Unit: Structural Biochemistry - URA 2185 CNRS
Director: ALZARI, Pedro M.
Our research activities are oriented towards the study of the three-dimensional structure and ligand-binding specificity of proteins by X-ray crystallography, protein biochemistry, microcalorimetry and molecular modeling techniques. Current research subjects in the lab include the structural and functional studies of (1) proteins involved in bacterial cell signaling, (2) glycosidases and glycosyltransferases of biomedical relevance, and (3) mycobacterial proteins that could represent new potential therapeutic targets for drug design. Some research subjects are described below; further information on these and other topics can be found in our Web page: http://www.pasteur.fr/recherche/unites/Bstruct.
Trypanosomal enzymes (P.M. Alzari)
Trans-sialidases (TS) are GPI-anchored surface enzymes expressed in specific developmental stages of trypanosome parasites like Trypanosoma cruzi, the etiologic agent of Chagas disease, and T. brucei, the agent of sleeping sickness. Trypanosomes are unable to synthesize sialic acid and use this enzyme to scavenge the monosaccharide from host glycoconjugates to sialylate mucin-like acceptor molecules present in the parasite plasma membrane, a reaction that is critical for T. cruzi survival and cell invasion capability. In collaboration with A.C. Frasch (Argentine) and S. Withers (Canada), we have carried out structural and biochemical studies of trypanosomal trans-sialidases from pathogenic (T. cruzi, T. brucei) and non-pathogenic (T. rangeli ) parasites in order to understand their unusual enzymatic activities and to provide a framework for the structure-based design of specific inhibitors with potential therapeutic applications.
We have previously determined the crystal structures of the recombinant T. rangeli sialidase (Buschiazzo et al, EMBO J, 2000) and T. cruzi trans-sialidase (Buschiazzo et al, Mol. Cell, 2002), and carried out mutagenesis studies of the enzymes from T. cruzi (Paris et al, Glycobiol., 2001) and T. brucei (Montagna et al, Eur. J. Biochem., 2002). These studies have shown that, in contrast with sialidases, the active site architecture of trans-sialidases is intrinsically flexible. Binding of sialylated substrates triggers a conformational switch, which modulates the affinity for the acceptor substrate and concomitantly creates the conditions for efficient transglycosylation. A detailed comparison with the closely related structure of T. rangeli sialidase reveals a highly conserved catalytic center, where subtle structural differences account for strikingly different enzymatic activities and inhibition properties. In addition, the overall mode of binding of a transition state-analog inhibitor to the active site cleft of trans-sialidase is similar to that observed in other viral and bacterial sialidases, dominated by the interactions of the inhibitor carboxylate with the conserved arginine triad. However, the interactions of the other pyranoside ring substituents (hydroxyl, N-acetyl and glycerol moieties) differ between trypanosomal, bacterial and viral sialidases, providing a structural basis for specific inhibitor design (Amaya et al, J. Mol. Biol., 2003).
Research efforts during the last few years have substantiated the pathogenic role of microbial sialidases in a number of infectious diseases, making of these enzymes attractive targets for drug design. Besides trypanosomes, a prominent example is influenza virus infection, for which structural biology and rational drug design have led to the development of at least three potent sialidase inhibitors clinically useful as anti-viral compounds. However, in spite of extensive studies on the structure and mechanism of sialidases, many fundamental questions on their catalytic mechanism still remain unanswered or controversial, in large part because no structural information was available on the different enzymatic species along the reaction coordinate. The T. cruzi trans-sialidase provides an ideal system to address this issue, since the higher anticipated affinities in the aglycone site improve the chances of observing bound or intermediate species. Thus, using an activated (fluor-containing) sialoside substrate, we were able to trap the reaction intermediate and show that both T. cruzi trans-sialidase and T. rangeli sialidase operate through a double displacement mechanism, involving the transient formation of a covalent sialyl-enzyme intermediate with a strictly conserved tyrosine residue (Watts et al, J. Am. Chem. Soc, 2003 ; Amaya et al, Structure, 2004 ; Watts et al, J. Biol. Chem., 2005). These results provide strong evidence that all sialidases likely operate through a similar mechanism involving the transient formation of a covalent sialylated enzyme. Furthermore, we believe that the ability to tune' the inactivation and reactivation rates of mechanism-based inactivators towards specific enzymes represents an important step towards developing this class of inactivators into therapeutically useful compounds.
The Trypanosoma cruzi proline racemase (P.M. Alzari)
Amino acid racemases catalyze the stereoinversion of the chiral a-carbon to produce the D-enantiomers that participate in biological processes such as cell wall construction in prokaryotes. To overcome the high energetic barrier of this chemical reaction, some racemases evolved to use pyridoxal phosphate (PLP) as cofactor, since formation of an imine PLP-substrate covalent bond greatly acidifies the chiral center by resonance. However, a second class of enzymes operates through a two-bases mechanism in a cofactor-independent manner. Within this class, bacterial proline racemases have been extensively studied as a model system. The proline racemase from the human parasite Trypanosoma cruzi is a secreted enzyme that triggers host B-cell polyclonal activation preventing specific humoral immune responses and is crucial for parasite evasion and fate. Previous studies have suggested an intriguing link between the mitogenic and catalytic activities of the protein, although the molecular basis of this correlation remained unclear.
