Structure of the UDP-glucosyltransferase GtfB that modifies the heptapeptide aglycone in the biosynthesis of vancomycin group antibiotics.

BACKGROUND: Members of the vancomycin group of glycopeptide antibiotics have an oxidatively crosslinked heptapeptide scaffold decorated at the hydroxyl groups of 4-OH-Phegly4 or beta-OH-Tyr6 with mono- (residue 6) or disaccharides (residue 4). The disaccharide in vancomycin itself is L-vancosamine-1,2-glucose, and in chloroeremomycin it is L-4-epi-vancosamine-1,2-glucose. The sugars and their substituents play an important role in efficacy, particularly against vancomycin-resistant pathogenic enterococci. RESULTS: The glucosyltransferase, GtfB, that transfers the glucose residue from UDP-glucose to the 4-OH-Phegly4 residue of the vancomycin aglycone, initiating the glycosylation pathway in chloroeremomycin maturation, has been crystallized, and its structure has been determined by X-ray analysis at 1.8 A resolution. The enzyme has a two-domain structure, with a deep interdomain cleft identified as the likely site of UDP-glucose binding. A hydrophobic patch on the surface of the N-terminal domain is proposed to be the binding site of the aglycone substrate. Mutagenesis has revealed Asp332 as the best candidate for the general base in the glucosyltransfer reaction. CONCLUSIONS: The structure of GtfB places it in a growing group of glycosyltransferases, including Escherichia coli MurG and a beta-glucosyltransferase from T4 phage, which together form a subclass of the glycosyltransferase superfamily and give insights into the recognition of the NDP-sugar and aglycone cosubstrates. A single major interdomain linker between the N- and C- terminal domains suggests that reprogramming of sugar transfer or aglycone recognition in the antibiotic glycosyltransferases, including the glycopeptide and also the macrolide antibiotics, will be facilitated by this structural information.

Use of medium-sized cycloalkyl rings to enhance secondary binding: discovery of a new class of human immunodeficiency virus (HIV) protease inhibitors.

A unique strategy for the enhancement of secondary binding of an inhibitor to an enzyme has been demonstrated in the design of new human immunodeficiency virus (HIV) protease inhibitors. When the planar benzene ring of a 4-hydroxycoumarin lead compound (1a, Ki = 0.800 microM) was replaced with medium-sized (i.e., 7-9), conformationally-flexible, alkyl rings, the enzyme inhibitory activity of the resulting compounds was dramatically improved, and inhibitors with more than 50-fold better binding (e.g., 5d, Ki = 0.015 microM) were obtained. X-ray crystal structures of these inhibitors complexed with HIV protease indicated the cycloalkyl rings were able to fold into the S1' pocket of the enzyme and fill it much more effectively than the rigid benzene ring of the 4-hydroxycoumarin compound. This work has resulted in the identification of a promising lead structure for the design of potent, deliverable HIV protease inhibitors. Compound 5d, a small (MW = 324), nonpeptidic structure, has already shown several advantages over peptidic inhibitors, including high oral bioavailability (91-99%), a relatively long half-life (4.9 h), and ease of synthesis (three steps).

Discovery of a novel, selective, and orally bioavailable class of thrombin inhibitors incorporating aminopyridyl moieties at the P1 position.

A novel class of thrombin inhibitors incorporating aminopyridyl moieties at the P1 position has been discovered. Four of these thrombin inhibitors (13b,c,e and 14d) showed nanomolar potency (Ki 0.8-12 nM), 300-1500-fold selectivity for thrombin compared with trypsin, and good oral bioavailability (F = 40-76%) in rats or dogs. The neutral P1 was expected to increase metabolic stability and oral absorption. Identification of this novel aminopyridyl group at P1 was a key step in our search for a clinical candidate.

Structure of the lysine-fibrin binding subsite of human plasminogen kringle 4.

Human plasminogen kringle 4, which crystallizes in the orthorhombic system a = 32.15(2), b = 49.01(2), c = 49.04(3) A, space group P2(1)2(1)2(1), four molecules per unit cell, protein volume fraction 0.62, has been determined at 1.9 A resolution. The structure was solved by rotation-translation methods using the structure of bovine prothrombin kringle 1 as a model and it has been refined at 1.9 A resolution to an R-value of 0.142. The root mean square (rms) deviation between the main-chain atoms of the two kringles is about 0.5 A while that between 31 conserved side chains is a surprisingly large 1.2 A. The structure of the lysine binding subsite of fibrin binding of kringle 4 is approximated well by prothrombin kringle 1 but with some notable exceptions. The latter transform the site from a non-binding kringle to one which recognizes lysine and other omega-amino-carboxylic acids. The binding site of the observed kringle 4 structure is also compared with one that was modelled from the structure of kringle 1 of prothrombin fragment 1 and NMR observations. Arginine residues of the binding site of a neighbouring molecule make ion pairs with aspartic acid residues in the binding site of another molecule in the kringle 4 structure.

