HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 19, 17079-17087, May 10, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/19/17079    most recent M108793200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Farrell, C. M. Articles by Gilbert, S. P. Search for Related Content PubMed PubMed Citation Articles by Farrell, C. M. Articles by Gilbert, S. P. Social Bookmarking                      What's this?

The Role of ATP Hydrolysis for Kinesin Processivity^*

Christopher M. Farrell^§, Andrew T. Mackey^¶, Lisa M. Klumpp, and Susan P. Gilbert

From the Department of Biological Sciences, University of Pittsburgh, Pittsburgh, Pennsylvania 15260

Received for publication, September 12, 2001, and in revised form, January 10, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Conventional kinesin is a highly processive, plus-end-directed microtubule-based motor that drives membranous organelles toward the synapse in neurons. Although recent structural, biochemical, and mechanical measurements are beginning to converge into a common view of how kinesin converts the energy from ATP turnover into motion, it remains difficult to dissect experimentally the intermolecular domain cooperativity required for kinesin processivity. We report here our pre-steady-state kinetic analysis of a kinesin switch I mutant at Arg²¹⁰ (NXXSSRSH, residues 205-212 in Drosophila kinesin). The results show that the R210A substitution results in a dimeric kinesin that is defective for ATP hydrolysis and a motor that cannot detach from the microtubule although ATP binding and microtubule association occur. We propose a mechanistic model in which ATP binding at head 1 leads to the plus-end-directed motion of the neck linker to position head 2 forward at the next microtubule binding site. However, ATP hydrolysis is required at head 1 to lock head 2 onto the microtubule in a tight binding state before head 1 dissociation from the microtubule. This mechanism optimizes forward movement and processivity by ensuring that one motor domain is tightly bound to the microtubule before the second can detach.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Kinesin is a highly processive, dimeric mechanoenzyme that travels along microtubules toward their plus-ends in discrete 8-nm steps, each step tightly coupled to a single ATP turnover (1-3). Recent evidence from a variety of experimental approaches has focused our attention to the proposal presented by Rice et al. (4) that ATP binding induces a pronounced conformational change in the neck linker region, which docks the neck linker onto the catalytic core and propels the unattached kinesin head forward to find the next binding site on the microtubule. This model is based on a disorder-to-order transition in the neck linker region for monomeric kinesin constructs. The neck linker of the Mt·K¹ complex was shown to be mobile in the presence of ADP, existing in an equilibrium with two predominant conformations trapped by cryo-electron microscopy. However, upon the addition of ATP or nonhydrolyzable ATP analogs to the Mt·K complex, the neck mobility ceased with the neck linker element tightly associated with the catalytic core. This ordered state was reversed by the addition of ADP or loss of nucleotide. In addition, the cryo-electron microscopy of this proposed ATP state revealed a single discrete orientation of the neck linker with the carboxyl terminus of the motor domain directed toward the plus-end of the microtubule (4).

Xing et al. (5) have reported for a monomeric kinesin motor domain two discrete structural transitions induced by ADP binding and another produced by ATP binding. These three conformations revealed by fluorescence resonance energy transfer were consistent with the results reported by Rice et al. (4). Furthermore, biochemical studies of dimeric kinesin have demonstrated that ATP binding (or nonhydrolyzable analogs of ATP) to one of the two kinesin heads will trigger ADP release from the other (6-8). These pre-steady-state kinetics were the basis of the alternating site ATP hydrolysis model for kinesin motility. Another important contribution to our understanding of kinesin stepping was advanced by a molecular force clamp study that revealed a load-dependent isomerization that followed ATP binding (9). These results eliminated models in which ATP hydrolysis triggered the major conformational change for the 8-nm step and most loose coupling models, which predict that the ATP coupling ratio will decline with load. Therefore, the results from a variety of experimental approaches are converging into a model for kinesin plus-end directed motility and processivity. However, these studies have provided information predominantly for ATP-induced structural transitions. The results presented here focus on the role of ATP hydrolysis for motor domain coordination and tight coupling of ATP turnover with kinesin stepping.

We present the kinetics of a dimeric kinesin motor construct in which the target amino acid, switch I Arg²¹⁰, has been mutated to an alanine. The mutant kinesin motor, R210A, can be expressed and purified; therefore, we can evaluate the importance of Arg²¹⁰ for ATP-dependent interactions that are required for ATP turnover and coordination of the motor domains. The results presented here show that the steady-state ATPase kinetics are dramatically reduced, yet ATP binding is comparable with wild type. R210A is defective for ATP hydrolysis, and the dissociation kinetics suggest that this mutant cannot detach from the microtubule, a step essential for microtubule-based movement. We propose a model in which ATP hydrolysis at the rearward head is required for the leading head to bind tightly to the microtubule, and this tight binding state of the forward head is required for rearward head dissociation. This strategy ensures forward motion of kinesin stepping and tight coupling of ATP turnover to movement.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials-- Radiolabeled ATP ([-³²P]ATP, >3000 Ci/mmol) was purchased from PerkinElmer Life Sciences, Paclitaxel (taxol, Taxus brevifolia) from Sigma, polyethyleneimine-cellulose F TLC plates (20 × 20 cm, plastic-backed; EM Science of Merck) from VWR Scientific (West Chester, PA). ATP, GTP, DEAE-Sephacel, and S-Sepharose from Amersham Biosciences. MantATP and mantADP were synthesized and characterized as described previously (8, 10, 11).

Kinetic Assay Conditions-- The steady and pre-steady-state kinetic experiments were performed in ATPase buffer (20 mM Hepes, pH 7.2, with KOH, 5 mM magnesium acetate, 0.1 mM EGTA, 0.1 mM EDTA, 50 mM potassium acetate, 1 mM dithiothreitol) at 22-25 °C. All concentrations reported are final concentrations after mixing.

