High Resolution Crystal Structures of the Wild Type and Cys-55 → Ser and Cys-59 → Ser Variants of the Thioredoxin-like [2Fe-2S] Ferredoxin from Aquifex aeolicus

Artigo Acesso aberto Revisado por pares

High Resolution Crystal Structures of the Wild Type and Cys-55 → Ser and Cys-59 → Ser Variants of the Thioredoxin-like [2Fe-2S] Ferredoxin from Aquifex aeolicus

2002; Elsevier BV; Volume: 277; Issue: 37 Linguagem: Inglês

10.1074/jbc.m205096200

ISSN

1083-351X

Autores

Andrew Yeh, Xavier Ambroggio, Susana L. A. Andrade, Oliver Einsle, Claire Chatelet, Jacques Meyer, Douglas C. Rees,

Tópico(s)

Enzyme function and inhibition

Resumo

The [2Fe-2S] ferredoxin (Fd4) from Aquifex aeolicus adopts a thioredoxin-like polypeptide fold that is distinct from other [2Fe-2S] ferredoxins. Crystal structures of the Cys-55 → Ser (C55S) and Cys-59 → Ser (C59S) variants of this protein have been determined to 1.25 Å and 1.05 Å resolution, respectively, whereas the resolution of the wild type (WT) has been extended to 1.5 Å. The improved WT structure provides a detailed description of the [2Fe-2S] cluster, including two features that have not been noted previously in any [2Fe-2S] cluster-containing protein, namely, pronounced distortions in the cysteine coordination to the cluster and a Cα-H–Sγ hydrogen bond between cluster ligands Cys-55 and Cys-9. These features may contribute to the unusual electronic and magnetic properties of the [2Fe-2S] clusters in WT and variants of this ferredoxin. The structures of the two variants of Fd4, in which single cysteine ligands to the [2Fe-2S] cluster are replaced by serine, establish the metric details of serine-ligated Fe-S active sites with unprecedented accuracy. Both the cluster and its surrounding protein matrix change in subtle ways to accommodate this ligand substitution, particularly in terms of distortions of the Fe2S2 inorganic core from planarity and displacements of the polypeptide chain. These high resolution structures illustrate how the interactions between polypeptide chains and Fe-S active sites reflect combinations of flexibility and rigidity on the part of both partners; these themes are also evident in more complex systems, as exemplified by changes associated with serine ligation of the nitrogenase P cluster. The [2Fe-2S] ferredoxin (Fd4) from Aquifex aeolicus adopts a thioredoxin-like polypeptide fold that is distinct from other [2Fe-2S] ferredoxins. Crystal structures of the Cys-55 → Ser (C55S) and Cys-59 → Ser (C59S) variants of this protein have been determined to 1.25 Å and 1.05 Å resolution, respectively, whereas the resolution of the wild type (WT) has been extended to 1.5 Å. The improved WT structure provides a detailed description of the [2Fe-2S] cluster, including two features that have not been noted previously in any [2Fe-2S] cluster-containing protein, namely, pronounced distortions in the cysteine coordination to the cluster and a Cα-H–Sγ hydrogen bond between cluster ligands Cys-55 and Cys-9. These features may contribute to the unusual electronic and magnetic properties of the [2Fe-2S] clusters in WT and variants of this ferredoxin. The structures of the two variants of Fd4, in which single cysteine ligands to the [2Fe-2S] cluster are replaced by serine, establish the metric details of serine-ligated Fe-S active sites with unprecedented accuracy. Both the cluster and its surrounding protein matrix change in subtle ways to accommodate this ligand substitution, particularly in terms of distortions of the Fe2S2 inorganic core from planarity and displacements of the polypeptide chain. These high resolution structures illustrate how the interactions between polypeptide chains and Fe-S active sites reflect combinations of flexibility and rigidity on the part of both partners; these themes are also evident in more complex