Enhancement of the thermostability of Hydrogenobacter thermophilus cytochrome c(552) through introduction of an extra methylene group into its hydrophobic protein interior

Careful scrutiny of the protein interior of Hydrogenobacter thermophilus cytochrome c(552) (HT) on the basis of its X-ray structure [Travaglini-Allocatelli, C., Gianni, S., Dubey, V. K., Borgia, A., Di Matteo, A., Bonivento, D., Cutruzzola, F., Bren, K. L., and Brunori, M. (2005) J. Biol. Chem. 280, 25729-25734] indicated that a void space, which is large enough to accommodate a methyl group, exists in the hydrophobic protein interior near the heme. We tried to reduce the void space through the replacement of a Val by Ile or Leu (Val/Ile or Val/Leu mutation), and then the structural and functional consequences of these two mutations were characterized in order to elucidate the relationship between the nature of the packing of hydrophobic residues and the functional properties of the protein. The study demonstrated striking differences in the structural and functional consequences between the two mutations. The Val/Ile mutation was found to cause further enhancement of the thermostability of the oxidized HT, as reflected in the increase of the denaturation temperature (T(m)) value by ～ 3 deg, whereas the thermostability of the reduced form was essentially unaffected. As a result, the redox potential (E(m)) of the Val/Ile mutant exhibited a negative shift of ～ 50 mV relative to that of the wild-type protein in an enthalpic manner, this being consistent with our previous finding that a protein with higher stability in its oxidized form exhibits a lower E(m) value [Terui, N., Tachiiri, N., Matsuo, H., Hasegawa, J., Uchiyama, S., Kobayashi, Y., Igarashi, Y., Sambongi, Y., and Yamamoto, Y. (2003) J. Am. Chem. Soc. 125, 13650-13651]. In contrast, the Val/Leu mutation led to a decrease in thermostability of both the redox forms of the protein, as reflected in the decreases of the T(m) values of the oxidized and reduced proteins by ～ 3 and ～ 5 deg, respectively, and the E(m) value of the Val/Leu mutant happened to be similar to that of the Val/Ile one. The E(m) value of the Val/Leu mutant could be reasonably interpreted in terms of the different effects of the mutation on the stabilities of the two different redox forms of the protein. Thus, the present study demonstrated that the stability of the protein is affected quite sensitively by the contextual stereochemical packing of hydrophobic residues in the protein interior and that the structural properties of the hydrophobic core in the protein interior are crucial for control of the redox function of the protein. These findings provide novel insights as to functional control of a protein, which could be utilized for tuning of the T(m) and E(m) values of the protein by means of protein engineering.     

Investigating mitochondrial redox potential with redox-sensitive green fluorescent protein indicators

Current methods for determining ambient redox potential in cells are labor-intensive and generally require destruction of tissue. This precludes single cell or real time studies of changes in redox poise that result from metabolic processes or environmental influences. By substitution of surface-exposed residues on the Aequorea victoria green fluorescent protein (GFP) with cysteines in appropriate positions to form disulfide bonds, reduction-oxidation-sensitive GFPs (roGFPs) have been created. roGFPs have two fluorescence excitation maxima at about 400 and 490 nm and display rapid and reversible ratiometric changes in fluorescence in response to changes in ambient redox potential in vitro and in vivo. Crystal structure analyses of reduced and oxidized crystals of roGFP2 at 2.0- and 1.9-A resolution, respectively, reveal in the oxidized state a highly strained disulfide and localized main chain structural changes that presumably account for the state-dependent spectral changes. roGFP1 has been targeted to the mitochondria in HeLa cells. Fluorometric measurements on these cells using a fluorescence microscope or in cell suspension using a fluorometer reveal that the roGFP1 probe is in dynamic equilibrium with the mitochondrial redox status and responds to membrane-permeable reductants and oxidants. The roGFP1 probe reports that the matrix space in HeLa cell mitochondria is highly reducing, with a midpoint potential near -360 mV (assuming mitochondrial pH approximately 8.0 at 37 degrees C). In other work (C. T. Dooley, T. M. Dore, G. Hanson, W. C. Jackson, S. J. Remington, and R. Y. Tsien, submitted for publication), it is shown that the cytosol of HeLa cells is also unusually reducing but somewhat less so than the mitochondrial matrix.     

