Low-Temperature Decoupling of Water and Protein Dynamics

Letter pubs.acs.org/JPCL

Low-Temperature Decoupling of Water and Protein Dynamics Measured by Neutron Scattering Antonio Benedetto*,†,‡ †

School of Physics, University College Dublin, Dublin 4, Ireland Laboratory for Neutron Scattering, Paul Scherrer Institut, Villigen, Switzerland

‡

ABSTRACT: Water plays a major role in biosystems, greatly contributing to determine their structure, stability, and function. It is well known, for instance, that proteins require a minimum amount of water to be fully functional. Despite many years of intensive research, however, the detailed nature of protein− hydration water interactions is still partly unknown. The widely accepted “protein dynamical transition” scenario is based on perfect coupling between the dynamics of proteins and that of their hydration water, which has never been probed in depth experimentally. I present here high-resolution elastic neutron scattering measurements of the atomistic dynamics of lysozyme in water. The results show for the first time that the dynamics of proteins and of their hydration water are actually decoupled at low temperatures. This important result challenges the “protein dynamical transition” scenario and requires a new model to link protein dynamics to the dynamics of its hydration water.

I

hydration water, the authors proposed to link the PDT to the functioning of proteins. The common way to measure such PDT experimentally is by elastic neutron scattering, which is collected as a function of the sample-temperature in what is called “fixed-window scan”. This plot shows an abrupt decrease for the hydrated protein with respect to the dry one at TPDT (Figure 2). Consequently, the research community has focused on understanding the protein−water relaxation mechanisms underlying the PDT and, in turn, the biochemical function. To the best of my knowledge, the only attempt to propose a unified model of protein dynamics was made by Frauenfelder and coworkers.15 In their paper they proposed that protein motions are modulated by the hydration shell and by the bulk solvent. Essentially, they proposed that (i) large-scale protein motions are coupled to fluctuations in the bulk solvent that are controlled by the solvent viscosity, which are absent in a solid environment, whereas (ii) internal protein motions are coupled to the beta fluctuations of the hydration shell, they are controlled by hydration, and are absent in dehydrated proteins. This intriguing scenario in which water molecules drive protein dynamics and, in turn, their functions has to be considered together with the relaxation dynamics of the protein hydration water itself. This protein hydration water relaxation has been measured by Chen and coworkers:16 By increasing the temperature, it revealed a crossover from Arrhenius (strong glass-like) to non-Arrhenius (fragile glass-like) behavior at TFSC ≈ 220 K. The picture that emerges is even more intriguing: The change in the hydration−water dynamics at TFSC triggers the

n 560 BC, Thales hypothesized that water is the primary essence of life,1 and nowadays this hypothesis is widely accepted. Since a few decades ago, it has become well established that water molecules actively interact with, and support, the biochemistry of different classes of biomolecules.2−11 In the case of proteins, for instance, the water molecules adsorbed at the protein surface play a major biological role, and it is well known that proteins require a minimum amount of water to be fully biologically active. These water molecules form the so-called protein hydration layers, usually consisting of one or two layers adsorbed at the protein surface and accounting for ∼20% of the hydrated protein weight (Figure 1a). These water molecules display physicochemical properties that are apparently different from molecules in pure bulk water. For instance, they are prevented from crystallizing just below 0 °C because the competition of their mutual interaction and their hydrogen bonding to the protein makes it difficult for them to arrange into the typical tetrahedral ice structure (Figure 1b, cycle 1). Remarkably, proteins’ denaturation restores water ability to crystallize (Figure 1b, cycle 2). Because the biological functions of proteins involve changes of conformation and sometimes chemical reactions, it is natural to expect a connection of these functions with dynamical properties of the coupled system made of proteins with their hydration water. An important thread in this discussion started in 1989 with the seminal neutron scattering work of Doster and coworkers12 pointing to a sharp increase in the atomic meansquare displacement of a hydrated model protein at around TPDT ≈ 220 K. The authors referred to this transition as the “protein dynamical transition” (PDT). Because proteins seem to become biologically active around this temperature13,14 and because this transition was not observed in the absence of © XXXX American Chemical Society

