Many proteins frequently undergo structural rearrangement to complete their functions. Ligand entry and binding are often associated with some degree of localized disorder. Indeed, low populations of disordered excited states may help drive such processes. Characterization of these states is vital to understanding the mechanisms of many biological functions.
Transitions from low energy, ground state conformations to higher energy, excited state conformations play a vital role in protein function. Consequently, to fully comprehend the biological mechanism of a protein, it is essential to obtain a full description of the energetics and kinetics of the processes involved in such transitions. The time-dependent conformational fluctuations of proteins, which generate excited states — albeit with low populations — can be exquisitely investigated by a powerful suite of relatively new heteronuclear NMR relaxation techniques1. Although initial NMR relaxation experiments and analyses concentrated on protein dynamics on the ps-ns timescale2,3 more recent methodologies have focused on μs-ms motions1.
Why the growing interest in such slower movements? As more evidence accumulates, it becomes clear that many biological processes exhibit rates that coincide with this timescale. For example, rates for enzyme catalysis and product release4,5, protein folding6,7 and allosteric transitions8 all occur in this time regime. Hand-in-hand with the continuing design and implementation of these novel NMR dynamics experiments has been the development of theoretical approaches for extracting thermodynamic and kinetic information from relaxation data9,10,11,12. In principle, therefore, we now have the capability to characterize both the thermodynamics and kinetics of the transition between the ground state and the excited state of a protein. NMR spectroscopy is unique in this regard, allowing the investigation of important biological events that are relatively inaccessible to other techniques.
On page 932 of this issue of Nature Structural Biology, Kay, Dahlquist and coworkers describe one fine example of such characterization13. These researchers thoroughly investigated the transition of the L99A mutant of T4 lysozyme between its ground state conformation, which is inaccessible to ligand binding, and its excited state, which allows for binding. By using recently developed relaxation dispersion NMR techniques performed at several magnetic field strengths and over a range of temperatures, Kay and coworkers have calculated the enthalpic and entropic contributions to the interconversion between states. In addition, they present evidence to bolster the suggestion that ligand binding to this mutant is related to the slower timescale dynamics of the protein. The partial 'unfolding' or 'local disorder' that the protein dynamics confer in the binding region of L99A is thought to be important for ligand recognition/binding in a variety of other proteins that target small hydrophobic molecules.
The 'hole' story
Why are the dynamics of this particular lysozyme mutant interesting? Why the myriad of crystallographic and NMR studies? The answer lies at the core of this protein. Most proteins possess a core of well-packed hydrophobic residues that have been regarded, until recently, as structurally static. In 1992, Brian Matthews and his group at the University of Oregon discovered that a 150 Å3 cavity could be engineered into the core of T4 lysozyme by replacing the large leucine residue at position 99 with a much smaller alanine residue14 (Fig. 1). This group then went on to demonstrate, using crystallographic data, that the protein structure did not relax to fill this cavity and that small hydrophobic ligands, such as benzene, could easily bind at this location. Crystallographic studies determined that the structure of the mutant with and without bound ligands is very similar to that of wild type lysozyme. Matthews' group also examined the structural aspects inherent in the binding of a whole host of ligands to the L99A mutant15,16. Perhaps the most intriguing finding is that the bound ligands are buried 7 Å into the core of the protein, leading to the supposition that the L99A mutant must possess backbone motions large enough to allow for the entry of these small ligands. It is also noteworthy that the required dynamic processes can occur in spite of the structural restraints imposed by the crystal lattice.
Shortly after these initial crystallographic studies, Dahlquist and coworkers17 used NMR-determined ligand exchange kinetics to examine the magnitude of structural fluctuations required for ligand binding and calculated the kinetics of binding. The results indicate that L99A must undergo large core backbone structural fluctuations (1–2 Å) on the μs–ms timescale in order to accommodate ligand entry. Further insights into the dynamics of this system languished for some time because a straightforward way to examine millisecond and microsecond movements of proteins at atomic resolution was not yet in anyone's experimental arsenal.
