-Asp-Asp-Asp-Asp-Lys), phosphoenol pyruvate, 2,4-dinitrophenyl-hydrazine, and pyruvate kinase were from Sigma. Phenyl-Sepharose and Sephacryl S-300 were purchased from Amersham Biosciences. Anti-M5T antibody, an antibody recognizing the C-terminal 22 amino acid residues of myosin Va, was provided by Dr. Jack L. Leonard, Department of Physiology, University of Massachusetts Medical School (25).
Expression and Purification of Myosin Va Constructs-GTD was expressed in Sf9 cells and purified by Ni-NTA-agarose affinity chromatography and size-exclusion chromatography. To express GTD, 400 ml of Sf9 cells (suspension culture) was infected with baculovirus expressing GTD, and cultured in 1 liter of medium at 27°C for 3 days. After harvesting and washing with 4 mM EGTA in TBS (0.15 M NaCl, and 50 mM Tris-HCl, pH 7.5), the cell pellets were lysed with sonication in 50 ml of lysis buffer (50 mM Tris-HCl, pH 7.5, 15 mM imidazole-HCl, pH7.5, 0.3 M NaCl, 5 mM MgCl 2 , 1 mM EGTA, 0.02% NaN 3 , 5 mM ATP, 5 mM 2-mercaptoethanol, 10 g/ml leupeptin, 0.2 mg/ml trypsin inhibitor (Egg), and 0.5% Triton X-100). After centrifugation at 120,000 ϫ g for 30 min, the supernatant was incubated with 4.0 ml of Ni-NTAagarose affinity resin in a 50-ml conical tube on a rotating wheel in a cold room for 2 h. The resin suspension was then loaded on a column (1 ϫ 10 cm) and washed with 30 ml of solution A (50 mM Tris-HCl, pH 7.5, 15 mM imidazole-HCl pH7.5, 0.3 M NaCl, 0.2 mM EGTA, 2 g/ml leupeptin, and 5 mM 2-mercaptoethanol). The protein bound to the resin was eluted with 0.25 M imidazole in solution A. SDS-PAGE showed that the eluate from the Ni-NTA-agarose column contained two bands with apparent molecular masses of 100 kDa and 50 kDa ( Fig.  2A). The calculated molecular mass of GTD is 50 kDa. It therefore appeared that the lower and upper bands in SDS-PAGE might be monomeric and dimeric GTD, respectively. Western blot showed that both bands were recognized by anti-His antibody and anti-M5T antibody (Fig. 2C). To separate the two bands, the eluate from Ni-NTA-agarose chromatography was subjected to size-exclusion chromatography. About 4 ml of the eluate from Ni-NTA-agarose was loaded onto a Sephacryl S-300 column (1.5 ϫ 60 cm) pre-equilibrated with Solution D (0.2 M NaCl, 5 mM Tris-HCl, pH 7.5, 1 mM DTT). The fractions 30 -54 (1.5 ml per fraction) were collected and subjected to SDS-PAGE (Fig. 2B). The fractions containing high concentrations of monomeric GTD were pooled. The concentration of GTD was determined by the absorbance at 280 nm (1 A 280 equal to 1.35 mg/ml GTD). About 10 mg of monomeric GTD could be obtained in one preparation.
The nature of the putative "dimeric GTD" is not clear. Because there are 8 cysteine residues in GTD, we suspected that some GTD might have dimerized by cross-linking through intermolecular disulfide bonds. However, the migration of the "dimeric GTD" in SDS-PAGE did not change even when it was treated with high concentrations of reducing reagents (0.1 M 2-mercaptoethanol, 0.1 M DTT, or 25 mM TCEP, data not shown). In size-exclusion chromatography (Fig. 2B), the dimeric GTD eluted far ahead of IgG (molecular mass of 150 kDa), suggesting that it may be an oligomer of dimeric GTD. Whatever the nature of the 100-kDa band, all following experiments were done with the monomeric GTD.
