Combinatorial formulation of biocatalyst preparations for increased activity in organic solvents: salt activation of penicillin amidase

A combinatorial experimental technique was used to identify salts and salt mixtures capable of activating penicillin amidase in organic solvents for the transesterification of phenoxyacetate methyl ester with 1-propanol. Penicillin amidase was lyophilized in the presence of various chloride and acetate salts within 96-deep-well plates and catalytic rates measured to determine lead candidates for highly salt-activated preparations. The kinetics of the most active formulations were then further evaluated. These studies revealed that a formulation consisting of 98% (w/w) of a 1:1 KAc:CsCl salt mixture, 1% (w/w) enzyme, and 1% (w/w) potassium phosphate buffer was approximately 35,000-fold more active than the salt-free formulation in hexane, as reflected in values of V(max)/K(m). This extraordinary activation could be extended to more polar solvents, including tert-amyl alcohol, and to formulations with lower total salt contents. A correlation was found between the kosmotropic/chaotropic behavior of the salts (as measured by the Jones-Dole B coefficients) and the observed activation. Strongly chaotropic cations combined with strongly kosmotropic anions yielded the greatest activation, and this is likely due to the influence of the ions on protein-water and protein-salt interactions.     

Optimizing the salt-induced activation of enzymes in organic solvents: effects of lyophilization time and water content

The addition of simple inorganic salts to aqueous enzyme solutions prior to lyophilization results in a dramatic activation of the dried powder in organic media relative to enzyme with no added salt. Activation of both the serine protease subtilisin Carlsberg and lipase from Mucor javanicus resulting from lyophilization in the presence of KCl was highly sensitive to the lyophilization time and water content of the sample. Specifically, for a preparation containing 98% (w/w) KCl, 1% (w/w) phosphate buffer, and 1% (w/w) enzyme, varying the lyophilization time showed a direct correlation between water content and activity up to an optimum, beyond which the activity decreased with increasing lyophilization time. The catalytic efficiency in hexane varied as much as 13-fold for subtilisin Carlsberg and 11-fold for lipase depending on the lyophilization time. This dependence was apparently a consequence of including the salt, as a similar result was not observed for the enzyme freeze-dried without KCl. In the case of subtilisin Carlsberg, the salt-induced optimum value of kcat/Km for transesterification in hexane was over 20,000-fold higher than that for salt-free enzyme, a substantial improvement over the previously reported enhancement of 3750-fold (Khmelnitsky, 1994). As was found previously for pure enzyme, the salt-activated enzyme exhibited greatest activity when lyophilized from a solution of pH equal to the pH for optimal activity in water. The active-site content of the lyophilized enzyme samples also depended upon lyophilization time and inclusion of salt, with opposite trends in this dependence observed for the solvents hexane and tetrahydrofuran. Finally, substrate selectivity experiments suggested that mechanism(s) other than selective partitioning of substrate into the enzyme-salt matrix are responsible for salt-induced activation of enzymes in organic solvents.     

Dramatic enhancement of enzymatic activity in organic solvents by lyoprotectants

When seven different hydrolytic enzymes (four proteases and three lipases) were lyophilized from aqueous solution containing a ligand, N-Ac-L-Phe-NH(2), their catalytic activity in anhydrous solvents was far greater (one to two orders of magnitude) than that of the enzymes lyophilized without the ligand. This ligand-induced activation was expressed regardless of whether the substrate employed in organic solvents structurally resembled the ligand. Furthermore, nonligand lyoprotectants [sorbitol, other sugars, and poly(ethylene glycol)] also dramaticaliy enhanced enzymatic activity in anhydrous solvents when present in enzyme aqueous solution prior to lyophilization. The effects of the ligand and of the lyoprotectants were nonadditive, suggesting the same mechanism of action. Excipient activated and nonactivated enzymes exhibited identical activities in water. Also, addition of the excipients directly to suspensions of nonactivated enzymes in organic solvents had no appreciable effect on catalytic activity. These observations indicate that the mechanism of the excipient-induced activation is based on the ability of the excipients to alleviate reversible denaturation of enzymes upon lyophilization. Activity enhancement induced by the excipients is displayed even after their removal by washing enzymes with anhydrous solvents. Subtilisin Carlsberg, lyophilized with sorbitol, was found to be a much more efficient practical catalyst than its "regular" counterpart. (c) 1993 John Wiley & Sons, Inc.     

