Functional and Structural Relationship between the Calmodulin-binding, Actin-binding, and Actomyosin-ATPase Inhibitory Domains on the C Terminus of Smooth Muscle Caldesmon*

Multiple functional domains responsible for calmodulin (CaM) binding and actin-binding/actomyosin ATPase inhibition are present in the region between residues 598–756 of the chicken gizzard smooth muscle caldesmon (CaD) molecule. To precisely localize these functional domains and to further elucidate the structural basis of these domains, we analyzed a series of purified mutants of chicken gizzard smooth muscle CaD generated by internal deletions of amino acid sequences and expression in a baculovirus expression system. Our results demonstrate that, in addition to a strong actin-binding site sequence between residues 718–723 (Wang, Z., and Chacko, S. (1996)J. Biol. Chem. 271, 25707–25714), two weak actin-binding motifs are present in the regions between residues 690–699 and 650–666. These weak actin-binding regions function independently and are associated with weak actomyosin inhibitory activity. Analysis of the CaM-binding sites A (residues 658–666) and B (residues 690–695), the major CaM-binding sites in the C-terminal region of CaD, provided direct evidence for the involvement of both CaM-binding sites in the CaM-mediated reversal of the inhibition of actomyosin ATPase activity by CaD and for the functional independence of the two CaM-binding sites. Furthermore, the sequences between residues 598–649, upstream of CaM-binding site A, and 700–717, downstream of CaM-binding site B, appear to have no effect on either actin-binding or CaM-binding. The data also suggest that both CaM-binding sites A and B structurally overlap or lie in close proximity to the adjacent weak actin-binding sites and weak actomyosin ATPase inhibitory determinants.

Multiple functional domains responsible for calmodulin (CaM) binding and actin-binding/actomyosin ATPase inhibition are present in the region between residues 598 -756 of the chicken gizzard smooth muscle caldesmon (CaD) molecule. To precisely localize these functional domains and to further elucidate the structural basis of these domains, we analyzed a series of purified mutants of chicken gizzard smooth muscle CaD generated by internal deletions of amino acid sequences and expression in a baculovirus expression system. Our results demonstrate that, in addition to a strong actinbinding site sequence between residues 718 -723 (Wang, Z., and Chacko, S. (1996) J. Biol. Chem. 271, 25707-25714), two weak actin-binding motifs are present in the regions between residues 690 -699 and 650 -666. These weak actin-binding regions function independently and are associated with weak actomyosin inhibitory activity. Analysis of the CaM-binding sites A (residues 658 -666) and B (residues 690 -695), the major CaM-binding sites in the C-terminal region of CaD, provided direct evidence for the involvement of both CaM-binding sites in the CaM-mediated reversal of the inhibition of actomyosin ATPase activity by CaD and for the functional independence of the two CaM-binding sites. Furthermore, the sequences between residues 598 -649, upstream of CaM-binding site A, and 700 -717, downstream of CaM-binding site B, appear to have no effect on either actin-binding or CaM-binding. The data also suggest that both CaM-binding sites A and B structurally overlap or lie in close proximity to the adjacent weak actinbinding sites and weak actomyosin ATPase inhibitory determinants.
Regulation of smooth muscle contraction is thought to occur primarily through myosin phosphorylation-dephosphorylation mechanisms (for review, see Refs. 1-3) complemented by a thin filament-mediated regulation that requires the actin/calmodulin-binding proteins, such as smooth muscle caldesmon (CaD) 1 (4 -9) and calponin (10,11). It has been established that CaD inhibits the actin-activated ATPase activity of phosphorylated smooth muscle myosin (for review, see Refs. 12 and 13), reduces the movement of actin filaments by phosphorylated smooth muscle myosin in in vitro motility assays (14 -16), and downregulates force generation in a chemically skinned fiber system (17). This CaD-induced inhibition of actomyosin ATPase activity is completely released by Ca 2ϩ -binding protein calmodulin (CaM) or caltropin (18) in the presence of Ca 2ϩ . Smooth muscle tropomyosin enhances the inhibition (5-7, and for review, see Refs. 12 and 13). It has also been suggested that the inhibitory function of CaD is modulated by phosphorylation of the CaD C-terminal region by mitogen-activated protein kinase or cdc2 kinase, leading to partial reversal of the CaD-induced inhibition of actomyosin ATPase activity (19,20). The exact inhibitory mechanism by which CaD inhibits actomyosin ATPase activity is unclear, but proposed models include competition of CaD with myosin for the same binding site in the actin molecule (21) and CaD-mediated effect on the catalytic step in the ATPase cycle (22,23).