Crystal and molecular structure of human plasminogen kringle 4 refined at 1.9-A resolution.

The crystal structure of human plasminogen kringle 4 (PGK4) has been solved by molecular replacement using the bovine prothrombin kringle 1 (PTK1) structure as a model and refined by restrained least-squares methods to an R factor of 14.2% at 1.9-A resolution. The K4 structure is similar to that of PTK1, and an insertion of one residue at position 59 of the latter has minimal effect on the protein folding. The PGK4 structure is highly stabilized by an internal hydrophobic core and an extensive hydrogen-bonding network. Features new to this kringle include a cis peptide bond at Pro30 and the presence of two alternate, perpendicular, and equally occupied orientations for the Cys75 side chain. The K4 lysine-binding site consists of a hydrophobic trough formed by the Trp62 and Trp72 indole rings, with anionic (Asp55/Asp57) and cationic (Lys35/Arg71) charge pairs at either end. With the adjacent Asp5 and Arg32 residues, these result in triply charged anionic and cationic clusters (pH of crystals at 6.0), which, in addition to the unusually high accessibility of the Trp72 side chain, serve as an obvious marker of the binding site on the K4 surface. A complex intermolecular interaction occurs between the binding sites of symmetry-related molecules involving a highly ordered sulfate anion of solvation in which the Arg32 side chain of a neighboring kringle occupies the binding site.

The refined structure of the epsilon-aminocaproic acid complex of human plasminogen kringle 4.

The crystallographic structure of the plasminogen kringle 4-epsilon-aminocaproic acid (ACA) complex (K4-ACA) has been solved by molecular replacement rotation-translation methods utilizing the refined apo-K4 structure as a search model (Mulichak et al., 1991), and it has been refined to an R value of 0.148 at 2.25-A resolution. The K4-ACA structure consists of two interkringle residues, the kringle along with the ACA ligand, and 106 water molecules. The lysine-binding site has been confirmed to be a relatively open and shallow depression, lined by aromatic rings of Trp62, Phe64, and Trp72, which provide a highly nonpolar environment between doubly charged anionic and cationic centers formed by Asp55/Asp57 and Lys35/Arg71. A zwitterionic ACA ligand molecule is held by hydrogen-bonded ion pair interactions and van der Waals contacts between the charged centers. The lysine-binding site of apo-K4 and K4-ACA have been compared: the rms differences in main-chain and side-chain positions are 0.25 and 0.69 A, respectively, both practically within error of the determinations. The largest deviations in the binding site are due to different crystal packing interactions. Thus, the lysine-binding site appears to be preformed, and lysine binding does not require conformational changes of the host. The results of NMR studies of lysine binding with K4 are correlated with the structure of K4-ACA and agree well.

Crystal structure of human alpha-thrombin complexed with hirugen and p-amidinophenylpyruvate at 1.6 A resolution.

Crystals of human alpha-thrombin complexed with hirugen and the alpha-keto acid thrombin inhibitor APPA (p-amidinophenylpyruvate) that diffract to 1.6 A resolution were obtained by soaking an alpha-thrombin-hirugen crystal in a solution of APPA. The crystal structure was determined using the difference Fourier method and refined to an R factor of 18.7% at 1.6 A resolution. This structure is the highest resolution structure of the thrombin molecule that is currently available. With the exception of the region near Arg77A-Asn78, the structures of the thrombin and hirugen molecules in the ternary complex are similar to those reported for the thrombin-hirugen binary complex. As previously determined for the APPA-trypsin complex, the carbonyl carbon atom of APPA forms a covalent bond with O gamma of Ser195 of thrombin to yield a "transition-state" analog of the tetrahedral intermediate. Comparison of the specificity pocket of the APPA complexes of thrombin and trypsin reveals differences in hydrogen bonding and shows for the first time that the S1 site of thrombin is larger than that of trypsin and as a result thrombin may be able to accommodate a bulkier P1 group than trypsin.

The structure of mammalian hexokinase-1.

