Protein Hydration Dynamics: Much Ado about Nothing? - The Journal

Oct 5, 2017 - Citation data is made available by participants in Crossref's Cited-by Linking service. For a more comprehensive list of citations to th...
0 downloads 0 Views 3MB Size
Viewpoint pubs.acs.org/JPCL

Protein Hydration Dynamics: Much Ado about Nothing?

W

e all know that water is essential for life on Earth.1−3 Its multifarious functions hardly need any further elaboration.4−7 It not only stabilizes proteins, DNA, and lipid membranes but also plays a crucial role in many, if not all, cellular processes.3,6 Because proteins are large molecules, they can contain more than 1000 water molecules within a thin layer of ∼5Å around its folded state. This layer is termed the protein hydration layer (hereafter referred to as PHL). These water molecules face the heterogeneous surface. Some water molecules face, at a given time, hydrophilic or polar amino acid side chains, some face hydrophobic groups, and some others face peptide backbone atoms. Moreover, these surface water molecules remain connected to the bulk, leading to continuous exchange8 of water molecules between the two regions. The latter tends to minimize the effects of water− protein surface interactions. It is however not clear how much the properties of hydration water can be different or altered due to the proximity to the protein surface. What makes the situation intriguing is that different experimental approaches seem to have given rise to different quantification of the nature of the hydration layer. Scattering experiments have suggested that the structure remains altered even up to ∼20 Å in water.5,9 That is, one needs to move away more than 20 Å to find bulk water. Several experimental estimates of the dielectric constant of the hydration layer have repeatedly given a value of approximately 40, instead of the bulk value of 78. This estimate is obtained from the magnitude of the total fluorescence Stokes shift and is to be regarded as an effective value. Yet, it is the issue of the time scales of water motions (rotation and translation) in the hydration layer that has drawn interest and given rise to controversy. It seems that two schools of thought have emerged. According to one such school, led mostly by experimentalists,7,10−12 this water is sufficiently distinct to merit a different name (“Biological water”). While the second school, based on computer simulation and NMR studies, advocates the idea that the water at the layer is just “bulk-like” and should not be treated separately.5,13,14 The purpose of this Viewpoint is to analyze the two divergent views. To this end, we briefly review the earlier studies in order to gain a broader perspective. Early NMR experiments measured the rotational correlation time of the myoglobin molecule and obtained a value of 45 ns.15 This could not be explained by using the Debye−Stokes−Einstein relation16 with the crystallographic radius of myoglobin. This latter, otherwise reliable procedure, gave only 14 ns!17 Similar problems arose in other proteins as well.18 To explain the discrepancy, an “iceberg model” was put forward where it was assumed that a rigid layer of water molecules ∼3 Å thick remained stuck to the protein as the latter rotated and increased the hydrodynamic radius of proteins.15 Subsequently, it was also realized that this simple hydrodynamic continuum model is not applicable for the PHL. A far more sophisticated approach that takes into account the roughness of the protein surface (by the use of the triaxial ellipsoid method) and © 2017 American Chemical Society

modified dielectric relaxation of the hydration layer provided a more satisfactory description of protein rotation.19 In a series of studies that have relevance here, Pethig,20 Grant,21 Mashimo,22 and co-workers reported, along with the usual 10 ps and 10 ns time scales, the observation of another “mysterious” component of the order of 40 ps. They termed this relaxation in aqueous protein solution “delta-dispersion” and attributed it to the water molecules that reside inside of the PHL. The NMR experiments reported by Wüthrich et al. suggested that the time scale’s relaxation of the layer must be less than ∼300 ps.23 This study is to be understood in the context of then prevailing iceberg model, a fact that is often ignored when criticizing this value. Subsequent NMR experiments found that the average time scales (that is, averaged over all water molecules) becomes slower by only a factor ranging between ∼2 and 5.24 This led to the conclusion that there was no noticeably slow water in the hydration layer, in contradiction to all of the earlier studies. More recently, timedependent fluorescence Stokes shift (TDFSS) studies again found several slow components, with the slowest component becoming slower by a factor of 10−100.7,11,12,25 We note that all of these experiments probe different time scales and length scales and also measure substantially different properties. Comparisons between these techniques have to be done with utmost care. In Figure 2, we summarize different

Figure 1. Illustration to emphasize the need for structural changes in the hydration layer and hence the necessity of water molecules as “essential assistants” in the vicinity of a protein for its activity.

