Relationship Between Amino Acid Properties and Protein Stability: Buried Mutations

In order to understand the mechanism of protein stability and to develop a simple method for predicting mutation-induced stability changes, we analyzed the relationship between stability changes caused by buried mutations and changes in 48 amino acid properties. As expected from the importance of hydrophobicity, properties reflecting hydrophobicity are strongly correlated with the stability of proteins. We found that subgroup classification based on secondary structure increased correlations significantly, and mutations within -strand segments correlated better than did those in -helical segments, which may result from stronger hydrophobicity of the -strands. Multiple regression analyses incorporating combinations of three properties from among all possible combinations of the 48 properties increased the correlation coefficient to 0.88 and by an average of 13% for all data sets. Analyzing the stability of tryptophan synthase mutants with Glu49 replaced by all other residues except Arg revealed that combining buriedness, solvent-accessible surface area for denatured protein, and unfolding Gibbs free energy change increased the correlation to 0.95. Consideration of sequence and structural information (neighboring residues in sequence and in space) did not significantly strengthen the correlations in buried mutations, suggesting that nonspecific interactions dominate in the interior of proteins.     

An inserted Gly residue fine tunes dynamics between mesophilic and thermophilic ribonucleases H

Dynamic processes are inherent properties of proteins and are crucial for a wide range of biological functions. To address how changes in protein sequence and structure affect dynamic processes, a quantitative comparison of microsecond-to-microsecond time scale conformational changes, measured by solution NMR spectroscopy, within homologous mesophilic and thermophilic ribonuclease H (RNase H) enzymes is presented. Kinetic transitions between the observed major state (high population) and alternate (low population) conformational state(s) of the substrate-binding handle region in RNase H from the mesophile Escherichia coli (ecRNH) and thermophile Thermus thermophilus (ttRNH) occur with similar kinetic exchange rate constants, but the difference in stability between exchanging conformers is smaller in ttRNH compared to ecRNH. The altered thermodynamic equilibrium between kinetically exchanging conformers in the thermophile is recapitulated in ecRNH by the insertion of a Gly residue within a putative hinge between alpha-helices B and C. This Gly insertion is conserved among thermophilic RNases H, and allows the formation of additional intrahelical hydrogen bonds. A Gly residue inserted between alpha-helices B and C appears to relieve unfavorable interactions in the transition state and alternate conformer(s) and represents an important adaptation to adjust conformational changes within RNase H for activity at high temperatures.     

A thermodynamic comparison of HPr proteins from extremophilic organisms

A thermodynamic stability study of five histidine-containing phosphocarrier protein (HPr) homologues derived from organisms inhabiting diverse environments is described. These HPr homologues are from Bacillus subtilis (Bs), Streptococcus thermophilus (St), Bacillus staerothermophilus (Bst), Bacillus halodurans (Bh), and Oceanobacillus iheyensis (Oi). Analyses of solvent and thermal denaturation experiments provide the cardinal thermodynamic parameters, like deltaG, deltaH, deltaS, T(m), and deltaC(p), that characterize the conformational stability for each homologue. The homologue from Bacillus staerothermophilus (BstHPr) was established as the most thermostable homologue and also the homologue with highest deltaG at all temperatures. A good correlation between habitat temperature of the organism and thermal stability of the protein is also seen. Stability curves (deltaG vs T) for every homologue are also reported; these reveal very similar deltaC(p) and temperature of maximum stability (T(S)) values for all HPr homologues. Stability curves show that the higher thermal stability of some homologues is not a result of change in curvature of the curve or a shift to higher temperature, but rather a displacement of the stability curves to higher deltaG values. Stability curves also allowed estimation of deltaG at habitat temperature of the organisms, and we find good agreement between homologues. Electrostatic contributions to stability of each homologue were investigated by measuring stability as a function of varying pH and NaCl concentration, and our results suggest that most HPr homologues share similar electrostatic contributions to stability.     