We have determined the structures of the glucose-6-phosphate (G6P)-inhibitable 100,000 Mr Type I hexokinase from rat and the G6P-sensitive 50,000 Mr hexokinase from Schistosoma mansoni at a resolution of 2.8 and 2.6 A respectively. The structures define the glucose and G6P binding sites in these enzymes, suggest the mechanisms of intradomain G6P inhibition and activity loss in the Type I hexokinase N-terminal half, and reveal the structure of the membrane targeting motif that integrates the Type I hexokinase into the outer mitochondrial membrane.

The role of arginine 120 of human prostaglandin endoperoxide H synthase-2 in the interaction with fatty acid substrates and inhibitors.

Arg-120 is located near the mouth of the hydrophobic channel that forms the cyclooxygenase active site of prostaglandin endoperoxide H synthases (PGHSs)-1 and -2. Replacement of Arg-120 of ovine PGHS-1 with a glutamine increases the apparent Km of PGHS-1 for arachidonate by 1,000-fold (Bhattacharyya, D. K., Lecomte, M., Rieke, C. J., Garavito, R. M., and Smith, W. L. (1996) J. Biol. Chem. 271, 2179-2184). This and other evidence indicate that the guanido group of Arg-120 forms an ionic bond with the carboxylate group of arachidonate and that this interaction is an important contributor to the overall strength of arachidonate binding to PGHS-1. In contrast, we report here that R120Q human PGHS-2 (hPGHS-2) and native hPGHS-2 have very similar kinetic properties, but R120L hPGHS-2 catalyzes the oxygenation of arachidonate inefficiently. Our data indicate that the guanido group of Arg-120 of hPGHS-2 interacts with arachidonate through a hydrogen bond rather than an ionic bond and that this interaction is much less important for arachidonate binding to PGHS-2 than to PGHS-1. The Km values of PGHS-1 and -2 for arachidonate are the same, and all but one of the core residues of the active sites of the two isozymes are identical. Thus, the results of our studies of Arg-120 of PGHS-1 and -2 imply that interactions involved in the binding of arachidonate to PGHS-1 and -2 are quite different and that residues within the hydrophobic cyclooxygenase channel must contribute more significantly to arachidonate binding to PGHS-2 than to PGHS-1. As observed previously with R120Q PGHS-1, flurbiprofen was an ineffective inhibitor of R120Q hPGHS-2. PGHS-2-specific inhibitors including NS398, DuP-697, and SC58125 had IC50 values for the R120Q mutant that were up to 1,000-fold less than those observed for native hPGHS-2; thus, the positively charged guanido group of Arg-120 interferes with the binding of these compounds. NS398 did not cause time-dependent inhibition of R120Q hPGHS-2, whereas DuP-697 and SC58125 were time-dependent inhibitors. Thus, Arg-120 is important for the time-dependent inhibition of hPGHS-2 by NS398 but not by DuP-697 or SC58125.

Crystal structure of SQD1, an enzyme involved in the biosynthesis of the plant sulfolipid headgroup donor UDP-sulfoquinovose.

The SQD1 enzyme is believed to be involved in the biosynthesis of the sulfoquinovosyl headgroup of plant sulfolipids, catalyzing the transfer of SO(3)(-) to UDP-glucose. We have determined the structure of the complex of SQD1 from Arabidopsis thaliana with NAD(+) and the putative substrate UDP-glucose at 1.6-A resolution. Both bound ligands are completely buried within the binding cleft, along with an internal solvent cavity which is the likely binding site for the, as yet, unidentified sulfur-donor substrate. SQD1 is a member of the short-chain dehydrogenase/reductase (SDR) family of enzymes, and its structure shows a conservation of the SDR catalytic residues. Among several highly conserved catalytic residues, Thr-145 forms unusually short hydrogen bonds with both susceptible hydroxyls of UDP-glucose. A His side chain may also be catalytically important in the sulfonation.

Crystallization and preliminary diffraction data of Escherichia coli ADP glucose pyrophosphorylase.

ADP glucose pyrophosphorylase from Escherichia coli has been crystallized from polyethylene glycol 8000 solutions. The crystals are: orthorhombic, a = 155(2), b = 153(2), c = 174(2) A, space group P2(1)2(1)2(1), four tetrameric molecules/unit cell. This gives a solvent fraction of about 75% consistent with the relatively poor diffraction quality of crystals (5.0-A resolution) and their sensitivity to x-ray exposure damage. Ways of circumventing the former and improving the latter are proposed.

Human plasminogen kringle 4. Crystallization and preliminary diffraction data of two different crystal forms.

