Search

A three-dimensional model for residues 142-427 of the ligand binding domain (LBD) of the human nuclear receptor for 1alpha, 25-dihydroxy-vitamin D(3) [VDR] has been generated based on the X-ray crystallographic atomic coordinates of the LBD of the rat alpha1 thyroid receptor (TR). The VDR LBD model is an elongated globular shape comprised of an antiparallel alpha-helical triple sandwich topology, made up of 12 alpha-helical elements linked by short loop structures; collectively these structural features are similar to the characteristic secondary and tertiary structures for six nuclear receptors with known X-ray structures. The model has been used to describe the interaction of the conformationally flexible natural hormone, 1alpha,25-dihydroxy-vitamin D(3) [1alpha, 25(OH)(2)D(3)], and a number of related analogs with the VDR LBD. The optimal orientation of the 1alpha,25(OH)(2)D(3) in the LBD is with its A-ring directed towards the interior and its flexible side chain pointing towards and interacting with helix-12, site of the activation function-2 domain (AF-2) of the VDR. Mapping of four natural and one experimental point mutations of the VDR LBD, which result in ligand-related receptor dysfunction, indicates the close proximity of these amino acids to the bound ligand., (Copyright 1999 Wiley-Liss, Inc.)

Clathrin is a triskelion-shaped cytoplasmic protein that polymerizes into a polyhedral lattice on intracellular membranes to form protein-coated membrane vesicles. Lattice formation induces the sorting of membrane proteins during endocytosis and organelle biogenesis by interacting with membrane-associated adaptor molecules. The clathrin triskelion is a trimer of heavy-chain subunits (1,675 residues), each binding a single light-chain subunit, in the hub domain (residues 1,074-1,675). Light chains negatively modulate polymerization so that intracellular clathrin assembly is adaptor-dependent. Here we report the atomic structure, to 2.6 A resolution, of hub residues 1,210-1,516 involved in mediating spontaneous clathrin heavy-chain polymerization and light-chain association. The hub fragment folds into an elongated coil of alpha-helices, and alignment analyses reveal a 145-residue motif that is repeated seven times along the filamentous leg and appears in other proteins involved in vacuolar protein sorting. The resulting model provides a three-dimensional framework for understanding clathrin heavy-chain self-assembly, light-chain binding and trimerization.

The determination of several structures of nuclear receptor ligand binding domains (LBD) has led to new insights into the mechanism of action of this very important class of receptors. This review describes and compares the different LBD structures and their relationship to the function of the nuclear receptors. The role of the ligand in the LBD structures and the implications of ligand structure on receptor activity are also discussed. Structural information regarding interactions between the LBD and coactivator proteins and the potential role of these interactions in ligand agonism and antagonism is reviewed. Different pathways for nuclear receptor signaling and the use of new ligands to investigate these pathways are also described.

Neurotrophin-3 (NT-3) is a cystine knot growth factor that promotes the survival, proliferation, and differentiation of developing neurons and is a potential therapeutic for neurodegenerative diseases. To clarify the structural basis of receptor specificity and the role of neurotrophin dimerization in receptor activation, the structure of the NT-3 homodimer was determined using X-ray crystallography. The orthorhombic crystals diffract to 2.4 A, with dimer symmetry occurring about a crystallographic 2-fold axis. The overall structure of NT-3 resembles that of the other neurotrophins, NGF and BDNF; each protomer forms a twisted four-stranded beta sheet, with three intertwined disulfide bonds. There are notable differences, however, between NT-3 and NGF in the surface loops and in three functionally important regions, shown in previous mutagenesis studies to be critical for binding. One such difference implies that NT-3's binding affinity and specificity depend on a novel hydrogen bond between Gln 83, a residue important for binding specificity with TrkC, and Arg 103, a residue crucial for binding affinity with TrkC. NT-3's extensive dimer interface buries much of the otherwise solvent-accessible hydrophobic surface area and suggests that the dimeric state is stabilized through the formation of this hydrophobic core. A comparison of the dimer interface between the NT-3 homodimer and the BDNF/NT-3 heterodimer reveals similar patterns of hydrogen bonds and nonpolar contacts, which reinforces the notion that the evolutionarily conserved neurotrophin interface resulted from the need for receptor dimerization in signal initiation.

1. Thyroid hormone receptors (TR) are expressed from two separate genes (alpha and beta) and belong to the nuclear receptor superfamily, which also contains receptors for steroids, vitamins and prostaglandins. 2. Unliganded TR are bound to DNA thyroid hormone response elements (TRE) predominantly as homodimers, or as heterodimers with retinoid X-receptors (RXR), and are associated with a complex of proteins containing corepressor proteins. Ligand binding promotes corepressor dissociation and binding of a coactivator. 3. Recent studies from our group have focused on the acquisition and use of X-ray crystallographic structures of ligand-binding domains (LBD) of both the rat (r) TR alpha and the human (h) TR beta bound to several different ligands. We have also developed ligands that bind selectively to the TR beta, which may provide ways to explore the differential functions of TR alpha compared with TR beta isoforms. 4. The LBD is comprised mostly of alpha-helices. The ligand is completely buried in the receptor and forms part of its hydrophobic core. Kinetic studies suggest that the limiting step in formation of high-affinity ligand-receptor complexes is the rate of folding of the receptor around the ligand. Ligands can be fitted tightly in the ligand-binding pocket and small differences in this fitting may explain many structure-activity relationships. Interestingly, analysis of the structures of antagonists suggests that they have chemical groups, 'extensions', that could impair receptor folding around them and, thus, prevent the agonist-induced conformation changes in the receptor. 5. The TR structures allowed us to see that the mutations that occur in the syndrome of generalized resistance to thyroid hormone are located in the vicinity of the ligand-binding pocket. 6. X-ray structure of the TR has also been used to guide construction of mutations in the TR surface that block binding of various proteins important for receptor function. Studies with these TR mutants reveal that the interfaces for homo- and heterodimerization map to similar residues in helix 10 and 11 and also allow the definition of the surface for binding of coactivators, which appears to be general for nuclear receptors. Formation of this surface, which involves packing of helix 12 of the TR into a scaffold formed by helices 3 and 5, appears to be the major change in the receptor structure induced by hormone occupancy.

