Crystal structures of the GH18 domain of the bifunctional peroxiredoxin-chitinase CotE from Clostridium difficile

CotE is a coat protein that is present in the spores of Clostridium difficile, an obligate anaerobic bacterium and a pathogen that is a leading cause of antibiotic-associated diarrhoea in hospital patients. Spores serve as the agents of disease transmission, and CotE has been implicated in their attachment to the gut epithelium and subsequent colonization of the host. CotE consists of an N-terminal peroxiredoxin domain and a C-terminal chitinase domain. Here, a C-terminal fragment of CotE comprising residues 349-712 has been crystallized and its structure has been determined to reveal a core eight-stranded β-barrel fold with a neighbouring subdomain containing a five-stranded β-sheet. A prominent groove running across the top of the barrel is lined by residues that are conserved in family 18 glycosyl hydrolases and which participate in catalysis. Electron density identified in the groove defines the pentapeptide Gly-Pro-Ala-Met-Lys derived from the N-terminus of the protein following proteolytic cleavage to remove an affinity-purification tag. These observations suggest the possibility of designing peptidomimetics to block C. difficile transmission.

Defining the remarkable structural malleability of a bacterial surface protein Rib domain implicated in infection

Streptococcus groups A and B cause serious infections, including early onset sepsis and meningitis in newborns. Rib domain-containing surface proteins are found associated with invasive strains and elicit protective immunity in animal models. Yet, despite their apparent importance in infection, the structure of the Rib domain was previously unknown. Structures of single Rib domains of differing length reveal a rare case of domain atrophy through deletion of 2 core antiparallel strands, resulting in the loss of an entire sheet of the β-sandwich from an immunoglobulin-like fold. Previously, observed variation in the number of Rib domains within these bacterial cell wall-attached proteins has been suggested as a mechanism of immune evasion. Here, the structure of tandem domains, combined with molecular dynamics simulations and small angle X-ray scattering, suggests that variability in Rib domain number would result in differential projection of an N-terminal host-colonization domain from the bacterial surface. The identification of 2 further structures where the typical B-D-E immunoglobulin β-sheet is replaced with an α-helix further confirms the extensive structural malleability of the Rib domain.

The Rhodococcus equi virulence protein VapA disrupts endolysosome function and stimulates lysosome biogenesis

Rhodococcus equi (R. equi) is an important pulmonary pathogen in foals that often leads to the death of the horse. The bacterium harbors a virulence plasmid that encodes numerous virulence‐associated proteins (Vaps) including VapA that is essential for intracellular survival inside macrophages. However, little is known about the precise function of VapA. Here, we demonstrate that VapA causes perturbation to late endocytic organelles with swollen endolysosome organelles having reduced Cathepsin B activity and an accumulation of LBPA, LC3 and Rab7. The data are indicative of a loss of endolysosomal function, which leads cells to upregulate lysosome biogenesis to compensate for the loss of functional endolysosomes. Although there is a high degree of homology of the core region of VapA to other Vap proteins, only the highly conserved core region of VapA, and not VapD of VapG, gives the observed effects on endolysosomes. This is the first demonstration of how VapA works and implies that VapA aids R. equi survival by reducing the impact of lysosomes on phagocytosed bacteria.

Structural characterisation of the virulence-associated protein VapG from the horse pathogen Rhodococcus equi

•

The 3-dimensional structure of a Rhodococcus equi virulence protein was determined.

•

VapG comprises a closed beta barrel domain preceded by a natively disordered region.

•

The structures of VapB, VapD and VapG are closely superimposable.

•

The VAP structures lack recognisable ligand or protein binding sites.

•

Phagosome-induced conformational changes may be required for virulence.

