Structures of glycosylated mammalian glutaminyl cyclases reveal conformational variability near the active center

Formation of N-terminal pyroglutamate (pGlu or pE) from glutaminyl or glutamyl precursors is catalyzed by glutaminyl cyclases (QC). As the formation of pGlu-amyloid has been linked with Alzheimer's disease, inhibitors of QCs are currently the subject of intense development. Here, we report three crystal structures of N-glycosylated mammalian QC from humans (hQC) and mice (mQC). Whereas the overall structures of the enzymes are similar to those reported previously, two surface loops in the neighborhood of the active center exhibit conformational variability. Furthermore, two conserved cysteine residues form a disulfide bond at the base of the active center that was not present in previous reports of hQC structure. Site-directed mutagenesis suggests a structure-stabilizing role of the disulfide bond. At the entrance to the active center, the conserved tryptophan residue, W(207), which displayed multiple orientations in previous structure, shows a single conformation in both glycosylated human and murine QCs. Although mutagenesis of W(207) into leucine or glutamine altered substrate conversion significantly, the binding constants of inhibitors such as the highly potent PQ50 (PBD150) were minimally affected. The crystal structure of PQ50 bound to the active center of murine QC reveals principal binding determinants provided by the catalytic zinc ion and a hydrophobic funnel. This study presents a first comparison of two mammalian QCs containing typical, conserved post-translational modifications.     

Papaya glutamine cyclotransferase shows a singular five-fold beta-propeller architecture that suggests a novel reaction mechanism

Cyclisation of N-terminal glutamine and/or glutamate to yield pyroglutamate is an essential posttranslational event affecting a plethora of bioactive peptides and proteins. It is directly linked with pathologies ranging from neurodegenerative diseases to inflammation and several types of cancers. The reaction is catalysed by ubiquitous glutaminyl cyclotransferases (QCs), which present two distinct prototypes. Mammalian QCs are zinc-dependent enzymes with an alpha/beta-hydrolase fold. Here we present the 1.6-A-resolution structure of the other prototype, the plant analogue from Carica papaya (PQC). The hatbox-shaped molecule consists of an unusual five-fold beta-propeller traversed by a central channel, a topology that has hitherto been described only for some sugar-binding proteins and an extracellular nucleotidase. The high resistance of the enzyme to denaturation and proteolytic degradation is explained by its architecture, which is uniquely stabilised by a series of tethering elements that confer rigidity. Strikingly, the N-terminus of PQC specifically interacts with residues around the entrance to the central channel of a symmetry-related molecule, suggesting that this location is the putative active site. Cyclisation would follow a novel general-acid/base working mechanism, pivoting around a strictly conserved glutamate. This study provides a lead structure not only for plant QC orthologues, but also for bacteria, including potential human pathogens causing diphtheria, plague and malaria.     

X-ray diffraction structure of a plant glycosyl hydrolase family 32 protein: fructan 1-exohydrolase IIa of Cichorium intybus

Fructan 1-exohydrolase, an enzyme involved in fructan degradation, belongs to the glycosyl hydrolase family 32. The structure of isoenzyme 1-FEH IIa from Cichorium intybus is described at a resolution of 2.35 A. The structure consists of an N-terminal fivefold beta-propeller domain connected to two C-terminal beta-sheets. The putative active site is located entirely in the beta-propeller domain and is formed by amino acids which are highly conserved within glycosyl hydrolase family 32. The fructan-binding site is thought to be in the cleft formed between the two domains. The 1-FEH IIa structure is compared with the structures of two homologous but functionally different enzymes: a levansucrase from Bacillus subtilis (glycosyl hydrolase family 68) and an invertase from Thermotoga maritima (glycosyl hydrolase family 32).     