experimental techniques that are used to obtain information about the structure and dynamics of the PHL. NMR, magnetic resonance dispersion (MRD), nuclear Overhauser effect (NOE), quasi-elastic neutron scattering (QENS), and so forth measure the dynamics at a single-particle level, but the outcome is averaged over all of the water molecules. Consequently, these techniques overshadow the small fraction (∼10−20%) of slow water molecules inside of the PHL. On the contrary, dielectric relaxation, depolarized light scattering Published: October 5, 2017 4878

DOI: 10.1021/acs.jpclett.7b02324 J. Phys. Chem. Lett. 2017, 8, 4878−4882

Viewpoint

The Journal of Physical Chemistry Letters

Figure 2. Summary of the different experimental techniques that are employed to study the structure and dynamics of the PHL. Different techniques may give rise to different information regarding time scales of relaxation because of the nature of the response that they measure as outputs.

Our main point is that the different views that have emerged from different experiments need not be incompatible. In order to resolve the issues involved, several simulations have been carried out that measured the residence time and rotational correlation time of water molecules in the layer. It is worth mentioning that there is uncertainty regarding the thickness of the PHL. As proteins do not provide a homogeneous surface, the thickness of the water layer varies from one site to the other. The positions of the first minima in the radial distribution function of amino acid−water pairs are not the same as those for hydrophobic and hydrophilic sites. Fortunately, in computer simulations, one can easily employ a geometric cutoff to distinguish between the PHL and bulk. The results, thus obtained, seem to differ from protein to protein. In general, one finds a slow tail in the distribution of relaxation times of individual water molecules, but the tail seems to account for only ∼5−10% of the total water molecules in the PHL. Stating explicitly, the simulations seem to find less than 10% of the water molecules that can be termed as “slow”. However, the average was found to be about 2−5 times slower, in agreement with NMR experiments24 that prompted the conclusion that water in the PHL is largely bulk-like. What could then be the origin of the slow component observed in dielectric relaxation and TDFSS measurements? A resolution of the paradox was offered by simulations that performed several “thought experiments” where the amino acid side chains (AASCs) were frozen while allowing water molecules to remain mobile. This led to a substantial reduction in the slow component in the TDFSS time correlation function. A computer experiment that combines freezing of AASCs with mutation of the charged groups in the neighborhood of the probe reveals that a certain degree of slowness in TDFSS arises due to the slow water itself (even in the absence of AASC motion).28 This latter slow component is sensitive to the local polar environment and typically characterized by time constants in the 20−50 ps range.28 This is in complete agreement with the results obtained by Zewail et al.11,26 and clearly demonstrates the origin of the 20−40 ps range slow component observed in those TDFSS experiments. One is thus led to conclude that there are multiple origins of observed slow components. They could arise not only from a coupled motion between the amino acid side-chain fluctuations and water molecules but also from other slow molecules in the hydration layer.29,30 The origin of water−amino acid side-chain coupling can qualitatively be explained in the following way. Protein atoms in its native state and water molecules around it

(DLS), the optical Kerr effect (OKE), and far-infrared (FIR) spectroscopy measure a collective response (that is, the measured response is the sum over all of the water molecules, mostly without any spatial constraints). TDFSS,26 three-pulse photon echo peak shift (3PEPS),10 and two-dimensional infrared (2D-IR)27 act as partially local probe techniques. In the short time, TDFSS contains contributions from ultrafast bulk solvent modes, while in the long time capture, even the responses from the motions of the biomolecule are included.7,11,16,28. Thus, the present scenario resembles the famous tale of “six blind men and an elephant”, where those blind men only examined certain parts of the elephant and arrived at contradictory conclusions. In the language of the immortal poem by John Godfrey Saxe on “The Blind Men (from Indostan) and the Elephant” (see Figure 3), we quote only the last part of the poem in this context.

Figure 3. Stanza and representation of the poem The Blind Men and the Elephant by John Godfrey Saxe. The elephant can be thought of as a protein and those men as the various experimental probes that provide information that are true but partial. We, hence, need to combine all possible techniques in order to understand the real scenario. 4879

DOI: 10.1021/acs.jpclett.7b02324 J. Phys. Chem. Lett. 2017, 8, 4878−4882

Viewpoint

The Journal of Physical Chemistry Letters

Figure 4. Distribution of first-rank rotational time constants of water molecules that have residence times more than 100 ps inside of the PHL for four proteins, namely, (a) lysozyme, (b) myoglobin, (c) monellin, and (d) protein-G, and the bulk (blue). [Ci1(t) = ⟨P1(μ̂i0,μ̂it)⟩, where P1(x) = x and ∞ i μ̂it are the unit vectors along any one O−H bond vector of the ith water molecule at time t. Hence, τ(i) 1 = ∫ 0 C1(t) dt is the first-rank rotational time constant of the ith water molecule.]

