The Calmodulin-binding Domain from a Plant Kinesin Functions as a Modular Domain in Conferring Ca2+-Calmodulin Regulation to Animal Plus- and Minus-end Kinesins*

Plant kinesin-like calmodulin-binding protein (KCBP) is a novel member of the kinesin superfamily that interacts with calmodulin (CaM) via its CaM-binding domain (CBD). Activated CaM (Ca2+-CaM) has been shown to inhibit KCBP interaction with microtubules (MTs) thereby abolishing its motor- and MT-dependent ATPase activities. To test whether the fusion of CBD to non-CaM-binding kinesins confers Ca2+-CaM regulation, we fused the CBD of KCBP to the N or C terminus of a minus-end (non-claret disjunction) or C terminus of a plus-end (Drosophila kinesin) motor. Purified chimeric kinesins bound CaM in a Ca2+-dependent manner whereas non-claret disjunction, Drosophila kinesin, and KCBP that lack a CBD did not. As in the case of KCBP with CBD, the interaction of chimeric motors with MTs, as well as their MT-stimulated ATPase activity, was inhibited by Ca2+-CaM. The presence of a spacer between the motor and CBD did not alter Ca2+-CaM regulation. However, KCBP interaction with MTs and its MT-stimulated ATPase activity were not inhibited when the motor domain and CBD were added separately, suggesting that Ca2+-CaM regulation of CaM-binding motors occurs only when the CBD is attached to the motor domain. These results show that the fusion of the CBD to animal motors confers Ca2+-CaM regulation and suggest that the CBD functions as a modular domain in disrupting motor-MT interaction. Our data also support the hypothesis that CaM-binding kinesins may have evolved by addition of a CBD to a kinesin motor domain.

ATP, a neck (determines motor directionality on MTs), a coiledcoil region (aids in dimerization), and a variable tail domain (involved in regulation of motor activity and/or in cargo carrying) (3,4). The motor domain can be at the N or C terminus or in the middle. The N-terminal motors move toward the fast growing plus-end of MTs whereas the C-terminal motors translocate to the slow growing minus-end of MTs (3,5,6).
Recent genome sequence projects, coupled with cell and molecular biological studies in animals and plants, revealed that although both plants and animals contain some common motors plants are unique in containing a large number of kinesins (61 in Arabidopsis) including several plant-specific ones (7). Studies with a few plant kinesins indicate that they are regulated by novel mechanisms and perform plant-specific functions (2,8,9). One of the plant kinesins, kinesin-like calmodulin-binding protein (KCBP), a minus-end motor, was isolated from a number of dicot and monocot plants as a calmodulin (CaM)-interacting protein (10 -13). The CaM-binding domain (CBD) in KCBP is mapped to the C terminus of the motor domain (10). Although a CaM-binding kinesin has not been identified in the completely sequenced genomes of Saccharomyces cerevisiae, Schizosaccharomyces pombe, Caenorhabditis elegans, Drosophila melanogaster, and Homo sapiens (7,14), a kinesin (kinesin-C) that binds to CaM has been isolated from sea urchin (15).
KCBP and kinesin-C are similar in having a conserved neck, motor domain at the C terminus, and a CBD through which these kinesins interact with CaM in a Ca 2ϩ -dependent manner. However, the KCBP and kinesin-C differ structurally in the following: (i) the length of coiled-coil region and (ii) the tail domain organization; KCBP contains a MyTH4 and a talin-like domain present in tail regions of some animal myosins (IV, X, and VIIa) whereas kinesin-C lacks these domains (7,16). Functionally KCBP is involved in plant-specific processes such as trichome morphogenesis and is associated with the preprophase band and phragmoplast in dividing cells, which are absent in animal cells (17)(18)(19)(20). Using KCBP-specific antibodies in microinjection studies, it has been shown that KCBP controls several aspects of cell division (21). Although the function of kinesin-C is not known, it is implicated in Ca 2ϩ -dependent events during embryonic development by maintaining MT cross-bridging in mitotic divisions (15).
The mechanisms that regulate the activity of kinesins, with a few exceptions, are poorly understood (22). Chimeric motor proteins have been used to study directionality determinants, motor velocity, and functions of various domains of kinesin family members from a number of species (5,(23)(24)(25)(26)(27)(28). Regions that control motor directionality have been determined by swapping neck regions between plus-and minus-end motors (23,25,26). These chimeric motors showed reversal of their normal motor polarity i.e. NCD and kinesin move plus-and minus-end direction, respectively. Further, mutation analyses of neck regions of chimeric motors revealed the critical residues necessary for directionality (5,23,29). However, using the chimeric constructs it has been demonstrated that the myosin motor core, but not neck, determines the directionality in myosins (30). A series of chimeric myosins have also been utilized to characterize function and regulation of members of the myosin family (31)(32)(33). For example, using a chimeric myosin composed of skeletal MHC motor domain and smooth muscle MHC long ␣-helix S1 and S2 regions, it has been shown that the actin-stimulated ATPase activity of chimeric MHC, but not the native skeletal MHC, is regulated by Ca 2ϩ -CaM-dependent myosin-light chain kinase phosphorylation (31).