Functional analysis of CaD fragments produced from limited proteolysis, chemical cleavage, and bacterial expression systems has led to the localization of various functional domains in the CaD molecule. The major myosin-binding site was mapped to the NH 2 -terminal region of CaD (21,24,25), whereas the ATPase inhibitory determinants and the crucial actin-/calmodulin-binding sites were first mapped to a 38-kDa chymotryptic fragment in the C terminus (22, 26 -28) and recently restricted to a smaller region between residues 658 -756 (29,30). More recently, our analysis of C-terminal deletion mutants of chicken gizzard CaD derived in our laboratory indicated the presence of multiple actin-binding sites and ATPase inhibitory determinants in the region between residues 658 -756 (31). Further analysis of internal deletion mutants of chicken gizzard CaD identified a strong actin-binding motif of six amino acid residues from Lys 718 to Glu 723 , which also forms the core sequence for CaD-induced inhibition of actomyosin ATPase activity (32). Our findings also suggested that the region between residues 690 and 717 is associated with the weak inhibition of actomyosin ATPase activity. This weak inhibitory determinant between residues 690 -717 and the strong inhibitory determinant between residues 718 -756 function independently (32).
It has been assumed that Ca 2ϩ -CaM is involved in the regulation of CaD function because Ca 2ϩ -CaM completely reverses the inhibition of actomyosin ATPase activity by CaD and, to some extent, displaces CaD bound to actin (4 -9, and for review, see Refs. 12 and 13). Comparison of data from biochemical studies using synthetic and chymotryptic peptides of CaD has defined two major CaM-binding sites in the C terminus of CaD, which have been precisely localized to a sequence between residues 658 -666 (CaM-binding site A) (33) and a sequence between residues 687-695 (CaM-binding site B) (34). Analysis of C-terminal truncated proteins of CaD further localized CaM-binding site B to a six-residue stretch from Asn 690 to Lys 695 (31). However, the functional role of the two calmodulin-binding sites in the interaction with Ca 2ϩ -CaM remains controversial.
To better understand the weak actin-binding and actomyosin ATPase inhibitory domains and the relationship between the CaM-binding sites and actin-binding/actomyosin ATPase inhibitory domains in the CaD molecule, we generated a series of internal deletion in the regions that encompass the weak actinbinding sites, and analyzed these mutants for their actin-binding, inhibition of actomyosin ATPase, and CaM binding. Our data clearly demonstrate that 1) two weak actin-binding/actomyosin ATPase inhibitory motifs are located in the regions between residues 650 -666 and 690 -699, respectively; that 2) both CaM-binding sites A and B are functionally involved in CaM-induced reversal of CaD-mediated inhibition of actomyosin ATPase activity; and that 3) both CaM-binding sites A and B overlap or are in close proximity with the adjacent weak actin-binding/ATPase inhibitory determinants.
Transfection and Isolation of Recombinant Baculovirus-Production of baculovirus recombinants by co-transfecting Sf9 cells with a mixture of recombinant baculovirus vector and wild-type Autographa californica nuclear polyhedrosis virus (AcNPV) DNA and isolation of the recombinants were carried out as described (31,32,36,37).