Expression and Purification of Kinesin Mutant R210A for Kinetic Analysis-- The R210A kinesin mutant plasmid was constructed by introducing a single amino acid change in the K401-wt plasmid (12) using the Chameleon Mutagenesis protocol (Stratagene). The arginine to alanine substitution at residue 210 was verified by DNA sequencing. The K401-wt motor contains the first 401 amino acids of the kinesin protein and when expressed is dimeric (13). The R210A plasmid was transformed into BL21(DE3)pLysS for expression in Escherichia coli and purification as described previously (14).

The R210A protein concentration was determined by the Bradford method using Bio-Rad Protein Assay with IgG as a protein standard. It was also measured spectrophotometrically at A₂₈₀ (12) based on the calculated extinction coefficient of 29,240 M¹ cm¹ (26,740 protein + 2,500 ADP) and M_r = 44,994 for R210A.

Bovine Brain Microtubule Preparation-- Microtubules were polymerized from tubulin and stabilized with 20 µM taxol as previously described (12). Sedimentation assays followed by SDS-PAGE confirmed that the microtubules were stable as the microtubule polymer. The concentrations of tubulin reported reflect the tubulin assembled into microtubules and stabilized with 20 µM taxol.

Active Site Measurement-- The active site experiments were based on the binding of [-³²P]ATP (15). R210A (K·ADP) at 5 µM was reacted with trace amounts of [-³²P]ATP, and the reaction was quenched with 5 M formic acid at various times ranging from 5 s to 100 min. The products [-³²P]ADP + P_i are separated from [-³²P]ATP by TLC and quantified. Because ADP product release is so slow, each active site under the conditions of the assay retains [-³²P]ADP. The data were fit to a single exponential function,

(Eq. 1)

where A is the amplitude and t is time. The rate constant, k_off, represents the rate of ADP release from the active site in the absence of microtubules, and the constant term C provides the active site concentration.

Steady-state ATPase Assays-- Steady-state ATPase measurements were determined by following the hydrolysis of [-³²P]ATP to form [-³²P]ADP·P_i as previously described (12).

Microtubule Equilibrium Binding Experiments-- These experiments were conducted as described previously (16). R210A at 2 µM was incubated with 0-20 µM microtubules in the absence of added nucleotides for 30 min, followed by centrifugation. The microtubule pellet was resuspended in ATPase buffer to equal the volume of the supernatant. Gel samples of the supernatant and resuspended pellet were prepared in 5× Laemmli sample buffer and resolved by SDS-PAGE (8% acrylamide, 2 M urea). The gel was stained with Coomassie Blue, analyzed by a Microtek Scan Maker X6EL scanner (Microtek, Redondo Beach, CA), and quantified using NIH Image version 1.62 to determine the fraction of R210A in the supernatant and pellet at each microtubule concentration. In Fig. 3 fractional binding, defined as the ratio of R210A in the pellet to total R210A, is plotted as a function of microtubule concentration. The data were fit to quadratic Equation 2,

(Eq. 2)

where [Mt·K]/[K] is the fraction of R210A sedimenting with microtubules, [K] is total R210A, [Mt₀] is the total tubulin concentration, and K_d is the dissociation constant.

Acid Quench ATPase Assay-- These experiments were performed to determine the pre-steady-state kinetics of ATP hydrolysis for the switch I mutant in comparison with K401-wt (17). The preformed Mt·R210A complex (syringe concentrations: 16 µM R210A, 30 µM microtubules, 40 µM taxol) was rapidly mixed in a chemical quench-flow instrument (Kintek Corp., Austin, TX) with [-³²P]ATP. The reaction was terminated with 5 M formic acid (syringe concentration) and expelled from the instrument. Radiolabeled ADP + P_i were separated from radiolabeled ATP by TLC, and the data were quantified. The concentration of [-³²P]ADP was determined for each reaction and plotted as a function of time (KaleidaGraph; Synergy Software, Reading, PA). The data were then fit to the burst equation,

(Eq. 3)

where A is the amplitude of the pre-steady-state burst phase which represents the formation of [-³²P]ADP·P_i at the active site during the first turnover; k_b is the rate constant of the exponential burst phase; t is time in seconds; and k_ss is the rate constant of the linear phase (µM ADP·s¹). The rate constant k_ss, when divided by enzyme concentration, corresponds to the rate of steady-state turnover at the same ATP and microtubule concentrations. Concentrations reported in the figure legends are final concentrations after mixing.

Stopped-flow Kinetics-- The pre-steady-state kinetics of mantATP binding, mantADP release, R210A binding to microtubules, and detachment of R210A were all conducted using the SF-2001 Kintek stopped-flow instrument in ATPase buffer at 25 °C. For the mantATP and mantADP experiments, excitation was set at 360 nm (Hg arc lamp) with emitted light measured through a 400-nm cut-off filter (mant _em = 450 nm). The mantATP binding data in Fig. 5A (inset) were fit to the following equation,

(Eq. 4)

where k_obs is the rate of first exponential increase in fluorescence, k₁ is the second-order rate constant for mantATP binding, and k_off obtained from the y intercept is the rate of mantATP dissociation from the Mt·R210A·ATP complex.