Received: August 28, 2017 Accepted: September 22, 2017 Published: September 22, 2017 4883

DOI: 10.1021/acs.jpclett.7b02273 J. Phys. Chem. Lett. 2017, 8, 4883−4886

Letter

The Journal of Physical Chemistry Letters

Figure 3. A useful analogy. To better probe if two objects moving in the same direction either have or do not have the same velocity, they should be observed by increasing the observation time. If they have two slightly different velocities (case a), then a longer observation time can probe the difference. The “observation time” is analogous to the “resolution time” in an elastic neutron experiment, which in this study is increased by a factor of 3 with respect to the past investigations.

experiment: one by hydrating with H2O and the other with D2O, in both cases with an amount of water corresponding to the protein hydration water as sketched in Figure 1a (i.e., between about 0.3 and 0.4 g of water per gram of protein). The elastic spectrum of the protein in the dry state also has to be collected to provide a baseline. The rationale is that neutrons are very sensitive to hydrogen atoms and much less sensitive to deuterium, enabling a distinction between the isotopes. The biochemical function, however, is only slightly affected by the use of heavy water instead of water. As a result, the D2O signal from a protein hydrated in D2O is negligible, and relaxations in the elastic spectra can be related to the protein itself, which also contains a high number of hydrogen atoms. In this way it is possible to probe the relaxation dynamics of protein alone, yet hydrated, and extract TPDT. In the case of a protein hydrated in H2O, the contributions of water and protein have very similar weights in the signal (i.e., the densities of hydrogen atoms in the protein and in hydration water are nearly equivalent). As a result, by measuring the protein hydrated in H2O, relaxations in the elastic spectra arise from relaxations of either the protein or its hydration water. By comparison of the spectra of the protein in D2O with that in H2O, it is possible to determine the relaxation process of hydration water alone and, in turn, extract its transition temperature. With the current state of the art, all experiments so far show proteins and their hydration water to have the same transition

Figure 1. Model protein lysozyme and its hydration water. (a) Crystal structure of human lysozyme (green) together with (∼280) hydration water molecules obtained by X-ray diffraction at 1.8 Å of spatial resolution.9 (b) Differential scan calorimetric heating profiles before (blue) and after (red) thermal denaturation (blue peak). The hydration water due to the interaction with the protein is not able to form ice, but after the thermal denaturation, due to a different state of the protein, part of the hydration water molecules recover their ability to form ice (red peak).

protein internal motions that also became active at TFSC. This would imply that at TFSC a sudden variation in the protein dynamics should occur as well. The scenario emerging from this logical implication is supported by the fact that TPDT ≈ TFSC ≈ 220 K, which is a strong coupling between the dynamics of proteins and of their hydration water. The coupling between the transition in the hydration−water dynamics and the PDT has been observed in all neutronscattering experiments on this subject so far. Again, as abovementioned, the elastic neutron scattering intensity versus temperature represents one of the most used observables to study such protein−hydration water coupled dynamics, which allows the dynamics of proteins and of their hydration water to be probed independently. To achieve this aim, two elasticscattering profiles of the hydrated protein are collected in my

Figure 2. Total elastic neutron scattered intensity versus temperature for D2O-hydrated lysozyme (green), H2O-hydrated lysozyme (blue), and dry lysozyme (red) at 1 μeV resolution collected on the IN10 neutron spectrometer (Q-range = [0.2÷1.9]Å−1) at the Institut Laue-Langevin in Grenoble, France.17 Comparison of hydrated and dry lysozyme clearly shows the strong coupling between the dynamics of the protein and of its hydration water that both start to relax at about TPDT = 200 K as from the fits in panels b and c. 4884

