Site-directed replacement of the invariant lysine 73 of Saccharomyces cerevisiae iso-1-cytochrome c with all ribosomally encoded amino acids

Structure of a mitochondrial cytochrome c conformer competent for peroxidase activity

At the onset of apoptosis, the peroxidation of cardiolipin at the inner mitochondrial membrane by cytochrome c requires an open coordination site on the heme. We report a 1.45-Å resolution structure of yeast iso-1-cytochrome c with the Met80 heme ligand swung out of the heme crevice and replaced by a water molecule. This conformational change requires modest adjustments to the main chain of the heme crevice loop and is facilitated by a trimethyllysine 72-to-alanine mutation. This mutation also enhances the peroxidase activity of iso-1-cytochrome c. The structure shows a buried water channel capable of facilitating peroxide access to the active site and of moving protons produced during peroxidase activity to the protein surface. Alternate positions of the side chain of Arg38 appear to mediate opening and closing of the buried water channel. In addition, two buried water molecules can adopt alternate positions that change the network of hydrogen bonds in the buried water channel. Taken together, these observations suggest that low and high proton conductivity states may mediate peroxidase function. Comparison of yeast and mammalian cytochrome c sequences, in the context of the steric factors that permit opening of the heme crevice, suggests that higher organisms have evolved to inhibit peroxidase activity, providing a more stringent barrier to the onset of apoptosis.

Role of Hydrogen Bond Networks and Dynamics in Positive and Negative Cooperative Stabilization of a Protein

Cooperativity mediated through hydrogen bond networks in yeast iso-1-cytochrome c was studied using a thermodynamic triple mutant cycle. Three known stabilizing mutations, Asn 26 to His, Asn 52 to Ile, and Tyr 67 to Phe, were used to construct the triple mutant cycle. The side chain of His 26, a wild-type residue, forms two hydrogen bonds that bridge two substructures of the wild-type protein, and Tyr 67 and Asn 52 are part of an extensive buried hydrogen bond network. The stabilities of all variants in the triple mutant cycle were determined by guanidine hydrochloride denaturation methods and used to determine the pairwise, Delta(2)G(int), and triple interaction energies. His 26 and Ile 52 interact cooperatively (Delta(2)G(int) is 1-2 kcal/mol), whereas the two other pairs of mutations interact anticooperatively (Delta(2)G(int) is -0.5 to -1.5 kcal/mol). Previously reported structural data for iso-1-cytochrome c variants containing these mutations show that changes in the strength of the His 26 to Glu 44 hydrogen bond, apparently caused by changes in main chain dynamics, provide a mechanism for the long distance (His 26 to Phe 67 and His 26 to Ile 52) propagation of pairwise interaction energies. Opposing changes in the strength of the His 26 to Glu 44 hydrogen bond caused by the N52I and Y67F mutations generate a negative triple interaction energy (-0.9 +/-0.7 kcal/mol) that combined with cancellation of cooperative and anticooperative pairwise interactions produce apparent additivity for the stabilizing effects of the single mutations in the triple mutant variant.

Conformational Properties of the Iso-1-Cytochrome c Denatured State: Dependence on Guanidine Hydrochloride Concentration

Production of seven single surface histidine variants of yeast iso-1-cytochrome c allowed measurement of the apparent pK(a), pK(a)(obs), for histidine-heme loop formation for loops of nine to 83 amino acid residues under varying denaturing conditions (2 M to 6 M guanidine hydrochloride, gdnHCl). A linear correlation between pK(a)(obs) and the log of the loop size is expected for a random coil, pK(a)(obs) proportional to k log(n), where k is a scaling factor and n is the number of monomers in the loop. For small loops of nine, 16, and 22 monomers, no dependence of pK(a)(obs) on loop size was observed at any denaturant concentration indicating effects from chain stiffness. For larger loops of 37, 56, 72, and 83 monomers, the dependence of pK(a)(obs) on log(n) was linear and the slope of that dependence decreased with increasing concentration of denaturant. The scaling factor obtained at 5 M and 6 M gdnHCl for the larger loop sizes was approximately -2.0, close to the value of -2.2 expected for a random coil with excluded volume. However, scaling factors obtained under less harsh denaturing conditions (2 M to 4.5 M gdnHCl) deviated strongly from that expected for a random coil, being in the range -3 to -4. The gdnHCl dependence of pK(a)(obs) at each loop size was also evaluated to obtain denaturant m-values. Short loops where chain stiffness dominates had similar m-values of approximately 0.25 kcal/mol M. For larger loops m-values decrease with increasing loop size indicating that less hydrophobic area is sequestered when larger loops form. It is known that the earliest events in protein folding involve the formation of simple loops. The data from these studies provide direct insight into the relative probability with which loops of different sizes will form, as well as the factors which affect loop formation.

