A metal switch for controlling the activity of molecular motor proteins.

Kinesins are molecular motors that require a divalent metal ion (for example, Mg(2+)) to convert the energy of ATP hydrolysis into directed force production along microtubules. Here we present the crystal structure of a recombinant kinesin motor domain bound to Mn(2+) and ADP and report on a serine-to-cysteine substitution in the switch 1 motif of kinesin that allows its ATP hydrolysis activity to be controlled by adjusting the ratio of Mn(2+) to Mg(2+). This mutant kinesin binds ATP similarly in the presence of either metal ion, but its ATP hydrolysis activity is greatly diminished in the presence of Mg(2+). In human kinesin-1 and kinesin-5 as well as Drosophila melanogaster kinesin-10 and kinesin-14, this defect is rescued by Mn(2+), providing a way to control both the enzymatic activity and force-generating ability of these nanomachines.

INTRODUCTION

Purine nucleotide-binding proteins, including small monomeric G-proteins, protein-synthesizing GTPases (elongation factor), heterotrimeric G-proteins, thymidine kinases, RNA helicases (CSFV NS3 helicase), rad50, RecA, myosin, dynein, and kinesin superfamilies, constitute a large portion of the hydrolases (EC 3.6) found in biological systems. These enzymes are crucial for a diverse range of cellular processes, but all use a phosphate-binding loop (P-loop or Walker A motif; consensus sequence GxxxxGKS/T) to bind the nucleotide phosphates and all require a metal ion (typically magnesium, Mg2+) in the active site [reviewed in [1]]. The metal ion coordination usually includes a conserved serine or threonine in the P-loop, an oxygen of the β-phosphate, and two water ligands. The remaining two ligands required to satisfy the octahedral coordination geometry vary depending on whether nucleotide diphosphate or triphosphate occupies the site. Therefore, the metal ion anchors an intricate network of interactions in the enzyme that are believed to be necessary to lower the free energy of the transition state and achieve an enhanced rate of hydrolysis.

Molecular motors, such as kinesins and myosins, as well as molecular switches, such as G-proteins, utilize the energy from nucleotide hydrolysis to perform various cellular tasks. In addition to the P-loop, these proteins use similar structural motifs to sense and respond to the presence or absence of the γ-phosphate of the nucleotide called switch 1 and switch 2 (NxxSSR and DxxGxE are consensus sequences in kinesins and myosins, respectively) and coordinate nucleotide hydrolysis[2]. In contrast to these structural similarities, there are major differences in the mechanistic timing for the “closing” and “opening” of these switches around the nucleotide during the NTPase cycle. The conformational changes of the switches are propagated beyond the active site to relay a structural signal that conveys the state of the nucleotide bound to the enzyme[3]. A complete understanding of the mechanisms of nucleotide binding and hydrolysis requires a detailed description of these conformational changes and other molecular interactions within the active site that initiate force production in molecular motors or turn a molecular switch “on” and “off”.

Thio-substitution and metal ion rescue experiments were developed to characterize the metal catalyzed reactions of different classes of ribozymes [reviewed in [4]]. This provided a powerful and direct probe of catalytically important metal-nucleotide interactions within RNA macromolecules[5-8], as well as protein-based enzymes such as DNA transposases[9-11]. Mg2+ is a relatively hard metal ion (high charge density, low polarizability) that participates in catalysis through coordination of oxygens from hydroxyls and/or phosphates. In previous metal rescue experiments, oxygen was substituted with sulfur, which does not bind as tightly to Mg2+ but has a higher affinity for softer transition metal ions with lower charge density and somewhat greater polarizability, such as Mn2+ or Zn2+(. Therefore, in previous studies, the addition of Mn2+ resulted in the catalytic “rescue” of the sulfur-substituted macromolecule, which provided evidence for metal binding sites that are required for the catalytic activity of many different ribozymes[4].

In this study, we set out to develop a similar strategy to probe metal interactions within kinesin molecular motors by taking advantage of the differential affinities of Mg2+ and Mn2+ for serine (−OH) and cysteine (−SH) amino acids. Substitution of an active site serine with cysteine is expected to diminish catalytic activity if the serine interaction with Mg2+ is critical for catalysis. Substituting Mg2+ with Mn2+, which interacts more strongly with cysteine, should therefore restore the metal interaction with this residue, thus rescuing catalysis.

