Thermodynamics of apocytochrome b5 unfolding.

Apocytochrome b5 from rabbit liver was studied by scanning calorimetry, limited proteolysis, circular dichroism, second derivative spectroscopy, and size exclusion chromatography. The protein is able to undergo a reversible two-state thermal transition. However, transition temperature, denaturational enthalpy, and heat capacity change are reduced compared with the holoprotein. Apocytochrome b5 stability in terms of Gibbs energy change at protein unfolding (delta G) amounts to delta G = 7 +/- 1 kJ/mol at 25 degrees C (pH 7.4) compared with delta G = 25 kJ/mol for the holoprotein. Apocytochrome b5 is a compact, native-like protein. According to the spectral data, the cooperative structure is mainly based in the core region formed by residues 1-35 and 79-90. This finding is in full agreement with NMR data (Moore, C.D. & Lecomte, J.T.J., 1993, Biochemistry 32, 199-207).     

Thermodynamics of the temperature-induced unfolding of globular proteins.

The heat capacity, enthalpy, entropy, and Gibbs energy changes for the temperature-induced unfolding of 11 globular proteins of known three-dimensional structure have been obtained by microcalorimetric measurements. Their experimental values are compared to those we calculate from the change in solvent-accessible surface area between the native proteins and the extended polypeptide chain. We use proportionality coefficients for the transfer (hydration) of aliphatic, aromatic, and polar groups from gas phase to aqueous solution, we estimate vibrational effects, and we discuss the temperature dependence of each constituent of the thermodynamic functions. At 25 degrees C, stabilization of the native state of a globular protein is largely due to two favorable terms: the entropy of non-polar group hydration and the enthalpy of interactions within the protein. They compensate the unfavorable entropy change associated with these interactions (conformational entropy) and with vibrational effects. Due to the large heat capacity of nonpolar group hydration, its stabilizing contribution decreases quickly at higher temperatures, and the two unfavorable entropy terms take over, leading to temperature-induced unfolding.     

Thermodynamics of unfolding of an integral membrane protein in mixed micelles

Quantitative studies of membrane protein folding and unfolding can be difficult because of difficulties with efficient refolding as well as a pronounced propensity to aggregate. However, mixed micelles, consisting of the anionic detergent sodium dodecyl sulfate and the nonionic detergent dodecyl maltoside facilitate reversible and quantitative unfolding and refolding. The 4-transmembrane helix protein DsbB from the inner membrane of Escherichia coli unfolds in mixed micelles according to a three-state mechanism involving an unfolding intermediate I. The temperature dependence of the kinetics of this reaction between 15 degrees and 45 degrees C supports that unfolding from I to the denatured state D is accompanied by a significant decrease in heat capacity. For water-soluble proteins, the heat capacity increases upon unfolding, and this is generally interpreted as the increased binding of water to the protein as it unfolds, exposing more surface area. The decrease in DsbB's heat capacity upon unfolding is confirmed by independent thermal scans. The decrease in heat capacity is not an artifact of the use of mixed micelles, since the water soluble protein S6 shows conventional heat-capacity changes in detergent. We speculate that it reflects the binding of SDS to parts of DsbB that are solvent-exposed in the native DM-bound state. This implies that the periplasmic loops of DsbB are relatively unstructured. This anomalous thermodynamic behavior has not been observed for beta-barrel membrane proteins, probably because they do not bind SDS so extensively. Thus the thermodynamic behavior of membrane proteins appears to be intimately connected to their detergent-binding properties.     

Investigating protein unfolding kinetics by pulse proteolysis

Investigation of protein unfolding kinetics of proteins in crude samples may provide many exciting opportunities to study protein energetics under unconventional conditions. As an effort to develop a method with this capability, we employed “pulse proteolysis” to investigate protein unfolding kinetics. Pulse proteolysis has been shown to be an effective and facile method to determine global stability of proteins by exploiting the difference in proteolytic susceptibilities between folded and unfolded proteins. Electrophoretic separation after proteolysis allows monitoring protein unfolding without protein purification. We employed pulse proteolysis to determine unfolding kinetics of E. coli maltose binding protein (MBP) and E. coli ribonuclease H (RNase H). The unfolding kinetic constants determined by pulse proteolysis are in good agreement with those determined by circular dichroism. We then determined an unfolding kinetic constant of overexpressed MBP in a cell lysate. An accurate unfolding kinetic constant was successfully determined with the unpurified MBP. Also, we investigated the effect of ligand binding on unfolding kinetics of MBP using pulse proteolysis. On the basis of a kinetic model for unfolding of MBP•maltose complex, we have determined the dissociation equilibrium constant (K_d) of the complex from unfolding kinetic constants, which is also in good agreement with known K_d values of the complex. These results clearly demonstrate the feasibility and the accuracy of pulse proteolysis as a quantitative probe to investigate protein unfolding kinetics.     