In the case of KCBP, activated CaM (Ca 2ϩ -CaM) inhibits the motor activity and its interaction with MTs (34 -36). Although kinesin-C contains a CBD, the significance of the domain in regulating motor activity is not known (15). Tubulins and the MT-interacting regions in kinesins are highly conserved in eukaryotes (2,6). To test whether the CBD of KCBP functions as a regulatory module in non-CaM-binding kinesins, we fused the CBD of KCBP to the N or C terminus of a minus-end (NCD) and to the C terminus of a plus-end (Drosophila kinesin, DK) motor with different length spacers between the motor and CBD. We have analyzed the chimeras for their MT binding and MT-stimulated ATPase activities in the presence of Ca 2ϩ and/or bovine CaM or Arabidopsis CaM isoforms. We show here that the MT binding and MT-stimulated ATPase activities of both N-and C-terminal chimeric motors are inhibited by either bovine CaM or Arabidopsis CaM isoforms in a Ca 2ϩ -dependent manner, suggesting that the CBD of plant KCBP acts as a regulatory module in animal chimeric motors. A spacer between the motor and CBD does not alter Ca 2ϩ -CaM regulation of chimeric motors. Furthermore, the CBD confers Ca 2ϩ -CaM regulation in cis but not in trans.
Plasmid Construction-The cDNA region encoding the CBD of Arabidopsis KCBP (aa 1211-1261) was cloned into EcoRI-HindIII digested pET28b (10). The NCD coding sequence (aa 209 -700) in pET3b MC1 was PCR-amplified using the 5Ј T7 forward primer and 3Ј NCD-specific reverse primer (5Ј-GCAGAATTCTTATCGAAATTGCCGCTG-3Ј) with an EcoRI site. The amplified product was digested with NdeI-EcoRI and ligated into NdeI-EcoRI-digested pET28b containing the CBD (aa 1211-1261) of KCBP to generate NCDϩCBD. The DK coding sequence (aa 1-375) in pET21d DK was PCR-amplified using the 5Ј T7 forward primer and 3Ј DK-specific reverse primer (5Ј-GCAGAATTCCGCG-CAAGCTCGATCTCC-3Ј) with an EcoRI site. The PCR product was digested with NcoI-EcoRI and ligated into NcoI-EcoRI-digested pET28b containing the CBD of KCBP to generate DKϩCBD. The NCD coding sequence (aa 209 -700) without the CBD was constructed by deleting the CBD from NCDϩCBD with EcoRI-XhoI. The DK (aa 1-375) coding sequence without the CBD was generated by digesting the DKϩCBD with NcoI-EcoRI, and the resultant fragment was ligated into NcoI-EcoRI-digested pET32b. Preparation of 1.5C, 1.0C, and 0.4C KCBP constructs was described previously (10,35). The calculated molecular mass of the NCDϩCBD, DKϩCBD, and KCBP with CBD, NCD, and DK, and KCBP without CBD are 64, 48, 69, 59, 61, and 45 kDa, respectively. The DK construct with a spacer and CBD (DKϩSϩCBD) was prepared by inserting the PstI-EcoRI-digested PCR-amplified DK sequence coding for aa 294 to 475 (forward primer, 5Ј-GATTC-CAAGCTAACGCGCAT-3Ј and reverse primer, 5Ј-ATGAGTTCCTCCT-GCTCCA-3Ј with EcoRI site) into the PstI-EcoRI-digested DKϩCBD construct. This DKϩSϩCBD construct has a 149-amino acid spacer comprising of neck and coiled-coil regions between the motor domain and CBD. The calculated molecular mass of the DKϩSϩCBD is 60 kDa. Another construct with the CBD at the N terminus of NCD (CBDϩSϩNCD) was prepared by inserting the NcoI-digested PCRamplified CBD sequence from KCBP 0.4C construct (T7 forward primer and KCBP-specific reverse primer, 5Ј-ACTATCTGCCTCATCTTT-TCGTGT-3Ј with an NcoI site: coding for amino acids 1211 to 1261 of KCBP) in the NcoI-digested NCD construct in pET28b. The calculated molecular mass of the CBDϩSϩNCD protein is 68 kDa. The NCDϩCBD, DKϩCBD, DKϩSϩCBD, and CBDϩSϩNCD fusions were verified by sequencing.