In collaboration with P. Minoprio's team at Pasteur Institute, we have carried out structural and thermodynamic studies of the trypanosomal racemase (Buschiazzo et al, PNAS, 2006). The enzyme is a homodimer, with each monomer folded in two symmetric a/b subunits separated by a deep crevice. The structure and solution properties of TcPRACA in complex with a transition-state analog inhibitor reveal the presence of one reaction center per monomer, in contrast with the previously accepted model that puts forward a unique active site per dimer. In each monomer, two Cys residues are optimally located to perform acid/base catalysis through a carbanion stabilization mechanism, Mutation of the catalytic cysteines abolishes the enzymatic activity but preserves the mitogenic properties of the protein. In contrast, inhibitor binding promotes the closure of the inter-domain crevice and completely abrogates B-cell proliferation, suggesting that the mitogenic properties of TcPRACA depend on the exposure of transient epitopes in the ligand-free enzyme.
Mycobacterial Ser/Thr protein kinases and phosphatases (P.M. Alzari)
Bacterial signaling involve primarily the action of two-component systems, a histidine protein kinase and a response regulator. However, during the last few years several bacterial genes coding for eukariotic-like protein kinases or phosphatases have been identified. In particular, the genome of M. tuberculosis includes 11 genes which code for putative Ser/Thr protein kinases and 3 genes which code for putative Tyr or Ser/Thr protein phosphatases. In collaboration with S.T. Cole's team at the IP, we are carrying out a biochemical and structural study of mycobacterial Ser/Thr protein kinases and phosphatases to investigate the molecular basis of their modes of action and their possible role in cell signalling. We have previously cloned several of these enzymes in appropriate expression vectors and produced the recombinant proteins for further biochemical and structural characterization.
More recently, we have focused our work on the study of two trans-membrane enzymes from Mycobacterium tuberculosis, the Ser/Thr protein kinase PknB and the protein phosphatase PstP, which are part of a conserved operon presumably involved in cell growth control. PstP was found to specifically dephosphorylate model phospho-Ser/Thr substrates in a Mn2+-dependent manner. Autophosphorylated PknB was shown to be a substrate for Pstp and its kinase activity was affected by PstP-mediated dephosphorylation. Our results indicate that, as for eukaryotic homologues, phosphorylation of the activation loop provides a regulation mechanism of mycobacterial kinases and strongly suggest that PknB and PstP could work as a functional pair in vivo to control mycobacterial cell growth (Boitel et al, Mol. Microbiol. 2003). The phosphorylation studies have been extended to other mycobacterial kinases, revealing a conserved phosphorylation pattern of the activation loop (essential for activity) as well as in the juxtamembrane region (possibly involved in signaling cascades) (Duran et al, BBRC, 2005). Furthermore, we identified GarA (Rv1827), a Forkhead-Associated (FHA) domain-containing protein, as a putative physiological substrate of PknB. Using a global proteomic approach, GarA was found to be the optimal PknB substrate in non-denatured whole cell protein extracts from M. tuberculosis and the saprophyte M. smegmatis (Villarino et al, J. Mol. Biol., 2005). The ensuing model of PknB-GarA interactions suggests a substrate recruitment mechanism that might also apply to other mycobacterial kinases bearing multiple phosphorylation sites in their activation loops. Further work is currently in progress to assess the physiological relevance of these findings.
Experimental and computational molecular structural biology (M. Delarue)
A. Crystal structures and protein crystallography
1. 6PGL (coll. PT6 and V. Stoven, Bioinformatique Structurale, IP). The structure of 6-phosphogluconolactonase (6PGL) of T. brucei has been refined at 2.1 Å resolution. Different data sets of diffraction data were collected at 1.9 Å for crystals soaked in concentrated solutions of inhibitors or substrate analogs. No extra density was revealed in the resulting 2Fo-Fc maps. Co-crystallisation experiments are under way to try to solve this problem.
2. Tdt and Pol mu (N. Expert-Bezancon, coll. F. Rougeon, IP). Numerous purifications and crystallisation experiments have been done on polymerase mu, in an attempt to solve the structure of this important enzyme closely related to TdT and involved in non-homologous end joining. Several complexes with DNA duplexes of various lengths and sequences have been tried; another construct is in the process of becoming available, which will allow us to get rid of the N-terminus His-tag peptide. Attempts to cleave it in the current construct resulted in a mixture of species because non-specific cleavage was observed in the (disordered) loop 1. We have now built a plausible model of pol mu interacting with its probable substrate (see Ramsden et al., Mol. Cell, 2005, 19:357), based on the structure of TdT. Several short dynamics simulations were run to check the relative stabilities of the different complexes that were built. Experiments are being performed in the lab to check our structural hypotheses.
B. Normal Modes and Structural Biology (E. Lindahl)
The work on Normal Modes has been put on line for the following options: (1) NMA and X-Ray refinement, (2) NMA and EM refinement, (3) NMA and docking of small molecules. The last part has been published this year in N.A.R. In addition, a collaboration with the group of J.P. Changeux (IP) on the application of NMA to the nicotinic receptor (a pentameric protein) using the Elastic Network Model has been published in Biophys. J. The refinement of models into cryo-EM envelopes has been completely revised and updated, in particular for macromolecular objects with non-crystallographic symmetry. Several test cases have been worked out. See http://lorentz.immstr.pasteur.fr/norma.php.
C. Solvent Structure and electrostatics (C. Azuara)
A new model of protein-solvent interactions was put on line, in collaboration with H. Orland (CEA). It allows to get rid of the constant dielectric medium hypothesis in Poisson-Boltzmann calculations and directly computes the (dipolar) solvent density around the solute (see http://lorentz.immstr.pasteur.fr/mutate_solvate.php). The following issues are currently being (computationally) addressed: (1) Incorporation of the protein polarisability using Normal Mode Analysis, (2) Study of denaturation of proteins by urea as a function of concentration, (3) Study of stability of proteins in organic solvents (and mixtures thereof).
Keywords: structural biology, X-ray diffraction, bacterial protein kinases, glycosidases and glycosyl transferases, tuberculosis, Trypanosoma cruzi, polymerases, structural genomics