NMR hydrogen exchange of the OB-fold protein LysN as a function of denaturant: the most conserved elements of structure are the most stable to unfolding.

The structure of LysN contains an OB-fold motif composed of a structurally conserved five-stranded beta-barrel capped by a poorly conserved alpha-helix between strands beta3 and beta4. Two additional alpha-helices, unique to the LysN structure, flank the N terminus of the OB-fold. The stability of LysN to unfolding has been investigated with NMR native state hydrogen exchange measurements as a function of guanidinium hydrochloride concentration, and equilibrium unfolding transitions monitored by ellipticity at 222 nm and fluorescence at 350 nm. The spectrophotometric measurements suggest an apparent two-state unfolding transition with DeltaGu(0) approximately 6 kcal/mol and m approximately 3 kcal/(molM). By contrast, NMR hydrogen exchange measurements manifest a distribution of DeltaGu(0) and m values which indicate that the protein can undergo subglobal unfolding. The largest DeltaGu(0) values from hydrogen exchange are for residues in the beta-sheet of the protein. These values, which reflect complete unfolding of the protein, are between 3 and 4 kcal/mol higher than those obtained from circular dichroism or fluorescence. This discrepancy may be due to the comparison of NMR hydrogen exchange parameters measured at residue-level resolution, with spectrophotometric parameters that reflect an unresolved super position of unfolding transitions of the alpha-helices and beta-strands. The largest DeltaGu(0) values obtained from hydrogen exchange for the subset of residues in the alpha-helices of the protein, agree with the DeltaGu(0) values obtained from circular dichroism or fluorescence. Based on the hydrogen exchange data, however, the three alpha-helices of LysN are on average 3 kcal/mol less stable than the beta-sheet. Consistent with the subglobal unfolding of LysN evinced by hydrogen exchange, a deletion mutant that lacks the first alpha-helix of the protein retains a cooperatively folded structure. Taken together with previous results on the OB-fold proteins SN and CspA, the present results for LysN suggest that the most conserved elements of structure in the OB-fold motif are the most resistant to denaturation. In all three proteins, stability to denaturation correlates with sequence hydrophobicity.     

Stabilizing the subtilisin BPN' pro-domain by phage display selection: How restrictive is the amino acid code for maximum protein stability?

We have devised a procedure using monovalent phage display to select for stable mutants in the pro-domain of the serine protease, subtilisin BPN'. In complex with subtilisin, the pro-domain assumes a compact structure with a four-stranded antiparallel beta-sheet and two three-turn alpha-helices. When isolated, however, the pro-domain is 97% unfolded. These experiments use combinatorial mutagenesis to select for stabilizing amino acid combinations at a particular structural locus and determine how many combinations are close to the maximum protein stability. The selection for stability is based on the fact that the independent stability of the pro-domain is very low and that binding to subtilisin is thermodynamically linked to folding. Two libraries of mutant pro-domains were constructed and analyzed to determine how many combinations of amino acids at a particular structural locus result in the maximum stability. A library comprises all combinations of four amino acids at a structural locus. Previous studies using combinatorial genetics have shown that many different combinations of amino acids can be accommodated in a selected locus without destroying function. The present results indicate that the number of sequence combinations at a structural locus, which are close to the maximum stability, is small. The most striking example is a selection at an interior locus of the pro-domain. After two rounds of phagemid selection, one amino acid combination is found in 40% of sequenced mutants. The most frequently selected mutant has a deltaG(unfolding) = 4 kcal/mol at 25 degrees C, an increase of 6 kcal/mol relative to the naturally occurring sequence. Some implications of these results on the amount of sequence information needed to specify a unique tertiary fold are discussed. Apart from possible implications on the folding code, the phage display selection described here should be useful in optimizing the stability of other proteins, which can be displayed on the phage surface.     

