Dynamics of lysozyme and its hydration water under electric field
aa r X i v : . [ c ond - m a t . s o f t ] D ec Journal of Biological Physics manuscript No. (will be inserted by the editor)
P.M. Favi, · Q. Zhang, · H. O’Neill, · E.Mamontov, · S.O. Diallo
Dynamics of lysozyme and its hydration waterunder electric field
Received: July 12, 2018/ Accepted: date
Abstract
The effects of static electric field on the dynamics of lysozyme and its hydration water havebeen investigated by means of incoherent quasi-elastic neutron scattering (QENS). Measurements wereperformed on lysozyme samples, hydrated respectively with heavy water (D O) to capture the proteindynamics, and with light water (H O), to probe the dynamics of the hydration shell, in the temperaturerange from 210 < T <
260 K. The hydration fraction in both cases was about ∼ .
38 gram of waterper gram of dry protein. The field strengths investigated were respectively 0 kV/mm and 2 kV/mm( ∼ × V/m) for the protein hydrated with D O and 0 kV and 1 kV/mm for the H O-hydratedcounterpart. While the overall internal protons dynamics of the protein appears to be unaffected bythe application of electric field up to 2 kV/mm, likely due to the stronger intra-molecular interactions,there is also no appreciable quantitative enhancement of the diffusive dynamics of the hydration water,as would be anticipated based on our recent observations in water confined in silica pores under fieldvalues of 2 . E c ∼ − Keywords
Quasi-Elastic Neutron Scattering · Protein Dynamics · Electric Field · Diffusion
PACS · · Interactions of proteins with charged surfaces are important in many applications such as chromato-graphic separation [1], biosensors [2], and design of biocompatible medical implants [3]. Knowledge ofthe interactions involving charged, polar and polarizable groups, and the hydration water in proteins
P.M. FaviQuantum Condensed Matter Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831, USA
Permanent address:
Department of Materials Science and Engineering, University of Tennessee, Knoxville,Tennessee 37996, USAQ. Zhang and H. O’NeillBiology and Soft Matter Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831, USAE. MamontovChemical and Engineering Materials Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831,USAS.O. DialloQuantum Condensed Matter Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831, USAE-mail: [email protected] P.M. Favi, et al. is thus of fundamental interests because it provides microscopic insights into biophysical molecularrecognition and protein folding mechanism [4].Many opened questions remain regarding the exact way in which proteins respond to external stress,including electric field. Since the function of proteins is critically linked to their three-dimensionalstructures and to their hydration water, exposure to any form of stress, thermal or non, which mayinduce changes in conformation can alter cellular function. This is particularly relevant in today’senvironment, with the increasingly use of portable electromagnetic devices, which raises many questionsabout the possible effects on human health.There is strong evidence that protein misfolding is responsible for a number of known diseases suchas prion diseases, Alzheimer’s disease, and so on [4,5]. Investigating the effects of the different con-tributing factors (such as temperature, hydration, pressure, pH, ionic strength) on the conformationalchanges of proteins is thus key to understanding this complex mechanism. Application of an electricfield is such a parameter since the protein itself has an internal dipole moment. The force exerted onthe protein dipole will result in a torque that will rotate the protein and is likely to affect the diffusionmotions of the hydration water, which are known to be on the nano- to pico-second time scales [6].Charged and polar groups are also expected to move little with the application of electric field [7].Considerable progress has already been made in structural studies involving intermediate statesalong the folding pathways of some proteins under electric field and the effects on their crystallization[8,9,10]. This body of work is supported by several molecular simulation techniques [11,12,13,14].Similarly, but to a lesser extent, there has been a variety of spectroscopy techniques, principally H-D exchange NMR, circular dichroism, and fluorescence spectroscopy and protein engineering, thathave provided some information about the changing environments and configurations of individualresidues during the folding process [5,10]. Unfortunately, these measurements are not able to resolvethe spatial differences in the dynamics of the hydration water nor are they able to dissociate thedifferent type of motions involved. For instance; Are the molecular motions localized or long-ranged?Are they translational or rotational in nature? Such questions can uniquely be answered by quasi-elastic neutron scattering (QENS), thanks to the dependence of the observed relaxation times on thewavevector Q , which can not be accessed with other spectroscopy techniques such as NMR [15,16].We have applied the QENS technique to study of the effects of applied electric field on the diffusivemotion of hydration water in a model protein system, Hen-Egg-White-Lysozyme (HEWL), a residueprotein found in secretions (e.g., saliva, sweat, and mucus), well-known for its protective propertyagainst certain bacterial or virus aggressions. We find no effect of external electric field on the dynamicsof Lysozyme, probably due to to its stronger intra-molecular interactions [17]. Similarly, but in contrastto our recent observations in water confined in silica pores under field values of 2 . O and another with D O [20]. The samples were lyophilized repeatedlybefore being hydrated. In both cases, the labile hydrogen atoms were exchanged for deuterium atoms bydissolving in heavy water (D O), prior to lyophilization. The samples were thus hydrated using isopi-estic conditions by incubation in a sealed container containing respectively 99.9% of H O and 99.9% ofD O. The level of hydration was controlled by varying the incubation time. The final hydration level h was determined by the relative change in the sample weight following humidity exposure, yieldingan h ≃
