Voltar
Artigo Acesso aberto Revisado por pares
BIBTEX RIS

A Structural Basis of Equisetum arvense Ferredoxin Isoform II Producing an Alternative Electron Transfer with Ferredoxin-NADP+ Reductase

2004; Elsevier BV; Volume: 280; Issue: 3 Linguagem: Inglês

10.1074/jbc.m408904200

ISSN

1083-351X

Autores

Genji Kurisu, Daisuke Nishiyama, Masami Kusunoki, Shinobu Fujikawa, Midori Katoh, Guy T. Hanke, Toshiharu Hase, Keizo Teshima,

Tópico(s)

Metalloenzymes and iron-sulfur proteins

Resumo

We have determined the crystal structure, at 1.2-Å resolution, of Equisetum arvense ferredoxin isoform II (FdII), which lacks residues equivalent to Arg39 and Glu28 highly conserved among other ferredoxins (Fds). In other Fds these residues form an intramolecular salt bridge crucial for stabilization of the [2Fe-2S] cluster, which is disrupted upon complex formation with Fd-NADP+ oxidoreductase (FNR) to form two intermolecular salt bridges. The overall structure of FdII resembles the known backbone structures of E. arvense isoform I (FdI) and other plant-type Fds. Dramatically, in the FdII structure a unique, alternative salt bridge is formed between Arg22 and Glu58. This results in a different relative orientation of the α-helix formed by Leu23-Glu29 and eliminates the possibility of forming three of the five intermolecular salt bridges identified on formation of a complex between maize FdI and maize FNR. Mutation of FdII, informed by structural differences with FdI, showed that the alternative salt bridge and the absence of an otherwise conserved Tyr residue are important for the alternative stabilization of the FdII [2Fe-2S] cluster. We also investigated FdI and FdII electron transfer to FNR on chloroplast thylakoid membranes. The Km and Vmax values of FdII are similar to those of FdI, contrary to previous measurements of the reverse reaction, from FNR to Fd. The affinity between reduced FdI and oxidized FNR is much greater than that between oxidized FdI and reduced FNR, whereas this is not the case with FdII. The pH dependence of electron transfer by FdI, FdII, and an FdII mutant with FdI features was measured and further indicated that the binding mode to FNR differs between FdI and FdII. Based on this evidence, we hypothesize that binding modes with other Fd-dependent reductases may also vary between FdI and FdII. The structural differences between FdI and FdII therefore result in functional differences that may influence partitioning of electrons into different redox metabolic pathways. We have determined the crystal structure, at 1.2-Å resolution, of Equisetum arvense ferredoxin isoform II (FdII), which lacks residues equivalent to Arg39 and Glu28 highly conserved among other ferredoxins (Fds). In other Fds these residues form an intramolecular salt bridge crucial for stabilization of the [2Fe-2S] cluster, which is disrupted upon complex formation with Fd-NADP+ oxidoreductase (FNR) to form two intermolecular salt bridges. The overall structure of FdII resembles the known backbone structures of E. arvense isoform I (FdI) and other plant-type Fds. Dramatically, in the FdII structure a unique, alternative salt bridge is formed between Arg22 and Glu58. This results in a different relative orientation of the α-helix formed by Leu23-Glu29 and eliminates the possibility of forming three of the five intermolecular salt bridges identified on formation of a complex between maize FdI and maize FNR. Mutation of FdII, informed by structural differences with FdI, showed that the alternative salt bridge and the absence of an otherwise conserved Tyr residue are important for the alternative stabilization of the FdII [2Fe-2S] cluster. We also investigated FdI and FdII electron transfer to FNR on