Expression, Purification, and Detection of Kinesins-The expression of the fusion proteins in E. coli BL21 (DE3) was induced by adding isopropyl-1-thio-␤-D-galactopyranoside to a final concentration of 0.5 mM. After incubation for 5 h at 30°C, the cells were pelleted and washed with 50 mM Tris-HCl, pH 8.0. The proteins were extracted by lysing the cells in 50 mM Tris-HCl, pH 8.0, 0.5 mM Mg-ATP, 0.5 mM ␤-mercaptoethanol, 0.5 mM dithiothreitol, 100 mM lysozyme, and Complete protease inhibitor on ice for 30 min, sonicating three times (10-s each), and centrifuging for 30 min at 4°C. The motor proteins containing the CBD were purified using a CaM-Sepharose affinity column. His-Bind affinity column was used to purify the motor proteins without the CBD. The supernatant fraction containing the motor proteins with a CBD was loaded onto a pre-equilibrated (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 0.5 mM Mg-ATP, and 1 mM CaCl 2 ) CaM-Sepharose column, washed with the same buffer, and eluted with 50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 0.5 mM Mg-ATP, and 1 mM EGTA. The supernatant fraction containing the motor proteins without a CBD was loaded onto a pre-equilibrated His-Bind affinity column with TN buffer (20 mM Tris-HCl, pH 7.9, 0.5 M NaCl, and 0.5 mM Mg-ATP) plus 5 mM imidazole, washed extensively with TN buffer plus 20 mM imidazole, and eluted with TN buffer containing 100 mM imidazole. All purified proteins were dialyzed against 50 mM Tris-HCl, pH 7.5, 150 mM NaCl, and 0.5 mM Mg-ATP at 4°C. Total soluble and purified proteins were separated on SDS-containing gels, stained with Coomassie Brilliant Blue or blotted and detected with horseradish peroxidase (HRP)-conjugated CaM, T7tag monoclonal antibody, or S-protein conjugated with alkaline phosphatase (37). The preparation and purification of HRP-CaM and Arabidopsis CaM isoforms was carried out as described (37). The concentration of motor and CaM proteins was calculated by the Bradford method and verified on SDS gels (38).
Cosedimentation Assay-Polymerization of ␣and ␤-tubulins (10 mg/ ml) was performed in a buffer (80 mM Pipes, pH 6.8, 1 mM MgCl 2 , 1 mM EGTA, 1 mM GTP, 1 M taxol, and 10% glycerol) at 35°C for 30 min (35). The polymerized MTs were stabilized by adding taxol to a final concentration of 10 M. After incubation for 3 h at 35°C, the polymerized MTs were centrifuged at 33,000 rpm at 35°C, and the MT pellet was dissolved in 10ϫ cosedimentation assay buffer (1ϫ ϭ 20 mM Pipes, pH 6.9, 1 mM MgCl 2 , 1 mM dithiothreitol, 150 mM NaCl, 20 M taxol). The MT-motor binding assay was performed in a 100-l reaction containing 2.5 M motor protein or KCBP 0.4C (encoding CBD) protein with or without taxol-stabilized 10 M MTs (35). Where appropriate, 100 M CaCl 2 , 15 M CaM (bovine or Arabidopsis CaM isoforms), 5 mM AMP-PNP, or 5 mM Mg-ATP were also added. After incubation at 22°C for 20 min, the tubes were centrifuged in a TLA 100.3 rotor (Beckman Instruments Inc.) for 20 min at 35°C at 100,000 ϫ g. The supernatant and pellet were separated, mixed with SDS loading buffer, and analyzed by SDS-PAGE. The tubulin subunits and motor proteins were visualized by staining the gels with Coomassie Brilliant Blue. Because KCBP 1.0C and DKϩCBD motor proteins co-migrate with tubulin subunits, duplicate gels were run, and one gel was stained, and the second gel was blotted and probed with T7-tag monoclonal antibody or HRP-CaM, respectively, as described above. The cosedimentation assay for each of the motor proteins was repeated at least two times.
ATPase Assay-The ATPase activity of the motors was performed in a 100-l reaction buffer (15 mM imidazole, pH 7.0, 2 mM MgCl 2 , and 1 mM dithiothreitol) containing 60 nM motor protein with or without taxol-stabilized MTs (2 M) and 3 mM Mg-ATP (39). CaCl 2 , AtCaM2, and EGTA to a final concentration of 100 M, 1 M, and 2 mM, respectively, were also added in appropriate reactions. Following incubation at 30°C for 20 min, 800 l of freshly prepared malachite green reagent and 100 l of 34% citric acid were added to the reactions, vortexed, and incubated further at room temperature for 10 min (39). The released P i was estimated colorimetrically using 1 OD 660 equals 9.45 nmol of P i . The ATPase activity of the motor proteins was expressed in mol of P i released per mg of motor protein per min. All reactions were performed in triplicate. Each experiment was repeated at least three independent times. Average values obtained from triplicates were analyzed for standard deviation using Microsoft Excel.

Expression and Purification of KCBP, NCD, and DK Motor
Proteins with or without the CBD-To determine whether the CBD of a plant motor functions as a modular domain, we fused the CBD of a plant KCBP to the C terminus of two Drosophila kinesins, NCD and DK, a C-terminal motor, and an N-terminal motor, respectively (Fig. 1A). The bacterial expressed motor proteins with a CBD (KCBP 1.5C, NCDϩCBD, and DKϩCBD) were purified using a CaM-Sepharose column, and motor proteins without a CBD (KCBP 1.0C, NCD, and DK) but with a His tag were purified using a His-Bind column. The crude bacterial-induced total proteins and purified motor proteins are shown in Fig. 1B. All the purified motor proteins yielded a single band (Fig. 1B, stained gel, lane P). Three other identical gels were blotted and probed with HRP-CaM, T7-tag antibody, or Sprotein (Fig. 1B). As expected NCDϩCBD, DKϩCBD, and KCBP 1.5C interacted with CaM whereas NCD, DK, and KCBP 1.0C that lack a CBD did not interact with CaM in blot overlay assay (Fig. 1B, HRP-CaM). Further, chimeric motors did not bind CaM in the presence of EGTA, a Ca 2ϩ chelator (data not shown). These results suggest that the chimeric motors NCDϩCBD and DKϩCBD bound CaM in a Ca 2ϩ -dependent manner, and the plant kinesin CBD confers CaM binding ability to animal kinesins. NCDϩCBD, NCD, and KCBP 1.0C that contain T7-tag were detected by T7-tag antibody (Fig. 1B, T7tag.Ab) whereas KCBP 1.5C and DK with S-tag were detected by S-protein (Fig. 1B, S-protein).