DOI: 10.1021/acs.jpclett.7b02273 J. Phys. Chem. Lett. 2017, 8, 4883−4886

Letter

The Journal of Physical Chemistry Letters

Figure 4. High-resolution elastic neutron data showing the decoupling between the dynamics of protein and of its hydration water. Total elastic neutron-scattered intensity (Q-range = [0.2÷1.9]Å−1) versus temperature for D2O hydrated lysozyme (green), H2O hydrated lysozyme (blue), and dry lysozyme (red). The Q-range perfectly matches the one accessed in Figure 2 by IN10. The comparison in panel a clearly shows the decoupling between the dynamics of lysozyme and of its hydration water. The protein starts to relax at TP = 195 K, decoupled from the transition in the dynamics of hydration water, which only starts to relax 16 K lower, that is, TW = 179 K. (b,c) High counting-statistic of the experiment also allows us to probe the effect of hydrating with heavy water: In this case the protein dynamics is slightly slower that when hydrated in isotopically normal water.

decoupling suggested by other studies points to some new scenario. I aim to shed light on this 30 year old puzzle by using elastic neutron scattering at the highest achievable resolution for this type of studies, that is, 0.3 μeV in place of the more usual 1 μeV used so far. Lysozyme still represents the protein component. The rationale is that (i) if the relaxation times of two (or more) processes are identical (i.e., perfect coupling of their dynamics), then their coupling has to be independent of the time resolution of the experiment; however, (ii) if their relaxation times are different but so similar that they cannot be separated at a given resolution, then an higher resolution can see their decoupling. For more details of elastic neutron scattering spectroscopy and energy resolution, refer to refs 27 and 28. However, a good analogy is the following: Imagine two objects moving with two different but very similar velocities in the same direction. The longer the observation time, the better the chance of seeing a difference in their motions. However, if the two velocities are identical, then no difference in their motions can be seen at any observation time (Figure 3). In Figure 4, the elastic intensity as a function of temperature for dry and hydrated lysozyme is presented. The data have been collected on the IN16B spectrometer 29 (Q-range = [0.2÷1.9]Å−1) at the Institut Laue-Langevin in Grenoble, France in its highest resolution setup, ΔωRES = 0.3 μeV. This highest energy resolution allows me to measure the decoupling between the dynamics of proteins and their hydration water for the first time. All previous experiments on hydrated proteins have been carried out at worse energy resolutions (∼1 μeV). Figure 4 clearly shows decoupling of the hydration-water motion and of the protein−relaxation dynamics: The onset of anharmonic relaxation is observed at TW = 179 ± 1 K in the hydration water dynamics and at TP = 195 ± 1 K for the protein, that is, 16 K above. (Figure 4c).

Figure 5. Measured mean-square displacements showing the decoupling between the dynamics of protein and of its hydration water.

temperature, which strongly supports the PDT scenario described above. That is, the dynamics of proteins and of their hydration water are strongly coupled. In Figure 2, the classic example of the model protein, lysozyme, is shown. On the contrary, there is a collection of experimental and computational results18−25 that is indeed in disagreement with the above-pictured PDT scenario.12−16,26 Several of these outof-scenario results are based on the fact that no changes in the dynamical behavior of proteins seem to occur at all at the PDT temperature, but they do seem to agree on a sort of dynamical crossover in the relaxation dynamics of the protein hydration− water. Furthermore, a recent work25 also shows that proteins exhibit a dynamical transition even in the dry state. In summary, the PDT scenario is supported by the experimentally measured perfect coupling between the dynamics of proteins and their hydration water, as shown in Figure 2, while the 4885