Hofmeister effects on activity and stability of alkaline phosphatase

We have studied the effects on alkaline phosphatase of adding high concentrations (normally 1.0 M) of simple salts. It is necessary to allow for significant effects of salts on the extinction coefficient of the reaction product, and on the apparent pH of the buffer. Both activity and stability of the enzyme correlate well with the Hofmeister series in terms of the salt's kosmotropic/chaotropic properties, which are assessed by the Jones–Dole viscosity B coefficients (B₊ for cations and B₋ for anions). The catalytic activity or V_max/K_m of the enzyme showed a bell-shaped relationship with the (B₋ − B₊) values of the salts present, being optimal with salts (such as NaCl, KCl, and KNO₃) where the anion and cation have similar kosmotropic/chaotropic properties. This effect is believed to be enzyme-specific and relates to the impact of both cations and anions on the enzyme's surface pH, active site, and catalytic mechanism. Anions play a more predominant role than cations in affecting enzyme stability. The rate of irreversible thermal inactivation is strongly reduced by addition of kosmotropic anions like SO₄²⁻ (half-life increased from 8 to 580 min at 60 °C). This effect is general and the mechanism probably involves the ability of the ions to affect the water solvation layer around the enzyme molecule and to interact with both the surface and internal structure of the enzyme.     

Activation and substrate specificity of caspase-14

Caspase-14 is a developmentally regulated and tissue restricted member of the caspase family present in mammals. It is mainly found in epidermal keratinocytes and has been hypothesized to be involved in a tissue-specific form of cell senescence, leading to the differentiation of keratinocytes that form the cornified cell layer. However, the substrate specificity, activation mechanism, and function of this caspase have yet to be revealed. We report that caspase-14, in contrast to other caspases, is not produced in active form following expression in Escherichia coli but can be activated by high concentrations of kosmotropic salts. Moreover, proteolytic cleavage is also required since the kosmotropic salts were only effective on the cleaved enzyme. We propose that caspase-14 requires proteolytic cleavage within the catalytic domain, followed by dimerization and ordering of mobile active site loops, to generate a competent enzyme. In the presence of kosmotropic salt, we were able to determine the substrate specificities of mouse and human caspase-14. Surprisingly, the substrate preferences for the human and mouse enzyme are dissimilar. The results obtained with human caspase-14 classify this enzyme as a cytokine activator, but the mouse enzyme shows preferences similar to apical apoptotic caspases.     

The Correlation of RNase A Enzymatic Activity with the Changes in the Distance between Nε2-His12 and Nδ1-His119 Upon Addition of Stabilizing and Destabilizing Salts

The effect of stabilizing and destabilizing salts on the catalytic behavior of ribonuclease A (RNase A) was investigated at pH 7.5 and 25°C, using spectrophotometric, viscometric and molecular dynamic methods. The changes in the distance between N_ε2 of His₁₂ and N_δ1 of His₁₁₉ at the catalytic center of RNase A upon the addition of sodium sulfate, sodium hydrogen sulfate and sodium thiocyanate were evaluated by molecular dynamic methods. The compactness and expansion in terms of Stokes radius of RNase A upon the addition of sulfate ions as kosmotropic salts, and thiocyanate ion as a chaotropic salt, were estimated by viscometric measurements. Enzyme activity was measured using cytidine 2′, 3′-cyclic monophosphate as a substrate. The results from the measurements of distances between N_ε2 of His₁₂ and N_δ1 of His₁₁₉ and Stokes radius suggest (i) that the presence of sulfate ions decreases the distance between the catalytic His residues and increases the globular compactness, and (ii) that there is an expansion of the enzyme surface as well as elongation of the catalytic center in the presence of thiocyanate ion. These findings are in agreement with activity measurements.     