Carica papaya glutamine cyclotransferase belongs to a novel plant enzyme subfamily: cloning and characterization of the recombinant enzyme

A full-length cDNA encoding Carica papaya glutamine cyclotransferase was cloned by RT-PCR on the basis of results from amino acid sequencing of tryptic fragments of the native enzyme. The cDNA of 1036 nucleotides encodes a typical 22-residue signal peptide and a mature protein of 266 residues with a calculated molecular mass of 30,923 Da. Five plant ESTs encoding putative QCs highly homologous to PQC were identified and the numbers and locations of cysteines and N-glycosylation sites are conserved. The plant QC amino acid sequences are very different from the known mammalian QC sequences and no clear homology was observed. The PQC cDNA was expressed in Escherichia coli as either His-tagged PQC, with three different signal peptides and in fusions with thioredoxin, glutathione S-transferase, and (pre-) maltose-binding protein. In all cases, the expressed protein was either undetectable or insoluble. Expression in Pichia pastoris of PQC fused to the alpha-factor leader resulted in low levels of PQC activity. Extracellular expression of PQC in the insect cell/baculovirus system was successful and 15-50 mg/liter of active PQCs with three different secretion signals was expressed and purified. Further, PQC N-terminally fused to a combined secretion signal/His-tag peptide was correctly processed by the host signal peptidase and the His-tag could subsequently be removed with dipeptidyl peptidase I. The expressed products were characterized by activity assays, SDS-PAGE, N-terminal amino acid sequencing, MALDI-TOF mass spectroscopy, and peptide mass fingerprint analysis.     

Isolation, catalytic properties, and competitive inhibitors of the zinc-dependent murine glutaminyl cyclase

Murine glutaminyl cyclase (mQC) was identified in the insulinoma cell line beta-TC 3 by determination of enzymatic activity and RT-PCR. The cloned cDNA was expressed in the secretory pathway of the methylotrophic yeast Pichia pastoris and purified after fermentation using a new three-step protocol. mQC converted a set of various substrates with very similar specificity to human QC, indicating a virtually identical catalytic competence. Furthermore, mQC was competitively inhibited by imidazole derivatives. A screen of thiol reagents revealed cysteamine as a competitive inhibitor of mQC bearing a Ki value of 42 +/-2 microM. Substitution of the thiol or the amino group resulted in a drastic loss of inhibitory potency. The pH dependence of catalysis and inhibition support that an uncharged nitrogen of the inhibitors and the substrate is necessary in order to bind to the active site of the enzyme. In contrast to imidazole and cysteamine, the heterocyclic chelators 1,10-phenanthroline, 2,6-dipicolinic acid, and 8-hydroxyquinoline inactivated mQC in a time-dependent manner. In addition, citric acid inactivated the enzyme at pH 5.5. Inhibition by citrate was abolished in the presence of zinc ions. A determination of the metal content by total reflection X-ray fluorescence spectrometry and atomic absorption spectroscopy in mQC revealed stoichiometric amounts of zinc bound to the protein. Metal ion depletion appeared to have no significant effect on protein structure as shown by fluorescence spectroscopy, suggesting a catalytic role of zinc. The results demonstrate that mQC and probably all animal QCs are zinc-dependent catalysts. Apparently, during evolution from an ancestral protease, a switch occurred in the catalytic mechanism which is mainly based on a loss of one metal binding site.     

Substrate specificity of glutaminyl cyclases from plants and animals

Glutaminyl cyclases (QC) catalyze the intramolecular cyclization of N-terminal glutamine residues of peptides and proteins. For a comparison of the substrate specificity of human and papaya QC enzymes, a novel continuous assay was established by adapting an existing discontinuous method. Specificity constants (kcat/Km) of dipeptides and dipeptide surrogates were higher for plant QC, whereas the selectivity for oligopeptides was similar for both enzymes. However, only the specificity constants of mammalian QC were dependent on size and composition of the substrates. Specificity constants of both enzymes were equally pH-dependent in the acidic pH-region, revealing a pKa value identical to the pKa of the substrate, suggesting similarities in the substrate conversion mode. Accordingly, both QCs converted the L-beta homoglutaminyl residue in the peptide H-beta homoGln-Phe-Lys-Arg-Leu-Ala-NH2 and the glutaminyl residues of the branched peptide H-Gln-Lys(Gln)-Arg-Leu-Ala-NH2 as well as the partially cyclized peptide H-Gln-cyclo(N epsilon-Lys-Arg-Pro-Ala-Gly-Phe). In contrast, only QC from C. papaya was able to cyclize a methylated glutamine residue, while this compound did not even inhibit human QC-catalysis, suggesting distinct substrate recognition pattern. The conversion of the potential physiological substrates [Gln1]-gastrin, [Gln1]-neurotensin and [Gln1]-fertilization promoting peptide indicates that human QC may play a key role in posttranslational modification of most if not all pGlu-containing hormones.     