A useful epitope tag derived from maltose binding protein

Maltose binding protein (MBP) is used in recombinant protein expression as an affinity and solubility tag. The monoclonal antibody B48 binds MBP tightly and has no cross‐reactivity to other proteins in an Escherichia coli lysate. This high level of specificity suggested that MBP contains an epitope that could prove useful as a purification and visualization tag for proteins expressed in E. coli. To discover the MBP epitope, a co‐crystal structure was determined for MBP bound to its antibody and four amino acids of MBP were identified as critical for the binding interaction. Fusions of various fragments of MBP to the glutathione S‐transferase protein were engineered in order to identify the smallest fragment still recognized by the α‐MBP antibody. Stabilization of the epitope via mutational engineering resulted in a minimized 14 amino‐acid tag.     

Denaturant mediated unfolding of both native and molten globule states of maltose binding protein are accompanied by large deltaCp's.

Maltose binding protein (MBP) is a large, monomeric two domain protein containing 370 amino acids. In the absence of denaturant at neutral pH, the protein is in the native state, while at pH 3.0 it forms a molten globule. The molten globule lacks a tertiary circular dichroism signal but has secondary structure similar to that of the native state. The molten globule binds 8-anilino-1-naphthalene sulfonate (ANS). The unfolding thermodynamics of MBP at both pHs were measured by carrying out a series of isothermal urea melts at temperatures ranging from 274-329 K. At 298 K, values of deltaGdegrees , deltaCp, and Cm were 3.1+/-0.2 kcal mol(-1), 5.9+/-0.8 kcal mol(-1) K(-1) (15.9 cal (mol-residue)(-1) K(-1)), and 0.8 M, respectively, at pH 3.0 and 14.5+/-0.4 kcal mol(-1), 8.3+/-0.7 kcal mol(-1) K(-1) (22.4 kcal (mol-residue)(-1) K(-1)), and 3.3 M, respectively, at pH 7.1. Guanidine hydrochloride denaturation at pH 7.1 gave values of deltaGdegrees and deltaCp similar to those obtained with urea. The m values for denaturation are strongly temperature dependent, in contrast to what has been previously observed for small globular proteins. The value of deltaCp per mol-residue for the molten globule is comparable to corresponding values of deltaCp for the unfolding of typical globular proteins and suggests that it is a highly ordered structure, unlike molten globules of many small proteins. The value of deltaCp per mol-residue for the unfolding of the native state is among the highest currently known for any protein.     

The interplay between effector binding and allostery in an engineered protein switch

The protein design rules for engineering allosteric regulation are not well understood. A fundamental understanding of the determinants of ligand binding in an allosteric context could facilitate the design and construction of versatile protein switches and biosensors. Here, we conducted extensive in vitro and in vivo characterization of the effects of 285 unique point mutations at 15 residues in the maltose‐binding pocket of the maltose‐activated β‐lactamase MBP317‐347. MBP317‐347 is an allosteric enzyme formed by the insertion of TEM‐1 β‐lactamase into the E. coli maltose binding protein (MBP). We find that the maltose‐dependent resistance to ampicillin conferred to the cells by the MBP317‐347 switch gene (the switch phenotype) is very robust to mutations, with most mutations slightly improving the switch phenotype. We identified 15 mutations that improved switch performance from twofold to 22‐fold, primarily by decreasing the catalytic activity in the absence of maltose, perhaps by disrupting interactions that cause a small fraction of MBP in solution to exist in a partially closed state in the absence of maltose. Other notable mutations include K15D and K15H that increased maltose affinity 30‐fold and Y155K and Y155R that compromised switching by diminishing the ability of maltose to increase catalytic activity. The data also provided insights into normal MBP physiology, as select mutations at D14, W62, and F156 retained high maltose affinity but abolished the switch's ability to substitute for MBP in the transport of maltose into the cell. The results reveal the complex relationship between ligand binding and allostery in this engineered switch.     

