Effects of naturally occurring osmolytes on protein stability and solubility: issues important in protein crystallization

Protein solubility and stability are issues of consideration in attempts to crystallize proteins. These two properties of proteins are also at issue in the cells of organisms that have adapted to water stress conditions that could ordinarily denature or inactivate some proteins. Most organisms that have adapted to environmental stresses have done so by production and accumulation of certain small organic molecules, known as osmolytes, that arose by natural selection and have the ability to stabilize intracellular proteins against the environmental stress. Here, concepts developed to understand the special properties of the naturally occurring osmolytes in effecting protein stability and solubility, and the principles that have come from studies of these compounds have been presented. Along with excluded volume and preferential interaction parameters, identification of the osmophobic effect and the attenuation of this effect by favorable interactions of solute with side-chains appear to contribute to the full set of effects protecting osmolytes have on protein stability and solubility. With these concepts in mind and the fact that urea interacts favorably with the peptide backbone we note that: (1) osmolyte-induced effects on protein stability ranging from denaturation to forcing proteins to fold can be achieved experimentally and the underlying principles understood at near molecular-level detail, and (2) osmolyte-mediated solubility effects ranging from protein precipitation to protein solubilization are predictable based on these principles. These effects are contrasted and compared with effects of 2-methyl-2,4-pentanediol and polyethylene glycol on proteins, and how the principles found for the naturally occurring osmolytes can be applied to these two commonly used protein crystallizing agents.     

Evaluation of selectivity changes in HIC systems using a preferential interaction based analysis

It is well established that salt enhances the interaction between solutes (e.g., proteins, displacers) and the weak hydrophobic ligands in hydrophobic interaction chromatography (HIC) and that various salts (e.g., kosmotropes, chaotropes, and neutral) have different effects on protein retention. In this article, the solute affinity in kosmotropic, chaotropic, and neutral mobile phases are compared and the selectivity of solutes in the presence of these salts is examined. Since solute binding in HIC systems is driven by the release of water molecules, the total number of released water molecules in the presence of various types of salts was calculated using the preferential interaction theory. Chromatographic retention times and selectivity reversals of both proteins and displacers were found to be consistent with the total number of released water molecules. Finally, the solute surface hydrophobicity was also found to have a significant effect on its retention in HIC systems.     

Compatible solutes as protectants for zymogens against proteolysis

Compatible solutes are small organic osmoprotectants that have the capability to stabilize proteins. In coupled assays, the effect of the solutes ectoine, hydroxyectoine and betaine on the activation of the zymogens trypsinogen and chymotrypsinogen, catalyzed by enteropeptidase and trypsin, respectively, was studied. To different extents, all solutes protected the zymogens against activation. Ectoine (800 mM) was the most potent solute in reducing the formation of trypsin to 4% of the control value and of chymotrypsin to 23%. In separate experiments, the ability of the solutes to preserve proteolytic activity during incubation was investigated. After 4 h, trypsin and chymotrypsin completely lost their activity, but in the presence of ectoine, approximately 50% residual activity was maintained. It is proposed that a conformational shift of the protein towards folded, native-like states induced by preferential exclusion of the solute is responsible for the stabilizing and chaperone-like effects.     

Static light scattering from concentrated protein solutions, I: General theory for protein mixtures and application to self-associating proteins

Exact expressions for the static light scattering of a solution containing up to three species of point-scattering solutes in highly nonideal solutions at arbitrary concentration are obtained from multicomponent scattering theory. Explicit expressions for thermodynamic interaction between solute molecules, required to evaluate the scattering relations, are obtained using an equivalent hard particle approximation similar to that employed earlier to interpret scattering of a single protein species at high concentration. The dependence of scattering intensity upon total protein concentration is calculated for mixtures of nonassociating proteins and for a single self-associating protein over a range of concentrations up to 200 g/l. An approximate semiempirical analysis of the concentration dependence of scattering intensity is proposed, according to which the contribution of thermodynamic interaction to scattering intensity is modeled as that of a single average hard spherical species. Simulated data containing pseudo-noise comparable in magnitude to actual experimental uncertainty are modeled using relations obtained from the proposed semiempirical analysis. It is shown that by using these relations one can extract from the data reasonably reliable information about underlying weak associations that are manifested only at very high total protein concentration.     