The microtubule association kinetics (Fig. 4) and the R210A dissociation kinetics (Fig. 7) were monitored by the change in turbidity at 340 nm. The exponential rate constants for microtubule association were plotted as a function of microtubule concentration and fit to Equation 5,

(Eq. 5)

where k_obs is the rate of exponential process, k₅ is the second-order rate constant for microtubule association, and k₅ obtained from the y intercept is the rate constant for motor dissociation from the Mt·R210A complex.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Active Site Measurement-- We began the analysis of R210A by evaluating the mutant motor in the absence of microtubules to determine whether the mutant retained the fundamental enzymatic features of a kinesin: the ability to bind and hydrolyze ATP and to retain ADP tightly bound at the active site (Fig. 1). R210A was incubated with a trace amount of [-³²P]ATP. During the incubation, ADP tightly bound at the active site should be released, followed by the binding and hydrolysis of [-³²P]ATP to yield a stable R210A·[-³²P]ADP intermediate. The results presented in Fig. 1 show that R210A exhibits the ability to bind and hydrolyze ATP. The rate constant of [-³²P]ADP release from the active site was 0.05 s¹, and this rate is somewhat faster than data reported previously for conventional kinesin at 0.006-0.01 s¹ (12, 18-20). This assay permitted the determination of R210A active site concentration at 4.6 µM. Therefore, these results document the ability of mutant R210A to bind and hydrolyze ATP followed by slow ADP release. Thus, in the absence of microtubules, R210A exhibits the characteristics expected of wild-type kinesin.

View larger version (11K): [in this window] [in a new window]   Fig. 1.   R210A active site determination. R210A at 5 µM (estimated by the Bio-Rad protein assay) with bound ADP was rapidly mixed with trace amounts of [-32P]ATP, and the reaction was quenched at various times. The data were fit to Equation 1, which provided the rate constant for ADP release from the active site in the absence of microtubules at 0.05 ± 0.003 s1. The concentration of R210A sites that were catalytic was 4.6 ± 0.09 µM.

Steady-state ATPase Assays-- The steady-state kinetics of R210A were determined in comparison with the kinetics of K401-wt (Fig. 2). The steady-state ATPase kinetics for the Mt·R210A complex were significantly altered in comparison with K401-wt as follows: R210A (seven preparations), k_cat = 0.12 ± 0.05 s¹ (0.07-0.15 s¹), K_m(ATP) = 118 ± 62.7 µM (75-211 µM) versus K401-wt, k_cat = 20-25 s¹, K_m(ATP) = 60-96 µM.

View larger version (21K): [in this window] [in a new window]   Fig. 2.   Steady-state ATPase kinetics of R210A. A, the Mt·R210A complex (1 µM R210A, 30 µM tubulin) was preformed and incubated with MgATP (0-2 mM). The rate of [-32P]ATP hydrolysis increased as a function of ATP concentration. The data were fit to a hyperbola, which provided the following steady-state kinetic parameters: kcat = 0.08 ± 0.005 s1, Km(ATP) = 163.3 ± 36.9 µM. B, comparison of the steady-state kinetics of R210A and K401-wt. Parameters for K401-wt are as follows: kcat = 24.5 ± 0.5 s1, Km(ATP) = 92.6 ± 6.8 µM.

There are several hypotheses that can account for the extremely low k_cat of R210A. The first is that there is a defect in ATP turnover. The second hypothesis is that there is a problem with microtubule binding that will affect release of ADP from the active site of the mutant. The third hypothesis is that the protein was inactive and the small amount of ATP hydrolysis seen was due to a few active motors still functioning. However, the third hypothesis appears unlikely based on the results of the active site assay (Fig. 1), which confirmed that R210A was active and exhibited the characteristics of wild type kinesin in the absence of microtubules. The experiments presented below evaluate ATP binding and ATP hydrolysis, microtubule association and detachment, and microtubule-activated product release.

Equilibrium Binding of R210A to the Microtubule-- One possible explanation for the depressed ATPase activity may be that microtubule binding and therefore Mt·R210A complex formation is aberrant. We evaluated formation of Mt·R210A complex by equilibrium binding (Fig. 3) and the pre-steady-state kinetics of Mt·R210A association (Fig. 4).

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Equilibrium binding of R210A to microtubules. R210A at 2 µM was incubated with varying concentrations of microtubules (0-20 µM tubulin) for 30 min in the absence of added nucleotides. The fraction of R210A bound to the microtubules was plotted as a function of total microtubule concentration. These data were fit to Equation 2, which yielded the apparent Kd(Mt) = 0.95 ± 0.028 µM with maximal fractional binding at 0.92 ± 0.15.

View larger version (27K): [in this window] [in a new window]   Fig. 4.   Pre-steady-state kinetics of microtubule-R210A·ADP association. R210A at 2 µM was rapidly mixed with varying concentrations of taxol-stabilized microtubules (1-10 µM) in the stopped-flow instrument. A, a representative stopped-flow transient, where 2 µM R210A was rapidly mixed with 3 µM microtubules. The data were fit to a single exponential function with a linear term where the exponential rate of microtubule association was 9.0 ± 1.1 s1. B, the exponential rate constants of the microtubule-dependent turbidity change increased as a function of microtubule concentration. The fit of the data to Equation 6 defined the apparent second-order rate constant for microtubule association, k+5 = 0.83 ± 0.04 µM1 s1, with the y intercept, k5 = 5.83 ± 0.26 s1.

The relative affinity of R210A for microtubules was determined by equilibrium binding in which R210A was incubated with increasing concentrations of microtubules, followed by centrifugation and analysis by SDS-PAGE. Fig. 3 shows that R210A partitioned with microtubules as a function of tubulin concentration, and the fit of the data provided an apparent K_d(Mt) = 0.95 µM tubulin with maximal fractional binding at 92%. The fact that the fractional binding is almost 100% suggests that the mutant motor can bind microtubules. However, the K_d at 0.95 µM for R210A is weaker than the K_d determined for the Mt·K401-wt complex at 37 nM (21).

Association Kinetics of R210A Binding to Microtubules-- R210A was rapidly mixed with microtubules in the stopped-flow instrument, and the turbidity signal was monitored to quantify Mt·R210A complex formation. The results presented in Fig. 4 show that the rate of microtubule association increased linearly as a function of microtubule concentration with the second-order rate constant, k₊₅ = 0.8 µM¹ s¹ and k₅ = 5.8 s¹ (Scheme 1, Table I). The kinetics for K401-wt have been reported at 10-20 µM¹ s¹ with no evidence of an off rate (Table I) (22-24). Therefore, the association kinetics clearly indicate that formation of the Mt·R210A complex is defective. Both the association kinetics and the equilibrium binding results show that the affinity of R210A for microtubules is weaker than observed for K401-wt.