chloroplast thylakoid membranes. The Km and Vmax values of FdII are similar to those of FdI, contrary to previous measurements of the reverse reaction, from FNR to Fd. The affinity between reduced FdI and oxidized FNR is much greater than that between oxidized FdI and reduced FNR, whereas this is not the case with FdII. The pH dependence of electron transfer by FdI, FdII, and an FdII mutant with FdI features was measured and further indicated that the binding mode to FNR differs between FdI and FdII. Based on this evidence, we hypothesize that binding modes with other Fd-dependent reductases may also vary between FdI and FdII. The structural differences between FdI and FdII therefore result in functional differences that may influence partitioning of electrons into different redox metabolic pathways. [2Fe-2S] ferredoxin (Fd) 1The abbreviations used are: Fd, ferredoxin; FNR, ferredoxin-NADP+ oxidoreductase; FTR, ferredoxin-thioredoxin reductase; NiR, nitrite reductase; SiR, sulfite reductase. electron transfer proteins distribute reducing equivalents derived from light energy to Fd-dependent reductases, such as Fd-NADP+ reductase (FNR), nitrite reductase (NiR), sulfite reductase (SiR), and Fd-thioredoxin reductase (FTR), for the assimilation of inorganic compounds and the regulation of carbon assimilation (1Knaff D.B. Ort D.R. Yocum C.F. Photosynthesis: The Light Reactions. Kluwer Academic Publishers Group, Dordrecht, Netherlands1996: 333-361Google Scholar, 2Blankenship R.E. Nat. Struct. Biol. 2001; 8: 94-95Crossref PubMed Scopus (15) Google Scholar). Recently, Fds have also been found to donate electrons to desaturases for biosynthesis of unsaturated fatty acids (3Schultz D.J. Suh M.C. Ohlrogge J.B. Plant Physiol. 2000; 124: 681-692Crossref PubMed Scopus (52) Google Scholar), choline monooxygenase for betaine biosynthesis (4Rathinasabapathi B. Burnet M. Russell B.L. Gage D.A. Liao P-C. Nye G.N. Scott P. Golbeck J.H. Hanson A.D. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 3454-3458Crossref PubMed Scopus (221) Google Scholar), and heme oxygenase for phytochrome biosynthesis (5Muramoto T. Tsurui N. Terry M.J. Kohchi T. Plant Physiol. 1995; 130: 1958-1966Crossref Scopus (146) Google Scholar), indicative of the contribution of Fds to various metabolic pathways in addition to assimilation of inorganic substances. Most intriguingly, higher plant species possess several isoforms of Fd, implying isoform specificity to different redox-metabolic pathways. For example, maize Fd isoforms, FdI and FdII, are distributed differentially in mesophyll and bundle-sheath cells and are found to be mainly involved in the formation of NADPH and ATP, respectively (6Kimata Y. Hase T. Plant Physiol. 1989; 89: 1193-1197Crossref PubMed Google Scholar). The functional difference between maize FdI and FdII is dependent solely on the replacement of Asp65 with Asn, which affects the FNR-binding site (7Matsumura T. Kimata-Ariga Y. Sakakibara H. Sugiyama T. Murata H. Takao T. Shimonishi Y. Hase T. Plant Physiol. 1999; 119: 481-488Crossref PubMed Scopus (57) Google Scholar). This variation results in distinct phenotypes of cyanobacteria (lacking an endogenous Fd gene) in which FdI and FdII are expressed (8Kimata-Ariga Y. Matsumura T. Kada S. Fujimoto H. Fujita Y. Endo T. Mano J. Sato F. Hase T. EMBO J. 2000; 19: 5041-5050Crossref PubMed Scopus (46) Google Scholar). Additionally, biochemical experiments indicate partial differences between the Fd-binding sites of the Fd-dependent reductases FNR, SiR, and FTR (9De Pascalis A.R. Schürmann P. Bosshard H.R. FEBS Lett. 1994; 337: 217-220Crossref PubMed Scopus (30) Google Scholar, 10Akashi T. Matsumura T. Ideguchi T. Iwakiri K. Kawakatsu T. Taniguchi I. Hase