Virulence and host range in Rhodococcus equi depends on the variable pathogenicity island of their virulence plasmids. Notable gene products are a family of small secreted virulence-associated proteins (Vaps) that are critical to intramacrophagic proliferation. Equine-adapted strains, which cause severe pyogranulomatous pneumonia in foals, produce a cell-associated VapA that is necessary for virulence, alongside five other secreted homologues. In the absence of biochemical insight, attention has turned to the structures of these proteins to develop a functional hypothesis. Recent studies have described crystal structures for VapD and a truncate of the VapA orthologue of porcine-adapted strains, VapB. Here, we crystallised the full-length VapG and determined its structure by molecular replacement. Electron density corresponding to the N-terminal domain was not visible suggesting that it is disordered. The protein core adopted a compact elliptical, anti-parallel β-barrel fold with β1–β2–β3–β8–β5–β6–β7–β4 topology decorated by a single peripheral α-helix unique to this family. The high glycine content of the protein allows close packing of secondary structural elements. Topologically, the surface has no indentations that indicate a nexus for molecular interactions. The distribution of polar and apolar groups on the surface of VapG is markedly uneven. One-third of the surface is dominated by exposed apolar side-chains, with no ionisable and only four polar side-chains exposed, giving rise to an expansive flat hydrophobic surface. Other surface regions are more polar, especially on or near the α-helix and a belt around the centre of the β-barrel. Possible functional significance of these recent structures is discussed.

Structure of the virulence-associated protein VapD from the intracellular pathogen Rhodococcus equi

Rhodococcus equi is a multi-host pathogen that infects a range of animals as well as immune-compromised humans. Equine and porcine isolates harbour a virulence plasmid encoding a homologous family of virulence-associated proteins associated with the capacity of R. equi to divert the normal processes of endosomal maturation, enabling bacterial survival and proliferation in alveolar macrophages. To provide a basis for probing the function of the Vap proteins in virulence, the crystal structure of VapD was determined. VapD is a monomer as determined by multi-angle laser light scattering. The structure reveals an elliptical, compact eight-stranded β-barrel with a novel strand topology and pseudo-twofold symmetry, suggesting evolution from an ancestral dimer. Surface-associated octyl-β-D-glucoside molecules may provide clues to function. Circular-dichroism spectroscopic analysis suggests that the β-barrel structure is preceded by a natively disordered region at the N-terminus. Sequence comparisons indicate that the core folds of the other plasmid-encoded virulence-associated proteins from R. equi strains are similar to that of VapD. It is further shown that sequences encoding putative R. equi Vap-like proteins occur in diverse bacterial species. Finally, the functional implications of the structure are discussed in the light of the unique structural features of VapD and its partial structural similarity to other β-barrel proteins.

Synthesis and evaluation of α-thymidine analogues as novel antimalarials

Plasmodium falciparum thymidylate kinase (PfTMPK) is a key enzyme in pyrimidine nucleotide biosynthesis. 3-Trifluoromethyl-4-chloro-phenyl-urea-α-thymidine has been reported as an inhibitor of Mycobacterium tuberculosis TMPK (MtTMPK). Starting from this point, we designed, synthesized and evaluated a number of thymidine analogues as antimalarials. Both 5′-urea-α- and β-thymidine derivatives were moderate inhibitors of PfTMPK and furthermore showed moderate inhibition of parasite growth. The structure of several enzyme–inhibitor complexes provides a basis for improved inhibitor design. However, we found that certain 5′-urea-α-thymidine analogues had antimalarial activity where inhibition of PfTMPK is not the major mode of action. Optimization of this series resulted in a compound with potent antimalarial activity (EC₅₀ = 28 nM; CC₅₀ = 29 μM).

Design, synthesis, and evaluation of 5'-diphenyl nucleoside analogues as inhibitors of the Plasmodium falciparum dUTPase

Deoxyuridine 5'-triphosphate nucleotidohydrolase (dUTPase) is a potential drug target for malaria. We previously reported some 5'-tritylated deoxyuridine analogues (both cyclic and acyclic) as selective inhibitors of the Plasmodium falciparum dUTPase. Modelling studies indicated that it might be possible to replace the trityl group with a diphenyl moiety, as two of the phenyl groups are buried, whereas the third is exposed to solvent. Herein we report the synthesis and evaluation of some diphenyl analogues that have lower lipophilicity and molecular weight than the trityl lead compound. Co-crystal structures show that the diphenyl inhibitors bind in a similar manner to the corresponding trityl derivatives, with the two phenyl moieties occupying the predicted buried phenyl binding sites. The diphenyl compounds prepared show similar or slightly lower inhibition of PfdUTPase, and similar or weaker inhibition of parasite growth than the trityl compounds.