A conserved hydrogen-bond network in the catalytic centre of animal glutaminyl cyclases is critical for catalysis

QCs (glutaminyl cyclases; glutaminyl-peptide cyclotransferases, EC 2.3.2.5) catalyse N-terminal pyroglutamate formation in numerous bioactive peptides and proteins. The enzymes were reported to be involved in several pathological conditions such as amyloidotic disease, osteoporosis, rheumatoid arthritis and melanoma. The crystal structure of human QC revealed an unusual H-bond (hydrogen-bond) network in the active site, formed by several highly conserved residues (Ser(160), Glu(201), Asp(248), Asp(305) and His(319)), within which Glu(201) and Asp(248) were found to bind to substrate. In the present study we combined steady-state enzyme kinetic and X-ray structural analyses of 11 single-mutation human QCs to investigate the roles of the H-bond network in catalysis. Our results showed that disrupting one or both of the central H-bonds, i.e., Glu(201)...Asp(305) and Asp(248)...Asp(305), reduced the steady-state catalysis dramatically. The roles of these two COOH...COOH bonds on catalysis could be partly replaced by COOH...water bonds, but not by COOH...CONH(2) bonds, reminiscent of the low-barrier Asp...Asp H-bond in the active site of pepsin-like aspartic peptidases. Mutations on Asp(305), a residue located at the centre of the H-bond network, raised the K(m) value of the enzyme by 4.4-19-fold, but decreased the k(cat) value by 79-2842-fold, indicating that Asp(305) primarily plays a catalytic role. In addition, results from mutational studies on Ser(160) and His(319) suggest that these two residues might help to stabilize the conformations of Asp(248) and Asp(305) respectively. These data allow us to propose an essential proton transfer between Glu(201), Asp(305) and Asp(248) during the catalysis by animal QCs.     

A Unique Carboxylic-Acid Hydrogen-Bond Network (CAHBN) Confers Glutaminyl Cyclase Activity on M28 Family Enzymes

Proteins with sequence or structure similar to those of di-Zn exopeptidases are usually classified as the M28-family enzymes, including the mammalian-type glutaminyl cyclases (QCs). QC catalyzes protein N-terminal pyroglutamate formation, a posttranslational modification important under many physiological and pathological conditions, and is a drug target for treating neurodegenerative diseases, cancers and inflammatory disorders. Without functional characterization, mammalian QCs and their orthologs remain indistinguishable at the sequence and structure levels from other M28-family proteins, leading to few reported QCs. Here, we show that a low-barrier carboxylic-acid hydrogen-bond network (CAHBN) is required for QC activity and discriminates QCs from M28-family peptidases. We demonstrate that the CAHBN-containing M28 peptidases deposited in the PDB are indeed QCs. Our analyses identify several thousands of QCs from the three domains of life, and we enzymatically and structurally characterize several. For the first time, the interplay between a CAHBN and the binuclear metal-binding center of mammalian QCs is made clear. We found that the presence or absence of CAHBN is a key discriminator for the formation of either the mono-Zn QCs or the di-Zn exopeptidases. Our study helps explain the possible roles of QCs in life.     