T. J. Biol. Chem. 1999; 274: 29399-29405Abstract Full Text Full Text PDF PubMed Scopus (71) Google Scholar). We therefore suspect that Fd isoforms might play an important role in the modulation of various redox-metabolic pathways in plant cells and that such modulation might be ascribed to specific noncovalent interactions between Fd isoforms and Fd-dependent partner proteins. Recently, we determined the three-dimensional structure of the maize mesophyll Fd (FdI)·FNR complex at 2.59-Å resolution (11Kurisu G. Kusunoki M. Katoh E. Yamazaki T. Teshima K. Onda Y. Kimata-Ariga Y. Hase T. Nat. Struct. Biol. 2001; 8: 117-121Crossref PubMed Scopus (285) Google Scholar), in which five intermolecular salt bridges are formed between Fd and FNR. This structure shows an intramolecular Fd salt bridge between Arg40-Glu29 is dynamically exchanged to form two intermolecular salt bridges with FNR residues in the process of the complex formation, suggesting the Arg-Glu pair has a role in electron transfer. Furthermore, this Arg-Glu pair is conserved among all but three of more than 70 plant-type Fd sequences (12Matsubara H. Saeki K. Adv. Inorg. Chem. 1992; 38: 223-280Crossref Scopus (173) Google Scholar). In Equisetum arvense the Arg-Glu pair is conserved in Fd isoform I (FdI) but missing from isoform II (FdII), and a comparison of electron transfer with FNR under excess NADPH shows obvious differences between the Km and kcat values of FdI and FdII. Electron transfer rates of FdII mutants, in which the Arg-Glu pair is introduced, indicate that these residues are crucial to these differences (13Teshima K. Fujita S. Hirose S. Nishiyama D. Kurisu G. Kusunoki M. Kimata-Ariga Y. Hase T. FEBS Lett. 2003; 546: 189-194Crossref PubMed Scopus (8) Google Scholar). These findings suggest that E. arvense FdII might vary from FdI in its electron transfer activity around photosystem I, because of its different interaction with FNR. When the Arg-Glu pair of E. arvense FdI is substituted with the corresponding noncharged residues of FdII (R39Q/E28S), the [2Fe-2S] cluster became unstable (13Teshima K. Fujita S. Hirose S. Nishiyama D. Kurisu G. Kusunoki M. Kimata-Ariga Y. Hase T. FEBS Lett. 2003; 546: 189-194Crossref PubMed Scopus (8) Google Scholar), indicating a crucial role for these residues in stabilizing Fd in the native state. However, it should be noted that the [2Fe-2S] cluster of wild type FdII is stable even in the absence of a salt bridge between Arg39 and Glu28. To understand the structural basis of the unique properties of E. arvense FdII and to elucidate the mechanism that stabilizes the [2Fe-2S] cluster, we have therefore determined its crystal structure at 1.2 Å resolution and compared this with the previously determined structure of E. arvense FdI (1.8 Å resolution) (14Ikemizu S. Bando M. Sato T. Morimoto Y. Tsukihara T. Fukuyama K. Acta Crystallogr. Sect. D Biol. Crystallogr. 1994; 50: 167-174Crossref PubMed Scopus (49) Google Scholar). We report that FdII has another different internal salt bridge between Arg22-Glu58 instead of that between Arg39-Glu28 and that the [2Fe-2S] cluster is stabilized by this alternative salt bridge, and by the deletion of one amino acid residue in a short loop connecting the cluster binding loop to the preceding α-helix region. Furthermore, we investigated the electron transfer capabilities of E. arvense FdI and FdII from photosystem I to FNR by using chloroplast thylakoid membranes. We will discuss a correlation between the structural differences of FdI and FdII and their electron transfer abilities with FNR. Preparation of Recombinant Fds—E. arvense recombinant FdII, FdI, and the FdII mutants were expressed and accumulated as the apo-form in Escherichia coli cells and successfully converted to the holo-form by chemical reconstitution of the [2Fe-2S] cluster as described previously (13Teshima K. Fujita S. Hirose S. Nishiyama D. Kurisu G. Kusunoki M. Kimata-Ariga Y. Hase T. FEBS Lett. 2003; 546: 189-194Crossref PubMed Scopus (8) Google Scholar). Further purification of the Fds was carried out as described previously (15Hase T. Mizutani S. Mukohata Y. Plant Physiol. 