Evaluation of cooperative interactions between substructures of iso-1-cytochrome c using double mutant cycles

A double mutant cycle has been used to evaluate interaction energies between the global stabilizer mutation asparagine 52 --> isoleucine (N52I) in iso-1-cytochrome c and mutations producing single surface histidines at positions 26, 33, 39, 54, 73, 89, and 100. These histidine mutation sites are distributed through the four cooperative folding units of cytochrome c. The double mutant cycle starts with the iso-1-cytochrome c variant AcTM, a variant with no surface histidines and with asparagine at position 52. Isoleucine is added singly at position 52, AcTMI52 variant, as are the surface histidines, AcHX variants, where X indicates the histidine sequence position. The double mutant variants, AcHXI52, provide the remaining corner of the double mutant cycle. The stabilities of all variants were determined by guanidine hydrochloride denaturation and interaction energies were calculated between position 52 and each histidine site. Six of the seven double mutants show additive (AcH33I52, AcH39I52, AcH54I52, AcH89I52, and AcH100I52) stability effects or weak interaction energies (AcH73I52) of the histidine mutations and the N52I mutation, consistent with cooperative effects on protein folding and stability being sparsely distributed through the protein structure. The AcH26I52 variant shows a strong favorable interaction energy, 2.0 +/- 0.5 kcal/mol, between the N52I mutation in one substructure and the addition of His 26 to an adjacent substructure. The data are consistent with an entropic stabilization of the intersubstructure hydrogen bond between His 26 and Glu 44 by the Ile 52 mutation.

Denatured state thermodynamics: residual structure, chain stiffness and scaling factors

A set of nine variants of yeast iso-1-cytochrome c with zero or one surface histidine have been engineered such that the N-terminal amino group is acetylated in vivo. N-terminal acetylation has been confirmed by mass spectral analysis of intact and proteolytically digested protein. The histidine-heme loop-forming equilibrium, under denaturing conditions (3 M guanidine hydrochloride), has been measured by pH titration providing an observed pK(a), pK(a)(obs), for each variant. N-terminal acetylation prevents the N-terminal amino group-heme binding equilibrium from interfering with measurements of histidine-heme affinity. Significant deviation is observed from the linear dependence of pK(a)(obs) on the log of the number of monomers in the loop formed, expected for a random coil denatured state. The maximum histidine-heme affinity occurs for a loop size of 37 monomers. For loop sizes of 37-83 monomers, histidine-heme pK(a)(obs) values are consistent with a scaling factor of -4.2+/-0.3. This value is much larger than the scaling factor of -1.5 for a freely jointed random coil, which is commonly used to represent the conformational properties of protein denatured states. For loop sizes of nine to 22 monomers, chain stiffness is likely responsible for the decreases in histidine-heme affinity relative to a loop size of 37. The results are discussed in terms of residual structure and sequence composition effects on the conformational properties of the denatured states of proteins.

pH dependence of formation of a partially unfolded state of a Lys 73 --> His variant of iso-1-cytochrome c: implications for the alkaline conformational transition of cytochrome c

The alkaline conformational transition of a lysine 73 --> histidine variant of iso-1-cytochrome c has been studied. The transition has been monitored at 695 nm, a band sensitive to the presence of the heme-methionine 80 bond, at the heme Soret band which is sensitive to the nature of the heme ligand, and by NMR methods. The guanidine hydrochloride dependence of the alkaline conformational transition has also been monitored. The histidine 73 protein has an unusual biphasic alkaline conformational transition at both 695 nm and the heme Soret band, consistent with a three-state process. The conformational transition is fully reversible. An equilibrium model has been developed to account for this behavior. With this model, it has been possible to obtain the acid constant for the trigger group, pK(H), of the low-pH phase from the equilibrium data. A pK(H) value of 6.6 +/- 0.1 in H(2)O was obtained, consistent with a histidine acting as the trigger group. The NMR data for the low-pH phase of the alkaline conformational transition are consistent with an imidazole ligand replacing Met 80. For the high-pH phase of the biphasic alkaline transition, the NMR data are consistent with lysine 79 being the heme ligand. Guanidine hydrochloride m values of 1.67 +/- 0.08 and 1.1 +/- 0.2 kcal mol(-1) M(-1) were obtained for the low- and high-pH phases of the biphasic alkaline transition of the histidine 73 protein, respectively, consistent with a greater structural disruption for the low-pH phase of the transition.

