Regular article
The time scale of the catalytic loop motion in triosephosphate isomerase1

https://doi.org/10.1006/jmbi.2001.4672Get rights and content

Abstract

Loop 6 in the active site of Triosephosphate Isomerase (Saccharomyces cerevisiae) moves in order to reposition key residues for catalysis. The timescale of the opening and closing of this loop has been measured, at temperatures from −15 to +45°C, using broadline solid state deuterium NMR of a single deuterated tryptophan located in the loop’s N terminal hinge. The rate of the loop opening and closing was best detected using samples containing subsaturating amounts of a substrate analogue dl-glycerol 3-phosphate so that the populations of the open and closed forms were roughly equal, and using temperatures optimal for enzymatic function (30–45°C). The T2 values were much shorter than for unligated samples, consistent with full opening and closing of the loop at a rate of order 104 s−1, and in good agreement with the rates estimated based on solution state 19F NMR. The loop motion appears to be partially rate limiting for chemistry in both directions.

Introduction

The catalytic reaction of triosephosphate isomerase (TIM) was one of the first clear demonstrations of the role of protein dynamics in promoting and controlling chemical reactivity. TIM catalyses the reversible isomerization of dihydroxyacetone phosphate (DHAP) to d-glyceraldehyde 3-phosphate (GAP), while suppressing elimination of orthophosphate Richard 1991, Rose 1962. TIM is highly efficient in the thermodynamically “downhill” direction, with a kcat value of 9 × 103 s−1; enhancing the reaction by ten orders of magnitude over the reaction carried out under optimized solution conditions. TIM has therefore been described as an optimally efficient enzyme, or even as an exemplar of diffusion controlled reactivity. A vast body of kinetic, mechanistic and structural studies Albery and Knowles 1976a, Albery and Knowles 1976b, Davenport et al 1991, Knowles 1991, Komives et al 1991, Komives et al 1995, Lodi and Knowles 1991, Lolis et al 1990, Lolis and Petsko 1990, Wierenga et al 1991 support a proton transfer-based mechanism involving a cis-enediol or enediolate intermediate Rieder and Rose 1959, Rose 1962 (Figure 1(a)). The free-energy profile for the reaction was deduced from a series of deuterium and tritium isotope exchange experiments between the enzyme-substrate complex and the solvent (Albery & Knowles, 1976b). The rate-limiting step, in the thermodynamically unfavorable but metabolically crucial direction from DHAP to GAP, is not the proton transfer step, nor the diffusion of substrate to the enzyme, but is the loss of GAP product from the enzyme or, as Albery & Knowles (1976b) proposed, a slow conformational change preceding such a loss (Maister et al., 1976). The relation between the proposed rate-limiting conformational change and the residence time of the intermediate or substrate suggests a role for conformational dynamics in coordination of the reaction (Alber et al., 1983). Our aim is to characterize these conformational dynamics, and understand the relation between rate and enzyme optimization.

Loop 6, which contacts the active site, appears to move substantially during function (Figure 1(b)) Alber et al 1981a, Joseph et al 1990, Lolis and Petsko 1990, Phillips et al 1976. The loop has two conformers, an open and a closed state, the mechanistic purposes of which appear to be clear: the open conformation facilitates binding and release while the closed conformation catalyzes the chemical reaction. The closed loop environment organizes several ionizable resides around the substrate to facilitate the core of the chemical reactions, a proton transfer between C1 and C2 methylene via the active site Glu165 with stabilization offered from the nearby neutral His95. The key catalytic residue, Glu165, moves in coordination with the loop. In the empty enzyme Glu165 is hydrogen bonded to Ser96 (Lolis et al., 1990); upon ligand binding this hydrogen bond is broken and a 2 Å shift positions Glu165 in contact with the substrate Davenport et al 1991, Lolis and Petsko 1990. In the closed loop environment the low dielectric constant in the active site increases the enzyme’s ability to moderate the stability of the transition state and intermediates.

While X-ray structural data indicate two strongly populated conformers, they provide no clear information about kinetic aspects. The relation between protein sequence, the rate of the loop motion and enzyme kinetics is relatively unexplored experimentally. The rate of the interconversion of the open and closed forms can be viewed as a separate issue from the relative stabilities of the open and closed forms. Although the ratios are constrained by the affinity, (koff/kon)(kopen/kclosed) = Kd, a barrier for the loop motion imposed by the protein can depress both the open and closed rates and thereby determine the order of magnitude of these rate constants without affecting the stabilities or binding constants. In absence of experimental data it is conceivable that the loop opening rate could be as slow as ∼100 μs, or as fast as ∼0.1 ns. The purpose of this work is to clearly document the timescale of the loop opening and test the hypothesis that it is rate determining for function.