Partially folded state of the disulfide-reduced N-terminal half-molecule of ovotransferrin as a renaturation intermediate

A previous report (Hirose, M., Akuta, T., and Takahashi, N. (1989) J. Biol. Chem. 264, 16867-16872) has shown that for the efficient oxidative refolding of disulfide-reduced ovotransferrin, a preincubation under reduced conditions at a low temperature is essential. To study the renaturation pathway, the disulfide-reduced N-terminal half-molecule of ovotransferrin was analyzed by CD spectrum. The reduced protein was found to take, at low temperatures, a partially folded conformation that can be distinguished from both the native and denatured states. The folded protein was in a metastable state with delta GD value of 2.2-2.8 kcal/mol at 6 degrees C. The conformation was variable depending on temperature conditions; its stability was decreased at a lower temperature (1.0-1.2 kcal/mol at 0 degrees C). Subsequent reoxidation at 6 degrees C by oxidized glutathione led efficiently the reduced protein to the correctly renatured form having the iron-binding capacity, indicating that the partially folded state is the immediate precursor to subsequent oxidative refolding.     

Importance of a conserved hydrogen-bonding network in cytochromes c to their redox potentials and stabilities.

To understand the determinants of redox potential and protein stability in c-type cytochromes, we have characterized two mutations to a highly conserved tyrosine group, tyrosine-75, of Rhodobacter capsulatus cytochrome c2. Mutant Y75F was designed to test the importance of the tyrosine hydroxyl group to the typically high redox potentials of the cytochromes c2 while maintaining a hydrophobic core. Mutant Y75C was designed to test the importance of a large hydrophobic group to redox potential by replacing an aromatic group with a small nonpolar group. Both mutants exhibit spectral and redox properties indicating that their heme environments have been perturbed. The kinetics of reduction by lumiflavin semiquinone and photooxidation by Rhodobacter sphaeroides photosynthetic reaction centers have been used to demonstrate that both mutants are structurally analogous to the wild-type protein at the active site of electron transfer. Different degrees of relative stability of the mutants toward a denaturant have been observed with the order being Y75C less than wt less than Y75F in the oxidized state and Y75C less than Y75F less than wt in the reduced state. These results are discussed in light of the recent structure determination of the R. capsulatus wild-type ferrocytochrome c2 to suggest that R. capsulatus tyrosine-75, or its equivalent in other species, is part of a conserved hydrogen-bonding network which plays an important role in maintaining high redox potentials and protein stability of cytochromes c in general.     

Specific mutagenesis of the rieske iron-sulfur protein in Rhodobacter sphaeroides shows that both the thermodynamic gradient and the pK of the oxidized form determine the rate of quinol oxidation by the bc(1) complex

In the Rieske iron-sulfur protein (ISP) of the ubiquinol:cytochrome c(2) oxidoreductase (bc(1) complex) of Rhodobacter sphaeroides, residue Tyr 156 is located close to the iron-sulfur cluster. Previous studies of the equivalent residue in both Saccharomyces cerevisiae [Denke, E., Merbitz-Zahradnik, T., Hatzfeld, O. M., Snyder, C. H., Link, T. A., and Trumpower, B. L. (1998) J. Biol. Chem. 273, 9085-9093] and Paracoccus denitrificans [Schroter, T., Hatzfeld, O. M., Gemeinhardt, S., Korn, M., Friedrich, T., Ludwig, B. , and Link, T. A. (1998) Eur. J. Biochem. 255, 100-106] have indicated that mutations at this site can lead to modifications in the redox potential of the ISP. To study the effect of similar modifications on the thermodynamic behavior and kinetics of partial reactions of the bc(1) complex upon flash activation, we have constructed four mutant strains of Rb. sphaeroides where Tyr 156 was mutated to His, Leu, Phe, or Trp. The bc(1) complex was assembled and able to support photosynthetic growth in all mutants. Three substitutions (Leu, Phe, Trp) led to alteration of the midpoint potential (E(m)) of the ISP and a slowing in rate of quinol oxidation, suggesting that electron transfer from quinol to the oxidized ISP controls the overall rate and that this step includes the high activation barrier. The Trp mutation led to an increase of approximately 1 pH unit in the pK value of the oxidized ISP. The pH dependence of the rate of quinol oxidation in this mutant was also shifted up by approximately 1 pH unit, showing the importance of the protonation state of the ISP for quinol oxidation. This provides support for a model in which the dissociated form of the oxidized ISP is required for formation of the enzyme-substrate complex [Ugulava, N., and Crofts, A. R. (1998) FEBS Lett. 440, 409-413].     

