Research Interests: Protein Engineering, Molecular Biology, Enzymology, Structural Biology, Molecular Modelling, Protein NMR, Organic Chemistry & Cellular Biology

Overview of Research Program

The complexity of enzyme catalysis, consisting of ligand recognition and exquisite discrimination followed by rapid catalytic turnover, has traditionally been difficult to study. Recent developments in molecular modelling of macromolecules and powerful molecular biological methodologies now offer new approaches for improving our understanding of enzyme structure-function relationships. While these tools are commonly applied to research on non-catalytic proteins, the research program underway combines these methodologies to explore the fundamentals of enzyme catalysis. The benefits of better understanding enzyme catalysis are numerous: in our laboratory, we A) are gaining a better understanding of enzyme-based drug resistances and B) are modifying enzymes for synthetic applications.

A) By gaining a better understanding of the ligand selectivity that frequently underlies drug resistance, we will provide additional information for drug design. Current approaches to small-molecule drug design include such approaches as structure-based, computer-assisted strategies and high-throughput small-molecule library screening. Knowledge of the molecular contacts that exist between the enzyme and a ligand – whether the ligand is a natural substrate or an inhibitor – provides information that is essential to the design of more advanced drug generations. Currently, there is a pressing for development of complementary strategies to provide greater detail about the nature of enzyme-ligand interactions.

B) In order to gain the knowledge and capacity to modify enzymes for synthetic purposes, we are gaining additional insight into the plasticity of enzyme active sites by distinguishing the elements that are indispensable for function from those that can be altered.

We are working to develop and apply methodologies that enhance the speed, power and information-content of the various steps implicated in this research. The methodologies should be generally-applicable to many enzymes so as to provide maximal benefit to the field of enzyme engineering.

Exploration and modification of the substrate specificity of a transglutaminase enzyme, toward the development of model substrates and inhibitors
Project in collaboration with Professor Jeffrey W. Keillor and Dr. Robert Lortie

Transglutaminase (TGase) catalyses the formation of gamma-Glu-epsilon-Lys isopeptide bonds from the gamma-carboxamide group of a glutamine (Gln) residue of one protein and the epsilon-amino group of a lysine (Lys) residue of another protein. The TGase-mediated cross-linking of proteins is implicated in blood clotting, endocytosis, apoptosis and cell growth regulation. It can also lead to physiological disorders such as cataract formation and Celiac disease and is thus a target for drug design. Our collaborative goal is to explore the specificity of the enzyme, toward the development of model substrates and inhibitors.

The Pelletier group has brought in expertise in recombinant DNA and protein expression methodologies. Our collaborative work has resulted in the expression in E. coli of a hexahistidine-tagged guinea pig liver tissue tranglutaminase (His6-tTGase) allowing rapid purification. Using this procedure we have obtained the highest reported specific activity (17 U/mg) combined with a high yield (22 mg/L of culture) for recombinant TGase using a single-step purification protocol. (S. M. F. G. Gillet, R. A. Chica, J. W. Keillor and J. N. Pelletier (2004) Prot. Exp. & Purif. 33: 256-264)

In order to modify substrate specificity, a better understanding of the mode of binding of model peptide substrates to the enzyme was required. Using a combined kinetic/molecular modelling approach, we generated a model for the binding of small acyl-donor peptide substrates to tissue transglutaminase. Our binding model of CBz-Gln-Gly on tissue TGase provides us with a greater level of detail regarding substrate docking at the active site and chemical steps in the acylation reaction. (R. A. Chica, P. Gagnon, J. W. Keillor and J. N. Pelletier (2004) Prot. Sci. 13(4):979-991)

Figure 1. Proposed binding cleft for the peptide substrate N-terminal functional group at the surface of TGase. The diagonal cleft running over the active site cavity is shaded while the Gln side chain of the CBz-Gln-Gly substrate that enters the active site cavity away from the viewer is represented by "x". Negative charges are in red, positive are in blue. The four positions tested for manual docking of the N-terminal CBz group are numbered. The CBz group is drawn to scale. Our results indicate that position 0 is favoured for binding.