Crystal structure of the autocatalytic initiator of glycogen biosynthesis,glycogenin

The crystal structure of a medium-chain NAD(H)-dependent alcohol dehydrogenase (ADH) from an archaeon has been solved by multiwavelength anomalous diffraction, using a selenomethionine-substituted enzyme. The protein (SsADH), extracted from the hyperthermophilic organism Sulfolobus solfataricus, is a homo-tetramer with a crystallographic 222 symmetry. Despite the low level of sequence identity, the overall fold of the monomer is similar to that of the other homologous ADHs of known structure. However, a significant difference is the orientation of the catalytic domain relative to the coenzyme-binding domain that results in a larger interdomain cleft. At the bottom of this cleft, the catalytic zinc ion is coordinated tetrahedrally and lacks the zinc-bound water molecule that is usually found in ADH apoform structures. The fourth coordination position is indeed occupied by a Glu residue, as found in bacterial tetrameric ADHs. Other differences are found in the architecture of the substrate pocket whose entrance is more restricted than in other ADHs. SsADH is the first tetrameric ADH X-ray structure containing a second zinc ion playing a structural role. This latter metal ion shows a peculiar coordination, with a glutamic acid residue replacing one of the four cysteine ligands that are highly conserved throughout the structural zinc-containing dimeric ADHs.     

A water-mediated salt link in the catalytic site of Escherichia coli alkaline phosphatase may influence activity

Escherichia coli alkaline phosphatase catalyzes the hydrolysis of a wide variety of phosphomonoesters at similar rates, and the reaction proceeds through a phosphoenzyme intermediate. The active site region is highly conserved between the E. coli and mammalian alkaline phosphatases. The three-dimensional structure of the E. coli enzyme indicates that Lys-328, which is replaced by histidine in all mammalian alkaline phosphatases, is bridged to the phosphate through a water molecule. This water molecule is also hydrogen bonded to Asp-327, a bidendate ligand of the one of the two zinc atoms. Here we report the use of site-specific mutagenesis to convert Lys-328 to both histidine and alanine. Steady-state kinetic studies above pH 7.0 indicate that both mutant enzymes have altered pH versus activity profiles compared to the profile for the wild-type enzyme. At pH 10.3, in the presence of Tris, the Lys-328----Ala enzyme is approximately 14-fold more active than the wild-type enzyme. At the same pH in the absence of Tris the Lys-328----Ala enzyme is still 6-fold more active than the wild-type enzyme. Both mutant enzymes have lower phosphate affinities than the wild-type enzyme at all pH values investigated. Pre-steady-state kinetics at pH 5.5 reveal that the Lys-328----Ala enzyme behaves very similar to the phosphate-free wild-type enzyme.(ABSTRACT TRUNCATED AT 250 WORDS)     

Crystal structure of Bacillus subtilis guanine deaminase: the first domain-swapped structure in the cytidine deaminase superfamily

Guanine deaminase, a key enzyme in the nucleotide metabolism, catalyzes the hydrolytic deamination of guanine into xanthine. The crystal structure of the 156-residue guanine deaminase from Bacillus subtilis has been solved at 1.17-A resolution. Unexpectedly, the C-terminal segment is swapped to form an intersubunit active site and an intertwined dimer with an extensive interface of 3900 A(2) per monomer. The essential zinc ion is ligated by a water molecule together with His(53), Cys(83), and Cys(86). A transition state analog was modeled into the active site cavity based on the tightly bound imidazole and water molecules, allowing identification of the conserved deamination mechanism and specific substrate recognition by Asp(114) and Tyr(156'). The closed conformation also reveals that substrate binding seals the active site entrance, which is controlled by the C-terminal tail. Therefore, the domain swapping has not only facilitated the dimerization but has also ensured specific substrate recognition. Finally, a detailed structural comparison of the cytidine deaminase superfamily illustrates the functional versatility of the divergent active sites found in the guanine, cytosine, and cytidine deaminases and suggests putative specific substrate-interacting residues for other members such as dCMP deaminases.     