The conserved active site proline determines the reducing power of Staphylococcus aureus thioredoxin

Nature uses thioredoxin-like folds in several disulfide bond oxidoreductases. Each of them has a typical active site Cys-X-X-Cys sequence motif, the hallmark of thioredoxin being Trp-Cys-Gly-Pro-Cys. The intriguing role of the highly conserved proline in the ubiquitous reducing agent thioredoxin was studied by site-specific mutagenesis of Staphylococcus aureus thioredoxin (Sa_Trx). We present X-ray structures, redox potential, pK(a), steady-state kinetic parameters, and thermodynamic stabilities. By replacing the central proline to a threonine/serine, no extra hydrogen bonds with the sulphur of the nucleophilic cysteine are introduced. The only structural difference is that the immediate chemical surrounding of the nucleophilic cysteine becomes more hydrophilic. The pK(a) value of the nucleophilic cysteine decreases with approximately one pH unit and its redox potential increases with 30 mV. Thioredoxin becomes more oxidizing and the efficiency to catalyse substrate reduction (k(cat)/K(M)) decreases sevenfold relative to wild-type Sa_Trx. The oxidized form of wild-type Sa_Trx is far more stable than the reduced form over the whole temperature range. The driving force to reduce substrate proteins is the relative stability of the oxidized versus the reduced form Delta(T(1/2))(ox/red). This driving force is decreased in the Sa_Trx P31T mutant. Delta(T(1/2))(ox/red) drops from 15.5 degrees C (wild-type) to 5.8 degrees C (P31T mutant). In conclusion, the active site proline in thioredoxin determines the driving potential for substrate reduction.     

Quantifying energetic contributions to ground state destabilization

Solution differential scanning calorimetry (DSC) of oxidized amicyanin, a Type I copper protein, at pH 7.5 reveals two thermal transitions. The major transition at 67.7 degrees C corresponds to the disruption of the Cys(92) thiolate to Cu(II) charge transfer as evidenced by a corresponding temperature-dependent loss of amicyanin visible absorbance. A minor transition at 75.5 degrees C describes the further irreversible protein unfolding. Reduced amicyanin exhibits a pH-dependent change of the copper ligand geometry. At pH 8.5 where the Type I tetrahedral geometry is maintained, DSC reveals two thermal transitions with T(m) values similar to that of oxidized amicyanin. At pH 6.2 where the Cu(I) coordination is trigonal planar, reduced amicyanin exhibits a single thermal transition with a lower T(m) of 64.0 degrees C. Apoamicyanin, from which copper has been removed, also exhibits a single thermal transition but with a much lower T(m) of 51.8 degrees C. Thus, the thermal stability of amicyanin is dictated both by the presence or absence of copper and its ligand geometry, but not its redox state. The physiological relevance of these data is discussed.     

pH- and metal ion-linked stability of the hemopexin-heme complex

Thermal denaturation of the human hemopexin-heme complex was investigated under a variety of solution conditions to identify factors that influence heme release. The midpoint temperature for the transition between the folded and folded states, T(m), of the hemopexin-ferriheme complex exhibits a significant dependence on pH. When the pH is reduced from 7 to 5 (50 mM BisTris buffer and 50 mM NaCl), T(m) decreases by approximately 23 degrees C despite the relatively higher chloride concentration that tends to stabilize the protein. The thermal stability of the hemopexin-ferroheme complex was examined at pH 7.4 to yield a T(m) that is 3.2 degrees C lower than that of the hemopexin-ferriheme complex under identical conditions. The effect of transition metal ions, which hemopexin has recently been shown to bind [Mauk, M. R., Rosell, F. I., Lelj-Garolla, B., Moore, G. R., and Mauk, A. G. (2005) Biochemistry 44, XXXX-XXXX], was also considered. Cu(2+) and Zn(2+) had the greatest effect, reducing T(m) for the transition by 4.8 and 6.5 degrees C, respectively, relative to the value for the protein in the absence of metal ions [T(m) = 64.9 degrees C [10 mM sodium phosphate buffer (pH 7.4)]]. These metal ions also interfered significantly with the recovery of the native state from the unfolded protein when the protein on returning to 20 degrees C. The current results demonstrate how the conditions within the endosomes of hepatocytes (pH approximately 5.0, [Cl(-)] approximately 60 mM) and the potential presence of transition metal ions or heme iron reduction contribute to the membrane receptor-mediated process of heme release from hemopexin.     

Outer sphere mutagenesis of Lactobacillus plantarum manganese catalase disrupts the cluster core. Mechanistic implications.