Binding Assays for Calmodulin and Smooth
Muscle Actin-CaMcoupled agarose (Sigma) was used to determine the binding of fulllength CaD or CaD mutants to CaM according to our published method (31). Nonspecific binding was estimated by incubation of agarose (without coupled calmodulin) with increasing concentrations of 14 C-labeled full-length CaD or CaD mutants, and these values were subtracted from each point on the binding curve.
Binding of 14 C-labeled CaD mutants to actin and tropomyosin-actin was determined by co-sedimentation using an Airfuge (Beckman Instruments) as described (31). The amounts of bound and unbound CaD mutants were estimated using a liquid scintillation counter (Beckman Instruments). The 14 C-labeled full-length CaD and CaD mutants were also used to determine actin binding in the presence or absence of tropomyosin, as described (31). All the binding assays were done in triplicate. The apparent dissociation constants for calmodulin, actin, and actin-tropomyosin bindings were determined by Scatchard analysis (43) and by weighted nonlinear least-squares curve fitting as described by Munson and Rodbard (44).
ATPase Assay-ATPase assays were carried out at 25°C as de-scribed (45). Specific assay conditions are described in the figure legends. ATPase assays for each experimental parameter were repeated four times using four different protein preparations. Data from a representative experiment are shown in each figure.

Localization of the Weak Actin-binding Motif and Weak
ATPase Inhibitory Determinant-Our previous studies suggested that the region between residues 690 -717 in the CaD molecule is associated with weak actin binding, weak inhibition of actomyosin ATPase activity, and strong CaM binding (31,32). However, the exact location of the functional motifs in the sequence between residues 690 -717 and its structural relationship with CaM-binding site B are unknown (36). To further define the functional motifs, we constructed several internal deletion mutants of chicken gizzard smooth muscle CaD, which progressively lacked the targeted sequences within residues 690 -717, and overexpressed them in a baculovirus system (36). Fig. 1 illustrates the schematic structures of these internal mutants. All the mutant proteins were purified to near homogeneity (Ͼ95% purity) using our published method (36).
Binding of 14 C-labeled CaD mutants to smooth muscle actin was determined by a co-sedimentation assay using full-length CaD as a control. The actin-binding of both CaD⌬700 -717 and CaD⌬710 -717, which lack 18 and 8 amino acids, respectively (see in Fig. 1), was similar to that of full-length CaD ( Fig. 2A). By contrast, binding of CaD⌬690 -717 and CaD⌬690 -699 to actin was slightly less (13% and 9%, respectively) than that of full-length CaD ( Fig. 2A). The actin-binding of both full-length CaD and CaD mutants was saturated at a CaD concentration of 5 M (molar ratio of CaD to actin, 1:4). As expected, the presence of smooth muscle tropomyosin increased the binding of both full-length CaD and CaD mutants to actin (Fig. 2B). Scatchard analysis of the binding affinity of both full-length CaD and the internal deletion mutants to actin or tropomyosinactin (43,44) showed that CaD⌬690 -717 and CaD⌬690 -699 bound to actin with lower affinity (0.65 Ϯ 0.033 ϫ 10 Ϫ6 M and 0.64 Ϯ 0.037 ϫ 10 Ϫ6 M, respectively) than did full-length CaD Effects of full-length CaD and the internal deletion mutants on the inhibition of the activation of smooth muscle myosin ATPase activity by smooth muscle actin or tropomyosin-actin were measured (in Fig. 3). CaD⌬710 -717, which lacks residues Leu 710 to Gly 717 , inhibited the actin or tropomyosin-actin activated ATPase hydrolysis of myosin as efficiently as did fulllength CaD, whereas CaD⌬700 -717, CaD⌬690 -717, and CaD⌬690 -699 were relatively less effective in this inhibition (Fig. 3A). Maximal inhibition obtained with CaD⌬690 -699, CaD⌬700 -717, and CaD⌬690 -717 was reduced by approximately 15%, 17%, and 29%, respectively, as compared with that of full-length CaD. As shown in Fig. 3B, the presence of smooth muscle tropomyosin enhanced the maximal inhibition caused by CaD 1-756 , CaD⌬690 -699, CaD⌬710 -717, CaD⌬700 -717, and CaD⌬690 -717.