Kosmotropic salt activation and substrate specificity of poliovirus protease 3C

Picornaviruses produce a large polyprotein, which is cleaved by virally encoded cysteine peptidases, picornain-2A and -3C. Picornain-3C has characteristics of both the serine peptidase chymotrypsin and the cysteine peptidase papain in that the 3D structure resembles chymotrypsin, but its nucleophile is a cysteine SH rather than a serine OH group. We investigated the specificity of poliovirus picornain-3C (PV3C) protease and the influence of kosmotropic salts on catalytic activity, using FRET peptides related to a cleavable segment of the virus polyprotein. The peptidase activity of PV3C was found to be 100-fold higher in the presence of 1.5 M sodium citrate. This activation was anion-dependent, following the Hofmeister series citrate(3-) > SO4(2-) > HPO4(2-) > acetate- > HCO3(-) > Cl-. The activation appeared to be independent of substrate sequence and arose primarily from an increase in kcat. A shift to higher pH was also observed for the pK1 of the enzyme pH-activity profile. Experiments with the fluorescent probe ANS (1-anilino-8-naphthalene sulfonate) showed that the protease bound the dye in the presence of 1 M sodium citrate but not in its absence or in the presence of 1 M NaCl. Structural changes in PV3C protease were detected using circular dichroism and the thermodynamic data indicated a more organized active site in the presence of sodium citrate. PV3C protease was also activated in D2O, which was added to the activation by citrate. These effects seem to be related to nonspecific interactions between the solvent and the protein. Our data show that the catalytic efficiency of PV3C protease is modulated by the composition of the environment and that this modulation may play a role in the optimal processing of polyprotein for the virus assembly that occurs inside specific vesicles formed in poliovirus-infected cells.     

Effects of organic solvents and substrate binding on trypsin in acetonitrile and hexane media

In this work, we used molecular dynamic (MD) simulation to study trypsin with and without a six-amino-acid peptide bound in three different solvents (water, acetonitrile and hexane) in order to provide molecular information for well understanding the structure and function of enzymes in non-aqueous media. The results show that the enzyme is more compact and less native-like in hexane than in the other two polar solvents. The substrate could stabilize the native protein structure in the two polar media, but not in the non-polar hexane. There are no significant differences in the conformation of the S1 pocket upon the substrate binding in water and acetonitrile media while a reverse behavior is observed in hexane media, implying a possible induced fit binding mechanism in the non-polar media. The substrate binding enhances the stability of catalytic H-bond network since it could expel the solvent molecules from the active site. The enzyme and the substrate appear to be more appropriate to the reactive conformation in the organic solvents compared with aqueous solution. There is much greater substrate binding strength in hexane media than the water and acetonitrile ones since the polar solvent significantly weakens electrostatic interactions, which are observed to be the main driving force to the binding. In addition, some residues of the S1 pocket could remain favorable contribution to the binding despite the solvent change, but with differences in the contribution extent, the number and the type of residues between the three media.

Enzyme activation for organic solvents made easy

Enzymes are highly selective catalysts that perform intricate chemistries at ambient temperatures and pressures. Although water is the solvent of life, it is a poor solvent for most synthetic organic reactions and, therefore, most chemists avoid aqueous solutions for synthetic applications. However, when removed from the aqueous environment and placed in an organic solvent, enzyme activity is reduced greatly. Here, we present a general overview of recent efforts to activate enzymes for use in nonaqueous media, giving particular focus to the use of simple salts as additives that result in significant biocatalytic improvements.     

Enzymatic production of ceramide from sphingomyelin

Due to its major role in maintaining the water-retaining properties of the epidermis, ceramide is of great commercial potentials in cosmetic and pharmaceutical industries such as in hair and skin care products. Chemical synthesis of ceramide is a costly process, and developments of alternative cost-efficient production methods are of great interest. Present study was the first attempt to perform a systematic study on the production of ceramide through enzymatic hydrolysis of sphingomyelin. Sphingomyelin hydrolysis proved to be more efficient in two-phase (water:organic solvent) system than in one-phase (water-saturated organic solvent) system. Among the screened phospholipase C, the Clostridium perfringens enzyme had the highest sphingomyelin conversion rate, with very small temperature dependence. Addition of ethanol to the system markedly enhanced the rate of ceramide formation, and a mixture of ethylacetate:hexane (50:50) was the best organic solvent tested. Other factors such as (NH(4))(2)SO(4), NaCl and CaCl(2) were also tested but excluded for further consideration. On the basis of the initial experiments, the reaction system was optimized using response surface methodology including five factors (enzyme amount, water amount, ethanol amount, reaction time and the hexane ratio of organic solvent). Water content and enzyme amount was shown to have the most significant influence on the hydrolysis reaction in the fitted quadratic model. The efficiency of sphingomyelin hydrolysis was dramatically improved through system evaluation and optimization, with the optimal conditions at 75 min reaction time, 3 Uml(-1) enzyme amount, 6% water amount, 1.8% ethanol amount and 46% hexane in ethylacetate.     