1991; 97: 1395-1401Crossref PubMed Scopus (43) Google Scholar, 16Matsumura T. Sakakibara H. Nakano R. Kimata Y. Sugiyama T. Hase T. Plant Physiol. 1997; 114: 653-660Crossref PubMed Scopus (54) Google Scholar). Site-specific and insertion mutants of FdII R22T, FdII with Tyr32 newly inserted (FdII(Y32)) and FdII(Y32)Q38R/S28E were prepared with the Quikchange site-directed mutagenesis kit (Stratagene) (13Teshima K. Fujita S. Hirose S. Nishiyama D. Kurisu G. Kusunoki M. Kimata-Ariga Y. Hase T. FEBS Lett. 2003; 546: 189-194Crossref PubMed Scopus (8) Google Scholar), and these mutation sites were confirmed by DNA sequencing. The concentration of Fd was determined spectrophotometrically based on a molar extinction coefficient of 9.68 mm-1 cm-1 at 422 nm (17Tagawa K. Arnon D.I. Biochim. Biophys Acta. 1968; 153: 602-613Crossref PubMed Scopus (204) Google Scholar). Crystallization—We searched extensively for crystallization conditions of E. arvense FdII by the hanging drop vapor diffusion method with Crystal Screen I and II (Hampton Research) and Wizard I, II, and cryo (Emerald Biostructures). Sodium phosphate and ammonium sulfate were also checked as precipitants. Needle-shape crystals of FdII were initially obtained at 4 °C from equal volumes of the protein (10 mg/ml) and reservoir solution (3.1 m NaH2PO4/K2HPO4 (pH 7.5)), but were too small for x-ray experiments. Flat plate-shaped crystals of appropriate size were successfully obtained with the same reservoir solution additionally containing benzamidine hydrochloride (2.0% (w/v)). The improved crystals were transferred to a reservoir solution containing sucrose (20% (w/v)) and immediately frozen with liquid nitrogen for x-ray experiments. Crystallographic Data Collection and Processing—X-ray data were collected at a wavelength of 0.9 Å at beamline BL44XU at the SPring-8 synchrotron in Hyogo, Japan. Diffraction images were collected at liquid nitrogen temperature (100 K) on a CCD-based PX210 detector system. Images were processed with the program DPS/MOSFLM (18Rossman M.G. Beeek C.G. Acta Crystallogr. Sect. D Biol. Crystallogr. 1999; 55: 1631-1640Crossref PubMed Scopus (177) Google Scholar) and SCALA in the CCP4 program package (19Collaborative Computational Project, Number 4Acta Crystallogr. Sect. D Biol. Crystallogr. 1994; 50: 760-763Crossref PubMed Scopus (19825) Google Scholar). Crystal data and crystallographic statistics are given in Table I.Table ICrystal data, data collection, and refinement statisticsCrystal dataSpace groupC2Unit cell dimensions a (Å)97.54 b (Å)29.41 c (Å)32.56 β (°)103.94ZaThe number of protein molecules in the unit cell4Vm (Å3/Da)bThe volume occupied by unit weight of protein in the unit cell2.1Vprot. (%)cThe ratio of protein volume in the unit cell59.7Data collection statistics Resolution (Å)16.5 to 1.2 (1.24 to 1.20) No. measured reflections97,597 No. unique reflections27,366 Completeness (%)dThe values in parentheses refer to the highest resolution shell97.2 (97.2) Rmeas (%)dThe values in parentheses refer to the highest resolution shell,eRmeas = Σh (nh/(nh – 1))1/2Σi|Ih – Ihi|/ΣhΣiIhi, where nh, Ih, and Ihi are the multiplicity, intensity, and ith intensity measurement of reflection h, respectively. Values in parentheses are the resolution range in Å for the outermost shell7.7 (12.3) I/σ (I)5.3Refinement