X-ray crystallography of the nonheme manganese catalase from Lactobacillus plantarum (LPC) [Barynin, V.V., Whittaker, M.M., Antonyuk, S.V., Lamzin, V.S., Harrison, P.M., Artymiuk, P.J. & Whittaker, J.W. (2001) Structure9, 725-738] has revealed the structure of the dimanganese redox cluster together with its protein environment. The oxidized [Mn(III)Mn(III)] cluster is bridged by two solvent molecules (oxo and hydroxo, respectively) together with a micro 1,3 bridging glutamate carboxylate and is embedded in a web of hydrogen bonds involving an outer sphere tyrosine residue (Tyr42). A novel homologous expression system has been developed for production of active recombinant LPC and Tyr42 has been replaced by phenylalanine using site-directed mutagenesis. Spectroscopic and structural studies indicate that disruption of the hydrogen-bonded web significantly perturbs the active site in Y42F LPC, breaking one of the solvent bridges and generating an 'open' form of the dimanganese cluster. Two of the metal ligands adopt alternate conformations in the crystal structure, both conformers having a broken solvent bridge in the dimanganese core. The oxidized Y42F LPC exhibits strong optical absorption characteristic of high spin Mn(III) in low symmetry and lower coordination number. MCD and EPR measurements provide complementary information defining a ferromagnetically coupled electronic ground state for a cluster containing a single solvent bridge, in contrast to the diamagnetic ground state found for the native cluster containing a pair of solvent bridges. Y42F LPC has less than 5% of the catalase activity and much higher Km for H2O2 ( approximately 1.4 m) at neutral pH than WT LPC, although the activity is slightly restored at high pH where the cluster is converted to a diamagnetic form. These studies provide new insight into the contribution of the outer sphere tyrosine to the stability of the dimanganese cluster and the role of the solvent bridges in catalysis by dimanganese catalases.     

Thioredoxin reductase-dependent insulin disulfide reduction by phage T7 DNA polymerase reflects dissociation of the enzyme into subunits

Phage T7 DNA polymerase contains Escherichia coli thioredoxin as a subunit and is a 1:1 complex with T7 gene 5 protein. The enzyme showed high thioredoxin activity in assays at 37 degrees C using reduction of insulin disulfides with NADPH and thioredoxin reductase, leading Randahl (Randahl, H. (1982) FEBS Lett. 150, 109-113) to propose that the thioredoxin dithiol active site is exposed in T7 DNA polymerase. However, T7 DNA polymerase and free thioredoxin differ in reactivity with iodoacetic acid after preincubation with dithiothreitol or incubation with insulin. Insulin reduction assays work at low temperatures even at 0 degrees C. The time and temperature dependence of the thioredoxin activity of T7 DNA polymerase demonstrated that dissociation into subunits at 25 or 37 degrees C accounts for the previously observed activity. Thus, T7 DNA polymerase contains the reduced form of thioredoxin with its active site SH groups masked by the subunit contact with the gene 5 protein in agreement with the results of Adler and Modrich (Adler, S., and Modrich, P. (1983) J. Biol. Chem. 258, 6956-6962). The subunit interaction of thioredoxin and gene 5 protein is salt-insensitive, but markedly temperature-dependent consistent with involvement of a hydrophobic surface area in reduced thioredoxin.     

Determination of the redox properties of the Rieske [2Fe-2S] cluster of bovine heart bc1 complex by direct electrochemistry of a water-soluble fragment.

The redox potential of the Rieske [2Fe-2S] cluster of the bc1 complex from bovine heart mitochondria was determined by cyclic voltammetry of a water-soluble fragment of the iron/sulfur protein. At the nitric-acid-treated bare glassy-carbon electrode, the fragment gave an immediate and stable quasireversible response. The midpoint potential at pH 7.2, 25 degrees C and I of 0.01 M was Em = +312 +/- 3 mV. This value corresponds within 20 mV to results of an EPR-monitored dye-mediated redox titration. With increasing ionic strength, the midpoint potential decreased linearly with square root of I up to I = 2.5 M. From the cathodic-to-anodic peak separation, the heterogeneous rate constant, k degrees, was calculated to be approximately 2 x 10(-3) cm/s at low ionic strength; the rate constant increased with increasing ionic strength. From the temperature dependence of the midpoint potential, the standard reaction entropy was calculated as delta S degrees = -155 J.K-1.mol-1. The pH dependence of the midpoint potential was followed over pH 5.5-10. Above pH 7, redox-state-dependent pK changes were observed. The slope of the curve, -120 mV/pH above pH9, indicated two deprotonations of the oxidized protein. The pKa values of the oxidized protein, obtained by curve fitting, were 7.6 and 9.2, respectively. A group with a pKa,ox of approximately 7.5 could also be observed in the optical spectrum of the oxidized protein. Redox-dependent pK values of the iron/sulfur protein are considered to be essential for semiquinone oxidation at the Qo center of the bc1 complex.     