Effects of water content on the tetrahedral intermediate of chymotrypsin - trifluoromethylketone in polar and non-polar media: observations from molecular dynamics simulation

The work uses MD simulation to study effects of five water contents (3 %, 10 %, 20 %, 50 %, 100 %?v/v) on the tetrahedral intermediate of chymotrypsin - trifluoromethyl ketone in polar acetonitrile and non-polar hexane media. The water content induced changes in the structure of the intermediate, solvent distribution and H-bonding are analyzed in the two organic media. Our results show that the changes in overall structure of the protein almost display a clear correlation with the water content in hexane media while to some extent U-shaped/bell-shaped dependence on the water content is observed in acetonitrile media with a minimum/maximum at 10–20 % water content. In contrast, the water content change in the two organic solvents does not play an observable role in the stability of catalytic hydrogen-bond network, which still exhibits high stability in all hydration levels, different from observations on the free enzyme system [Zhu L, Yang W, Meng YY, Xiao X, Guo Y, Pu X, Li M (2012) J Phys Chem B 116(10):3292–3304]. In low hydration levels, most water molecules mainly distribute near the protein surface and an increase in the water content could not fully exclude the organic solvent from the protein surface. However, the acetonitrile solvent displays a stronger ability to strip off water molecules from the protein than the hexane. In a summary, the difference in the calculated properties between the two organic solvents is almost significant in low water content (<10 %) and become to be small with increasing water content. In addition, some structural properties at 10?~?20 %?v/v hydration zone, to large extent, approach to those in aqueous solution.

The systematic characterization by aqueous column chromatography of solutes which affect protein stability

We have systematically characterized, by aqueous column chromatography on a size exclusion cross-linked dextran gel (Sephadex G-10), 12 solutes, 11 of which are known to affect protein stability. Six are chaotropes (water structure breakers) and destabilize proteins, while five are polar kosmotropes (polar water structure makers) and stabilize proteins. Analysis of the chromatographic behavior of these neutral (ethylene glycol, urea), positively charged (Tris, guanidine, as the hydrochloride salts) and negatively charged (SO2-4, HPO2-4, F-, Cl-, Br-, Cl3CCO-2, I-, SCN-, as the sodium salts, in order of elution) solutes at pH 7 as a function of sample concentration (up to 0.6 M), supporting electrolyte, and temperature yields four conclusions, based largely on the behavior of the anions. Chaotropes adsorb to the gel according to their position in the Hofmeister series, with the most chaotropic species adsorbing most strongly. ++Chaotropes adsorb to the gel less strongly in the presence of chaotropes (a salting in effect) and more strongly in the presence of polar kosmotropes (a salting out effect). Polar kosmotropes do not adsorb to the gel, and are sieved through the gel according to their position in the Hofmeister series, with the most kosmotropic species having the largest relative hydrodynamic radii. The hydrodynamic radii of polar kosmotropes is increased by chaotropes and decreased by polar kosmotropes. These results suggest that a chaotrope interacts with the first layer of immediately adjacent water molecules somewhat less strongly than would bulk water in its place; a polar kosmotrope, more strongly.     

Biocatalysis with hydroperoxide lyase in extracts from Penicillium camemberti in neat organic solvent media

Abstract

Biocatalysis with hydroperoxide lyase (HPL) in extracts from Penicillium camemberti, in neat organic solvent media has been investigated. The effects of reaction conditions including organic solvent mixtures, initial water activity (a_w) and reaction temperature as well as the effect of the lyoprotectants, KCl and dextran 1 kDa, on HPL activity were studied. The addition of KCl to the enzymatic extract (70:1 protein, w/w) prior to lyophilization, enhanced HPL activity 6.53-fold. In contrast, the presence of dextran at a ratio of 8:1 decreased the enzymatic activity. Using hexane as the reaction medium, with an initial a_w of 0.1 and 0.5, the HPL specific activity was determined to be as 6.3 and 65.9 nmol converted 10-HPOD/mg protein/min, for the enzymatic extract without and with KCl present, respectively. Although HPL enzymatic extract with KCl showed a relatively low optimum reaction temperature (45°C) compared to 55°C without KCl, it exhibited a 2.51- and 2.78-fold higher thermal stability at 60 and 80°C, respectively. The kinetic results indicated that the highest HPL catalytic efficiency, V_max/K_m, of 6.58 × 10^?2 mL/mg protein/min, was obtained in the presence of KCl.     