Full-length myosin Va and the truncated mutants were expressed in Sf9 cells by co-infection with recombinant baculovirus expressing myosin Va and calmodulin. Expressed myosin Va was purified by anti-FLAG M2 affinity gel as described previously (13), except that the purified myosin Va samples were dialyzed against 5 mM Tris-HCl (pH 7.5 at 20°C), 0.2 M NaCl, 1 mM EGTA, 1 mM DTT, and 10% sucrose on ice overnight. The purified myosin Va samples, in an aliquot of 100 l in PCR tube, were quick-frozen in liquid nitrogen and stored at Ϫ80°C. The concentration of myosin Va was determined by Coomassie Brilliant Blue R250 staining of SDS-PAGE gels (7.5-20%) using smooth muscle myosin heavy chain as a standard as described previously (13). To calculate the molar concentration of the expressed protein, the following molecular masses   Electron Microscopy-Rotary metal-shadowing electron microscopy of myosin Va was performed as described previously (13). Briefly, myosin Va samples, diluted to about 4 nM in EGTA-600 solution (1 mM MgCl 2 , 30% glycerol, 1 mM EGTA, 600 mM ammonium acetate, pH 7.0), were absorbed onto a freshly cleaved mica surface for 30 s. Unbound proteins were rinsed away, and then the specimen was stabilized by brief exposure to uranyl acetate (26) before shadowing. For negative staining, full-length or truncated myosin Va was diluted to 20 nM with 0.15 M sodium acetate, 1 mM EGTA, 2 mM MgCl 2 , 10 mM MOPS (pH 7.0) at room temperature. Five microliters of the solution were applied to UV-treated or glowdischarged, carbon-coated grids, and immediately stained with 1% uranyl acetate (27). The same procedure was used for the specimens made by mixing 20 nM truncated myosin Va with 2 M GTD for 10 min. Negatively stained grids were examined in a CM120 electron microscope (FEI, Hillsboro, OR), fitted with an anti-contamination device cooled by liquid nitrogen, and operated at 120 kV. Micrographs were recorded on an F224HD slow scan CCD camera (TVIPS, Gauting, Germany) at a magnification of 65,000 (0.37 nm/pixel).
Other Assays-The ATPase activity of myosin Va was measured using an ATP regeneration system at 25°C as described before (13). Reaction solution, except ATP, was mixed and preincubated at 25°C for 10 min before adding ATP to start the reaction. SDS-PAGE and Western blot were as described previously (13).

RESULTS
Exogenous GTD Inhibits the ATPase Activity of GTD-deleted Myosin Va (McM5⌬T)-As described in the introduction, the actin-activated ATPase activity of truncated myosin Va without the C-terminal GTD is not regulated by Ca 2ϩ , whereas that of full-length myosin Va is (15,16). Regulation is abolished by deleting just the GTD (the construct of McM5⌬T, see Fig. 1) (14). These results suggest that the GTD is necessary to inhibit the actin-activated ATPase activity of myosin Va. To determine whether exogenous GTD inhibits the function of the myosin Va motor domain, we measured the actin-activated ATPase activity of McM5⌬T (Fig. 1) in the presence of exogenous GTD. The inhibition of ATPase activity was dosedependent, and the apparent affinity between McM5⌬T and GTD was 0.88 M (Fig. 3A). The actin-activated ATPase activity of McM5⌬T in the presence of 8 M GTD was similar to that of full-length myosin Va under EGTA conditions, which was 1.75 head Ϫ1 s Ϫ1 under similar conditions (13), suggesting that McM5⌬T was changed to an inhibited state in the presence of exogenous GTD. Similar to full-length myosin Va, the activity of McM5⌬T was not inhibited by GTD under high Ca 2ϩ conditions (Fig. 3A).
The inhibitory effect of GTD on McM5⌬T was very sensitive to ionic strength. As shown in Fig. 3B, with an increase of NaCl from 100 mM to 400 mM, the ATPase activity of McM5⌬T in the absence of GTD sharply decreased, whereas that of McM5⌬T in the presence of GTD was bell-shaped with an increase in ionic strength. Overall, the inhibitory effect of GTD on McM5⌬T decreases with an increase in ionic strength. When the NaCl concentration is over 300 mM, GTD has virtually no inhibitory effect. These results suggest that GTD-dependent inhibition of the ATPase activity of McM5⌬T depends on ionic interaction.