Amino acid substitutions affecting protein dynamics in eglin C do not affect heat capacity change upon unfolding

The heat capacity change upon unfolding (deltaC(p)) is a thermodynamic parameter that defines the temperature dependence of the thermodynamic stability of proteins; however, physical basis of the heat capacity change is not completely understood. Although empirical surface area-based calculations can predict heat capacity changes reasonably well, accumulating evidence suggests that changes in hydration of those surfaces is not the only parameter contributing to the observed heat capacity changes upon unfolding. Because packing density in the protein interior is similar to that observed in organic crystals, we hypothesized that changes in protein dynamics resulting in increased rigidity of the protein structure might contribute to the observed heat capacity change upon unfolding. Using differential scanning calorimetry we characterized the thermodynamic behavior of a serine protease inhibitor eglin C and two eglin C variants with altered native state dynamics, as determined by NMR. We found no evidence of changes in deltaC(p) in either of the variants, suggesting that changes in rigidity do not contribute to the heat capacity change upon unfolding in this model system.     

Thermodynamics of unfolding for turkey ovomucoid third domain: thermal and chemical denaturation.

We have used thermal and chemical denaturation to characterize the thermodynamics of unfolding for turkey ovomucoid third domain (OMTKY3). Thermal denaturation was monitored spectroscopically at a number of wave-lengths and data were subjected to van't Hoff analysis; at pH 2.0, the midpoint of denaturation (Tm) occurs at 58.6 +/- 0.4 degrees C and the enthalpy of unfolding at this temperature (delta Hm) is 40.8 +/- 0.3 kcal/mol. When Tm was perturbed by varying pH and denaturant concentration, the resulting plots of delta Hm versus Tm yield a mean value of 590 +/- 120 cal/(mol.K) for the change in heat capacity upon unfolding (delta Cp). A global fit of the same data to an equation that includes the temperature dependence for the enthalpy of unfolding yielded a value of 640 +/- 110 cal/(mol.K). We also performed a variation of the linear extrapolation method described by Pace and Laurents, which is an independent method for determining delta Cp (Pace, C.N. & Laurents, D., 1989, Biochemistry 28, 2520-2525). First, OMTKY3 was thermally denatured in the presence of a variety of denaturant concentrations. Linear extrapolations were then made from isothermal slices through the transition region of the denaturation curves. When extrapolated free energies of unfolding (delta Gu) were plotted versus temperature, the resulting curve appeared linear; therefore, delta Cp could not be determined. However, the data for delta Gu versus denaturant concentration are linear over an extraordinarily wide range of concentrations. Moreover, extrapolated values of delta Gu in urea are identical to values measured directly.     

Denaturant m values and heat capacity changes: relation to changes in accessible surface areas of protein unfolding.

Denaturant m values, the dependence of the free energy of unfolding on denaturant concentration, have been collected for a large set of proteins. The m value correlates very strongly with the amount of protein surface exposed to solvent upon unfolding, with linear correlation coefficients of R = 0.84 for urea and R = 0.87 for guanidine hydrochloride. These correlations improve to R = 0.90 when the effect of disulfide bonds on the accessible area of the unfolded protein is included. A similar dependence on accessible surface area has been found previously for the heat capacity change (delta Cp), which is confirmed here for our set of proteins. Denaturant m values and heat capacity changes also correlate well with each other. For proteins that undergo a simple two-state unfolding mechanism, the amount of surface exposed to solvent upon unfolding is a main structural determinant for both m values and delta Cp.     

How maltose influences structural changes to bind to maltose‐binding protein: Results from umbrella sampling simulation

A well‐studied periplasmic‐binding protein involved in the abstraction of maltose is maltose‐binding protein (MBP), which undergoes a ligand‐induced conformational transition from an open (ligand‐free) to a closed (ligand‐bound) state. Umbrella sampling simulations have been us to estimate the free energy of binding of maltose to MBP and to trace the potential of mean force of the unbinding event using the center‐of‐mass distance between the protein and ligand as the reaction coordinate. The free energy thus obtained compares nicely with the experimentally measured value justifying our theoretical basis. Measurement of the domain angle (N‐terminal‐domain – hinge – C‐terminal‐domain) along the unbinding pathway established the existence of three different states. Starting from a closed state, the protein shifts to an open conformation during the initial unbinding event of the ligand then resides in a semi‐open conformation and later resides predominantly in an open‐state. These transitions along the ligand unbinding pathway have been captured in greater depth using principal component analysis. It is proposed that in mixed‐model, both conformational selection and an induced‐fit mechanism combine to the ligand recognition process in MBP. Proteins 2013. © 2012 Wiley Periodicals, Inc.     