Human plasminogen kringle 4 has been crystallized in two different crystal forms: monoclinic, a = 32.78(3), b = 49.17(2), c = 46.27(3) A, beta = 100.67 degrees, space group P2(1), four molecules/unit cell, two molecules/asymmetric unit; orthorhombic, a = 32.09(7), b = 49.14(6), c = 49.47(9) A, space group P2(1)2(1)2, four molecules/unit cell. Both crystal forms have a large protein fraction (66% for monoclinic and 62% for orthorhombic) and diffract x-rays to 2.0 A resolution. A self-rotation function has been calculated with monoclinic data indicating a non-crystallographic 2-fold rotation approximately parallel to a* (peak height of 14.3 x sigma). Cross-rotation function calculations are in progress utilizing the coordinates of the conserved structure of kringle 1 of prothrombin and plasminogen kringle 4.

Peroxidase activity in prostaglandin endoperoxide H synthase-1 occurs with a neutral histidine proximal heme ligand.

Prostaglandin endoperoxide H synthases-1 and -2 (PGHS-1 and -2) convert arachidonic acid to prostaglandin H(2) (PGH(2)), the committed step in prostaglandin and thromboxane formation. Interaction of peroxides with the heme sites in PGHSs generates a tyrosyl radical that catalyzes subsequent cyclooxygenase chemistry. To study the peroxidase reaction of ovine oPGHS-1, we combined spectroscopic and directed mutagenesis data with X-ray crystallographic refinement of the heme site. Optical and Raman spectroscopy of oxidized oPGHS-1 indicate that its heme iron (Fe(3+)) exists exclusively as a high-spin, six-coordinate species in the holoenzyme and in heme-reconstituted apoenzyme. The sixth ligand is most likely water. The cyanide complex of oxidized oPGHS-1 has a six-coordinate, low-spin ferric iron with a v[Fe-CN] frequency at 445 cm(-)(1); a monotonic sensitivity to cyanide isotopomers that indicates the Fe-CN adduct has a linear geometry. The ferrous iron in reduced oPGHS-1 adopts a high-spin, five-coordinate state that is converted to a six-coordinate, low-spin geometry by CO. The low-frequency Raman spectrum of reduced oPGHS-1 reveals two v[Fe-His] frequencies at 206 and 222 cm(-)(1). These vibrations, which disappear upon addition of CO, are consistent with a neutral histidine (His388) as the proximal heme ligand. The refined crystal structure shows that there is a water molecule located between His388 and Tyr504 that can hydrogen bond to both residues. However, substitution of Tyr504 with alanine yields a mutant having 46% of the peroxidase activity of native oPGHS-1, establishing that bonding of Tyr504 to this water is not critical for catalysis. Collectively, our results show that the proximal histidine ligand in oPGHS-1 is electrostatically neutral. Thus, in contrast to most other peroxidases, a strongly basic proximal ligand is not necessary for peroxidase catalysis by oPGHS-1.

Mutational and X-ray crystallographic analysis of the interaction of dihomo-gamma -linolenic acid with prostaglandin endoperoxide H synthases.

Prostaglandin endoperoxide H synthases-1 and -2 (PGHSs) catalyze the committed step in prostaglandin biosynthesis. Both isozymes can oxygenate a variety of related polyunsaturated fatty acids. We report here the x-ray crystal structure of dihomo-gamma-linolenic acid (DHLA) in the cyclooxygenase site of PGHS-1 and the effects of active site substitutions on the oxygenation of DHLA, and we compare these results to those obtained previously with arachidonic acid (AA). DHLA is bound within the cyclooxygenase site in the same overall L-shaped conformation as AA. C-1 and C-11 through C-20 are in the same positions for both substrates, but the positions of C-2 through C-10 differ by up to 1.74 A. In general, substitutions of active site residues caused parallel changes in the oxygenation of both AA and DHLA. Two significant exceptions were Val-349 and Ser-530. A V349A substitution caused an 800-fold decrease in the V(max)/K(m) for DHLA but less than a 2-fold change with AA; kinetic evidence indicates that C-13 of DHLA is improperly positioned with respect to Tyr-385 in the V349A mutant thereby preventing efficient hydrogen abstraction. Val-349 contacts C-5 of DHLA and appears to serve as a structural bumper positioning the carboxyl half of DHLA, which, in turn, positions properly the omega-half of this substrate. A V349A substitution in PGHS-2 has similar, minor effects on the rates of oxygenation of AA and DHLA. Thus, Val-349 is a major determinant of substrate specificity for PGHS-1 but not for PGHS-2. Ser-530 also influences the substrate specificity of PGHS-1; an S530T substitution causes 40- and 750-fold decreases in oxygenation efficiencies for AA and DHLA, respectively.