Solute concentration effect on osmotic reflection coefficient

A theory for the effect of concentration on osmotic reflection coefficient, correct to first order, was developed at the molecular level by considering the effect of solute-solute interactions on solute concentration and the fluid stress tensor within a solvent-filled pore. The solvent was modeled as a continuous fluid and potential energies between solute molecules and the pore wall were assumed to be pairwise additive. Although the theory is more general, calculations are presented only for excluded volume effects (hard-sphere for solute, hard-wall for pore). The relationship between the first-order concentration effect and the infinite dilution value of reflection coefficient appears to be geometry independent. The theory is discussed in light of experimental studies of osmotic flow that have recently appeared in the literature.     

Application of the local-bulk partitioning and competitive binding models to interpret preferential interactions of glycine betaine and urea with protein surface

Two thermodynamic models have been developed to interpret the preferential accumulation or exclusion of solutes in the vicinity of biopolymer surface and the effects of these solutes on protein processes. The local-bulk partitioning model treats solute (and water) as partitioning between the region at/or near the protein surface (the local domain) and the bulk solution. The solvent exchange model analyzes a 1:1 competition between water and solute molecules for independent surface sites. Here we apply each of these models to interpret thermodynamic data for the interactions of urea and the osmoprotectant glycine betaine (N,N,N-trimethylglycine; GB) with the surface exposed in unfolding the marginally stable lacI HTH DNA binding domain. The partition coefficient K(P) quantifying accumulation of urea at this protein surface (K(P) approximately equal 1.1) is only weakly dependent on urea concentration up to 6 M urea. However, K(P) quantifying exclusion of GB from the vicinity of this protein surface increases from 0.83 (extrapolated to 0 M GB) to 1.0 (indicating that local and bulk GB concentrations are equal) at 4 M GB (activity > 40 M). We interpret the significant concentration dependence of K(P) for GB, predicted to be general for excluded, nonideal solutes such as GB, as a modest (8%) attenuation of the GB concentration dependence of solute nonideality in the local domain relative to that in the bulk solution. Above 4 M, K(P) for the interaction of GB with the surface exposed in protein unfolding is predicted to exceed unity, which explains the maximum in thermal stability observed for RNase and lysozyme at 4 M GB (Santoro, M. M., Liu, Y. F., Khan, S. M. A., Hou, L. X., and Bolen, D. W. (1992) Biochemistry 31, 5278-5283). Both thermodynamic models provide good two-parameter fits to GB and urea data for lacI HTH unfolding over a wide concentration range. The solute partitioning model allows for a full spectrum of attenuation effects in the local domain, encompasses the cases treated by the competitive binding model, and provides a somewhat better two-parameter fit of effects of high GB concentration on lacI HTH stability. Parameters of this fit should be applicable to isothermal and thermal unfolding data for all proteins with similar compositions of surface exposed in unfolding.     

Effect of the compatible solute ectoine on the stability of the membrane proteins

Mechanical single molecule techniques offer exciting possibilities for investigating protein folding and stability in native environments at sub-nanometer resolutions. Compatible solutes show osmotic activity which even at molar concentrations do not interfere with cell metabolism. They are known to protect proteins against external stress like temperature, high salt concentrations and dehydrating conditions. We studied the impact of the compatible solute ectoine (1M) on membrane proteins by analyzing the mechanical properties of Bacteriorhodopsin (BR) in its presence and absence by single molecule force spectroscopy. The unfolding experiments on BR revealed that ectoine decreases the persistence length of its polypeptide chain thereby increasing its tendency to coil up. In addition, we found higher unfolding forces indicating strengthening of those intra molecular interactions which are crucial for stability. This shows that force spectroscopy is well suited to study the effect of compatible solutes to stabilize membrane proteins against unfolding. In addition, it may lead to a better understanding of their detailed mechanism of action.     