Building the right experiment
Until very recently, biomolecular NMR experimentalists have focused the majority of their attention on phenomena that occur on the ps–ns timescale, and many established experimental protocols are in place to characterize such events (for reviews of these methods and their applications, see refs 2,3,18,19). Complementary methods for measuring very slow chemical exchange processes have also been available. These experiments track 'long-lasting' longitudinal magnetization as it is transferred between nuclear sites in a protein. Referred to as ZZ-exchange methods, these experiments study exchange in the 0.1–10 s−1 range20,21,22. However, the ability to monitor μs–ms timescale kinetic processes remained somewhat more elusive. Detailed analysis of NMR lineshapes was one approach used in an attempt to extract these types of data. Although much good work was done in this area, a completely reliable method for extracting the required transverse relaxation rate constants remained out of reach for reasons not yet completely understood23.
The breakthrough for investigating μs–ms exchange phenomena came from Akke and Palmer24 in 1996. They introduced a 15N relaxation experiment that employed an off-resonance radiofrequency field to allow all the chemically shifted 15N nuclei in a protein to be studied simultaneously. The strength of this off-resonance radiofrequency field dictates the rates of the exchange process that can be detected. The experiment also contained a constant time relaxation period that removes unwanted resonance offset effects. This pulse sequence, referred to as the 'R1ρ–R1 constant relaxation time experiment', is able to effectively measure exchange processes on the microsecond to submillisecond timescale (Kex < 1 × 105 s−1). The use of this experiment is limited primarily by hardware restrictions resulting from the need to use ever-increasing off-resonance field strengths. An important practical concern introduced by this need is that radiofrequency (RF) pulses with increased field strengths and long duration may heat up a sample and cause the protein to denature. Over the next two years, the basic experiment was modified to incorporate adiabatic pulses for better alignment of the nuclear spins with the effective field, thereby increasing sensitivity. In addition, gradient coherence selection was added to aid in artifact suppression and in the reduction of experimental NMR time25.
Even after such modifications, a 'millisecond' gap still remained. In 1999, Palmer and coworkers introduced a modified 15N spin-echo experiment to fill this void26. The modification was elegantly simple, involving the introduction of a new element into the existing pulse sequence. The result of this new element was to average the differential relaxation properties of two different types of magnetization present in the system during the course of the pulse sequence. Without this averaging effect, it would be impossible to examine exchange kinetics from the data since there would be too many interdependent parameters to solve for. This modified experiment, referred to as the 'relaxation-compensated CPMG' sequence, allowed exchange events to be measured in the critical 0.5–5 ms time range. Again, the ability to generate high-powered pulse with appropriate fast repetition rate imposes limits on the experiment. Nevertheless, the combination of these powerful experiments allows the accurate assessment of exchange processes on the microsecond and millisecond timescales.
The relaxation-compensated CPMG experiments, noted above, were designed to study backbone motions. These experiments were recently extended to include the ability to study 0.5–5 millisecond methyl side chain dynamics by a variety of research groups, particularly those of Kay27,28 and Torchia29. All of these backbone and side chain experiments are usually referred to as 'dispersion' experiments as they monitor relaxation rates as a function of other critical variables. One often used critical variable is the radiofrequncy field strength, but other variables can include temperature and magnetic field strength. The utilization of these 15N and 13C backbone and side chain relaxation dispersion experiments allowed the groups of Kay and Dahlquist to revisit the T4 lysozyme story.