Protein engineering to improve the thermostability of glucoamylase from Aspergillus awamori based on molecular dynamics simulations

Twelve mutations were constructed to improve the thermostability of glucoamylase from Aspergillus awamori based on the results of molecular dynamics simulations. The thermal unfolding of the catalytic domain followed a putative hierarchical behavior. In addition, the unfolding of the 13 alpha-helices obeyed the random ordered mechanism, in which the alpha-helices 8, 1 and 11 unfolded more rapidly than the others. The catalytic center was well protected by the (alpha/alpha)(6)-barrel at simulation temperatures up to 600 K, whereas the catalytic base, E400, migrated from its original interior pocket to the surface of the catalytic domain by surmounting the hydrophobic barrier provided by alpha-helices 12 and 13 at 800 K. The disulfide bonds engineered to 'lock' the alpha-helix 11 on the surface of the catalytic domain dramatically increased the thermostability. Substituting G396 and G407 with Ala residues slightly increased the thermostability, whereas their specific activity and catalytic efficiency were reduced. This indicates that the introduced residues with higher hydrophobicity were favorable in the loop between alpha-helices 12 and 13, whereas they partially destroyed the hydrogen bond and salt linkage network in the catalytic center. Alpha-helices 12 and 13 can be stabilized by introducing residues with higher hydrophobicity, except for the H391M mutation.     

Effects of glycosylation on protein conformation and amide proton exchange rates in RNase B.

Assignment of most of the proton NMR resonances of bovine pancreatic RNase B has been achieved using standard NMR techniques and by comparison with the published assignments for RNase A. A comparison of the NMR spectra of RNase B with RNase A shows that glycosylation of the enzyme has little overall effect on the conformation of the protein in solution. Comparisons of hydrogen-deuterium solvent exchange rates for the NH protons of RNase A and RNase B were made using two-dimensional 1H correlation spectroscopy. In the case of the glycosylated enzyme the exchange rates decreased for the NH protons of residues 9-14, 23-24, 32, 34-35, 39-40, 43-44, 48-49, 60, 71, 75-76, 80, 83-85, 100-101, 107, 111 and 122, relative to the unglycosylated RNase A. These results are consistent with the presence of the oligosaccharide inducing enhanced global dynamic stability and consequent changes to the unfolding equilibrium of the enzyme. The enhanced stability is observed not only for residues in the vicinity of the glycosylation site, asparagine-34, but also at residues remote from this site, as much as 30 A away.     

Role of structural and sequence information in the prediction of protein stability changes: comparison between buried and partially buried mutations.

Predicting mutation-induced changes in protein stability is one of the greatest challenges in molecular biology. In this work, we analyzed the correlation between stability changes caused by buried and partially buried mutations and changes in 48 physicochemical, energetic and conformational properties. We found that properties reflecting hydrophobicity strongly correlated with stability of buried mutations, and there was a direct relation between the property values and the number of carbon atoms. Classification of mutations based on their location within helix, strand, turn or coil segments improved the correlation of mutations with stability. Buried mutations within beta-strand segments correlated better than did those in alpha-helical segments, suggesting stronger hydrophobicity of the beta-strands. The stability changes caused by partially buried mutations in ordered structures (helix, strand and turn) correlated most strongly and were mainly governed by hydrophobicity. Due to the disordered nature of coils, the mechanism underlying their stability differed from that of the other secondary structures: the stability changes due to mutations within the coil were mainly influenced by the effects of entropy. Further classification of mutations within coils, based on their hydrogen-bond forming capability, led to much stronger correlations. Hydrophobicity was the major factor in determining the stability of buried mutations, whereas hydrogen bonds, other polar interactions and hydrophobic interactions were all important determinants of the stability of partially buried mutations. Information about local sequence and structural effects were more important for the prediction of stability changes caused by partially buried mutations than for buried mutations; they strengthened correlations by an average of 27% among all data sets.     