Compatible solute effects on thermostability of glutamine synthetase and aspartate transcarbamoylase from Methanococcus jannaschii

Methanococcus jannaschii accumulates alpha- and beta-glutamate as osmolytes. The effect of these and other solutes on the thermostability of two multisubunit metabolic enzymes from M. jannaschii, aspartate transcarbamoylase catalytic trimer (ATCase C3) and glutamine synthetase (GS), has been measured and compared to solute effects on bacterial mesophilic counterparts in order to explore if osmolytes accumulated by each organism can preferentially stabilize the proteins to thermal unfolding. For both ATCase enzymes and for the B. subtilis GS, the solutes normally accumulated by the organism were very effective in protecting the enzyme from losing activity at high temperatures, although solute effects on loss of secondary structure did not necessarily correlate with this thermoprotection of activity. The recombinant M. jannaschii GS exhibited quite different behavior. The pure enzyme had a thermal unfolding transition with a midpoint temperature (Tm) less than 60 degrees C, well under the growth temperature of the organism (85 degrees C). None of the small molecule solutes tested (including the K+-glutamate isomers accumulated by M. jannaschii) significantly stabilized the protein to incubation at 85 degrees C. Instead, protein-protein interactions, as illustrated by E. coli GroEL or ribosomal protein L2 stabilization of GS, appeared to be the dominant factor in stabilizing this archaeal enzyme at the growth temperature.     

Crowding effects in binary mixtures of rod-like and spherical particles

Crowding effects relevant to the phase stability of binary mixtures of rod-like and spherical particles are investigated by means of Monte Carlo simulations in the isobaric NPT ensemble. The two types of particles are represented, respectively, by freely rotating hard spherocylinders of a moderate aspect ratio (L/sigma = 5) and hard spheres of the same diameter sigma. Molar fractions of spheres ranging xHS = 0.00-0.37 are considered with the aim of characterizing the crowding effects on the liquid crystal phases of the hard spherocylinder fluid induced by the spherical component as depleting agent. We find that the addition of the spherical crowder is beneficial for the stabilization of the layers of the rod-like particles characteristic of the smectic phase. On the contrary, the addition of spheres has a negative impact upon the stability of the nematic phase, where the rod-like particles tend to align collectively parallel to each other. Interestingly, the spheres tend to arrange forming rod-like clusters in the nematic phase and lamellar structures in the smectic phase, which is compensated by the entropy gained by the spherocylinder particles in each phase. The main results are in qualitative agreement with recent experimental and theoretical studies and serve to test the prediction of current equations of state for these types of binary mixtures.     

Measurement of thermodynamic nonideality arising from volume-exclusion interactions between proteins and polymers

The effective thermodynamic radii of 23 ribosomal proteins from the 50 S subunit have been determined by gel chromatography on Sephadex G-50, thereby supporting the contention that most of the proteins of the 50 S ribosomal unit exhibit reasonably globular structures. To investigate further the usefulness of modelling proteins as spheres, the second virial coefficient describing excluded volume interactions of some ribosomal proteins with two inert polymers, polyethylene glycol (PEG) and dextran, has been determined by gel chromatography and/or sedimentation equilibrium techniques. Protein-polymer excluded volumes obtained with PEG 20000 and Dextran T70 as the space-filling solute are shown to conform reasonably well with a quantitative expression describing interaction between an impenetrable sphere and an ideal Brownian path (K.M. Jansons and C.G. Phillips, J. Colloid Interface Sci., 137 (1990) 75).     

Cation-pi interactions as determinants for binding of the compatible solutes glycine betaine and proline betaine by the periplasmic ligand-binding protein ProX from Escherichia coli

Compatible solutes such as glycine betaine and proline betaine are accumulated to exceedingly high intracellular levels by many organisms in response to high osmolarity to offset the loss of cell water. They are excluded from the immediate hydration shell of proteins and thereby stabilize their native structure. Despite their exclusion from protein surfaces, the periplasmic ligand-binding protein ProX from the Escherichia coli ATP-binding cassette transport system ProU binds the compatible solutes glycine betaine and proline betaine with high affinity and specificity. To understand the mechanism of compatible solute binding, we determined the high resolution structure of ProX in complex with its ligands glycine betaine and proline betaine. This crystallographic study revealed that cation-pi interactions between the positive charge of the quaternary amine of the ligands and three tryptophan residues forming a rectangular aromatic box are the key determinants of the high affinity binding of compatible solutes by ProX. The structural analysis was combined with site-directed mutagenesis of the ligand binding pocket to estimate the contributions of the tryptophan residues involved in binding.     