Binding to DNA protects Neisseria meningitidis fumarate and nitrate reductase regulator (FNR) from oxygen

Here, we report the overexpression, purification, and characterization of the transcriptional activator fumarate and nitrate reductase regulator from the pathogenic bacterium Neisseria meningitidis (NmFNR). Like its homologue from Escherichia coli (EcFNR), NmFNR binds a 4Fe-4S cluster, which breaks down in the presence of oxygen to a 2Fe-2S cluster and subsequently to apo-FNR. The kinetics of NmFNR cluster disassembly in the presence of oxygen are 2-3x slower than those previously reported for wild-type EcFNR, but similar to constitutively active EcFNR* mutants, consistent with earlier work in which we reported that the activity of FNR-dependent promoters in N. meningitidis is only weakly inhibited by the presence of oxygen (Rock, J. D., Thomson, M. J., Read, R. C., and Moir, J. W. (2007) J. Bacteriol. 189, 1138-1144). NmFNR binds to DNA containing a consensus FNR box sequence, and this binding stabilizes the iron-sulfur cluster in the presence of oxygen. Partial degradation of the 4Fe-4S cluster to a 3Fe-4S occurs, and this form remains bound to the DNA. The 3Fe-4S cluster is converted spontaneously back to a 4Fe-4S cluster under subsequent anaerobic reducing conditions in the presence of ferrous iron. The finding that binding to DNA stabilizes FNR in the presence of oxygen such that it has a half-life of approximately 30 min on the DNA has implications for our appreciation of how oxygen switches off FNR activatable genes in vivo.

Structural characterization of a beta-diketone hydrolase from the cyanobacterium Anabaena sp. PCC 7120 in native and product-bound forms, a coenzyme A-independent member of the crotonase suprafamily

The gene alr4455 from the well-studied cyanobacterium Anabaena sp. PCC 7120 encodes a crotonase orthologue that displays beta-diketone hydrolase activity. Anabaena beta-diketone hydrolase (ABDH), in common with 6-oxocamphor hydrolase (OCH) from Rhodococcus sp. NCIMB 9784, catalyzes the desymmetrization of bicyclo[2.2.2]octane-2,6-dione to yield [(S)-3-oxocyclohexyl]acetic acid, a reaction unusual among the crotonase superfamily as the substrate is not an acyl-CoA thioester. The structure of ABDH has been determined to a resolution of 1.5 A in both native and ligand-bound forms. ABDH forms a hexamer similar to OCH and features one active site per enzyme monomer. The arrangement of side chains in the active site indicates that while the catalytic chemistry may be conserved in OCH orthologues, the structural determinants of substrate specificity are different. In the active site of ligand-bound forms that had been cocrystallized with the bicyclic diketone substrate bicyclo[2.2.2]octane-2,6-dione was found the product of the asymmetric enzymatic retro-Claisen reaction [(S)-3-oxocyclohexyl]acetic acid. The structures of ABDH in both native and ligand-bound forms reveal further details about structural variation and modes of coenzyme A-independent activity within the crotonases and provide further evidence of a wider suprafamily of enzymes that have recruited the crotonase fold for the catalysis of reactions other than those regularly attributed to canonical superfamily members.