Arabidopsis CaM Isoforms Inhibited Binding of KCBP 1.5C and Chimeric NCDϩCBD and DKϩCBD Motors with MTs-KCBP from plants (10) and kinesin-C from sea urchin (15) are the only known CaM-binding kinesins. Previously we have shown that the activity of KCBP is modulated by bovine CaM (34 -36). Here, we tested whether the fusion of plant kinesin CBD to phylogenetically divergent animal plus-and minus-end motors confers Ca 2ϩ -CaM regulation. The NCDϩCBD has a 36-amino acid spacer whereas in the DKϩCBD the CBD is separated from the motor by the neck and a short coiled-coil region of about 49 amino acids. Initially, we tested the binding of all three motor proteins (KCBP 1.5C, NCDϩCBD, and DKϩCBD) to MTs using a cosedimentation assay in the presence of ATP, AMP-PNP (a nonhydrolyzable ATP analog), Ca 2ϩ , CaM, or Ca 2ϩ -CaM. As shown in Fig. 2, the KCBP 1.5C motor protein cosedimented with MTs in the pellet fraction in the presence of AMP-PNP and Ca 2ϩ or CaM. In the presence of Ca 2ϩ -CaM (either bovine or Arabidopsis CaM) and AMP-PNP, the motor protein did not cosediment with the MTs and remained in the soluble fraction suggesting that motor protein interaction with MTs is inhibited by Ca 2ϩ -CaM (Fig. 2, K 1.5C). Because the KCBP 1.0C motor co-migrates with the tubulin subunits, a duplicate gel was blotted and probed with T7-tag antibody (Fig. 2, T7- Using the cosedimentation assay we analyzed the interaction of chimeric NCDϩCBD and native NCD motor with MTs. As shown in Fig. 3, although NCDϩCBD or NCD in the absence of MTs remained in the soluble fraction even after high speed ultracentrifugation (Fig. 3, No Mt), when combined with MTs they cosedimented with MTs in the presence of AMP-PNP and Ca 2ϩ or CaM (Fig. 3). However, when NCDϩCBD was incubated with MTs in the presence of AMP-PNP and both Ca 2ϩ and CaM or in the presence of ATP alone, the motor protein did not interact and cosediment with the MTs (Fig. 3, compare Mt/Ca, Mt/CaM, and Mt/Ca CaM). In contrast, NCD without CBD (Fig. 3, NCD) interacted and cosedimented with the MTs despite the presence of Ca 2ϩ -CaM. These results show clearly that NCDϩCBD and NCD are regulated in the same manner as KCBP 1.5C and KCBP 1.0C by Ca 2ϩ -CaM, respectively.
We then tested the effect of Ca 2ϩ -CaM on the chimeric DKϩCBD in which the CBD is fused to a plus-end motor and native DK motor domain in cosedimentation assays. Because the DKϩCBD co-migrates with MTs, a duplicated gel was blotted and probed with HRP-CaM to detect the DKϩCBD motor protein (Fig. 4, HRP-CaM). Like KCBP 1.5C and NCDϩCBD, the DKϩCBD is also regulated by Ca 2ϩ -CaM (Fig.  4). In the presence of both Ca 2ϩ -CaM, but not in the presence of Ca 2ϩ or CaM alone, DKϩCBD chimeric motor did not bind MTs, and the motor protein remained in the soluble fraction (Fig. 4, DKϩCBD; see Mt/Ca CaM). In contrast to DKϩCBD, DK bound with MTs in the presence of both Ca 2ϩ -CaM (Fig. 4,  DK). Our results with DK without the CBD in pelleting assay are in agreement with the MT-binding activity of the DK motor (aa 1-337) reported previously (40). Because we fused plant CBD to animal motors and used bovine CaM to show the negative regulation of chimeric NCDϩCBD and DKϩCBD motors with MTs, we tested whether Arabidopsis CaM2, -4, and -6 isoforms regulate the chimeric motors in a Ca 2ϩ -dependent manner. Although these AtCaM isoforms differ in a few amino acids (38) These results indicate that bovine and all Arabidopsis CaM isoforms are able to interact with CBD-containing motors and inhibit their interaction with MTs. These data suggest that activated CaM regulates the two chimeric (NCDϩCBD and DKϩCBD) motors, as well as KCBP 1.5C in a similar fashion although NCD and DK are minus-and plus-end-directed motors that perform different functions in Drosophila. DK is involved in membrane transport and is essential for neuromuscular function whereas NCD is involved in stabilizing bipolar spindle assembly during metaphase and chromosome distribution in female meiosis and early mitosis in oocytes and embryos (1,22).