Association of the Region between Residues 650 and 666 with Weak Actin-binding and Weak Inhibition of Actomyosin ATPase Activity-Previous studies using a 7.3-kDa CaD Cterminal peptide produced from limited proteolysis indicated that the region between residues 597-665 in the chicken gizzard smooth muscle CaD molecule is critical for both actin-binding and inhibition of actomyosin ATPase activity (46). However, studies using truncated CaD mutants suggested that the region between residues 658 -689 is only associated with weak actin-binding and weak inhibition of actomyosin ATPase activity and that the region between residues 598 -657 has no detectable activity for inhibition of actomyosin ATPase (31). To provide more precise analysis of the weak actin-binding and inhibitory domains, we generated various CaD internal deletion mutants (Fig. 1), focusing on the region between residues 609 -666.
Deletion of the region between residues 609 -628 did not alter the binding of CaD to actin, nor did removal of residues 629 -639 (CaD⌬629 -639) or residues 629 -649 (CaD⌬629 -649) (Fig. 4uA). However, further deletion of a segment between residues 650 -666 resulted in a 16% decrease in the binding to actin compared with the actin-binding produced by full-length CaD (Fig. 4A). Moreover, the binding of CaD⌬650 -666 to actin was similar to that of CaD⌬629 -666, and dual deletions of residues 650 -666 and 690 -699 (CaD⌬650 -666\590 -699) resulted in an additive effect on the decreased level of binding to actin (Fig. 4A). Apparently, the region between residues 650 -666 contains a functional motif responsible for weak actin-binding. The apparent dissociation constants of CaD⌬609 -628, CaD⌬629 -639, and CaD⌬629 -649 were identical to that of full-length CaD, whereas the apparent K d value for CaD⌬629 -666 was 0.68 Ϯ 0.037 ϫ 10 Ϫ6 M, indicating a 36% increase over that of full-length CaD. The apparent K d value of the actin-binding of CaD⌬650 -666\690 -699 was 1.07 Ϯ 0.07 ϫ 10 Ϫ6 M. The presence of tropomyosin increased the binding of full-length CaD and the CaD mutants to actin (Fig. 4B). The apparent K d values of CaD⌬629 -666 and CaD⌬650 -666\690 -699 for tropomyosin-actin binding were 0.4 Ϯ 0.027 and 0.61 Ϯ 0.053 ϫ 10 Ϫ6 M, respectively. The apparent K d values of CaD⌬650 -666 for both actin binding and tropomyosin-actin binding were indistinguishable from those of CaD⌬629 -666.