I222 crystal form of despentapeptide (B26-B30) insulin provides new insights into the properties of monomeric insulin

Despentapeptide (des-B26-B30) insulin (DPI), an active modified insulin, has been crystallized in the presence of 20% acetic acid pH 2. A crystal structure analysis to 1.8 A spacing (space group I222) revealed that the DPI molecule, which is unable to make beta-strand interactions for physiological dimer formation and is apparently monomeric in solution, formed an alternative lattice-generated dimer. The formation of this dimer involved interactions between surfaces which included the B9-B19 alpha-helices (usually buried by the dimer-dimer contacts within the native hexamer). The two crystallographically independent molecules within the dimer were essentially identical and were similar in conformation to T-state insulin as seen in the T(6) insulin hexamer. An unusual feature of each molecule in the dimer was the presence of two independent conformations at the B-chain C-terminus (residues B20-B25). Both conformations were different from that of native insulin, involving a 3.5 A displacement of the B20-B23 beta-turn and a repositioning of residue PheB25 such that it made close van der Waals contact with the main body of the molecule, appearing to stabilize the B-chain C-terminus.

dUTPase as a platform for antimalarial drug design: structural basis for the selectivity of a class of nucleoside inhibitors

Pyrimidine metabolism is a major route for therapeutic intervention against malaria. Here we report inhibition and structural studies on the deoxyuridine nucleotidohydrolase from the malaria parasite Plasmodium falciparum (PfdUTPase). We have identified a series of triphenylmethane derivatives of deoxyuridine with antimalarial activity in vitro which inhibit specifically the Plasmodium dUTPase versus the human enzyme. A 2.4 Angstrom crystal structure of PfdUTPase in complex with one of these inhibitors reveals an atypical trimeric enzyme in which the triphenylmethane derivative can be seen to select for PfdUTPase by way of interactions between the trityl group and the side chains of residues Phe46 and Ile117. Immunofluorescence microscopy studies of parasitized red blood cells reveal that enzyme concentrations are highest during the trophozoite/schizont stages, suggesting that PfdUTPase has a major role in DNA replication. Taken together the data show that PfdUTPase may be considered as an antimalarial drug target.

Toward the insulin-IGF-I intermediate structures: functional and structural properties of the [TyrB25NMePheB26] insulin mutant

The origins of differentiation of insulin from insulin-like growth factor I (IGF-I) are still unknown. To address the problem of a structural and biological switch from the mostly metabolic hormonal activity of insulin to the predominant growth factor activities of IGF-I, an insulin analogue with IGF-I-like structural features has been synthesized. Insulin residues Phe(B25) and Tyr(B26) have been swapped with the IGF-I-like Tyr(24) and Phe(25) sequence with a simultaneous methylation of the peptide nitrogen of residue Phe(B26). These modifications were expected to introduce a substantial kink in the main chain, as observed at residue Phe(25) in the IGF-I crystal structure. These alterations should provide insight into the structural origins of insulin-IGF-I structural and functional divergence. The [Tyr(B25)NMePhe(B26)] mutant has been characterized, and its crystal structure has been determined. Surprisingly, all of these changes are well accommodated within an insulin R6 hexamer. Only one molecule of each dimer in the hexamer responds to the structural alterations, the other remaining very similar to wild-type insulin. All alterations, modest in their scale, cumulate in the C-terminal part of the B-chain (residues B23-B30), which moves toward the core of the insulin molecule and is associated with a significant shift of the A1 helix toward the C-terminus of the B-chain. These changes do not produce the expected bend of the main chain, but the fold of the mutant does reflect some structural characteristics of IGF-1, and in addition establishes the CO(A19)-NH(B25) hydrogen bond, which is normally characteristic of T-state insulin.