Expression of aChlamydia anticarbohydrate single-chain antibody as a maltose binding fusion protein

A single-chain antibody fragment has been constructed for an antibody that binds to theChlamydia specific carbohydrate structure of the lipopolysaccharide. Single-chain protein was expressed and secreted into the periplasmic space ofE. coli as a fusion protein with the maltose binding protein. The fusion protein was purified in one step by virtue of its ability to bind to maltose. In a sandwich ELISA, the eluted protein boundChlamydia lipopolysaccharide, which demonstrates that the single-chain protein domain will function as part of a fusion protein. The expression of maltose binding fusion proteins into the periplasmic space could be used for production of other single-chain antibodies or protein fragments requiring appropriate folding and disulfide bond formation.     

Thermodynamics of DNA binding and distortion by the hyperthermophile chromatin protein Sac7d

Sac7d is a hyperthermophile chromatin protein which binds non-specifically to the minor groove of duplex DNA and induces a sharp kink of 66 degrees with intercalation of valine and methionine side-chains. We have utilized the thermal stability of Sac7d and the lack of sequence specificity to define the thermodynamics of DNA binding over a wide temperature range. The binding affinity for poly(dGdC) was moderate at 25 degrees C (Ka = 3.5(+/-1.6) x 10(6) M(-1)) and increased by nearly an order of magnitude from 10 degrees C to 80 degrees C. The enthalpy of binding was unfavorable at 25 degrees C, and decreased linearly from 5 degrees C to 60 degrees C. A positive binding heat at 25 degrees C is attributed in part to the energy of distorting DNA, and ensures that the temperature of maximal binding affinity (75.1+/-5.6 degrees C) is near the growth temperature of Sulfolobus acidocaldarius. Truncation of the two intercalating residues to alanine led to a decreased ability to bend and unwind DNA at 25 degrees C with a small decrease in binding affinity. The energy gained from intercalation is slightly greater than the free energy penalty of bending duplex DNA. Surprisingly, reduced distortion from the double alanine substitution did not lead to a significant decrease in the heat of binding at 25 degrees C. In addition, an anomalous positive DeltaCp of binding was observed for the double alanine mutant protein which could not be explained by the change in polar and apolar accessible surface areas. Both the larger than expected binding enthalpy and the positive heat capacity can be explained by a temperature dependent structural transition in the protein-DNA complex with a Tm of 15-20 degrees C and a DeltaH of 15 kcal/mol. Data are discussed which indicate that the endothermic transition in the complex is consistent with DNA distortion.     

Thermodynamics of protein‐cation interaction: Ca+2 and Mg+2 binding to the fifth binding module of the LDL receptor

The ligand binding domain of the LDL receptor (LDLR) contains seven structurally homologous repeats. The fifth repeat (LR5) is considered to be the main module responsible for the binding of lipoproteins LDL and β‐VLDL. LR5, like the other homologous repeats, is around 40‐residue long and contains three disulfide bonds and a conserved cluster of negatively charged residues surrounding a hexacoordinated calcium ion. The calcium coordinating cage is formed by the backbone oxygens of W193 and D198, and side‐chain atoms of D196, D200, D206, and E207. The functionality of LDLR is closely associated with the presence of calcium. Magnesium ions are to some extent similar to calcium ions. However, they appear to be involved in different physiological events and their concentrations in extracellular and intracellular compartments are regulated by different mechanisms. Whether magnesium ions can play a role in the complex cycle of LDLR internalization and recycling is not known. We report here a detailed study of the interaction between LR5 and these two cations combining ITC, emission fluorescence, high resolution NMR, and MD simulations, at extracellular and endosomal pHs. Our results indicate that the conformational stability and internal dynamics of LR5 are strongly modulated by the specific bound cation. It appears that the difference in binding affinity for these cations is somewhat compensated by their different concentrations in late LDL‐associated endosomes. While the mildly acidic and calcium‐depleted environment in late endosomes has been proposed to contribute significantly to LDL release, the presence of magnesium might assist in efficient LDLR recycling. Proteins 2010. © 2009 Wiley‐Liss, Inc.     