Identification and SAR for a selective, nonpeptidyl thrombin inhibitor.

A novel, nonpeptidyl thrombin inhibitor, L-636,619 (1), was identified via topological similarity searching over the Merck Corporate Sample Database. X-ray crystallographic studies determined the geometry for ligand binding to the enzyme. Chemical modification of the P1 and P3 segments of the ligand resulted in enhanced potency and improvement in the chemical stability of the lead. Analog 9 proved to be the most interesting lead from this structurally novel series.

The crystallographic structure of the protease from human immunodeficiency virus type 2 with two synthetic peptidic transition state analog inhibitors.

The crystal structure of human immunodeficiency virus (HIV) type 2 protease has been determined in complexes with peptidic inhibitors Noa-His-Cha psi [CH(OH)CH(OH)]Val-Ile-Amp (U75875) and Qnc-Asn-Cha psi [CH(OH)CH2]Val-Npt(U92163) (where Noa is naphthyloxyacetyl, Cha is cyclohexylalanine, Amp is 2-aminomethylpyridine, Qnc is quinoline-2-carbonyl, and Npt is neopentylamine), which have dihydroxyethylene and hydroxyethylene moieties, respectively, in place of the normal scissile bond of the natural ligand. The complexes crystallize in space group P2(1)2(1)2(1), with one dimer-inhibitor complex per asymmetric unit and average cell dimensions of a = 33.28 A, b = 45.35 A, c = 135.84 A. Data were collected to approximately 2.5-A resolution. The model structures were refined with resulting R-factors of around 0.19. As expected, the HIV-2 protease structure is approximately C2-symmetric with a gross structure very similar to that of the HIV-1 enzyme. The inhibitors bind in an extended conformation positioned lengthwise in the binding cleft in a manner similar to that found in the HIV-1 protease-inhibitor complexes previously reported. The substitution of the bulkier Ile82 side chain in the HIV-2 protease may help explain the better ability of HIV-2 protease to bind and hydrolyze ligands with small P1 and P1' side groups. It appears that differences in specificity between the proteases of HIV-1 and HIV-2 are not merely a result of simple side chain substitutions, but may be complicated by differences in main chain flexibility as well.

Prostaglandin endoperoxide H synthase-1: the functions of cyclooxygenase active site residues in the binding, positioning, and oxygenation of arachidonic acid.

Prostaglandin endoperoxide H synthases (PGHSs) catalyze the committed step in the biosynthesis of prostaglandins and thromboxane, the conversion of arachidonic acid, two molecules of O(2), and two electrons to prostaglandin endoperoxide H(2) (PGH(2)). Formation of PGH(2) involves an initial oxygenation of arachidonate to yield PGG(2) catalyzed by the cyclooxygenase activity of the enzyme and then a reduction of the 15-hydroperoxyl group of PGG(2) to form PGH(2) catalyzed by the peroxidase activity. The cyclooxygenase active site is a hydrophobic channel that protrudes from the membrane binding domain into the core of the globular domain of PGHS. In the crystal structure of Co(3+)-heme ovine PGHS-1 complexed with arachidonic acid, 19 cyclooxygenase active site residues are predicted to make a total of 50 contacts with the substrate (Malkowski, M. G, Ginell, S., Smith, W. L., and Garavito, R. M. (2000) Science 289, 1933-1937); two of these are hydrophilic, and 48 involve hydrophobic interactions. We performed mutational analyses to determine the roles of 14 of these residues and 4 other closely neighboring residues in arachidonate binding and oxygenation. Mutants were analyzed for peroxidase and cyclooxygenase activity, and the products formed by various mutants were characterized. Overall, the results indicate that cyclooxygenase active site residues of PGHS-1 fall into five functional categories as follows: (a) residues directly involved in hydrogen abstraction from C-13 of arachidonate (Tyr-385); (b) residues essential for positioning C-13 of arachidonate for hydrogen abstraction (Gly-533 and Tyr-348); (c) residues critical for high affinity arachidonate binding (Arg-120); (d) residues critical for positioning arachidonate in a conformation so that when hydrogen abstraction does occur the molecule is optimally arranged to yield PGG(2) versus monohydroperoxy acid products (Val-349, Trp-387, and Leu-534); and (e) all other active site residues, which individually make less but measurable contributions to optimal catalytic efficiency.