Crystallographic and solution studies of N-lithocholyl insulin: a new generation of prolonged-acting human insulins

The addition of specific bulky hydrophobic groups to the insulin molecule provides it with affinity for circulating serum albumin and enables it to form soluble macromolecular complexes at the site of subcutaneous injection, thereby securing slow absorption of the insulin analogue into the blood stream and prolonging its half-life once there. N-Lithocholic acid acylated insulin [Lys(B29)-lithocholyl des-(B30) human insulin] has been crystallized and the structure determined by X-ray crystallography at 1.6 A resolution to explore the molecular basis of its assembly. The unit cell in the crystal consists of an insulin hexamer containing two zinc ions, with two m-cresol molecules bound at each dimer-dimer interface stabilizing an R(6) conformation. Six covalently bound lithocholyl groups are arranged symmetrically around the outside of the hexamer. These form specific van der Waals and hydrogen-bonding interactions at the interfaces between neighboring hexamers, possibly representing the kinds of interactions which occur in the soluble aggregates at the site of injection. Comparison with an equivalent nonderivatized native insulin hexamer shows that the addition of the lithocholyl group disrupts neither the important conformational features of the insulin molecule nor its hexamer-forming ability. Indeed, binding studies show that the affinity of N-lithocholyl insulin for the human insulin receptor is not significantly diminished.

The 2-A crystal structure of 6-oxo camphor hydrolase. New structural diversity in the crotonase superfamily

6-Oxo camphor hydrolase (OCH) is an enzyme of the crotonase superfamily that catalyzes carbon-carbon bond cleavage in bicyclic beta-diketones via a retro-Claisen reaction (Grogan, G., Roberts, G. A., Bougioukou, D., Turner, N. J., and Flitsch, S. L. (2001) J. Biol. Chem. 276, 12565-12572). The native structure of OCH has been solved at 2.0-A resolution with selenomethionine multiple wave anomalous dispersion and refined to a final R(free) of 19.0. The structure of OCH consists of a dimer of trimers that resembles the "parent" enzyme of the superfamily, enoyl-CoA hydratase. In contrast to enoyl-CoA hydratase, however, two octahedrally coordinated sodium atoms are found at the 3-fold axis of the hexamer of OCH, and the C-terminal helix of OCH does not form a discrete domain. Models of the substrate, 6-oxo camphor, and a proposed enolate intermediate in the putative active site suggest possible mechanistic roles for Glu-244, Asp-154, His-122, His-45, and His-145.

Insulin at pH 2: structural analysis of the conditions promoting insulin fibre formation

When insulin solutions are subjected to acid, heat and agitation, the normal pattern of insulin assembly (dimers-->tetramers-->hexamers) is disrupted; the molecule undergoes conformational changes allowing it to follow an alternative aggregation pathway (via a monomeric species) leading to the formation of insoluble amyloid fibres. To investigate the effect of acid pH on the conformation and aggregation state of the protein, the crystal structure of human insulin at pH 2.1 has been determined to 1.6 A resolution. The structure reveals that the native fold is maintained at low pH, and that the molecule is still capable of forming dimers similar to those found in hexameric insulin structures at higher pH. Sulphate ions are incorporated into the molecule and the crystal lattice where they neutralise positive charges on the protein, stabilising its structure and facilitating crystallisation. The sulphate interactions are associated with local deformations in the protein, which may indicate that the structure is more plastic at low pH. Transmission electron microscopy analysis of insulin fibres reveals that the appearance of the fibres is greatly influenced by the type of acid employed. Sulphuric acid produces distinctive highly bunched, truncated fibres, suggesting that the sulphate ions have a sophisticated role to play in fibre formation, rather as they do in the crystal structure. Analytical ultracentrifugation studies show that in the absence of heating, insulin is predominantly dimeric in mineral acids, whereas in acetic acid the equilibrium is shifted towards the monomer. Hence, the effect of acid on the aggregation state of insulin is also complex. These results suggest that acid conditions increase the susceptibility of the molecule to conformational change and dissociation, and enhance the rate of fibrillation by providing a charged environment in which the attractive forces between the protein molecules is increased.