Measuring denatured state energetics: deviations from random coil behavior and implications for the folding of iso-1-cytochrome c

The changes in the free energy of the denatured state of a set of yeast iso-1-cytochrome c variants with single surface histidine residues have been measured in 3 M guanidine hydrochloride. The thermodynamics of unfolding by guanidine hydrochloride is also reported. All variants have decreased stability relative to the wild-type protein. The free energy of the denatured state was determined in 3 M guanidine hydrochloride by evaluating the strength of heme-histidine ligation through determination of the pK(a) for loss of histidine binding to the heme. The data are corrected for the presence of the N-terminal amino group which also ligates to the heme under similar solution conditions. Significant deviations from random coil behavior are observed. Relative to a variant with a single histidine at position 26, residual structure of the order of -1.0 to -2.5 kcal/mol is seen for the other variants studied. The data explain the slower folding of yeast iso-1-cytochrome c relative to the horse protein. The greater number of histidines and the greater strength of ligation are expected to slow conversion of the histidine-misligated forms to the obligatory aquo-heme intermediate during the ligand exchange phase of folding. The particularly strong association of histidine residues at positions 54 and 89 may indicate regions of the protein with strong energetic propensities to collapse against the heme during early folding events, consistent with available data in the literature on early folding events for cytochrome c.

Effect of pH on formation of a nativelike intermediate on the unfolding pathway of a Lys 73 --> His variant of yeast iso-1-cytochrome c

Previous work on a Lys 73 --> His (H73) variant of iso-1-cytochrome c at pH 7.5 [Godbole et al. (1997) Biochemistry 36, 119-126] showed that this variant unfolds through a nativelike intermediate that has properties consistent with replacement of the Met 80 heme ligand by His 73. Here, the pH dependence of the equilibrium unfolding of the wild type (WT) and H73 proteins have been investigated, since a characteristic pH dependence is expected for the stability of an intermediate stabilized by histidine-heme ligation. Stability has been evaluated using guanidine hydrochloride and pH denaturation methods. Above pH 5, the m-values from guanidine hydrochloride denaturation of the WT and H73 variants remain significantly different, consistent with continued population of this intermediate. At pH 4.5 the m-values for the two proteins are within error the same. To assess stability at lower pH, acid denaturation was carried out. The midpoint is about 3.3 for both proteins but the transition is broader for the H73 protein, suggestive of intermediates again being populated during the unfolding of the H73 protein at this lower pH. Heme ligation by Met 80 was monitored (695 nm absorbance) during gdnHCl (pH 4.5 and 5.0) and acid denaturation, confirming, respectively, the absence and presence of intermediates. A thermodynamic analysis demonstrates that this complex pH dependence for the presence of histidine ligation induced intermediates is expected and implicates a titratable group with a pKa of approximately 6.6. The analysis also demonstrates when the pH dependences of global stability and stability of an intermediate differ significantly, population of folding intermediates as a function of pH will show novel behavior.

Stability effects of increasing the hydrophobicity of solvent-exposed side chains in staphylococcal nuclease

A total of fifty single site surface phenylalanine substitution mutants have been made in the model protein staphylococcal nuclease. The fifty residues that were replaced with phenylalanine were chosen to give a broad sampling of solvent accessibility, secondary structure, and backbone conformations. The change in the stability of each mutant protein relative to wild type was measured by guanidine hydrochloride denaturation. These results were compared to previous results obtained when these same sites were substituted with an alanine and a glycine. By this means, changes in the stability due to the loss of interactions of the wild-type side chain can be separated from the effects of introducing the bulky, hydrophobic phenylalanine in these solvent-exposed positions. In general, our results agree with the conventional wisdom that placing a hydrophobic residue in a solvent-exposed position is destabilizing in most cases, but less destabilizing than most changes in the hydrophobic core of the protein. However, the degree to which a hydrophobic surface substitution destabilizes or stabilizes a globular protein is highly context-dependent, with some mutations being as destabilizing as those in the core. This indicates that steric and packing considerations are also important on the surface of a globular protein but generally are not as important as in the interior. No evidence for the widespread occurrence of the so-called reverse hydrophobic effect at solvent-exposed sites was found. In addition, this survey of numerous sites suggests that previous measurements of alpha-helix "propensities" often seriously underestimate the importance of the environment of the side chain.