38% for each sample. Neutron-scattering measurements were performed on the backscatteringsilicon spectrometer (BASIS) at the 1 MW Spallation Neutron Source, Oak Ridge National Laboratory(ORNL), USA [21].Each protein sample was mounted between two Al plates separated with a Teflon gasket (with ∼ ynamics of lysozyme and its hydration water under electric field 3 Fig. 1
Temperature dependence of the elastically scattered neutron intensity obtained from lysozyme hydratedwith D O (left panel) and with H O (right panel), at selected Q values of 0.3 ˚A − and 1.1 ˚A − in the absence offield (E= 0 kV/mm). The stronger suppression in intensity in the H O hydrated protein arises from the largermean-squared proton displacement h r ( T ) i in the Debye-Waller factor compared to the D O hydrated sample.The associated uncertainties on the data points are around 6-8%, and are much smaller than the symbols. S ( Q , E ) (or DSF), which is related to the probability for an incident neutron to be scatteredwith a wavevector transfer Q and an energy transfer E to the sample. The E and Q -dependence of S ( Q , E ) provides information on the characteristic correlation times ( t ∼ ¯ h/E ) and on the geometry( r ∼ /Q ) of the molecular motions within the sample, respectively. For an isotropic system, as is thecase here, the DSF allows for a ‘powder averaging’ of the signal, and consequently it depends effectivelyon the magnitude Q of the wavevector transfer rather than on the vector Q . In most cases, the DSFcontains both coherent and incoherent scattering contributions, arising from pair- and self-correlations, P.M. Favi, et al.
Fig. 2
Superposition of the QENS response obtained from the protein hydrated with H O (black solid circles)and D O (red solid circles), at temperature T =240 K and Q =0.9 ˚A − . The larger broadening in the H O-protein sample is due to the hydration shell. The corresponding fits are shown as solid blue lines and theinstrument resolution (measured at 50 K using the exact same sample) is shown in dashed line for comparison. respectively. However, and thanks to the large incoherent cross-section of hydrogen over that of deu-terium (and all other elements), it is possible to mask the dynamics in part of the sample with selectivedeuteration. In this event, the dynamics seen by the neutron comes largely from the incoherent part,yielding information on self-diffusion processes.To probe such dynamics, which are typically in the pico-to-nano second regime, a state-of-the-arthigh energy resolution neutron scattering instrument such as BASIS [21] is required. The wide accessibledynamic range ∆ E= ± µ eV combined with the excellent energy resolution of 3.5 µ eV (Full-Widthat Half Maximum or FWHM) at the elastic position makes BASIS an ideal spectrometer for probingthe pico-nano second dynamics in lysozyme and its hydration water. The Q -range investigated herevaries from 0.3 to 1.1 ˚A − in step of ∆Q =0.2 ˚A − . Incoherent elastic signal
For diagnostic purposes, we perform rapid standard ‘elastic’ scans on bothlysozyme samples hydrated with H O and D O. The aim was to check the quality of the neutron signalfrom the samples, and to provide a calibrated temperature range for the subsequent QENS measure-ments, which requires high counting times. Data were collected with 10 K temperature increments oncooling from 270 K to 210 K. Fig. 1 shows the normalized elastic intensity (with respect to the max-imum lowest temperature value I ( T )) as a function of temperature for the D O and H O hydrated ynamics of lysozyme and its hydration water under electric field 5
Fig. 3
Field dependence of the QENS response obtained on the H O-hydrated protein at temperature T =230K and wavevector Q =0.9 ˚A − . The black solid symbols represent the 0 kV/mm applied field data, and thered solid circles the 1 kV/mm data. The corresponding fits are shown as solid blue lines and the instrumentresolution is shown as a dashed line for comparison. proteins, at two selected Q values, lowest and highest Q values investigated. The elastic intensity foreach temperature was obtained by integrating the corresponding spectrum over a very narrow energyrange of ± µ eV, corresponding to the elastic resolution. For an isotropic powder sample, the elasticintensity is expected to have a Debye-Waller behavior, I ( T ) ∼ e − Q h r ( T ) i / , where h r ( T ) i is the meansquare amplitude vibration of the molecule. As the sample cools down, the molecular diffusion start toslow down and h r ( T ) i decreases, yielding an increase in the elastic line. The elastic intensity withinthe 3.5 µ eV energy resolution effectively increases with decreasing temperature but never reaches amaximum plateau region down to the lowest temperature investigated, suggesting that both the pro-tein and its hydration water are still mobile below 220 K. The stronger suppression in intensity in theprotein hydrated with H O with increasing temperature arises from the larger mean-squared protondisplacement h r ( T ) i in the Debye-Waller factor of the hydration water in the H O hydrated protein.