Crystallization and preliminary crystallographic investigation of a low-pH native insulin monomer with flexible behaviour

Insulin naturally aggregates as dimers and hexamers, whose structures have been extensively analysed by X-ray crystallography. Structural determination of the physiologically relevant insulin monomer, however, is an unusual challenge owing to the difficulty in finding solution conditions in which the concentration of insulin is high enough for crystallization yet the molecule remains monomeric. By utilizing solution conditions known to inhibit insulin assembly, namely 20% acetic acid, crystals of insulin in the monomeric state have been obtained. The crystals are strongly diffracting and a data set extending to 1.6 A has recently been collected. The crystals nominally belong to the space group I422, with unit-cell parameters a = b = 57.80, c = 54.61 A, giving rise to one molecule in the asymmetric unit. Preliminary electron-density maps show that whilst most of the insulin monomer is well ordered and similar in conformation to other insulin structures, parts of the B-chain C-terminus main chain adopt more than one conformation.

Sulfatide promotes the folding of proinsulin, preserves insulin crystals, and mediates its monomerization

Sulfatide is a glycolipid that has been associated with insulin-dependent diabetes mellitus. It is present in the islets of Langerhans and follows the same intracellular route as insulin. However, the role of sulfatide in the beta cell has been unclear. Here we present evidence suggesting that sulfatide promotes the folding of reduced proinsulin, indicating that sulfatide possesses molecular chaperone activity. Sulfatide associates with insulin by binding to the insulin domain A8--A10 and most likely by interacting with the hydrophobic side chains of the dimer-forming part of the insulin B-chain. Sulfatide has a dual effect on insulin. It substantially reduces deterioration of insulin hexamer crystals at pH 5.5, conferring stability comparable to those in beta cell granules. Sulfatide also mediates the conversion of insulin hexamers to the biological active monomers at neutral pH, the pH at the beta-cell surface. Finally, we report that inhibition of sulfatide synthesis with chloroquine and fumonisine B1 leads to inhibition of insulin granule formation in vivo. Our observations suggest that sulfatide plays a key role in the folding of proinsulin, in the maintenance of insulin structure, and in the monomerization process.

Structural consequences of the B5 histidine --> tyrosine mutation in human insulin characterized by X-ray crystallography and conformational analysis

The addition of phenols to hexameric insulin solutions produces a particularly stable hexamer, resulting from a rearrangement in which residues B1-B8 change from an extended conformation (T-state) to form an alpha-helix (R-state). The R-state is, in part, stabilized by nonpolar interactions between the phenolic molecule and residue B5 His at the dimer-dimer interface. The B5 His --> Tyr mutant human insulin was constructed to see if the tyrosine side chain would mimic the effect of phenol binding in the hexamer and induce the R-state. In partial support of this hypothesis, the molecule crystallized as a half-helical hexamer (T(3)R(3)) in conditions that conventionally promote the fully nonhelical (T6) form. As expected, in the presence of phenol or resorcinol, the B5 Tyr hexamers adopt the fully helical (R6) conformation. Molecular modeling calculations were performed to investigate the conformational preference of the T-state B5 Tyr side chain in the T(3)R(3) form, this side chain being associated with structural perturbations of the A7-A10 loop in an adjacent hexamer. For an isolated dimer, several different orientations of the side chain were found, which were close in energy and readily interconvertible. In the crystal environment only one of these conformations remains low in energy; this conformation corresponds to that observed in the crystal structure. This suggests that packing constraints around residue B5 Tyr result in the observed structural rearrangements. Thus, rather than promoting the R-state in a manner analogous to phenol, the mutation appears to destabilize the T-state. These studies highlight the role of B5 His in determining hexamer conformation and in mediating crystal packing interactions, properties that are likely be important in vivo.

Interactions of phenol and m-cresol in the insulin hexamer, and their effect on the association properties of B28 pro --> Asp insulin analogues

