Intermolecular interactions in highly concentrated protein solutions upon compression and the role of the solvent S. Grobelny, M. Erlkamp, J. Möller, M. Tolan, and R. Winter Citation: The Journal of Chemical Physics 141, 22D506 (2014); doi: 10.1063/1.4895542 View online: http://dx.doi.org/10.1063/1.4895542 View Table of Contents: http://scitation.aip.org/content/aip/journal/jcp/141/22?ver=pdfcov Published by the AIP Publishing Articles you may be interested in Accurate small and wide angle x-ray scattering profiles from atomic models of proteins and nucleic acids J. Chem. Phys. 141, 22D508 (2014); 10.1063/1.4896220 Molecular near-field antenna effect in resonance hyper-Raman scattering: Intermolecular vibronic intensity borrowing of solvent from solute through dipole-dipole and dipole-quadrupole interactions J. Chem. Phys. 140, 204506 (2014); 10.1063/1.4879058 Theoretical study of interactions of BSA protein in a NaCl aqueous solution J. Chem. Phys. 138, 115103 (2013); 10.1063/1.4794919 Observation of electric-field induced aggregation in crystallizing protein solutions by forward light scattering Appl. Phys. Lett. 99, 153701 (2011); 10.1063/1.3648114 Forward light scattering for highly sensitive detection of aggregation in crystallizing protein solutions Appl. Phys. Lett. 98, 263701 (2011); 10.1063/1.3603932
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 128.248.155.225 On: Fri, 21 Nov 2014 19:31:34
THE JOURNAL OF CHEMICAL PHYSICS 141, 22D506 (2014)
Intermolecular interactions in highly concentrated protein solutions upon compression and the role of the solvent S. Grobelny,1 M. Erlkamp,1 J. Möller,2 M. Tolan,2 and R. Winter1 1
Faculty of Chemistry, Physical Chemistry-Biophysical Chemistry, TU Dortmund, Otto-Hahn Str. 6, 44227 Dortmund, Germany 2 Fakultät Physik/DELTA, TU Dortmund, Maria-Goeppert-Mayer-Str. 2, 44227 Dortmund, Germany
(Received 30 April 2014; accepted 9 June 2014; published online 23 September 2014) The influence of high hydrostatic pressure on the structure and protein-protein interaction potential of highly concentrated lysozyme solutions up to about 370 mg ml−1 was studied and analyzed using small-angle X-ray scattering in combination with a liquid-state theoretical approach. In the concentration region below 200 mg ml−1 , the interaction parameters of lysozyme solutions are affected by pressure in a nonlinear way, which is probably due to significant changes in the structural properties of bulk water, i.e., due to a solvent-mediated effect. Conversely, for higher concentrated protein solutions, where hydration layers below ∼4 water molecules are reached, the interaction potential turns rather insensitive to compression. The onset of transient (dynamic) clustering is envisaged in this concentration range. Our results also show that pressure suppresses protein nucleation, aggregation and finally crystallization in supersaturated condensed protein solutions. These findings are of importance for controlling and fine-tuning protein crystallization. Moreover, these results are also important for understanding the high stability of highly concentrated protein solutions (as they occur intracellularly) in organisms thriving under hydrostatic pressure conditions such as in the deep sea, where pressures up to the kbar-level are reached. © 2014 AIP Publishing LLC. [http://dx.doi.org/10.1063/1.4895542] I. INTRODUCTION
The investigation of concentrated protein solutions and protein condensation phenomena is not only a challenging subject in liquid-state theory of soft condensed matter systems, a detailed understanding of the properties of such systems is also the prerequisite for a number of biological and biotechnological applications. From a biological point of view, there has been an increased interest in studies of the behavior of proteins in highly crowded solutions in order to mimic intracellular environments. In biological cells, crowding and confinement are able to cause major alterations of the conformational stability and biological activity of proteins.1–4 Knowledge of protein-protein interactions in condensed fluid phases is also crucial for understanding the functional and structural stability of proteins and to yield molecular insights into processes such as protein crystallization, aggregation, and fibrillation.5–11 The latter involves diseases that occur due to undesired protein nucleation and aggregate formation. Classic examples are formation of polymer fibers of sickle hemoglobin molecules within the red blood cells that produces sickle cell anemia, age-related cataracts produced by the undesired aggregation of γ -crystallin, or the formation of amyloid fibres via cross-β-sheet formation in a series of debilitating diseases such as Alzheimer’s, Parkinson’s, CreutzfeldJakob disease, or diabetes mellitus type 2.12 To combat such diseases, strategies for preventing nucleation of protein fibers must be developed. The structural properties of highly concentrated protein solutions and the phase diagram of proteins depend on a num-
0021-9606/2014/141(22)/22D506/8/$30.00
ber of control parameters, such as temperature, pressure, pH, ionic strength, and the presence of osmolytes. This parameter space and the underlying intermolecular interaction potential of the proteins as a function of these parameters are hardly explored today.13, 15–29 The use of the pressure variable is least understood so far and is hence in the focus of this study. Here we report the pressure dependence of the interaction potential of concentrated lysozyme solutions over a wide concentration range, up to about 370 mg ml−1 , including concentrations where protein crystallization sets in. Figure 1 exhibits the temperature-protein concentration dependent phase diagram of lysozyme at corresponding solution conditions as well as the temperature-protein concentration conditions selected in our study. Conditions were chosen such that the protein crystallization kinetics is still very slow, allowing us to explore protein-protein interactions in the highly condensed fluid phase which mimics also conditions encountered intracellularly. Large supersaturation is required to overcome the activation energy barrier for crystal formation of lysozyme. In the absence of crystal nuclei, depending on the protein concentration, it can take more than a month to reach thermal equilibrium.6, 30 The appearance of Bragg reflections due to the onset of crystallization has been observed for the highest protein concentration chosen in our study, only. In order to determine the structure and intermolecular interactions of such highly concentrated protein solutions, small-angle X-ray scattering (SAXS) experiments have been carried out in the pressure range from 1 bar to 3 kbar. The results are of interest for exploring the effects of pressure on the phase behavior of protein solutions, including self-crowding
141, 22D506-1
© 2014 AIP Publishing LLC
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 128.248.155.225 On: Fri, 21 Nov 2014 19:31:34
22D506-2
Grobelny et al.
FIG. 1. Temperature-concentration phase diagram of lysozyme. (Red circles) Experimental conditions chosen in our study. Liquid–liquid phase coexistence according to Taratuta et al.14 The solubility curve was calculated according to Muschol et al.15
conditions similar to conditions met in a biological cell. Investigating pressure effects is also of relevance for understanding the hydration behavior of biological matter under extreme environmental conditions, such as in the deep sea where pressures up to the kbar-level (1 kbar = 100 MPa) and beyond are encountered.31–34 II. MATERIALS AND METHODS
Lysozyme is a small enzyme that protects humans from bacterial infection. It attacks the protective cell walls of bacteria, breaking the carbohydrate chains in the walls, thereby destroying the structural integrity of the bacteria’s cell walls. It is a tightly packed globular protein that consists of 129 amino acid residues. It has a molecular weight of about 14 kDa and an isoelectric point at pI = 11.16. The protein is an ideal candidate for investigations of its phase diagram, nucleation, and crystal growth kinetics, as its equilibrium properties are well established under ambient conditions.12 We used hen egg white lysozyme, which was purchased from Sigma Aldrich and used without further purification. For the highest protein concentrations, samples were prepared in dialyR , MWCO = 2 kDa, Thermo Sciensis bags (Slyde-A-Lyzer tific, Rockford, IL, USA) applying osmotic pressure over different timescales at 20 ◦ C under constant stirring, which is a mild method to prepare also highly supersaturated metastable protein solutions.35 Polyethyleneglycol (PEG, Sigma Aldrich, Schnelldorf, Germany) with a molar mass of 35 kDa was used as main component of the dialysis buffer, which was adjusted to pH 7. The initial concentration was amounted to 80 mg ml−1 . As buffer solution, Bis-Tris (Sigma Aldrich, Schnelldorf, Germany) at a concentration of 50 mM was used, which is a pressure stable buffer over a wide pressure range.36 This relatively high buffer concentration was needed to keep the pH value constant at 7 even at the highest protein concentrations used. The final protein concentration was determined by UV-Vis spectroscopy using an extinction coefficient of 2.64 ml mg−1 cm−1 .37 The small-angle X-ray Scattering (SAXS) experiments were carried out at the beamline I22 at Diamond Light Source (Didcot, Oxfordshire, UK) with a custom-made high pressure
J. Chem. Phys. 141, 22D506 (2014)
cell.38 The X-ray energy was 18 keV, which corresponds to a wavelength, l, of 0.096 nm. The scattering curves were recorded with a PILATUS 2M detector at a sample to detector distance of 5.2 m, thereby covering a range of momentum transfer, q, from 0.2 to 4.5 nm−1 . The exposure time was 15 s. A total of 20 μl of the sample was filled into the pressure cell. The time for thermal equilibration before each measurement was set to 15 min. High hydrostatic pressures up to 3 kbar – with an accuracy of 0.5 % – at steps of 250 bars were applied at a temperature of 20 ◦ C, using water as pressurising medium. Temperature control (with an accuracy of ±0.2 ◦ C) was achieved by a water circulation system from a thermostat through the temperature controlled jacket of the pressure cell. The intensity of the diffraction pattern was plotted as a function of the scattering vector q (q = (4π/l) sinθ , with 2θ being the scattering angle). The scattering curves were background corrected for scattering of the pure solvent, taking into account the different absorption factors. Further analysis was performed with the Fit2D software.39 The scattering intensity of a highly concentrated protein solution (eventually after further subtraction of a constant incoherent background term) can be described as a product of the form factor, P(q), which considers the scattering of a single particle, and the effective structure factor, Seff (q), which takes the intermolecular interactions of the proteins into account: I (q) = P (q) · Seff (q).