the PHL and ignoring cross-terms with the outer layer) is about 2−3 times that of bulk water.31 Another important input is given by the total dipole moment fluctuation that leads to an effective local dielectric constant.31 We obtain a value of around 45, which is substantially reduced from the bulk value (68 for SPC/E water) but in general agreement with TDFSS estimates. The results that are depicted in Figure 4 are quite general, seen in all of the dynamical properties calculated (including translational diffusion and second-rank rotational time constants).31 On the basis of the above results, we offer an altered view of the PHL. The water molecules that inhabit the hydration layer exhibit both faster and slower than bulk dynamics. There are ∼5−10% of so-called slow water molecules with relaxation times that are 5−10 times slower in terms of mobility. However, there are faster water molecules with relaxation times smaller by even a factor of 2. These values can change depending on the protein. From the fastest to the slowest, we have a difference in relaxation time that varies by a factor of 10! In bulk water, this variation is just about 10%. It is not difficult to appreciate that this difference can be exploited in biological function. If we consider the interior of a cell, the average distance between macromolecules inside of cytoplasm is ∼1 nm. Therefore, the water molecules are always under the influence of several biomolecules and can exhibit uniqueness as mentioned earlier and cannot be entirely “bulklike”. The PHL, by virtue of its low dielectric constant, may allow proteins to interact among themselves and with substrates/ ligands. We can also anticipate that the larger width in energy distribution resulting in a larger specific heat can serve to stabilize the protein by reducing temperature fluctuation of the water around the protein.31 In addition to the water molecules that surround a protein, water molecules that reside inside of cavities or channels help in the activity. For example, we look into the case of the membrane protein bacteriorhodopsin that acts as a proton

both try to remain close to an energy-minimized configuration that resists change, resulting in slow water dynamics. As already mentioned, the nature of the protein side chains varies considerably from protein to protein. For example, lysozyme contains 11 arginine residues and 6 lysine residues, whereas protein-G has no arginine. However, the charge on arginine is distributed over three nitrogen atoms, thus lowering the hydrogen bond-accepting capability as compared to that of lysine. Still, it is seen to slow down the rotation of surrounding water molecules. The first layer of molecular water gets directly perturbed by the biomolecule via hydrogen bonding and electrostatic interactions. They can in turn perturb the subsequent layers. Therefore, both the structure and dynamics are altered. Hence, the search for universal and general features of the hydration water molecules remains elusive. We discuss this aspect in terms of recent findings below. In Figure 4, we have shown the distribution of first-rank rotational correlation time constants (τ1) of water molecules inside of the PHL of four proteins and compared it with that in the bulk.31 Note that the two are strikingly different. While the distribution is sharp and Gaussian in the bulk water, the same distribution is broad and log-normal in the PHL. The reason for the occurrence of the log-normal distribution can be understood as follows. The water molecules that are neighbors of the polypeptide backbone atoms are faster than the bulk, and the molecules near the charged atoms are slower than those in the bulk.32 In addition, the extent of slowness is determined by the energy barriers. The barriers themselves are from hydrogen bonding. The distribution of the strengths of the barriers can be Gaussian. These combinations can lead to a log-normal distribution of the relaxation times. This type of distribution persists even when side chains are immobile. Indeed, this broad distribution of water dynamic response is reflected in a larger than bulk value of the effective specific heat of the hydration layer. We find that the specific heat (calculated from self-energy fluctuation of all the water molecules inside 4880