systems, as exemplified by changes associated with serine ligation of the nitrogenase P cluster. ferredoxin(s) [2Fe-2S] ferredoxin 4 from A. aeolicus 4-morpholineethanesulfonic acid wild type The low potential iron-sulfur (Fe-S) electron carriers known as ferredoxins (Fds)1 are found in three distinct classes, the [3Fe-4S]/[4Fe-4S] bacterial type Fds, the plant and mammalian type [2Fe-2S] Fds, and the thioredoxin-like [2Fe-2S] Fds (1Meyer J. FEBS Lett. 2001; 509: 1-5Crossref PubMed Scopus (67) Google Scholar). The first two classes were discovered nearly 40 years ago (2Mortenson L.E. Valentine R.C. Carnahan J.E. Biochem. Biophys. Res. Commun. 1962; 7: 448-452Crossref PubMed Scopus (167) Google Scholar, 3Tagawa K. Arnon D.I. Nature. 1962; 195: 537-543Crossref PubMed Scopus (361) Google Scholar) and have since been characterized in considerable detail, including by high resolution x-ray crystallography (4Dauter Z. Wilson K.S. Sieker L.C. Meyer J. Moulis J.M. Biochemistry. 1997; 36: 16065-16073Crossref PubMed Scopus (129) Google Scholar, 5Morales R. Charon M.H. Hudry-Clergeon G. Peátillot Y. Nørager S. Medina M. Frey M. Biochemistry. 1999; 38: 15764-15773Crossref PubMed Scopus (127) Google Scholar). The third class of Fd is more sparsely distributed and, therefore, has not been investigated as thoroughly (1Meyer J. FEBS Lett. 2001; 509: 1-5Crossref PubMed Scopus (67) Google Scholar). The best characterized members of that group are [2Fe-2S] Fds from the bacteria Clostridium pasteurianum (6Golinelli M.P. Chatelet C. Duin E.C. Johnson M.K. Meyer J. Biochemistry. 1998; 37: 10429-10437Crossref PubMed Scopus (29) Google Scholar),Azotobacter vinelandii (7Chatelet C. Meyer J. J. Biol. Inorg. Chem. 1999; 4: 311-317Crossref PubMed Scopus (19) Google Scholar), and Aquifex aeolicus(8Chatelet C. Gaillard J. Peátillot Y. Louwagie M. Meyer J. Biochem. Biophys. Res. Commun. 1999; 261: 885-889Crossref PubMed Scopus (24) Google Scholar). The high level of similarity of these proteins allows for easy transfer of structural information among them. For instance, many properties of molecular variants of the C. pasteurianum[2Fe-2S] Fd (6Golinelli M.P. Chatelet C. Duin E.C. Johnson M.K. Meyer J. Biochemistry. 1998; 37: 10429-10437Crossref PubMed Scopus (29) Google Scholar, 9Meyer J. Fujinaga J. Gaillard J. Lutz M. Biochemistry. 1994; 33: 13642-13650Crossref PubMed Scopus (71) Google Scholar, 10Golinelli M.P. Akin L.A. Crouse B.R. Johnson M.K. Meyer J. Biochemistry. 1996; 35: 8995-9002Crossref PubMed Scopus (36) Google Scholar) could be rationalized from the crystal structure of A. aeolicus Fd4 (11Yeh A.P. Chatelet C. Soltis S.M. Kuhn P. Meyer J. Rees D.C. J. Mol. Biol. 2000; 300: 587-595Crossref PubMed Scopus (50) Google Scholar). This structure established the unexpected thioredoxin-like fold of these Fds and confirmed that they are distinct from the other two ferredoxin classes (1Meyer J. FEBS Lett. 2001; 509: 1-5Crossref PubMed Scopus (67) Google Scholar). The 2.3 Å resolution of the Fd4 structure, however, was not among the highest currently reported (about 1 Å) for metalloenzymes. Indeed, such high resolution structures are of utmost interest because they bring forth precise geometries of metal sites (4Dauter Z. Wilson K.S. Sieker L.C. Meyer J. Moulis J.M. Biochemistry. 1997; 36: 16065-16073Crossref PubMed Scopus (129) Google Scholar) and may allow description of redox transitions (5Morales R. Charon M.H. Hudry-Clergeon G. Peátillot Y. Nørager S. Medina M. Frey M. Biochemistry. 1999; 38: 15764-15773Crossref PubMed Scopus (127) Google Scholar). Additional efforts have therefore been made, both on wild type (WT) thioredoxin-like Fds and on several of the molecular variants that were produced over the years (6Golinelli M.P. Chatelet C. Duin E.C. Johnson M.K. Meyer J. Biochemistry. 