Analysis of the ability of the internal deletion mutants to inhibit actomyosin ATPase activity revealed a similar degree of inhibition for both actin and tropomyosin-actin activated myosin ATPase activities with CaD 1-756 , CaD⌬609 -628, CaD⌬629 -639, and CaD⌬629 -649, whereas deletion of residues 629 -666 caused a 37 and 19% decrease in the inhibition of actin-activated and tropomyosin-actin-activated myosin ATPase activity, respectively, as compared with that of fulllength CaD (Fig. 5, A and B). In the presence of CaD⌬650 -666, the inhibition of actin-activated and tropomyosin-actin-acti- vated myosin ATPase activities was similar to that of CaD⌬629 -666 (Fig. 5, A and B). The CaD dual internal deletion mutant, CaD⌬650 -666\690 -699, caused a less inhibitory effect on both actin-activated and tropomyosin-actin-activated myosin ATPase activities compared with those generated by either CaD⌬629 -666 or CaD⌬690 -717. A CaD internal deletion mutant lacking residues 597-608 was also generated using the baculovirus expression system; this mutant showed the same actin binding and inhibitory function as full-length CaD (data not shown). Thus, the weak inhibitory determinant is present in the amino acid stretch from Glu 650 to Ser 666 . Fig. 6, A and B, deletion of the regions between residues 690 -699 and between residues 650 -666 led to a 40 and 24% decrease, respectively, in binding to CaM as compared with that of full-length CaD. CaD⌬650 -666\690 -699 lacking residues 650 -666 and 690 -699 was unable to bind to CaM (Fig. 6B). The CaM-binding capacity of CaD⌬650 -666 was similar to that of CaD⌬629 -666 (data not shown). The sequences between residues 700 -717 and residues 609 -649 were not responsible for the binding of CaD to CaM because CaD⌬609 -628, CaD⌬629 -639, CaD⌬629 -649, CaD⌬710 -717, and CaD⌬700 -717 bound to CaM as tightly as full-length CaD. Moreover, deletion of the sequence between residues 680 -689 did not affect the binding of CaD to CaM (data not shown). Scatchard analysis showed that the CaD⌬710 -717, CaD⌬700 -717, CaD⌬609 -628, CaD⌬629 -639, and 629 -649 bound to CaM with the same affinity as full-length CaD (0.98 Ϯ 0.061 ϫ 10 Ϫ6 M). In contrast, the affinity of both CaD⌬690 -699 and CaD⌬690 -717 to CaM (2.01 Ϯ 0.063 and 1.96 Ϯ 0.058 ϫ 10 Ϫ6 M, respectively) was reduced ϳ2-fold compared with that generated by full-length CaD. As expected, CaD⌬629 -666 displayed a markedly reduced affinity for CaM (1.45 Ϯ 0.048 ϫ 10 Ϫ6 M). The apparent K d value of the CaM-binding of CaD⌬650 -666 was indistinguishable from that of CaD⌬629 -666.

Functional Involvement of Both CaM-binding Sites A and B in CaM-induced Reversal of Inhibition of CaD-induced Actomyosin ATPase Activity-As shown in
Analysis to determine whether both CaM-binding sites A and B were functionally involved in CaM-induced reversal of CaD-mediated inhibition of actomyosin ATPase activity showed that inhibition by both CaD⌬629 -666 and CaD⌬690 -717 was reduced with increasing concentrations of Ca 2ϩ -CaM (Fig. 7). At 1:1 molar ratios of CaM to actin or tropomyosinactin, the percentage of reversal by CaM for the inhibition caused by CaD⌬629 -666 was slightly lower than that of both full-length CaD and CaD⌬690 -717. At a stoichiometry of 0.075 mol CaD/mol CaM, the inhibition induced by both CaD⌬629 -666 and CaD⌬690 -717 was completely reversed, similar to that for full-length CaD (Fig. 7). Ca 2ϩ -CaM also reversed the inhibition by CaD⌬629 -666 in a pattern similar to that of CaD⌬690 -717. Additionally, tropomyosin did not affect the CaM-induced reversal of the inhibition by either full-length CaD or the internal deletion mutants consistent with our recent findings (31,32). Together, the results indicate that both CaM-binding sites A and B are functionally involved in the CaM-reversal for inhibition of actomyosin ATPase activity by CaD and both function independently.