Activation of enzymes for nonaqueous biocatalysis by denaturing concentrations of urea

Urea is one of the most commonly used denaturants of proteins. However, herein we report that enzymes lyophilized from denaturing concentrations of aqueous urea exhibited much higher activity in organic solvents than their native counterparts. Thus, instead of causing deactivation, urea effected unexpected activation of enzymes suspended in organic media. Activation of subtilisin Carlsberg (SC) in the organic solvents (hexane, tetrahydrofuran, and acetone) increased with increasing urea concentrations up to 8 M. Active-site titration results and activity assays indicated the presence of partially unfolded but catalytically active SC in 8 M urea; however, the urea-modified enzyme retained high enantioselectivity and was ca. 80 times more active than the native enzyme in anhydrous hexane. Likewise, the activity of horseradish peroxidase (HRP) lyophilized from 8 M urea was ca. 56 times and 350 times higher in 97% acetone and water-saturated hexane, respectively, than the activity of HRP lyophilized from aqueous buffer. Compared with the native enzyme, the partially unfolded enzyme may have a more pliant and less rigid conformation in organic solvents, thus enabling it to retain higher catalytic activity. However, no substantial activation was observed for alpha-chymotrypsin lyophilized from urea solutions in which the enzyme retained some activity, illustrating that the activation effect is not completely general.     

Activity,stability and enantioselectivity of lipase-coated microcrystals of inorganic salts in organic solvents

Lipase-coated microcrystals of inorganic salts were prepared by dissolving enzymes in buffers and then mixing with 3 volumes of saturated salt solutions followed by drop-wise addition into polar precipitating organic solvents. The Mucor javanicus lipase-coated microcrystals did not show any activity for esterification of lauric acid with 1-propanol in isooctane when NaCl and Na₂SO₄ were used as the salts but showed much higher activity than the enzyme powder when KCl (10.0 times) and K₂SO₄ (5.8 times) were used as the salts and precipitated in 1-propanol. Acetonitrile was found to be the best precipitating solvent for preparing M. javanicus lipase-coated microcrystals, with enzyme activities 26.2 and 22.4 times higher than that of the enzyme powder when KCl and K₂SO₄ were used as precipitating salts, respectively. The presence of water in the precipitating solvents markedly decreased the enzyme activity. The M. javanicus lipase-coated microcrystals prepared using K₂SO₄ as the salt and acetonitrile as the precipitating solvent was as active at 80°C as at 40°C. No significant improvement in enantioselectivity of Candida rugosa lipase-coated microcrystals was observed for transesterification of 1-phenylethanol with vinyl acetate in hexane when the microcrystals were prepared by dissolving the enzymes in salt solutions containing 25% (v/v) of acetone or 2-propanol before precipitating in polar solvents.     

The influence of kosmotropic and chaotropic salts on the functional properties of Mucuna pruriens protein isolate

The influence of chaotropic and kosmotropic salts on Mucuna pruriens protein isolates was investigated. Protein solubility profile indicated that solubility was minimal at the isoelectric point of the protein isolate (4.0) while the solubility was maximal at pH 10.0 in all salt solutions. Chaotropes (I(-), ClO(4)(-) and SCN(-)) exhibit better protein solubility than the kosmotropes (SO(4)(2-), Cl(-) and Br(-)). Increase in protein solubility follows the Hofmeister series: NaSO(4)相似文献   

Measuring enzyme hydration in nonpolar organic solvents using NMR

A very sensitive NMR method has been developed for measuring deuterated water bound to proteins suspended in nonpolar solvents. This has been used to determine the amount of bound water as a function of water activity for subtilisin Carlsberg suspended in hexane, benzene, and toluene and for alpha-chymotrypsin in hexane. The adsorption isotherms for subtilisin in the three solvents are very similar showing that water activity can be usefully employed to predict the amount of water bound to proteins in nonpolar organic media. Comparison of the degree of enzyme hydration reached in nonpolar solvents with that obtained in air shows that adsorption of strongly bound water is hardly affected by the low dielectric medium, but adsorption of loosely bound water is significantly reduced. This suggests that the hydrophobic regions of the protein surface are preferentially solvated by solvent molecules, and that in a nonpolar environment formation of a complete monolayer of water over the protein surface is thermodynamically unfavorable. (c) 1995 John Wiley & Sons, Inc.     