Quasi-Elastic Neutron Scattering
Generally, the observed incoherent DSF is a convolution of the trans-lational DSF and the rotational one. By using only the spectra at low Q values where rotations aregenerally not observed, (for water molecules on BASIS, generally Q ≤ − ) the rotational contribu-tions can be conveniently neglected. The QENS data were thus investigated for wavevector transfers Q ,0.3 ≤ Q ≤ − , in steps of ∆Q = 0.2 ˚A − , within the same temperatures range as the elastic scans P.M. Favi, et al. above. The measurements were performed with and without the application of external electric field.QENS Measurements were initially taken on the sample hydrated with D O at field values of E = 0and 2 kV/mm, followed by the measurements on the H O-hydrated protein sample at E = 0 and 1kV/mm. These field strengths are the maximum achievable values below which electrical breakdowndoes not occur within our experimental set-up. The likelihood of such an event was carefully monitoredwith an oscilloscope via a simultaneous measurement of the current and voltage across the sample. To analyze the data, we begin by qualitatively comparing the different spectra collected on the samplehydrated with H O with that hydrated with D O to assess the relative contribution of the hydrationwater with respect to that of the protein. Fig. 2 shows the observed QENS spectra at temperature T = 240 K at selected Q = 0 . − in the absence of field. The overlaid solid lines are model fits, asdescribed below. The dashed line is the instrument resolution taken with the H O-hydrated sample at T = 50 K, where all relevant molecular motions are frozen out. There is clearly an appreciable excessbroadening of the QENS signal in the protein hydrated with H O compared with that hydrated withD O, indicating a discernible dynamics quite different to that of the protein itself.3.1 Proteins dynamicsThe internal dynamics of protein is complex, and not well captured by ‘standard’ Lorentzian line fitsused in QENS analysis. Rather, a stretched exponential function in the form e ( − tτ ) β , with no singleactivation energy generally works best [16]. It is also common to study the ‘mean’ square amplitudevibration of the protons inside the proteins, which comes to a large extent from the side groups(hydrogen atoms on the methyl groups for example). In the present case, this is reflected in themonotonic reduction of the elastic intensity with increasing temperature, as thermal fluctuations affect h r ( T ) i . We would like to emphasize here that our primary objective is to study the effects of thestatic electric field on the hydration water. A qualitative conclusion regarding the protein dynamicscan be drawn by visually inspecting the difference in the response from the D O hydrated samplewhen the field is applied and when it is not. Our main observation is that there is no obvious differencebetween the two, within the experimental precision. We conclude that the application of field (up tothe maximum field applied here; i.e., ∼ V/m ( ∼
100 kV/mm).3.2 Hydration waterTo uniquely characterize the dynamics of the hydration water (and exclude the protein contributions),we subtracted the spectra of the H O-hydrated sample ( S L H ( Q, E )) from that of the D O hydratedprotein ( S L D ( Q, E )) using the correct mass ratio and the relative sample transmissions, followingmethods described elsewhere [22,23]. The resulting spectra represent the net signal from the hydrationwater, as discussed below. Based on the relative signal strength between the two samples, as highlightedfor example in Fig.2, we expect the overall signal obtained on the sample hydrated with H O to bedominated by an ‘average’ broad Lorentzian term. Another Lorentzian function, somewhat narrower,was required to fully reproduce the observed QENS spectra. Within the Q -range investigated, themodel using a double Lorentzian model captured the signal from the hydration shell data reasonablywell, in agreement with previous arguments [24] and findings [18]. The dynamical structure factor forthe hydration water S wat ( Q, E ) is thus approximated by, S wat ( Q, E ) = S L H ( Q, E ) − ηS L D ( Q, E ) (1)= (1 − p ( Q )) 1 π Γ ( Q ) Γ ( Q ) + E + p ( Q ) 1 π Γ ( Q ) Γ ( Q ) + E ynamics of lysozyme and its hydration water under electric field 7 Fig. 4