The First Long Coiled-coil Segment Is Essential for Inhibition of the ATPase Activity of Myosin Va Motor Domain-
The motor domain and neck of myosin Va are connected to the GTD by a series of coiled-coil domains interrupted by a number of flexible regions (Fig. 1). The number and length of predicted coiled-coils varies among different isoforms of myosin Va. For McM5, five coiled-coils separated by four flexible regions are predicted by Paircoil (28). To investigate which portion of the tail domain is essential for regulation by GTD, we made a series of C-terminal truncation mutants of myosin Va (Fig. 1), and examined the inhibitory effect of GTD on the actin-activated ATPase activity of the constructs. As shown in Fig. 4, the ATPase activity of M5-1234 was potently inhibited by GTD with an apparent affinity of 1.2 M, which is similar to that of McM5⌬T. The inhibition of the actin-activated ATPase activity of M5-1105 by GTD was slightly less than that of M5-1234. By contrast, the actin-activated ATPase activities of M5-928, M5-975, and M5-1047 were not effectively inhibited by GTD. These results indicate that the first long coiled-coil (coil-1) is critical for regulation, whereas the distal tail portion, i.e. coil-2 to coil-5, is not essential for inhibition by GTD. These results could be explained in three ways. The first is that the specific sequence of the C-terminal portion of coil-1 is essential for regulation by GTD. Second, the double-headed structure is critical for regulation and deletion of the C-terminal portion of coil-1 destabilizes the double-headed structure. Third, the length of coil-1 is critical for GTD-induced inhibition of myosin Va. To examine whether the regulation of truncated myosin Va is related to the double-headed state, we visualized the molecules by rotary metal-shadowing electron microscopy. We found that the majority of the molecules of McM5⌬T, M5-1234, M5-1105, and M5-1047 were double-headed, whereas about 75% of M5-975 and most of M5-928 were singleheaded (Fig. 5). These results suggest that single-headed myosin Va S1 is not regulated. On the other hand, although both M5-1105 and M5-1047 are mostly double-headed, GTD markedly inhibited the actin-activated ATPase activity of M5-1105, but only marginally inhibited that of M5-1047, suggesting that the double-headed structure is not sufficient for the regulation by GTD.
To further clarify the role of the double-headed structure, we constructed M5S1-LZ1, an artificially dimerized myosin Va S1, by replacing the C-terminal 33 amino acid residues of M5-975 with 32 amino acid residues of leucine zipper, producing a continuous, in-register coiled-coil (Fig. 1). As shown in Fig. 5G, the majority of M5S1-LZ1 molecules are double-headed. However, the actin-activated ATPase activity of M5S1-LZ1 is not effectively inhibited by GTD, being similar to that of M5-975 (Fig. 4), indicating that the double-headed structure is not sufficient for the regulation of myosin Va by GTD.
To investigate the role of the coil-1/hinge region on the regulation of myosin Va, we produced three McM5⌬T constructs with various deletions in this region (Fig. 1). Deletion of 1048 -1059 (McM5⌬T⌬2) moderately decreased the inhibitory effect of GTD on the ATPase activity, whereas deletion of 1048 -1105 (McM5⌬T⌬3) or 1048 -1150 (McM5⌬T⌬4) essentially eliminated the inhibitory effect of GTD (Fig. 6). The results support the idea that the C-terminal portion of coil-1 is important for GTD-induced inhibition of ATPase activity.
In the Presence of GTD, Truncated Myosin Va Forms a Triangular Shape, Which Is Similar to Full-length Myosin Va-It was reported that full-length myosin Va forms a triangular shape at low Ca 2ϩ and low ionic strength (12). This suggested that GTD directly contacts the motor domain, thus inhibiting its motor function. Because exogenous GTD inhibited the ATPase activity of McM5⌬T, we examined whether exogenous GTD and McM5⌬T formed this triangular shape. As shown in Fig. 7B, McM5⌬T had a T-or Y-shape in the absence of GTD (at 150 mM sodium acetate), but formed a triangular shape in its presence (Fig. 7D), similar to that of full-length myosin Va, McM5 (Fig. 7A).
Similar to McM5⌬T, M5-1105 also formed a triangular shape in the presence of GTD (Fig. 7E), but a T-or Y-shape in its absence (Fig. 7C). In contrast, we did not observe a triangular shape for either McM5⌬T⌬3 or M5-1047 even in the presence of GTD under the same conditions (data not shown). These results, in conjunction with the results that the ATPase activi- The Relative Geometric Position of the Tail and the Head Is Critical for the Regulation of Myosin Va-Because the above results suggest that the triangular form represents the inhibitory state, we hypothesized that the length of the neck domain and the first long coiled-coil, as well as the proper orientation of the tail and the head are critical for inhibition by the tail binding to the head. To evalu-ate this hypothesis, we produced a number of constructs with various deletions of the neck region or coil-1 region (Fig. 1), and investigated their regulation by GTD. As shown in Fig.  8, McM5⌬T⌬IQ34, a myosin Va mutant with its neck shortened by deleting the 3rd and 4th IQ motifs was not effectively inhibited by GTD. This result could mean that the 3rd and 4th IQ motifs, specifically, are essential for regulation, or that the intact neck is important for regulation. To evaluate these possibilities, we produced McM5⌬T⌬IQ12, and McM5⌬T⌬IQ56, by deleting IQ1 and IQ2 or IQ5 and IQ6 of myosin Va, respectively (Fig. 1). Similar to McM-5⌬T⌬IQ34, the actin-activated AT-Pase activity of neither McM5⌬T⌬-IQ12 nor McM5⌬T⌬IQ56 was effectively inhibited by GTD (data not shown), indicating that the deletion of any two IQ motifs of neck disrupts the interaction between the motor domain and GTD.