Modeling hydration mechanisms of enzymes in nonpolar and polar organic solvents

A comprehensive study of the hydration mechanism of an enzyme in nonaqueous media was done using molecular dynamics simulations in five organic solvents with different polarities, namely, hexane, 3-pentanone, diisopropyl ether, ethanol, and acetonitrile. In these solvents, the serine protease cutinase from Fusarium solani pisi was increasingly hydrated with 12 different hydration levels ranging from 5% to 100% (w/w) (weight of water/weight of protein). The ability of organic solvents to 'strip off' water from the enzyme surface was clearly dependent on the nature of the organic solvent. The rmsd of the enzyme from the crystal structure was shown to be lower at specific hydration levels, depending on the organic solvent used. It was also shown that organic solvents determine the structure and dynamics of water at the enzyme surface. Nonpolar solvents enhance the formation of large clusters of water that are tightly bound to the enzyme, whereas water in polar organic solvents is fragmented in small clusters loosely bound to the enzyme surface. Ions seem to play an important role in the stabilization of exposed charged residues, mainly at low hydration levels. A common feature is found for the preferential localization of water molecules at particular regions of the enzyme surface in all organic solvents: water seems to be localized at equivalent regions of the enzyme surface independently of the organic solvent employed.     

Steady-state kinetic investigation of specific anion effects on the catalytic activity of yeast glutathione reductase

The catalytic activity of yeast glutathione reduetase at pH 7.6 is sensitive to the sodium phosphate buffer concentration and the presence of monovalent sodium salts in the assay medium. Low concentrations of sodium phosphate activate and high concentrations inhibit enzymatic activity. The optimal concentration is at about 0.06 m sodium phosphate. In the presence of 0.06 m sodium phosphate, addition of a variety of monovalent sodium salts results in inhibition of enzymatic activity, the inhibition being competitive with respect to NADPH and noncompetitive with respect to oxidized glutathione. At suboptimal concentrations of sodium phosphate, addition of monovalent sodium salts activates enzymatic activity. In addition, at suboptimal sodium phosphate concentration Lineweaver-Burk plots of initial velocity at constant NADPH concentration with oxidized glutathione as the variable substrate are nonlinear, being concave down. The nonlinear behavior can be eliminated by addition of 0.1 m sodium chloride. It is concluded that there are at least two specific anion binding sites at or near the enzyme active site. The anion inhibition is explained in terms of an ordered sequential mechanism for glutathione reduetase. The anion activation is analyzed in terms of a change of reaction pathway, the reactive enzyme species being dependent upon the oxidized glutathione concentration.     

Structural stability of beta-lactoglobulin in the presence of kosmotropic salts. A kinetic and thermodynamic study

The thiol group of beta-lactoglobulin reacted very sluggishly with dithio-bis-nitro-benzoic acid as compared to that of glutathione at pH 6.85. The pKapp value of the thiol group of the protein was 9.35. In the presence of 3 M urea, the thiol group reacted completely with dithio-bis-nitrobenzoic acid at pH 6.85. Heating (from 50 degrees to 80 degrees) increased the exposure of the thiol by dissociating the dimer unit. From the pseudo-first order rate constants of heat-exposure of thiol, thermodynamic activation parameters, delta G++, delta H++, and delta S++, for the heat-dissociation of beta-lactoglobulin dimer were estimated to be 23,290 cal/mol, 31,160 cal/mol, and 22.9 e.u. (at 70 degrees), respectively. Addition of kosmotropic salts, chloride, tartrate, sulfate, phosphate, and citrate (0.2 M) decreased the heat-induced exposure of the thiol group (at 70 degrees), probably by decreasing the dissociation of the dimer at pH 6.85. The relative change in free energy of activation for the dissociation of the dimer, delta(delta G++dimer), in the presence of the salts was positive, suggesting that these additives increase the stability of the dimer against heat. These salts also increased the conformational stability of beta-lactoglobulin as revealed by an increase in -delta(delta G0conf) values in their presence. Both delta(delta G++dimer) and -delta(delta G0conf) values followed the order, chloride less than tartrate less than sulfate less than phosphate less than citrate. These salts seem to manifest their structure-stabilizing effect by increasing both inter- and intramolecular hydrophobic interactions via changes in structure of water.