A fast method for the quantitative estimation of the distribution of hydrophobic and hydrophilic segments in alpha-helices of membrane proteins

The work presents a fast quantitative approach for estimating the orientations of hydrophilic and hydrophobic regions in the helical wheels of membrane-spanning alpha-helices of transmembrane proteins. The common hydropathy analysis provides an estimate of the integral hydrophobicity in a moving window which scans an amino acid sequence. The new parameter, orientation hydrophobicity, is based on the estimate of hydrophobicity of the angular segment that scans the helical wheel of a given amino acid sequence. The corresponding procedure involves the treatment of transmembrane helices as cylinders with equal surface elements for each amino acid residue. The orientation hydrophobicity, P(phi), phi = 0-360 degrees, of a helical cylinder is given as a sum of hydrophobicities of individual amino acids which are taken as the S-shaped functions of the angle between the centre of amino acid surface element and the centre of the segment. Non-zero contribution to P(phi) comes only from the amino acids belonging to the angular segment for a given angle phi. The size of the angular segment is related to the size of the channel pore. The amplitudes of amino acid S-functions are calibrated in the way that their maximum values (reached when the amino acid is completely exposed into the pore) are equal to the corresponding hydropathy index in the selected scale (here taken as Goldman-Engelman-Steitz hydropathy scale). The given procedure is applied in the studies of three ionic channels with well characterized three-dimensional structures where the channel pore is formed by a bundle of alpha-helices: cholera toxin B, nicotinic acetylcholine homopentameric alpha7 receptor, and phospholamban. The estimated maximum of hydrophilic properties at the helical wheels are in a good agreement with the spatial orientations of alpha-helices in the corresponding channel pores.     

Eukaryotic ribonucleases HI and HII generate characteristic hydrolytic patterns on DNA-RNA hybrids: further evidence that mitochondrial RNase H is an RNase HII

RNase H activities from HeLa cells (either of cytoplasmic or mitochondrial origin), and from mitochondria of beef heart and Xenopus ovaries, have been tested with RNA-DNA substrates of defined length (20 bp) and sequence. Substrates were either blunt-ended, or presented DNA or RNA overhangs. The hydrolysis profiles obtained at early times of the digestion showed a good correlation between the class of RNase H, either type I or II assigned according to biochemical parameters, whatever the organism. Consequently, the pattern of primary cuts can be considered as a signature of the predominant RNase H activity. For a given sequence, hydrolysis profiles obtained are similar, if not identical, for either blunt-ended substrates or those presenting overhangs. However, profiles showed variations depending on the sequence used. Of the three sequences tested, one appears very discriminatory, class I RNases H generating a unique primary cut 3 nt from the 3' end of the RNA strand, whereas class II RNases H generated two simultaneous primary cuts at 6 and at 8 nt from the 5' end of the RNA strand. Hydrolysis profiles further confirm the assignation of the mitochondrial RNase H activity from HeLa cells, beef heart and Xenopus oocytes to the class II.     

Low molecular mass pectate lyase from Fusarium moniliforme: similar modes of chemical and thermal denaturation

A low molecular mass pectate lyase from Fusarium moniliforme was unfolded reversibly by urea and Gdn-HCl at its optimum pH of 8.5, as monitored by intrinsic fluorescence, circular dichroism, and enzymatic activity measurements. Equilibrium unfolding studies yielded a deltaG(H(2)O) of 1.741 kcal/mol, D1/2 of 2.3M, and m value of 0.755kcal/molM with urea and a deltaG(H(2)O) of 1.927kcal/mol, D1/2 of 1.52M, and m value of 1.27 kcal/molM with Gdn-HCl as the denaturant. Thermal denaturation of the pectate lyase at, pH 8.5, was also reversible even after exposure to 75 degrees C for 10 min. Thermodynamic parameters calculated from thermal denaturation curves at pH values from 5.0 to 8.5 yielded a deltaCp of 0.864kcal/(molK). The deltaG(25 degrees C) at, pH 8.5, was 2.06kcal/mol and was in good agreement with the deltaG(H(2)O) values obtained from chemical denaturation curves. There was no exposure of hydrophobic pockets during chemical or thermal denaturation as indicated by the inability of ANS to bind the pectate lyase.     