Ca 2ϩ -CaM Abolished MT-stimulated ATPase Activity of KCBP 1.5C and Chimeric NCDϩCBD and DKϩCBD Kinesins-To further confirm the Ca 2ϩ -CaM negative regulation of chimeric motors, we determined MT-stimulated ATPase activity of these motors in the presence of Ca 2ϩ , CaM, or both. Because all Arabidopsis CaM isoforms showed inhibition of motor interaction with MTs, and because AtCaM2 has 2-fold higher affinity for KCBP 1.5C (38), we used AtCaM2 in ATPase assays. KCBP, NCD, and DK with or without CBD showed basal ATPase activity in the absence of MTs (Fig. 5). The ATPase activity of KCBP, NCD, and DK with and without CBD in the presence of MTs is increased by 15-21-fold (Fig. 5). The MT-stimulated ATPase activity of motors with or without the CBD was not affected in the presence of Ca 2ϩ or AtCaM2 alone (Fig. 5), suggesting that there is no influence of the fusion of the CBD to NCD and DK motors on the ATPase function of the chimeras. Interestingly, as in the MT-pelleting assay, the MT-stimulated ATPase activity of KCBP 1.5C, NCDϩCBD, and DKϩCBD was reduced by 90, 89, and 88%, respectively, in the presence of both Ca 2ϩ and AtCaM2 (Fig.  5A). However, the MT-stimulated ATPase activity of motors without the CBD was not inhibited in the presence of Ca 2ϩ and AtCaM2 (Fig. 5B). Furthermore, Ca 2ϩ -CaM-inhibited MTstimulated ATPase activity of KCBP 1.5C, NCDϩCBD, and DKϩCBD was reversed to normal levels in the presence of EGTA (Fig. 5A). These results show that the MT-stimulated ATPase activity of KCBP 1.5C and chimeric NCDϩCBD and DKϩCBD motors is inhibited by Ca 2ϩ -CaM through the CBD. The differences in MT-dependent ATPase activity of KCBP (ϳ17-fold), NCD (ϳ15-fold), and kinesin (ϳ19-fold) may be because of the presence of different lengths of coiled-coil region in the constructs and/or different tag fusions. For example, MT-stimulated ATPase activity with NCD (209 -700 aa) has been shown to be in the range of 8-to 28-fold (41,42). The MT-dependent ATPase activity of full-length (1-975 aa) and N401 (1-401 aa) of DK is also varied (43).

Insertion of a Spacer between the Motor Domain and the CBD or CBD Fusion to the N Terminus of the Motor Domain Did Not Alter Ca 2ϩ -CaM Regulation-Studies with DKϩCBD and
NCDϩCBD that have a short spacer (Ͻ50 amino acids) have indicated that CBD-mediated inhibition can be affective at a distance from the motor. To test whether a longer spacer between the motor and the CBD retains Ca 2ϩ -CaM regulation we prepared DKϩSϩCBD construct. In this construct, we inserted the coiled-coil region of DK (from aa 376 to 475) between the motor and the CBD (Fig. 6A). The bacterial-expressed protein bound CaM2 and was purified using a CaM-Sepharose affinity column in a Ca 2ϩ -dependent manner (Fig. 6B), suggesting that the increased spacer between the motor and the CBD does not affect chimeric motor interaction with CaM. To test the regulation of this motor protein by Ca 2ϩ -CaM, the purified chimeric motor (DKϩSϩCBD) was used in cosedimentation and ATPase assays. As shown in Fig. 6C, the DKϩSϩCBD protein cosedimented with the MTs in the presence of AMP-PNP and Ca 2ϩ or CaM2 alone. But in the presence of AMP-PNP and Ca 2ϩ -CaM2, it remained in the supernatant fraction (Fig. 6C, lane Mt/Ca CaM2). Also, the Mt-stimulated ATPase activity of the chimeric motor was inhibited by 97% (Fig. 6D). The inhibitory effect of Ca 2ϩ -CaM on MT-stimulated ATPase activity was abolished by the addition of EGTA. These results suggest that a spacer (149 amino acids) between the motor and the CBD does not influence Ca 2ϩ -CaM regulation of this chimeric motor.
All the constructs that we used so far have the CBD at the C terminus of the motor domain. To test whether the CBD fusion to the N terminus of the motor confers Ca 2ϩ -CaM regulation,  Drosophila kinesin with spacerϩCBD). B, detection of DKϩSϩCBD motor protein expressed in E. coli. The soluble proteins from the induced cultures (I) were purified (P) using CaM-Sepharose column and were separated on two SDS-containing gels. One gel was stained (Stained gel), and the second gel was blotted and probed with HRP-CaM. V, induced proteins (I) from E. coli cells containing pET28b vector without any insert. Note: the DKϩSϩCBD protein migrates as a higher molecular mass protein than its calculated molecular mass of 60 kDa. C, Ca 2ϩ -CaM inhibits DKϩSϩCBD protein interaction with MTs in the pelleting assay. The purified DKϩSϩCBD protein was incubated with MTs in the presence or absence of various test compounds and processed as described in the legend for Fig. 2. Abbreviations are as explained for Fig. 2. D, Ca 2ϩ -CaM inhibits MT-stimulated ATPase activity of the DKϩSϩCBD chimeric protein. The ATPase activity was performed with motor or motor with the MTs in the presence of CaCl 2 or CaM2 alone or Ca 2ϩ -CaM or Ca 2ϩ -CaM with EGTA as indicated by plus or minus signs for the presence or absence of tested components in each reaction. DϩSϩC refers to DKϩSϩCBD.
