The tethered motor domain of a kinesin-microtubule complex catalyzes reversible synthesis of bound ATP

Abstract

Although the steps for the forward reaction of ATP hydrolysis by the motor protein kinesin have been studied extensively, the rates for the reverse reactions and thus the energy changes at each step are not as well defined. Oxygen isotopic exchange between water and P(i) was used to evaluate the reverse rates. The fraction of the kinesin x ADP x P(i) complex that reverts to ATP before release of P(i) during net hydrolysis was approximately 0 and approximately 2.6% in the absence and presence of microtubules (MTs), respectively. The rate of synthesis of bound ATP from free P(i) and the MT x kinesin x ADP complex was approximately 1.7 M(-1) x s(-1) (K0.5 ADP = 70 microM) with monomeric kinesin in the absence of net hydrolysis. Synthesis of bound ATP from the ADP of the tethered head of a dimer-MT complex was 20-fold faster than for the monomer-MT complex. This MT-activated ATP synthesis at the tethered head is in marked contrast to the lack of MT stimulation of ADP release from the same site. The more rapid ATP synthesis with dimers suggests that the tethered head binds behind the strongly attached head, because this positions the neck linker of the tethered head toward the plus end of the MT and would thus facilitate its docking on synthesis of ATP. The observed rate of ATP synthesis also puts limits on the overall energetics that suggest that a significant fraction of the free energy of ATP hydrolysis is available to drive the docking of the neck linker on binding of ATP.

Scheme 1.

Minimal scheme for hydrolysis of ATP by the complex of kinesin (E) with MTs.

Scheme 2.

Mechanism for intermediate exchange during net hydrolysis.

Scheme 3.

Mechanism for exchange of water-derived oxygens into P_i of the medium by reversible synthesis of bound ATP. Note that reversible formation of a pentacoordinate intermediate from ADP and P_i would not produce exchange with water, because no dehydration occurs until ATP is formed. Exchange could in principle occur by reversible dehydration of P_i to generate metaphosphate without formation of ATP. Metaphosphate, however, would be a highly unstable intermediate that would likely have an energy close to that for the transition state leading to ATP synthesis, and thus both processes would be expected to have similar rates. Also, exchange by a metaphosphate route would require equilibration of the released water with the bulk water pool during the extremely short lifetime of this unstable intermediate, rather than during the much longer lifetime of the ATP state.

Fig. 1.

Medium P_i = water oxygen exchange by kinesin. (a) Change in ratio of ¹⁸O₃-P_i:¹⁸O₄-P_i species versus time for monomeric BKinM. Full reaction with 2 μM BKinM, 4 μM MTs, 1 mM MgADP, and 1 mM ¹⁸O-P_i (diamonds). Control reactions had one component omitted. Minus MTs (square), minus ADP (circle), and minus BKinM (triangle). (b) Dependence of exchange rate on [ADP]. As in a, except for 0.5 mM P_i and variation in [ADP]. Reaction for 154 min. A K_d of 70 μM was determined by fitting to a hyperbolic binding model. (c) Dependence of exchange rate on [P_i]. As in a at 1 mM MgADP and variable [P_i]. Reaction for 168 min. (d) Progress of exchange reaction for dimeric DKH405. Reaction as 2 μM DKH405 (per head concentration), 4 μM MTs, and 1 mM ¹⁸O-P_i. No ADP was added, and the free [ADP] was 2 μM that resulted from half-site release of ADP on binding of DKH405 to the MT and from carryover of free ADP in the stock of DKH405. Control reactions had extra ADP added to a total free concentration of 3 μM (circle) or 5 μM (triangle). Results are reported as total amount of exchange, because the exchange is more extensive than in a with multiple cycles of exchange having occurred. Consequently, the ¹⁸O₃-P_i:¹⁸O₄-P_i ratio no longer scales linearly with the amount of exchange. For comparison, the point at 25 min has a ¹⁸O₃-P_i:¹⁸O₄-P_i ratio of 0.097 that is already off the scale of a.

Fig. 2.

Influence of added salts on binding of 2′(3′)-O-(N-methylanthraniloyl)ADP (mantADP) to nucleotide-free K349. Stopped-flow analysis of the rate of the fluorescent transient on binding of mantADP to nucleotide-free human K349 monomer. Final concentrations of 0.24 μM K349 and 2.9 μM mantADP in A25 buffer with 25 mM KCl plus additional potassium salts as indicated. No addition, diamond; P_i, circles; sulfate, squares; KCl to 50 mM total, triangle. Hyperbolic fits to data yield K_0.5 values of 3.1 and 1.8 mM and maximum fractional inhibitions of 61 and 81% for P_i and sulfate, respectively.

Scheme 4.

Comparison of ADP release and ATP resynthesis by monomers and the tethered head of a kinesin dimer bound to a MT. Motor domains are represented by circles with the MT indicated by a solid bar with the plus end at the right. ADP and ATP are indicated by D and T, respectively, with the three species on each line being the nucleotide-free, ADP, and ATP forms. The NL is indicated by a line originating near position of I325 (human kinesin numbering) with a curved line depicting an undocked linker and a straight arrow indicating the forward directed position of the docked linker in the presence of ATP (see ref. for a more complete description of this schematic representation). The increased freedom of the undocked NL with monomers is indicated by the multiple conformations of the NL in A. An intact neck coil is indicted by long parallel lines, and a partially uncoiled neck coil is indicated by shorter parallel lines and single lines leading to the ends of the NLs. The weight of the arrows in lines A and B is only intended to convey the relative changes in rate between monomers and dimers for each step. As discussed in the text, the equilibrium for ATP synthesis still strongly favors hydrolysis at 1 mM P_i even for dimers.

Scheme 5.

Model for ATP-induced movement of the tethered head from the rear to the lead position. The starting configuration is that of B in Scheme 4 that is consistent with the acceleration of medium P_i = water exchange with dimers. ATP binding to the lead head would favor NL docking, and this would displace the trailing head with ADP toward the lead position with stimulation of ADP release. * marks a fixed position on the MT.