To understand how the photoreceptor protein rhodopsin performs in its role as a receptor, its structure needs to be determined at the atomic level. Upon receiving a photon of light, rhodopsin undergoes a change in conformation that allows it to bind and activate the C-protein, transducin. An important future goal should be to determine the structure of both the inactive and the photoactivated state of rhodopsin, R*. This should provide the groundwork necessary for experiments on how rhodopsin achieves its signaling state R*, and how R* functions to activate transducin. To do this, the crystal structure of both rhodopsin and R* must be determined. Few membrane proteins have been successfully crystallized, so this is not a trivial undertaking. Two- or three dimensional crystals of rhodopsin must be prepared that are well ordered, to produce a high-resolution structure. Rhodopsin must be purified to homogeneity and the appropriate detergent(s) selected for crystallization experiments. Long-term thermal stability of the rhodopsin-detergent complex must be achieved in the presence of a precipitant. Two-dimensional crystals may offer advantages in investigating the structure of R*, but the structure obtained may be limited in resolution. It is necessary to work with rhodopsin in the dark, unless suitable light-stable retinal derivatives are developed. Protein engineering of rhodopsin offers attractive opportunities to improve its ability to crystallize, but is presently hindered by the absence of a high-yielding expression system. Knowledge of the structure of rhodopsin will have general importance. Because rhodopsin is a member of the family of C-protein-coupled receptors, knowledge of the structure and the mechanism of action of rhodopsin suggests by analogy how other members of the receptor family may function.

The crystal structure of ribosomal protein L5 from Thermus thermophilus complexed with a 34-nt fragment comprising helix III and loop C of Escherichia coli 5S rRNA has been determined at 2.5 Å resolution. The protein specifically interacts with the bulged nucleotides at the top of loop C of 5S rRNA. The rRNA and protein contact surfaces are strongly stabilized by intramolecular interactions. Charged and polar atoms forming the network of conserved intermolecular hydrogen bonds are located in two narrow planar parallel layers belonging to the protein and rRNA, respectively. The regions, including these atoms conserved in Bacteria and Archaea, can be considered an RNA–protein recognition module. Comparison of the T. thermophilus L5 structure in the RNA-bound form with the isolated Bacillus stearothermophilus L5 structure shows that the RNA-recognition module on the protein surface does not undergo significant changes upon RNA binding. In the crystal of the complex, the protein interacts with another RNA molecule in the asymmetric unit through the β-sheet concave surface. This protein/RNA interface simulates the interaction of L5 with 23S rRNA observed in the Haloarcula marismortui 50S ribosomal subunit.

The RNA/DNA 14-mer, (gguauuucgguaCc)2 with consecutive uridine bulges (underlined) on each strand has been determined in two crystal forms, spermine bound (Sp-form) and spermine free (Sp-free). The former was solved by the MAD method with three-wavelength data collected at Brookhaven National Laboratory (BNL); the later isomorphous structure was solved by the molecular replacement method using data collected on our Raxis IIc imaging plate system. The two crystal forms belong to the space group C2 with one molecule of double-stranded 14 mer in the asymmetric unit. The Sp-form has cell constants, a = 60.06, b = 29.10, c = 52.57 Å, β = 120.79° and was refined to 1.7 Å resolution with a final Rwork/Rfree of 19.8%/22.7% using 8,549 independent reflections. The Sp-free structure has cell constants, a = 60.06, b = 29.58, c = 52.50 Å, β = 120.85° and was refined to 1.8 Å with a final Rwork/Rfree of 20.8%/23.2% using 6,285 unique reflections. The two structures are identical, except that the Sp-form has a spermine bound in the major groove, parallel to the RNA helical axis. One of the uridine bulges forms a novel intramolecular U*(A[bull ]U) base triple. The helices are in the C3′-endo conformation (A-form), but the bulges adopt the C2′-endo sugar pucker. Furthermore, the bulges induce a kink (30°) in the helix axis and a very large twist (55°) between the base pairs flanking the bulges. The Sp-form has one Mg2+ ion whereas the Sp-free form has two Mg2+ ions.

Ribosomal protein L5 is a 5S rRNA binding protein in the large subunit and plays an essential role in the promotion of a particular conformation of 5S rRNA. The crystal structure of the ribosomal protein L5 from Bacillus stearothermophilus has been determined at 1.8 Å resolution. The molecule consists of a five-stranded antiparallel β-sheet and four α-helices, which fold in a way that is topologically similar to the ribonucleoprotein (RNP) domain. The molecular shape and electrostatic representation suggest that the concave surface and loop regions are involved in 5S rRNA binding. To identify amino acid residues responsible for 5S rRNA binding, we made use of Ala-scanning mutagenesis of evolutionarily conserved amino acids occurring in the β-strands and loop regions. The mutations of Asn37 at the β1-strand and Gln63 at the loop between helix 2 and β3-strand as well as that of Phe77 at the tip of the loop structure between the β2- and β3-strands caused a significant reduction in 5S rRNA binding. In addition, the mutations of Thr90 on the β3-strand and Ile141 and Asp144 at the loop between β4- and β5-strands moderately reduced the 5S rRNA-binding affinity. Comparison of these results with the more recently analyzed structure of the 50S subunit from Haloarcula marismortui suggests that there are significant differences in the structure at N- and C-terminal regions and probably in the 5S rRNA binding.

