Crystallographic studies of V44 mutants of Clostridium pasteurianum rubredoxin: effects of side-chain size on reduction potential.

Understanding the structural origins of differences in reduction potentials is crucial to understanding how various electron transfer proteins modulate their reduction potentials and how they evolve for diverse functional roles. Here, the high-resolution structures of several Clostridium pasteurianum rubredoxin (Cp Rd) variants with changes in the vicinity of the redox site are reported in order to increase this understanding. Our crystal structures of [V44L] (at 1.8 A resolution), [V44A] (1.6 A), [V44G] (2.0 A) and [V44A, G45P] (1.5 A) Rd (all in their oxidized states) show that there is a gradual decrease in the distance between Fe and the amide nitrogen of residue 44 upon reduction in the size of the side chain of residue 44; the decrease occurs from leucine to valine, alanine or glycine and is accompanied by a gradual increase in their reduction potentials. Mutation of Cp Rd at position 44 also changes the hydrogen-bond distance between the amide nitrogen of residue 44 and the sulfur of cysteine 42 in a size-dependent manner. Our results suggest that residue 44 is an important determinant of Rd reduction potential in a manner dictated by side-chain size. Along with the electric dipole moment of the 43-44 peptide bond and the 44-42 NH--S type hydrogen bond, a modulation mechanism for solvent accessibility through residue 41 might regulate the redox reaction of the Rds.

The unique hydrogen bonded water in the reduced form of Clostridium pasteurianum rubredoxin and its possible role in electron transfer.

Rubredoxin is a small iron-sulfur (FeS4) protein involved in oxidation-reduction reactions. The side chain of Leu41 near the iron-sulfur center has two conformations, which we suggested previously serve as a gate for a water molecule during the electron transfer process. To establish the role of residue 41 in electron transfer, an [L41A] mutant of Clostridium pasteurianum rubredoxin was constructed and crystallized in both oxidation states. Despite the lack of the gating side chain in this protein, the structure of the reduced [L41A] rubredoxin reveals a specific water molecule in the same position as observed in the reduced wild-type rubredoxin. In contrast, both the wild-type and [L41A] rubredoxins in the oxidized state do not have water molecules in this location. The reduction potential of the [L41A] variant was approximately 50 mV more positive than wild-type. Based on these observations, it is proposed that the site around the Sgamma of Cys9 serves as a port for an electron acceptor. Lastly, the Fe-S distances of the reduced rubredoxin are expanded, while the hydrogen bonds between Sgamma of the cysteines and the backbone amide nitrogens are shortened compared to its oxidized counterpart. This small structural perturbation in the Fe(II)/Fe(III) transition is closely related to the small energy difference which is important in an effective electron transfer agent.

Contribution of the [FeII(SCys)4] site to the thermostability of rubredoxins.

The thermostabilities of Fe(2+) ligation in rubredoxins (Rds) from the hyperthermophile Pyrococcus furiosus (Pf) and the mesophiles Clostridium pasteurianum (Cp) and Desulfovibrio vulgaris (Dv) were compared. Residue 44 forms an NH.S(Cys) hydrogen bond to one of the cysteine ligands to the [Fe(SCys)(4)] site, and substitutions at this location affect the redox properties of the [Fe(SCys)(4)] site. Both Pf Rd and Dv Rd have an alanine residue at position 44, whereas Cp Fd has a valine residue. Wild-type proteins were examined along with V44A and A44V "exchange" mutants of Cp and Pf Rds, respectively, in order to assess the effects of the residue at position 44 on the stability of the [Fe(SCys)(4)] site. Stability of iron ligation was measured by temperature-ramp and fixed-temperature time course experiments, monitoring iron release in both the absence and presence of more thiophilic metals (Zn(2+), Cd(2+)) and over a range of pH values. The thermostability of the polypeptide fold was concomitantly measured by fluorescence, circular dichroism, and (1)H NMR spectroscopies. The A44V mutation strongly lowered the stability of the [Fe(II)(SCys)(4)] site in Pf Rd, whereas the converse V44A mutation of Cp Rd significantly raised the stability of the [Fe(II)(SCys)(4)] site, but not to the levels measured for wild-type Dv Rd. The region around residue 44 is thus a significant contributor to stability of iron coordination in reduced Rds. This region, however, made only a minor contribution to the thermostability of the protein folding, which was found to be higher for hyperthermophilic versus mesophilic Rds, and largely independent of the residue at position 44. These results, together with our previous studies, show that localized charge density, solvent accessibility, and iron site/backbone interactions control the thermostability of the [Fe(SCys)(4)] site. The iron site thermostability does make a minor contribution to the overall Rd thermostability. From a mechanistic standpoint, we also found that attack of displacing ions (H(+), Cd(2+)) on the Cys42 sulfur ligand at the [Fe(SCys)(4)] site occurs through the V8 side and not the V44 side of the iron site.