View larger version (6K): [in this window] [in a new window]   Scheme 1.

ATP Binding and Hydrolysis Kinetics-- The kinetics of ATP binding were evaluated by rapidly mixing in the stopped-flow instrument the Mt·motor complex (15 µM tubulin plus 2 µM R210A or K401-wt) with increasing concentrations of the fluorescent ATP analog, mantATP (Fig. 5). The kinetics reveal a biphasic fluorescence enhancement. Because there is an increase in fluorescence as mantATP enters the more hydrophobic environment of the active site, we assume the initial rapid phase of fluorescence enhancement is mantATP binding to the active site. At low mantATP concentrations (<50 µM), the observed rate of the first exponential phase increased linearly as a function of mantATP concentration and provided the second-order rate constant, k₊₁ = 0.82 µM¹ s¹ with a dissociation rate of 26.9 s¹ (Scheme 1, Table I). For the Mt·K401-wt complex, mantATP binding was reported at 1.1 µM¹ s¹ with a dissociation rate of 9.8 s¹ (23).

View larger version (29K): [in this window] [in a new window]   Fig. 5.   Pre-steady-state kinetics of mantATP binding to Mt·R210A complex. The preformed Mt·R210A complex was rapidly mixed in the stopped-flow instrument with increasing concentrations of mantATP (5-1000 µM). A, a representative transient is shown with the final concentration of mantATP at 500 µM. The fit of the data to a double exponential function provided the initial rapid rate of fluorescence enhancement at 80.8 ± 2.3 s1 (relative amplitude 0.17 ± 0.003) and the slower second phase at 12.4 ± 0.34 s1 (relative amplitude 0.23 ± 0.003). Inset, the exponential rate constants of the first phase increased linearly as a function of mantATP concentration from 0-50 µM mantATP. The data were fit to Equation 4, providing the second-order rate constant for mantATP binding, k1 = 0.71 ± 0.08 µM1 s1, with koff = 26.9 ± 1.6 s1. B, the mantATP concentration dependence of the initial fast phase () and the second slow phase () were plotted and fit to hyperbolae that provided the maximum rates for the initial fast phase at 80.7 ± 4.2 s1 and the slower second phase at 12.1 ± 0.6 s1. The Mt·R210A complexes were preformed at 2 µM R210A + 15 µM tubulin for mantATP concentrations 5-100 µM, and 15 µM R210A + 30 µM tubulin for mantATP concentrations 50-1000 µM.

ATP binding for wild type kinesin is believed to involve two steps (Scheme 2) based on our pulse-chase rapid quench kinetics with K401-wt (17) and the mantATP binding kinetics reported by Ma and Taylor for human kinesin K379 (25). In the first step, the collision complex is formed (Mt·K·ATP), followed by a rate-limiting conformational change at 200 s¹ to form the Mt·K*·ATP intermediate that proceeds toward ATP hydrolysis. The kinetics for R210A indicate that the required conformational change does occur; however, the rate constant observed is 81 s¹. These data suggest that the ATP-driven structural transition required for ATP hydrolysis is slowed significantly in the mutant.

Although the mantATP binding results indicate that the mutant was able to bind ATP effectively, the chemistry step of ATP hydrolysis was clearly aberrant. For the ATP hydrolysis kinetics (Fig. 6), a preformed Mt·R210A complex was rapidly mixed with [-³²P]ATP in the rapid quench instrument, followed by an acid quench to terminate the reaction and release nucleotide at the active site. The kinetics for K401-wt showed the expected, dramatic exponential burst of ADP·P_i product formation at the active site during the first turnover because ATP binding and hydrolysis are fast steps for kinesin relative to the rate-limiting step in the pathway (17, 25). Note that R210A showed no evidence of an exponential burst, and the rate constant for ATP hydrolysis (k₂; Scheme 1) was extremely low and similar to steady-state turnover. These results clearly indicate that ATP hydrolysis is defective in R210A.

View larger version (18K): [in this window] [in a new window]   Fig. 6.   Acid quench kinetic comparison of R210A with K401-wt. [-32P]ATP (100 or 200 µM) was rapidly mixed with a preformed Mt·R210A complex (8 µM R210A, 15 µM tubulin) in the rapid quench instrument for 0.005-0.2 s, followed by the acid quench. This experiment was repeated for K401-wt. The data for K401-wt were fit to the burst equation (Equation 3). At 100 µM MgATP, the burst amplitude of K401-wt was 2.0 ± 0.34 µM, kburst = 81.8 ± 34.0 s1, and kss = 55.4 ± 3.07 µM·s1/8 µM sites (6.9 s1). At 200 µM MgATP, K401-wt had a burst amplitude of 2.69 ± 0.52 µM, kburst = 389 s1 and kss = 83.0 ± 6.5 µM·s1/8 µM sites (10.4 s1). No pre-steady-state burst of product formation was observed for R210A at either 100 or 200 µM ATP. The linear fit of the data provided kobs = 0.2 s1 at 200 µM MgATP.

The steady-state ATPase kinetics for R210A in combination with the rapid quench kinetics suggests that the intermediate that accumulates is the prehydrolysis M·K*·ATP intermediate (Scheme 2). However, the maximal rate constant for mantATP binding (Scheme 2, Mt·K·ATP  Mt·K*·ATP isomerization) and the high steady-state K_m(ATP) exhibited by R210A suggest that the mutation is affecting the ability of the motor to generate the structural transition(s) required to reach the Mt·K*·ATP state for catalysis from the Mt·K·ATP collision complex.

View larger version (5K): [in this window] [in a new window]   Scheme 2.