(1)
The form factor of the lysozyme can be modeled by an ellipsoidal shape:22 2 1 j q a 2 + x 2 (b2 − a 2 ) 1 P (q) = (2) 4 dx, q a 2 + x 2 (b2 − a 2 ) 0 with a and b being the semiaxes of the ellipsoid and j1 the first-order Bessel function. Lysozyme is approximately an ellipsoid with a volume of (π/6) × 4.5 × 3.0 × 3.0 nm3 . The effective structure factor, Seff (q), for high protein concentrations can be obtained within the decoupling approximation:22 Seff (q) = 1 +
F (q)2 (S(q) − 1) . P (q)
(3)
F (q) is the calculated spherical average of the Fouriertransform of the protein’s electron density and S(q) is the intermolecular structure factor, the Fourier-transform of the pair-correlation function. By measuring the scattering pattern at both highly concentrated and highly diluted concentrations, the structure factor Seff (q) can be determined (Eqs. (1)–(3)). The theoretical approaches used to calculate the structure factor from effective pair potentials (and vice versa) are extensions of developments made in the framework of the theory of the structure of simple liquids and colloidal suspensions. The forces between protein molecules in solution typically include screened repulsive Coulomb interactions and attractive van der Waals interactions. In addition, proteinion dispersion and hydration forces may play a role. Subtle mutations on the surface of these molecules may have
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 128.248.155.225 On: Fri, 21 Nov 2014 19:31:34
22D506-3
Grobelny et al.
J. Chem. Phys. 141, 22D506 (2014)
profound effects on their intermolecular interactions as well but can be disregarded in the case of lysozyme under the conditions chosen here. Generally, rather simple models are used to describe protein-protein interactions, which are spatially isotropic and next to Coulomb repulsion include shortrange attractive interactions, only. In that sense, proteins can be also viewed as valuable model colloids, as they are monodisperse and are of the order of a few nanometers in diameter. In our study, the interaction potential, V(r), was obtained by using a modified DLVO (Derjaguin-LandauVerwey-Overbeek)-model in the mean spherical approximation (MSA) from fits to the structure factor data.12, 18, 21, 22, 41–43 The potential consists of a sum of three contributions. The first part, the hard-sphere potential, VHS (r), can be described as ∞, r ≤ σ , (4) VHS (r) = 0, r>σ where σ is the hard-sphere diameter of the protein. The repulsive Coulomb potential, VC (r), can be expressed as ⎧ ⎪ r≤σ ⎨ 0, 2 2 −κ(r−σ ) . (5) VC (r) = Z e e ⎪ , r>σ ⎩ 4πε0 εr (1 + 0.5κσ )2 r Here, Z is the protein charge, e is the elementary charge, ε0 is the vacuum dielectric constant, εr is the dielectric permittivity of the solution, and κ is the reciprocal Debye-Hückel screening length. The pressure and temperature dependence of the dielectric permittivity of the solution was taken into account as well.44 The attractive part is modeled by a Yukawa potential, VY (r), which can be calculated as ⎧ r≤σ ⎨ 0, −(r−σ )/d . (6) VY (r) = ⎩ −J σ e , r>σ r The effective protein hard-sphere diameter, σ , was set to 2.99 nm and the effective protein charge in the model was kept constant at a value of Z = 8 for pH 7.45 A width of the attractive part of VY (r) of d = 0.3 nm was used.25 From the fits to the experimental data, the depth of the attractive part of the interaction potential, J (Eq. (6)), was obtained. The calculations were performed by a MATLAB-based fitting routine according to Liu et al.40 III. RESULTS AND DISCUSSION
SAXS data on lysozyme solutions were taken at 20 ◦ C covering a wide range of protein concentrations, c, up to ∼370 mg ml−1 , which is about twice as high as reported before.22 In order to check if lysozyme is still in its native conformation over the whole pressure range covered, SAXS data were also measured for a low lysozyme concentration (8.57 mg ml−1 ), i.e., under conditions where intermolecular interactions are negligible and the radius of gyration, RG , could be determined from the Guinier-plot or the pair distance distribution function, p(r), the latter yielding more ac-
FIG. 2. (a) Pressure dependent SAXS data of lysozyme at a concentration of 8.57 mg ml−1 and ambient temperature (T = 20 ◦ C). With increasing pressure, the contrast and hence the scattering intensity decreases. (b) Pressure dependence of the radius of gyration, RG, as obtained from the pair distribution function, p(r).
curate data (Figs. 2(a) and 2(b)).22, 46 As shown in Fig. 2(b), RG amounts to 1.48 nm at ambient pressure, which is in good agreement with literature data.20, 47 In the pressure range up to 3 kbar, no changes in RG are observed, indicating that the monomeric lysozyme molecule does not unfold in this pressure range, which is also in agreement with high-pressure FTIR spectroscopic literature data.22, 48 Figure 3 depicts small-angle X-ray scattering curves at lysozyme concentrations, ranging from 87 to 367 mg ml−1 at ambient pressure for T = 20 ◦ C. The appearance of a correlation peak, at momentum transfer qcorr , which originates from repulsive protein-protein interaction of the highly positively charged (Z = 8) lysozyme molecules, is clearly visible; qcorr values range from 0.48 nm−1 to 0.8 nm−1 . As expected, the correlation peak shifts linearly to higher q-values with increasing protein concentration, c, corresponding to smaller distances between the particles (Fig. 3, right). Assuming a (c)1/3 -dependent shift of qmax to larger q-values, which represents a reasonable approximation for strongly repulsive
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 128.248.155.225 On: Fri, 21 Nov 2014 19:31:34
22D506-4
Grobelny et al.
J. Chem. Phys. 141, 22D506 (2014)
FIG. 3. (a) Small-angle X-ray scattering curves of lysozyme solutions at different concentrations at 1 bar and 20 ◦ C. For better visualization, the scattering curves are vertically shifted. (b) Position of the correlation peaks, qcorr , as a function of protein concentration at 1 bar and 20 ◦ C.
systems as studied here, the average distance between particles can be estimated from 2π/qmax .8, 15 Using this approximation, the shift of qmax from 0.48 to 0.8 nm−1 corresponds to changes in intermolecular distances from 13.10 to 7.85 nm. At protein concentrations above approximately 150 mg ml−1 , a shift of qmax is no longer observable; qmax seems even to slightly decrease. Such behavior indicates changes in intermolecular correlations, such as the onset of (dynamic) clustering of lysozyme molecules. For the highest concentration measured (c = 366 mg ml−1 ) partial crystallization of the sample is observed. Strong Bragg reflections are superimposed on the correlation peak, typical for formation of a liquid/crystal two-phase coexistence region in this highly supersaturated solution. Typical pressure-dependent SAXS curves are depicted in Fig. 4 for a lysozyme concentration of 112 mg ml−1 at 20 ◦ C. With increasing pressure, after a small increase, a pressureinduced shift of the correlation peak, qcorr , to lower values is observed. The pressure dependence of qmax for all homogeneous protein samples is depicted in Fig. 4 (right). A more or less shallow maximum of qmax (p) is observed in the 1 kbar pressure range for all protein concentrations up to ∼300 mg ml−1 . A strong decrease of qmax at high pressures is observed for concentrations below 100 mg ml−1 , only. Generally, upon compression in the kbar range, a small increase of qmax would be expected. The shallow maximum and subsequent decrease of qmax must be related to solventmediated changes in intermolecular correlations22, 49 and/or some kind of (dynamic) clustering of the protein molecules, which has in fact been observed in concentrated lysozyme solutions.26, 28, 61 Further information about the origin of these structural changes upon compression can be obtained from
the calculation of the pressure dependence of the attractive potential, J, within the DLVO approximation, as discussed below. The pressure dependence of the attractive part, J, of the interaction potential, as obtained from the fit of the DLVOmodel within the MSA approximation to the experimental data (see Sec. II), is displayed in Fig. 5 for different protein concentrations. It can be clearly seen that J decreases, as expected, with increasing protein concentration, resulting in stronger repulsive interactions (lysozyme has a constant net charge of 8 under our solvent conditions). Applying pressure leads first to an increase in the repulsion, i.e., a decrease of the attractive potential, which arises from the higher density of the solution. At a pressure of about 1500–2000 bars, however, the attractive potential, J, increases again. In previous studies, carried out for the lower concentration region,22 it was concluded that this nonlinear behavior and minimum in J(p) might arise from the evolving collapse of the second hydration shell of water and the penetration of non-hydrogen bonded water molecules into the first hydration shell upon pressurization,50 i.e., due to changes in the medium-range structure of bulk water that starts to rearrange in this pressure range. Notably, with increasing protein concentration, the minimum in J(p) becomes shallower and almost disappears for concentrations of about 200 mg ml−1 . For these high protein concentrations, J(p) remains essentially pressure-insensitive above ∼1500 bars, indicating that a pressure-induced change in the water structure is less effective in such highly concentrated protein solutions and/or changes in interaction forces (e.g., hydration repulsion, many-body effects) at these short distances. Note, that for concentrations above about
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 128.248.155.225 On: Fri, 21 Nov 2014 19:31:34
22D506-5
Grobelny et al.