DISCUSSION
Based on our previous studies (31,32), we postulated the presence of a weak actin-binding motif in the region between residues 690 -717. The present study supports this suggestion and further defines the weak actin-binding motif to a 10-residue sequence from Asn 690 to Gly 699 . The sequence between residues 700 -717 does not contribute to actin binding since removal of this region had very little effect on the binding of CaD to actin (Fig. 2). We therefore conclude that the 67-amino acid region between residues 690 -756 in the C terminus of CaD contains two discontinuous actin-binding motifs, i.e. the strong actin-binding motif consisting of six residues from Lys 718 to Glu 723 (32) and the weak actin-binding motif located between residues 690 -699. The two functional motifs are separated by an 18-residue stretch from Asn 700 to Gly 717 (Fig. 2). The presence of tropomyosin did not modify these binding sites except for the increased amounts of CaD bound per actin (Fig.  2B). The 10-residue sequence from Asn 690 to Gly 699 was also associated with very weak inhibition of actomyosin ATPase activity (Fig. 3). Interestingly, the region between residues 700 -709, which is not required for actin-binding, appears to be necessary for the weak inhibition of actomyosin ATPase activity because deletion of residues 710 -717 (CaD⌬710 -717) had no effect on the inhibition of actomyosin ATPase activity by CaD, and further deletion of the stretch from Asn 700 to Asp 709 (CaD⌬700 -717) slightly lowered the inhibition of actomyosin ATPase activity by CaD (Fig. 3). Again, tropomyosin enhanced this effect (Fig. 3B). These results strongly suggest that the weak inhibitory determinant is located in the 20-residue stretch from Asn 690 to Asp 709 .
Our previous analysis of CaD C-terminal truncated mutants (31) showed that the region between residues 598 -657, responsible for weak actin-binding, has no detectable activity on inhibition of actomyosin ATPase, whereas residues 658 -689 are needed for both weak actin-binding and weak inhibition of actomyosin ATPase activity (31). Experiments using CaD Cterminal internal deletion mutants in the region between residues 629 -666 (Fig. 1) showed that deletion of the region between residues 629 -649 has no effect on the actin/tropomyosin-actin binding and CaD-induced inhibition of actin-activation of myosin ATPase. However, further deletion of the sequence between residues 650 -666 affected both the actin/ tropomyosin-actin binding and CaD-induced inhibition of actin-activation of myosin ATPase (Figs. 4 and 5). It may be argued that the effect of deletion of residues 629 -666 on actinbinding and on the CaD-induced inhibition of actomyosin ATPase activity is due to an abnormal conformational change rather than deletion of a functional motif present in the region (residues 629 -666). Deletion of residues 609 -628 and 629 -649 had no effect on both actin binding and the CaD-induced inhibition, whereas deletion of a 17-amino acid stretch (residues 650 -666) resulted in a decrease in both actin-binding and inhibition of actomyosin ATPase activity similar to those of CaD⌬629 -666. These data strongly suggest the presence of a functional motif in residues 650 -666 in the CaD molecule. Together with our previous results (31), the present data further define the functional domain to the segment from Met 658 to Ser 666 . Moreover, the region between residues 629 -649 (Figs. 4 and 5), upstream of the segment from Glu 650 to Ser 666 , is not involved in either actin-binding or inhibition of actomyosin ATPase activity since the actin-binding and inhibitory activities of CaD were unchanged with deletion up to 21 residues from Val 629 to Ala 649 .
The data from the present study also suggested that the actin-binding motifs present in residues 650 -666 and 690 -699 function independently since deletion of either residues 650 -666 or 690 -699 reduced the binding of CaD to actin. Furthermore, the CaD dual internal deletion mutant, CaD⌬650 -666\690 -699, had a lower binding capacity to actin when compared with those generated by either CaD⌬650 -666 or CaD⌬690 -699 (Fig. 4).