The stopped-flow transients for mantATP binding best fit a double exponential function (Fig. 5). The second phase of mantATP binding was slower and represents a first order isomerization at 12 s¹. Although biphasic fluorescence transients have been observed for wild type kinesin, the second phase has always shown a decrease in fluorescence intensity rather than an increase as observed for R210A (5, 23, 25). We cannot assign the 12 s¹ isomerization to a specific step in the R210A ATPase pathway at this time.

ATP-promoted Dissociation Kinetics of Mt·R210A-- The dissociation kinetics were measured by rapidly mixing a Mt·K complex with MgATP and following the decrease in turbidity associated with motor detachment from the microtubule (Fig. 7). For the experiments in Fig. 7A, 100 mM KCl was included in the ATP syringe. The added salt was required to measure the kinetics of dissociation because of kinesin's processivity, and the additional salt does not alter the kinetics of the first ATP turnover (see Fig. 2; Ref. 22). The observed rate (k₃) of ATP-promoted dissociation of K401-wt was 16 s¹ and consistent with results published previously (8, 25). In contrast, the mutant R210A showed no significant change in turbidity, although there appeared to be a slow decrease in turbidity comparable with steady-state turnover. These results demonstrate that R210A cannot detach from the microtubule at ATP and salt conditions that lead to K401-wt dissociation. This inability to detach from the microtubule suggests that ATP hydrolysis is required to reach a state that normally occurs to weaken the affinity of kinesin for the microtubule.

View larger version (28K): [in this window] [in a new window]   Fig. 7.   ATP-promoted dissociation kinetics of Mt·R210A in comparison with Mt·K401-wt. A, the Mt·R210A complex (6 µM R210A, 6 µM tubulin) or the Mt·K401 complex (4 µM K401, 3.75 µM tubulin) was rapidly mixed with 1 mM MgATP plus 100 mM KCl. The Mt·K401 wild type data were fit to a double exponential function that provided the observed rate of dissociation at 16.3 ± 0.7 s1. The R210A transient did not show a significant change in turbidity. B, the Mt·R210A complex (6 µM R210A, 6 µM tubulin) or the Mt·K401 complex (4 µM K401, 3.75 µM tubulin) was rapidly mixed with 1 mM MgATP but in the absence of the additional 100 mM KCl. The dissociation kinetics of K401-wt at 1.14 s1 indicate that the wild type motor was in association with the microtubule for 0.88 s (transit time = 1/kobs). In contrast, R210A showed only a small turbidity change during the 30-s period of observation.

The Mt·R210A dissociation kinetics (Fig. 7A) indicating tight binding to the microtubule appear to be inconsistent with the microtubule association kinetics and equilibrium binding results, which suggest that R210A affinity for microtubules is relatively weak (Figs. 3 and 4). We reasoned that the difference in the results was due to the presence of ATP, which, during ATP turnover, induced a stable Mt·R210A species that does not readily detach from the microtubule. To explore this hypothesis, we repeated the ATP-promoted dissociation experiment at the low salt conditions and extended the period of observation to 30 s (Fig. 7B). Note that K401-wt does exhibit dissociation kinetics, yet the turbidity signal of R210A shows a slow decrease in turbidity. The steady-state k_cat for R210A at 0.11 s¹ indicates that during the 30-s period of observation, R210A would turn over ~3 ATP molecules (1/0.1 s¹ provides transit time for 1 turnover = 10 s; 30 s/10 s = ~3 ATP). In contrast, the dissociation kinetics for K401-wt (Fig. 7B) indicate that 34 molecules of ATP will by hydrolyzed. The prediction is that if ATP hydrolysis were required for R210A dissociation from the microtubule, then the R210A dissociation kinetics should occur at a rate comparable with steady-state turnover, and the amplitude associated with the dissociation kinetics should be ~8-10% (3 ATP/34 ATP) of the wild type signal. The kinetics presented in Fig. 7 are consistent with this interpretation.

MantADP Release from Both Heads of the R210A·MantADP Complex-- R210A was incubated with mantADP at a 1:2 ratio in order to exchange the ADP at the active sites of the protein with mantADP. This newly formed R210A·mantADP complex was then rapidly mixed in the stopped-flow instrument with varying concentrations of microtubules plus MgATP (Fig. 8). For these experiments, the stopped-flow instrument was used to monitor the decrease in fluorescence as mantADP was released from the active site to the buffer and the fluorescence was quenched. The MgATP included with the microtubules served to block rebinding of mantADP to the active sites of the motor. Fig. 8 shows that the exponential rate of mantADP release increased as a function of microtubule concentration with the maximum rate constant, k₆ = 57 s¹ with K_1/2(Mt) = 16 µM. For wild type kinesin motors, mantADP release has been measured at >100 s¹ (4, 7, 8, 22). The K_1/2(Mt) at 16 µM is comparable with the constant determined for K401-wt at 15 µM. These results indicate that although the microtubule association kinetics may be aberrant (Fig. 4), the Mt·R210A collision complex formed does activate mantADP release. Our studies were extended to separate the kinetics of mantADP release from each motor domain of the dimer, and the kinetics of mantADP release from the high affinity site are presented next.

View larger version (26K): [in this window] [in a new window]   Fig. 8.   Pre-steady-state mantADP release from both heads of the Mt·R210A·mantADP complex. A preformed R210A·mantADP complex (2.5 µM R210A, 5 µM mantADP) was rapidly mixed in the stopped-flow instrument with varying concentrations of taxol-stabilized microtubules (2.5-40 µM tubulin plus 1 mM MgATP). A, a representative stopped-flow transient of the change in fluorescence due to the release of mantADP from the Mt·R210A·mantADP complex when a solution of 2.5 µM R210A and 5 µM mantADP was rapidly mixed with a solution of 10 µM microtubules and 1 mM MgATP. The data were fit to a single exponential function where the exponential rate was kobs = 22.8 ± 0.7 s1. B, the exponential rate constants of the microtubule-dependent fluorescence change were plotted as a function of microtubule concentration, and the data were fit to a hyperbola. The maximum rate constant of mantADP release from the Mt·R210A·mantADP complex was 57.2 ± 2.9 s1.