we generated a CBDϩSϩNCD construct. In this construct, we fused the CBD to the N terminus of the motor domain, and the CBD is separated from the motor by a spacer of 140 amino acids (the entire coiled-coil and neck regions of NCD) (Fig. 7A). The bacterial expressed chimeric protein bound CaM2 and CaM-Sepharose in a Ca 2ϩ -dependent manner (Fig. 7B), suggesting that the location of the CBD and the increased spacer between the CBD and motor do not affect the interaction of this motor with CaM. To analyze the regulation of this chimeric motor function by Ca 2ϩ -CaM, the purified motor protein (CBDϩSϩNCD) was used in MT-pelleting and ATPase assays. Surprisingly, Ca 2ϩ -CaM also regulated motor functions of CBDϩSϩNCD negatively. The CBDϩSϩNCD protein cosedimented with the MTs in the presence of AMP-PNP and Ca 2ϩ or CaM2 alone (Fig. 7C). However, in the presence of AMP-PNP and Ca 2ϩ -CaM2, it remained in the supernatant fraction (Fig.  7C, lane Mt/Ca CaM2). The MT-stimulated ATPase activity of the CBDϩSϩNCD was also inhibited in the presence of Ca 2ϩ -CaM by 89% (Fig. 7D). The inhibitory effect of Ca 2ϩ -CaM on ATPase activity was reversed in the presence of EGTA. These data suggest that neither increased distance between the CBD and the motor nor the location of the CBD in the protein (N or C terminus to the motor) influence Ca 2ϩ -CaM regulation of the chimeric motor.
CBD Did Not Confer Ca 2ϩ -CaM Regulation in Trans-To test whether the CBD acts as a modular domain in trans, we used purified 0.4C (CBD) and 1.0C (motor domain) KCBP proteins in cosedimentation and ATPase assays. As shown in Fig.  8A, the purified CBD protein remained in the supernatant fraction in the cosedimentation assay in the presence of AMP-FIG. 7. Ca 2؉ -CaM interacts and inhibits Drosophila chimeric NCD protein with an N terminus CBD and a spacer between the CBD and motor. A, schematic diagram of the Drosophila chimeric NCD with plant kinesin CBD at its N terminus and a long spacer that consists of coiled-coil and neck regions (140 amino acids) between the CBD and the motor (CBDϩSϩNCD). B, detection of CBDϩSϩNCD motor protein expressed in E. coli. The soluble proteins from the induced cultures (I) were purified (P) using a CaM-Sepharose column and were separated on three SDS-containing gels. One gel was stained (Stained gel), and two gels were blotted and probed with CaM (HRP-CaM) or T7-tag antibody (T7-tag.Ab). V, induced proteins (I) from E. coli containing pET28b vector without any insert. Note: the CBDϩSϩNCD protein migrates as a higher molecular mass protein than its calculated molecular mass of 68 kDa. C, Ca 2ϩ -CaM inhibits the interaction of CBDϩSϩNCD protein with MTs in the pelleting assay. The purified CBDϩSϩNCD protein was incubated with MTs in the presence or absence of various test compounds and processed as described in the legend for Fig. 2. Abbreviations are as explained in Fig. 2. D, Ca 2ϩ -CaM inhibits MT-stimulated ATPase activity of the CBDϩSϩNCD chimeric protein. The ATPase activity was performed with motor or motor with the MTs in the presence of CaCl 2 or CaM2 alone or Ca 2ϩ -CaM or Ca 2ϩ -CaM with EGTA. The presence or absence of tested components in each reaction is indicated by plus or minus signs. CϩSϩN refers to CBDϩSϩNCD.

FIG. 8. CBD protein does not confer Ca 2؉ -CaM regulation in trans.