The deletion of two IQ motifs shortens the neck by 7.2 nm (48 amino acid residues ϫ 0.15 nm for each residue in the ␣-helix). One possible scenario is that the shortening of the neck makes it difficult for GTD, which is located 30-nm away from the joint of the two heads, to interact with the motor domain. If so, deletion of an appropriate length of the proximal region of coil-1 may restore the interaction between the motor domain and GTD.
As shown in Fig. 7, full-length myosin Va forms a triangular shape with an angle of about 40 degrees between the two heads. Therefore, shortening of the proximal coiledcoil region by about 6.77 nm (7.2 nm ⅐ cos (40°/2)) would restore the triangular shape thus restoring the interaction between motor domain and GTD. The length of 6.77 nm is equivalent to 45 amino acid residues in an ␣-helix coiled-coil. Because the helical period of the coiled-coil is 7 amino acid residues, deletion of 49 or 42 residues would produce a continuous, in-register coiled-coil. Therefore, we deleted 950 -998 (49 residues) and 950 -991 (42 residues) of McM5⌬T⌬IQ34 at the coil-1 region, producing McM5⌬T⌬IQ34⌬49aa and McM5⌬T⌬IQ34⌬42aa, respectively (Fig. 1). We found that deletion of 42 aa partially restored the regulation of McM5⌬T⌬IQ34 by GTD (Fig. 8), whereas deletion of 49 aa did not (data not shown). In contrast,  (Fig. 8). These results support the idea that the appropriate length of neck and coil-1 of myosin Va are critical for the interaction between motor domain and GTD, and thus the regulation of myosin Va.

DISCUSSION
The Structure of the Tail Domain of Myosin Va-We examined the structure of various truncated constructs of myosin Va by electron microscopy. One of the most intriguing findings is that the tail length is much shorter than that predicted based upon a series of coiled-coil domains in the tail. The proximal portion (residues 909 -1467) of the tail of myosin Va are predicted to be five coiled-coils separated by four flexible regions by Paircoil (28). The length of coil-1 (residues of 909 -1105) is predicted to be 29.4 nm, and those of the other four coiled-coil helices (coil-2 to coil-5) are 6.5, 6.3, 6.8, and 5.4 nm, respectively. If these coiled-coil domains are linearly arranged to form a long stalk, the length between the joint of the two-heads to GTD would be ϳ55 nm, much longer than the observed 30-nm length of the stalk of myosin Va (Fig. 5 and Ref. 3). We found a small globular structure at the end of McM5⌬T (Fig. 5A), which is not found for the tail of M5-1105 or M5-1234. The length of the stalk of McM5⌬T is essentially equal to that of M5-1105 tail (30.5 nm versus 31.3 nm) (Fig. 5). These results suggest that the four short putative coil domains do not serve as a stalk but constitute the small globular domain. This view is consistent with the earlier images of chicken myosin Va visualized by quickfreeze deep-etch electron microscopy. Cheney et al. (3) observed a small globule besides the globular tail domain in electron micrographs of chick brain myosin Va, and sug-gested that this structure probably corresponds to residues of 1107-1420 of chick brain myosin Va (corresponding to 1106 -1467 of McM5) (3). It has been observed that the length of tail of myosin Va is variable. Earlier images of myosin Va observed by quick-freeze deep-etch electron microscopy showed the length from the head-rod junction to the end of GTD to vary from ϳ30 nm to a maximum of 75 nm (3). Rotary metal-shadowed images of mouse myosin Va show that the distance between the head-rod junction and GTD is shorter in low ionic strength than in high ionic strength (14). Those observations suggest that the small globular domain formed by coil-2 to coil-5 might be extended under certain conditions.