1998; 37: 10429-10437Crossref PubMed Scopus (29) Google Scholar, 10Golinelli M.P. Akin L.A. Crouse B.R. Johnson M.K. Meyer J. Biochemistry. 1996; 35: 8995-9002Crossref PubMed Scopus (36) Google Scholar,12Chatelet C. Meyer J. Biochim. Biophys. Acta. 2001; 1549: 32-36Crossref PubMed Scopus (8) Google Scholar), with the aim of improving the crystallographic data. In some modified forms of C. pasteurianum Fd, cysteine ligands of the Fe-S cluster were replaced by serine (9Meyer J. Fujinaga J. Gaillard J. Lutz M. Biochemistry. 1994; 33: 13642-13650Crossref PubMed Scopus (71) Google Scholar, 13Fujinaga J. Gaillard J. Meyer J. Biochem. Biophys. Res. Commun. 1993; 194: 104-111Crossref PubMed Scopus (39) Google Scholar). In their reduced [2Fe-2S]+ level, these serine-ligated active sites were found to assume a delocalized mixed valence state having a ground spin state of 9/2 (14Crouse B.R. Meyer J. Johnson M.K. J. Am. Chem. Soc. 1995; 117: 9612-9613Crossref Scopus (78) Google Scholar, 15Achim C. Golinelli M.P. Bominaar E.L. Meyer J. Mu¨nck E. J. Am. Chem. Soc. 1996; 118: 8168-8169Crossref Scopus (73) Google Scholar). This unprecedented occurrence in binuclear iron-sulfur clusters has stimulated experimental and theoretical work (16Bominaar E.L. Achim C. Borshch S.A. J. Chem. Phys. 1999; 110: 11411-11422Crossref Scopus (8) Google Scholar, 17Achim C. Bominaar E.L. Meyer J. Peterson J. Mu¨nck E. J. Am. Chem. Soc. 1999; 121: 3704-3714Crossref Scopus (52) Google Scholar). The development of these investigations has been hampered, however, by the absence of structural data on serine-ligated [2Fe-2S] active sites and their environment. Because only the Fd4 from A. aeolicus has been crystallized, we have repeated amino acid substitutions in that protein previously performed on C. pasteurianum Fd and produced the C55S and C59S variants that contain serine-ligated [2Fe-2S] clusters. We report here the high resolution structures for both of these variants as well as WT protein which provide accurate metric details for serine-ligated Fe-S clusters in proteins. Fd4 from A. aeolicus was purified as described by Chatelet et al. (8Chatelet C. Gaillard J. Peátillot Y. Louwagie M. Meyer J. Biochem. Biophys. Res. Commun. 1999; 261: 885-889Crossref PubMed Scopus (24) Google Scholar). The C55S and C59S variants were prepared by site-directed mutagenesis as described for the C56S and C60S counterparts from C. pasteurianum(13Fujinaga J. Gaillard J. Meyer J. Biochem. Biophys. Res. Commun. 1993; 194: 104-111Crossref PubMed Scopus (39) Google Scholar). The mutagenic oligonucleotides were 5′-cacgcgttcatgGaaccggtgggag-3′ (hybridizing to the coding strand, mutated base in underlined uppercase) and 5′-ggttgcatgaacgcgtCtatgatgggaccg-3′ (hybridizing to the noncoding strand), for C55S and C59S, respectively. The mutated genes were overexpressed in Escherichia coli, and the C55S and C59S proteins were purified as described for the WT (8Chatelet C. Gaillard J. Peátillot Y. Louwagie M. Meyer J. Biochem. Biophys. Res. Commun. 1999; 261: 885-889Crossref PubMed Scopus (24) Google Scholar). Crystals of oxidized WT, C55S, and C59S Fd4 were prepared by the sitting drop vapor diffusion method. Although anaerobic conditions were employed to minimize degradation of the cluster by exposure to atmospheric oxygen, no reductants were present during the crystallizations, so that the proteins should remain in the oxidized state. In the case of C55S, crystals were obtained by equilibrating 2 μl of reservoir solution and 2 μl of ∼83 mg/ml Fd4 C55S (in 20 mm Tris-HCl buffer at pH 8.0 and 0.2m NaCl) against a reservoir solution containing 30% (w/v) polyethylene glycol 4000, 0.2 m ammonium acetate, and 0.1m sodium acetate at pH 4.6. Crystals of Fd4 C59S were obtained by equilibrating 2 μl of reservoir solution and 2 μl of ∼67 mg/ml Fd4 C59S (in 20 