Measuring the hole dynamics
As is characteristic of these researchers the current set of studies have been comprehensive and far-reaching27,28,30. They have centered on the use of relaxation dispersion experiments to meticulously probe the flexibility and ligand exchange of the buried cavity in the L99A T4 lysozyme mutant13. Here heteronuclear spin-echo experiments are used to access the thermodynamic and kinetic properties of the transition of the L99A T4 lysozyme mutant from its ligand-inaccessible ground state to a 3% populated excited state. This excited state likely facilitates ligand entry. Dispersion profiles were measured for a large set of nuclei, including backbone nitrogens and side chain methyl carbons, at different magnetic field strengths and over a range of temperatures. Analysis of these data reveals that more than half the nuclei are not involved in an exchange process; however, a significant number of methyls and amides were adequately fit by a two-site exchange phenomenon — the ground to excited state interconversion. Most interesting, the methyl and amide groups that display conformational exchange on this timescale are located around the cavity.
Beyond visualizing the sites in the polypeptide chain undergoing transitions, a tremendously appealing aspect of this work is the straightforward manner in which the thermodynamic parameters were obtained. Rate constants for the exchange are easily generated from the dispersion profiles and the equilibrium constant for the interconversion is calculated (K). By virtue of recording the relaxation data at different temperatures, the temperature dependence of ln (K) can be very simply related to the entropy and enthalpy of the process, assuming the reaction follows the Arrhenius behavior. Although a large amount of experimental NMR time is required to produce the necessary data, the clear-cut thermodynamic analysis makes these efforts extremely worthwhile.
The L99A T4 lysozyme mutant is not unique with regard to its backbone and side chain dynamics, and the relationship between these processes and the function of the protein. In fact, many biological processes proceed in the μs–ms regime. For example, the folding of the peripheral subunit-binding domain of the pyruvate dehydrogenase complex31 occurs on this timescale. Relaxation-compensated CPMG data offer the promise of a detailed analysis of this process. In addition to protein folding, many protein–ligand interactions occur on this timescale. For example, in the study of the insulin receptor substrate 1 phosphotyrosine binding domain and its binding to an interleukin 4 receptor phosphopeptide substrate, Fesik and coworkers lament that they were forced to estimate the rates of the chemical exchange processes32. The appropriate tools did not exist to measure and analyze them directly at that time. An attention-grabbing example of the type of systems ripe for study by the methodologies discussed here is the recently published inhibition of HIV protease by an inorganic cluster. These novel polyoxometalates present a new mechanism of inhibition because they are shown to bind to the surface of the protease, where they decrease the mobility of flaps that cover the active site of the protein33 (Fig. 2). Surely, direct measurements of the dynamics of this entirely new inhibition mechanism must prove invaluable in the fight against this protein that somehow seems to find a way of becoming resistant to other types of inhibitors. For these reasons, and many more yet to come, the methods presented in the Kay and Dahlquist paper13 will find their way into every NMR spectroscopist's toolbox.
Palmer, A.G., Kroenke, C.D. & Loria, J.P Methods Enzymol. 339, 204–238 (2001).
Kay, L.E. Nature Struct. Biol. 5, 513–517 (1998).
Palmer, A.G. Curr. Opin. Struct. Biol. 7, 732–737 (1997).
Voet, D. & Voet, J.G., Biochemistry (John Wiley & Sons, New York; 1990).
Fersht, A. Enzyme structure and mechanism, 2nd ed. (Freeman & Co., New York; 1995).
Jackson, S.E. Folding Des. 3, 81–91 (1998).
Roder, H. & Shastry, M.C.R. Curr. Opin. Struct. Biol. 9, 620–626 (1999).
McCammon, J.A. & Harvey, S.C., Dynamics of proteins and nucleic acids (Cambridge University Press, Cambridge, UK; 1987).
Akke, M., Bruschweiler, R. & Palmer, A.G. J. Am. Chem. Soc. 115, 9832–9833 (1993).
Yang, D. & Kay, L.E. J. Mol. Biol. 263, 369–382 (1996)
Li, Z., Raychaudhuri, S. & Wand, A.J. Protein Sci. 5, 2647–2650 (1996).