Recent studies have shown surprising structural and functional similarities between the motor domains of kinesin and myosin. Common features have also been described for motor proteins and G proteins. Despite these similarities, the evolutionary relationships between these proteins, even among the motor proteins, has not been obvious, since the topological connectivities of the core overlapping structural elements in these transducing proteins are not identical to one another. Using secondary structure topology, comparison of functional domains and active site chemistry as criteria for relatedness, we propose a set of rules for determining potential evolutionary relationships between proteins showing little or no sequence identity. These rules were used to explore the evolutionary relationship between kinesin and myosin, as well as between motor proteins and other phosphate-loop (P-loop) containing nucleotide-binding proteins. We demonstrate that kinesin and myosin show significant chemical conservations within and outside of the active site, and present an evolutionary scheme that produce their respective topologies from a hypothetical ancestral protein. We also show that, when compared with various other P-loop-containing proteins, the cytoskeletal motors are most similar to G proteins with respect to topology and active site chemistry. We conclude that kinesin and myosin, and possibly G proteins, are probably directly related via divergent evolution from a common core nucleotide-binding motif, and describe the likely topology of this ancestor. These proteins use similar chemical and physical mechanisms to both sense the state of the nucleotide bound in the active site, and then transmit these changes to protein partners. The different topologies can be accounted for by unique genetic insertions that add to the edge of a progenitor protein structure and do not disrupt the hydrophobic core.

The hypoxanthine phosphoribosyltransferase (HPRT) from Trypanosoma cruzi, etiologic agent of Chagas' disease, was cocrystallized with the inosine analogue Formycin B (FmB) and the structure determined to 1.4 A resolution. This is the highest resolution structure yet reported for a phosphoribosyltransferase (PRT), and the asymmetric unit of the crystal contains a dimer of closely associated, nearly identical subunits. A conserved nonproline cis peptide in one active-site loop exposes the main-chain nitrogen to the enzyme active site, while the adjacent lysine side chain interacts with the other subunit of the dimer, thereby providing a possible mechanism for communication between the subunits and their active sites. The three-dimensional coordinates for the invariant Ser103-Tyr104 dipeptide are reported here for the first time. These are the only highly conserved residues in a second active-site loop, termed the long flexible loop, which is predicted to close over the active site of HPRTs to protect a labile transition state [Eads et al. (1994) Cell 78, 325-334]. This structure represents a major step forward in efforts to design/discover potent selective inhibitors of the HPRT of T. cruzi.

Motor proteins of the kinesin superfamily transport intracellular cargo along microtubules. Although different kinesin proteins share 30-50% amino-acid identity in their motor catalytic cores, some move to the plus end of microtubules whereas others travel in the opposite direction. Crystal structures of the catalytic cores of conventional kinesin (a plus-end-directed motor involved in organelle transport) and ncd (a minus-end-directed motor involved in chromosome segregation) are nearly identical; therefore, the structural basis for their opposite directions of movement is unknown. Here we show that the ncd 'neck' made up of 13 class-specific residues next to the superfamily-conserved catalytic core, is essential for minus-end-directed motility, as mutagenesis of these neck residues reverses the direction of ncd motion. By solving the 2.5 A structure of a functional ncd dimer, we show that the ncd neck (a coiled-coil) differs from the corresponding region in the kinesin neck (an interrupted beta-strand), although both necks interact with similar elements in the catalytic cores. The distinct neck architectures also confer different symmetries to the ncd and kinesin dimers and position these motors with appropriate directional bias on the microtubule.

The crystallographic structure of the Fab fragment of the catalytic antibody, 29G11, complexed with an (S)-norleucine phenyl phosphonate transition state analog was determined at 2.2 A resolution. The antibody catalyzes the hydrolysis of norleucine phenyl ester with (S)-enantioselectivity. The shape and charge complementarity of the binding pocket for the hapten account for the preferential binding of the (S)-enantiomer of the substrate. The structure is compared to that of the more catalytically efficient antibody, 17E8, induced by the same hapten transition state analog. 29G11 has different residues from 17E8 at eight positions in the heavy chain, including four substitutions in the hapten-binding pocket: A33V, S95G, S99R and Y100AN, and four substitutions at positions remote from the catalytic site, I28T, R40K, V65G and F91L. The two antibodies show large differences in the orientations of their variable and constant domains, reflected by a 32 degrees difference in their elbow angles. The VL and VH domains in the two antibodies differ by a rotation of 8.8 degrees. The hapten binds in similar orientations and locations in 29G11 and 17E8, which appear to have catalytic groups in common, though the changes in the association of the variable domains affect the precise positioning of residues in the hapten-binding pocket., (Copyright 1998 Academic Press.)

Resistance to thyroid hormone (RTH) has hitherto been associated with thyroid hormone beta receptor (TRbeta) mutations which cluster in two regions (alphaalpha 310-353 and alphaalpha 429-461) of the hormone-binding domain and closely approximate the ligand-binding cavity. Here, we describe a third cluster of RTH mutations extending from alphaalpha 234-282 which constitute a third boundary of the ligand pocket. One mutant, T277A, exhibits impaired transactivation which is disproportionate to its mildly reduced ligand affinity (Ka). T3-dependent recruitment of coactivators (SRC-1, ACTR) by mutant receptor-RXR heterodimers was reduced in comparison with wild-type. Cotransfection of SRC-1 restored transactivation by T277A. In the TRbeta crystal structure this helix 3 residue is surface-exposed and is in close proximity to residues L454 and E457 in helix 12 which are known to be critical for coactivator interaction, suggesting that they all constitute part of a receptor-coactivator interface. The transcriptional function of other mutants (A234T, R243W/Q, A268D, Delta276I, A279V, R282S) in this cluster correlated with their reduced Ka and they inhibited wild-type TRbeta action in a dominant negative manner. DNA binding, heterodimerization and corepressor recruitment were preserved in all mutants, signifying the importance of these attributes for dominant negative activity and correlating with the absence of natural mutations in regions bordering the third cluster which mediate these functions.

The interaction between ecotin and target proteases with trypsin-like specificity has been systematically dissected to understand the structural basis of ecotin's broad inhibitory specificity and the role of the secondary binding site. Site-directed and region-specific mutagenesis were preformed at ecotin's primary site P1 residue (84), the C-terminal dimer interface (133 to 142), and two surface loops of the secondary binding site (67 to 70, 108 to 113). Substitutions at the P1 position resulted in less than fivefold difference in the potency of ecotin binding to rat trypsin, suggesting that the extended binding site is important in binding. A ten amino acid C-terminal truncation variant showed threefold weaker self-association but remained a dimer. The interactions of the secondary binding site of ecotin with bovine trypsin, rat trypsin and human urokinase-type plasminogen activator (uPA) were investigated with alanine substitutions in ecotin at Trp67, Gly68, Tyr69, Asp70, Arg108, Asn110, Lys112 and Leu113, which formed contacts between the inhibitor and protease. By combining these mutations at the secondary binding site with mutations in the primary binding site the molecular recognition between ecotin and its target serine proteases was probed. The contrast in the Ki values of the various ecotin variants towards bovine trypsin, rat trypsin and human uPA established the role of ecotin's secondary binding site in recognizing these homologous serine proteases. Ecotin binds to proteases with a chymotrypsin fold through a combination of primary and secondary site surface loops and is amenable to redesign of its potency and specificity for this class of enzymes., (Copyright 1998 Academic Press Limited.)