J. Chem. Phys. 141, 22D506 (2014)
FIG. 4. (a) Pressure-dependent SAXS curves of lysozyme solutions at a concentration of 112 mg ml−1 at T = 20 ◦ C. The scattering curves are shifted vertically to aid better visualization. (b) Pressure dependent shift of the position, qmax , of the correlation peak at different protein concentrations.
FIG. 5. (a) Depth of the attractive well, J, of the interaction potential V(r) of highly concentrated lysozyme solutions as a function of pressure at different protein concentrations (T = 20 ◦ C). The lines are shown to guide the eye, only. (b) Representative scattering curves for low and high protein concentrations at 1 bar and 20 ◦ C with corresponding fitting curves using the DLVO-potential within the MSA approximation superimposed on the experimental data.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 128.248.155.225 On: Fri, 21 Nov 2014 19:31:34
22D506-6
Grobelny et al.
200 mg ml−1 , the fits of the scattering curves are less accurate (Fig. 5, right) and hence the results have to be discussed with care. In fact, the DLVO-potential is expected to be less suitable for describing protein-protein interactions at very high protein concentrations, where additional interaction terms (hydration repulsion, interaction of hydrophobic patches, polarization terms, etc.) as well as multibody forces might come into play, similar to those observed in dense colloidal suspensions.51–53 As a consequence, also an increased tendency towards (dynamic) clustering of protein molecules may be considered, a phenomenon observed before in lysozyme solutions even at lower protein concentrations. However, the formation and extent of clustering at high lysozyme concentrations is still a controversial issue in the current literature.8, 26, 28, 54–61 Neutron-spin-echo (NSE) and small-angle neutron scattering (SANS) experiments revealed the existence of dynamic clusters at high protein concentrations, and a transition from a monomer to a clusterdominated scenario was revealed.58–60 Interestingly, the concentration of about 200 mg ml−1 , where the maximum of qmax (c) and the disappearance of the pronounced minimum in J(p) is observed, is similar to the critical concentration (ccrit ≈ 230 mg ml−1 ) of the liquid-liquid two-phase region, which is observed at a 15 ◦ C lower temperature, however. In this concentration range, separation into a condensed and a dilute protein solution takes place. Hence, the assumption of (dynamic) clustering taking place in this concentration region at 20 ◦ C seems to be not unreasonable. Moreover, in this concentration range, the concentration dependence of J seems to change sign, becoming more attractive. Increasing the protein concentration to 366 mg ml−1 , a concentration is reached, where, within the duration of the experiment (∼3 h), nucleation and partial crystallization of the supersaturated sample takes place.62–66 Figure 6 displays the q-range where Bragg reflections are observed. Indexing of the Bragg reflections indicates a primitive tetragonal unit cell (space group P43 21 2) with axis a = b = 7.9 nm and c = 3.8 nm, which is in good agreement with literature data.67–71 The application of pressure leads to an anisotropic compression along the axis of the unit cell, i.e., a compression of the a axis, whereas the c axis length stays approx-
FIG. 6. Pressure dependent SAXS curves of lysozyme solutions at a concentration of 366 mg ml−1 (T = 20 ◦ C), showing the intermediate q-range where Bragg reflections are observed. With increasing pressure, protein crystal formation is suppressed.
J. Chem. Phys. 141, 22D506 (2014)
imately constant.25 These data are in good agreement with pressure dependent measurements on lysozyme crystals.68 Interestingly, the intensity of the Bragg peaks decreases upon compression, which indicates suppression of protein crystallization upon hydrostatic compression of the solution. (Of note, the pressure-induced decrease of the scattering intensity due to a decrease in scattering contrast upon compression of the solution cannot explain this marked decrease alone.) This observation implies that formation of lysozyme crystals is connected to an overall volume increase of the system. IV. CONCLUSIONS
The influence of pressure on the structure and proteinprotein interaction potential of highly concentrated protein solutions up to about 370 mg ml−1 was studied and analyzed using small-angle X-ray scattering in combination with a liquid-state theoretical approach. The highest concentrations could be obtained using osmotic pressure methology. Under the preparation conditions chosen, the protein solutions where stable against crystallization for days. Only at a protein concentration as high as 366 mg ml−1 , partial crystallization of the metastable protein solution was observed within the time period of the experiment. In the concentration region below 200 mg ml−1 , in agreement with literature data,22 the interaction parameters of the lysozyme solutions are affected by pressure in a nonlinear way, which is probably due to significant changes in the structural properties of bulk water that occur above 1–2 kbar, i.e., is a solvent-mediated effect. In this pressure regime, where the coordination number of water increases markedly due to a collapsed second hydration shell, protein-protein interactions are modified. Interestingly, a peculiar pressure dependence has also been observed in transport and related thermodynamic properties of bulk water, e.g., shallow maxima in the diffusion constant and configurational entropy around 1.5 kbar at room temperature.72, 73 Conversely, for higher concentrated protein solutions, exceeding ∼200 mg ml−1 , no pronounced minimum in J(p) is observed anymore, and the interaction potential turns rather insensitive to compression. Our data show that in this pressure range of 200–366 mg ml−1 , bulk-water-mediated effects on protein-protein interactions as observed in the lower protein concentration range seem to have largely vanished. This might not be too surprising, as in this concentration range – assuming a homogeneous distribution of protein particles – intermolecular distances have decreased to 5 nm at c = 200 mg ml−1 and 4 nm at c = 366 mg ml−1 , respectively. Taking the size of the lysozyme molecule into account, this translates into water layers between adjacent protein molecules of about 4 and 1, only. For such thin hydration layers, whose structural properties are also influenced by the surface properties of the protein molecules (e.g., by electrostriction effects upon hydration of surface charges), no bulk-like solvent properties can be expected anymore. Hence, the structural properties of bulk water that influence the spatial organization of the more dilute protein solutions upon compression, cannot be as effective anymore for protein concentrations exceeding 200 mg ml−1 . Consequently, no
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 128.248.155.225 On: Fri, 21 Nov 2014 19:31:34
22D506-7
Grobelny et al.
pronounced minimum in J(p) is observed. This is also the concentration range, where the intermolecular interaction potential does not become repulsive anymore with increasing protein concentration, and the concentration dependence of J seems to even change sign at the highest protein concentration, becoming more attractive. However, these data have to be taken with care, as the DLVO-potential is expected to be less suitable for describing protein-protein interactions at these very high protein concentrations, where additional interaction terms including many-body forces might come into play. The absence of a SAXS intensity increase for q → 0 shows that no irreversible aggregation of lysozyme takes place. Rather, – possibly transient (dynamic) – protein clustering takes place, which is also apparent from the maximum of qmax reached in this concentration range. Future NSE experiments in this concentration range might support these conclusions. Our results also show that pressure suppresses protein nucleation, aggregation, and finally crystallization in supersaturated condensed lysozyme solutions. These findings are thus also of importance for understanding the high stability of concentrated protein solutions (as they occur intracellularly) in biosystems thriving under hydrostatic pressure conditions such as in the deep sea, where pressures up to the kbar-level and beyond are reached.74, 75 In the crowded cell, where highly concentrated protein solutions (200–300 mg ml−1 )1, 76–78 are encountered and macromolecular crowding reaches levels of 30 vol.% and more, pressure is expected to effectively suppress protein aggregation. ACKNOWLEDGMENTS
We acknowledge the Diamond Light Source for providing synchrotron radiation. Financial support from the DFG Research Unit FOR 1979 and in part of the Cluster of Excellence RESOLV (EXC 1069) funded by the Deutsche Forschungsgemeinschaft (DFG) is gratefully acknowledged. J.M. thanks the BMBF (05K10PEC) for financial support. 1 A.
P. Minton, J. Cell Sci. 119, 2863 (2006). J. Ellis, Adv. Exp. Med. Biol. 594, 1 (2007). 3 S. Zhao, H. Gies, and R. Winter, Z. Phys. Chem. 221, 139 (2007). 4 M. Erlkamp, S. Grobelny, and R. Winter, Phys. Chem. Chem. Phys. 16, 5965 (2014). 5 C. Haas and J. Drenth, J. Cryst. Growth 196, 388 (1999). 6 N. Asherie, Methods 34, 266 (2004). 7 R. Ravindra, S. Zhao, H. Gies, and R. Winter, J. Am. Chem. Soc. 126, 12224 (2004). 8 A. Stradner, H. Sedgwick, F. Cardinaux, W. C. K. Poon, S. U. Egelhaaf, and P. Schurtenberger, Nature (London) 432, 492 (2004). 9 A. C. Dumetz, A. M. Chockla, E. W. Kaler, and A. M. Lenhoff, Biophys. J. 94, 570 (2008). 10 R. Mishra and R. Winter, Angew. Chem., Int. Ed. 47, 6518 (2008). 11 J. Seeliger, A. Werkmüller, and R. Winter, PLoS One 8, e69652 (2013). 12 J. D. Gunton, A. Shiryayev, D. L. Pagan, Protein Condensation (Cambridge University Press, Cambridge, 2007). 13 N. Javid, K. Vogtt, C. Krywka, M. Tolan, and R. Winter, ChemPhysChem 8, 679 (2007). 14 V. G. Taratuta, A. Holschbach, G. M. Thurston, D. Blankschtein, and G. B. Benedek, J. Phys. Chem. 94, 2140 (1990). 15 M. Muschol and F. Rosenberger, J. Chem. Phys. 107, 1953 (1997). 16 A. Ducruix, J. Guilloteau, M. Riès-Kautt, and A. Tardieu, J. Cryst. Growth 168, 28 (1996). 17 A. Tardieu, A. Le Verge, M. Malfois, F. Bonnete, S. Finet, M. Ries-Kautt, and L. Belloni, J. Cryst. Growth 196, 193 (1999). 2 R.