mm Tris-HCl buffer at pH 8.0 and 0.2 m NaCl) against a reservoir solution containing 1.0m 1,6-hexanediol, 0.01 m cobalt chloride and 0.1 m sodium acetate at pH 4.6. WT Fd4 was crystallized by equilibrating 2 μl of ∼10 mg/ml Fd4 (in 10 mm Tris-HCl buffer at pH 8.0 and 0.2 m NaCl) and 2 μl of reservoir solution against a reservoir solution containing 0.01 mzinc sulfate heptahydrate, 0.1 m MES buffer at pH 6.5, and 25% polyethylene glycol monomethyl ether 550. Despite the different crystallization conditions, in all three cases nearly isomorphous crystals were obtained in space group C2 (C55S:a = 67.3 Å, b = 59.8 Å,c = 46.9 Å, β = 109.8o; C59S:a = 67.3 Å, b = 59.8 Å,c = 46.8 Å, β = 109.3o; WT:a = 67.2 Å, b = 59.8 Å,c = 47.2 Å, β = 110.3o), with one dimeric Fd4 molecule/asymmetric unit. Diffraction data to 1.25 Å resolution for C55S and 1.05 Å resolution for C59S were collected under cryogenic conditions on beamline 9-2 at the Stanford Synchrotron Radiation Laboratory on an Area Detector Systems Corp. Quantum-4 CCD detector controlled by the distributed control system software BLU-ICE. C55S and C59S data sets were processed and scaled using DENZO and SCALEPACK (18Otwinowski Z. Minor W. Methods Enzymol. 1997; 276: 307-326Crossref PubMed Scopus (38617) Google Scholar). Diffraction data to 1.5 Å resolution for WT Fd4 were collected under cryogenic conditions at the Stanford Synchrotron Radiation Laboratory beamline 9-2 on an Area Detector Systems Corp. Quantum-315 CCD detector and were processed and scaled using MOSFLM and SCALA (19Bailey S. Acta Crystallogr. Sect. D Biol. Crystallogr. 1994; 50: 760-763Crossref PubMed Scopus (42) Google Scholar). A summary of the data collection statistics is listed in TableI.Table ISummary of data collection and refinement statisticsWTC55SC59SData collection statistics Maximum resolution (Å)1.501.251.05 Wavelength (Å)0.99180.88600.9580 Total reflections98,554183,676358,550 Unique reflections27,75848,25677,950 Completeness (%)1-aNumbers in parentheses correspond to values in the highest resolution shell.98.9 (97.1)99.7 (99.5)95.6 (92.0) I/ς(I)7.8 (1.5)28.5 (3.8)38.2 (6.4) Rsym(%)1-bRsym = (ΣhklΣi‖Ii(hkl)− < I(hkl) > ‖)/(ΣhklΣi I(hkl)).5.1 (40.4)4.7 (26.1)3.5 (21.6)Refinement statistics Resolution limits (Å)31.5–1.5044.3–1.2531.2–1.05 R-factor1-cR-factor = Σ(‖Fobs‖ − ‖Fcalc)/Σ‖Fobs‖,R-free is the R-factor calculated for a 3% test set of reflections excluded from the refinement calculation.0.1840.1440.138 R-free0.2160.1960.162 Estimated coordinate error (Å)1-dCoordinate errors were obtained from the diffraction component precision index, calculated from the values of R-free by the method of Cruickshank (42).0.090.050.03 RMS deviations from ideal values Bond lengths (Å)0.0240.0150.017 Bond angles (°)2.1923.0562.666 Dihedral angles (°)25.1126.4826.38 Improper torsion angles (°)1.601.641.83 Average temperature factor (Å2) Protein22.1, 20.621.0, 19.918.1, 16.1 Iron-sulfur13.0, 12.812.9, 12.29.5, 9.7 Water36.839.331.9 Zinc37.1 Sulfate44.3 Ramachandran plot,1-eAs determined by PROCHECK (43). residues in Most favored regions (%)90.293.293.2 Additional allowed regions (%)9.16.86.8 Generously allowed regions (%)0.60.00.0 Disallowed regions (%)0.00.00.01-a Numbers in parentheses correspond to values in the highest resolution shell.1-b Rsym = (ΣhklΣi‖Ii(hkl)− < I(hkl) > ‖)/(ΣhklΣi I(hkl)).1-c R-factor = Σ(‖Fobs‖ − ‖Fcalc)/Σ‖Fobs‖,R-free is the R-factor calculated for a 3% test set of reflections excluded from the refinement calculation.1-d Coordinate errors were obtained from the diffraction component precision index, calculated from the values of R-free by the method of Cruickshank (42Cruickshank D.W.J. Acta Crystallogr. Sect. D Biol. Crystallogr. 1999; 55: 583-601Crossref PubMed Scopus (501) Google Scholar).1-e As determined by PROCHECK (43Laskowski R.A. MacArthur M.W. Moss D.S. Thornton J.M. J. Appl. Crystallogr. 