MantADP Release from the Second Head of the R210A·MantADP Complex-- An equilibrium Mt·R210A·mantADP complex (2 µM R210A, 1 µM mantADP, 15 µM tubulin) was preformed and then rapidly mixed with MgATP in the stopped- flow (Fig. 9). The experimental design assumes that when the complex is preformed with half the concentration of mantADP as active sites of R210A, the mantADP will partition to the head that holds ADP more tightly (7). This head is assumed to be weakly bound to the microtubule (7, 8). Upon the addition of MgATP, ATP binds to the empty site, leading to mantADP release from the high affinity site (4, 7, 8). The observed kinetics of mantADP release presented in Fig. 9 indicate that mantADP was released from the high affinity site at a maximum rate of 34 s¹, which is significantly less than reported previously for wild type kinesin at >100 s¹ (4, 7, 8). Interestingly, this 34 s¹ rate constant determined for R210A was quite similar to the rate constants (30-40 s¹) reported for wild type kinesin constructs when mantADP release was initiated by ATP analogs AMP-PNP and ATPS (7, 8). We pursued experiments to evaluate whether the slow mantADP release kinetics observed in the absence of ATP hydrolysis for K401-wt and R210A may be revealing the same structural transition.

View larger version (31K): [in this window] [in a new window]   Fig. 9.   MantADP Release from the high affinity site of the Mt·R210A·mantADP complex initiated by MgATP. The Mt·R210A·mantADP complex (2 µM R210A, 1 µM mantADP, 15 µM tubulin) was rapidly mixed with varying concentrations of MgATP (1-1000 µM). A, a representative stopped-flow transient at 20 µM MgATP. The data were fit to a single exponential function plus a linear term where the exponential rate was kobs = 33.0 ± 2.9 s1. B, the exponential rate constants of the MgATP-dependent fluorescence change were plotted as a function of MgATP concentration. The fit of the data to a hyperbola yielded a maximum rate of 33.8 ± 0.8 s1 for mantADP release from the high affinity site. The inset shows the data from 0-50 µM MgATP.

The Mt·R210A·mantADP complex was preformed (2 µM R210A, 1 µM mantADP, 15 µM tubulin) and rapidly mixed with MgAMP-PNP in the stopped-flow instrument (Fig. 10). Note that the maximum rate of mantADP release initiated by AMP-PNP was 40 s¹, consistent with the interpretation that the R210A ATP hydrolysis mutant releases mantADP from the high affinity site with kinetics comparable with mantADP release from wild type kinesin when ATP hydrolysis is prevented (AMP-PNP) or significantly slowed (ATPS). These kinetics suggest that AMP-PNP induces the same structural transition species as ATP for R210A, and this intermediate for wild type kinesin is trapped by AMP-PNP binding. In addition, the results indicate that ATP hydrolysis at head 1 signals head 2 (Fig. 11) and affects the structural transitions that occur for head 2 microtubule binding and subsequent rapid mantADP release (discussed below). When the experiment was repeated, but using 1 mM MgADP to initiate mantADP release, the rate constant obtained for mantADP release from head 2 was significantly faster at 25 s¹ than observed with K401-wt at 6 s¹ (Table I) (7, 8). These results indicate that the active site of R210A has lost the structural precision required to discriminate between nucleotide intermediates, and/or the response of the active site to coordinate the motor domains is disrupted by the amino acid substitution at Arg²¹⁰.

View larger version (25K): [in this window] [in a new window]   Fig. 10.   MantADP release from the high affinity site of the Mt·R210A·mantADP complex initiated by AMP-PNP. The Mt·R210·mantADP complex (2 µM R210A, 1 µM mantADP, 15 µM tubulin) was rapidly mixed with varying concentrations of MgAMP-PNP (0.25-1000 µM) in the stopped flow. A, a representative transient when the Mt·R210A·mantADP complex was rapidly mixed with 1 mM AMP-PNP. The data were fit to a single exponential plus a linear term that yielded the initial exponential rate, kobs = 30.0 ± 3.9 s1. B, the exponential rate constants were plotted as a function of AMP-PNP concentration, and the fit of the data to a hyperbola yielded the maximum rate constant of mantADP release at 40.6 ± 3.5 s1. The inset shows the data from 0-50 µM MgAMP-PNP.

View larger version (17K): [in this window] [in a new window]   Fig. 11.   Mechanistic model for the role of ATP hydrolysis. This model is framed in the context of recent proposals by Rice et al. (4) and Schnitzer et al. (9), although the kinetics for K401-wt and R210A do not necessarily exclude inchworm models in which head 1 is always forward with head 2 rearward. The cycle begins as head 1 binds the microtubule with rapid ADP dissociation. ATP binding at head 1 leads to the plus-end-directed motion of the neck linker to position head 2 forward at the next microtubule binding site. ATP binding at head 1 is sufficient to promote head 2 association with the microtubule followed by rapid ADP release. However, ATP hydrolysis at head 1 is required to lock head 2 onto the microtubule in a tight binding state. The strain induced by the tight binding of head 2 weakens the affinity of head 1 and results in its detachment from the microtubule concomitant with Pi release. The active site of head 2 is now accessible for ATP binding, and the cycle is repeated.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

We have used a switch I kinesin mutant that is defective for ATP hydrolysis to examine the role of ATP hydrolysis in coordinating the motor domains of the dimer for kinesin processivity. The experimentally determined constants (Scheme 1) for R210A and K401-wt are reported in Table I, and the model for kinesin motility is presented in Fig. 11.