J. Chem. Phys. 141, 22D506 (2014) 18 J.
Narayanan and X. Y. Liu, Biophys. J. 84, 523 (2003). Pellicane, D. Costa, and C. Caccamo, J. Phys. Condens. Matter 15, 375 (2003). 20 G. Pellicane, D. Costa, and C. Caccamo, J. Phys. Condens. Matter 16, S4923 (2004). 21 M. Niebuhr and M. H. J. Koch, Biophys. J. 89, 1978 (2005). 22 M. Schroer, J. Markgraf, D. Wieland, C. Sahle, J. Möller, M. Paulus, M. Tolan, and R. Winter, Phys. Rev. Lett. 106, 178102 (2011). 23 M. Schroer, Y. Zhai, D. C. F. Wieland, C. J. Sahle, J. Nase, M. Paulus, M. Tolan, and R. Winter, Angew. Chem., Int. Ed. 50, 11413 (2011). 24 M. Schroer, M. Tolan, and R. Winter, Phys. Chem. Chem. Phys. 14, 9486 (2012). 25 J. Möller, M. Schroer, M. Erlkamp, S. Grobelny, M. Paulus, S. Tiemeyer, F. J. Wirkert, M. Tolan, and R. Winter, Biophys. J. 102, 2641 (2012). 26 A. J. Chinchalikar, V. K. Aswal, J. Kohlbrecher, and A. G. Wagh, Phys. Rev. E 87, 062708 (2013). 27 J. Möller, S. Grobelny, J. Schulze, S. Bieder, A. Steffen, M. Erlkamp, M. Paulus, M. Tolan, and R. Winter, Phys. Rev. Lett. 112, 028101 (2014). 28 A. Stradner, F. Cardinaux, and P. Schurtenberger, J. Phys. Chem. B 110, 21222 (2006). 29 F. Zhang, M. W. A. Skoda, R. M. J. Jacobs, R. A. Martin, C. M. Martin, and F. Schreiber, J. Phys. Chem. B 111, 251 (2007). 30 M. Carpineti and R. Piazza, Phys. Chem. Chem. Phys. 6, 1506 (2004). 31 F. Meersman, I. Daniel, D. H. Bartlett, R. Winter, R. Hazael, and P. F. McMillan, Rev. Mineral. Geochem. 75, 607 (2013). 32 S. Bachali, X. Bailly, J. Jolles, P. Jolles, and J. S. Deutsch, Eur. J. Biochem. 271, 237 (2004). 33 S. Bachali, M. Jager, A. Hassanin, F. Schoentgen, P. Jollès, A. FialaMedioni, and J. S. Deutsch, J. Mol. Evol. 54, 652 (2002). 34 J. Jollès, A. Fiala-Médioni, and P. Jollès, J. Mol. Evol. 43, 523 (1996). 35 A. Bouchoux, P.-E. Cayemitte, J. Jardin, G. Gésan-Guiziou, and B. Cabane, Biophys. J. 96, 693 (2009). 36 Y. Kitamura and T. Itoh, J. Sol. Chem. 16, 715 (1987). 37 R. F. Steiner, Biochim. Biophys. Acta 79, 51 (1964). 38 N. J. Brooks, B. L. L. E. Gauthe, N. J. Terrill, S. E. Rogers, R. H. Templer, O. Ces, and J. M. Seddon, Rev. Sci. Instrum. 81, 064103 (2010). 39 A. P. Hammersley, S. O. Svensson, M. Hanfland, A. N. Fitch, and D. Häusermann, High Press. Res. 14, 235 (1996). 40 Y. Liu, W.-R. Chen, and S.-H. Chen, J. Chem. Phys. 122, 044507 (2005). 41 B. V. Derjaguin and L. V. Landau, Acta Physicochim. USSR 14, 633 (1941). 42 E. J. W. Verwey, J. Chem. Phys. 51, 631 (1947). 43 E. J. W. Verwey and J. T. G. Overbeek, Theory of the Stability of Lyophobic Colloids (Elsevier, New York, 1948). 44 W. B. Floriano and M. A. C. Nascimento, Braz. J. Phys. 34, 38 (2004). 45 D. E. Kuehner, J. Engmann, F. Fergg, M. Wernick, H. W. Blanch, and J. M. Prausnitz, J. Phys. Chem. B 103, 1368 (1999). 46 D. I. Svergun and M. H. Koch, Rep. Prog. Phys. 66, 1735 (2003). 47 D. I. Svergun, S. Richard, M. H. J. Koch, Z. Sayers, S. Kuprin, and G. Zaccai, Proc. Natl. Acad. Sci. U.S.A. 95, 2267 (1998). 48 L. Smeller, F. Meersman, and K. Heremans, Biochim. Biophys. Acta 1764, 497 (2006). 49 T. Zemb and P. Lindner, Neutron, X-Rays and Light. Scattering Methods Applied to Soft Condensed Matter (North Holland, Amsterdam, 2002). 50 A. Soper and M. Ricci, Phys. Rev. Lett. 84, 2881 (2000). 51 J. Crocker and D. Grier, Phys. Rev. Lett. 77, 1897 (1996). 52 A. E. Larsen and D. G. Grier, Nature (London) 385, 230 (1997). 53 J. J. Gray, B. Chiang, and R. T. Bonnecaze, Nature (London) 402, 750 (1999). 54 A. Stradner, F. Cardinaux, S. U. Egelhaaf, and P. Schurtenberger, Proc. Natl. Acad. Sci. U.S.A. 105, E75 (2008). 55 A. Stradner, F. Cardinaux, and P. Schurtenberger, Phys. Rev. Lett. 96, 219801 (2006). 56 A. Shukla, E. Mylonas, E. Di Cola, S. Finet, P. Timmins, T. Narayanan, and D. I. Svergun, Proc. Natl. Acad. Sci. U.S.A. 105, 5075 (2008). 57 A. Shukla, E. Mylonas, E. Di Cola, S. Finet, P. Timmins, T. Narayanan, and D. I. Svergun, Proc. Natl. Acad. Sci. U.S.A. 105, E76 (2008). 58 L. Porcar, P. Falus, W.-R. Chen, A. Faraone, E. Fratini, K. Hong, P. Baglioni, and Y. Liu, J. Phys. Chem. Lett. 1, 126 (2010). 59 F. Cardinaux, E. Zaccarelli, A. Stradner, S. Bucciarelli, B. Farago, S. U. Egelhaaf, F. Sciortino, and P. Schurtenberger, J. Phys. Chem. B 115, 7227 (2011). 60 Y. Liu, L. Porcar, J. Chen, W.-R. Chen, P. Falus, A. Faraone, E. Fratini, K. Hong, and P. Baglioni, J. Phys. Chem. B 115, 7238 (2011). 19 G.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 128.248.155.225 On: Fri, 21 Nov 2014 19:31:34
22D506-8 61 W.
Grobelny et al.
Pan, P. G. Vekilov, and V. Lubchenko, J. Phys. Chem. B 114, 7620 (2010) 62 L. F. Filobelo, O. Galkin, and P. G. Vekilov, J. Chem. Phys. 123, 014904 (2005). 63 E. Cacioppo and M. L. Pusey, J. Cryst. Growth 114, 286 (1991). 64 F. Rosenberger, S. B. Howard, J. W. Sowers, and T. A. Nyce, J. Cryst. Growth 129, 1 (1993). 65 W. C. K. Poon, S. U. Egelhaaf, P. A. Beales, A. Salonen, and L. Sawyer, J. Phys. Condens. Matter 12, L569 (2000). 66 M. Muschol and F. Rosenberger, J. Chem. Phys. 103, 10424 (1995). 67 C. C. F. Blake, G. A. Mair, A. C. T. North, Proc. R. Soc. London, Ser. B 167, 365 (1967). 68 R. Fourme, R. Kahn, M. Mezouar, E. Girard, C. Hoerentrup, T. Prangé, and I. Ascone, J. Synchrotron Radiat. 8, 1149 (2001).
J. Chem. Phys. 141, 22D506 (2014) 69 C.
Sauter, F. Otalora, J. Gavira, O. Vidal, R. Giege, and J. Garcia-Ruiz, Acta Crystallogr., Sect. D 57, 1119 (2001). 70 Y. Cheng, R. F. Lobo, S. Sandler, and A. M. Lenhoff, Biotechnol. Bioeng. 94, 177 (2006). 71 C. Kundrot and F. M. Richards, J. Appl. Crystallogr. 19, 208 (1986). 72 F. Mallamace, C. Corsaro, D. Mallamace, C. Vasi, and H. E. Stanley, Faraday Discuss. 167, 95 (2013). 73 R. Ludwig, Angew. Chem., Int. Ed. 40, 1808 (2001). 74 I. Daniel, P. Oger, and R. Winter, Chem. Soc. Rev. 35, 858 (2006). 75 F. M. Lauro and D. H. Bartlett, Extremophiles 12, 15 (2008). 76 B. van den Berg, R. Wain, C. M. Dobson, and R. J. Ellis, EMBO J. 19, 3870 (2000). 77 G. Rivas, F. Ferrone, and J. Herzfeld, EMBO Rep. 5, 23 (2004). 78 L. Stagg, S.-Q. Zhang, M. S. Cheung, and P. Wittung-Stafshede, Proc. Natl. Acad. Sci. U.S.A. 104, 18976 (2007).