Fit parameters characterizing the dynamics of the hydration water. Panel on the left summarizes theparameters with no field and the panel on the right the results when the field is applied. The parameter p ( Q )denotes the overall elastic fraction arising from all immobile atoms, as seen by the spectrometer window, Γ the broad Lorentzian width associated with the fast molecular dynamics, Γ the narrow Lorentzian widthassociated with the slower molecular dynamics, and p ( Q ) is the relative weight of the narrow Lorentzian, asdescribed in the text. The relatively low quality of the fits reflects the limited statistics of our QENS data P.M. Favi, et al. where p is the relative weight of the narrow component, Γ , are the HWHMs associated with thedynamics of the hydration shell, η is the ratio between the protein mass contained in the H O hydratedsample and that contained in the D O-hydrated sample scaled to the relative neutron transmissionfactors ( η ≃ I ( Q, E ) is, I ( Q, E ) = A ( Q )[ p ( Q ) δ ( E ) + (1 − p ( Q )) S wat ( Q, E )] O R ( Q, E ) + B ( Q, E ) (2)
Here R ( Q, E ) is the instrument resolution, measured with the H O-hydrated sample at temperatureof 50 K where all observable dynamics within our sample are frozen out. The weight p is the fractionof water molecules elastically scattered by the neutrons plus those that appear to be ‘immobile’ on ourspectrometer time window. The term B ( Q, E ) = a + bE is a small correction required to account for theresidual function left after the subtraction in Eq. 2. The protein-water samples are likely to lead to un-wanted multiple scattering effects, which are non-trivial to account for. The subtraction method above,when properly done [22], yields a net hydration water signal with negligible contribution from multiplescattering. Any residual multiple scattering effect would be buried under the small Q -independent term a in the function B ( Q, E ).We note that the convolution in Eq. 2 leads to fitted lines with statistics limited by those of themeasured resolution function. This effect tends to be more apparent in the background regions, as canbe observed in Figs. 2 and 3. While the data can be smoothed to improve visualization, we chose topresent our data using the native binning of BASIS ( ∆ E=0.4 µ eV).We attribute the broader of the two Lorentzians to the ‘caged’ motion of water molecules (fasttransient motion inside a molecular cage), and the narrow component to the ‘cage-breaking’ watermolecules which diffuse comparatively slower [18,24]. Recent compelling arguments for using Eq. 2 todescribe the dynamics of confined water can be found in Ref. [24]. To illustrate the quality of the fitsobtained with Eqs. 2 and 2, we show (as an example) the fits obtained at selected temperature and Q values, as solid lines in Fig. 2 and Fig.3. A summary of the temperature dependence, as well asthe wave-vector dependence of the fit parameters is shown in Fig. 4. The observed Γ , ( Q ) at eachtemperature and field value were fit using the following jump diffusion expression, Γ , ( Q ) = ¯ h D , Q D , Q τ , (3)to determine the average residence time τ between conformation jumps, and D the correspondingdiffusion coefficients. The subscript denotes whether it is the fast (1) or slow dynamics (2). Fig. 4.From the D = h r i / τ , it is possible to estimate the mean squared diffusion jump length h r i . To agood approximation, the diffusion coefficient D is behaves as ¯ hQ at low Q , and as ¯ h/τ at the higher Q values. The fit results are summarized in Table 1.In summary, the present QENS study reveals two type of motions for the hydration water (a fastand slow motion). Both processes are very much temperature dependent, increasing with increasingtemperature with an Arrhenius behavior. To quantify this temperature evolution, we first attemptedto determine the diffusion parameter D at each temperature T . To a first approximation, D can beobtained from a series expansion of Eq. 3 around Q ≃