β-Lactamases are involved in bacterial resistance. Members of the metallo-enzyme class are now found in many pathogenic bacteria and are becoming thus of major clinical importance. Despite the availability of Zn-β-lactamase X-ray structures their mechanism of action is still unclear. One puzzling observation is the presence of one or two zincs in the active site. To aid in assessing the role of zinc content in β-lactam hydrolysis, the replacement by Ser of the zinc-liganding residue Cys168 in the Zn-β-lactamase from Bacillus cereus strain 569/H/9 was carried out: the mutant enzyme (C168S) is inactive in the mono-Zn form, but active in the di-Zn form. The structure of the mono-Zn form of the C168S mutant has been determined at 1.85 Å resolution. Ser168 occupies the same position as Cys168 in the wild-type enzyme. The protein residues mostly affected by the mutation are Asp90–Arg91 and His210. A critical factor for the activity of the mono-Zn species is the distance between Asp90 and the Zn ion, which is controlled by Arg91: a slight movement of Asp90 impairs catalysis. The evolution of a large superfamily including Zn-β-lactamases suggests that they may not all share the same mechanism.

On the basis of crystal structures of the pyrroloquinoline quinone (PQQ) dependent enzymes methanol dehydrogenase (MDH) and soluble glucose dehydrogenase (s-GDH), different catalytic mechanisms have been proposed. However, several lines of biochemical and kinetic evidence are strikingly similar for both enzymes. To resolve this discrepancy, we have compared the structures of these enzymes in complex with their natural substrates in an attempt to bring them in line with a single reaction mechanism. In both proteins, PQQ is located in the center of the molecule near the axis of pseudo-symmetry. In spite of the absence of significant sequence homology, the overall binding of PQQ in the respective active sites is similar. Hydrogen bonding interactions are made with polar protein side chains in the plane of the cofactor, whereas hydrophobic stacking interactions are important below and above PQQ. One Arg side chain and one calcium ion are ligated to the ortho-quinone group of PQQ in an identical fashion in either active site, in agreement with their proposed catalytic function of polarizing the PQQ C5-O5 bond. The substrates are bound in a similar position above PQQ and within hydrogen bond distance of the putative general bases Asp297 (MDH) and His144 (s-GDH). On the basis of these similarities, we propose that MDH and s-GDH react with their substrates through an identical mechanism, comprising general base-catalyzed hydride transfer from the substrate to PQQ and subsequent tautomerization of the PQQ intermediate to reduced PQQ.

Eucaryotes possess one or more NADP-dependent methylene-THF dehydrogenases as part of multifunctional enzymes. In addition, yeast expresses an unusual monofunctional NAD-dependent enzyme, yMTD. We report X-ray structures for the apoenzyme and its complex with NAD+ at 2.8 and 3.0 Å resolution, respectively. The protein fold resembles that seen for the human and Escherichia coli dehydrogenase/cyclohydrolase bifunctional enzymes. The enzyme has two prominent domains, with the active site cleft between them. yMTD has a noncanonical NAD-binding domain that has two inserted strands compared with the NADP-binding domains of the bifunctional enzymes. This insert precludes yMTD from dimerizing in the same way as the bifunctional enzymes. yMTD functions as a dimer, but the mode of dimerization is novel. It does not appear that the difference in dimerization accounts for the difference in cofactor specificity or for the loss of cyclohydrolase activity. These functional differences are probably accounted for by minor differences within the tertiary structure of the active site of the monomeric protein.

The X-ray structure of chitinase from the fungal pathogen Coccidioides immitis has been solved to 2.2 Å resolution. Like other members of the class 18 hydrolase family, this 427 residue protein is an eight-stranded β/α-barrel. Although lacking an N-terminal chitin anchoring domain, the enzyme closely resembles the chitinase from Serratia marcescens. Among the conserved features are three cis peptide bonds, all involving conserved active site residues. The active site is formed from conserved residues such as tryptophans 47, 131, 315, 378, tyrosines 239 and 293, and arginines 52 and 295. Glu171 is the catalytic acid in the hydrolytic mechanism; it was mutated to a Gln, and activity was abolished. Allosamidin is a substrate analog that strongly inhibits the class 18 enzymes. Its binding to the chitinase hevamine has been observed, and we used conserved structural features of the two enzymes to predict the inhibitors binding to the fungal enzyme.