1993; 26: 283-291Crossref Google Scholar). Open table in a new tab The Fd4 C55S structure was solved by molecular replacement using EPMR (20Kissinger C.R. Gehlhaar D.K. Fogel D.B. Acta Crystallogr. Sect. D Biol. Crystallogr. 1999; 55: 484-491Crossref PubMed Scopus (691) Google Scholar) with the original WT Fd4 structure determined at 2.3 Å resolution (11Yeh A.P. Chatelet C. Soltis S.M. Kuhn P. Meyer J. Rees D.C. J. Mol. Biol. 2000; 300: 587-595Crossref PubMed Scopus (50) Google Scholar) as the search model. Multiple rounds of positional refinement and individual isotropic B-factor refinement with crystallography NMR software (21Bru¨nger A.T. Adams P.D. Clore G.M. DeLano W.L. Gros P. Grosse-Kunstleve R.W. Jiang J.-S. Kuszewski J. Nilges M. Pannu N.S. Read R.J. Rice L.M. Simonson T. Warren G.L. Acta Crystallogr. Sect. D Biol. Crystallogr. 1998; 54: 905-921Crossref PubMed Scopus (16979) Google Scholar) were alternated with model rebuilding in the molecular graphics program O (22Jones T.A. Zou J.Y. Cowan S.W. Kjeldgaard M. Acta Crystallogr. Sect. A. 1991; 47: 110-119Crossref PubMed Scopus (13014) Google Scholar) against 2 Fo − Fcςa-weighted and Fo − Fcςa-weighted maps (23Read R.J. Acta Crystallogr. Sect. A. 1986; 42: 140-149Crossref Scopus (2037) Google Scholar). The [2Fe-2S] cluster geometry was not restrained during refinement. Upon solvent addition and completion of refinement with CNS, positional and anisotropic B-factor refinements of the model were performed using the programs SHELX97 (24Sheldrick G.M. Schneider T.R. Methods Enzymol. 1997; 277: 319-343Crossref PubMed Scopus (1892) Google Scholar) and REFMAC5 (25Murshudov G.N. Vagin A.A. Dodson E.J. Acta Crystallogr. 1997; 53: 240-255Crossref PubMed Scopus (13914) Google Scholar), which resulted in a finalR-factor and R-free of 14.4 and 19.6%, respectively. A final round of refinement in SHELX97 yielded the standard uncertainties in atomic coordinates, bond lengths, and angles. The final Fd4 C55S model comprises two subunits (2 × 101 residues, 1,575 atoms), two [2Fe-2S] clusters (8 atoms), and 215 water molecules. Because of the absence of electron density for the first 2 residues at the N terminus and the last 7 residues at the C terminus, these residues were not modeled. The Fd4 C55S model without water molecules was used as the starting model for the Fd4 C59S model. The Fd4 C59S model was refined using a protocol similar to that outlined above for Fd4 C55S to anR-factor and R-free of 13.8 and 16.2%, respectively. As with Fd4 C55S, no electron density was present for the first 2 residues at the N terminus, and the last 7 residues at the C terminus were absent. The final Fd4 C59S model consists of two subunits (2 × 101 residues, 1,576 atoms), two [2Fe-2S] clusters (8 atoms), and 198 water molecules. The Fd4 C55S model without water molecules was also used as the starting model for the high resolution WT structure. The model was refined using CNS (21Bru¨nger A.T. Adams P.D. Clore G.M. DeLano W.L. Gros P. Grosse-Kunstleve R.W. Jiang J.-S. Kuszewski J. Nilges M. Pannu N.S. Read R.J. Rice L.M. Simonson T. Warren G.L. Acta Crystallogr. Sect. D Biol. Crystallogr. 1998; 54: 905-921Crossref PubMed Scopus (16979) Google Scholar) with a protocol similar to that described above to an R-factor and R-free of 18.4 and 21.6%, respectively. The cluster geometry was not restrained during refinement. During the solvent addition process, 4 zinc ions and 1 sulfate anion were modeled into difference electron density peaks that were significantly higher than those corresponding to the water molecules. In contrast to the variant structures, the WT temperature factors were not refined anisotropically because of the lower resolution of the data. The final model comprises two subunits (103 residues in subunit A, 101 residues in subunit B, 1,590 atoms), two [2Fe-2S] clusters (8 atoms), 187 water molecules, 4 zinc ions, and 1 sulfate anion. Final refinement statistics for all three structures are listed in Table I. The WT and serine-substituted forms of Fd4 analyzed in this study were crystallized in a monoclinic space group,C2, which is distinct from the original tetragonal form solved at 2.3 Å resolution (11Yeh A.P. Chatelet C. Soltis S.M. Kuhn P. Meyer J. Rees D.C. J. Mol. Biol. 2000; 300: 587-595Crossref PubMed Scopus (50) Google Scholar). As observed originally, the current structure of WT Fd4 determined at 1.5 Å resolution exists as a homodimer, with each monomer adopting a thioredoxin-like fold (Fig.1A). The two noncrystallographic symmetry-related subunits are nearly identical, with a root mean square deviation of 0.27 Å between 101 Cα atoms. As a consequence of the differences in crystal packing between the original and present WT structures, the two monomers in the Fd4 dimer undergo a slight rigid body shift relative to each other (Fig.1A). With the A subunits of both WT forms superimposed, the shift in the B subunit of the new WT form relative to that of the old form can be characterized quantitatively as a 3.8° rotation about an axis oriented ∼74° from the 2-fold rotation axis relating subunits in the dimer. The axis about which this 3.8° rotation occurs passes near residue Thr-B53, which along with Pro-B52, Gly-B54, and the corresponding residues in subunit A, form a short stretch of antiparallel β-sheet which stabilizes the dimer interface. As a result of this change in dimer packing, the hydrogen bonding geometries of residues in this antiparallel arrangement of β-strands are slightly modified. Because cluster ligand Cys-55 is adjacent to this region, it is possible that these alterations in subunit-subunit packing could be coupled to changes in cluster environment. The differences in crystal interactions are also reflected in changes in conformations in two flexible loop regions spanning residues 13–20 and residues 39–46 (Fig. 1A), the former of which is near the cluster ligands Cys-9 and Cys-22. The [2Fe-2S] cluster, located near the surface of each monomer, is coordinated by 4 cysteines, with Cys-9 and Cys-22 ligating Fe1 and Cys-55 and Cys-59 ligating Fe2 (Fig.1B). At the resolution of the current study, it was possible to conduct the refinement without restraining the cluster geometry, resulting in stereochemical parameters that are both more accurate and also less biased than in the previous WT model. A [2Fe-2S] cluster coordinated by four sulfhydryl groups may be idealized as a framework consisting of two edge-sharing tetrahedral iron sites with a planar Fe2S2 inorganic core. The geometries of real [2Fe-2S] clusters, as observed in both model compounds and in proteins, generally reflect this expectation, although deviations from this idealization are evident (26Berg J.M. Holm R.H. Spiro T.G. Iron Sulfur Proteins. John Wiley & Sons, New York1982: 1-66Google Scholar). TableII summarizes the average stereochemical parameters for [2Fe-2S] clusters in the Fd4 structures described here as well as in proteins refined at high resolution and in model compounds. In general, the bond distances and angles observed in protein-bound [2Fe-2S] clusters agree well with each other and with model compounds. A notable deviation from the idealized symmetry in the protein-bound clusters, however, is the nonplanarity of the Fe2S2 core, which can be characterized by the average absolute value of ∼175° for the Fe-S-S-Fe torsion angle, where 180° would correspond to exact planarity. For further comparison, the Fe2S2 unit present in [4Fe-4S] clusters is significantly more nonplanar than in [2Fe-2S] clusters, with an average value for this torsion angle of only 162.0° (27Strop P. Takahara P.M. Chiu H.