0, a limit in which Γ ≃ ¯ hDQ , as explained above.Unfortunately, D is simply not a well determined fit parameter due to significant small angle scatteringsignal at low Q . This limitation is not specific to our system, but quite commonly encountered withpowder samples. The parameter τ on the other hand is a well determined quantity in the high Q limit.Moreover because the two identified dynamics processes are well separated in time scales (1-2 orderof magnitude), the corresponding relaxation times ( τ and τ ) are clearly distinguishable on BASIS,as depicted in Fig. 5. Overall however, the water relaxation processes are largely unaffected by theapplication of external field (up to the maximum achieved 1 kV/mm value). The observed relaxationtime versus temperature are summarized in Fig. 5, for each field condition (on or off). Bulk waterdata from Ref. [6] are shown for comparison. The relaxation times of the hydration layer dynamics,as observed in the H O-hydrated protein are larger than in the bulk at high temperatures (certainlyfor
T > K ), consistent with the known idea that confinement tends to suppress dynamics. Below250 K, the relaxation times for the faster component starts to overlap with that of the bulk liquid,indicating dynamics with comparable time scales . Fig. 5 conveys a key message; that is the appliedfield of 1 kV/mm does not alter the relaxation times of the adsorbed water in any meaningful way, in ynamics of lysozyme and its hydration water under electric field 9 Fig. 5 (Color) Temperature behavior of the average residence time τ of water surrounding lysozyme with(closed black symbols) and without field (open red symbols). The two types of observed relaxations (fast andslow) exhibit both an Arrhenius temperature behavior (i.e. e − EaRT ). No significant effect of the field is observed.Bulk water data (open blue squares) are shown for comparison [6]. The inset compares the τ of the fastcomponent observed here (open and solid squares) with those of water adsorbed in tight silica pores (filleddotted circles), as reported in Ref. [18].0 P.M. Favi, et al. Table 1
Temperature dependence of the observed relaxation time τ , with and without field, obtained fromthe hydration water. The subscripts 1 and 2 refer to the fast and slow diffusion processes associated with thehydration shell, respectively. T (K) E (kV/mm) τ (ps) τ (ps)220 0 38 261230 0 24.5 1771 26.5 239240 0 19.5 1621 20 194250 0 19.2 171270 0 18.7 131 contrast to our recent observations in water confined in silica pores [18]. The inset figure compares thepresent observed fast relaxation time τ (open and solid squares) with those of water adsorbed in tightsilica pores [18]. We can not definitely rule out the possibility of field-induced enhanced diffusion inproteins because the maximum field achieved in the present study is just 40% of that reached in thehydrated silica experiment. We note here that the electrical breakdown which sets this maximum fieldis highly sensitive to factors such as cell geometry and assembly, sample dielectric strength, defects,hydration, surrounding pressure and so on. A small change in any of these factors can have significantimpact on the maximum attainable field. We found in our case that no matter how meticulously weprepare our sample assembly, we always find that the protein+water sample breaks down at a muchlower field strength (around 1 kV/mm) than in the silica case, at comparable hydration level. This isclearly due to a lower dielectric strength of our sample compared to the silica+water system.To test the universality of an electric field induced enhanced water diffusion, and to confirm theexistence of an onset field value E c , further work (both MD and experiments) are needed. We anticipatehowever the intrinsic E c of confined water to be in the range 2-3 kV/mm, in concordance with ourobservations in hydrated silica. Field dependence studies of the QENS signal of water adsorbed inother less polarizable proteins (such as Myelin basic protein), would be key in determining the onsetfield value E c . A key scientific goal is to clarify the mechanism by which this enhanced diffusion takesplace at the molecular level, and how it is