Crystal Structures of Glutaminyl Cyclases (QCs) from Drosophila melanogaster Reveal Active Site Conservation between Insect and Mammalian QCs

Glutaminyl cyclases (QCs), which catalyze the formation of pyroglutamic acid (pGlu) at the N-terminus of a variety of peptides and proteins, have attracted particular attention for their potential role in Alzheimer's disease. In a transgenic Drosophila melanogaster (Dm) fruit fly model, oral application of the potent competitive QC inhibitor PBD150 was shown to reduce the burden of pGlu-modified Aβ. In contrast to mammals such as humans and rodents, there are at least three DmQC species, one of which (isoDromeQC) is localized to mitochondria, whereas DromeQC and an isoDromeQC splice variant possess signal peptides for secretion. Here we present the recombinant expression, characterization, and crystal structure determination of mature DromeQC and isoDromeQC, revealing an overall fold similar to that of mammalian QCs. In the case of isoDromeQC, the putative extended substrate binding site might be affected by the proximity of the N-terminal residues. PBD150 inhibition of DromeQC is roughly 1 order of magnitude weaker than that of the human and murine QCs. The inhibitor binds to isoDromeQC in a fashion similar to that observed for human QCs, whereas it adopts alternative binding modes in a DromeQC variant lacking the conserved cysteines near the active center and shows a disordered dimethoxyphenyl moiety in wild-type DromeQC, providing an explanation for the lower affinity. Our biophysical and structural data suggest that isoDromeQC and human QC are similar with regard to functional aspects. The two Dm enzymes represent a suitable model for further in-depth analysis of the catalytic mechanism of animal QCs, and isoDromeQC might serve as a model system for the structure-based design of potential AD therapeutics.     

β-Propeller repeats and a PDZ domain in the tricorn protease: predicted self-compartmentalisation and C-terminal polypeptide-binding strategies of substrate selection

Prokaryotic proteases demonstrate a variety of substrate-selection strategies that prevent uncontrolled protein degradation. Proteasomes and ClpXP-like proteases form oligomeric structures that exclude large substrates from central solvated chambers containing their active sites. Monomeric prolyl oligopeptidases have been shown to contain beta-propeller structures that similarly reduce access to their catalytic residues. By contrast, Tsp-like enzymes contain PDZ domains that are thought to specifically target C-terminal polypeptides. We have investigated the sequence of Thermoplasma acidophilum tricorn protease using recently-developed database search methods. The tricorn protease is known to associate into a 20 hexamer capsid enclosing an extremely large cavity that is 37 nm in diameter. It is unknown, however, how this enzyme selects its small oligopeptide substrates. Our results demonstrate the presence in tricorn protease of a PDZ domain and two predicted six-bladed beta-propeller domains. We suggest that the PDZ domain is involved in targeting non-polar C-terminal peptides, similar to those generated by the T. acidophilum proteasome, whereas the beta-propeller domains serve to exclude large substrates from the tricorn protease active site in a similar manner to that previously indicated for prolyl oligopeptidase.     

Dipeptidyl aminopeptidase IV from Stenotrophomonas maltophilia exhibits activity against a substrate containing a 4-hydroxyproline residue

The crystal structure of dipeptidyl aminopeptidase IV from Stenotrophomonas maltophilia was determined at 2.8-A resolution by the multiple isomorphous replacement method, using platinum and selenomethionine derivatives. The crystals belong to space group P4(3)2(1)2, with unit cell parameters a = b = 105.9 A and c = 161.9 A. Dipeptidyl aminopeptidase IV is a homodimer, and the subunit structure is composed of two domains, namely, N-terminal beta-propeller and C-terminal catalytic domains. At the active site, a hydrophobic pocket to accommodate a proline residue of the substrate is conserved as well as those of mammalian enzymes. Stenotrophomonas dipeptidyl aminopeptidase IV exhibited activity toward a substrate containing a 4-hydroxyproline residue at the second position from the N terminus. In the Stenotrophomonas enzyme, one of the residues composing the hydrophobic pocket at the active site is changed to Asn611 from the corresponding residue of Tyr631 in the porcine enzyme, which showed very low activity against the substrate containing 4-hydroxyproline. The N611Y mutant enzyme was generated by site-directed mutagenesis. The activity of this mutant enzyme toward a substrate containing 4-hydroxyproline decreased to 30.6% of that of the wild-type enzyme. Accordingly, it was considered that Asn611 would be one of the major factors involved in the recognition of substrates containing 4-hydroxyproline.     