A, the purified CBD (K 0.4C) protein did not bind MTs. B, addition of the CBD protein to pelleting assays with KCBP motor domain did not inhibit motor interaction with MTs. Purified motor and/or CBD of KCBP were incubated with MTs in the absence or presence Ca 2ϩ , CaM, or Ca 2ϩ -CaM and processed as described in the legend for Fig. 2. Because the K 1.0C protein comigrates with MTs, one gel was stained (Stained gel), and another gel was blotted and probed with T7-tag antibody (T7-tag.Ab). The position of 0.4C, CaM, 1.0C KCBP motor, and MTs were shown. Abbreviations are as explained in the legend for Fig. 2. C, Ca 2ϩ -CaM did not inhibit MT-stimulated ATPase activity of KCBP motor in the presence of the CBD protein. The ATPase assays were performed as described under "Experimental Procedures." The presence or absence of various components in ATPase reaction is indicated by plus or minus signs, respectively. PNP and irrespective of the presence of Ca 2ϩ or AtCaM2 or both. These results suggest that the CBD protein alone does not interact with MTs. We then tested CBD and 1.0C KCBP (motor without CBD) in cosedimentation (Fig. 8B) and ATPase (Fig. 8C) assays. In the cosedimentation assay, the motor (1.0C) and CBD of KCBP remained in the supernatant fraction in the absence of MTs (Fig. 8B, No Mt). However, when incubated with MTs in the presence of AMP-PNP, the 1.0C cosedimented with MTs, but CBD remained in the supernatant fraction (Fig. 8B, Mt), suggesting that Ca 2ϩ ⅐CaM⅐CBD and motor do not form a complex. This interaction of 1.0C KCBP protein with MTs was not affected in the presence of the CBD protein, Ca 2ϩ , CaM2, or all of them (Fig. 8B, Mt/Ca CaM2). These data indicate that the CBD does not confer Ca 2ϩ -CaM regulation in trans, as it does in cis (motor with CBD; see Fig. 2, K 1.5C). The Mt-stimulated ATPase activity of 1.0C KCBP motor protein was also not inhibited in the presence of CBD protein in trans, Ca 2ϩ , CaM2, or all together (Fig. 8C). These results confirm that the CBD of KCBP acts as a modular domain only in cis (Fig. 2, K 1.5C) but not in trans (Fig. 8).

The CaM-binding Domain of Plant KCBP Functions as a Regulatory Module in N-or C-terminal Drosophila NCD and
DK Motors-In this study we used C-and N-terminal motors, because the primary sequence and structural elements are highly conserved between them (6,29,44). We added the CBD to either the C terminus of DK (DKϩCBD with a 49-amino acid spacer and DKϩSϩCBD with a 149-amino acid spacer) or C terminus (NCDϩCBD with a 36-amino acid non-coiled-coil spacer) or N terminus (CBDϩSϩNCD with a 140-amino acid coiled-coil and neck spacer) of NCD. All chimeric kinesins with a CBD bound CaM in a Ca 2ϩ -dependent manner. In blot overlay assay, crude and purified chimeric proteins with a CBD, but not without a CBD, bound HRP-CaM (see Figs. 1B, 6B, and 7B) and 35  , indicating that the CBD is effective at a distance from the motor domain in inhibiting motor-MT interaction. We used various lengths of coiled-coil spacer (DKϩCBD, DKϩSϩCBD, and CBDϩSϩNCD) or non-coiled-coil spacer (NCDϩCBD), and both showed similar results (see Figs. 3 to 7). Interestingly, Ca 2ϩ -CaM regulation of chimeric motors was not affected by the location of the CBD. The presence of a CBD either at the C or N terminus to the motor with or without spacer conferred similar regulation where the interaction of motor with MTs was disrupted (see Figs. 2 to 7). However, no Ca 2ϩ -CaM regulation was observed if motor domain and CBD were added separately to the pelleting assay or ATPase assay (Fig. 8), indicating that the CBD functions only in cis but not in trans. The phylogenetic relationship of the motors revealed that NCD and KCBP fall in C-terminal and DK falls in Nterminal kinesin subfamilies (7) with opposite polarities and different functions (1,22). Despite this, all chimeric motors are regulated in a similar manner. However, the x-ray crystal structures of DK and NCD (44 -46) have shown that N and C termini of these motors are close together and on the opposite side from the MT interface. Hence the CBD in our constructs is expected to be on the opposite side of the MT-interacting elements. Based on these results, we conclude the CBD in CaMbinding kinesins functions as a module in conferring Ca 2ϩ -CaM regulation to both minus-and plus-end motors likely through disruption of MT-motor-interacting elements.
Bovine CaM and three Arabidopsis CaM isoforms were effective in regulating chimeric motors. Arabidopsis has seven conserved CaM isoforms and many CaM-like proteins (2,47). AtCaM2 differs in five (D7E, R74K, D122E, K126R, and Y138H) and two (T117S and K126R) amino acids from CaM4 and CaM6, respectively, whereas CaM4 differs from CaM6 in six amino acids (E7D, K74R, T117S, E122D, H138Y, and I144V). Bovine CaM differs from Arabidopsis CaM2, CaM4, and CaM6 in 19, 20, and 18 amino acids, respectively. Despite these differences at the amino acid level, plant and animal CaMs are similarly effective in regulating chimeric motor functions, MT-binding and MT-stimulated ATPase activities. Kinesins interaction with MTs, and their MT-stimulated ATPase activity has been the subject of intensive research to address the mechanisms of regulation of the motor/MT-interacting elements and ATP hydrolysis-associated force generation. To address the structural changes that take place in the motor domain during kinesin interaction with MTs and ATP hydrolysis, several biophysical and molecular techniques such as atomic resolution, alanine-scanning mutagenesis, proteolytic mapping, and cryoelectron microscopy of native and mutant kinesins have been used extensively (6, 24, 44, 45, 48 -51). These studies revealed that MT-interacting elements L8, L11, ␣-helix 4, L12, and ␣-helix 5 and nucleotide-interacting elements N1 to N4 are highly conserved in all kinesins. Mutations in these regions have confirmed their role in motor functions. For example, mutations in highly conserved amino acid residues in ␣-helix 5 (T291M), L11 (N600K), switch II (E585A), and switch I (R552A) regions have been shown to alter motor functions (52)(53)(54)(55). The decoupling mutants (N600K and E585A in NCD) retain MT-binding and basal ATPase activities but showed no MT-stimulated ATPase activity (52)(53)(54). Thus they are defective in communicating the signal from MT binding sites (L11 and ␣-helix 4) to ATP binding sites (switch II-switch I) in the proposed structural signaling pathway (MT-binding 3 salt bridge 3 active site) (53).