RESULTS

Kinesin•MnADP structure shows that Mn2+ replaces Mg2+

In the present study, we have investigated monomeric and dimeric proteins of different kinesin families (Kin1, Kin5, Kin10, and Kin14) to characterize our metal switch technology across the superfamily. As with other protein-based systems, certain kinesin proteins afford the ability to answer different structural, biochemical, and mechanistic questions related to the ATPase cycle given their unique biophysical properties. In this manuscript, monomeric Kin5 (Eg5) was predominantly used for biochemical assays since this motor can be purified nucleotide-free without loss of stability or activity[13]. This kinesin provides the ideal model system for designing specific experiments to assess the complex ATPase mechanism of kinesin motors in the presence of different metals such as Mg2+ or Mn2+. However X-ray crystallographic studies and in vitro MT motility assays were conducted using Kin10 (Nod) and Kin1 (conventional kinesin), respectively.

We began by purifying and crystallizing the motor domain of a Drosophila melanogaster kinesin (Nod) with Mn2+ and ADP in order to solve its high resolution structure using X-ray crystallography. The identity of the bound metal in our kinesin crystals was determined by electron paramagnetic resonance (EPR) spectroscopy (Fig. 1a) and F difference mapping using the Kin10•MgADP model for initial refinement (Fig. 1b)[14]. EPR data for the washed and denatured crystalline sample had the distinctive six-peak spectrum typical of Mn2+ and a positive density peak in the F map was observed at the metal binding site. The Kin10•MnADP structure was refined to 2.6 Å (Table 1; Supplementary Figure 1) and aligned with Kin10•MgADP with a root mean square deviation (RMSD) at 0.32 Å2 (Fig. 1c), suggesting that the overall structure was very similar, despite Mn2+ being bound at the active site. The Mn2+ ion showed octahedral oxygen coordination consisting of the nucleotide β-phosphate, the P-loop serine and four waters (Fig. 1d,e).

Figure 1

Crystal structure of Kin10•MnADP and the ATP hydrolysis-competent state for kinesin. (a) EPR spectra of washed and denatured Kin10•MnADP crystals and mock control (no crystals) are shown. (b) The active site of Kin10•MnADP is shown. Water molecules are shown as red spheres and Mn2+ as gray sphere. The green cage represents the positive peak in the F map (3 σ) with the MnADP data being refined with the Kin10•MgADP structure (3DC4). (c) Alignment of main chain atoms of Kin10•MnADP with Kin10•MgADP. (d,e) Shown are detailed views of the active site of Kin10•MnADP and Kin10•MgADP, respectively. Mg2+ is shown as a green sphere. (f) Model of the ATP hydrolysis-competent state of kinesin. AMPPNP from the Kin5 structure [3HQD; [15]] was replaced with ATP from Myo2 structure [1FMW; [27]] after alignment using the P-loop residues.

TABLE 1

Data collection and refinement statistics.

	MnADP a
Data Collection
Space Group	P212121
Cell Dimensions
a, b, c (Å)	47.6, 75.7, 94.2
α, β, γ,(°)	90, 90, 90
Resolution (Å)	20-2.6 (2.76-2.60)
Rsym (%)b,c	9.4 (42.9)
I / σ(I) b	24.5 (4.3)
Completeness (%) b	99.8 (99.9)
Redundancy b	7.10 (7.22)
Refinement
Resolution (Å)	20 – 2.6
No. Reflections b	77485 (15497)
Rwork / Rfreed (%)	22.7 / 27.0
No. atoms
Protein	2247
Ligand/ion	29
Water	28
B-factors
Protein	32.7
Ligand/ion	25.9
Water	44.7
R.M.S. Deviations
Bond Length (Å)	0.009
Bond Angle (°)	1.3

PDB ID is 3PXN. One crystal used for structure determination.

Data in parentheses represent the highest resolution shell.

R = 100 × ΣhΣi | I(h) – I(h) | / ΣhΣi I(h), where I(h) and I(h) values are the ith and mean measurements of the intensity of reflection h.

R is for 5 % of total reflections not included in the refinement.

Switch 1 substitution shows similar metal and ATP binding

Wild-type kinesins have a conserved serine in switch 1 (NxxSR) that is thought to coordinate the metal during ATP hydrolysis (Fig. 1f)[15]. Site-directed mutagenesis was used to replace this metal-interacting serine residue in human Kin5(WT) with cysteine [Kin5(SC)]. Since EPR signals from Mn2+ bound to proteins are often not observed at common X-band frequencies[16], EPR detection of free Mn2+ was used to monitor Mn2+ binding to Kin5(WT) and Kin5(SC) (Fig. 2a; Supplementary Fig. 2). Similar weak binding of MnADP (reduced free Mn2+ signal), which was displaced by excess Mg2+, was found with both enzymes. These results indicate that the switch 1 serine-to-cysteine substitution does not alter metal binding in the presence of ADP.