Our initial set of observations from the active site experiment shows that the mutant R210A can bind ATP, hydrolyze it to ADP·P_i, and release the products ADP + P_i. These results demonstrate that the R210A protein is active and that ADP is kept tightly bound at the active site in the absence of microtubules. Thus, in the absence of microtubules, R210A exhibits the minimal characteristics expected of a kinesin motor protein.

However, the results also show that R210A-microtubule interactions are altered. The steady-state kinetic analysis demonstrated that the Mt·R210A complex had severe difficulty in ATP turnover (k_cat = 0.11 s¹ in comparison with 20-25 s¹ for K401-wt). We used pre-steady-state kinetic approaches to probe specific steps of the ATPase pathway to determine which were disrupted due to the mutation and to analyze disrupted motor domain cooperativity.

Kinetics of ATP Binding and ATP Hydrolysis-- The stopped-flow experiments revealed that the Mt·R210A complex could bind mantATP (Fig. 5, Table I). However, the acid quench transients showed no exponential burst of ADP·P_i product formation during the first ATP turnover (Fig. 6). The absence of the exponential burst was indicative of impaired ATP hydrolysis. Furthermore, the rate constant of ATP hydrolysis was similar to steady-state turnover, suggesting that the rate-limiting step in the R210A pathway has become ATP hydrolysis. These initial kinetic experiments clearly demonstrated that the switch I Arg²¹⁰ is a critical amino acid necessary for ATP hydrolysis in kinesin.

Comparison with Myosin Mutants-- The loss of the ability to hydrolyze ATP due to a mutation in the switch I loop is not unique to kinesin. Shimada et al. (26) reported their scanning alanine mutagenesis study of the conserved switch I region (NXNSSRFG) using Dictyostelium discoideum myosin II. One particular mutant, R238A,² is the analogous mutant in myosin II to R210A in kinesin. This myosin mutant exhibited very low steady-state ATPase activity, no evidence of an exponential burst of ATP hydrolysis, and an inability to support actin filament sliding in vitro. In addition, development of the Dictyostelium mutant stopped at the mound state and did not proceed through morphogenesis. These studies have been extended by several groups to understand the structural and mechanistic role of the switch I arginine for ATP hydrolysis and mechanochemistry (27-34).

Structural Role of Residue Arg²¹⁰-- A structural explanation can be given as to why this particular residue may play such a key role in ATP hydrolysis. Switch I Arg²¹⁰ represents one-half of a salt bridge with residue Glu²⁴³ that is located within the active site of the motor (Fig. 12). This salt bridge is thought to attract a water molecule to attack the -phosphate on the nucleotide. By subsequently disrupting this ionic interaction at Arg²¹⁰ through the mutation of the arginine to an alanine, the water molecule would be unable to coordinate properly in the active site. Evidence in support of this hypothesis has been published for kinesin, myosin, and G-proteins (34-40). First, there are four structural elements (N1-N4) that form the kinesin nucleotide-binding pocket, and these are highly conserved for myosins,³ kinesins,⁴ and G-proteins both in amino acid sequence and structure. Second, a conserved structural element revealed in the kinesin and myosin crystal structures is the salt bridge between the switch I arginine (NXXSSRSH) and the switch II glutamate (DLAGXE) (34, 40-44). Third, mutation of the switch II glutamate results in motors defective in ATP hydrolysis for both myosin II and kinesin (4, 29-31, 45). Fourth, mutagenesis of the myosin II switch I arginine to glutamate and the switch II glutamate to arginine results in a myosin double mutant with an inverted salt bridge that can support efficient ATP hydrolysis and normal myosin function (31). This Dictylostelium myosin II mutant also rescued myosin null cells. The transformants were able to undergo cytokinesis and proceed through morphogenesis to form fruiting bodies and viable spores (31). Last, the crystal structure and biochemical characterization of the switch I salt bridge mutant, Kar3 R598A, shows that the mutation destabilized the conformation surrounding switch I (34). The structural changes were also correlated with the functional behavior of Kar3 R598A. The steady-state ATPase activity of the Mt·Kar3 R598A complex was depressed to basal ATPase levels, and motor domain affinity for microtubules was weakened.

View larger version (70K): [in this window] [in a new window]   Fig. 12.   Rat kinesin monomer model 2KIN (44). This detailed view of the nucleotide binding site shows ADP (yellow) at the active site with the switch I arginine (white) and the switch II glutamate (green) highlighted. The proposed interswitch salt bridge (3 Å) between Drosophila Arg210 and Glu243 (rat Arg204-Glu237) is indicated by the dashed lines.

Therefore, our results for the kinesin R210A motor are consistent with the growing evidence that the switch I arginine-switch II glutamate salt bridge is required for ATP hydrolysis and essential for mechanochemistry and motor function.

Structural Transitions and Nucleotide Binding State-- The R210A mutant has provided new information to order steps in the ATPase cycle (Fig. 11). The mantADP release kinetics (Figs. 9 and 10) from the high affinity site indicate that ADP release from head 2 occurs prior to ATP hydrolysis of head 1 (states 1-4; Fig. 11). First, R210A is defective for ATP hydrolysis. Second, R210A can accumulate mantADP on head 2, and its release occurs in the absence of ATP hydrolysis. Third, the K_1/2(ATP) and K_1/2(AMP-PNP) at ~0.4 µM reveal that a very low concentration of ATP or AMP-PNP is required to initiate mantADP release. These results imply that it is the Mt·K·ATP intermediate that leads to mantADP release from head 2 (intermediate 4; Fig. 11).