NMR structure of the N-terminal domain of Saccharomyces cerevisiae RNase HI reveals a fold with a strong resemblance to the N-terminal domain of ribosomal protein L9.

In addition to the conserved and well-defined RNase H domain, eukaryotic RNases HI possess either one or two copies of a small N-terminal domain. The solution structure of one of the N-terminal domains from Saccharomyces cerevisiae RNase HI, determined using NMR spectroscopy, is presented. The 46 residue motif comprises a three-stranded antiparallel beta-sheet and two short alpha-helices which pack onto opposite faces of the beta-sheet. Conserved residues involved in packing the alpha-helices onto the beta-sheet form the hydrophobic core of the domain. Three highly conserved and solvent exposed residues are implicated in RNA binding, W22, K38 and K39. The beta-beta-alpha-beta-alpha topology of the domain differs from the structures of known RNA binding domains such as the double-stranded RNA binding domain (dsRBD), the hnRNP K homology (KH) domain and the RNP motif. However, structural similarities exist between this domain and the N-terminal domain of ribosomal protein L9 which binds to 23 S ribosomal RNA.     

Structural properties of the histidine-containing loop in HIV-1 RNase H

The isolated HIV-1 RNase H domain is inactive. This inactivity has been linked to the lack of structure in the C-terminus of the isolated domain. Thermodynamic stability experiments on the RNase H domain as well as a deletion mutant lacking the C-terminal helix have implied that this region is structured. His539 residing in a loop preceding the C-terminal helix was studied by NMR to determine the stability and conformational properties of this region. The stability of the structural environment of His539 matches that of the entire RNase H domain. Furthermore, His539 is locked into a defined tautometric state in the folded protein and its pK(a) is shifted compared to a freely accessible His, suggesting that this region is structured. The data support the view that the overall dynamics rather than the lack of structure in a small portion of the protein render activity of the isolated HIV-1 RNase H.     

Polyproline helices in protein structures: A statistical survey

A statistical survey of polyproline II (PPII) helices extracted from protein crystal structures is here reported. The average hydrophobicity of these helices is intermediate between those displayed by beta-strands and coil regions and is similar to that of alpha-helices. PPII helices with amphipathic properties have been identified and classified. Amino acid propensities for PPII helices derived in this study differ significantly from those previously reported. They show a little albeit significant correlation with propensities for alpha-helices whereas they are fully non-correlated to propensities for beta-sheets. Finally, PPII propensities have been correlated with amino acid frequencies in structural proteins, such as collagen and extensins.     

Hydrophobicity scales and computational techniques for detecting amphipathic structures in proteins