DOI: 10.1021/acs.jpclett.7b02324 J. Phys. Chem. Lett. 2017, 8, 4878−4882

The Journal of Physical Chemistry Letters



pump. Here, internal water molecules form a “proton wire” that in turn helps in the proton transport across the membrane.33 This is essential for the conformational change in retinal that helps in vision. In earlier studies, it was noticed that the absence of water molecules from the ion channels of many transmembrane proteins prevents the transport of ions.34 The water molecules present in the insulin hexamer cavity stabilize it from the core by the formation of bridging hydrogen bonds across residues of different monomer units.35 The antifreeze proteins serve as another class of example where water molecules, as an active participant, bind to ice crystals in order to control nucleation and ice growth.36 These all provide examples of slow and active water molecules in the vicinity of proteins. According to a unified model of protein dynamics,37 the fluctuations of the protein that are essential for its activity are subject to the fluctuations of the hydration shell and the bulk solvent (Figure 1 provides an illustration of this depiction). It is clear from the above discussion that hydration water exhibits many unique static and dynamic properties. The heterogeneity of the biomolecular surface gives rise to dynamical heterogeneity in the PHL that results in the nearly universal broad log-normal distributions. Additionally, water molecules in the vicinity of proteins actively participate in its function. Hydration water can be seen as a lubricant of a protein’s conformational motions. However, these altered properties can be a mere manifestation of rather general phenomena of surface chemistry and electrostatic interactions.5 For example, intrinsically disordered proteins seem to indeed exhibit slower response of the water layer surrounding them. However, the Terran life evolved with water in its surroundings, and proteins are naturally adapted to it. The presence of water is still considered a sign of life in a barren planet. It is known that protein’s function and flexibility increase with the level of hydration. While there remain many open questions, we should humbly acknowledge the fact that, until now, we could not properly understand a unified picture that is hidden behind these puzzles. If we want to abandon the use of the term “biological water”, there is no other way to convey how unique and important water is for biomolecules and, of course, for life!

REFERENCES

(1) Bagchi, B. Water in Biological and Chemical Processes: From Structure and Dynamics to Function; Cambridge University Press, 2013. (2) Szent-Györgyi, A. Science 1956, 124, 873−875. (3) Ball, P. Water as an active constituent in cell biology. Chem. Rev. 2008, 108, 74−108. (4) Onuchic, J. N.; Wolynes, P. G. Theory of protein folding. Curr. Opin. Struct. Biol. 2004, 14, 70−75. (5) Laage, D.; Elsaesser, T.; Hynes, J. T. Water Dynamics in the Hydration Shells of Biomolecules. Chem. Rev. 2017, 117, 10694− 10725. (6) Bagchi, B. Water dynamics in the hydration layer around proteins and micelles. Chem. Rev. 2005, 105, 3197−3219. (7) Bhattacharyya, K. Solvation dynamics and proton transfer in supramolecular assemblies. Acc. Chem. Res. 2003, 36, 95−101. (8) Tarek, M.; Tobias, D. J. The dynamics of protein hydration water: a quantitative comparison of molecular dynamics simulations and neutron-scattering experiments. Biophys. J. 2000, 79, 3244−3257. (9) Ebbinghaus, S.; Kim, S. J.; Heyden, M.; Yu, X.; Heugen, U.; Gruebele, M.; Leitner, D. M.; Havenith, M. An extended dynamical hydration shell around proteins. Proc. Natl. Acad. Sci. U. S. A. 2007, 104, 20749−20752. (10) Jordanides, X. J.; Lang, M. J.; Song, X.; Fleming, G. R. Solvation dynamics in protein environments studied by photon echo spectroscopy. J. Phys. Chem. B 1999, 103, 7995−8005. (11) Pal, S. K.; Peon, J.; Bagchi, B.; Zewail, A. H. Biological water: femtosecond dynamics of macromolecular hydration. J. Phys. Chem. B 2002, 106, 12376−12395. (12) Qin, Y.; Jia, M.; Yang, J.; Wang, D.; Wang, L.; Xu, J.; Zhong, D. Molecular Origin of Ultrafast Water−Protein Coupled Interactions. J. Phys. Chem. Lett. 2016, 7, 4171−4177. (13) Halle, B. Protein hydration dynamics in solution: a critical survey. Philos. Trans. R. Soc., B 2004, 359, 1207−1224. (14) Halle, B.; Nilsson, L. Does the dynamic Stokes shift report on slow protein hydration dynamics? J. Phys. Chem. B 2009, 113, 8210− 8213. (15) Kuntz, I. D.; Kauzmann, W. Hydration of proteins and polypeptides. Adv. Protein Chem. 1974, 28, 239−345. (16) Bagchi, B. Molecular Relaxation in Liquids; Oxford University Press USA, 2012. (17) Robinson, R. A.; Stokes, R. H. Electrolyte Solutions; Courier Corporation, 2002. (18) Buck, M.; Boyd, J.; Redfield, C.; MacKenzie, D. A.; Jeenes, D. J.; Archer, D. B.; Dobson, C. M. Structural determinants of protein dynamics: analysis of 15N NMR relaxation measurements for mainchain and side-chain nuclei of hen egg white lysozyme. Biochemistry 1995, 34, 4041−4055. (19) Mukherjee, A.; Bagchi, B. Chem. Phys. Lett. 2005, 404, 409−413. (20) Pethig, R. Protein-water interactions determined by dielectric methods. Annu. Rev. Phys. Chem. 1992, 43, 177−205. (21) Grant, E. The dielectric method of investigating bound water in biological material: An appraisal of the technique. Bioelectromagnetics 1982, 3, 17−24. (22) Mashimo, S.; Kuwabara, S.; Yagihara, S.; Higasi, K. Dielectric relaxation time and structure of bound water in biological materials. J. Phys. Chem. 1987, 91, 6337−6338. (23) Wüthrich, K.; Billeter, M.; Güntert, P.; Luginbühl, P.; Riek, R.; Wider, G. NMR studies of the hydration of biological macromolecules. Faraday Discuss. 1996, 103, 245−253. (24) Mattea, C.; Qvist, J.; Halle, B. Dynamics at the protein-water interface from 17 O spin relaxation in deeply supercooled solutions. Biophys. J. 2008, 95, 2951−2963. (25) Zhang, L.; Wang, L.; Kao, Y.-T.; Qiu, W.; Yang, Y.; Okobiah, O.; Zhong, D. Mapping hydration dynamics around a protein surface. Proc. Natl. Acad. Sci. U. S. A. 2007, 104, 18461−18466. (26) Pal, S. K.; Peon, J.; Zewail, A. H. Biological water at the protein surface: dynamical solvation probed directly with femtosecond resolution. Proc. Natl. Acad. Sci. U. S. A. 2002, 99, 1763−1768.