Contribution of conformational stability and reversibility of unfolding to the increased thermostability of human and bovine superoxide dismutase mutated at free cysteines

The conformational stability and reversibility of unfolding of the human dimeric enzyme Cu Zn superoxide dismutase (HSOD) and the three mutant enzymes constructed by replacement of Cys6 by Ala and Cys111 by Ser, singly and in combination, were determined by differential scanning calorimetry. The differential scanning calorimetry profile of wild-type HSOD consists of two components, which probably represent the unfolding of the oxidized and reduced forms of the enzyme, with denaturation temperatures (Tm) of 74.9 and 83.6 degrees C, approximately 7 degrees lower than those for bovine superoxide dismutase (BSOD). The conformational stabilities of the two components of the mutant HSOD's differ only slightly from those of the wild type (delta delta Gs of -0.2 to +0.8 kcal/mol of dimer), while replacement of the BSOD Cys6 by Ala is somewhat destabilizing (delta delta G of -0.7 to -1.3 kcal/mol of dimer). These small alterations in conformational stability do not correlate with the large increases in resistance to thermal inactivation following substitution of free Cys in both HSOD and BSOD (McRee, D.E., Redford, S.M., Getzoff, E.D., Lepock, J.R., Hallewell, R.A., and Tainer, J.A. (1990) J. Biol. Chem. 265, 14234-14241 and Hallewell, R.A., Imlay, K.C., Laria, I., Gallegos, C., Fong, N., Irvine, B., Getzoff, E.D., Tainer, J.A., Cubelli, D.E., Bielski, B.H.J., Olson, P., Mallenbach, G.T., and Cousens, L.S. (1991) Proteins Struct. Funct. Genet., submitted for publication). The reversibility of unfolding was determined by scanning part way through the profile, cooling, rescanning, and calculating the amount of protein irreversibly unfolded by the first scan. The order of reversibility at a constant level of unfolding is the same as the order of resistance to inactivation for both the HSOD and BSOD wild-type and mutant enzymes. Thus, the greater resistance to thermal inactivation of the superoxide dismutase enzymes with free Cys replaced by Ala or Ser is dominated by a greater resistance to irreversible unfolding and relatively unaffected by changes in conformational stability.     

The superoxide dismutase activity of desulfoferrodoxin from Desulfovibrio desulfuricans ATCC 27774.

Desulfoferrodoxin (Dfx), a small iron protein containing two mononuclear iron centres (designated centre I and II), was shown to complement superoxide dismutase (SOD) deficient mutants of Escherichia coli [Pianzzola, M.J., Soubes M. & Touati, D. (1996) J. Bacteriol. 178, 6736-6742]. Furthermore, neelaredoxin, a protein from Desulfovibrio gigas containing an iron site similar to centre II of Dfx, was recently shown to have a significant SOD activity [Silva, G., Oliveira, S., Gomes, C.M., Pacheco, I., Liu, M.Y., Xavier, A.V., Teixeira, M., Le Gall, J. & Rodrigues-Pousada, C. (1999) Eur. J. Biochem. 259, 235-243]. Thus, the SOD activity of Dfx isolated from the sulphate-reducing bacterium Desulfovibrio desulfuricans ATCC 27774 was studied. The protein exhibits a SOD activity of 70 U x mg-1, which increases approximately 2.5-fold upon incubation with cyanide. Cyanide binds specifically to Dfx centre II, yielding a low-spin iron species with g-values at 2.27 (g perpendicular) and 1.96 (g parallel). Upon reaction of fully oxidized Dfx with the superoxide generating system xanthine/xanthine oxidase, Dfx centres I and II become partially reduced, suggesting that Dfx operates by a redox cycling mechanism, similar to those proposed for other SODs. Evidence for another SOD in D. desulfuricans is also presented - this enzyme is inhibited by cyanide, and N-terminal sequence data strongly indicates that it is an analogue to Cu,Zn-SODs isolated from other sources. This is the first indication that a Cu-containing protein may be present in a sulphate-reducing bacterium.     

Medium-chain acyl-coenzyme A dehydrogenase bound to a product analogue, hexadienoyl-coenzyme A: effects on reduction potential, pK(a), and polarization