Excluded volume in solvation: sensitivity of scaled-particle theory to solvent size and density

Changes in solvent environment greatly affect macromolecular structure and stability. To investigate the role of excluded volume in solvation, scaled-particle theory is often used to calculate delta G(tr)(ev), the excluded-volume portion of the solute transfer free energy, delta G(tr). The inputs to SPT are the solvent radii and molarities. Real molecules are not spheres. Hence, molecular radii are not uniquely defined and vary for any given species. Since delta G(tr)(ev) is extremely sensitive to solvent radii, uncertainty in these radii causes a large uncertainty in delta G(tr)(ev)-several kcal/mol for amino acid solutes transferring from water to aqueous mixtures. This uncertainty is larger than the experimental delta G(tr) values. Also, delta G(tr)(ev) can be either positive or negative. Adding neutral crowding molecules may not necessarily reduce solubility. Lastly, delta G(tr)(ev) is very sensitive to solvent density, rho. A few percent error in rho may even cause qualitative deviations in delta G(tr)(ev). For example, if rho is calculated by assuming the hard-sphere pressure to be constant, then delta G(tr)(ev) values and uncertainties are now only tenths of a kcal/mol and are positive. Because delta G(tr)(ev) values calculated by scaled-particle theory are strongly sensitive to solvent radii and densities, determining the excluded-volume contribution to transfer free energies using SPT may be problematic.     

Effects of protein perturbants on phospholipid bilayers

Series of alcohols, amides, ureas, and sulfoxides with increasingly longer hydrocarbon chains have been shown to lower progressively the thermal denaturation temperature of proteins. This effect is presumably due to a hydrophobic interaction between the solute and nonpolar domains of the protein. Theoretically, these interactions should occur between the solute and any macromolecular structure having a nonpolar region to which the solute has access. A recent review by Arakawa et al. has summarized evidence for such an interaction between organic solutes and proteins and suggested that these interactions are favored at higher temperatures. The present study investigates the effects of several classes of compounds on the stability of phospholipid vesicles. The results show that many compounds that are known to perturb protein function also destabilize phospholipid bilayers as reflected by solute-induced loss of vesicle contents.     

A review of experimental measurements of effective diffusive permeabilities and effective diffusion coefficients in biofilms

Experimental measurements of effective diffusive permeabilities and effective diffusion coefficients in biofilms are reviewed. Effective diffusive permeabilities, the parameter appropriate to the analysis of reaction-diffusion interactions, depend on solute type and biofilm density. Three categories of solute physical chemistry with distinct diffusive properties were distinguished by the present analysis. In order of descending mean relative effective diffusive permeability (De/Daq) these were inorganic anions or cations (0.56), nonpolar solutes with molecular weights of 44 or less (0.43), and organic solutes of molecular weight greater than 44 (0.29). Effective diffusive permeabilities decrease sharply with increasing biomass volume fraction suggesting a serial resistance model of diffusion in biofilms as proposed by Hinson and Kocher (1996). A conceptual model of biofilm structure is proposed in which each cell is surrounded by a restricted permeability envelope. Effective diffusion coefficients, which are appropriate to the analysis of transient penetration of nonreactive solutes, are generally similar to effective diffusive permeabilities in biofilms of similar composition. In three studies that examine diffusion of very large molecular weight solutes (>5000) in biofilms, the average ratio of the relative effective diffusion coefficient of the large solute to the relative effective diffusion coefficient of either sucrose or fluorescein was 0.64, 0.61, and 0.36. It is proposed that large solutes are effectively excluded from microbial cells, that small solutes partition into and diffuse within cells, and that ionic solutes are excluded from cells but exhibit increased diffusive permeability (but decreased effective diffusion coefficients) due to sorption to the biofilm matrix.     