statistics Resolution limits (Å)16.5 to 1.2 Rcryst (%)fRfactor = Σ||Fo| – |Fc||/Σ|Fo|11.75 Rfree (%)fRfactor = Σ||Fo| – |Fc||/Σ|Fo|15.45No. non-hydrogen atoms Protein696 Benzamidine54 Iron2 Sulfur2 Water208r.m.s.gr.m.s., root mean square deviations bond length (Å)0.015 Bond angle (Å)0.033Ramachandran plot Most favored (%)89.6 Allowed (%)10.4a The number of protein molecules in the unit cellb The volume occupied by unit weight of protein in the unit cellc The ratio of protein volume in the unit celld The values in parentheses refer to the highest resolution shelle Rmeas = Σh (nh/(nh – 1))1/2Σi|Ih – Ihi|/ΣhΣiIhi, where nh, Ih, and Ihi are the multiplicity, intensity, and ith intensity measurement of reflection h, respectively. Values in parentheses are the resolution range in Å for the outermost shellf Rfactor = Σ||Fo| – |Fc||/Σ|Fo|g r.m.s., root mean square Open table in a new tab Structural Determination and Refinement—The structure was solved by molecular replacement method using E. arvense FdI structure (14Ikemizu S. Bando M. Sato T. Morimoto Y. Tsukihara T. Fukuyama K. Acta Crystallogr. Sect. D Biol. Crystallogr. 1994; 50: 167-174Crossref PubMed Scopus (49) Google Scholar) as a search model with the program CNS (20Brünger 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 (16991) Google Scholar). First, the structure of FdII was refined with isotropic temperature factors using the program CNS. After convergence with CNS, we next used the program SHELXL in the SHELX97 program package to model atomic anisotropy. Manual revision of the atomic model was carried out with the program O (21Jones T.A. Zou J.Y. Cowan S.W. Kjeldgaard M. Acta Crystallogr. Sect. A. 1991; 47: 110-119Crossref PubMed Scopus (13016) Google Scholar). On careful inspection of the |Fo| - |Fc| and 2|Fo| - |Fc| maps, we added benzamidine and water molecules into the model and modeled dual conformers of amino acid side chains. Subsequent refinement was carried out up to 1.2 Å resolution without any structural restraints on the [2Fe-2S] cluster. The stereochemical geometry of the model was checked with the program PROCHECK (22Vaguine A.A. Richelle J. Wodak S.J. Acta Crystallogr. Sect. D Biol. Crystallogr. 1999; 55: 191-205Crossref PubMed Scopus (862) Google Scholar). Stability Assays—The degradation rates of the [2Fe-2S] clusters of FdII and FdII mutants were measured by monitoring the decrease of A420 due to the presence of the [2Fe-2S] cluster for 36 h at pH 7.5, 100 mm NaCl, and 4 °C. The initial values of A420 were set at 0.5-1.2 for all the samples. The stabilities of their [2Fe-2S] clusters were evaluated by their half-lives estimated on the assumption that degradation of the [2Fe-2S] cluster obeys first-order kinetics. Enzyme Assays—NADP+ photoreduction ability of Fd by purified thylakoid membranes from spinach chloroplast was assayed as described previously (23Hanke G.T. Kimata-Ariga Y. Taniguchi I. Hase T. Plant Physiol. 2004; 134: 255-264Crossref PubMed Scopus (119) Google Scholar). The reaction mixture contained, in a total volume of 1.0 ml, 0.2 mm NADP+, thylakoid membranes (about 0.01 μg of chlorophyll), 1-10 μm Fd, 50 mm Hepes (pH 6.4-8.1) and 100 mm NaCl. The reaction was measured in a spectrometer and was initiated by irradiating with red light perpendicular to the spectrophotometric light beam. The increase of A340 due to the photoreduction of NADP+ was monitored. The pH dependences of the abilities of FdI, FdII, and FdII mutant between pH 6.4 and 7.7 were analyzed using the equation, v = V0 + ΣΔVi × Ka,i/((H+) + Ka,i), where V0 is the ability when all the ionizable groups participating are protonated, and Ka,i is the dissociation constant of the corresponding ionizable group participating, and