D-trehalose/D-maltose-binding protein from the hyperthermophilic archaeon Thermococcus litoralis: the binding of trehalose and maltose results in different protein conformational states

In this work, we used fluorescence spectroscopy, molecular dynamics simulation, and Fourier transform infrared spectroscopy for investigating the effect of trehalose binding and maltose binding on the structural properties and the physical parameters of the recombinant D-trehalose/D-maltose binding protein (TMBP) from the hyperthermophilic archaeon Thermococcus litoralis. The binding of the two sugars to TMBP was studied in the temperature range 20 degrees-100 degrees C. The results show that TMBP possesses remarkable temperature stability and its secondary structure does not melt up to 90 degrees C. Although both the secondary structure itself and the sequence of melting events were not significantly affected by the sugar binding, the protein assumes different conformations with different physical properties depending whether maltose or trehalose is bound to the protein. At low and moderate temperatures, TMBP possesses a structure that is highly compact both in the absence and in the presence of two sugars. At about 90 degrees C, the structure of the unliganded TMBP partially relaxes whereas both the TMBP/maltose and the TMBP/trehalose complexes remain in the compact state. In addition, Fourier transform infrared results show that the population of alpha-helices exposed to the solvent was smaller in the absence than in the presence of the two sugars. The spectroscopic results are supported by molecular dynamics simulations. Our data on dynamics and stability of TMBP can contribute to a better understanding of transport-related functions of TMBP and constitute ground for targeted modifications of this protein for potential biotechnological applications.     

Analysis of allosteric signal transduction mechanisms in an engineered fluorescent maltose biosensor

We previously reported the construction of a family of reagentless fluorescent biosensor proteins by the structure-based design of conjugation sites for a single, environmentally sensitive small molecule dye, thus providing a mechanism for the transduction of ligand-induced conformational changes into a macroscopic fluorescence observable. Here we investigate the microscopic mechanisms that may be responsible for the macroscopic fluorescent changes in such Fluorescent Allosteric Signal Transduction (FAST) proteins. As case studies, we selected three individual cysteine mutations (F92C, D95C, and S233C) of Escherichia coli maltose binding protein (MBP) covalently labeled with a single small molecule fluorescent probe, N-((2-iodoacetoxy)ethyl)-N-methyl)amino-7-nitrobenz-2-oxa-1,3-diazole (NBD), each giving rise to a robust FAST protein with a distinct maltose-dependent fluorescence response. The fluorescence emission intensity, anisotropy, lifetime, and iodide-dependent fluorescence quenching were determined for each conjugate in the presence and absence of maltose. Structure-derived solvent accessible surface areas of the three FAST proteins are consistent with experimentally observed quenching data. The D95C protein exhibits the largest fluorescence change upon maltose binding. This mutant was selected for further characterization, and residues surrounding the fluorophore coupling site were mutagenized. Analysis of the resulting mutant FAST proteins suggests that specific hydrogen-bonding interactions between the fluorophore molecule and two tyrosine side-chains, Tyr171 and Tyr176, in the open state but not the closed, are responsible for the dramatic fluorescence response of this construct. Taken together these results provide insights that can be used in future design cycles to construct fluorescent biosensors that optimize signaling by engineering specific hydrogen bonds between a fluorophore and protein.     

Signal sequence non-optimal codons are required for the correct folding of mature maltose binding protein

Non-optimal codons are generally characterised by a low concentration of isoaccepting tRNA and a slower translation rate compared to optimal codons. In a previous study, we reported a 20-fold reduction in maltose binding protein (MBP) level when the non-optimal codons in the signal sequence were optimised. In this study, we report that the 20-fold reduction is rescued when MBP is expressed at 28 °C instead of 37 °C, suggesting that the signal sequence optimised MBP protein (MBP-opt) may be misfolded, and is being degraded at 37 °C. Consistent with this idea, transient induction of the heat shock proteases prior to MBP expression at 28 °C restores the 20-fold difference, demonstrating that the difference in production levels is due to post-translational degradation of MBP-opt by the heat-shock proteases. Analysis of the structure of purified MBP-wt and MBP-opt grown at 28 °C showed that although they have similar secondary structure content, MBP-opt is more resistant to thermal unfolding than is MBP-wt. The two proteins also exhibit different tryptic fragment profiles, further confirming that they are folded into conformationally different states. This is the first study to demonstrate that signal sequence non-optimal codons can influence the folding of the mature exported protein.     

Kinetics and energetics of the translocation of maltose binding protein folding mutants

Protein translocation in Escherichia coli is mediated by the translocase that, in its minimal form, comprises a protein-conducting pore (SecYEG) and a motor protein (SecA). The SecYEG complex forms a narrow channel in the membrane that allows passage of secretory proteins (preproteins) in an unfolded state only. It has been suggested that the SecA requirement for translocation depends on the folding stability of the mature preprotein domain. Here we studied the effects of the signal sequence and SecB on the folding and translocation of folding stabilizing and destabilizing mutants of the mature maltose binding protein (MBP). Although the mutations affect the folding of the precursor form of MBP, these are drastically overruled by the combined unfolding stabilization of the signal sequence and SecB. Consequently, the translocation kinetics, the energetics and the SecA and SecB dependence of the folding mutants are indistinguishable from those of wild-type preMBP. These data indicate that unfolding of the mature domain of preMBP is likely not a rate-determining step in translocation when the protein is targeted to the translocase via SecB.     

Thermodynamics of zinc binding to hepatitis C virus NS3 protease: A folding by binding event

The hepatitis C virus (HCV) nonstructural protein 3 (NS3) protease is responsible for the processing of the non‐structural region of the viral precursor polyprotein in infected hepatic cells. HCV NS3 is a zinc‐dependent serine protease. The zinc ion, which is bound far away from the active site and considered to have a structural role, is essential for the structural integrity of the protein; furthermore, the ion is required for the hydrolytic activity. Consequently, the NS3 zinc binding site has been considered for a long time as a possible target for drug discovery. As a first step towards this goal, the energetics of the NS3‐zinc interaction and its effect on the NS3 conformation must be established and discussed. The thermodynamic characterization of zinc binding to NS3 protease by isothermal titration calorimetry and spectroscopy is presented here. Spectroscopic and calorimetric results suggest that a considerable conformational change in the protein is coupled to zinc binding. The energetics of the conformational change is comparable to that of the folding of a protein of similar size. Therefore, zinc binding to NS3 protease can be considered as a “folding by binding” event. Proteins 2009. © 2009 Wiley‐Liss, Inc.     

Thermodynamics of melittin tetramerization determined by circular dichroism and implications for protein folding.

The tetramerization of melittin, a 26-amino acid peptide from Apis mellifera bee venom, has been studied as a model for protein folding. Melittin converts from a monomeric random coil to an alpha-helical tetramer as the pH is raised from 4.0 to 9.5, as ionic strength is increased, as temperature is raised or lowered from about 37 degrees C, or as phosphate is added. The thermodynamics of this tetramerization (termed "folding") are explored using circular dichroism. The melittin tetramer has two pKa values of 7.5 and 8.5 corresponding to protonation of the N-terminus and Lys 23, respectively. pKa values calculated with the program DelPhi (Gilson, M.K., Sharp, K.A., & Honig, B.H., 1987, J. Comp. Chem. 9, 327-335; Gilson, M.K. & Honig, B.H., 1988a, Proteins 3, 32-52; Gilson, M.K. & Honig, B.H., 1988b, Proteins 4, 7-18) agree with experimental titration data. Greater electrostatic repulsion of these protonated groups destabilizes the tetramer by 3.6 kcal/mol at pH 4.0 compared to pH 9.5. Increasing the concentration of NaCl in the solution from 0 to 0.5 M stabilizes the tetramer by 5-6 kcal/mol at pH 4.0. The effect of NaCl is modeled with a ligand-binding approach. The melittin tetramer is found to have a temperature of maximum stability ranging from 35.5 to 43 degrees C depending on the pH, unfolding above and below that temperature. delta Cp0 for folding ranges from -0.085 to -0.102 cal g-1 K-1, comparable to that of other small globular proteins (Privalov, P.L., 1979, Adv. Protein Chem. 33, 167-241). delta H0 and delta S0 are found to decrease with temperature, presumably due to the hydrophobic effect (Kauzmann, W., 1959, Adv. Protein Chem. 14, 1-63). Phosphate is found to perturb the equilibrium substantially with a maximal effect at 150 mM, stabilizing the tetramer at pH 7.4 and 25 degrees C by 4.6 kcal/mol. The enthalpy change due to addition of phosphate (-7.5 kcal/mol at 25 degrees C) can be accounted for by simple dielectric screening. Both circular dichroism and crystallographic results suggest that phosphate may bind Lys 23 at the ends of the elongated tetramer. These detailed measurements give insight into the relative importance of various forces for the stability of melittin in the folded form and may provide an experimental standard for future tests of computational energetics on this simple protein system.