Results from refinement of the crystal structures of P. aerogenes ferredoxin and C pasteurianum rubredoxin determined by x-ray diffraction show that there are 15-18 NHOS bonds in the former and six in the latter with lengths in the range 3.1-3.9 A. Earlier tritium exchange experiments are consistent with the presence of these hydrogen onds in the ferredoxin structure and show that more peptide hydrogen atoms are available for exchange in apoferredoxin than in intact ferredoxin. Four types of NH-S bonds are observed and two of these are geometrically similar to the two.types of 310 NH-O bonds. The existence of more NH-S bonds in ferredoxin than in high potential iron protein suggests why the -2 form of the Fe4S4 cluster is preferred in ferredoxin over the -1 form found in high potential iron protein. From comparison of Cys-X-Y-Cys sequences in rubredoxin, terredoxin, and high potential iron protein we suggest that two Cys-X-Y-Cys-Z sequences, where Z may have conformation angles similar to glycine, are required to make a oneiron cluster, no more than one Cys-X-Y-Cys-Z-Gly sequence is required to form a Fe2S2 ferredoxin, and a Cys-X-Y-Cys-Gly sequence where Y has a conformation such that the cysteines bond to different iron atoms is necessary to form the tetrameric cluster.The properties of iron-sulfur proteins raise an intriguing question: How does the protein structure contribute to the various observed forms of iron-sulfur centers? Iron-sulfur proteins include: (i) rubredoxin (Rb) with one Fe, no inorganic sulfur, and four cysteines; (ii) plant-type ferredoxins with two irons, two inorganic sulfur atoms, and at least four cysteines; and (iii) bacterial-type ferredoxin with one or more tetrameric clusters having four iron atoms, four inorganic sulfur atoms, and four cysteinyl sulfur atoms (1). The latter clusters are found both in Peptococcus aerogenes ferredoxin (Fd) (2) and in high potential iron protein (HiPIP) (3), a bacterial protein with unknown function which is reduced approximately +300 mV, in contrast to ferredoxin, which is reduced at the very low potential of about -400 mV. According to the three-state hypothesis (4), three oxidation states are accessible to the tetrameric cluster, one pair of which is preferred in HiPIP, another pair, of lower potential, in Fd. It has been demonstrated that the cluster in HiPIP can be "superreduced" in the presence of the denaturant dimethyl sulfoxide to the lowest state accessible to Fd (5), while the cluster in ferredoxin can be "superoxidized" in the presence of K3Fe(CN)6 (6). Studies with model compounds for tetrameric clusters have indicated that the -2/-3 reduction step analogous to reduction of ferredoxin operates at a much more negative potential than is observed for the proteins (7). Thus the protein portion of the iron-sulfur proteins has at least two functions that are important for cluster formation and activity: It provides an environment which favors one form of iron-sulfur center over another (i.e., 1 Fe, 2 Fe-2 S, 4 Fe-4 S) and, in the case o...

No abstract

No abstract

No abstract

The phases for the x‐ray data for crystals of azurin from P. aeruginosa have been refined and extended from 3 Å to 2.7 Å resolution by a method based on the direct space averaging of the electron density of the four molecules in the asymmetric unit. The electron density in the present map is much improved over the earlier one, and most of the side‐chains are now evident. In the detailed fitting of a model to the present map, an alternative interpretation, conforming to the plastocyanin model and reversing the chain directions in strands one and two of the original azurin model, was found to better represent certain features of the present map. The surface of the molecule closest to the copper consists of an extended region of invariant or semiconserved hydrophobic residues. There is not a region of conserved charge that is not paired with a compensating charge.

No abstract

The crystal structure of the oxidized form of flavodoxin from Desulfovibrio vulgaris has been studied at 2.0-A resolution, and a detailed description of the region around the flavin mononucleotide binding site is now available. The flavin is between a tyrosine group, roughly parallel to it on one side, and a tryptophan, about 450 from being parallel, on the other side. The two carbonyl groups and two nitrogen atoms of the flavin are hydrogen bonded to the peptide chain of the protein, while the two methyl groups are exposed at the surface of the protein.The phosphate group of the flavin mononucleotide is inside the protein and extensively hydrogen bonded to it. The ribityl group is hydrogen bonded both to the protein and to water on the surface of the protein.Flavin-protein interactions have been studied extensively in recent years. Through kinetic, thermodynamic, various spectral techniques, and other physico-chemical methods, attempts have been made to understand flavin binding and electron transport in flavoproteins. Any studies by these techniques would be greatly aided by knowledge of the threedimensional structure of a flavoprotein and, in particular, the binding and environment of FMN (riboflavin-5'-phosphate) in it. The crystal structure of the oxidized form of flavodoxin from the sulfate-reducing bacterium Desulfovibrio vulgaris (strain Hildenborough, NCIB 8303) has now been extended to 2.0-A resolution. The improvement of the electron density map over the 2.5-resolution map enables us to make a much more certain and detailed interpretation of the FMN environment than that already reported (1).A similar investigation on the flavodoxin from Clostridium MP is underway by Martha Ludwig and coworkers at the University of Michigan (2). ExperimentalCrystallization and the method of data collection are the same as previously described (1). Flavodoxin from D. vulgaris crystallizes in space group P432,2 with unit cell dimensions a = b = 51.6 A, c = 139.6 A, based on Xcus,, = 1.5418 A and unit cell volume of 372,000 A3. There is one molecule of the protein of approximate molecular weight 16,000 per asymmetric unit. Ni-filtered CuKa radiation was used with a takeoff angle of 3.5°. A five-step (0.080 per step) co/20 scan across the top of each peak was fit by least-squares with a Gaussian curve and used to obtain the integrated intensity. Over 13,000 reflections (6669 Friedel pairs on the native crystal) were collected on both the native crystal and a Sm+3 derivative in the region 0.2 < sin O/X < 0.25 (2.0 A < d < 2.6 A). If The phases for the new data were determined by use of a single Sm+s derivative. The Sm+a site was the same as that found in the 2.5A determination and the phase ambiguity of the single derivative was resolved by use of anomalous scattering from Sm+a. The additional data were scaled and merged with the previously phased data to 2.5-A d spacing, giving a complete data set from co to 2.0A d spacing, nearly doubling the number of terms used in calculating the electron density map. The qua...

Refinement of triclinic lysozyme by restrained least squares against the 2 A resolution X-ray data is described, beginning with the model from cycle 17 of the preceding paper [Hodsdon, Brown, Sieker & Jensen (1990). Acta Cryst. B46, 54-62]. After 20 refinement cycles, R stood at 0.172. Nevertheless, serious errors involving both main-chain and side-chain atoms still remained, requiring numerous model rebuilding sessions interleaved with refinement cycles. After 63 cycles R = 0.124 for the model which includes all protein atoms, 249 water oxygen sites and five nitrate ions. Although the overall B is relatively low, 10.5 A2, B's for atoms in the region of residues 101-103, toward the termini of some of the longer side chains, and in the region of the C terminus of the main chain exceed 20 A2, indicating relatively high atomic mobilities, disorder, or remaining errors in the model.