2,4-Hexadienoyl-coenzyme A (HD-CoA) has been used to investigate the redox and ionization properties of medium-chain acyl-CoA dehydrogenase (MCAD) from pig kidney. HD-CoA is a thermodynamically stabilized product analogue that binds tightly to oxidized MCAD (K(dox) = 3.5 +/- 0.1 microM, pH 7.6) and elicits a redox potential shift that is 78% of that observed with the natural substrate/product couple [Lenn, N. D., Stankovich, M. T., and Liu, H. (1990) Biochemistry 29, 3709-3715]. The midpoint potential of the MCAD.HD-CoA complex exhibits a pH dependence that is consistent with the redox-linked ionization of two key glutamic acids as well as the flavin adenine dinucleotide (FAD) cofactor. The estimated ionization constants for Glu376-COOH (pK(a,ox) approximately 9.3) and Glu99-COOH (pK(a,ox) approximately 7.4) in the oxidized MCAD.HD-CoA complex indicate that while binding of the C(6) analogue makes Glu376 a stronger catalytic base (pK(a,ox) approximately 6.5, free MCAD), it has little effect on the pK of Glu99 (pK(a,ox) approximately 7.5, free MCAD) [Mancini-Samuelson, G. J., Kieweg, V., Sabaj, K. M., Ghisla, S., and Stankovich, M. T. (1998) Biochemistry 37, 14605-14612]. This finding is in agreement with the apparent pK of 9.2 determined for Glu376 in the human MCAD.4-thia-octenoyl-CoA complex [Rudik, I., Ghisla, S., and Thorpe, C. (1998) Biochemistry 37, 8437-8445]. The pK(a)s estimated for Glu376 and Glu99 in the reduced pig kidney MCAD.HD-CoA complex, 9.8 and 8.6, respectively, suggest that both of these residues remain protonated in the charge-transfer complex under physiological conditions. Polarization of HD-CoA in the enzyme active site may contribute to the observed pK(a) and redox potential shifts. Consequently, the electronic structures of the product analogue in its free and MCAD-bound forms have been characterized by Raman difference spectroscopy. Binding to either the oxidized or reduced enzyme results in localized pi-electron polarization of the hexadienoyl C(1)=O and C(2)=C(3) bonds. The C(4)=C(5) bond, in contrast, is relatively unaffected by binding. These results suggest that, upon binding to MCAD, HD-CoA is selectively polarized such that partial positive charge develops at the C(3)-H region of the ligand, regardless of the oxidation state of the enzyme.     

13C NMR and fluorescence analysis of tryptophan dynamics in wild-type and two single-Trp variants of Escherichia coli thioredoxin.

The rotational motion of tryptophan side chains in oxidized and reduced wild-type (WT) Escherichia coli thioredoxin and in two single-tryptophan variants of E. coli thioredoxin was studied in solution in the temperature range 20-50 degrees C from 13C-NMR relaxation rate measurements at 75.4 and 125.7 MHz and at 20 degrees C from steady-state and time-resolved trp fluorescence anisotropy measurements. Tryptophan enriched with 13C at the delta 1 and epsilon 3 sites of the indole ring was incorporated into WT thioredoxin and into two single-trp mutants, W31F and W28F, in which trp-28 or trp-31 of WT thioredoxin was replaced, respectively, with phenylalanine. The NMR relaxation data were interpreted using the Lipari and Szabo "model-free" approach (G. Lipari and A. Szabo. 1982. J. Amer. Chem. Soc. 104:4546-4559) with trp steady-state anisotropy data included for the variants at 20 degrees C. Values for the correlation time for the overall rotational motion (tau m) from NMR of oxidized and reduced WT thioredoxin at 35 degrees C agree well with those given by Stone et al. (Stone, M. J., K. Chandrasekhar, A. Holmgren, P. E. Wright, and H. J. Dyson. 1993. Biochemistry. 32:426-435) from 15N NMR relaxation rates, and the dependence of tau m on viscosity and temperature was in accord with the Stokes-Einstein relationship. Order parameters (S2) near 1 were obtained for the trp side chains in the WT proteins even at 50 degrees C. A slight increase in the amplitude of motion (decrease in S2) of trp-31, which is near the protein surface, but not of trp-28, which is partially buried in the protein matrix, was observed in reduced relative to oxidized WT thioredoxin. For trp-28 in W31F, order parameters near 1 (S2 > or = 0.8) at 20 degrees C were found, whereas trp-31 in W28F yielded the smallest order parameters (S2 approximately 0.6) of any of the cases. Analysis of time-resolved anisotropy decays in W28F and W31F yielded S2 values in good agreement with NMR, but gave tau m values about 60% smaller. Generally, values of tau e, the effective correlation time for the internal motion, were < or = 60 ps from NMR, whereas somewhat longer times were obtained from fluorescence. The ability of NMR and fluorescence techniques to detect subnanosecond motions in proteins reliably is examined.     

Structure of a stabilizing disulfide bridge mutant that closes the active-site cleft of T4 lysozyme.