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 128.248.155.225 On: Fri, 21 Nov 2014 19:31:34
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 128.248.155.225 On: Fri, 21 Nov 2014 19:31:34
THE JOURNAL OF CHEMICAL PHYSICS 141, 22D506 (2014)
Intermolecular interactions in highly concentrated protein solutions upon compression and the role of the solvent S. Grobelny,1 M. Erlkamp,1 J. Möller,2 M. Tolan,2 and R. Winter1 1
Faculty of Chemistry, Physical Chemistry-Biophysical Chemistry, TU Dortmund, Otto-Hahn Str. 6, 44227 Dortmund, Germany 2 Fakultät Physik/DELTA, TU Dortmund, Maria-Goeppert-Mayer-Str. 2, 44227 Dortmund, Germany
(Received 30 April 2014; accepted 9 June 2014; published online 23 September 2014) The influence of high hydrostatic pressure on the structure and protein-protein interaction potential of highly concentrated lysozyme solutions up to about 370 mg ml−1 was studied and analyzed using small-angle X-ray scattering in combination with a liquid-state theoretical approach. In the concentration region below 200 mg ml−1 , the interaction parameters of lysozyme solutions are affected by pressure in a nonlinear way, which is probably due to significant changes in the structural properties of bulk water, i.e., due to a solvent-mediated effect. Conversely, for higher concentrated protein solutions, where hydration layers below ∼4 water molecules are reached, the interaction potential turns rather insensitive to compression. The onset of transient (dynamic) clustering is envisaged in this concentration range. Our results also show that pressure suppresses protein nucleation, aggregation and finally crystallization in supersaturated condensed protein solutions. These findings are of importance for controlling and fine-tuning protein crystallization. Moreover, these results are also important for understanding the high stability of highly concentrated protein solutions (as they occur intracellularly) in organisms thriving under hydrostatic pressure conditions such as in the deep sea, where pressures up to the kbar-level are reached. © 2014 AIP Publishing LLC. [http://dx.doi.org/10.1063/1.4895542] I. INTRODUCTION
The investigation of concentrated protein solutions and protein condensation phenomena is not only a challenging subject in liquid-state theory of soft condensed matter systems, a detailed understanding of the properties of such systems is also the prerequisite for a number of biological and biotechnological applications. From a biological point of view, there has been an increased interest in studies of the behavior of proteins in highly crowded solutions in order to mimic intracellular environments. In biological cells, crowding and confinement are able to cause major alterations of the conformational stability and biological activity of proteins.1–4 Knowledge of protein-protein interactions in condensed fluid phases is also crucial for understanding the functional and structural stability of proteins and to yield molecular insights into processes such as protein crystallization, aggregation, and fibrillation.5–11 The latter involves diseases that occur due to undesired protein nucleation and aggregate formation. Classic examples are formation of polymer fibers of sickle hemoglobin molecules within the red blood cells that produces sickle cell anemia, age-related cataracts produced by the undesired aggregation of γ -crystallin, or the formation of amyloid fibres via cross-β-sheet formation in a series of debilitating diseases such as Alzheimer’s, Parkinson’s, CreutzfeldJakob disease, or diabetes mellitus type 2.12 To combat such diseases, strategies for preventing nucleation of protein fibers must be developed. The structural properties of highly concentrated protein solutions and the phase diagram of proteins depend on a num-
0021-9606/2014/141(22)/22D506/8/$30.00
ber of control parameters, such as temperature, pressure, pH, ionic strength, and the presence of osmolytes. This parameter space and the underlying intermolecular interaction potential of the proteins as a function of these parameters are hardly explored today.13, 15–29 The use of the pressure variable is least understood so far and is hence in the focus of this study. Here we report the pressure dependence of the interaction potential of concentrated lysozyme solutions over a wide concentration range, up to about 370 mg ml−1 , including concentrations where protein crystallization sets in. Figure 1 exhibits the temperature-protein concentration dependent phase diagram of lysozyme at corresponding solution conditions as well as the temperature-protein concentration conditions selected in our study. Conditions were chosen such that the protein crystallization kinetics is still very slow, allowing us to explore protein-protein interactions in the highly condensed fluid phase which mimics also conditions encountered intracellularly. Large supersaturation is required to overcome the activation energy barrier for crystal formation of lysozyme. In the absence of crystal nuclei, depending on the protein concentration, it can take more than a month to reach thermal equilibrium.6, 30 The appearance of Bragg reflections due to the onset of crystallization has been observed for the highest protein concentration chosen in our study, only. In order to determine the structure and intermolecular interactions of such highly concentrated protein solutions, small-angle X-ray scattering (SAXS) experiments have been carried out in the pressure range from 1 bar to 3 kbar. The results are of interest for exploring the effects of pressure on the phase behavior of protein solutions, including self-crowding
141, 22D506-1
© 2014 AIP Publishing LLC
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 128.248.155.225 On: Fri, 21 Nov 2014 19:31:34
22D506-2
Grobelny et al.
FIG. 1. Temperature-concentration phase diagram of lysozyme. (Red circles) Experimental conditions chosen in our study. Liquid–liquid phase coexistence according to Taratuta et al.14 The solubility curve was calculated according to Muschol et al.15
conditions similar to conditions met in a biological cell. Investigating pressure effects is also of relevance for understanding the hydration behavior of biological matter under extreme environmental conditions, such as in the deep sea where pressures up to the kbar-level (1 kbar = 100 MPa) and beyond are encountered.31–34 II. MATERIALS AND METHODS
Lysozyme is a small enzyme that protects humans from bacterial infection. It attacks the protective cell walls of bacteria, breaking the carbohydrate chains in the walls, thereby destroying the structural integrity of the bacteria’s cell walls. It is a tightly packed globular protein that consists of 129 amino acid residues. It has a molecular weight of about 14 kDa and an isoelectric point at pI = 11.16. The protein is an ideal candidate for investigations of its phase diagram, nucleation, and crystal growth kinetics, as its equilibrium properties are well established under ambient conditions.12 We used hen egg white lysozyme, which was purchased from Sigma Aldrich and used without further purification. For the highest protein concentrations, samples were prepared in dialyR , MWCO = 2 kDa, Thermo Sciensis bags (Slyde-A-Lyzer tific, Rockford, IL, USA) applying osmotic pressure over different timescales at 20 ◦ C under constant stirring, which is a mild method to prepare also highly supersaturated metastable protein solutions.35 Polyethyleneglycol (PEG, Sigma Aldrich, Schnelldorf, Germany) with a molar mass of 35 kDa was used as main component of the dialysis buffer, which was adjusted to pH 7. The initial concentration was amounted to 80 mg ml−1 . As buffer solution, Bis-Tris (Sigma Aldrich, Schnelldorf, Germany) at a concentration of 50 mM was used, which is a pressure stable buffer over a wide pressure range.36 This relatively high buffer concentration was needed to keep the pH value constant at 7 even at the highest protein concentrations used. The final protein concentration was determined by UV-Vis spectroscopy using an extinction coefficient of 2.64 ml mg−1 cm−1 .37 The small-angle X-ray Scattering (SAXS) experiments were carried out at the beamline I22 at Diamond Light Source (Didcot, Oxfordshire, UK) with a custom-made high pressure
J. Chem. Phys. 141, 22D506 (2014)
cell.38 The X-ray energy was 18 keV, which corresponds to a wavelength, l, of 0.096 nm. The scattering curves were recorded with a PILATUS 2M detector at a sample to detector distance of 5.2 m, thereby covering a range of momentum transfer, q, from 0.2 to 4.5 nm−1 . The exposure time was 15 s. A total of 20 μl of the sample was filled into the pressure cell. The time for thermal equilibration before each measurement was set to 15 min. High hydrostatic pressures up to 3 kbar – with an accuracy of 0.5 % – at steps of 250 bars were applied at a temperature of 20 ◦ C, using water as pressurising medium. Temperature control (with an accuracy of ±0.2 ◦ C) was achieved by a water circulation system from a thermostat through the temperature controlled jacket of the pressure cell. The intensity of the diffraction pattern was plotted as a function of the scattering vector q (q = (4π/l) sinθ , with 2θ being the scattering angle). The scattering curves were background corrected for scattering of the pure solvent, taking into account the different absorption factors. Further analysis was performed with the Fit2D software.39 The scattering intensity of a highly concentrated protein solution (eventually after further subtraction of a constant incoherent background term) can be described as a product of the form factor, P(q), which considers the scattering of a single particle, and the effective structure factor, Seff (q), which takes the intermolecular interactions of the proteins into account: I (q) = P (q) · Seff (q).