Figure 2

Metal and nucleotide binding to kinesin. (a) Partial Mn2+ EPR spectra for Kin5(WT) and Kin5(SC) under various conditions: Mn2+ alone (black), Mn2+ and nucleotide-free Kin5 (cyan), MnADP (red), MnADP plus Kin5 (green), and Kin5•MnADP plus excess Mg2+ (purple). Final concentrations: 50 µM MnCl2, 50 µM ADP, 50 µM Kin5, 5 mM MgCl2. (b) Tryptophan fluorescence enhancement of Kin5(WT) and Kin5(SC) upon rapid mixing in a stopped-flow instrument with MgATP or MnATP (as indicated). Final concentrations: 5 µM Kin5, 500 µM ATP, 5 mM MgCl2 or MnCl2. (c) Observed rates of each fluorescent transient were plotted against ATP concentration. Each data set was fit to a hyperbola to determine the K (mean ± s.e.m.): WT+Mg2+ = 6.2 ± 1.4 µM; WT+Mn2+ = 2.5 ± 0.5 µM; SC+Mg2+ = 10.4 ± 3.2 µM; SC+Mn2+ = 21.8 ± 7.0 µM. (d) Relative amplitudes were plotted as a function of ATP concentration. The data were fit to a hyperbola: WT+Mg2+ = 1.4 ± 0.1 µM; WT+Mn2+ = 2.7 ± 0.2 µM; SC+Mg2+ = 78.6 ± 4.4 µM; SC+Mn2+ = 21.6 ± 1.5 µM.

Similar rates of MgATP or MnATP binding to Kin5(WT) and Kin5(SC) were observed by monitoring changes in intrinsic tryptophan fluorescence, which changes upon ATP binding (Fig. 2b)[13]. When these experiments were performed as a function of ATP concentration, the overall ATP affinity was comparable for Kin5(WT) with Mg2+ or Mn2+ (K = 6.2 µM and 2.5 µM, respectively; Fig. 2c). However, for Kin5(SC) the binding affinity was slightly weakened with Mg2+ or Mn2+ (10.4 µM and 21.7 µM, respectively). When comparing the amplitudes of each transient as a function of ATP concentration, differences in the overall thermodynamics for substrate binding were observed (Fig. 2d), indicating a substantial change in the SC kinesin ATPase cycle downstream from the MgATP-dependent isomerization of the enzyme.

Kinesin activity controlled using [Mn2+]/[Mg2+] ratio

The steady-state ATPase activity of Kin5(WT) and Kin5(SC) was monitored in the absence and presence of microtubules (MTs)[17]. For Kin5(WT), the maximum rate of MgATP turnover was similar to that of MnATP in both the absence and presence of MTs (Fig. 3a,b, respectively). For Kin5(SC), however, the rate of MgATP turnover showed between 83–94 percent reduction, but could be rescued to WT levels with MnATP. At a high metal concentration (5 mM), the ATPase activity of Kin5(SC) could be controlled by varying the [Mn2+]/[Mg2+] ratio in the absence (Fig. 3c) or presence (Fig. 3d) of MTs.

Figure 3

Mn2+ rescue of kinesin (SC) ATPase in the absence and presence of MTs. (a) Time course of MgATP and MnATP hydrolysis by Kin5(WT) and Kin5(SC) in the absence of MTs and (b) in the presence of MTs. Final concentrations: 0.5 µM Kin5 (no MTs), 50 nM Kin5 plus 2 µM MTs, 5 µM taxol, 200 µM ATP, 5 mM MgCl2 or MnCl2. (c,d) The observed rate of ATP hydrolysis for Kin5(SC) plotted against [Mn2+]/[Mg2+] ratio in the absence and presence of MTs. (e) Time course of MgATP and MnATP hydrolysis by Kin1, Kin10, and Kin14 as indicated. Final concentrations: 4 µM Kin1, 10 µM Kin10, 10 µM Kin14, 200 µM ATP. Representative data shown for n=3 experiments for each time course.