Sequence requirement for trimethylation of yeast cytochrome c

Lysine 72 (using the vertebrate numbering system) is trimethylated in cytochromes c from fungi and plants but not from higher animals. We have investigated the characteristics of an amino acid sequence required for trimethylation of lysine 72 by examining 21 altered iso-1-cytochromes c from Saccharomyces cerevisiae having single replacements in the region encompassing residues 67 through 77. These results indicated that tyrosine 74 is critical for trimethylation of lysine 72, whereas replacements at other positions did not produce significant diminutions. Various replacements of tyrosine 74 resulted in different levels of inhibition, with the Y74F replacement causing no significant reduction, and the Y74E and Y74K replacements completely or almost completely preventing trimethylation of lysine 72. However, other similarly spaced lysine and tyrosine residues at other sites in the protein did not result in trimethylation of the lysine residue. Thus, a properly situated aromatic residue, determined by the overall conformation of apocytochrome c in the vicinity of lysine 72, appears to be essential for trimethylation.

Thermal denaturation of iso-1-cytochrome c variants: comparison with solvent denaturation

Thermal denaturation studies as a function of pH were carried out on wild-type iso-1-cytochrome c and three variants of this protein at the solvent-exposed position 73 of the sequence. By examining the enthalpy and Tm at various pH values, the heat capacity increment (delta Cp), which is dominated by the degree of change in nonpolar hydration upon protein unfolding, was found for the wild type where lysine 73 is normally present and for three variants. For the Trp 73 variant, the delta Cp value (1.15 +/- 0.17 kcal/mol K) decreased slightly relative to wild-type iso-1-cytochrome c (1.40 +/- 0.06 kcal/mol K), while for the Ile 73 (1.65 +/- 0.07 kcal/mol K) and the Val 73 (1.50 +/- 0.06 kcal/mol K) variants, delta Cp increased slightly. In previous studies, the Trp 73, Ile 73, and Val 73 variants have been shown to have decreased m-values in guanidine hydrochloride denaturations relative to the wild-type protein (Hermann L, Bowler BE, Dong A, Caughey WS. 1995. The effects of hydrophilic to hydrophobic surface mutations on the denatured state of iso-1-cytochrome c: Investigation of aliphatic residues. Biochemistry 34:3040-3047). Both the m-value and delta Cp are related to the change in solvent exposure upon unfolding and other investigators have shown a correlation exists between these two parameters. However, for this subset of variants of iso-1-cytochrome c, a lack of correlation exists which implies that there may be basic differences between the guanidine hydrochloride and thermal denaturations of this protein. Spectroscopic data are consistent with different denatured states for thermal and guanidine hydrochloride unfolding. The different response of m-values and delta Cp for these variants will be discussed in this context.

A lysine 73-->histidine variant of yeast iso-1-cytochrome c: evidence for a native-like intermediate in the unfolding pathway and implications for m value effects

In this paper we report thermodynamic studies on a variant of yeast iso-1-cytochrome c in which a surface lysine residue at position 73 has been replaced with a histidine (H73). Guanidine hydrochloride denaturation studies monitored by circular dichroism spectroscopy indicated decreased thermodynamic stability (a lower delta G(o)(u)H20) and a smaller m value for the H73 protein as compared to the wild type (WT) protein. Further investigations to probe the causes for the thermodynamic stability differences between the two proteins involved guanidine hydrochloride and urea denaturations monitored by tryptophan fluorescence. The stability of heme ligation in the denatured state in the presence of either guanidine hydrochloride or urea was monitored by the spin-state transition of the heme iron induced by pH. None of these studies supported the hypothesis that the decreased m value was due to heme-His73 ligation in the denatured state. Guanidine hydrochloride denaturations monitored by the change in the extinction coefficient at 695 nm, which is sensitive to the presence of heme-Met80 ligation, revealed a native-like intermediate for the H73 protein, probably caused by displacement of the Met80 heme ligand by histidine 73 at guanidine hydrochloride concentrations much lower than required for full cooperative unfolding. Presence of the native-like intermediate is most likely the cause of the smaller m value and decreased thermodynamic stability for the CD-monitored H73 protein unfolding as compared to the unfolding of the WT protein. Guanidine hydrochloride denaturations in the presence of 200 mM imidazole provide further evidence in support of the proposed mechanism.