Splicing factor SF3a60 is the mammalian homologue of PRP9 of S.cerevisiae: the conserved zinc finger-like motif is functionally exchangeable in vivo.

A cDNA encoding the 60 kDa subunit of mammalian splicing factor SF3a has been isolated. The deduced protein sequence reveals a 30% identity to the PRP9 splicing protein of the yeast S.cerevisiae. The highest homology is present in a zinc finger-like region in the C-terminal domain of both proteins. The PRP9 zinc finger-like motif has been replaced by the equivalent region of mammalian SF3a60. The chimeric protein rescues the temperature-sensitive phenotype of the prp9-1 mutant strain demonstrating that not only the structure but also the function of this domain has been conserved during evolution.     

Evidence for essential histidines in human pituitary glutaminyl cyclase

Glutaminyl cyclase (QC, EC 2.3.2.5) catalyzes the formation of the pyroglutamyl residue present at the amino terminus of numerous secretory peptides and proteins. Treatment with diethyl pyrocarbonate inactivated recombinant human QC with the apparent modification of three essential histidine residues. Comparisons of the protein sequences of QC from a variety of eukaryotic species show four completely conserved histidine residues. Mutation of each of these residues to glutamine resulted in two mutant enzymes that were inactive (H140Q and H330Q), suggesting a role in catalysis, and two that exhibited increased Km values (H307Q and H319Q), suggesting a role in substrate binding. Consistent with these results is the prediction that QC possesses a zinc aminopeptidase domain in which the four histidines identified here are present in the active site. Mammalian glutaminyl cyclases may, therefore, have structural and catalytic similarities to a family of bacterial zinc aminopeptidases.     

Crystal structure of full length topoisomerase I from Thermotoga maritima

DNA topoisomerases are a family of enzymes altering the topology of DNA by concerted breakage and rejoining of the phosphodiester backbone of DNA. Bacterial and archeal type IA topoisomerases, including topoisomerase I, topoisomerase III, and reverse gyrase, are crucial in regulation of DNA supercoiling and maintenance of genetic stability. The crystal structure of full length topoisomerase I from Thermotoga maritima was determined at 1.7A resolution and represents an intact and fully active bacterial topoisomerase I. It reveals the torus-like structure of the conserved transesterification core domain comprising domains I-IV and a tightly associated C-terminal zinc ribbon domain (domain V) packing against domain IV of the core domain. The previously established zinc-independence of the functional activity of T.maritima topoisomerase I is further supported by its crystal structure as no zinc ion is bound to domain V. However, the structural integrity is preserved by the formation of two disulfide bridges between the four Zn-binding cysteine residues. A functional role of domain V in DNA binding and recognition is suggested and discussed in the light of the structure and previous biochemical findings. In addition, implications for bacterial topoisomerases I are provided.     

Crystal structure of Toxoplasma gondii porphobilinogen synthase: insights on octameric structure and porphobilinogen formation

Porphobilinogen synthase (PBGS) is essential for heme biosynthesis, but the enzyme of the protozoan parasite Toxoplasma gondii (TgPBGS) differs from that of its human host in several important respects, including subcellular localization, metal ion dependence, and quaternary structural dynamics. We have solved the crystal structure of TgPBGS, which contains an octamer in the crystallographic asymmetric unit. Crystallized in the presence of substrate, each active site contains one molecule of the product porphobilinogen. Unlike prior structures containing a substrate-derived heterocycle directly bound to an active site zinc ion, the product-bound TgPBGS active site contains neither zinc nor magnesium, placing in question the common notion that all PBGS enzymes require an active site metal ion. Unlike human PBGS, the TgPBGS octamer contains magnesium ions at the intersections between pro-octamer dimers, which are presumed to function in allosteric regulation. TgPBGS includes N- and C-terminal regions that differ considerably from previously solved crystal structures. In particular, the C-terminal extension found in all apicomplexan PBGS enzymes forms an intersubunit β-sheet, stabilizing a pro-octamer dimer and preventing formation of hexamers that can form in human PBGS. The TgPBGS structure suggests strategies for the development of parasite-selective PBGS inhibitors.