Unlike decoupling mutants that block intramolecular relay between MT-binding and salt bridge regions (53), Ca 2ϩ -CaM may regulate the structural confirmation of MT-binding regions reversibly in CBD-containing kinesins resulting in disruption of interaction between motors and MTs. These Ca 2ϩ -CaM disrupted structural changes in the motor precede the intramolecular structural signaling pathway between MTbinding and ATP-hydrolysis cycle. Thus, binding of Ca 2ϩ -CaM to kinesin prevents motor interaction with MTs. The CBD containing kinesins could be useful in dissecting CaM-induced structural changes in the MT-binding sites. Structure at atomic resolution in free and Ca 2ϩ -CaM-bound states, coupled with proteolytic mapping studies of these motors in the presence or absence of Ca 2ϩ -CaM and MTs, are necessary to elucidate the mechanisms by which Ca 2ϩ -CaM regulates CaM-binding kinesins negatively.
Possible modes of Ca 2ϩ -CaM Regulation of Chimeric Motors with MTs-Our results (see Figs. 2 to 8), together with our previous studies (34 -36,38), suggest that the inhibitory effect of Ca 2ϩ -CaM is likely because of the disruption of MT-binding elements in the motor. This disruption of motor-MT interaction by activated CaM could be achieved through two possible mechanisms: (i) once activated CaM binds the CBD the CBD⅐Ca 2ϩ ⅐CaM complex may mask the MT-interacting sites on the motor thereby blocking the interaction of motor with the MTs, or (ii) the binding of CaM to CBD results in structural change in the motor and alters charges in MT-binding sites and thereby prevents the charge-based interaction between the Cterminal acidic region of the tubulin subunits and the MTinteracting basic regions of the motor. These mechanisms are not mutually exclusive.
The fact that the CBD does not confer Ca 2ϩ -CaM regulation in trans suggests that the CBD⅐Ca 2ϩ ⅐CaM complex does not bind the motor domain. It is also possible that the binding of the CBD⅐Ca 2ϩ ⅐CaM complex to the motor domain may be weak but can occur in cis when this complex is close to the motor domain. The CBD⅐Ca 2ϩ ⅐CaM complex in cis could achieve the effect of a very high level of local concentration of CBD⅐Ca 2ϩ ⅐CaM around each motor that is sufficient to bind and block the motor even though the binding is weak. This would suggest that the CBD itself has sufficient flexibility to permit the binding regardless of its attachment to the N or C terminus. Moreover, short or long linkers fused to the CBD would appear to permit a similar flexibility. Alternatively, the CBD⅐Ca 2ϩ ⅐CaM complex in cis may block MT-interacting sites on the motor domain.
The change in charge of the MT-interacting elements could be because of either retraction of positive charges or display of negative charges or both around the CBD region of the motor. Both the CBD region and MT-interacting elements of the motor contain positively charged regions that interact with negatively charged regions of their respective counterparts, Ca 2ϩ -CaM and MTs (10,50). Pelleting and MT-dependent ATPase assays with KCBP 1.5C (motor with CBD) in the presence of high salt (1 M NaCl) showed an inhibitory effect that is similar to Ca 2ϩ -CaM (data not shown), suggesting Ca 2ϩ -CaM may alter the charge-based interaction between the motor and MTs. Because of higher affinity between the CBD and Ca 2ϩ -CaM than between motor and MT (K d 12.8 versus 65 nM) (36,38), the motor⅐Ca 2ϩ ⅐CaM complex so formed not only prevents chimeric motor interaction with MTs (see Figs. 3 to 7) but also dissociates KCBP 1.5C motor from the preformed motor⅐MT complex (36). However, in the absence of Ca 2ϩ -CaM, there is no influence of the CBD region on the interaction of the motor with MTs, as motors with or without a CBD show similar MTbinding and MT-stimulated ATPase activities (see Figs. 2 to 8). Moreover, the CBD region alone does not bind MTs (Fig. 8A). Therefore, the CBD in the presence of activated CaM may disrupt the positive-negative charge-based interaction between motor and MT by either one or both mechanisms.
In summary, we have demonstrated that the CBD functions as a module in conferring Ca 2ϩ -CaM regulation to CBD-containing motors, and this regulation is not dependent on location or proximity of the CBD with respect to the motor domain. Furthermore, the CBD functions only in cis. Our studies also suggest that naturally occurring CaM-binding kinesins in plants (KCBP) and animals (kinesin-C) may have evolved from ancestral kinesins by fusion of a CBD.