Prediction of reduction potential changes in rubredoxin: a molecular mechanics approach.

Predicting the effects of mutation on the reduction potential of proteins is crucial in understanding how reduction potentials are modulated by the protein environment. Previously, we proposed that an alanine vs. a valine at residue 44 leads to a 50-mV difference in reduction potential found in homologous rubredoxins because of a shift in the polar backbone relative to the iron site due to the different side-chain sizes. Here, the aim is to determine the effects of mutations to glycine, isoleucine, and leucine at residue 44 on the structure and reduction potential of rubredoxin, and if the effects are proportional to side-chain size. Crystal structure analysis, molecular mechanics simulations, and experimental reduction potentials of wild-type and mutant Clostridium pasteurianum rubredoxin, along with sequence analysis of homologous rubredoxins, indicate that the backbone position relative to the redox site as well as solvent penetration near the redox site are both structural determinants of the reduction potential, although not proportionally to side-chain size. Thus, protein interactions are too complex to be predicted by simple relationships, indicating the utility of molecular mechanics methods in understanding them.

Hydrogen bonds in rubredoxins from mesophilic and hyperthermophilic organisms.

The extent and strength of the hydrogen bond networks in rubredoxins from the hyperthermophile Pyrococcus furiosus (PfRd), and its mesophilic analogue Clostridium pasteurianum (CpRd), are examined and compared using NMR spectroscopy. NMR parameters examined in this study include through-hydrogen bond (h3)J(NC)(') scalar couplings and (1)H, (13)C, and (15)N chemical shifts, as well as covalent (1)J(NH) and (1)J(NC)(') scalar couplings. These parameters have allowed the characterization in solution of 12 hydrogen bonds in each protein. Despite a 83% sequence homology and a low RMSD for the backbone heavy atoms (0.648 A) in the crystalline state, subtle, but definite, changes have been identified in the detailed hydrogen-bonding patterns. CpRd shows an increased number of hydrogen bonds in the triple-stranded beta-sheet and an additional hydrogen bond in the multiple-turn segment including residues 14-32. On the other hand, PfRd exhibits an overall strengthening of N-H...O=C hydrogen bonds in the loops involved at the metal binding site as well as evidence for an additional NH...S(Cys) hydrogen bond involving the alanine residue 44. These data, as well as temperature dependence of the NMR parameters, suggest that the particular NMR hydrogen bond pattern found in the hyperthermophile rubredoxin leads to an increased stabilization at the metal binding pocket. It seems to result from a subtle redistribution of hydrogen-bonding interactions between the triple-stranded beta-sheet and the actual metal binding site.

The [4Fe-4S](2+) cluster in reconstituted biotin synthase binds S-adenosyl-L-methionine.