DOI: 10.1021/acs.jpclett.7b02273 J. Phys. Chem. Lett. 2017, 8, 4883−4886

Letter

The Journal of Physical Chemistry Letters

(10) Chaplin, C. Do We Underestimate the Importance of Water in Cell Biology? Nat. Rev. Mol. Cell Biol. 2006, 7, 861−866. (11) Ball, P. Water as an Active Constituent in Cell Biology. Chem. Rev. 2008, 108, 74−108. (12) Doster, W.; Cusack, S.; Petry, W. Dynamical Transition of Myoglobin Revealed by Inelastic Neutron Scattering. Nature 1989, 337, 754−756. (13) Rasmussen, B. F.; Stock, A. M.; Ringe, D.; Petsko, G. A. Crystalline Ribonuclease A Loses Function Below the Dynamical Transition at 220 K. Nature 1992, 357, 423. (14) Ostermann, A.; Waschipky, R.; Parak, F. G.; Nienhaus, G. U. Ligand Binding and Conformational Motions in Myoglobin. Nature 2000, 404, 205−208 (2000).. (15) Frauenfelder, H.; Chen, G.; Berendzen, J.; Fenimore, P. W.; Jansson, H.; McMahon, B. H.; Stroe, I. R.; Swenson, J.; Young, R. D. A Unified Model of Protein Dynamics. Proc. Natl. Acad. Sci. U. S. A. 2009, 106, 5129−5134. (16) Chen, S. H.; Liu, L.; Fratini, E.; Baglioni, P.; Faraone, A.; Mamontov, E. Observation of Fragile-to-Strong Dynamic Crossover in Protein Hydration Water. Proc. Natl. Acad. Sci. U. S. A. 2006, 103, 9012−9016. (17) See http://www.ill.eu/instruments-support/instrumentsgroups/instruments/in10. (18) Khodadadi, S.; Pawlus, S.; Roh, J. H.; Garcia Sakai, V.; Mamontov, E.; Sokolov, A. P. The Origin of the Dynamic Transition in Proteins. J. Chem. Phys. 2008, 128, 195106. (19) Pawlus, S.; Khodadadi, S.; Sokolov, A. P. Conductivity in Hydrated Proteins: No Signs of the Fragile-to-Strong Crossover. Phys. Rev. Lett. 2008, 100, 108103. (20) Becker, T.; Smith, J. C. Energy Resolution and Dynamical Heterogeneity Effects on Elastic Incoherent Neutron Scattering From Molecular Systems. Phys. Rev. E: Stat. Phys., Plasmas, Fluids, Relat. Interdiscip. Top. 2003, 67, 021904. (21) Becker, T.; Hayward, J. A.; Finney, J. L.; Daniel, R. M.; Smith, J. C. Neutron Frequency Windows and The Protein Dynamical Transition. Biophys. J. 2004, 87, 1436−1444. (22) Magazu, S.; Migliardo, F.; Benedetto, A. Puzzle of Protein Dynamical Transition. J. Phys. Chem. B 2011, 115, 7736−7743. (23) Young, R. D.; Frauenfelder, H.; Fenimore, P. Mossbauer Effect in Proteins. Phys. Rev. Lett. 2011, 107, 158102. (24) Nandi, P. K.; English, N. J.; Futera, Z.; Benedetto, A. HydrogenBond Dynamics at the Bio−Water Interface in Hydrated Proteins: a Molecular-Dynamics Study. Phys. Chem. Chem. Phys. 2017, 19, 318− 329. (25) Liu, Z.; Huang, J.; Tyagi, M.; O’Neill, H.; Zhang, Q.; Mamontov, E.; Jain, N.; Wang, Y.; Zhang, J.; Smith, J. C.; Hong, L. Dynamical Transition of Collective Motions in Dry Proteins. Phys. Rev. Lett. 2017, 119, 048101. (26) Schirò, G.; Fichou, Y.; Gallat, F.-X.; Wood, K.; Gabel, F.; Moulin, M.; Härtlein, M.; Heyden, M.; Colletier, J.-P.; Orecchini, A.; et al. Translational Diffusion of Hydration Water Correlates With Functional Motions in Folded and Intrinsically Disordered Proteins. Nat. Commun. 2015, 6, 6490. (27) Magazu’, S.; Migliardo, F.; Benedetto, A. Elastic Incoherent Neutron Scattering Operating by Varying Instrumental Energy Resolution: Principle, Simulations, and Experiments of the Resolution Elastic Neutron Scattering (RENS). Rev. Sci. Instrum. 2011, 82, 105115. (28) Benedetto, A.; Kearley, G. J. Elastic Scattering Spectroscopy (ESS): an Instrument-Concept for Dynamics of Complex (Bio-) Systems From Elastic Neutron Scattering. Sci. Rep. 2016, 6, 34266. (29) See http://www.ill.eu/instruments-support/instrumentsgroups/instruments/in16b.