(1)
The form factor of the lysozyme can be modeled by an ellipsoidal shape:22 2 1 j q a 2 + x 2 (b2 − a 2 ) 1 P (q) = (2) 4 dx, q a 2 + x 2 (b2 − a 2 ) 0 with a and b being the semiaxes of the ellipsoid and j1 the first-order Bessel function. Lysozyme is approximately an ellipsoid with a volume of (π/6) × 4.5 × 3.0 × 3.0 nm3 . The effective structure factor, Seff (q), for high protein concentrations can be obtained within the decoupling approximation:22 Seff (q) = 1 +
F (q)2 (S(q) − 1) . P (q)
(3)
F (q) is the calculated spherical average of the Fouriertransform of the protein’s electron density and S(q) is the intermolecular structure factor, the Fourier-transform of the pair-correlation function. By measuring the scattering pattern at both highly concentrated and highly diluted concentrations, the structure factor Seff (q) can be determined (Eqs. (1)–(3)). The theoretical approaches used to calculate the structure factor from effective pair potentials (and vice versa) are extensions of developments made in the framework of the theory of the structure of simple liquids and colloidal suspensions. The forces between protein molecules in solution typically include screened repulsive Coulomb interactions and attractive van der Waals interactions. In addition, proteinion dispersion and hydration forces may play a role. Subtle mutations on the surface of these molecules may have
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 128.248.155.225 On: Fri, 21 Nov 2014 19:31:34
22D506-3
Grobelny et al.
J. Chem. Phys. 141, 22D506 (2014)
profound effects on their intermolecular interactions as well but can be disregarded in the case of lysozyme under the conditions chosen here. Generally, rather simple models are used to describe protein-protein interactions, which are spatially isotropic and next to Coulomb repulsion include shortrange attractive interactions, only. In that sense, proteins can be also viewed as valuable model colloids, as they are monodisperse and are of the order of a few nanometers in diameter. In our study, the interaction potential, V(r), was obtained by using a modified DLVO (Derjaguin-LandauVerwey-Overbeek)-model in the mean spherical approximation (MSA) from fits to the structure factor data.12, 18, 21, 22, 41–43 The potential consists of a sum of three contributions. The first part, the hard-sphere potential, VHS (r), can be described as ∞, r ≤ σ , (4) VHS (r) = 0, r>σ where σ is the hard-sphere diameter of the protein. The repulsive Coulomb potential, VC (r), can be expressed as ⎧ ⎪ r≤σ ⎨ 0, 2 2 −κ(r−σ ) . (5) VC (r) = Z e e ⎪ , r>σ ⎩ 4πε0 εr (1 + 0.5κσ )2 r Here, Z is the protein charge, e is the elementary charge, ε0 is the vacuum dielectric constant, εr is the dielectric permittivity of the solution, and κ is the reciprocal Debye-Hückel screening length. The pressure and temperature dependence of the dielectric permittivity of the solution was taken into account as well.44 The attractive part is modeled by a Yukawa potential, VY (r), which can be calculated as ⎧ r≤σ ⎨ 0, −(r−σ )/d . (6) VY (r) = ⎩ −J σ e , r>σ r The effective protein hard-sphere diameter, σ , was set to 2.99 nm and the effective protein charge in the model was kept constant at a value of Z = 8 for pH 7.45 A width of the attractive part of VY (r) of d = 0.3 nm was used.25 From the fits to the experimental data, the depth of the attractive part of the interaction potential, J (Eq. (6)), was obtained. The calculations were performed by a MATLAB-based fitting routine according to Liu et al.40 III. RESULTS AND DISCUSSION
SAXS data on lysozyme solutions were taken at 20 ◦ C covering a wide range of protein concentrations, c, up to ∼370 mg ml−1 , which is about twice as high as reported before.22 In order to check if lysozyme is still in its native conformation over the whole pressure range covered, SAXS data were also measured for a low lysozyme concentration (8.57 mg ml−1 ), i.e., under conditions where intermolecular interactions are negligible and the radius of gyration, RG , could be determined from the Guinier-plot or the pair distance distribution function, p(r), the latter yielding more ac-
FIG. 2. (a) Pressure dependent SAXS data of lysozyme at a concentration of 8.57 mg ml−1 and ambient temperature (T = 20 ◦ C). With increasing pressure, the contrast and hence the scattering intensity decreases. (b) Pressure dependence of the radius of gyration, RG, as obtained from the pair distribution function, p(r).
curate data (Figs. 2(a) and 2(b)).22, 46 As shown in Fig. 2(b), RG amounts to 1.48 nm at ambient pressure, which is in good agreement with literature data.20, 47 In the pressure range up to 3 kbar, no changes in RG are observed, indicating that the monomeric lysozyme molecule does not unfold in this pressure range, which is also in agreement with high-pressure FTIR spectroscopic literature data.22, 48 Figure 3 depicts small-angle X-ray scattering curves at lysozyme concentrations, ranging from 87 to 367 mg ml−1 at ambient pressure for T = 20 ◦ C. The appearance of a correlation peak, at momentum transfer qcorr , which originates from repulsive protein-protein interaction of the highly positively charged (Z = 8) lysozyme molecules, is clearly visible; qcorr values range from 0.48 nm−1 to 0.8 nm−1 . As expected, the correlation peak shifts linearly to higher q-values with increasing protein concentration, c, corresponding to smaller distances between the particles (Fig. 3, right). Assuming a (c)1/3 -dependent shift of qmax to larger q-values, which represents a reasonable approximation for strongly repulsive
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 128.248.155.225 On: Fri, 21 Nov 2014 19:31:34
22D506-4
Grobelny et al.
J. Chem. Phys. 141, 22D506 (2014)
FIG. 3. (a) Small-angle X-ray scattering curves of lysozyme solutions at different concentrations at 1 bar and 20 ◦ C. For better visualization, the scattering curves are vertically shifted. (b) Position of the correlation peaks, qcorr , as a function of protein concentration at 1 bar and 20 ◦ C.
systems as studied here, the average distance between particles can be estimated from 2π/qmax .8, 15 Using this approximation, the shift of qmax from 0.48 to 0.8 nm−1 corresponds to changes in intermolecular distances from 13.10 to 7.85 nm. At protein concentrations above approximately 150 mg ml−1 , a shift of qmax is no longer observable; qmax seems even to slightly decrease. Such behavior indicates changes in intermolecular correlations, such as the onset of (dynamic) clustering of lysozyme molecules. For the highest concentration measured (c = 366 mg ml−1 ) partial crystallization of the sample is observed. Strong Bragg reflections are superimposed on the correlation peak, typical for formation of a liquid/crystal two-phase coexistence region in this highly supersaturated solution. Typical pressure-dependent SAXS curves are depicted in Fig. 4 for a lysozyme concentration of 112 mg ml−1 at 20 ◦ C. With increasing pressure, after a small increase, a pressureinduced shift of the correlation peak, qcorr , to lower values is observed. The pressure dependence of qmax for all homogeneous protein samples is depicted in Fig. 4 (right). A more or less shallow maximum of qmax (p) is observed in the 1 kbar pressure range for all protein concentrations up to ∼300 mg ml−1 . A strong decrease of qmax at high pressures is observed for concentrations below 100 mg ml−1 , only. Generally, upon compression in the kbar range, a small increase of qmax would be expected. The shallow maximum and subsequent decrease of qmax must be related to solventmediated changes in intermolecular correlations22, 49 and/or some kind of (dynamic) clustering of the protein molecules, which has in fact been observed in concentrated lysozyme solutions.26, 28, 61 Further information about the origin of these structural changes upon compression can be obtained from
the calculation of the pressure dependence of the attractive potential, J, within the DLVO approximation, as discussed below. The pressure dependence of the attractive part, J, of the interaction potential, as obtained from the fit of the DLVOmodel within the MSA approximation to the experimental data (see Sec. II), is displayed in Fig. 5 for different protein concentrations. It can be clearly seen that J decreases, as expected, with increasing protein concentration, resulting in stronger repulsive interactions (lysozyme has a constant net charge of 8 under our solvent conditions). Applying pressure leads first to an increase in the repulsion, i.e., a decrease of the attractive potential, which arises from the higher density of the solution. At a pressure of about 1500–2000 bars, however, the attractive potential, J, increases again. In previous studies, carried out for the lower concentration region,22 it was concluded that this nonlinear behavior and minimum in J(p) might arise from the evolving collapse of the second hydration shell of water and the penetration of non-hydrogen bonded water molecules into the first hydration shell upon pressurization,50 i.e., due to changes in the medium-range structure of bulk water that starts to rearrange in this pressure range. Notably, with increasing protein concentration, the minimum in J(p) becomes shallower and almost disappears for concentrations of about 200 mg ml−1 . For these high protein concentrations, J(p) remains essentially pressure-insensitive above ∼1500 bars, indicating that a pressure-induced change in the water structure is less effective in such highly concentrated protein solutions and/or changes in interaction forces (e.g., hydration repulsion, many-body effects) at these short distances. Note, that for concentrations above about
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 128.248.155.225 On: Fri, 21 Nov 2014 19:31:34
22D506-5
Grobelny et al.
J. Chem. Phys. 141, 22D506 (2014)
FIG. 4. (a) Pressure-dependent SAXS curves of lysozyme solutions at a concentration of 112 mg ml−1 at T = 20 ◦ C. The scattering curves are shifted vertically to aid better visualization. (b) Pressure dependent shift of the position, qmax , of the correlation peak at different protein concentrations.