To this end, we have been developing NMR measurements to document loop motion rates. Solid state NMR is capable of providing information about motions on time scales ranging continuously from seconds to pico-seconds, while preserving details of orientation dependence, and therefore was used to study TIM loop motion Palmer et al 1996, Spiess 1978. Previously, deuterium NMR was used to probe the dynamics of the flexible loop in TIM, particularly the motion timescale and the ligands dependence (Williams & McDermott, 1995). Small angle motions on a nanosecond timescale were proposed and larger angle motions on the microsecond timescale were evident. A model was presented that embodied information from the crystallographic studies and the NMR studies. Putatively, the loop opens and closes on a time scale slightly faster than the enzyme turnover rate and the two predominant conformers, open and closed coordinates from crystallography, are unequally populated. Here, we show detailed temperature dependent and ligand dependent data that characterize the motion in the presence of the substrate or substrate analogues, and allow us to measure the rate constants much more accurately. Along with additional solution studies in the accompanying paper (Rozovsky et al., 2001), these data allow us to demonstrate that loop motion is probably partially rate determining for the catalytic reaction.

Section snippets

Sample conditions for NMR experiments

All experiments involved a mutant of S. cerevisiae TIM (Trp90Tyr Trp157Phe), in which only one tryptophan residue, Trp168, was retained and labeled with deuterons in its indole ring. This reporter tryptophan residue is conserved in all TIM sequences and is located on the N-terminal hinge of the active site flexible loop. Upon loop closure the reporter tryptophan swings by about 45° and hence is a sensitive indicator of the loop conformation and dynamics. The Trp90Tyr Trp157Phe mutant displays

Conclusions

The opening and closing motion of the active site loop 6 in TIM has a rate constant that closely matches the turnover time for catalysis. This motion causes dramatic changes in the NMR dephasing (T2) times when the open and closed states are both substantially populated. The motion is also strongly temperature dependent. We conclude that the loop motion is likely to be partially rate determining for the chemical reaction, consistent with the observation that modest isotope effects are seen, and

Materials

All reagents used were purchased from Sigma-Aldrich Co. with the exception of glycerol-3-phosphate dehydrogenase from Boehringer Mannheim Ltd. and deuterium depleted water from Cambridge Isotope Laboratories, Inc. L-[2,4,5,6,7-2H5] Tryptophan was synthesized as described by Matthews et al. (1977) from l-tryptophan by acid catalyzed exchange.

Preparation of yeast mutant triosephosphate isomerase

The expression vector, a modified pKK223-3 containing S. cerevisiae (baker’s yeast) Triosephosphate isomerase gene with a Trp90Tyr Trp157Phe mutation was a

Acknowledgements

The authors thank Dr Jim Frye from Varian, Inc., for valuable advice regarding broadband deuterium NMR experiments, Dr Edward O’Connor and Dr John Williams for helpful discussions and Dr Greg Petsko, of Brandeis University, for stimulating conversations regarding the TIM mechanism. This work was supported by National Institutes of Health grant GM49964.

References (43)

  • W.J. Albery et al.

    Free-energy profile of the reaction catalyzed by triosephosphate isomerase

    Biochemistry

    (1976)
  • C.A. Browne et al.

    Studies of triosephosphate isomerase by hydrogen exchange

    Biochem. J.

    (1974)
  • I.D. Campbell et al.

    Enzyme-substrate and enzyme-inhibitor complexes of triose phosphate isomerase studied by 31P nuclear magnetic resonance

    Biochem. J.

    (1979)
  • R.C. Davenport et al.

    Structure of the triosephosphate isomerase-phosphoglycolohydroxamate complexan analogue of the intermediate on the reaction pathway

    Biochemistry

    (1991)
  • T.K. Harris et al.

    Proton transfer in the mechanism of triosephosphate isomerase

    Biochemistry

    (1998)
  • R.B. Jones et al.

    Spectrophotometric studies on the interaction between triose phosphate isomerase and inhibitors

    Biochem. J.

    (1979)
  • D. Joseph et al.

    Anatomy of a conformational changehinged “lid” motion of the triosephosphate isomerase loop

    Science

    (1990)
  • J.R. Knowles

    To build an enzyme

    Philos. Trans. R. Soc. London, ser. B

    (1991)
  • E.A. Komives et al.

    Electrophilic catalysis in triosephosphate isomerasethe role of histidine-95

    Biochemistry

    (1991)
  • E.A. Komives et al.

    The structural basis for pseudoreversion of the E165D lesion by the secondary S96P mutation in triosephosphate isomerase depends on the positions of active site water molecules

    Biochemistry

    (1995)
  • P.J. Kraulis

    MOLSCRIPTa program to produce both detailed and schematic plots of protein structures

    J. Appl. Crystallog.

    (1991)
  • Cited by (114)

    View all citing articles on Scopus
    1

    Edited by P. E. Wright

    View full text