Metal rescue observed for divergent kinesin motors

To test this metal rescue in other members of the kinesin superfamily, the corresponding switch 1 serine was substituted with cysteine in the motor domain of Homo sapiens kinesin-1 (Kin1 or conventional kinesin), D. melanogaster kinesin-10 (Kin10 or Nod), and D. melanogaster kinesin-14 (Kin14 or Ncd) (Supplementary Table 1). These well-characterized kinesins represent diverse properties of the superfamily, including different locations of the motor domain (Kin1, Kin5, Kin10 are N-terminal; Kin14 is C-terminal), differences in processivity (Kin1 and Kin5 are processive; Kin10 and Kin14 are non-processive), and differences in motility (Kin1, Kin5, Kin14 are motile; Kin10 is non-motile). Metal rescue was observed for all kinesins tested (Fig. 3e–g), indicating that this phenomenon was found throughout the superfamily and that metal-based control of the catalytic activity of specific kinesin motors is feasible.

ATP hydrolysis reaction affected by switch 1 substitution

Acid-quench experiments were performed to measure the presteady-state kinetics of MgATP and MnATP hydrolysis in order to determine which step of the mechanism was defective in the SC kinesin mutant. These experiments were conducted at high enzyme concentration (75 µM) to monitor phosphate production during the first and subsequent ATP hydrolysis events. If ATP hydrolysis is defective in Kin5(SC), then linear phosphate production over time would be observed. However, if phosphate release is affected, then an exponential burst of phosphate product would be expected. Fig. 4a shows linear kinetics of phosphate formation in the presence of MgATP. However, a robust burst of phosphate production was observed for MnATP, as previously observed for Kin5(WT)[13]. These results indicate that the catalytic step of ATP hydrolysis is aberrant in the SC kinesin mutant with Mg2+, and that hydrolysis can be rescued by Mn2+, presumably due to stronger interaction with the cysteine.

Figure 4

Acid-quench and phosphate release kinetics. (a) Acid quench kinetics of phosphate product formation by Kin5(SC) with MgATP and MnATP (as indicated). Dash line corresponds to expected kinetics if the phosphate release step were slowed. Final concentrations: 75 µM Kin5, 200 µM ATP, 5 mM MgCl2 or MnCl2. (b,c) Phosphate release kinetics by Kin5(WT) and Kin5(SC) with MgATP and MnATP in the absence and presence of MTs using the MDCC-PBP coupled assay. Final concentrations: 1 µM Kin5, 0 or 2 µM MTs, 0 or 5 µM taxol, 10 µM MDCC-PBP, 0.05 U/ml PNPase, 75 µM MEG, 200 µM ATP, 5 mM MgCl2 or MnCl2.

To directly measure the kinetics of phosphate release in real time, we performed stopped-flow experiments using the MDCC-PBP assay[18]. In this coupled assay, ATP binds the active site of Kin5 followed by ATP hydrolysis. When Pi is released to the solution, MDCC-PBP binds the Pi rapidly and tightly resulting in a fluorescence enhancement (Fig. 4b). Similar phosphate release kinetics were observed for Kin5(WT) with MgATP and MnATP. Kin5(SC) showed a slow linear rate of Pi release with Mg2+ but this was rescued to WT levels by Mn2+ in the absence (Fig. 4b) and presence (Fig. 4c) of MTs. The slow linear rate of Pi release for Kin5(SC) with Mg2+ was consistent with slow linear rate of ATP hydrolysis that precedes the Pi release step.

Metal rescue of microtubule gliding by mutant kinesin

Motility assays with dimeric Kin1(WT) and Kin1(SC) were undertaken to determine whether the SC mutant motor promotes MT gliding in vitro (Supplemental Movie 1)[19]. Kin1 shows rapid and robust motility under similar conditions as our biochemical assays[19], thus it provides an excellent model system for investigating the mechanical force generation by a kinesin motor. After the Kin1 and rhodamine-labeled MTs were perfused into a microscope cover glass chamber, MT gliding was initiated by addition of MgATP or MnATP. MT movement of 37 ± 9 and 29 ± 8 µm/min (mean ± s.d.) was found for Kin1(WT) with Mg2+ and Mn2+, respectively (Fig. 5a; Supplementary Table 2). However, Kin1(SC) showed very slow motility of 0.4 ± 0.2 µm/min with Mg2+ (Supplemental Movie 2), and partial rescue to 3.5 ± 1.4 µm/min by Mn2+ (Fig. 5b). Lack of full rescue to WT velocity could be due to slower ATPase activity for Kin1(SC) with MnATP (Fig. 5c), but cooperativity between the two motor domains could also be altered. Nevertheless, these results demonstrate that Mn2+ is able to rescue MT-based motility for the SC kinesin mutant.