The ligand-binding domain of nuclear receptors contains a transcriptional activation function (AF-2) that mediates hormone-dependent binding of coactivator proteins. Scanning surface mutagenesis on the human thyroid hormone receptor was performed to define the site that binds the coactivators, glucocorticoid receptor-interacting protein 1 (GRIP1) and steroid receptor coactivator 1 (SRC-1). The residues involved encircle a small surface that contains a hydrophobic cleft. Ligand activation of transcription involves formation of this surface by folding the carboxyl-terminal alpha helix against a scaffold of three other helices. These features may represent general ones for nuclear receptors.

Resistance to thyroid hormone (RTH) is characterized by elevated serum thyroid hormones, failure to suppress pituitary TSH secretion, and variable T3 responsiveness in peripheral tissues. The disorder is associated with diverse mutations that cluster within three areas of the thyroid hormone beta(TR beta) receptor. Here, we report a novel RTH mutation (R383H), which is located in a region not known to harbor naturally occurring mutations. Although the R383H mutant receptor activated positively regulated genes to an extent comparable to wild-type (WT), negative transcriptional regulation of human TSH alpha and TRH promoters was impaired in either TR beta 1 or TR beta 2 contexts, and WT receptor function was dominantly inhibited. T3-dependent changes in basal transcription with R383H were also impaired: on the TRH promoter, basal activation by unliganded R383H was not reversed by T3 to the same extent as WT; similarly transcriptional silencing by an unliganded Gal4-R383H fusion was not relieved at a T3 concentration that derepressed WT. In keeping with this, ligand-dependent corepressor release by R383H, either in a protein-protein interaction assay or as a DNA-bound heterodimer with retinoid X receptor on either positive or negative thyroid hormone response elements, was disproportionately impaired relative to its ligand-binding affinity, whereas its T3-dependent recruitment of coactivator was unimpaired. These properties were shared by another previously described RTH mutant (R429Q), and in the crystal structure of TR alpha the homologous residues interact in a polar invagination. Our data indicate a role for these residues in mediating negative transcriptional regulation and facilitating corepressor release and suggest that predominant impairment of these functions may be the minimal requirements for causation of RTH.

All parasitic protozoa obtain purine nucleotides solely by salvaging purine bases and/or nucleosides from their host. This observation suggests that inhibiting purine salvage may be a good way of killing these organisms. To explore this idea, we attempted to block the purine salvage pathway of the parasitic protozoan Tritrichomonas foetus. T. foetus is a good organism to study because its purine salvage depends primarily on a single enzyme, hypoxanthine-guanine-xanthine phosphoribosyltransferase (HGXPRTase), and could provide a good model for rational drug design through specific enzyme inhibition. Guided by the crystal structure of T. foetus HGXPRTase, we used structure-based drug design to identify several non-purine compounds that inhibited this enzyme without any detectable effect on human HGPRTase. One of these compounds, 4-[N-(3, 4-dichlorophenyl)carbamoyl]phthalic anhydride (referred to as TF1), was selected for further characterization. TF1 was shown to be a competitive inhibitor of T. foetus HGXPRTase with respect to both guanine (in the forward reaction; Ki = 13 microM) and GMP (in the reverse reaction; Ki = 10 microM), but showed no effect on the homologous human enzyme at concentrations of up to 1 mM. TF1 inhibited the in vitro growth of T. foetus with an EC50 of approximately 40 microM. This inhibitory effect was associated with a decrease in the incorporation of exogenous guanine into nucleic acids, and could be reversed by supplementing the growth medium with excess exogenous hypoxanthine or guanine. Thus, rationally targeting an essential enzyme in a parasitic organism has yielded specific enzyme inhibitors capable of suppressing that parasite's growth.

Phosphotriesterase homology protein (PHP) is a member of a recently discovered family of proteins related to phosphotriesterase, a hydrolytic, bacterial enzyme with an unusual substrate specificity for synthetic organophosphate triesters and phosphorofluoridates, which are common constituents of chemical warfare agents and agricultural pesticides. No natural substrate has been identified for phosphotriesterase, and it has been suggested that the enzyme may have evolved the ability to hydrolyze synthetic compounds in bacteria under selective pressure to meet nutritional needs. PHP, which has 28% sequence identity with phosphotriesterase, may belong to the family of proteins from which phosphotriesterase evolved. Here we report the cloning, expression, initial characterization, and high-resolution X-ray crystallographic structure of PHP. Biochemical analysis shows that PHP is monomeric and binds two zinc ions per monomer. Unlike phosphotriesterase, PHP does not catalyze the hydrolysis of nonspecific phosphotriesters. The structure, similar to that of phosphotriesterase, consists of a long, elliptical alpha/beta barrel and has a binuclear zinc center in a cleft at the carboxy end of the barrel at the location of the presumptive active site.

The DAX1 protein is an orphan nuclear hormone receptor based on sequence similarity in the putative ligand-binding domain (LBD). DAX1 mutations result in X-linked adrenal hypoplasia congenita (AHC). Our objective was to identify DAX1 mutations in a series of families, to determine the types of mutations resulting in AHC and to locate single-amino-acid changes in a DAX1 structural model. The 14 new mutations identified among our 17 families with AHC brought the total number of families with AHC to 48 and the number of reported mutations to 42; 1 family showed gonadal mosaicism. These mutations included 23 frameshift, 12 nonsense, and six missense mutations and one single-codon deletion. We mapped the seven single-amino-acid changes to a homology model constructed by use of the three-dimensional crystal structures of the thyroid-hormone receptor and retinoid X receptor alpha. All single-amino-acid changes mapped to the C-terminal half of the DAX1 protein, in the conserved hydrophobic core of the putative LBD, and none affected residues expected to interact directly with a ligand. We conclude that most genetic alterations in DAX1 are frameshift or nonsense mutations and speculate that the codon deletion and missense mutations give insight into the structure and function of DAX1.

This review summarizes the studies conducted in our laboratory on the mechanisms of thyroid hormone action over the past two decades. We have attempted to place our studies on thyroid hormone receptors (TRs) in perspective with the work conducted by other investigators that established their nuclear localization, DNA-binding properties, DNA response elements, and the role of other proteins involved in TR-mediated regulation of gene transcription. Recently, our crystallographic studies of the TR ligand binding domain (LBD) revealed that the ligand has a structural role in the folding of the receptor's hydrophobic core. The analysis of the structure led to biochemical and genetic studies that have defined the surfaces on the TR LBD required for dimerization and binding of coactivator proteins. Placement of the mutations found in patients with the syndrome of generalized resistance to thyroid hormone on the TR LBD revealed that they were restricted to amino acids in the vicinity of the binding pocket for thyroid hormone. The insights gained from the elucidation of the TR LBD structure will provide the basis for the design of compounds with selective agonistic or antagonistic activities.