-J. Hayley C. Angove C. Burgess B.K. Rees D.C. Biochemistry. 2001; 40: 651-656Crossref PubMed Scopus (111) Google Scholar). The pronounced puckering of this unit in [4Fe-4S] clusters reflects primarily a more compressed Fe-S-Fe angle (71.7°) relative to [2Fe-2S] clusters (75.5°).Table IIAverage stereochemical parameters for Fd4 structures (from Table I)ParameterWTC55SC59SProtein standardsModel compoundBonds (Å) Fe-S2.232.242.232.23 ± 0.032.201 Fe-Fe2.732.692.692.73 ± 0.022.691 S-S3.523.593.553.51 ± 0.043.483Angles (°) S-Fe-S104.5106.2105.6104.3 ± 1.8104.6 Fe-S-Fe75.473.574.475.5 ± 1.075.4 9Sγ-Fe-22Sγ104.6105.1106.1105.1 ± 1.7111.2 55S/Oγ-Fe-59S/Oγ90.596.499.1 ‖Fe-S-S-Fe‖ torsion angle173171177175180Protein standards are the average values for four independent [2Fe-2S] clusters coordinated by four cysteine ligands in protein structures refined at resolutions ≤1.4 Å. These structures include PDB entries 1QT9, 1AWD, and the two distinct clusters in 1HLR(5Morales R. Charon M.H. Hudry-Clergeon G. Peátillot Y. Nørager S. Medina M. Frey M. Biochemistry. 1999; 38: 15764-15773Crossref PubMed Scopus (127) Google Scholar, 33Bes M.T. Parisini E. Inda L.A. Saraiva L.M. Peleato M.L. Sheldrick G.M. Structure. 1999; 7: 1201-1211Abstract Full Text Full Text PDF PubMed Scopus (51) Google Scholar, 38Rebelo J.M. Dias J.M. Huber R. Moura J.J.G. Romao M.J. J. Biol. Inorg. Chem. 2001; 6: 791-800Crossref PubMed Scopus (83) Google Scholar). The model compound parameters are derived from the structure of (Fe2S2(SC6H4CH3)4)2−(44Mayerle J.J. Denmark S.E. DePamphilis B.V. Ibers J.A. Holm R.H. J. Am. Chem. Soc. 1975; 97: 1032-1045Crossref Scopus (249) Google Scholar, 45Parisini E. Capozzi F. Lubini P. Lamzin V. Luchinat C. Sheldrick G. Acta Crystallogr. Sect. D Biol. Crystallogr. 1999; 55: 1773-1784Crossref PubMed Scopus (61) Google Scholar). Open table in a new tab Protein standards are the average values for four independent [2Fe-2S] clusters coordinated by four cysteine ligands in protein structures refined at resolutions ≤1.4 Å. These structures include PDB entries 1QT9, 1AWD, and the two distinct clusters in 1HLR(5Morales R. Charon M.H. Hudry-Clergeon G. Peátillot Y. Nørager S. Medina M. Frey M. Biochemistry. 1999; 38: 15764-15773Crossref PubMed Scopus (127) Google Scholar, 33Bes M.T. Parisini E. Inda L.A. Saraiva L.M. Peleato M.L. Sheldrick G.M. Structure. 1999; 7: 1201-1211Abstract Full Text Full Text PDF PubMed Scopus (51) Google Scholar, 38Rebelo J.M. Dias J.M. Huber R. Moura J.J.G. Romao M.J. J. Biol. Inorg. Chem. 2001; 6: 791-800Crossref PubMed Scopus (83) Google Scholar). The model compound parameters are derived from the structure of (Fe2S2(SC6H4CH3)4)2−(44Mayerle J.J. Denmark S.E. DePamphilis B.V. Ibers J.A. Holm R.H. J. Am. Chem. Soc. 1975; 97: 1032-1045Crossref Scopus (249) Google Scholar, 45Parisini E. Capozzi F. Lubini P. Lamzin V. Luchinat C. Sheldrick G. Acta Crystallogr. Sect. D Biol. Crystallogr. 1999; 55: 1773-1784Crossref PubMed Scopus (61) Google Scholar). A detailed examination of the geometrical parameters of the [2Fe-2S] cluster in WT Fd4 (Table III) reveals two pronounced outliers: (i) the relatively small Cys-55 Sγ-Fe2 Cys-59 Sγ bond angle and (ii) a relatively long Cys-55 Sγ-Fe2 bond. The Cys-55 Sγ-Fe2 Cys-59 Sγ angle averages 90.5° in the two crystallographically independent subunits of WT Fd4, a value that is significantly smaller than the average Cys-9 Sγ-Fe2-Cys-22 Sγ angle of 104.6° observed in the same structure and the consensus value of 105.1° observed in other well refined [2Fe-2S] protein structures (Table II). The more compressed Cys-55 Sγ-Fe2-Cys-59 Sγ angle has not been observed previously in other well refined ferredoxins with Cys ligands but is similar to the His Nδ-Fe-His Nδ bond angles of ∼94° observed in Rieske type [2Fe-2S] clusters (28Colbert C.L. Couture M.M.J. Eltis L.D. Bolin J.T. Structure. 