The combination of resonance Raman, electron paramagnetic resonance and Mössbauer spectroscopies has been used to investigate the effect of S-adenosyl-l-methionine (SAM) on the spectroscopic properties of the [4Fe-4S]2+ cluster in biotin synthase. The results indicate that SAM interacts directly at a unique iron site of the [4Fe-4S]2+ cluster in BioB and support the hypothesis of a common inner-sphere mechanism for the reductive cleavage of SAM in the radical SAM family of Fe-S enzymes.

Recombinant Escherichia coli biotin synthase is a [2Fe-2S](2+) protein in whole cells.

EPR and Mössbauer spectroscopies have been used to determine the type and properties of the iron-sulfur clusters present in homologously expressed recombinant Escherichia coli BioB in whole cells prior to purification. Difference EPR spectra of samples of whole cells from a strain over-expressing E. coli BioB and a strain containing the same plasmid but without the bioB insertion showed an axial S=1/2 resonance that was attributed to the [2Fe-2S](+) cluster of the E. coli iron-sulfur cluster assembly 2Fe ferredoxin, based on principal g-values, linewidths and relaxation behavior. Comparison of the Mössbauer spectra of whole cells with and without the bioB insertion revealed that the E. coli cells with over-expressed BioB contain an additional species that exhibits a spectrum identical to that of the [2Fe-2S](2+) cluster in purified recombinant BioB. The concentration of this [2Fe-2S](2+) species in the whole cell sample was quantified using a Mössbauer standard and found to be approximately 260 microM, which was comparable to the BioB protein concentration estimated for the cell paste. The results demonstrate that the [2Fe-2S](2+) cluster found in purified samples of recombinant BioB is not an artifact of the protein purification procedure, and indicate that recombinant BioB is over-expressed in an inactive form during aerobic growth.

Thermal stability of the [Fe(SCys)(4)] site in Clostridium pasteurianum rubredoxin: contributions of the local environment and Cys ligand protonation.

Thermal denaturation of the mesophilic rubredoxin from Clostridium pasteurianum occurs through a number of temperature-dependent steps, the last and irreversible one being release of iron from the [Fe(2+)(SCys)(4)] site. We show here that thermally induced [Fe(2+)(SCys)(4)] site destruction is largely determined by the local environment, and not directly connected to thermostability of the native polypeptide fold of rubredoxin. Hydrophobic residues on the protein surface, V8 and L41, that shield the [Fe(SCys)(4)] site from solvent and form N-H(.)S hydrogen bonds to the metal-coordinating sulfurs, were mutated to residues with both uncharged and charged side chains. On these mutated rubredoxins the temperature dependence was measured for: (1) global unfolding of the protein by NMR, (2) loss of Fe(2+)at various ionic strengths and pH values, (3) the rates of non-denaturing displacement of Fe(2+) by Cd(2+) or Zn(2+). For reversible temperature-dependent changes in the global protein folding that occur prior to loss of iron, no thermostability differences were found among the wild-type, V8A, V8D, L41R, and L41D rubredoxins. However, for irreversible loss of iron from the [Fe(2+)(SCys)(4)] site, relative to the wild-type protein, L41R was more thermostable, V8A was somewhat less thermostable, and the acidic mutants L41D, V8D and [V8D, L41D] showed dramatically lowered thermostability. Lower pH facilitated - both kinetically and thermodynamically - thermally induced iron release, likely through protonation of ligand cysteines' thiols. For all of the rubredoxins a direct correlation was found between the midpoint temperature for thermally induced Fe(2+) loss and the rate of non-denaturing Fe(2+) displacement by Cd(2+) or Zn(2+) at room temperature. A mechanism is proposed involving transient movement of residue-8 and -41 side chains, allowing, and, in the case of negatively charged side chains, also facilitating, attack of a ligand cysteine by the incoming positively charged species (H(+), Cd(2+), or Zn(2+)). Thus, localized charge density and solvent accessibility modulate the stability of Fe(2+) ligation in rubredoxin. However, the reduced [Fe(SCys)(4)] site does not control the thermostability of the native polypeptide fold of rubredoxin.