Here we report the first structure of a mammalian 15-lipoxygenase. The protein is composed of two domains; a catalytic domain and a previously unrecognized beta-barrel domain. The N-terminal beta-barrel domain has topological and sequence identify to a domain in the mammalian lipases, suggesting that these domains may have similar functions in vivo. Within the C-terminal domain, the lipoxygenase substrate binding site is a hydrophobic pocket defined by a bound inhibitor. Arachidonic acid can be docked into this deep hydrophobic pocket with the methyl end extending down into the bottom of the pocket and the acid end tethered by a conserved basic residue on the surface of the enzyme. This structure provides a unifying hypothesis for the positional specificity of mammalian lipoxygenases.

Background: Glycogen phosphorylases (GPs) catalyze the conversion of the storage form of carbohydrate (glycogen) to the readily usable form (glucose-1-phosphate) to provide cellular energy. Members of this enzyme family have evolved diverse regulatory mechanisms that control a conserved catalytic function. The mammalian and yeast GPs are expressed as inactive forms requiring phosphorylation for activation. Phosphorylation of yeast GP occurs at a distinct site from that of mammalian GP. This work addresses the structural basis by which distinct activation signals relay to the conserved catalytic site in yeast and mammalian GPs. Such knowledge may help understand the principles by which diverse biological regulation evolves., Results: We have compared the crystal structures of the unphosphorylated and phosphorylated forms of yeast GP and propose a relay which links phosphorylation to enzyme activation. Structural components along the activation relay becomes more conserved within the GP family downstream along the relay, towards the catalytic center. Despite distinct upstream activation signals, a response element downstream of the relay leading to the catalytic center is conserved in all GPs. The response element consists of ten hydrophobic residues dispersed over two subunits of the homodimer. Phosphorylation induces hydrophobic condensation of these residues via structural rearrangement, which triggers conformation change of the active site GATE loop, leading to enzyme activation., Conclusions: Members of the GP family with diverse activation mechanisms have evolved from a constitutively active ancestral enzyme which has the TOWER hydrophobic response element in the active position. Diverse regulation evolved as a result of evolutionary constraint on the downstream response element in the active state, coupled with flexibility and variability in elements of the upstream relays.

Inosine-5'-monophosphate dehydrogenase (IMPDH) is an attractive drug target for the control of parasitic infections. The enzyme catalyzes the oxidation of inosine monophosphate (IMP) to xanthosine monophosphate (XMP), the committed step in de novo guanosine monophosphate (GMP) biosynthesis. We have determined the crystal structures of IMPDH from the protozoan parasite Tritrichomonas foetus in the apo form at 2.3 A resolution and the enzyme-XMP complex at 2.6 A resolution. Each monomer of this tetrameric enzyme is comprised of two domains, the largest of which includes an eight-stranded parallel beta/alpha-barrel that contains the enzyme active site at the C termini of the barrel beta-strands. A second domain, comprised of residues 102-220, is disordered in the crystal. IMPDH is expected to be active as a tetramer, since the active site cavity is formed by strands from adjacent subunits. An intrasubunit disulfide bond, seen in the crystal structure, may stabilize the protein in a less active form, as high concentrations of reducing agent have been shown to increase enzyme activity. Disorder at the active site suggests that a high degree of flexibility may be inherent in the catalytic function of IMPDH. Unlike IMPDH from other species, the T. foetus enzyme has a single arginine that is largely responsible for coordinating the substrate phosphate in the active site. This structural uniqueness may facilitate structure-based identification and design of compounds that specifically inhibit the parasite enzyme.

The structure of cruzain, an essential protease from the parasite Trypanosoma cruzi, was determined by X-ray crystallography bound to two different covalent inhibitors. The cruzain S2 specificity pocket is able to productively bind both arginine and phenylalanine residues. The structures of cruzain bound to benzoyl-Arg-Ala-fluoromethyl ketone and benzoyl-Tyr-Ala-fluoromethyl ketone at 2.2 and 2.1 A, respectively, show a pH-dependent specificity switch. Glu 205 adjusts to restructure the S2 specificity pocket, conferring right binding to both hydrophobic and basic residues. Kinetic analysis of activated peptide substrates shows that substrates placing hydrophobic residues in the specificity pocket are cleaved at a broader pH range than hydrophilic substrates. These results demonstrate how cruzain binds both basic and hydrophobic residues and could be important for in vivo regulation of cruzain activity.

Kinesin motors convert chemical energy from ATP hydrolysis into unidirectional movement. To understand how kinesin motors bind to and move along microtubules, we fit the atomic structure of the motor domain of Ncd (a kinesin motor involved in meiosis and mitosis) into three-dimensional density maps of Ncd-microtubule complexes calculated by cryo-electron microscopy and image analysis. The model reveals that Ncd shares an extensive interaction surface with the microtubule, and that a portion of the binding site involves loops that contain conserved residues. In the Ncd dimer, the microtubule-bound motor domain makes intimate contact with its partner head, which is dissociated from the microtubule. This head-head interaction may be important in positioning the dissociated head to take a step to the next binding site on the microtubule protofilament.

Phosphorylation of glycogen phosphorylase at residue Ser14 triggers a conformational transition that activates the enzyme. The N-terminus of the protein, in response to phosphorylation, folds into a 310 helix and moves from its location near a cluster of acidic residues on the protein surface to a site at the dimer interface where a pair of arginine residues form charged hydrogen bonds with the phosphoserine. Site-directed mutagenesis was used to replace Ser14 with Asp and Glu residues, analogs of the phosphoserine, that might be expected to participate in ionic interactions with the arginine side chains at the dimer interface. Kinetic analysis of the mutants indicates that substitution of an acidic residue in place of Ser14 at the site of regulatory phosphorylation partially activates the enzyme. The S14D mutant shows a 1.6-fold increase in Vmax, a 10-fold decrease in the apparent dissociation constant for AMP, and a 3-fold decrease in the S0.5 for glucose 1-phosphate. The S14E mutant behaves similarly, showing a 2.2-fold increase in Vmax, a 6-fold decrease in the apparent dissociation constant for AMP, and a 2-fold decrease in the S0.5 for glucose 1-phosphate. The ability of the mutations to enhance binding of AMP and glucose 1-phosphate and to raise catalytic activity suggests that the introduction of a carboxylate side chain at position 14 promotes docking of the N-terminus at the subunit interface and concomitant stabilization of the activated conformation of the enzyme. Like the native enzyme, both mutants show significant activity only in the presence of the activator, AMP. Full activation, analogous to that provided by covalent phosphorylation of the enzyme, likely is not achieved because of differences in the charge and the geometry of ionic interactions at the phosphorylation site.