The sequence of steps to ADP release from head 2 as presented in Fig. 11 are consistent with a number of kinesin motility models (4, 5, 7-9). However, because the rate constant for ATP hydrolysis at ~100 s¹ is so similar to the rate constant for mantADP release from head 2 (>100 s¹), it has been argued that ATP hydrolysis on head 1 occurs prior to ADP release from head 2 (46, 47). The results presented here as well as similar experiments with another kinesin mutant that is defective for ATP hydrolysis (human E236A (4)) provide a compelling argument that mantADP release from head 2 occurs prior to ATP hydrolysis on head 1. Human kinesin switch II mutant E236A (corresponding to Drosophila E243) shows very low microtubule-activated ATPase activity and no phosphate burst kinetics for ATP hydrolysis but effective mantADP release from head 2 (4).

Microtubule-R210A Interactions-- The ATP-promoted dissociation kinetics (Fig. 7) show that R210A is incapable of microtubule dissociation at conditions in which K401-wt exhibits dissociation kinetics. However, in the absence of ATP (species 1; Fig. 11), R210A appears to be more weakly bound to the microtubule than K401-wt (Figs. 3 and 4; Table I). These results imply that upon binding ATP, the Mt·K·ATP species formed cannot proceed to a weakly bound state for dissociation. One hypothesis to account for these data is that ATP hydrolysis at head 1 is required for kinesin to proceed to a weakly bound state for dissociation. In fact, AMP-PNP leads to K401-wt kinetics similar to the R210A transients in Fig. 7, consistent with the interpretation that ATP hydrolysis is required for motor detachment from the microtubule. We propose that ATP binding for R210A and AMP-PNP binding for wild type kinesin lead to accumulation of intermediate 4 (Fig. 11). Head 2 of this intermediate may represent the highly mobile microtubule-bound state of the kinesin monomer trapped by Sosa et al. in the presence of ADP (48).

Model for Kinesin Motility-- The model we propose in Fig. 11 is framed in the context of recent proposals by Rice et al. (4) and Schnitzer et al. (9), although the kinetics for K401-wt and R210A do not necessarily exclude all inchworm models in which head 1 is always forward with head 2 rearward. The cycle begins as the first motor domain binds the microtubule with rapid ADP release. ATP binding at head 1 leads to the series of conformational changes to dock the neck linker of head 1 onto the motor core and to propel head 2 forward to the next binding site on the microtubule (species 4). Microtubule association activates ADP release from head 2, but ATP hydrolysis on head 1 is required for head 2 to bind tightly to the microtubule (species 5). We propose that head 2 must lock down onto the microtubule before head 1 can undergo dissociation. This mechanism optimizes processivity by ensuring that one motor domain is tightly bound to the microtubule before the second can detach. The strain generated within the dimer weakens the affinity of head 1, resulting in concomitant dissociation and phosphate release as proposed by Xing et al. (5).

Our model, based on the kinetics of R210A and wild type kinesin (6-8, 22), predicts that ATP cannot bind at head 2 (species 4 and 5) until head 1 dissociates from the microtubule. This hypothesis implies that the nucleotide binding pocket at head 2 is inaccessible to nucleotide because of structural transitions transmitted by head 1 to head 2 and controlled by the nucleotide state at head 1. This mechanism of alternating site ATP hydrolysis also minimizes rearward stepping and/or slippage and ensures tight coupling of one ATP turnover per 8-nm step. These predictions are supported by the mechanical data for kinesin as proposed by both S. M. Block and co-workers (2, 9, 49) and J. Gelles and co-workers (3).

In summary, our studies with R210A have shown that this switch I arginine is required for ATP hydrolysis directly. Furthermore, this analysis has provided an understanding of the specific step of ATP hydrolysis for one series of structural transitions that occur for processive movement along the microtubule. Last, the kinetics have revealed the importance of the post-ATP hydrolysis state for head-head communication that must occur during kinesin motility.

	ACKNOWLEDGEMENTS

We thank Dr. John Rosenberg for assistance in the molecular modeling and Dr. Steve Rosenfeld (University of Alabama at Birmingham) for thoughtful review of the manuscript.

	FOOTNOTES

^* This work was supported by National Institutes of Health Grant GM 54141 (to S. P. G.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

These authors contributed equally to this work.

^§ A University of Pittsburgh Honors College Scholar and the recipient of a Howard Hughes Medical Institute Fellowship, a Barry M. Goldwater Scholarship, and a Norman H. Horowitz Award for Undergraduate Research.

^¶ The recipient of an Andrew Mellon Predoctoral Fellowship.

To whom correspondence should be addressed: Dept. of Biological Sciences, 518 Langley Hall, University of Pittsburgh, Pittsburgh, PA 15260. Tel.: 412-624-5842; Fax: 412-624-4759; E-mail: spg1+@pitt.edu.

Published, JBC Papers in Press, February 25, 2002, DOI 10.1074/jbc.M108793200

² Switch I-switch II salt bridge residues are as follows: Drosophila kinesin, Arg²¹⁰-Glu²⁴³; rat kinesin, Arg²⁰⁴-Glu²³⁷; human kinesin, Arg²⁰³-Glu²³⁶; Saccharomyces cerevisiae Kar3, Arg⁵⁹⁸-Glu⁶³¹; Dictyostelium discoideum myosin II, Arg²³⁸-Glu⁴⁵⁹; chicken gizzard smooth muscle myosin II, Arg²⁴⁷-Glu²⁴³.

³ The Myosin Home Page (www.mrc-lmb.cam.ac.uk/myosin/myosin.htm).

⁴ The Kinesin Home Page (www.blocks.fhcrc.org/~kinesin/).

	ABBREVIATIONS

The abbreviations used are: Mt·K, microtubule-kinesin complex; Mt, microtubule; K401-wt, kinesin heavy chain construct containing the N-terminal 401 amino acids of the Drosophila kinesin heavy chain gene; mantADP, 2'(3')-O-(N-methylanthraniloyl)adenosine 5'-diphosphate; mantATP, 2'(3')-O-(N-methylanthraniloyl)adenosine 5'-triphosphate; AMP-PNP, 5'-adenylyl imidodiphosphate; ATPS, adenosine 5'-O-(thiotriphosphate).

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES