Solvent-induced collapse of alpha-synuclein and acid-denatured cytochrome c.

The effects of solution conditions on protein collapse were studied by measuring the hydrodynamic radii of two unfolded proteins, alpha-synuclein and acid-denatured ferricytochrome c, in dilute solution and in 1 M glucose. The radius of alpha-synuclein in dilute solution is less than that predicted for a highly denatured state, and adding 1 M glucose causes further collapse. Circular dichroic data show that alpha-synuclein lacks organized structure in both dilute solution and 1 M glucose. On the other hand, the radius of acid-denatured cytochrome c in dilute solution is consistent with that of a highly denatured state, and 1 M glucose induces collapse to the size and structure of native cytochrome c. Taken together, these data show that alpha-synuclein, a natively unfolded protein, is collapsed even in dilute solution, but lacks structure.

Characterization of horse cytochrome c expressed in Escherichia coli.

We have expressed horse cytochrome c in Escherichia coli. The gene was designed with E. coli codon bias and assembled by using a recursive polymerase chain reaction method. The far-ultraviolet and near-ultraviolet/Soret circular dichroism (CD) spectra show that the structure of recombinant horse cytochrome c is the same as that of the authentic protein. CD-detected thermal denaturation studies were used to measure the thermodynamic parameters associated with two-state denaturation. The free energy of denaturation for the recombinant protein is 10.0 +/- 2.3 kcal mol(-1) at pH 4.6 and 25 degrees C, which agrees with the value for the authentic protein. The expression system will help advance our understanding of the roles of cytochrome c in electron transfer, oxidative stress, and apoptosis by allowing the production of protein variants.

Interpreting the effects of small uncharged solutes on protein-folding equilibria.

Proteins are designed to function in environments crowded by cosolutes, but most studies of protein equilibria are conducted in dilute solution. While there is no doubt that crowding changes protein equilibria, interpretations of the changes remain controversial. This review combines experimental observations on the effect of small uncharged cosolutes (mostly sugars) on protein stability with a discussion of the thermodynamics of cosolute-induced nonideality and critical assessments of the most commonly applied interpretations. Despite the controversy surrounding the most appropriate manner for interpreting these effects of thermodynamic nonideality arising from the presence of small cosolutes, experimental advantage may still be taken of the ability of the cosolute effect to promote not only protein stabilization but also protein self-association and complex formation between dissimilar reactants. This phenomenon clearly has potential ramifications in the cell, where the crowded environment could well induce the same effects.

Second virial coefficients as a measure of protein--osmolyte interactions.

The cytoplasm contains high concentrations of cosolutes. These cosolutes include macromolecules and small organic molecules called osmolytes. However, most biophysical studies of proteins are conducted in dilute solutions. Two broad classes of models have been used to describe the interaction between osmolytes and proteins. One class focuses on excluded volume effects, while the other focuses on binding between the protein and the osmolyte. To better understand protein--smolyte interactions, we have conducted sedimentation equilibrium analytical ultracentrifugation experiments using ferricytochrome c as a model protein. From these experiments, we determined the second virial coefficients for a series of osmolytes. We have interpreted the second virial coefficient as a measure of both excluded volume and protein--osmolyte binding. We conclude that simple models are not sufficient to understand the interactions between osmolytes and proteins.

Effects of crowding by mono-, di-, and tetrasaccharides on cytochrome c-cytochrome c peroxidase binding: comparing experiment to theory.

In cells, protein-protein interactions occur in an environment that is crowded with other molecules, but in vitro studies are almost exclusively performed in dilute solution. To gain information about the effects of crowding on protein complex formation, we used isothermal titration calorimetry to measure the stoichiometry, the free energy change, and the enthalpy change for the binding of yeast iso-1-ferricytochrome c to yeast ferricytochrome c peroxidase in dilute solution and in solutions crowded with the sugars glucose, sucrose, and stachyose. The stoichiometry is 1:1 under all conditions. The sugars stabilize the complex, but by only 0.1-0.5 kcal.mol(-)(1), and the increased stability is not correlated with the change in enthalpy of complex formation. We then compared the measured stability changes to values obtained from several analyses that are currently used to predict crowding-induced changes in biomolecular equilibria. None of the analyses are completely successful by themselves, and the results suggest that a complete analysis must account for both excluded-volume and chemical interactions.

Interactions between yeast iso-1-cytochrome c and its peroxidase.

Isothermal titration calorimetry was used to study the formation of 19 complexes involving yeast iso-1-ferricytochrome c (Cc) and ferricytochrome c peroxidase (CcP). The complexes comprised combinations of the wild-type proteins, six CcP variants, and three Cc variants. Sixteen protein combinations were designed to probe the crystallographically defined interface between Cc and CcP. The data show that the high-affinity sites on Cc and CcP coincide with the crystallographically defined sites. Changing charged residues to alanine increases the enthalpy of complex formation by a constant amount, but the decrease in stability depends on the location of the amino acid substitution. Deleting methyl groups has a small effect on the binding enthalpy and a larger deleterious effect on the binding free energy, consistent with model studies of the hydrophobic effect, and showing that nonpolar interactions also stabilize the complex. Double-mutant cycles were used to determine the coupling energies for nine Cc-CcP residue pairs. Comparing these energies to the crystal structure of the complex leads to the conclusion that many of the substitutions induce a rearrangement of the complex.

Osmolyte-induced changes in protein conformational equilibria.

Examining solute-induced changes in protein conformational equilibria is a long-standing method for probing the role of water in maintaining protein stability. Interpreting the molecular details governing the solute-induced effects, however, remains controversial. We present experimental and theoretical data for osmolyte-induced changes in the stabilities of the A and N states of yeast iso-1-ferricytochrome c. Using polyol osmolytes of increasing size, we observe that osmolytes alone induce A-state formation from acid-denatured cytochrome c and N state formation from the thermally denatured protein. The stabilities of the A and N states increase linearly with osmolyte concentration. Interestingly, osmolytes stabilize the A state to a greater degree than the N state. To interpret the data, we divide the free energy for the reaction into contributions from nonspecific steric repulsions (excluded volume effects) and from binding interactions. We use scaled particle theory (SPT) to estimate the free energy contributions from steric repulsions, and we estimate the contributions from water-protein and osmolyte-protein binding interactions by comparing the SPT calculations to experimental data. We conclude that excluded volume effects are the primary stabilizing force, with changes in water-protein and solute-protein binding interactions making favorable contributions to stability of the A state and unfavorable contributions to the stability of the N state. The validity of our interpretation is strengthened by analysis of data on osmolyte-induced protein stabilization from the literature, and by comparison with other analyses of solute-induced changes in conformational equilibria.

Patterned library analysis: a method for the quantitative assessment of hypotheses concerning the determinants of protein structure.

Site-directed mutagenesis and combinatorial libraries are powerful tools for providing information about the relationship between protein sequence and structure. Here we report two extensions that expand the utility of combinatorial mutagenesis for the quantitative assessment of hypotheses about the determinants of protein structure. First, we show that resin-splitting technology, which allows the construction of arbitrarily complex libraries of degenerate oligonucleotides, can be used to construct more complex protein libraries for hypothesis testing than can be constructed from oligonucleotides limited to degenerate codons. Second, using eglin c as a model protein, we show that regression analysis of activity scores from library data can be used to assess the relative contributions to the specific activity of the amino acids that were varied in the library. The regression parameters derived from the analysis of a 455-member sample from a library wherein four solvent-exposed sites in an alpha-helix can contain any of nine different amino acids are highly correlated (P < 0.0001, R(2) = 0. 97) to the relative helix propensities for those amino acids, as estimated by a variety of biophysical and computational techniques.

Equilibrium thermodynamics of a physiologically-relevant heme-protein complex.

We used isothermal titration calorimetry to study the equilibrium thermodynamics for formation of the physiologically-relevant redox protein complex between yeast ferricytochrome c and yeast ferricytochrome c peroxidase. A 1:1 binding stoichiometry was observed, and the binding free energies agree with results from other techniques. The binding is either enthalpy- or entropy-driven depending on the conditions, and the heat capacity change upon binding is negative. Increasing the ionic strength destabilizes the complex, and both the binding enthalpy and entropy increase. Increasing the temperature stabilizes the complex, indicating a positive van't Hoff binding enthalpy, yet the calorimetric binding enthalpy is negative (-1.4 to -6.2 kcal mol(-)(1)). We suggest that this discrepancy is caused by solvent reorganization in an intermediate state. The measured enthalpy and heat capacity changes are in reasonable agreement with the values estimated from the surface area change upon complex formation. These results are compared to those for formation of the horse ferricytochrome c/yeast ferricytochrome c peroxidase complex. The results suggest that the crystal and solution structures for the yeast complex are the same, while the crystal and solution structures for horse cytochrome c/yeast cytochrome c peroxidase are different.

Expression of 15N-labeled eukaryotic cytochrome c in Escherichia coli.

We present a simple and inexpensive method for producing 15N-labeled Saccharomyces cerevisiae iso-1-cytochrome c in Escherichia coli. The labeled protein gives excellent NMR spectra.

Partially formed native tertiary interactions in the A-state of cytochrome c.

Considerable insight into protein structure, stability, and folding has been obtained from studies of non-native states. We have studied the extent of native tertiary contacts in one such molecule, the A-state of yeast iso-1-ferricytochrome c. Previously, we showed that the interface between the N and C-terminal helices is completely formed in the A-state. Here, we focus on interactions essential for forming the heme pocket of eukaryotic cytochromes c. To determine the extent of these interactions, we used saturation mutagenesis at the evolutionarily invariant residue leucine 68, and measured the free energy of denaturation for the native states and the A-states of functional variants. We show that, unlike the interaction between the terminal helices, the native interactions between the 60s helix and the rest of the protein are not completely formed in the A-state.

Nonideality and protein thermal denaturation.

We studied the thermal denaturation of eglin c by using CD spectropolarimetry and differential scanning calorimetry (DSC). At low protein concentrations, denaturation is consistent with the classical two-state model. At concentrations greater than several hundred microM, however, the calorimetric enthalpy and the midpoint transition temperature increase with increasing protein concentration. These observations suggested the presence of intermediates and/or native state aggregation. However, the transitions are symmetric, suggesting that intermediates are absent, the DSC data do not fit models that include aggregation, and analytical ultracentrifugation (AUC) data show that native eglin c is monomeric. Instead, the AUC data show that eglin c solutions are nonideal. Analysis of the AUC data gives a second virial coefficient that is close to values calculated from theory and the DSC data are consistent with the behavior expected for nonideal solutions. We conclude that the concentration dependence is caused by differential nonideality of the native and denatured states. The nondeality arises from the high charge of the protein at acid pH and is exacerbated by low buffer concentrations. Our conclusion may explain differences between van't Hoff and calorimetric denaturation enthalpies observed for other proteins whose behavior is otherwise consistent with the classical two-state model.

The cytoplasmic region of mouse Fc gamma RIIb1, but not Fc gamma RIIb2, binds phospholipid membranes.

The cytoplasmic regions of the mouse low-affinity Fc gamma RII isoforms, Fc gamma RIIb1 and Fc gamma RIIb2, play key roles in signal transduction by mediating different cellular functions. The Fc gamma RIIb1 (94 residues) and Fc gamma RIIb2 (47 residues) cytoplasmic regions are generated by differential mRNA splicing in which a single aspartic acid residue in Fc gamma RIIb2 is replaced by a 48-residue insert in Fc gamma RIIb1. In previous work, quantities of the mFc gamma RIIb1 and mFc gamma RIIb2 cytoplasmic regions were generated, and their secondary structures were examined in different solutions with circular dichroism [Chen, L., Thompson, N. L., and Pielak, G. J. (1997) Protein Sci. 6, 1038-1046]. In the work described here, steady-state light scattering was used to investigate possible interactions of the two isolated cytoplasmic regions with phospholipid vesicles. Three phospholipid compositions were examined: phosphatidylserine/phosphatidylcholine (PS/PC) (25/75, mol/mol); phosphatidylinositol bisphosphate/phosphatidylcholine (PIP2/PC) (25/75, mol/mol); and pure phosphatidylcholine (PC). Binding was examined in the presence and absence of Ca2+. The mFc gamma RIIb1 cytoplasmic peptide binds PS/PC vesicles weakly in the absence of Ca2+ and more strongly in the presence of Ca2+. For PIP2/PC vesicles, the behavior is reversed; binding is weak in the presence of Ca2+ and stronger in its absence. The mFc gamma RIIb1 peptide also weakly binds pure PC vesicles, in a Ca2+-independent manner. The mFc gamma RIIb2 cytoplasmic peptide does not bind, in the presence or absence of Ca2+, to PS/PC, PIP2/PC, or PC vesicles. The implications of these results for the mechanisms of signal transduction mediated by the two mFc gamma RII cytoplasmic regions are discussed.

Sugar-induced molten-globule model.

Proteins denature at low pH because of intramolecular electrostatic repulsions. The addition of salt partially overcomes this repulsion for some proteins, yielding a collapsed conformation called the A-state. A-states have characteristics expected for the molten globule, a notional kinetic protein folding intermediate. Here we show that the addition of neutral sugars to solutions of acid-denatured equine ferricytochrome c induces formation of the A-state in the absence of added salt. We characterized the structure and stability of the sugar-induced A-state with circular dichroism spectropolarimetry (CD) and NMR-monitored hydrogen-deuterium exchange experiments. We also examined the stability of the sugar-induced A-state as a function of sugar size and concentration. The results are interpreted using several models and we conclude that the stabilizing effect is consistent with increased steric repulsion between the protein and the sugar solutions.

Baseline length and automated fitting of denaturation data.

To understand relationships between protein sequence and stability, we often compare data from proteins that differ by the substitution of one amino acid. Frequently, an amino acid change causes the cooperative denaturation transitions to shift to lower temperatures, diminishing the signal from the native state. Here we show that apparent stability changes, i.e., the free energy of denaturation, deltaGD, can also be caused by a deficiency of points in the low temperature end of the transition. In addition, we suggest a method for overcoming this problem.

Native tertiary structure in an A-state.

The A-state is an equilibrium species that is thought to represent the molten globule, an on-pathway protein folding intermediate with native secondary structure and non-native, fluctuating tertiary structure. We used yeast iso-1-ferricytochrome c to test for an evolutionary-invariant tertiary interaction in its A-state. Thermal denaturation monitored by circular dichroism (CD)spectropolarimetry was used to determine A-state and native-state stabilities, delta GA reversible D and delta GN reversible D. We examined the wild-type protein, seven variants with substitutions at the interface between the N and C-terminal helices, and four control variants. The controls have the same amino acid changes as the interface variants, but the changes are close to, not at, the interface. We also examined the pH and sulfate concentration dependencies and found that while these factors affect the far-UV CD spectra of the least stable variants, they do not alter the difference in stability between the wild-type protein and the variants. A delta GA reversible D versus-delta GN reversible D plot for the interface variants has a slope near unity and the control variants have near-wild-type stability. These results show that the helix-helix interaction stabilizes the A-state and the native state to the same degree, confirming our preliminary report. We determined that the heat capacity change for A-state denaturation is approximately 60% of the value for native-state denaturation, indicating that the A-state interior is native-like. We discuss our results in relation to ferricytochrome c folding kinetics.

A genetic approach for identifying critical residues in the fingers and palm subdomains of HIV-1 reverse transcriptase.

By using oligonucleotide-directed saturation mutagenesis, we collected 366 different single amino acid substitutions in a 109-aa segment (residues 95-203) in the fingers and palm subdomains of the HIV-1 reverse transcriptase (RT), the enzyme that replicates the viral genome. After expression in Escherichia coli, two phenotypic assays were performed. The first assay tested for RNA-dependent DNA polymerase activity. The other assay used Western blot analysis to estimate the stability of each mutant protein by measuring the processing of the RT into its mature heterodimeric form, consisting of a 66-kDa subunit and a 51-kDa subunit. The resulting phenotypic data provided a "genetic" means to identify amino acid side chains that are important for protein function or stability, as well as side chains located on the protein surface. Several HIV-1 RT crystal structures were used to evaluate the mutational analysis. Our genetic map correlates well with the crystal structures. Combining our phenotype data with crystallographic data allowed us to study the genetically defined critical residues. The important functional residues are found near the enzyme active site. Many residues important for the stability of the RT participate in potential hydrogen bonding or hydrophobic interactions in the protein interior. In addition to providing a better understanding of the HIV-1 RT, this work demonstrates the utility of saturation mutagenesis to study the function, structure, and stability of proteins in general. This strategy should be useful for studying proteins for which no crystallographic data are available.

Design, synthesis, expression, and characterization of the genes for mouse Fc gamma RIIb1 and Fc gamma RIIb2 cytoplasmic regions.

The cytoplasmic regions of the mouse low-affinity Fc gamma RII isoforms, mFc gamma RIIb1, and mFc gamma RIIb2, play a key role in signal transduction by mediating different cellular functions. mFc gamma RIIb1 has a 94-residue cytoplasmic region, whereas mFc gamma RIIb2 has a 47-residue cytoplasmic region. Genes encoding the cytoplasmic regions of mFc gamma RIIb1 (b1-94) and mFc gamma RIIb2 (b2-47) were designed, synthesized, and expressed as fusion proteins in Escherichia coli. A sequence-specific protease, thrombin, was used to release the b1-94 peptide, which was purified by using HPLC. The b2-47 peptide was synthesized chemically. CD spectropolarimetry was employed to examine the secondary structures of b1-94 and b2-47. These studies were conducted in aqueous solution, in mixtures of water and trifluoroethanol or methanol, and as a function of temperature. The results indicate that the b1-94 and b2-47 structures are sensitive functions of the solvent environment, and that nonaqueous solvents induce significant alpha-helical structure.

Control of formation and dissociation of the high-affinity complex between cytochrome c and cytochrome c peroxidase by ionic strength and the low-affinity binding site.

A new ruthenium photoreduction technique was used to measure the formation and dissociation rate constants kf and kd of the high-affinity complex between yeast iso-1-cytochrome c (yCc) and cytochrome c peroxidase compound I (CMPI) over a wide range of ionic strength. These studies utilized Ru-39-Cc, which contains trisbipyridylruthenium attached to the cysteine residue in the H39C, C102T variant of yCc, and has the same reactivity with CMPI as native yCc. kd and kf were measured by photoreducing a small concentration of Ru-39-Cc in the presence of the oxidized yCcIII: CMPI complex, which must dissociate before Ru-39-CcII can bind to CMPI and reduce the radical action. The value of kd for the 1:1 high-affinity complex is very small at low ionic strength, < 5 s-1 but is increased significantly by binding yCc to a second low-affinity site. However, the low-affinity yCc binding site is not active in direct electron transfer to either the radical cation or the oxyferryl heme in CMPI, and is too weak to play a role in the kinetics at ionic strengths above 70 mM. The value of kd increases to 4000 s-1 at 150 mM ionic strength, while kf decreases from > 3 x 10(9) M-1 s-1 at low ionic strength to 1.3 x 10(9) M-1 s-1 at 150 mM ionic strength. These studies indicate that the rate-limiting step in enzyme turnover is product dissociation below 150 mM ionic strength and intracomplex electron transfer to the oxyferryl heme at higher ionic strength. The interaction between yCc and CcP is optimized at physiological ionic strength to provide the largest possible complex formation rate constant kf without allowing product dissociation to be rate-limiting. The effects of surface mutations on the kinetics provided evidence that the high-affinity binding site used for the reaction in solution is similar to the one identified in the yCc:CcP crystal structure.