The crystal structure of fiddler crab collagenase complexed with the dimeric serine protease inhibitor ecotin at 2.5 A resolution reveals an extended cleft providing binding sites for at least 11 contiguous substrate residues. Comparison of the positions of nine intermolecular main chain hydrogen bonding interactions in the cleft, with the known sequences at the cleavage site of type I collagen, suggests that the protease binding loop of ecotin adopts a conformation mimicking that of the cleaved strand of collagen. A well-defined groove extending across the binding surface of the enzyme readily accommodates the two other polypeptide chains of the triple-helical substrate. These observations permit construction of a detailed molecular model for collagen recognition and cleavage by this invertebrate serine protease. Ecotin undergoes a pronounced internal structural rearrangement which permits binding in the observed conformation. The capacity for such rearrangement appears to be a key determinant of its ability to inhibit a wide range of serine proteases.

Crab collagenolytic serine protease 1 efficiently cleaves peptide bonds directly C-terminal to basic, polar, and hydrophobic amino acids. The crystal structure of this enzyme complexed to the protein inhibitor ecotin at 2.5 A resolution reveals a large primary binding pocket punctuated on one wall by the side chain of aspartate-226. Removal or relocation of this negatively charged group by site-directed mutagenesis generates variant enzymes which retain very high activities toward selected substrates. Full retention of activity toward hydrophobic substrates in collagenase D226G is accompanied by a 10-100-fold reduction in k(cat)/Km toward basic residues. In contrast, restoration of the negative charge in a trypsin-like position in collagenase D226G/G189D regenerates nearly full activity toward basic substrates while introducing a 5-fold decrease in k(cat)/Km toward hydrophobic amino acids. These results imply that the collagenase S1 pocket has multiple distinct binding sites for different amino acid side chains, a suggestion supported by molecular modeling studies based on the crystal structure. The ease of specificity modification in the primary binding site of this serine protease parallels similar observations with the bacterial enzymes alpha-lytic protease and subtilisin, and stands in sharp distinction to the extensive mutagenesis required to alter specificity in trypsin.

Lysine was substituted for a conserved arginine at position 199 of the schistosomal hypoxanthine phosphoribosyltransferase (HPRT). This resulted in a > or = 35-fold increase in the K(M) for binding phosphoribosyl-pyrophosphate (PRPP). The possible functional role of R199 in tertiary structure, as well as in the binding of PRPP, is interpreted in the context of the reported three dimensional structure for the human HPRT.

The kinesin superfamily comprises a large and structurally diverse group of microtubule-based motor proteins that produce a variety of force-generating activities within cells. This review addresses how the structures of kinesin proteins provide clues as to their biological functions and motile properties. We discuss structural features common to all kinesin motors, as well as specialized features that enable subfamilies of related motors to carry out specialized activities. We also discuss how the kinesin motor domain uses chemical energy from ATP hydrolysis to move along microtubules.

A phosphorylation-initiated mechanism of local protein refolding activates yeast glycogen phosphorylase (GP). Refolding of the phosphorylated amino-terminus was shown to create a hydrophobic cluster that wedges into the subunit interface of the enzyme to trigger activation. The phosphorylated threonine is buried in the allosteric site. The mechanism implicates glucose 6-phosphate, the allosteric inhibitor, in facilitating dephosphorylation by dislodging the buried covalent phosphate through binding competition. Thus, protein phosphorylation-dephosphorylation may also be controlled through regulation of the accessibility of the phosphorylation site to kinases and phosphatases. In mammalian glycogen phosphorylase, phosphorylation occurs at a distinct locus. The corresponding allosteric site binds a ligand activator, adenosine monophosphate, which triggers activation by a mechanism analogous to that of phosphorylation in the yeast enzyme.

The functional role in allosteric regulation of a flexible loop (residues 280-288) located near the active site of muscle glycogen phosphorylase was investigated. Mutations were made in residues 283-285 based on crystallographic studies that indicate that the loop functions as a gate controlling access of substrates to the active site and that these specific residues play distinct roles in mimicking the substrate and binding inhibitors when the enzyme is in an inactive conformation. Substitution of Ala or Asn for Asp-283, the putative substrate mimic, results in a 15-fold decrease in Vmax, a 10-fold decrease in the S0.5 for glucose 1-phosphate, a 10-fold increase in the Ka for AMP, and a 10-20-fold increase in the Ki for glucose. Substitution of Ala for Asn-284, which normally forms a hydrogen bond with the inhibitor glucose, reduces Vmax 3-fold, increases the Ki for glucose 2-fold, but has little effect on AMP or glucose 1-phosphate binding or cooperativity. Substitution of Asp at 284, on the other hand, reduces Vmax 10-fold, elevates the Ki for glucose 10-fold, decreases AMP cooperativity, but has little effect on the affinity of AMP or the cooperativity and binding of glucose 1-phosphate. Substitution of Leu for Phe-285, which forms aromatic stacking interactions with purine inhibitors, reduces Vmax 2-fold, decreases the affinity for caffeine at least 10-fold, raises the Ka for AMP 3-fold, and decreases AMP cooperativity but has little effect on glucose 1-phosphate binding or cooperativity. The results of the mutagenesis studies show the importance of the 280's loop for inhibitor binding and modulation of substrate affinity and suggest a role for the loop in allosteric activation. The propagation of allosteric effects across the domain interface may depend upon specific contacts between residues of the 280's loop and the C-terminal domain.

To streamline the preclinical phase of pharmaceutical development, we have explored the utility of structural data on the molecular target and synergy between computational and medicinal chemistry. We have concentrated on parasitic infectious diseases with a particular emphasis on the development of specific noncovalent inhibitors of proteases that play a key role in the parasites' life cycles. Frequently, the structure of the enzyme target of pharmaceutical interest is not available. In this setting we have modeled the structure of the relevant enzyme by virtue of its sequence similarity with proteins of known structure. For example, we have constructed a homology-based model of falcipain, the trophozoite cysteine protease, and used the computational ligand identification algorithm DOCK to identify in compuo enzyme inhibitors including oxalic bis(2-hydroxy-1-naphthyl-methylene)hydrazide (1) [Ring, C. S.; Sun, E.; McKerow, J. H.; Lee, G.; Rosenthal, P. J., Kuntz, I. D.; Cohen, F. E., Proc. Natl Acad. Sci. U.S.A. 1993, 90, 3583]. Compound 1 inhibits falcipain (IC50 6 microM) and the organism in vitro as judged by hypoxanthine uptake (IC50 7 microM). Following this lead, to date, we have identified potent bis arylacylhydrazides (IC50 150 nM) and chalcones (IC50 200 nM) that are active against both chloroquine-sensitive and chloroquine-resistant strains of malaria. In a second example, cruzain, the crystallographically determined structure of a papain-like cysteine protease, resolved to 2.35 A, was available. Aided by DOCK, we have identified a family of bis-arylacylhydrazides that are potent inhibitors of cruzain (IC50 600 microM). These compounds represent useful leads for pharmaceutical development over strict enzyme inhibition criteria in a structure-based design program.