2000; 8: 1267-1278Abstract Full Text Full Text PDF PubMed Scopus (102) Google Scholar, 29Iwata S. Saynovits M. Link T.A. Michel H. Structure. 1996; 4: 567-579Abstract Full Text Full Text PDF PubMed Scopus (281) Google Scholar). The other unusual feature of the Fd4 cluster geometry involves the Cys-55 Sγ-Fe2 bond, which is longer on average by ∼0.06 Å than the other three Fe-Sγ bonds (Table III). Constraints on the position of Cys-55 may contribute to this elongated bond because the residues around Cys-55 are well defined and appear to be relatively rigid, as reflected by the lower average temperature factors in this region (14.9 Å2 for residues 53–57 versus 21.4 Å2 for all protein atoms). An associated phenomenon may be the increased distortion of the peptide bond torsion angles (ω) of the residues surrounding Cys-55, particularly Thr-53, Met-56, and Ala-58, whose ω torsion angles are on average 166.0°, 171.8°, and 189.1°, respectively (2.4, 1.4, and 1.6 standard deviations from the ideal value of 180°). With an average Fe1-S1-S2-Fe2 torsion angle of –173°, the extent of nonplanarity of the [2Fe-2S] cluster in WT Fd4 is comparable with those observed in other well refined [2Fe-2S] protein structures.Table IIICluster geometry in molecules A and B for WT, C55S, and C59S Fd4ParameterResidueWTC55SC59SABABABBonds (Å) Fe1-Fe22.742.722.6922.6812.6972.690 S1-S23.523.523.5903.5923.5483.546 Fe1-S12.242.222.262(7)2.244(7)2.216(4)2.214(3) Fe1-S22.212.192.215(9)2.222(8)2.215(4)2.217(4) Fe2-S12.232.242.20(1)2.200(9)2.232(4)2.220(4) Fe2-S22.252.252.305(8)2.308(8)2.217(4)2.253(4) Sγ-Fe1C92.272.262.330(7)2.328(6)2.303(4)2.302(4) Sγ-Fe1C222.312.282.302(8)2.306(8)2.302(4)2.304(4) Sγ-Fe2C552.342.371.97(1)2.01(2)2.318(4)2.315(5) Sγ-Fe2C592.292.332.300(9)2.296(9)1.940(9)1.942(8)Angles (°) S1-Fe1-S2104.8106.0106.7(3)107.1(3)106.4(2)106.3(2) S1-Fe2-S2103.7103.3105.6(3)105.6(3)104.6(2)104.9(2) Fe1-S1-Fe275.675.174.274.274.774.7 Fe1-S2-Fe275.775.373.172.574.374.0 Sγ-Fe1-SγC9/C22105.0104.3105.7(3)104.6(2)106.1(1)106.1(1) O/Sγ-Fe2-O/SγC/S55-C/S5990.590.496.6(5)96.3(5)100.0(3)98.2(3) Sγ-Fe1-S1C9105.4106.4103.9(3)104.6(3)104.2(1)104.4(1) Sγ-Fe1-S2C9116.7116.2115.7(3)115.2(2)116.5(1)116.1(1) Sγ-Fe1-S1C22115.5113.7115.7(3)114.5(3)115.4(1)115.4(1) Sγ-Fe1-S2C22109.9110.6109.5(3)110.9(3)108.5(1)108.9(1) O/Sγ-Fe2-S1C/S55115.4115.7109.2(6)109.6(6)114.7(2)114.9(2) O/Sγ-Fe2-S2C/S55112.7113.5112.7(5)112.0(5)112.8(1)112.7(2) O/Sγ-Fe2-S1C/S59113.2112.2110.8(4)111.8(4)107.6(3)107.9(3) O/Sγ-Fe2-S2C/S59121.8122.3121.6(3)121.2(3)117.5(3)118.6(3)Torsion Angle (°) Fe1-S1-S2-Fe2−174.4−172.0−170.8−170.6−176.7−176.7Numbers in parentheses correspond to standard uncertainties in the last digit. Because of the lower resolution of the WT structure, coordinate uncertainties of the individual atoms were not calculated. Open table in a new tab Numbers in parentheses correspond to standard uncertainties in the last digit. Because of the lower resolution of the WT structure, coordinate uncertainties of the individual atoms were not calculated. Cys-55 also participates in an unusual interaction involving the probable formation of a Cα-H-Sγ hydrogen bond between the Cys-55 Cα-H and Cys-9 Sγ (Fig.2A). This interaction is identified on the basis of the Cys-55 Cα- Cys-9 Sγ and the Cys-55 Cα-H-Cys-9 Sγ distances of 3.6 and 2.7 Å, respectively, with a Cα-H-Sγ angle of 149°. For the purposes of this calculation, hydrogen positions were generated with the CCP4 program HGEN (19Bailey S. Acta Crystall

Ver no editor

Altmetric

PlumX

Entrar

Lembrar minha senha

Receber meu e-mail de confirmação

High Resolution Crystal Structures of the Wild Type and Cys-55 → Ser and Cys-59 → Ser Variants of the Thioredoxin-like [2Fe-2S] Ferredoxin from Aquifex aeolicus