ΔVi is the change in the ability when the ionizable group with a pKa value of pKa,i is deprotonated. The analysis was not corrected with the effect of a change of the net charge, because only a few ionizable groups are dissociated in the neutral pH region. Quality of E. arvense FdII Structure—Crystals of FdII belong to space group C2, with unit cell parameters a = 97.54 Å, b = 29.41 Å, c = 32.56 Å, and β = 103.94° (Table I). The refined model of FdII contains 1 FdII molecule, 6 benzamidines, and 208 water molecules in the asymmetric unit. A total of nine residues in the final model have dual conforming side chains. The final crystallographic R factor and free R factor with anisotropic temperature factors are 11.75 and 15.45%, respectively, for 27,366 unique reflections in the resolution range of 16-1.2 Å. All amino acid residues of FdII except glycine and proline residues are in the stereochemically most favored regions in the Ramachandran plot. Refinement statistics are summarized in Table I. Overall Structure—The overall structures of E. arvense FdII and FdI (14Ikemizu S. Bando M. Sato T. Morimoto Y. Tsukihara T. Fukuyama K. Acta Crystallogr. Sect. D Biol. Crystallogr. 1994; 50: 167-174Crossref PubMed Scopus (49) Google Scholar) are shown as ribbon drawings in Fig. 1, A and B, respectively. A superimposition of the main chain structures of FdII and FdI is shown in Fig. 1C. The structures are superimposed along the C-α atoms of residues 1-93 and the [2Fe-2S] cluster in FdII, and C-α atoms of residues 1-31 and 33-94 and the [2Fe-2S] cluster in FdI, because FdII lacks the 32nd and the C-terminal 95th residues in comparison to FdI, as shown in the sequence alignment (Fig. 2). The root mean square deviation value for the C-α atom between the two structures is 1.17 Å. The overall structure of FdII was fairly similar to that of FdI and also resembled the other known backbone structures of plant-type Fds (25Binda C. Coda A. Aliverti A. Zanetti G. Mattevi A. Acta Crystallogr. Sect. D Biol. Crystallogr. 1998; 54: 1353-1358Crossref PubMed Google Scholar, 29Fukuyama K. Hase T. Matsumoto S. Tsukihara T. Katsube Y. Tanaka N. Kakudo M. Wada K. Matsubara H. Nature. 1980; 286: 522-524Crossref Scopus (144) Google Scholar, 30Rypniewdki W.R. Breiter D.R. Benning M.M. Wesenberg G. Oh B.-H. Markley J.L. Rayment I. Holden H.M. Biochemistry. 1991; 30: 4126-4131Crossref PubMed Scopus (173) Google Scholar, 31Bes M.T. Parisini E. India L.A. Saraiva L.M. Peleato M.L. Sheldrick G.M. Structure (Lond.). 1999; 7: 1201-1211Abstract Full Text Full Text PDF PubMed Scopus (51) Google Scholar).Fig. 2Sequence alignment of E. arvense FdII, FdI (26Hase T. Wada K. Matsubara H. J. Biochem. (Tokyo). 1977; 82: 277-286Crossref PubMed Scopus (34) Google Scholar), maize FdI (27Hase T. Kimata Y. Yonekura K. Matsumura T. Sakakibara H. Plant Physiol. 1991; 96: 77-83Crossref PubMed Scopus (71) Google Scholar), and spinach FdI (28Matsubara H. Sasaki R.M. J. Biol. Chem. 1968; 243: 1732-1757Abstract Full Text PDF PubMed Google Scholar). Residues different between E. arvense FdII and FdI and conserved among the four Fds are colored red and blue, respectively. Boxed residues represent those that are observed to form an intramolecular salt bridge in either E. arvense FdI or FdII. Residues with filled circles correspond to the residues that are involved in intermolecular salt bridges between maize FdI and FNR (11Kurisu G. Kusunoki M. Katoh E. Yamazaki T. Teshima K. Onda Y. Kimata-Ariga Y. Hase T. Nat. Struct. Biol. 2001; 8: 117-121Crossref PubMed Scopus (285) Google Scholar).View Large Image Figure ViewerDownload Hi-res image Download (PPT) However, it should be noted that the backbone conformation connecting the α-helix of