The crystal structure of the hypoxanthine-guanine-xanthine phosphoribosyltransferase (HGXPRTase) from Tritrichomonas foetus has been determined and refined against X-ray data to 1.9 A resolution. T. foetus HGXPRTase crystallizes as an asymmetric dimer, with GMP bound to only one of the two molecules that form the asymmetric unit. Each molecule of HGXPRTase is formed by two lobes joined by a short "hinge" region, and the GMP binds in a cavity between the two lobes. A comparison of the two molecules in the asymmetric unit shows that the hinge region is flexible and that ligand binding affects the relative positions of the two lobes. The binding of GMP brings the two lobes closer together, rotating one lobe by about 5 degrees relative to the other. T. foetus appears to depend on HGXPRTase for its supply of GMP, making this enzyme a target for antiparasite drug design. A comparison of the structures of T. foetus HGXPRTase and human HGPRTase reveals that, while these enzymes retain a similar polypeptide fold, there are substantial differences between the active sites of these two homologs. These differences suggest that it will be possible to find compounds that selectively inhibit the parasite enzyme.

The three-dimensional structures of complexes of trypsin N143H, E151H bound to ecotin A86H are determined at 2.0 A resolution via X-ray crystallography in the absence and presence of the transition metals Zn2+, Ni2+, and Cu2+. The binding site for these transition metals was constructed by substitution of key amino acids with histidine at the trypsin-ecotin interface in the S2'/P2' pocket. Three histidine side chains, two on trypsin at positions 143 and 151 and one on ecotin at position 86, anchor the metals and provide extended catalytic recognition for substrates with His in the P2' pocket. Comparisons of the three-dimensional structures show the different geometries that result upon the binding of metal in the engineered tridentate site and suggest a structural basis for the kinetics of the metal-regulated catalysis. Of the three metals, the binding of zinc results in the most favorable binding geometry, not dissimilar to those observed in naturally occurring zinc binding proteins.

Recognition for proteolysis by trypsin depends almost exclusively on tight binding of arginine or lysine side chains by the primary substrate specificity pocket. Although extended subsite interactions are important for catalysis, the majority of binding energy is localized in the P1 pocket. Analysis of the interactions of trypsin with the P1 residue of the bound inhibitors ecotin and bovine pancreatic trypsin inhibitor suggested that the mutation D189S would improve metal-assisted trypsin N143H, E151H specificity toward peptides that have a Tyr at P1 and a His at P2'. In the presence of transition metals, the catalytic efficiency of the triple mutant Tn N143H, E151H, D189S improved toward the tyrosine-containing peptide AGPYAHSS. Trypsin N143H, E151H, D189S exhibits a 25-fold increase in activity with nickel and a 150-fold increase in activity with zinc relative to trypsin N143H, E151H on this peptide. In addition, activity of trypsin N143H, E151H, D189S toward an arginine-containing peptide, YLVGPRGHFYDA, is enhanced by copper, nickel, and zinc. With this substrate, copper yields a 30-fold, nickel a 70-fold, and zinc a 350-fold increase in activity over background hydrolysis without metal. These results demonstrate that the engineering of multiple substrate binding subsites in trypsin can be used to delocalize protease specificity by increasing relative substrate binding contributions from alternate engineered subsites.

The major, and possible only, component of the infectious prion is the scrapie prion protein (PrPSc); the protease resistant core of PrPSc is PrP 27-30, a protein of approximately 142 amino acids. PrPSc is derived from the cellular PrP isoform (PrPC) by a post-transliatonal process in which a profound conformational change occurs. Syrian hamster (SHa) PrP genes of varying length ranging from the N- and C- terminally truncated 90-228 up to the full-length mature protein 23-231 were inserted into various secretion and intracellular expression vectors that were transformed into Escherichia coli deficient for proteases. Maximum expression was obtained for a truncated SHaPrP containing residues 90-231, which correspond to the sequence of PrP 27-30; disruption of the bacteria using a microfluidizer produced the highest yields of this protein designated rPrP. After solubilization of rPrP in 8 M GdnHC1, it was purified by size exclusion chromatography and reversed phase chromatography. During purification the recovery was approximately 50%, and from each liter of E. coli culture, approximately 50 mg of purified rPrP was obtained. Expression of the longer species containing the basic N-terminal region was less successful and was not pursued further. The primary structure of rPrP was verified by Edman sequencing and mass spectrometry, and secondary structure determined by circular dichroism and Fourier transform infrared spectroscopy. When rPrP was purified under reducing conditions, it had a high beta-sheet content and relatively low solubility similar to PrPSc, particularly at pH values > 7. Refolding of rPrP by oxidation to form a disulfide bond between the two Cys residues of this polypeptide produced a soluble protein with a high alpha-helical content similar to PrPC. These multiple conformations of rPrP are reminiscent of the structural plurality that characterizes the naturally occurring PrP isoforms. The high levels of purified rPrP which can now be obtained should facilitate determination of the multiple tertiary structures that Prp can adopt.

Background: Glycogen phosphorylases consist of a conserved catalytic core onto which different regulatory sites are added. By comparing the structures of isozymes, we hope to understand the structural principles of allosteric regulation in this family of enzymes. Here, we focus on the differences in the glucose 6-phosphate (Glc-6-P) binding sites of two isozymes., Results: We have refined the structure of Glc-6-P inhibited yeast phosphorylase b to 2.6 A and compared it with known structures of muscle phosphorylase. Glc-6-P binds in a novel way, interacting with a distinct set of secondary elements. Structural links connecting the Glc-6-P binding sites and catalytic sites are conserved, although the specific contacts are not., Conclusions: Our comparison reveals that the Glc-6-P binding site was modified over the course of evolution from yeast to vertebrates to become a bi-functional switch. The additional ability of muscle phosphorylase to be activated by AMP required the recruitment of structural elements into the binding site and sequence changes to create a binding subsite for adenine, whilst maintaining links to the catalytic site.