Denisov, V.P., Venu, K., Peters, J., Horlein, H.D. & Halle, B. J. Phys. Chem. 101, 9380–9389 (1997).
Mulder, F.A.A., Mittenmaier, A., Hon, B., Dahlquist, F.W. & Kay, L.E. Nature Struct. Biol. 8, 932–935 (2001).
Eriksson, A.E., Baase, W.A., Wozniak, J.A. & Matthews, B.W. Nature 355, 371–373 (1992).
Morton, A. & Matthews, B.W. Biochemistry 34, 8576–8588 (1995).
Zhang, X-j., Wozniak, J.A. & Matthews, B.W. J. Mol. Biol. 250, 527–532 (1995).
Feher, V.A., Baldwin, E.P. & Dahlquist, F.W. Nature Struct. Biol. 3, 516–521 (1996).
Daragan, V.A. & Mayo, K.H. Prog. Nucl. Magn. Reson. Spectrosc. 31, 63–105 (1997).
Fischer, M.W.F., Majumdar, A. & Zuiderweg, E.R.P. Prog. Nucl. Magn. Reson. Spectrosc. 33, 207–272 (1998).
Montelione, G.T. & Wagner, G. J. Am. Chem. Soc. 111, 3096–3098 (1989).
Farrow, N., Zhang, O., Forman-Kay, J.D. & Kay L.E. J. Biomol. NMR 4, 727–734 (1994).
Wider, G., Neri, D. & Wuthrich, K. J. Biomol. NMR 1, 93–98 (1991).
Rao, B.D.N. Methods Enzymol. 176, 279–311 (1989).
Akke, M. & Palmer, A.G. J. Am. Chem Soc. 118, 911–912 (1996).
Zinn-Justin, S., Berthault, P., Guenneugues, M. & Desvaux, H. J. Biomol. NRM 10, 363–372 (1997).
Loria, J. P., Rance, M. & Palmer, A.G. J. Am. Chem. Soc., 121, 2331–2332 (1999).
Skrynnikov, N.R., Mulder, F.A.A., Hon, B., Dahlquist, F. & Kay, L.E. J. Am. Chem. Soc. 123, 4556–4566 (2001).
Mulder, F.A.A., Hon, B., Muhandiram, D.R., Dahlquist, F.W. & Kay, L.E. Biochemistry 39, 12614–12622 (2000).
Ishima, R., Louis, J. & Torchia, D.A. J. Am. Chem. Soc. 121, 11589–11590 (1999).
Goto, N.K., Skrynnikov, N.R., Dahlquist, F.W. & Kay, L.E. J. Mol. Biol. 308, 745–764 (2001).
Vugmeyster, L., Kroenke, C.D., Picart, F., Palmer III, A.G. & Raleigh, D.P. J. Am. Chem. Soc. 122, 5387–5388 (2000).
Olejniczak, E.T., Zhou, M-M. & Fesik, S.W. Biochemistry 36, 4118–4224 (1997).
Judd, D.A. et al. J. Am. Chem. Soc. 123, 886–897 (2001).
The authors would like to thank B. Matthews and C. Hill for kindly allowing us to use their figures for illustrative purposes and the appropriate journals for permission to reproduce them, and Mark Rance for a fruitful discussion.
About this article
Cite this article
Cavanagh, J., Venters, R. Protein dynamic studies move to a new time slot. Nat Struct Mol Biol 8, 912–914 (2001). https://doi.org/10.1038/nsb1101-912
The Journal of Physical Chemistry B (2020)
Characterizing micro-to-millisecond chemical exchange in nucleic acids using off-resonance R1ρ relaxation dispersion
Progress in Nuclear Magnetic Resonance Spectroscopy (2019)
Longer simulations sample larger subspaces of conformations while maintaining robust mechanisms of motion
Proteins: Structure, Function, and Bioinformatics (2012)
BMC Bioinformatics (2011)
Journal of Magnetic Resonance (2009)