Insulin's natural tendency to form dimers and hexamers is significantly reduced in a mutant insulin B28 Pro --> Asp, which has been designed as a monomeric, rapid-acting hormone for therapeutic purposes. This molecule can be induced to form zinc hexamers in the presence of small phenolic derivatives which are routinely used as antimicrobial agents in insulin preparations. Two structures of B28 Asp insulin have been determined from crystals grown in the presence of phenol and m-cresol. In these crystals, insulin exists as R6 zinc hexamers containing a number of phenol or m-cresol molecules associated with aromatic side chains at the dimer-dimer interfaces. At the monomer-monomer interfaces, the B28 Pro --> Asp mutation leads to increased conformational flexibility in the B chain C termini, resulting in the loss of important intermolecular van der Waals contacts, thus explaining the monomeric character of B28 Asp insulin. The structure of a cross-linked derivative of B28 Asp insulin, containing an Ala-Lys dipeptide linker between residues B30 Ala and A1 Gly, has also determined. This forms an R6 zinc hexamer containing several m-cresol molecules. Of particular interest in this structure are two m-cresol molecules whose binding disrupted the beta-strand in one of the dimers. This observation suggests that the cross-link introduces mechanical strain on the B chain C terminus, thereby weakening the monomer-monomer interactions.

A model of insulin fibrils derived from the x-ray crystal structure of a monomeric insulin (despentapeptide insulin)

The crystal structure of despentapeptide insulin, a monomeric insulin, has been refined at 1.3 A spacing and subsequently used to predict and model the organization in the insulin fibril. The model makes use of the contacts in the densely packed despentapeptide insulin crystal, and takes into account other experimental evidence, including binding studies with Congo red. The dimensions of this model fibril correspond well with those measured experimentally, and the monomer-monomer contacts within the fibril are in accordance with the known physical chemistry of insulin fibrils. Using this model, it may be possible to predict mutations in insulin that might alleviate problems associated with fibril formation during insulin therapy.

Crystal structure of a prolonged-acting insulin with albumin-binding properties

The fatty acid acylated insulin, Lys(B29)-tetradecanoyl, des-(B30) human insulin, has been crystallized and the structure determined by X-ray crystallography. The fatty acid substituent on residue B29 Lys binds reversibly to circulating albumin protein in vivo, and by this mechanism the hormone's action is prolonged. Crystals of the fatty acid insulin grow in space group R3, with two dimers in the asymmetric unit, and diffract to 1.8 A spacing. The structure has been solved by molecular replacement and refined using a maximum likelihood method. The crystal structure consists of R6 zinc insulin hexamers which contain phenol. The fatty acids can be seen bound between the hexamers, making specific interactions with the side chains of residue B1 Phe; however, the lysine side chains to which the fatty acids are covalently attached are mostly disordered. The mode of binding of the fatty acids appears to be determined by crystal packing, and whether or not they interact with the protein in this way in solution remains uncertain.

X-ray crystallographic studies on hexameric insulins in the presence of helix-stabilizing agents, thiocyanate, methylparaben, and phenol

Three X-ray crystallographic studies have been carried out on pig insulin in the presence of three ligands, thiocyanate, methylparaben (methyl p-hydroxybenzoate), and phenol. In each case, rhombohedral crystals were obtained, which diffracted to 1.8, 1.9, and 2.3 A, respectively. Each crystal structure was very similar to that of 4-zinc pig insulin, which was used as a starting model for PROLSQ refinement (Collaborative Computational Project, Number 4, 1994). The R factors for the refined structures of thiocyanate insulin, methylparaben insulin, and phenol insulin were 19.6, 18.4, and 19.1, respectively. Each crystal structure consists of T3R3f insulin hexamers with two zinc ions per hexamer. In the R3f trimer of the thiocyanate insulin hexamer, one thiocyanate ion is coordinated to the zinc on the hexamer 3-fold axis, but there is no evidence of zinc ion binding in the off-axis zinc ion sites seen in the 4-zinc pig insulin structure. In the methylparaben insulin and phenol insulin hexamers, the phenolic ligands are bound at the dimer-dimer interfaces in the R3f trimers in a manner similar to that of phenol in R6 phenol insulin. The binding of methylparaben appears to make the hexamer more compact by drawing the A and the B chains closer together in the binding site. In all three structures presented herein, the conformations of the first three residues of the B chain in the R3f trimer are extended rather than alpha-helical, as is seen in R6 phenol insulin. The energetics of ligand binding in the insulin hexamer are discussed.