Kinesin is the founding member of a superfamily of microtubule based motor proteins that perform force-generating tasks such as organelle transport and chromosome segregation. It has two identical approximately 960-amino-acid chains containing an amino-terminal globular motor domain, a central alpha-helical region that enables dimer formation through a coiled-coil, and a carboxy-terminal tail domain that binds light chains and possibly an organelle receptor. The kinesin motor domain of approximately 340 amino acids, which can produce movement in vitro, is much smaller than that of myosin (approximately 850 amino acids) and dynein (1,000 amino acids), and is the smallest known molecular motor. Here, we report the crystal structure of the human kinesin motor domain with bound ADP determined to 1.8-A resolution by X-ray crystallography. The motor consists primarily of a single alpha/beta arrowhead-shaped domain with dimensions of 70 x 45 x 45 A. Unexpectedly, it has a striking structural similarity to the core of the catalytic domain of the actin-based motor myosin. Although kinesin and myosin have virtually no amino-acid sequence++ identity, and exhibit distinct enzymatic and motile properties, our results suggest that these two classes of mechanochemical enzymes evolved from a common ancestor and share a similar force-generating strategy.

Microtubule-based ATPases of the kinesin superfamily provide the motile force for many animated features of living cells. Kinesin motors differ in their direction of movement along microtubules. Kinesin and ncd, a kinesin-related motor involved in formation and maintenance of mitotic and meiotic spindles, move in opposite directions along microtubules, even though their motor domains are 40% identical in amino-acid sequence. Here we report the crystal structure of the MgADP complex of the Drosophila ncd motor domain determined to 2.5A by X-ray crystallography, and compare it to the kinesin structure. The ncd and kinesin motor domains are remarkably similar in structure, and the locations of conserved surface amino acids suggest these motors share a common microtubule-binding site. Moreover, structural and functional comparisons of ncd, kinesin, myosin and G proteins indicate that these NTPases may have a similar strategy of changing conformation between NTP and NDP states. We propose a general model for converting a common gamma-phosphate-sensing mechanism into opposite polarities of movement for kinesin and ncd.

The crystal structure of the rat alpha 1 thyroid hormone receptor ligand-binding domain bound with a thyroid hormone agonist reveals that ligand is completely buried within the domain as part of the hydrophobic core. In addition, the carboxy-terminal activation domain forms an amphipathic helix, with its hydrophobic face constituting part of the hormone binding cavity. These observations suggest a structural role for ligand, in establishing the active conformation of the receptor, that is likely to underlie hormonal regulation of gene expression for the nuclear receptors.

ncd is a kinesin-related motor protein from Drosophila that moves in the opposite direction along microtubules to kinesin. To learn more about the ncd mechanism, ncd motor domain (R335-K700) was expressed in Escherichia coli and its enzymatic characteristics were studied. The ncd motor domain was purified from the cell lysate by S-Sepharose chromatography, and trace amounts of contaminants were removed by passing through a MonoQ column. The yield was 20 mg from a 500 mL culture of E. coli. The purified ncd motor domain exhibited an unusual UV spectrum with a broad peak around 272-275 nm, which was at least partly due to the bound nucleotide. Upon incubation with radioactive ATP, 3H at adenine but not 32P at gamma-phosphate was retained by the protein on gel filtration, indicating it bound ADP but not ATP. Thus, like kinesin, nucleotide binding to the ncd motor domain is tight, although there is an equilibrium between the protein and free nucleotide. We also used a fluorescent ATP analogue, mantATP, for the kinetic study of ncd motor domain. MantATP was turned over by ncd motor domain slowly in the absence of microtubules, but microtubules activated the turnover to a similar extent to that of ATP. Upon incubation with ncd motor domain, the fluorescent intensity of mantATP increased at 0.005 s-1, which is likely to reflect the release of endogenous ADP and incorporation of mantATP into the protein. The fluorescence intensity of the ncd motor domain having bound mantADP, likewise, decreased upon mixing with ATP, representing the mantADP release.(ABSTRACT TRUNCATED AT 250 WORDS)

Certain neurodegenerative diseases in humans and animals are caused by small proteinaceous infectious particles called prions. Limited proteolysis and detergent extraction of the prions containing PrPSc generate prion rods that are composed of a polypeptide having an apparent molecular mass of 27 to 30 kDa. This polypeptide, termed prion protein PrP 27-30, has a ragged N terminus that begins at about residue 90, but retains scrapie infectivity. Moreover, the findings in a patient having an inherited prion disease of a truncated PrP with its C terminus at residue 145 suggest that the residues 90 to 145 may be of particular importance in the pathogenesis of prion diseases. To determine the three-dimensional organization of prion rods and to identify the core region involved in amyloid formation, we recorded X-ray diffraction patterns from rods purified from scrapie-infected Syrian hamster (SHa) brains which contain PrP 27-30, and from synthetic SHaPrP peptides. Three peptides were studied corresponding to residues 113 to 120 (peptide A8A, an octamer composed of glycines and alanines), 109 to 122 (H1, the first predicted alpha-helical region of PrPC), and 90 to 145 (a 56 residue peptide containing both H1 and the second predicted alpha-helical region of PrPC, H2). Electron microscopy, carried out in parallel with the X-ray measurements, revealed that all the samples formed linear polymers which were approximately 60 to approximately 200 A wide, with fibrillar or ribbon-like morphology. Gels and dried preparations of prion rods gave X-ray patterns that indicated a beta-sheet conformation, in which the hydrogen bond distance was 4.72 A and the intersheet distance was 8.82 A. For the three PrP peptides, the intersheet spacings varied widely, owing to the side-chains of the residues involved in the formation of the beta-sheet interactions, i.e., 5.13 A for A8A, 5.91 A for lyophilized H1, 7.99 A from solubilized and dried H1 and 9.15 A for the peptide SHa 90-145. The intersheet distance of PrP 27-30 was thus within the observed range for the peptides, and suggests that the amyloidogenic core of PrP is closely modeled by the peptide SHa 90-145.

The kinesin superfamily is a class of microtubule-based mechano-enzymes involved in intracellular transport and chromosome movements. Molecules that move towards either the plus end or the minus end of microtubules are represented within the family. The motor domains of these molecules exhibit considerable sequence homology and contain both the ATP- and microtubule-binding sites (reviewed in refs 1, 2). Here we focus on non-claret disjunctional (ncd), a minus-end-directed motor involved in chromosome segregation in meiosis and early mitosis in Drosophila. We have calculated a three-dimensional map of tubulin sheets decorated with monomeric recombinant ncd motor domains by negative-stain electron microscopy and image analysis. Comparisons with a control structure of tubulin alone reveal that each motor domain binds to the crest of a single protofilament, making extensive contacts with both the alpha and beta tubulin monomers. Binding of the motor domain results in significant conformational changes in both of the tubulin monomers.