Excluded volume effects on the partition of single- and double-stranded oligodeoxynucleotides between two liquid phases.

The distribution coefficients of single- and double-stranded oligodeoxynucleotides in a PEG 8000/phosphate two-phase system are a function of their chain length. Values of the distribution coefficients are in general agreement with a simple extension of a model for excluded volume effects (the "available volume model") which was applied previously to the distribution of proteins in this system. The current results therefore provide a second set of examples for molecules of very different geometry where the distribution added molecules is controlled by excluded volume interactions between those molecules and the PEG 8000 of the two-phase system.     

Toward an accurate theoretical framework for describing ensembles for proteins under strongly denaturing conditions

Our focus is on an appropriate theoretical framework for describing highly denatured proteins. In high concentrations of denaturants, proteins behave like polymers in a good solvent and ensembles for denatured proteins can be modeled by ignoring all interactions except excluded volume (EV) effects. To assay conformational preferences of highly denatured proteins, we quantify a variety of properties for EV-limit ensembles of 23 two-state proteins. We find that modeled denatured proteins can be best described as follows. Average shapes are consistent with prolate ellipsoids. Ensembles are characterized by large correlated fluctuations. Sequence-specific conformational preferences are restricted to local length scales that span five to nine residues. Beyond local length scales, chain properties follow well-defined power laws that are expected for generic polymers in the EV limit. The average available volume is filled inefficiently, and cavities of all sizes are found within the interiors of denatured proteins. All properties characterized from simulated ensembles match predictions from rigorous field theories. We use our results to resolve between conflicting proposals for structure in ensembles for highly denatured states.     

Explaining glomerular pores with fiber matrices. A visualization study based on computer modeling.

The extracellular space of the glomerular capillary wall is occupied by a complex meshwork of fibrous molecules. Little is understood about how the size, shape, and charge recognition properties of glomerular ultrafiltration arise from this space-filling fiber matrix. We studied the problem of size recognition by visualizing the void volume accessible to hard spheres in computer-generated three-dimensional homogeneous random fiber matrices. The spatial organization of the void volume followed a complex "blob-and-throat" pattern in which circumscribed cavities of free space within the matrix ("blobs") were joined to adjacent cavities by narrower throats of void space. For sufficiently small solutes, chains of blobs and throats traversed the matrix, providing pathways for trans-matrix permeation. The matrices showed threshold or gating properties with respect to permeation: solutes whose radius exceeded a critical value, at which a throat on the last connected trans-matrix pathway pinched off, could not cross, whereas smaller solutes had nonzero permeability. The thresholds may give the glomerular fiber matrix porelike response properties and explain why pore models have been such a useful means of treating permselectivity.     

Counteracting effects of renal solutes on amyloid fibril formation by immunoglobulin light chains

In primary (light chain-associated) amyloidosis, immunoglobulin light chains deposit as amyloid fibrils in vital organs, especially the kidney. Because the kidney contains high concentrations of urea that can destabilize light chains as well as solutes such as betaine and sorbitol that serve as protein stabilizers, we investigated the effects of these solutes on in vitro amyloid fibril formation and thermodynamic stability of light chains. Two recombinant light chain proteins, one amyloidogenic and the other nonamyloidogenic, were used as models. For both light chains, urea enhanced fibril formation by reducing the nucleation lag time and diminished protein thermodynamic stability. Conversely, betaine or sorbitol increased thermodynamic stability of the proteins and partially inhibited fibril formation. These solutes also counteracted urea-induced reduction in protein thermodynamic stability and accelerated fibril formation. Betaine was more effective than sorbitol. A model is presented to explain how the thermodynamic effects of the solutes on protein state equilibria can alter nucleation lag time and, hence, fibril formation kinetics. Our results provide evidence that renal solutes control thermodynamic and kinetic stability of light chains and thus may modulate amyloid fibril formation in the kidney.