modified by the substrate interaction. Recent analyticaltheory and calculations have investigated the dipolar response in various hydrated proteins [25]. Inthose studies, which included lysozyme, ubiquitin, and cytochrome C and B , Matyushov found aremarkable variation of the dielectric constant between the different proteins. Of particular relevanceto our work, it also indicates a strong influence of the coupling of the protein charge surface to thehydration water on the protein overall dipolar response. The protein-water dipolar correlations extendto large distances in all non-neutral proteins and cannot be neglected when evaluating the total dipolarresponse. In ubiquitin, the only neutral protein of all, the protein self-correlations nearly cancel outthe protein-water correlations. We thus anticipate the net effect of field on hydrated ubiquitin to comelargely from the hydration layer. This makes ubiquitin an excellent substrate candidate for futuremeasurements, with the caveats that this protein can be deuterated to yield an observable QENSsignal of the hydration water. We have investigated the impact of static electric field on the dynamics of lysozyme and its hydra-tion water. Our aim was to probe the lysozyme response, a highly polar molecule, and to test theuniversality of the enhanced diffusivity of water under field, observed with silica substrate [18]. Ourmeasurements reveal that the nano- to pico-second dynamics of the protein are unaffected by the field,due possibly to the stronger intra-molecular interactions compared to the maximum achieved fieldstrength of 1 kV/mm. There is also no appreciable quantitative enhancement of the diffusive dynamics ynamics of lysozyme and its hydration water under electric field 11 of the hydration water, as compared to our observations with water in silica pores in which a field of2 . E c ∼ − Acknowledgements
We acknowledge the use of the DAVE software in part of the data analysis [32]. Wethank C. Stanley at ORNL for stimulating discussions. It is also pleasure to acknowledge R. Goyette, R. Mills,D. Maierhafer, R. Moody, and M. Loguillo at SNS for valuable technical support. PF acknowledges the GEMfellowship program at UTK. HON and QZ acknowledge the support of the Center for Structural MolecularBiology at ORNL supported by the U.S. DOE, Office of Science, Office of Biological and EnvironmentalResearch Project ERKP291. Work at ORNL and SNS is sponsored by the Scientific User Facilities Division,Office of Basic Energy Sciences, U.S. DOE.
References
1. Roth, C., Lenhoff, A.: Quantitative Modelling of Protein Adsorption. In: M. Malmsten (ed.) Biopolymersat Interfaces,
Surfactant Science Series , vol. 75, pp. 89–118. Marcel Dekker, New-York (1998)2. Talasaz, A.A.H., Nemat-Gorgani, M., Liu, Y., Stahl, P., Dutton, R.W., Ronaghi, M., Davis, R.W.: Predic-tion of protein orientation upon immobilization on biological and nonbiological surfaces. Proc. Nat. Acad.Sci. U.S.A. (40), 14,773 (2006)3. Subrahmanyam, S., Piletsky, S.A., Turner, A.P.F.: Application of natural receptors in sensors and assays.Anal. Chem. (16), 3942–3951 (2002)4. Ojeda-May, P., Garcia, M.E.: Electric field-driven disruption of a native β -sheet protein conformation andgeneration of a helix-structure. Biophys. J. (2), 595–599 (2010)5. Lange, A., Gattin, Z., Van Melckebeke, H., Wasmer, C., Soragni, A., van Gunsteren, W.F., Meier, B.H.:A combined solid-state nmr and md characterization of the stability and dynamics of the het-s(218-289)prion in its amyloid conformation. Chem. Bio. Chem. , 1657–1665 (2009)6. Teixeira, J., Bellissent-Funel, M. C., Chen, S. H., Dianoux, A. J.: Experimental-Determination of theNature of Diffusive Motions of Water-Molecules. Phys. Rev. A , 1913 (1985)7. Xu, D., Phillips, J.C., Schulten, K.: Protein response to external electric fields: Relaxation, hysteresis, andecho. J. Phys. Chem. , 12,108–12,112 (1996)8. Taleb, M., Didierjean, C., Jelsch, C., Mangeot, J., Aubry, A.: Equilibrium kinetics of lysozyme crystalliza-tion under an external electric field. J. Crys. Growth , 250–255 (2001)9. Nanev, C.N., Penkova, A.: Nucleation of lysozyme crystals under external electric and ultrasonic