Leu23-Glu29 and cluster binding loop of Pro34-Leu46 in FdII is significantly different from the corresponding conformation of FdI and also that the orientation of α-helix of Leu23-Glu29 for FdII is different from that of the corresponding α-helix for FdI by an angle of 27.6°. These differences may be ascribed to the deletion of Tyr32 in FdII (Fig. 2). In the region of FdII corresponding to the salt bridge formed between Arg39 and Glu28 in FdI (Fig. 1B), the corresponding Gln38 and Ser28 side chains do not even form a hydrogen bond. Instead, Arg22 and Glu58 form a unique, alternative salt bridge in FdII (Fig. 1A), which has no counterpart in the corresponding Thr22 and Glu59 of FdI. On refinement of this region, we found that both Arg22 and Glu58 have two rotational conformations. In the major side chain conformation, the major electron densities were interpreted as forming a salt bridge, whereas the minor conformations were exposed to the bulk solvent region containing 1.6 m phosphate as a precipitant. Therefore, we consider that these rotational conformations corresponding to minor electron densities were generated because of the very high ionic strength of the surrounding environment and that under the usual physiological conditions the Arg22 and Glu58 form an intramolecular salt bridge. Structure of the [2Fe-2S] Cluster and Its Surroundings— Inter-atomic distances and bond angles for the [2Fe-2S] cluster in FdII are summarized in Tables II and III, respectively, and also compared with the data for E. arvense FdI (14Ikemizu S. Bando M. Sato T. Morimoto Y. Tsukihara T. Fukuyama K. Acta Crystallogr. Sect. D Biol. Crystallogr. 1994; 50: 167-174Crossref PubMed Scopus (49) Google Scholar). These values are quite similar between FdII and FdI.Table IIInter-atomic distances and their average values for the [2Fe-2S] clusters in FdII and FdI(14Ikemizu S. Bando M. Sato T. Morimoto Y. Tsukihara T. Fukuyama K. Acta Crystallogr. Sect. D Biol. Crystallogr. 1994; 50: 167-174Crossref PubMed Scopus (49) Google Scholar) Residue numbers and atom names for FdII are given on the left column and those for FdI in parentheses.Atom pairDistances (Å) FdIIFdIaThe data are cited from Ref. 14Molecule AMolecule B37Sγ-Fe1 (38Sγ-Fe1)2.332.232.3142Sγ-Fe1 (43Sγ-Fe1)2.302.222.2445Sγ-Fe2 (46Sγ-Fe2)2.312.262.2575Sγ-Fe2 (76Sγ-Fe2)2.312.232.24Mean distance for Sγ-Fe2.312.242.26Fe1-S12.292.172.11Fe1-S22.272.332.24Fe2-S12.252.292.27Fe2-S22.192.252.30Mean distance for Fe-S2.252.262.23Fe1-Fe22.772.812.75S1-S23.543.543.5137Sγ-37Cβ (38Sγ-38Cβ)1.821.881.8442Sγ-42Cβ (43Sγ-43Cβ)1.821.821.8645Sγ-45Cβ (46Sγ-46Cβ)1.831.861.8475Sγ-75Cβ (76Sγ-76Cβ)1.811.781.85Mean distance for Sγ-Cβ1.821.841.85a The data are cited from Ref. 14Ikemizu S. Bando M. Sato T. Morimoto Y. Tsukihara T. Fukuyama K. Acta Crystallogr. Sect. D Biol. Crystallogr. 1994; 50: 167-174Crossref PubMed Scopus (49) Google Scholar Open table in a new tab Table IIIBond angles and their average values for the [2Fe-2S] clusters in FdII and FdI(14Ikemizu S. Bando M. Sato T. Morimoto Y. Tsukihara T. Fukuyama K. Acta Crystallogr. Sect. D Biol. Crystallogr. 1994; 50: 167-174Crossref PubMed Scopus (49) Google Scholar) Residue numbers and atom names for FdII are listed on the left column and those for FdI in parentheses.Atoms defining anglesAngles (°)FdIIFdIaThe data are cited from Ref. 14Molecule AMolecule BS1-Fe1-S1101.9103.7107.2S1-Fe2-S2105.7102.3100.2Mean S-Fe-S103.8103.0103.737Sγ-Fe1-S1 (38Sγ-Fe1-S1)116.7121.3118.237Sγ-Fe1-S2 (38Sγ-Fe1-S2)102.399.3101.042Sγ-Fe1-S1 (43Sγ-Fe1-S1)110.5108.7109.542Sγ-Fe1-S2 (43Sγ-Fe1-S2)117.7114.9115.145Sγ-Fe2-S1 (46Sγ-Fe2-S1)114.3114.8117.845Sγ-Fe2-S2 (46Sγ-Fe2-S2)109.3106.6104.075Sγ-Fe2-S1 (76Sγ-Fe2-S1)115.1113.3114.075Sγ-Fe2-S2 (76Sγ-Fe2-S2)106.6110.2107.9Mean Sγ-Fe-S111.6111.2110.9Fe1-S1-Fe275.278.077.6Fe1-S2-Fe276.975.776.7Mean Fe-S-Fe76.176.977.2Fe1–37Sγ-37Cβ (Fe1–38Sγ-38Cβ)115.6120.6120.9Fe1–42Sγ-42Cβ (Fe1–43Sγ-43Cβ)113.1114.1123.5Fe2–45Sγ-45Cβ (Fe2–46Sγ-46Cβ)101.6107.0107.8Fe2–75Sγ-75Cβ (Fe2–76Sγ-76Cβ)106.8105.7101.8Mean Fe-Sγ-Cβ109.3111.9113.537Sγ-Fe1–42Sγ (38Sγ-Fe1–43Sγ)107.8108.9106.045Sγ-Fe2–75Sγ (46Sγ-Fe2–76Sγ)105.5109.3111.3Mean Sγ-Fe-Sγ106.7109.1108.7a The data are cited from Ref. 14Ikemizu S. Bando M. Sato T. Morimoto Y. Tsukihara T. Fukuyama K. Acta Crystallogr. Sect. D Biol. Crystallogr. 1994; 50: 167-174Crossref PubMed Scopus (49) Google Scholar Open table in a new tab Fe and S atoms of the cluster and C-α atoms of the cluster binding loop (Pro34-Leu46 for FdII and Pro35-Leu47 for FdI) of FdII and FdI are superimposed with a root mean square deviation value of 0.21 Å. The general arrangement of the [2Fe-2S] cluster and its surroundings show high structural homology between the FdI and FdII. Apart from Leu35, Gln38, and Thr44; the local side chain conformation of the cluster binding loop in FdII is almost identical to that of the corresponding side chains in FdI. The hydrogen bonding networks around the [2Fe-2S] clusters of FdII and FdI are shown in Fig. 3. In FdII the O-γ atom of Thr44 is hydrogen-bonded to the S-γ atom of Cys42, and an equivalent bond also occurs in FdI between the O-γ atom of Ser45 and the S-γ atom of Cys43. By contrast, FdII has no counterpart to the hydrogen bond between the O-γ atom of Ser37 and the S-γ atom of Cys43 in FdI, because both dual conformations of the side chain of Ser36 for FdII face the bulk solvent region. Therefore, the number of hydrogen bonds between the hydroxyl group of Ser (or Thr) or the backbone amide group and the S atom of the [2Fe-2S] cluster or the sulfhydryl group of the Cys coordinating the cluster is 1 less for FdII than for FdI. In general, it is known that hydrogen bonds between backbone amide groups and cluster S atoms or coordinating sulfhydryl groups decrease the redox potential (32Sheridan R.P. Allen L.C. Carter Jr., C.W. J. Biol. Chem. 1981; 256: 5052-5057Abstract Full Text PDF PubMed Google Scholar, 33Backes G. Mino Y. Loehr T.M. Meyer T.E. Cusanovich M.A. Sweeney W.V. Adman E.T. Sanders-Loehr J. J. Am. Chem. Soc. 1991; 113: 2055-2064Crossref Scopus (253) Google Scholar). We also determined the redox potentials of these Fds by cyclic voltammetry demonstrating that the potential (-403.8 ± 0.4 mV) of FdII is slightly lower than that (-396.6 ± 1.3 mV) of FdI. As exposure of the cluster to the bulk solvent and the dielectric constant around the [2Fe-2S] cluster are also crucial factors in determining the redox potential, we are currently refining E. arvense FdI structure at higher resolution (∼1.2 Å higher than the known 1.8 Å) to compare more precisely the microenvironment around the [2Fe-2S] cluster at atomic resolution between FdII and FdI. We will report in detail on the relationship between the redox potential and the microenvironment around the cluster, together with the redox potential measurements, elsewhere. [2Fe-2S] Cluster Stability of FdII and Its Mutants—We reported previously that an E. arvense FdI mutant, substituted at Arg39 and Glu28 with the corresponding noncharged residues of FdII, was unstable and unfolded, releasing its [2Fe-2S] cluster within 15 h (13Teshima K. Fujita S. Hirose S. Nishiyama D. Kurisu G. Kusunoki M. Kimata-Ariga Y. Hase T. FEBS Lett. 2003; 546: 189-194Crossref PubMed Scopus (8) Google Scholar), whereas the [2Fe-2S] cluster of FdII was stable despite the absence of these salt bridge-forming residues. The present x-ray analysis data show that the conformation of the [2Fe-2S] cluster and

Referência(s)