Parvalbumins constitute a class of calcium-binding proteins characterized by the presence of several helix-loop-helix (EF-hand) motifs. In a previous study (Revett SP, King G, Shabanowitz J, Hunt DF, Hartman KL, Laue TM, Nelson DJ, 1997, Protein Sci 7:2397–2408), we presented the sequence of the major parvalbumin isoform from the silver hake (Merluccius bilinearis) and presented spectroscopic and structural information on the excised “EF-hand” portion of the protein. In this study, the X-ray crystal structure of the silver hake major parvalbumin has been determined to high resolution, in the frozen state, using the molecular replacement method with the carp parvalbumin structure as a starting model. The crystals are orthorhombic, space group C2221, with a = 75.7 Å, b = 80.7 Å, and c = 42.1 Å. Data were collected from a single crystal grown in 15% glycerol, which served as a cryoprotectant for flash freezing at −188 °C. The structure refined to a conventional R-value of 21% (free R 25%) for observed reflections in the range 8 to 1.65 Å [I > 2σ(I)]. The refined model includes an acetylated amino terminus, 108 residues (characteristic of a β parvalbumin lineage), 2 calcium ions, and 114 water molecules per protein molecule. The resulting structure was used in molecular dynamics (MD) simulations focused primarily on the dynamics of the ligands coordinating the Ca2+ ions in the CD and EF sites. MD simulations were performed on both the fully Ca2+ loaded protein and on a Ca2+ deficient variant, with Ca2+ only in the CD site. There was substantial agreement between the MD and X-ray results in addressing the issue of mobility of key residues in the calcium-binding sites, especially with regard to the side chain of Ser55 in the CD site and Asp92 in the EF site.

Turkey ovomucoid third domain (OMTKY3) is a canonical inhibitor of serine proteinases. Upon complex formation, the inhibitors fully exposed P1 residue becomes fully buried in the preformed cavity of the enzyme. All 20 P1 variants of OMTKY3 have been obtained by recombinant DNA technology and their equilibrium association constants have been measured with six serine proteinases. To rationalize the trends observed in this data set, high resolution crystal structures have been determined for OMTKY3 P1 variants in complex with the bacterial serine proteinase, Streptomyces griseus proteinase B (SGPB). Four high resolution complex structures are being reported in this paper; the three β-branched variants, Ile18I, Val18I, and Thr18I, determined to 2.1, 1.6, and 1.7 Å resolution, respectively, and the structure of the Ser18I variant complex, determined to 1.9 Å resolution. Models of the Cys18I, Hse18I, and Ape18I variant complexes are also discussed. The β-branched side chains are not complementary to the shape of the S1 binding pocket in SGPB, in contrast to that of the wild-type γ-branched P1 residue for OMTKY3, Leu18I. χ1 angles of approximately 40° are imposed on the side chains of Ile18I, Val18I, and Thr18I within the S1 pocket. Dihedral angles of +60°, −60°, or 180° are more commonly observed but 40° is not unfavorable for the β-branched side chains. Thr18I Oγ1 also forms a hydrogen bond with Ser195 Oγ in this orientation. The Ser18I side chain adopts two alternate conformations within the S1 pocket of SGPB, suggesting that the side chain is not stable in either conformation.

The YER057c/YIL051c/YjgF protein family is a set of 24 full-length homologs, each ∼130 residues in length, and each with no known function or relationship to proteins of known structure. To determine the function of this family, the structure of one member—the YjgF protein from Escherichia coli—was solved and refined at a resolution of 1.2 Å. The YjgF molecule is a homotrimer with exact threefold symmetry. Its tertiary and quaternary structures are related to that of Bacillus subtilis chorismate mutase, although their active sites are completely different. The YjgF protein has an active site curiously similar to protein tyrosine phosphatases, including a covalently modified cysteine, but it is unlikely to be functionally related. The lessons learned from this attempt to deduce function from structure may be useful to future projects in structural genomics.

Catalysis by thioredoxin reductase (TrxR) from Escherichia coli requires alternation between two domain arrangements. One of these conformations has been observed by X-ray crystallography (Waksman G, Krishna TSR, Williams CH Jr, Kuriyan J, 1994, J Mol Biol 236:800–816). This form of TrxR, denoted FO, permits the reaction of enzyme-bound reduced FAD with a redox-active disulfide on TrxR. As part of an investigation of conformational changes and intermediates in catalysis by TrxR, an X-ray structure of the FO form of TrxR with both the FAD and active site disulfide reduced has been determined. Reduction after crystallization resulted in significant local conformation changes. The isoalloxazine ring of the FAD cofactor, which is essentially planar in the oxidized enzyme, assumes a 34° “butterfly” bend about the N(5)–N(10) axis in reduced TrxR. Theoretical calculations reported by others predict ring bending of 15–28° for reduced isoalloxazines protonated at N(1). The large bending in reduced TrxR is attributed in part to steric interactions between the isoalloxazine ring and the sulfur of Cys138, formed by reduction of the active site disulfide, and is accompanied by changes in the positions and interactions of several of the ribityl side-chain atoms of FAD. The bending angle in reduced TrxR is larger than that for any flavoprotein in the Protein Data Bank. Distributions of bending angles in published oxidized and reduced flavoenzyme structures are different from those found in studies of free flavins, indicating that the protein environment has a significant effect on bending.