Protein segments that form amphipathic alpha-helices in their native state have periodic variation in the hydrophobicity values of the residues along the segment, with a 3.6 residue per cycle period characteristic of the alpha-helix. The assignment of hydrophobicity values to amino acids (hydrophobicity scale) affects the display of periodicity. Thirty-eight published hydrophobicity scales are compared for their ability to identify the characteristic period of alpha-helices, and an optimum scale for this purpose is computed using a new eigenvector method. Two of the published scales are also characterized by eigenvectors. We compare the usual method for detecting periodicity based on the discrete Fourier transform with a method based on a least-squares fit of a harmonic sequence to a sequence of hydrophobicity values. The two become equivalent for very long sequences, but, for shorter sequences with lengths commonly found in alpha-helices, the least-squares procedure gives a more reliable estimate of the period. The analog to the usual Fourier transform power spectrum is the "least-squares power spectrum", the sum of squares accounted for in fitting a sinusoid of given frequency to a sequence of hydrophobicity values. The sum of the spectra of the alpha-helices in our data base peaks at 97.5 degrees, and approximately 50% of the helices can account for this peak. Thus, approximately 50% of the alpha-helices appear to be amphipathic, and, of those that are, the dominant frequency at 97.5 degrees rather than 100 degrees indicates that the helix is slightly more open than previously thought, with the number of residues per turn closer to 3.7 than 3.6. The extra openness is examined in crystallographic data, and is shown to be associated with the C terminus of the helix. The alpha amphipathic index, the key quantity in our analysis, measures the fraction of the total spectral area that is under the 97.5 degrees peak, and is a characteristic of hydrophobicity scales that is consistent for different sets of helices. Our optimized scale maximizes the amphipathic index and has a correlation of 0.85 or higher with nine previously published scales. The most surprising feature of the optimized scale is that arginine tends to behave as if it were hydrophobic; i.e. in the crystallographic data base it has a tendency to be on the hydrophobic face of teh amphipathic helix. Although the scale is optimal only for predicting alpha-amphipathicity, it also ranks high in identifying beta-amphipathicity and in distinguishing interior from exterior residues in a protein.(ABSTRACT TRUNCATED AT 400 WORDS)     

Effects of mutation at methionine-42 of Escherichia coli dihydrofolate reductase on stability and function: implication of hydrophobic interactions

Methionine-42, distal to the active site of Escherichia coli dihydrofolate reductase, was substituted by site-directed mutagenesis with 14 amino acids (Ala, Cys, Glu, Gln, Gly, His, Ile, Leu, Pro, Ser, Thr, Trp, Tyr, and Val) to elucidate its role in the stability and function of this enzyme. Far-ultraviolet circular dichroism spectra of these mutants showed a distinctive negative peak at around 230 nm beside 220 nm, depending on the hydrophobicity of the amino acids introduced. The fluorescence intensity also increased in an order similar to that of the amino acids. These spectroscopic data suggest that the mutations do not affect the secondary structure, but strongly perturb the exciton coupling between Trp47 and Trp74. The free energy of urea unfolding, deltaG(o)u, increased with increases in the side-chain hydrophobicity in the range 2.96-6.40 kcal x mol(-1), which includes the value for the wild-type enzyme (6.08 kcal x mol(-1)). The steady-state kinetic parameters, Km and kcat, also increased with increases in the side-chain hydrophobicity, with the M42W mutant showing the largest increases in Km (35-fold) and kcat (4.3-fold) compared with the wild-type enzyme. These results demonstrate that site 42 distal to the active site plays an important role in the stability and function of this enzyme, and that the main effect of the mutations is to modify of hydrophobic interactions with the residues surrounding this position.     

Strong nucleic acid binding to the Escherichia coli RNase HI mutant with two arginine residues at the active site

The biochemical properties of the mutant protein D10R/E48R of Escherichia coli RNase HI, in which Asp(10) and Glu(48) are both replaced by Arg, were characterized. This mutant protein has been reported to have metal-independent RNase H activity at acidic pH [Casareno et al. (1995) J. Am. Chem. Soc. 117, 11011-11012]. The far- and near-UV CD spectra of this mutant protein were similar to those of the wild-type protein, suggesting that the protein conformation is not markedly changed by these mutations. Nevertheless, we found that this mutant protein did not show any RNase H activity in vitro. Instead, it showed high-nucleic-acid-binding affinity. Protein footprinting analyses suggest that DNA/RNA hybrid binds to or around the presumed substrate-binding site of the protein. In addition, this mutant protein did not complement the temperature-sensitive growth phenotype of the rnhA mutant strain, E. coli MIC3001, even at pH 6.0, suggesting that it does not show RNase H activity in vivo as well. These results are consistent with a current model for the catalytic mechanism of the enzyme, in which Glu(48) is not responsible for Mg(2+) binding but is involved in the catalytic function.