Figure 5

Mn2+ rescue of in vitro MT gliding. (a,b) Dimeric Kin1(WT) and dimeric Kin1(SC) promoted MT gliding in MgATP and MnATP. Asterisk highlights leading MT end. (c) Time course of MgATP and MnATP hydrolysis by dimeric Kin1(WT) and Kin1(SC) in the presence of MTs. Final concentrations: 10 nM Kin1, 2 µM MTs, 5 µM taxol, 200 µM ATP, 5 mM MgCl2 or MnCl2. WT+Mg2+ (mean ± s.e.m.): 25.9 ± 1.5 s−1, WT+Mn2+: 40.4 ± 1.9 s−1, SC+Mg2+: 2.1 ± 0.9 s−1, and SC+Mn2+: 16.6 ± 0.6 s−1. Representative data shown for n=3 experiments for each time course.

DISCUSSION

We have defined the minimal kinesin ATPase mechanism in the presence of different divalent metals when the conserved switch 1 serine was substituted with cysteine. This analysis has revealed a direct link between the switch 1 serine interaction with the metal and the hydrolytic step in the cycle. When this switch 1 serine was substituted with cysteine, its interaction with Mg2+ was presumed to be weakened due to sulfur being a poorer ligand for the hard Mg2+ ion. Very slow ATP hydrolysis was observed for the cysteine mutants of all kinesins tested, both in the absence and presence of MTs, suggesting that the mutant enzymes still possess the ability to catalyze ATP hydrolysis, albeit at a markedly reduced rate. However, normal hydrolysis rates were found for the SC mutant with Mn2+ due to stronger interaction of sulfur with this softer metal ion.

Mn2+ capable of replacing Mg2+ for kinesin

Mg2+ and Mn2+ share several chemical properties, including ionic radius and preferred octahedral coordination geometry, yet they differ in electronegativity, charge density and polarizability. For many Mg2+-dependent ATPases, Mn2+ is a suitable alternative metal cofactor for their biochemical reactions. We provide a high resolution structure of Kin10(WT) bound to MnADP, and comparison of this structure with the MgADP state[14] reveals that there is virtually no difference between the structures. These structural data are consistent with ATPase measurements that show very similar rates of ATP turnover for (WT) kinesin with Mg2+ or Mn2+ (Fig. 2), as well as presteady-state MgADP and MnADP release experiments that show similar kinetics for mantADP release from Kin5(WT) and Kin5(SC) in the absence and presence of MTs (Supplementary Fig. 3). Together, results presented in this study are consistent with Mn2+ structurally and functionally replacing Mg2+ as a metal cofactor for ATP hydrolysis in kinesin motors.

Switch 1-metal interaction during kinesin ATPase cycle

It is commonly thought that switch 1 must close around the nucleotide to reach the hydrolysis-competent state[15]. A key interaction that occurs during switch 1 closure is coordination of the central serine (NxxSR) to the metal in the active site. When non-hydrolyzable nucleotide analogues are bound to myosin[20] and kinesin[15], as well as small G-proteins[21], this closed switch 1 conformation has been observed with the conserved switch 1 serine (or threonine for G-proteins) coordinated to the metal. In kinesins, a recent study by Parke et al.[15] revealed many new insights into the mechanism of ATP hydrolysis by kinesins. Although the “two-waters mechanism” proposed by Parke et al. seems highly favorable, the research presented in this manuscript does not directly address this hypothesis. However, we believe the metal coordination plays a key role in establishing the proper conformation of the kinesin active site for ATP hydrolysis to occur rapidly. Not only does the switch 1 serine coordinate the metal in the ATP hydrolysis-competent state, but the preceding switch 1 serine (SRSH) hydrogen bonds to the γ-phosphate of the nucleotide, the polypeptide backbone makes critical interactions with the γ-phosphate and nucleophilic water, and the switch 1 arginine forms a salt bridge with the conserved switch 2 glutamic acid. Our results support the hypothesis that switch 1 serine coordination of the metal represents the keystone interaction that facilitates rapid closure of Sw1 to promote rapid ATP hydrolysis.