The engineered disulfide bridge between residues 21 and 142 of phage T4 lysozyme spans the active-site cleft and can be used as a switch to control the activity of the enzyme (Matsumura, M. & Matthews, B.W., 1989, Science 243, 792-794). In the oxidized form the disulfide increases the melting temperature of the protein by 11 degrees C at pH 2. The crystal structure of this mutant lysozyme has been determined in both the reduced and oxidized forms. In the reduced form, the crystal structure of the mutant is shown to be extremely similar to that of wild type. In the oxidized form, however, the formation of the disulfide bridge causes the alpha-carbons of Cys 21 and Cys 142, on opposite sides of the active-site cleft, to move toward each other by 2.5 A. In association with this movement, the amino-terminal domain of the protein undergoes a rigid-body rotation of 5.1 degrees relative to the carboxy-terminal domain. This rotation occurs about an axis passing through the junction of the amino-terminal and carboxy-terminal domains and is also close to the axis that best fits the apparent thermal motion of the amino-terminal domain seen previously in crystals of wild-type lysozyme. Even though the engineered Cys 21-Cys 142 disulfide links together the amino-terminal and carboxy-terminal domains of T4 lysozyme, it does not reduce the apparent mobility of the one domain relative to the other. The pronounced "hinge-bending" mobility of the amino-terminal domain that is suggested by the crystallographic thermal parameters of wild-type lysozyme persists in the oxidized (and reduced) mutant structures. In the immediate vicinity of the introduced disulfide bridge the mutant structure is more mobile (or disordered) than wild type, so much so that the exact conformation of Cys 21 remains obscure. As with the previously described disulfide bridge between residues 9 and 164 of T4 lysozyme (Pjura, P.E., Matsumura, M., Wozniak, J.A., & Matthews, B.W., 1990, Biochemistry 29, 2592-2598), the engineered cross-link substantially enhances the stability of the protein without making the folded structure more rigid.     

Redox properties of the iron-sulfur clusters in activated Fe-hydrogenase from Desulfovibrio vulgaris (Hildenborough).

The periplasmic Fe-hydrogenase from Desulfovibrio vulgaris (Hildenborough) contains three iron-sulfur prosthetic groups: two putative electron transferring [4Fe-4S] ferredoxin-like cubanes (two F-clusters), and one putative Fe/S supercluster redox catalyst (one H-cluster). Combined elemental analysis by proton-induced X-ray emission, inductively coupled plasma mass spectrometry, instrumental neutron activation analysis, atomic absorption spectroscopy and colorimetry establishes that elements with Z > 21 (except for 12-15 Fe) are present in 0.001-0.1 mol/mol quantities, not correlating with activity. Isoelectric focussing reveals the existence of multiple charge conformers with pI in the range 5.7-6.4. Repeated re-chromatography results in small amounts of enzyme of very high H2-production activity determined under standardized conditions (approximately 7000 U/mg). The enzyme exists in two different catalytic forms: as isolated the protein is 'resting' and O2-insensitive; upon reduction the protein becomes active and O2-sensitive. EPR-monitored redox titrations have been carried out of both the resting and the activated enzyme. In the course of a reductive titration, the resting protein becomes activated and begins to produce molecular hydrogen at the expense of reduced titrant. Therefore, equilibrium potentials are undefined, and previously reported apparent Em and n values [Patil, D. S., Moura, J. J. G., He, S. H., Teixeira, M, Prickril, B. C., DerVartanian, D. V., Peck, H. D. Jr, LeGall, J. & Huynh, B.-H. (1988) J. Biol. Chem. 263, 18,732-18,738] are not thermodynamic quantities. In the activated enzyme an S = 1/2 signal (g = 2.11, 2.05, 2.00; 0.4 spin/protein molecule), attributed to the oxidized H cluster, exhibits a single reduction potential, Em,7 = -307 mV, just above the onset potential of H2 production. The midpoint potential of the two F clusters (2.0 spins/protein molecule) has been determined either by titrating active enzyme with the H2/H+ couple (E,m = -330 mV) or by dithionite-titrating a recombinant protein that lacks the H-cluster active site (Em,7.5 = -340 mV). There is no significant redox interaction between the two F clusters (n approximately 1).     

Contribution to the thermodynamics of protein folding from the reduction in water-accessible nonpolar surface area

Protein folding and the transfer of hydrocarbons from a dilute aqueous solution to the pure liquid phase are thermodynamically similar in that both processes remove nonpolar surface from water and both are accompanied by anomalously large negative heat capacity changes. On the basis of a limited set of published surface areas, we previously proposed that heat capacity changes (delta C degrees p) for the transfer of hydrocarbons from water to the pure liquid phase and for the folding of globular proteins exhibit the same proportionality to the reduction in water-accessible nonpolar surface area (delta Anp) [Spolar, R.S., Ha, J.H., & Record, M.T., Jr. (1989) Proc. Natl. Acad. Sci. U.S.A. 86, 8382-8385]. The consequence of this proposal is that the experimental delta C degrees p for protein folding can be used to obtain estimates of delta Anp and of the contribution to the stability of the folded state from removal of a nonpolar surface from water. In this paper, a rigorous molecular surface area algorithm [Richmond, T.J. (1984) J. Mol. Biol. 178, 63-89] is applied to obtain self-consistent values of the water-accessible nonpolar surface areas of the native and completely denatured states of the entire set of globular proteins for which both crystal structures and delta C degrees p of folding have been determined and for the set of liquid and liquefiable hydrocarbons for which delta C degrees p of transfer are known. Both processes (hydrocarbon transfer and protein folding) exhibit the same direct proportionality between delta C degrees p and delta Anp. We conclude that the large negative heat capacity changes observed in protein folding and other self-assembly processes involving proteins provide a quantitative measure of the reduction in the water-accessible nonpolar surface area and of the contribution of the hydrophobic effect to the stability of the native state and to protein assembly.     