Allosteric switching between inactive and active conformational states in muscle glycogen phosphorylase alters the pattern of van der Waals contacts and hydrogen bonds between two helices located at the dimer interface of the enzyme. Alanine was substituted for residues N270, N274, and R277 to perturb helix interactions, which differ in inactive and active conformations. In addition, the entire alpha-helix in each subunit was exchanged with the analogous region from yeast phosphorylase. The N274A mutant shows increased affinity and reduced cooperativity for the activator, AMP, and reduced cooperativity for the substrate, glucose 1-phosphate. The N270A and R277A mutants, in contrast, show reduced binding and cooperativity for AMP and relatively little change in binding or cooperativity for glucose 1-phosphate. The substitution of the helix from the yeast enzyme results in an 8-fold reduction in Vmax, a loss in cooperativity for both AMP and glucose 1-phosphate, but little change in the affinities of either ligand. Crystallographic analyses of the N274A and R277A mutants show that these substitutions cause only small changes in the structure of the unliganded, inactive form of phosphorylase. The substitution at N274 eliminates intersubunit interactions which selectively stabilize the enzyme in an inactive conformation. The kinetic results indicate that the mutations at N270 and R277, on the contrary, perturb packing interactions at the dimer interface of the activated enzyme and weaken binding of AMP. The relatively modest effects of the replacement with the helices from the yeast enzyme indicate that the helices are not crucial for catalytic function.

Thioesterase II from rat mammary gland has been crystallized in the presence of decanoic acid by the vapor diffusion method. The crystals belong to the orthorhombic space group P2(1)2(1)2(1), and have cell dimensions, a = 52.7 A, b = 78.0 A, and c = 133.6 A. The asymmetric unit likely consists of two protein monomers based on predictions from its calculated Matthews coefficient. Crystals typically diffract to at least 2.5 A resolution and are suitable for X-ray crystallographic analysis.

Trypanosoma cruzi, a protozoan parasite, is the etiologic agent of American trypanosomiasis or Chagas' disease. Chagas' disease afflicts more than 24 million individuals in South and Central America producing a debilitating life-long disease. It is the leading cause of heart failure in many Latin American countries. Currently, there is no satisfactory treatment for this parasitic infection. Cruzain (also known as cruzipain, gp 57/51), the major cysteine protease present in T. cruzi, is critical for the development and survival of the parasite within the host cells, making this enzyme a target for potential trypanocidal drugs. Here we report the X-ray crystal structure of cruzain complexed with the potent inhibitor Z-Phe-Ala-fluoromethyl ketone. The structure was determined at 2.35 A (Rcryst = 0.15) by molecular replacement using a modified papain as the search model. The refined structure is compared to papain. Features which distinguish cruzain from papain are discussed since they may aid in the design of specificity inhibitors. Fluorescence microscopy shows that a biotinylated form of the bound inhibitor does not effectively reach host proteases in their lysosomal compartment, but is selectively taken up by the parasite. The inhibitor greatly reduces parasitemia in a cell culture system, without adverse effects to mammalian cells. This biological selectivity can be exploited, in conjunction with unique active site features revealed by the crystal structure, to develop chemotherapy for Chagas' disease.

Histidine substrate specificity has been engineered into trypsin by creating metal binding sites for Ni2+ and Zn2+ ions. The sites bridge the substrate and enzyme on the leaving-group side of the scissile bond. Application of simple steric and geometric criteria to a crystallographically derived enzyme-substrate model suggested that histidine specificity at the P2' position might be achieved by a tridentate site involving amino acid residues 143 and 151 of trypsin. Trypsin N143H/E151H hydrolyzes a P2'-His-containing peptide (AGPYAHSS) exclusively in the presence of nickel or zinc with a high level of catalytic efficiency. Since cleavage following the tyrosine residue is normally highly disfavored by trypsin, this result demonstrates that a metal cofactor can be used to modulate specificity in a designed fashion. The same geometric criteria applied in the primary S1 binding pocket suggested that the single-site mutation D189H might effect metal-dependent His specificity in trypsin. However, kinetic and crystallographic analysis of this variant showed that the design was unsuccessful because His189 rotates away from substrate causing a large perturbation in adjacent surface loops. This observation suggests that the reason specificity modification at the trypsin S1 site requires extensive mutagenesis is because the pocket cannot deform locally to accommodate alternate P1 side chains. By taking advantage of the extended subsites, an alternate substrate specificity has been engineered into trypsin.

Converting the specificity of trypsin to that of chymotrypsin has been shown to require the exchange of amino acids in multiple portions of the protein, including two surface loops which do not directly contact the substrate. Crystallographic analysis of two mutant trypsins possessing chymotrypsin-like specificity now reveals that these distal surface loops alter function by directly determining the structure of the primary binding site. Efficient acylation of cognate substrates correlates with a distinct backbone conformation of the conserved Gly216 residue. This amino acid is located on the surface of the specificity pocket and forms two main-chain hydrogen bonds with a nonspecific portion of substrate. By contrast, the improvement in substrate binding affinity effect by the substitution of the distal Tyr172 residue with Trp derives from structural rearrangements at the extreme base of the pocket. Together, the kinetic and crystallographic data strongly suggest that both Asp189 and Gly216 must be considered as primary determinants of substrate specificity in trypsin.

Ecotin, an Escherichia coli periplasmic protein of 142 amino acids, has been shown to be a potent inhibitor of a group of homologous serine proteases with widely differing substrate recognition. It is highly effective against a number of enzymes, including both pancreatic and neutrophil-derived elastases, chymotrypsin, trypsin, factor Xa, and kallikrein. Recent structural and functional studies on ecotin and its interactions with different serine proteases have clarified these initial observations and revealed the remarkable features of this protein in inhibiting a strikingly large subset of the chymotrypsin family of serine proteases. The structures of the ecotin:serine protease complexes provide the first examples of protein-protein recognition where the concept of specificity of interactions needs to be reexamined. The binding sites show a fluidity of protein contacts derived from ecotin's innate flexibility in fitting itself to proteases while strongly interfering with their function.

The motor domain of the kinesin homolog ncd has been crystallized in the presence of MgATP by the vapor diffusion method using polyethylene glycol as the precipitant. The crystals belong to the orthorhombic space group I222 with unit cell dimensions a = 127.1 A, b = 122.3 A, c = 68.0 A, and there is one ncd molecule per asymmetric unit. The crystals diffract X-ray to at least 2.3 A and are appropriate for high-resolution structure determination.