While the major CaM-binding sites in CaD molecule are well established, there is controversy about the functional significance of CaM-binding sites A and B in CaM-induced reversal for inhibition of actomyosin ATPase activity by CaD (34,47). The data from Marston et al. (34) favor CaM-binding site B, but not CaM-binding site A, as the function-associated CaM-binding site because a synthetic peptide corresponding to the sequence Ser 657 -Gly 670 did not release the CaM-induced reversal of the inhibition of actomyosin ATPase activity by CaD. By contrast, Zhuang et al. (47), using a similar research approach, concluded that CaM-binding site A, but not CaM-binding site B, is involved in the CaM-CaD interaction for reversal of the CaD-induced inhibition. One reason for this discrepancy may be the lack of amino acids adjacent to CaM binding sites A or B needed for the protein conformation required for the proteinprotein interaction in these studies. To clarify this issue, we analyzed the CaD internal deletion mutants in which CaMbinding sites A and B were nest-deleted, respectively, for their inhibition of actomyosin ATPase activity in the presence of Ca 2ϩ -CaM. Our data clearly show that both CaM-binding sites A and B are involved in the CaM-induced reversal of the inhibition of actomyosin ATPase activity by CaD (Fig. 7) and behave as independent functional domains despite the fact that the affinity of CaM-binding site B to CaM is approximately 1.3-fold higher than that of CaM-binding site A (Fig. 6). However, when inhibition was induced by the deletion mutant lacking CaM-binding site A (the region between residues 650 -666), the release of the inhibition by CaM was lower than the CaM-induced release of the inhibition caused by the deletion mutants lacking the CaM-binding site B (the region between residues 690 -699), indicating that CaM-binding site A is functionally more active than CaM-binding site B.
The discrepancies between our observations and those of others most likely rests in the use of synthetic CaD peptides, including VG29C (47) and M73 (34) that lack the necessary conformation for functional and structural interaction with CaM. Our present data also indicate that CaM-binding sites A and B each structurally overlap or lie in close proximity to the adjacent weak actin-binding motifs and weak inhibitory domains. The structural features of both CaM-binding sites A and B support the assumption that CaM competes with actin for binding to the same common contact sites in the CaD molecule. Note that the NMR measurements of bacterially generated human CaD fragments by Marston et al. (34) suggested the presence of a CaM-binding motif in a region corresponding to the sequence between residues 717-725 in the chicken gizzard smooth muscle CaD molecule. However, our previous results show that the deletion of the sequence between residues 718 -756 has no effect on the binding of CaD to CaM (31). Moreover, a CaD internal deletion mutant lacking the region between residues 717-725 bound to CaM with the same affinity as did full-length CaD (data not shown). These results directly rule out the possibility that a CaM-binding site is present in the region between residues 717-725. Thus, it is unlikely that CaM-induced reversal of inhibition of actomyosin ATPase activity by CaD is caused in part through interaction of residues 717-725 in the CaD molecule with CaM.
The presence of three functional domains for actin-binding/ actomyosin ATPase inhibition in the regions between residues 718 -731, 690 -699, and 650 -666 (Fig. 8) is consistent with the multiple-site model for actin-binding/actomyosin ATPase inhibitory domains proposed by us and others (28,31,32). Both the regions between residues 650 -666 and 690 -699 consist of 70% uncharged and 30% charged amino acid residues, suggesting that, along with charge-charge interaction and hydrogen bonding, hydrophobic interaction may be involved in the interaction of the CaD and actin that is essential for inhibition of actomyosin ATPase activity by CaD.
In summary, the last 99 amino acid residues at the C terminus of CaD contain three functionally independent motifs responsible for actin-binding/actomyosin ATPase inhibition. The three functional domains are structurally discontinuous and are located in the regions between residues 718 -731, 690 -699, and 650 -666, in which the regions between residues 690 -699 and 650 -666 each also contain a strong CaM-binding site. This precise localization should help to better understand the regu- FIG. 8. Localization of functional domains at the C terminus of chicken gizzard smooth muscle CaD. The region between residues 508 -565 is homologous to the tropomyosin-binding site on the troponin T molecule and is considered to be a tropomyosin-binding site on the CaD molecule (28,35). High-affinity CaM-binding site A was first demonstrated by Wang and co-workers (33), and high-affinity CaM-binding site B was derived from our present and previous data (31) and from the data of Mezgueldi et al. (48). The low affinity CaM-binding site, the tropomyosinbinding site at the C-terminal end of CaD, and the actin-binding/inhibitory site 3 are suggested by our previous data (31). Actin-binding/inhibitory sites 1 and 2 are demonstrated in the present study. latory mechanisms involved in the interaction of CaD with actin and the regulation of actin-myosin interaction and of actomyosin ATPase.