Chaperonins facilitate the in vitro folding of monomeric mitochondrial rhodanese.

In vitro refolding of the monomeric mitochondrial enzyme, rhodanese (thiosulfate sulfurtransferase; EC 2.8.1.1) is facilitated by molecular chaperonins. The four components: two proteins from Escherichia coli, chaperonin 60 (groEL) and chaperonin 10 (groES), MgATP, and K+, are necessary for the in vitro folding of rhodanese. These were previously shown to be necessary for the in vitro folding of ribulose-1,5-bisphosphate carboxylase at temperatures in excess of 25 degrees C (Viitanen, P. V., Lubben, T. H., Reed, J., Goloubinoff, P., O'Keefe, D. P., and Lorimer, G. H. (1990) Biochemistry 29, 5665-5671). The labile folding intermediate, rhodanese-I, which rapidly aggregates at 37 degrees C in the absence of the chaperonins, can be stabilized by forming a binary complex with chaperonin 60. The discharge of the binary chaperonin 60-rhodanese-I complex, results in the formation of active rhodanese, and requires the presence of chaperonin 10. Optimal refolding is associated with a K(+)-dependent hydrolysis of ATP. At lower protein concentrations and 25 degrees C, where aggregation is reduced, a fraction of the rhodanese refolds to an active form in the absence of the chaperonins. This spontaneous refolding can be arrested by chaperonin 60. There is some refolding (approximately equal to 20%) when ATP is replaced by nonhydrolyzable analogs, but there is no refolding in the presence of ADP or AMP. ATP analogs may interfere with the interaction of rhodanese-I with the chaperonins. Nondenaturing detergents facilitate rhodanese refolding by interacting with exposed hydrophobic surfaces of folding intermediates and thereby prevent aggregation (Tandon, S., and Horowitz, P. (1986) J. Biol. Chem. 261, 15615-15618). The chaperonin proteins appear to play a similar role in as much as they can replace the detergents. Consistent with this view, chaperonin 60, but not chaperonin 10, binds 2-3 molecules of the hydrophobic fluorescent reporter, 1,1'-bi(4-anilino)naphthalene-S,5'-disulfonic acid, indicating the presence of hydrophobic surfaces on chaperonin 60. The number of bound probe molecules is reduced to 1-2 molecules when chaperonin 10 and MgATP are added. The results support a model in which chaperonins facilitate folding, at least in part, by interacting with partly folded intermediates, thus preventing the interactions of hydrophobic surfaces that lead to aggregation.     

Spectroelectrochemical evaluation of redox potentials of cysteine tryptophylquinone and two hemes c in quinohemoprotein amine dehydrogenase from Paracoccus denitrificans

Quinohemoprotein amine dehydrogenase (QH-AmDH) from Paracoccus denitrificans has a novel cofactor cysteine tryptophylquinone (CTQ) in the smallest gamma subunit and two hemes c in the largest alpha subunit [Datta, S., Mori, Y., Takagi, K., Kawaguchi, K., Chen, Z., Okajima, T., Kuroda, S., Ikeda, T., Kano, K., Tanizawa, K., and Mathews, F. S. (2001) Proc. Natl. Acad. Sci. U.S.A. 98, 14268-14273]. The spectral change of QH-AmDH was assigned to the redox reaction of the hemes c alone. The redox potentials of the two hemes c with His and Met as the second axial ligands, respectively, were determined to be 0.149 and 0.235 V versus SHE at pH 7.0 by a mediator-assisted continuous-flow column electrolytic spectroelectrochemistry (MCES). The monomeric gamma subunit of QH-AmDH was isolated from urea-treated QH-AmDH. The fully oxidized and reduced forms of the gamma subunit exhibited a unique absorption band centered at 380 nm and a shoulder band around 315 nm, respectively, at neutral pH. The two-electron redox potential of CTQ in the isolated gamma subunit was evaluated to be 65 mV at pH 7.0 by MCES. The redox reaction was linked to the two-proton transfer at pH <8.6 and to a single-proton transfer at pH >8.6. The pK(a) value (K(a) being the acid dissociation constant) of 8.6 was assigned to one of the phenolic OH groups of the quinol form. Upon deprotonation, the red shift of the shoulder band was observed. The gamma subunit adsorbed on a glassy carbon electrode, and gave a direct but quasi-reversible electrochemical signal. Intra- and interprotein electron transfers of QH-AmDH are discussed from thermodynamic and structural points of view.