One explanation for our results has switch 1 in the SC mutant remaining predominantly in the open conformation upon ATP binding, and thus not coordinating the Mg2+. This open switch 1 conformation in the presence of Mg2+ would explain the very slow hydrolysis rate. However, in the presence of Mn2+, switch 1 can close through its interaction with this metal and achieve the ATP-hydrolysis competent state. Alternatively, switch 1 in the SC mutant does close upon ATP binding and the cysteine coordinates the Mg2+, but the sulfur–Mg2+ interaction modulates the Lewis acidity of the Mg2+ thereby altering the phosphate charge density required for rapid hydrolysis. When Mn2+ is substituted as the metal cofactor, the electrostatics in the active site are restored, and thus the kinetics of the hydrolysis reaction are rescued. Future investigations will be directed to test these hypotheses.

Switch 1 as “sensor” for productive kinesin-1 stepping

We performed in vitro MT gliding assays using dimeric kinesin-1 WT and SC in the presence of Mg2+ and Mn2+. For WT, we saw comparable MT gliding rates with either metal and the observed rate was similar to previously published work[19]. For SC in the presence of Mg2+, we observed very slow MT motility (~two orders of magnitude reduction compared to WT), even though we only measured an ~order of magnitude reduction in MgATPase rate. However, upon addition of Mn2+, we observed an ~order of magnitude enhancement of both MT motility and MnATPase activity. Since we utilized an ATP-regeneration system and performed these experiments at high ATP concentration (5 mM), these results are likely not associated with weaker binding of ATP to SC. This result was dependent neither on motor density on the surface of the coverslip (half concentration and double concentration tested) nor on the length of MTs used (shorter MTs showed a comparable rate as longer MTs) nor on the specific protein preparation used (n=2). We propose that a switch 1 conformational change during the kinesin-1 ATPase cycle plays a critical role in coupling ATPase activity in the motor to productive force production that leads to stepping of the motor along the MT. Future studies using dimeric kinesin-5 and kinesin-10 SC mutants will determine if this phenomenon is common to processive motors (kinesin-1 and kinesin-5) or is shared among non-processive motors as well (kinesin-14).

Potential Broad Applications

To the best of our knowledge, this is first demonstration of metal rescue in a protein-based hydrolytic enzyme. Given the generality of this metal rescue approach across the kinesin family, it should be applicable to other nucleotide triphosphorylase (NTPase) systems, including myosins and G-proteins. Currently, one of the most powerful chemical-genetic methods available involves an approach pioneered by Shokat and colleagues in which a protein kinase target is mutated to create an additional pocket in the nucleotide binding region such that an inhibitor resembling a nucleotide with a bulky side group can be accommodated[22,23]. While such stable mutants have been successfully created in a number of different systems[24,25], in proteins that do not tolerate introduction of an additional cavity, a potential advantage of the metal rescue technique is that it only requires a conservative serine to cysteine mutation, which is unlikely to have an effect on the structure of the target enzyme. A further distinction between the two approaches is that the metal rescue does not require inhibitors or modified nucleotides; one can simply vary the ratio of Mg2+ and Mn2+ ions, which in turn can modulate target enzyme activity from inactive to wild type levels. Additionally, the ability to alter serine residues in different regions of the active site allows one to inhibit different steps in the catalytic cycle of an enzyme, thereby dissecting its mechanism. For example, while conserved serines in the P-loop are critical for nucleotide binding, the switch 1 serine affects nucleotide hydrolysis without affecting binding, as demonstrated in this work.

There are several relevant and important applications of our metal switch approach that will substantially impact the molecular motor community as well as more broadly to the P-loop NTPase research community. Examples of such applications include using the metal switch to control the activity of a single catalytic head of a dimeric motor (e.g. kinesin-1, myosin-5) to investigate biophysical questions of inter-molecular head-head communication, mechanical versus chemical gating, and motor stepping behavior. The metal switch could also be utilized to control the activity of a single type of molecular motor in multiple motor assays to distinguish the distinct roles of various motors involved in cytoskeletal structures (e.g. reconstituted mitotic spindles). Ultimately, this technology could be incorporated into the development of an in vivo system that tolerates high concentrations of Mn2+ (e.g. Saccharomyces cerevisiae[26]) to study various cell biological phenomena. Given the structural similarity between molecular motors (e.g. kinesins and myosins) and molecular switch proteins (e.g. small G-proteins), such a metal switch should work in other NTPase systems to provide a way to control their activity in vitro and in vivo. Finally, the substitution of the conserved P-loop serine or threonine (consensus: GxxxxGKS/T) with a cysteine could provide for metal rescue of nucleotide binding and hence restoration of activity in P-loop containing hydrolases (EC 3.6). This protein mega-family is made up of at least 10 superfamilies representing tens-of-thousands of enzymes found in virtually all biological systems.