FIG. 5. (a) Depth of the attractive well, J, of the interaction potential V(r) of highly concentrated lysozyme solutions as a function of pressure at different protein concentrations (T = 20 ◦ C). The lines are shown to guide the eye, only. (b) Representative scattering curves for low and high protein concentrations at 1 bar and 20 ◦ C with corresponding fitting curves using the DLVO-potential within the MSA approximation superimposed on the experimental data.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 128.248.155.225 On: Fri, 21 Nov 2014 19:31:34
22D506-6
Grobelny et al.
200 mg ml−1 , the fits of the scattering curves are less accurate (Fig. 5, right) and hence the results have to be discussed with care. In fact, the DLVO-potential is expected to be less suitable for describing protein-protein interactions at very high protein concentrations, where additional interaction terms (hydration repulsion, interaction of hydrophobic patches, polarization terms, etc.) as well as multibody forces might come into play, similar to those observed in dense colloidal suspensions.51–53 As a consequence, also an increased tendency towards (dynamic) clustering of protein molecules may be considered, a phenomenon observed before in lysozyme solutions even at lower protein concentrations. However, the formation and extent of clustering at high lysozyme concentrations is still a controversial issue in the current literature.8, 26, 28, 54–61 Neutron-spin-echo (NSE) and small-angle neutron scattering (SANS) experiments revealed the existence of dynamic clusters at high protein concentrations, and a transition from a monomer to a clusterdominated scenario was revealed.58–60 Interestingly, the concentration of about 200 mg ml−1 , where the maximum of qmax (c) and the disappearance of the pronounced minimum in J(p) is observed, is similar to the critical concentration (ccrit ≈ 230 mg ml−1 ) of the liquid-liquid two-phase region, which is observed at a 15 ◦ C lower temperature, however. In this concentration range, separation into a condensed and a dilute protein solution takes place. Hence, the assumption of (dynamic) clustering taking place in this concentration region at 20 ◦ C seems to be not unreasonable. Moreover, in this concentration range, the concentration dependence of J seems to change sign, becoming more attractive. Increasing the protein concentration to 366 mg ml−1 , a concentration is reached, where, within the duration of the experiment (∼3 h), nucleation and partial crystallization of the supersaturated sample takes place.62–66 Figure 6 displays the q-range where Bragg reflections are observed. Indexing of the Bragg reflections indicates a primitive tetragonal unit cell (space group P43 21 2) with axis a = b = 7.9 nm and c = 3.8 nm, which is in good agreement with literature data.67–71 The application of pressure leads to an anisotropic compression along the axis of the unit cell, i.e., a compression of the a axis, whereas the c axis length stays approx-
FIG. 6. Pressure dependent SAXS curves of lysozyme solutions at a concentration of 366 mg ml−1 (T = 20 ◦ C), showing the intermediate q-range where Bragg reflections are observed. With increasing pressure, protein crystal formation is suppressed.
J. Chem. Phys. 141, 22D506 (2014)
imately constant.25 These data are in good agreement with pressure dependent measurements on lysozyme crystals.68 Interestingly, the intensity of the Bragg peaks decreases upon compression, which indicates suppression of protein crystallization upon hydrostatic compression of the solution. (Of note, the pressure-induced decrease of the scattering intensity due to a decrease in scattering contrast upon compression of the solution cannot explain this marked decrease alone.) This observation implies that formation of lysozyme crystals is connected to an overall volume increase of the system. IV. CONCLUSIONS
The influence of pressure on the structure and proteinprotein interaction potential of highly concentrated protein solutions up to about 370 mg ml−1 was studied and analyzed using small-angle X-ray scattering in combination with a liquid-state theoretical approach. The highest concentrations could be obtained using osmotic pressure methology. Under the preparation conditions chosen, the protein solutions where stable against crystallization for days. Only at a protein concentration as high as 366 mg ml−1 , partial crystallization of the metastable protein solution was observed within the time period of the experiment. In the concentration region below 200 mg ml−1 , in agreement with literature data,22 the interaction parameters of the lysozyme solutions are affected by pressure in a nonlinear way, which is probably due to significant changes in the structural properties of bulk water that occur above 1–2 kbar, i.e., is a solvent-mediated effect. In this pressure regime, where the coordination number of water increases markedly due to a collapsed second hydration shell, protein-protein interactions are modified. Interestingly, a peculiar pressure dependence has also been observed in transport and related thermodynamic properties of bulk water, e.g., shallow maxima in the diffusion constant and configurational entropy around 1.5 kbar at room temperature.72, 73 Conversely, for higher concentrated protein solutions, exceeding ∼200 mg ml−1 , no pronounced minimum in J(p) is observed anymore, and the interaction potential turns rather insensitive to compression. Our data show that in this pressure range of 200–366 mg ml−1 , bulk-water-mediated effects on protein-protein interactions as observed in the lower protein concentration range seem to have largely vanished. This might not be too surprising, as in this concentration range – assuming a homogeneous distribution of protein particles – intermolecular distances have decreased to 5 nm at c = 200 mg ml−1 and 4 nm at c = 366 mg ml−1 , respectively. Taking the size of the lysozyme molecule into account, this translates into water layers between adjacent protein molecules of about 4 and 1, only. For such thin hydration layers, whose structural properties are also influenced by the surface properties of the protein molecules (e.g., by electrostriction effects upon hydration of surface charges), no bulk-like solvent properties can be expected anymore. Hence, the structural properties of bulk water that influence the spatial organization of the more dilute protein solutions upon compression, cannot be as effective anymore for protein concentrations exceeding 200 mg ml−1 . Consequently, no
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 128.248.155.225 On: Fri, 21 Nov 2014 19:31:34
22D506-7
Grobelny et al.
pronounced minimum in J(p) is observed. This is also the concentration range, where the intermolecular interaction potential does not become repulsive anymore with increasing protein concentration, and the concentration dependence of J seems to even change sign at the highest protein concentration, becoming more attractive. However, these data have to be taken with care, as the DLVO-potential is expected to be less suitable for describing protein-protein interactions at these very high protein concentrations, where additional interaction terms including many-body forces might come into play. The absence of a SAXS intensity increase for q → 0 shows that no irreversible aggregation of lysozyme takes place. Rather, – possibly transient (dynamic) – protein clustering takes place, which is also apparent from the maximum of qmax reached in this concentration range. Future NSE experiments in this concentration range might support these conclusions. Our results also show that pressure suppresses protein nucleation, aggregation, and finally crystallization in supersaturated condensed lysozyme solutions. These findings are thus also of importance for understanding the high stability of concentrated protein solutions (as they occur intracellularly) in biosystems thriving under hydrostatic pressure conditions such as in the deep sea, where pressures up to the kbar-level and beyond are reached.74, 75 In the crowded cell, where highly concentrated protein solutions (200–300 mg ml−1 )1, 76–78 are encountered and macromolecular crowding reaches levels of 30 vol.% and more, pressure is expected to effectively suppress protein aggregation. ACKNOWLEDGMENTS
We acknowledge the Diamond Light Source for providing synchrotron radiation. Financial support from the DFG Research Unit FOR 1979 and in part of the Cluster of Excellence RESOLV (EXC 1069) funded by the Deutsche Forschungsgemeinschaft (DFG) is gratefully acknowledged. J.M. thanks the BMBF (05K10PEC) for financial support. 1 A.
P. Minton, J. Cell Sci. 119, 2863 (2006). J. Ellis, Adv. Exp. Med. Biol. 594, 1 (2007). 3 S. Zhao, H. Gies, and R. Winter, Z. Phys. Chem. 221, 139 (2007). 4 M. Erlkamp, S. Grobelny, and R. Winter, Phys. Chem. Chem. Phys. 16, 5965 (2014). 5 C. Haas and J. Drenth, J. Cryst. Growth 196, 388 (1999). 6 N. Asherie, Methods 34, 266 (2004). 7 R. Ravindra, S. Zhao, H. Gies, and R. Winter, J. Am. Chem. Soc. 126, 12224 (2004). 8 A. Stradner, H. Sedgwick, F. Cardinaux, W. C. K. Poon, S. U. Egelhaaf, and P. Schurtenberger, Nature (London) 432, 492 (2004). 9 A. C. Dumetz, A. M. Chockla, E. W. Kaler, and A. M. Lenhoff, Biophys. J. 94, 570 (2008). 10 R. Mishra and R. Winter, Angew. Chem., Int. Ed. 47, 6518 (2008). 11 J. Seeliger, A. Werkmüller, and R. Winter, PLoS One 8, e69652 (2013). 12 J. D. Gunton, A. Shiryayev, D. L. Pagan, Protein Condensation (Cambridge University Press, Cambridge, 2007). 13 N. Javid, K. Vogtt, C. Krywka, M. Tolan, and R. Winter, ChemPhysChem 8, 679 (2007). 14 V. G. Taratuta, A. Holschbach, G. M. Thurston, D. Blankschtein, and G. B. Benedek, J. Phys. Chem. 94, 2140 (1990). 15 M. Muschol and F. Rosenberger, J. Chem. Phys. 107, 1953 (1997). 16 A. Ducruix, J. Guilloteau, M. Riès-Kautt, and A. Tardieu, J. Cryst. Growth 168, 28 (1996). 17 A. Tardieu, A. Le Verge, M. Malfois, F. Bonnete, S. Finet, M. Ries-Kautt, and L. Belloni, J. Cryst. Growth 196, 193 (1999). 2 R.