Consistently with the elastic intensities of Figure 4, the dynamical decoupling occurs in the measured mean-square displacements as well (Figure 5). In conclusion, with the limited energy resolutions used in the past (up to ∼1 μeV on the best high-resolution spectrometers), proteins and their hydration water appeared to undergo a dynamical transition at the same temperature. This prompted the erroneous conclusion of a perfect-coupled water−protein relaxation dynamics, supporting, in turn, the conjecture that transitions in hydration−water dynamics are responsible for protein dynamics. My high-resolution experimental data (Figure 4 and 5) reveal for the first time that there is decoupling between the relaxation of a model protein and the relaxation of its hydration water. This is an important milestone toward the solution of this 30 year old puzzle.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Antonio Benedetto: 0000-0002-9324-8595 Notes

The author declares no competing financial interest.

■

ACKNOWLEDGMENTS A.B. thanks Pietro Ballone and Gordon J. Kearley for fruitful discussions. A.B. acknowledges support from (i) the European Community under the Marie-Curie Fellowship Grants HYDRA (No. 301463) and PSI-FELLOW (No. 290605) and from (ii) Science Foundation Ireland (SFI) under the Start Investigator Research Grant 15-SIRG-3538, with additional support provided by the School of Physics, University College Dublin, Ireland, and the Laboratory for Neutron Scattering, Paul Scherrer Institute (PSI), Switzerland. A.B. acknowledges the Institut Laue-Langevin (Grenoble, France) for the access to IN10 and IN16B spectrometers and Drs. Miguel Gonzales and Bernard Frick, in particular, for the help during the neutron beam-time. A.B. acknowledges Drs. Ekaterina Pomjakushina and Antonietta Gasperina (PSI) for the access to the differential scanning calorimetry laboratory and the biolab, respectively.

■

REFERENCES

(1) Aristotle. Metaphysics. 983 b6 8−11, 17−21. (2) Kuntz, I. D., Jr.; Brassfield, T. S.; Law, G. D.; Purcell, G. V. Hydration of Macromolecules. Science 1969, 163, 1329−1331. (3) Kuntz, D.; Kauzmann, W. Hydration of Proteins and Polypeptides. Adv. Protein Chem. 1974, 28, 239−345. (4) Cooke, R.; Kuntz, I. D. The Properties of Water in Biological Systems. Annu. Rev. Biophys. Bioeng. 1974, 3, 95. (5) Careri, G.; Gratton, E.; Yang, P. H.; Rupley, J. A. Correlation of IR Spectroscopy, Heat Capacity, Diamagnetic Susceptibility and Enzymatic Measurements on Lysozyme Powder. Nature 1980, 284, 572−573. (6) Rupley, J. A.; Careri, G. Protein Hydration and Function. Adv. Protein Chem. 1991, 41, 37−172. (7) Otting, G.; Liepinsh, E.; Wuthrich, K. Protein Hydration in Aqueous Solution. Science 1991, 254, 974−980. (8) Israelachvili, J.; Wennerstrom, H. Role of Hydration Water Structure in Biological and Colloidal Interactions. Nature 1996, 379, 219−225. (9) Funahashi, J.; Takano, K.; Yamagata, Y.; Yutani, K. Positive Contribution of Hydration Structure on the Surface of Human Lysozyme to the Conformational Stability. J. Biol. Chem. 2002, 277, 21792−21800. 4886

DOI: 10.1021/acs.jpclett.7b02273 J. Phys. Chem. Lett. 2017, 8, 4883−4886