Sayantan Mondal Saumyak Mukherjee Biman Bagchi*



Viewpoint

Solid State and Structural Chemistry Unit, Indian Institute of Science, Bengaluru, Karnataka 560 012, India

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Biman Bagchi: 0000-0002-7146-5994 Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS We thank the Department of Science and Technology (DST, India) for partial support of this work. B. Bagchi acknowledges a Sir J C Bose fellowship. S. Mondal thanks UGC, India for providing a research fellowship, and S. Mukherjee thanks DST, India for providing an INSPIRE fellowship. 4881

DOI: 10.1021/acs.jpclett.7b02324 J. Phys. Chem. Lett. 2017, 8, 4878−4882

Viewpoint

The Journal of Physical Chemistry Letters (27) King, J. T.; Arthur, E. J.; Brooks, C. L., III; Kubarych, K. J. Sitespecific hydration dynamics of globular proteins and the role of constrained water in solvent exchange with amphiphilic cosolvents. J. Phys. Chem. B 2012, 116, 5604−5611. (28) Mondal, S.; Mukherjee, S.; Bagchi, B. On the origin of multiple time scales in the relaxation of protein hydration layer: A molecular dynamics simulation study. arXiv preprint arXiv:1708.01210, 2017. (29) Mondal, S.; Mukherjee, S.; Bagchi, B. Decomposition of total solvation energy into core. side-chains and water contributions: Role of cross correlations and protein conformational fluctuations in dynamics of hydration layer. Chem. Phys. Lett. 2017, 683, 29−37. (30) Li, T.; Hassanali, A. A.; Kao, Y.-T.; Zhong, D.; Singer, S. J. Hydration dynamics and time scales of coupled water-protein fluctuations. J. Am. Chem. Soc. 2007, 129, 3376−3382. (31) Mukherjee, S.; Mondal, S.; Bagchi, B. Distinguishing dynamical features of protein hydration layer: Distributions reveal more than average. J. Chem. Phys. 2017, 147, 024901−1. (32) Jana, B.; Pal, S.; Bagchi, B. Hydration dynamics of protein molecules in aqueous solution: Unity among diversity. J. Chem. Sci. 2012, 124, 317−325. (33) Kandori, H. Role of internal water molecules in bacteriorhodopsin. Biochim. Biophys. Acta, Bioenerg. 2000, 1460, 177−191. (34) Anishkin, A.; Sukharev, S. Water dynamics and dewetting transitions in the small mechanosensitive channel MscS. Biophys. J. 2004, 86, 2883−2895. (35) Blundell, T.; Dodson, G.; Hodgkin, D.; Mercola, D. Insulin: The Structure in the Crystal and its Reflection in Chemistry and Biology by. Adv. Protein Chem. 1972, 26, 279−402. (36) Sun, T.; Lin, F.-H.; Campbell, R. L.; Allingham, J. S.; Davies, P. L. An antifreeze protein folds with an interior network of more than 400 semi-clathrate waters. Science 2014, 343, 795−798. (37) 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.

4882

DOI: 10.1021/acs.jpclett.7b02324 J. Phys. Chem. Lett. 2017, 8, 4878−4882