Short chain l-3-hydroxyacyl CoA dehydrogenase (SCHAD) is a soluble dimeric enzyme critical for oxidative metabolism of fatty acids. Its primary sequence has been reported to be conserved across numerous tissues and species with the notable exception of the pig heart homologue. Preliminary efforts to solve the crystal structure of the dimeric pig heart SCHAD suggested the unprecedented occurrence of three enzyme subunits within the asymmetric unit, a phenomenon that was thought to have hampered refinement of the initial chain tracing. The recently solved crystal coordinates of human heart SCHAD facilitated a molecular replacement solution to the pig heart SCHAD data. Refinement of the model, in conjunction with the nucleotide sequence for pig heart SCHAD determined in this paper, has demonstrated that the previously published pig heart SCHAD sequence was incorrect. Presented here are the corrected amino acid sequence and the high resolution crystal structure determined for pig heart SCHAD complexed with its NAD+ cofactor (2.8 Å; Rcryst = 22.4%, Rfree = 28.8%). In addition, the peculiar phenomenon of a dimeric enzyme crystallizing with three subunits contained in the asymmetric unit is described.

A retinoic acid binding protein isolated from the lumen of the rat epididymis (ERABP) is a member of the lipocalin superfamily. ERABP binds both the all-trans and 9-cis isomers of retinoic acid, as well as the synthetic retinoid (E)-4-[2-(5,6,7,8)-tetrahydro-5,5,8,8-tetramethyl-2 napthalenyl-1 propenyl]-benzoic acid (TTNPB), a structural analog of all-trans retinoic acid. The structure of ERABP with a mixture of all-trans and 9-cis retinoic acid has previously been reported. To elucidate any structural differences in the protein when bound to the all-trans and 9-cis isomers, the structures of all-trans retinoic acid–ERABP and 9-cis retinoic acid ERABP were determined. Our results indicate that the all-trans isomer of retinoic acid adopts an 8-cis structure in the binding cavity with no concomitant conformational change in the protein. The structure of TTNPB-ERABP is also reported herein. To accommodate this all-trans analog, which cannot readily adopt a cis-like structure, alternative positioning of critical binding site side chains is required. Consequently, both protein and ligand adaption are observed in the formation of the various holo-proteins.

The X-ray structure of the complex formed between yeast 5-aminolaevulinic acid dehydratase (ALAD) and the inhibitor laevulinic acid has been determined at 2.15 Å resolution. The inhibitor binds by forming a Schiff base link with one of the two invariant lysines at the catalytic center: Lys263. It is known that this lysine forms a Schiff base link with substrate bound at the enzyme's so-called P-site. The carboxyl group of laevulinic acid makes hydrogen bonds with the side-chain–OH groups of Tyr329 and Ser290, as well as with the main-chain >NH group of Ser290. The aliphatic moiety of the inhibitor makes hydrophobic interactions with surrounding aromatic residues in the protein including Phe219, which resides in the flap covering the active site. Our analysis strongly suggests that the same interactions will be made by P-side substrate and also indicates that the substrate that binds at the enzyme's A-site will interact with the enzyme's zinc ion bound by three cysteines (133, 135, and 143). Inhibitor binding caused a substantial ordering of the active site flap (residues 217–235), which was largely invisible in the native electron density map and indicates that this highly conserved yet flexible region has a specific role in substrate binding during catalysis.

Cytochrome cH is the electron donor to the oxidase in methylotrophic bacteria. Its amino acid sequence suggests that it is a typical Class I cytochrome c, but some features of the sequence indicated that its structure might be of special interest. The structure of oxidized cytochrome cH has been solved to 2.0 Å resolution by X-ray diffraction. It has the classical tertiary structure of the Class 1 cytochromes c but bears a closer gross resemblance to mitochondrial cytochrome c than to the bacterial cytochrome c2. The left-hand side of the haem cleft is unique; in particular, it is highly hydrophobic, the usual water is absent, and the “conserved” Tyr67 is replaced by tryptophan. A number of features of the structure demonstrate that the usual hydrogen bonding network involving water in the haem channel is not essential and that other mechanisms may exist for modulation of redox potentials in this cytochrome.