J. Chem. Phys. 141, 22D506 (2014) 18 J.
Narayanan and X. Y. Liu, Biophys. J. 84, 523 (2003). Pellicane, D. Costa, and C. Caccamo, J. Phys. Condens. Matter 15, 375 (2003). 20 G. Pellicane, D. Costa, and C. Caccamo, J. Phys. Condens. Matter 16, S4923 (2004). 21 M. Niebuhr and M. H. J. Koch, Biophys. J. 89, 1978 (2005). 22 M. Schroer, J. Markgraf, D. Wieland, C. Sahle, J. Möller, M. Paulus, M. Tolan, and R. Winter, Phys. Rev. Lett. 106, 178102 (2011). 23 M. Schroer, Y. Zhai, D. C. F. Wieland, C. J. Sahle, J. Nase, M. Paulus, M. Tolan, and R. Winter, Angew. Chem., Int. Ed. 50, 11413 (2011). 24 M. Schroer, M. Tolan, and R. Winter, Phys. Chem. Chem. Phys. 14, 9486 (2012). 25 J. Möller, M. Schroer, M. Erlkamp, S. Grobelny, M. Paulus, S. Tiemeyer, F. J. Wirkert, M. Tolan, and R. Winter, Biophys. J. 102, 2641 (2012). 26 A. J. Chinchalikar, V. K. Aswal, J. Kohlbrecher, and A. G. Wagh, Phys. Rev. E 87, 062708 (2013). 27 J. Möller, S. Grobelny, J. Schulze, S. Bieder, A. Steffen, M. Erlkamp, M. Paulus, M. Tolan, and R. Winter, Phys. Rev. Lett. 112, 028101 (2014). 28 A. Stradner, F. Cardinaux, and P. Schurtenberger, J. Phys. Chem. B 110, 21222 (2006). 29 F. Zhang, M. W. A. Skoda, R. M. J. Jacobs, R. A. Martin, C. M. Martin, and F. Schreiber, J. Phys. Chem. B 111, 251 (2007). 30 M. Carpineti and R. Piazza, Phys. Chem. Chem. Phys. 6, 1506 (2004). 31 F. Meersman, I. Daniel, D. H. Bartlett, R. Winter, R. Hazael, and P. F. McMillan, Rev. Mineral. Geochem. 75, 607 (2013). 32 S. Bachali, X. Bailly, J. Jolles, P. Jolles, and J. S. Deutsch, Eur. J. Biochem. 271, 237 (2004). 33 S. Bachali, M. Jager, A. Hassanin, F. Schoentgen, P. Jollès, A. FialaMedioni, and J. S. Deutsch, J. Mol. Evol. 54, 652 (2002). 34 J. Jollès, A. Fiala-Médioni, and P. Jollès, J. Mol. Evol. 43, 523 (1996). 35 A. Bouchoux, P.-E. Cayemitte, J. Jardin, G. Gésan-Guiziou, and B. Cabane, Biophys. J. 96, 693 (2009). 36 Y. Kitamura and T. Itoh, J. Sol. Chem. 16, 715 (1987). 37 R. F. Steiner, Biochim. Biophys. Acta 79, 51 (1964). 38 N. J. Brooks, B. L. L. E. Gauthe, N. J. Terrill, S. E. Rogers, R. H. Templer, O. Ces, and J. M. Seddon, Rev. Sci. Instrum. 81, 064103 (2010). 39 A. P. Hammersley, S. O. Svensson, M. Hanfland, A. N. Fitch, and D. Häusermann, High Press. Res. 14, 235 (1996). 40 Y. Liu, W.-R. Chen, and S.-H. Chen, J. Chem. Phys. 122, 044507 (2005). 41 B. V. Derjaguin and L. V. Landau, Acta Physicochim. USSR 14, 633 (1941). 42 E. J. W. Verwey, J. Chem. Phys. 51, 631 (1947). 43 E. J. W. Verwey and J. T. G. Overbeek, Theory of the Stability of Lyophobic Colloids (Elsevier, New York, 1948). 44 W. B. Floriano and M. A. C. Nascimento, Braz. J. Phys. 34, 38 (2004). 45 D. E. Kuehner, J. Engmann, F. Fergg, M. Wernick, H. W. Blanch, and J. M. Prausnitz, J. Phys. Chem. B 103, 1368 (1999). 46 D. I. Svergun and M. H. Koch, Rep. Prog. Phys. 66, 1735 (2003). 47 D. I. Svergun, S. Richard, M. H. J. Koch, Z. Sayers, S. Kuprin, and G. Zaccai, Proc. Natl. Acad. Sci. U.S.A. 95, 2267 (1998). 48 L. Smeller, F. Meersman, and K. Heremans, Biochim. Biophys. Acta 1764, 497 (2006). 49 T. Zemb and P. Lindner, Neutron, X-Rays and Light. Scattering Methods Applied to Soft Condensed Matter (North Holland, Amsterdam, 2002). 50 A. Soper and M. Ricci, Phys. Rev. Lett. 84, 2881 (2000). 51 J. Crocker and D. Grier, Phys. Rev. Lett. 77, 1897 (1996). 52 A. E. Larsen and D. G. Grier, Nature (London) 385, 230 (1997). 53 J. J. Gray, B. Chiang, and R. T. Bonnecaze, Nature (London) 402, 750 (1999). 54 A. Stradner, F. Cardinaux, S. U. Egelhaaf, and P. Schurtenberger, Proc. Natl. Acad. Sci. U.S.A. 105, E75 (2008). 55 A. Stradner, F. Cardinaux, and P. Schurtenberger, Phys. Rev. Lett. 96, 219801 (2006). 56 A. Shukla, E. Mylonas, E. Di Cola, S. Finet, P. Timmins, T. Narayanan, and D. I. Svergun, Proc. Natl. Acad. Sci. U.S.A. 105, 5075 (2008). 57 A. Shukla, E. Mylonas, E. Di Cola, S. Finet, P. Timmins, T. Narayanan, and D. I. Svergun, Proc. Natl. Acad. Sci. U.S.A. 105, E76 (2008). 58 L. Porcar, P. Falus, W.-R. Chen, A. Faraone, E. Fratini, K. Hong, P. Baglioni, and Y. Liu, J. Phys. Chem. Lett. 1, 126 (2010). 59 F. Cardinaux, E. Zaccarelli, A. Stradner, S. Bucciarelli, B. Farago, S. U. Egelhaaf, F. Sciortino, and P. Schurtenberger, J. Phys. Chem. B 115, 7227 (2011). 60 Y. Liu, L. Porcar, J. Chen, W.-R. Chen, P. Falus, A. Faraone, E. Fratini, K. Hong, and P. Baglioni, J. Phys. Chem. B 115, 7238 (2011). 19 G.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 128.248.155.225 On: Fri, 21 Nov 2014 19:31:34
22D506-8 61 W.
Grobelny et al.
Pan, P. G. Vekilov, and V. Lubchenko, J. Phys. Chem. B 114, 7620 (2010) 62 L. F. Filobelo, O. Galkin, and P. G. Vekilov, J. Chem. Phys. 123, 014904 (2005). 63 E. Cacioppo and M. L. Pusey, J. Cryst. Growth 114, 286 (1991). 64 F. Rosenberger, S. B. Howard, J. W. Sowers, and T. A. Nyce, J. Cryst. Growth 129, 1 (1993). 65 W. C. K. Poon, S. U. Egelhaaf, P. A. Beales, A. Salonen, and L. Sawyer, J. Phys. Condens. Matter 12, L569 (2000). 66 M. Muschol and F. Rosenberger, J. Chem. Phys. 103, 10424 (1995). 67 C. C. F. Blake, G. A. Mair, A. C. T. North, Proc. R. Soc. London, Ser. B 167, 365 (1967). 68 R. Fourme, R. Kahn, M. Mezouar, E. Girard, C. Hoerentrup, T. Prangé, and I. Ascone, J. Synchrotron Radiat. 8, 1149 (2001).
J. Chem. Phys. 141, 22D506 (2014) 69 C.
Sauter, F. Otalora, J. Gavira, O. Vidal, R. Giege, and J. Garcia-Ruiz, Acta Crystallogr., Sect. D 57, 1119 (2001). 70 Y. Cheng, R. F. Lobo, S. Sandler, and A. M. Lenhoff, Biotechnol. Bioeng. 94, 177 (2006). 71 C. Kundrot and F. M. Richards, J. Appl. Crystallogr. 19, 208 (1986). 72 F. Mallamace, C. Corsaro, D. Mallamace, C. Vasi, and H. E. Stanley, Faraday Discuss. 167, 95 (2013). 73 R. Ludwig, Angew. Chem., Int. Ed. 40, 1808 (2001). 74 I. Daniel, P. Oger, and R. Winter, Chem. Soc. Rev. 35, 858 (2006). 75 F. M. Lauro and D. H. Bartlett, Extremophiles 12, 15 (2008). 76 B. van den Berg, R. Wain, C. M. Dobson, and R. J. Ellis, EMBO J. 19, 3870 (2000). 77 G. Rivas, F. Ferrone, and J. Herzfeld, EMBO Rep. 5, 23 (2004). 78 L. Stagg, S.-Q. Zhang, M. S. Cheung, and P. Wittung-Stafshede, Proc. Natl. Acad. Sci. U.S.A. 104, 18976 (2007).
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 128.248.155.225 On: Fri, 21 Nov 2014 19:31:34