fields. J.Crys. Growth , 285–293 (2001)10. Pompa, P.P., Bramanti, A., Maruccio, G., Cingolani, R., Rienzo, F.D., Corni, S., Felice, R.D., Rinaldi, R.:Retention of nativelike conformation by proteins embedded in high external electric fields. J. Chem. Phys. , 181,102 (2005)11. Marklund, E.G., Larsson, D.S.D., van der Spoel, D., Patriksson, A., Caleman, C.: Structural stabilityof electrosprayed proteins: temperature and hydration effects. Phys. Chem. Chem. Phys. , 8069–8078(2009)12. Abrikossov, A.: Computer simulations: Orientation of Lysozyme in vacuum under the influence of anelectric field. Student Thesis, Uppsala Universitet (2011)13. Sun, D.P., Liao, D.I., Remington, S.J.: Electrostatic fields in the active sites of lysozymes. Proc. Nat.Acad. Sci. U.S.A. (14), 5361–5365 (1989)14. Budi, A., Legge, F.S., Treutlein, H., Yarovsky, I.: Effect of frequency on insulin response to electric fieldstress. The Journal of Physical Chemistry B (20), 5748–5756 (2007)15. Varga, B: Study of solvent-protein coupling effects by neutron scattering. J. Bio. Phys. , 207 (2010)16. Roh, J., Curtis, J., Azzam, S., Novikov, V., Peral, I., Chowdhuri, Z., Gregory, R., Sokolov, A.: Influenceof hydration on the dynamics of lysozyme. Biophys. J. (7), 2573 – 2588 (2006)2 P.M. Favi, et al.17. Bon, C., Dianoux, A., Ferrand, M., Lehmann, M.: A model for water motion in crystals of lysozyme basedon an incoherent quasielastic neutron-scattering study. Biophys. J. (3), 1578–1588 (2002)18. Diallo, S.O., Mamontov, E., Wada, N., Inagaki, S., Fukushima, Y.: Enhanced translational diffusion ofconfined water under electric field. Phys. Rev. E , 021,506 (2012)19. Vegiri, A.: Translational dynamics of a cold water cluster in the presence of an external uniform electricfield. J. Chem. Phys. , 8786 (2002)20. Mamontov, E., O’Neill, H., Zhang, Q.: Mean-squared atomic displacements in hydrated lysozyme, nativeand denatured. J. Bio. Phys. (3) (2010)21. Mamontov, E., Herwig, K.: A time-of-flight backscattering spectrometer at the spallation neutron source,basis. Rev. Sci. Inst. , 085,109 (2011)22. Orecchini, A., Paciaroni, A., De Francesco, A., Sani, L., Marconi, M., Laloni, A., Guarini E., Formisano, F.,Petrillo, C., Sacchetti F.: Brillouin spectroscopy of protein hydration water: new experimental potentialitiesopened up by the thermal neutron spectrometer brisp. Meas. Sci. and Tech. (3), 034,026 (2008)23. Orecchini, A., Paciaroni, A., De Francesco, A., Petrillo, C., Sacchetti, F.: Collective dynamics of proteinhydration water by brillouin neutron spectroscopy. J. Am. Chem. Soc. (13), 4664 (2009)24. Qvist, J., Schober, H., Halle, B.: Structural dynamics of supercooled water from quasielastic neutronscattering and molecular simulations. J. Chem. Phys. , 144,508 (2011)25. Matyushov, D.V.: Dipolar response of hydrated proteins. J. Chem. Phys. (8), 085102 (2012)26. Liu, L., Chen, S.H., Faraone, A., Yen, C., Mou, C.Y.: Pressure dependence of fragile-to-strong transitionand a possible second critical point in supercooled confined water. Phys. Rev. Lett. , 117,802 (2005)27. Mamontov, E.: Observation of Fragile-to-Strong Liquid Transition in Surface Water in CeO . J. Chem.Phys. , 1711,011 (2005)28. Liu, L., Chen, S-H, Faraone, A., Yen, C-W. , Mou, C-Y., Kolesnikov, A.I., Mamontov, E., Leao, J.:Quasielastic and inelastic neutron scattering investigation of fragile-to-strong crossover in deeply super-cooled water confined in nanoporous silica matrices. J. Phys. Condens. Mat. , S2261–S2284 (2006)29. Doster, W., Busch, S., Gaspar, A. M., Appavou, M.-S., Wuttke, J., Scheer, H.: Dynamical transition ofprotein-hydration water. Phys. Rev. Lett. , 098,101 (2010)30. Hong, L., Smolin, N., Lindner, B., Sokolov, A.P., Smith, J.C.: Three classes of motion in the dynamicneutron-scattering susceptibility of a globular protein. Phys. Rev. Lett. , 148,102 (2011)31. Khodadadi, S., Curtis, J.E., Sokolov, A.P.: Nanosecond relaxation dynamics of hydrated proteins: Waterversus protein contributions. J. Phys. Chem. B (19), 6222–6226 (2011)32. Azuah, R.T., Kneller, L.R., Qiu, Y., Tregenna-Piggott, P.L.W., Brown, C.M., Copley, J.R.D., Dimeo, R.M.:J. Res. Natl. Inst. Stan. Technol.114