The Structure of the Inhibited State of Myosin Va-In the present study, we found that exogenous GTD inhibits the actinactivated ATPase activity of unregulated HMM-like myosin Va constructs. The results clearly indicate that GTD functions as an inhibitor domain of myosin Va and is sufficient to explain the inhibition of full-length myosin Va motor activity under physiological ionic conditions (12)(13)(14).
The GTD-deleted constructs such as McM5⌬T and M5-1105 formed a T-or Y-shaped structure in the absence of GTD, but these molecules adopt a triangular shape in the presence of GTD (Fig. 7) that is similar to the triangularshaped conformation of full-length myosin Va in the inhibited state. The GTD-induced formation of the triangular conformation was correlated with the GTD-dependent inhibition of the actin-activated ATPase activity of the truncated myosin Va. These results support the notion that the triangular conformation represents the inhibited form of myosin Va (12). This is further supported by the results showing that neither McM5⌬T⌬3 nor M5-1047 forms the triangular conformation in the presence of exogenous GTD, and this is reflected by the lack of inhibition by GTD. The results also suggest that GTD binds to the C-terminal region of coil-1. The lengths of coil-1 and the head domain of myosin Va are about 30 nm. A triangular conformation could be formed, if GTD simultaneously binds to the C terminus of coil-1 and the head domain. If so, the proper length of the neck and coil-1 would be critical for the inhibited state of myosin Va. This view is supported by the finding that myosin Va constructs with a shortened neck (McM5⌬T⌬IQ12, McM5⌬T⌬IQ34, or McM5⌬T⌬IQ56) are not inhibited by GTD. If GTD binds to the C terminus of coil-1, the head domain will not reach the GTD when the neck length is shortened. Furthermore, myosin Va with a shortened coil-1 (such as McM5⌬T⌬42aa) is not inhibited by GTD. This suggests that, because the neck length is sufficient to reach the GTD at the C-terminal end of the coil-1, the neck domain and the coil-1 are rigid and the matching of the length of the two domains is critical for the interaction between the motor domain and the GTD domain, thus inactivating the motor activity. Supporting this idea, the inhibition of McM5⌬T⌬IQ34 by GTD was partially restored by deleting 42 amino acid residues of coil-1 (McM5⌬T⌬IQ34⌬42aa in Fig. 8). The complete restoration of regulation may require appropriate orientation of the GTD and the motor domain in addition to the matched length of the coil-1 and the head domain.
Based upon the findings of the present study, we propose a model for the formation of the inhibited state of myosin Va under normal low Ca 2ϩ conditions in vivo as follows. GTD binds to the C-terminal end of coil-1. The neck-tail junction of myosin Va is flexible, and the long neck enables the head domain to reach the GTD associated at the end of coil-1. Once the heads interact with the GTD, the triangularinhibited conformation is stabilized. The binding of the head domain to GTD prevents its conformational change during ATP turnover or interaction with actin, thus inhibiting the actin-activated ATPase activity of the motor domain (Fig. 9). There are two scenarios to destabilize the inhibited conformation. Cargo binding to the tail domain of myosin Va in the inhibited state may interrupt the interaction between GTD and the head of myosin Va, thus shifting the conformational equilibrium toward the active state. This possibility is supported by the finding that the actin-activated ATPase activity of myosin Va was enhanced by its cargo-binding protein, melanophilin (17). Whereas a number of cargo molecules can bind to myosin Va, it is unlikely that these cargo molecules always disrupt the inhibited form. Ca 2ϩ -dependent activation may play a role for the transport of such cargo molecules. It has been shown that micromolar concentrations of Ca 2ϩ stimulates the actinactivated ATPase activity and induces an extended conformation of myosin Va (12)(13)(14). The activation by Ca 2ϩ is presumably through calmodulin bound to the IQ motifs. It is possible that the neck region, especially the calmodulin bound to the 1st IQ motif, is involved in the binding of GTD to the head. High Ca 2ϩ may change the conformation of calmodulin at this motif, thus disrupting its interaction with GTD. Alternatively, the triangular conformation of myosin Va may require the proper interaction between calmodulins bound to the 6th IQ motif at the joint of the each heads, which is disrupted by high Ca 2ϩ . Another possibility is that high Ca 2ϩ induces the dissociation of calmodulin from IQ motifs, thus compromising the structure of neck and disrupting the triangular conformation. Further studies are required to clarify this issue.
Currently, the crystal structures of three components of myosin Va, i.e. the motor domain (29,30), the neck (31,32), and GTD (33), have been solved at high resolution. The fitting of these crystal structures into the triangular conformation of myosin Va should provide detailed structural information on the inhibited state of myosin Va.