R67 dihydrofolate reductase: an enzyme with a ‘binding hot-spot’

Mutating individual active-site residues is frequently deleterious to function, resulting in a large effect on enzyme viability, or fitness. This is a result of ‘functional coupling’ between residues, which defines the degree to which residues must be altered simultaneously when their mutation results in non-additive effects on function. Consequently, a productive means of active site modification should be the co-variation of the coupled, active site residues. To illustrate this strategy, we undertook the combinatorial sequence exploration of all the principal amino acids that constitute the active site of Type II R67 dihydrofolate reductase (R67 DHFR), a bacterial enzyme identified as a result of its resistance to the widely-administered antibiotic trimethoprim (TMP). It has been suggested that the active site cavity is a ‘binding hot-spot’ of relatively low specificity since it can accommodate various folate analogs at the same time as a hydride donor such as NADPH. Our goal is to modify and exploit this ‘hot-spot’ to create new redox-active enzymes for efficient preparation of folate analogs of synthetic interest for drug design.

By molecular modelling using the crystal structure coordinates of R67 DHFR, we defined the amino acids having potential ligand contacts. We combinatorially introduced active-site mutations, creating 1536 active-site variants and selected three new variants that fulfill the chemical and steric requirements of DHFR catalysis and TMP resistance. Importantly, the active-site sequences of these three functional variants differ entirely from the native enzyme. The negative free energy values revealed that the mutations are highly coupled. Our results support a model for catalysis where the transition state is mainly stabilized by productive positioning of the two ligands relative to each other with no direct participation of specific side-chains to catalysis. (A. R. Schmitzer, F. Lépine and J. N. Pelletier (2005) Prot. Eng. Des. Sel., in press)

Figure 2. Schematic representation of the sequence space covered by the 1536 R67 DHFR variants created. The central column represents WT R67 DHFR; column height (z axis) is a non-quantitative representation of fitness. Mutational distance is measured in the (x, y) plane where each concentric circle is one mutation away from the WT R67 DHFR. The library encodes 4 point mutants (approximately 0.2% of the clones) of no detectable fitness, shown as black circles on the first concentric ring. Double mutants where two positions are mutated occur at a frequency of approximately 5% (second ring). Triple and quadruple mutants (third and fourth rings, respectively) occur in 34% and in almost 60%, respectively, of all variants. Two triple (M10; M36) and one quadruple mutant (M8_30) of native-like fitness were identified and characterized. Variants NS15 and NS19 were identified prior to selection and have no detectable activity.

Selectivity of beta-lactamase toward its antibiotic substrates

beta-Lactamase production has become a world-wide problem in bacterial strain resistance to widely-used clinical antibiotics such as penicillins and cephalosporins. Class A serine hydrolase beta-lactamases have become extensively studied model enzymes. The high rate of occurrence of mutated enzymes capable of hydrolysing higher generation cephalosporins has stimulated research of beta-lactamase adaptation to these new substrates in order to understand the molecular basis of this evolutionary chain of events. Class-A conserved residue Tyr105 delineates one of the edges of the active site cavity of TEM-1 beta-lactamase from E. coli as a result of its position near the substrate. Although early chemical modification studies targeting Tyr105 in other Class-A beta-lactamases suggested a role for this residue in catalysis, these hypotheses have since been refuted.

We substituted Tyr105 by saturation mutagenesis in TEM-1 beta-lactamase in order to clarify its role in enzyme activity and in substrate stabilization and discrimination. Minimum inhibitory concentrations (MICs) in E. coli were determined for each Tyr105X mutant in presence of various penicillin and cephalosporin antibiotics. Only aromatic residues as well as asparagine conferred high in vivo survival rates for all substrates tested. At position 105, the small residues alanine and glycine provide weak substrate discrimination as evidenced by the difference in benzylpenicillin hydrolysis relative to cephalothin, two typical penicillin and cephalosporin antibiotics. In vitro kinetic analyses revealed that the Tyr105X replacements have a greater effect on Km than kcat, highlighting the importance of Tyr105 in substrate recognition and confirming that Tyr105 is not a catalytic residue. Finally, by performing a short Molecular Dynamics study on a restricted set of Tyr105X mutants, we found that the strong aromatic bias observed at position 105 is primarily defined by a structural requirement, selecting planar residues that form a stabilizing ‘wall’ to the active site (see figure below). The adopted conformation of residue 105 prevents detrimental steric interactions with the substrate in the active site cavity and provides a rationalization for the strong aromatic bias found in nature at this position among Class-A beta-lactamases. (N. Doucet, P.-Y. De Wals and J. N. Pelletier (2004) J. Biol. Chem. 279 (44): 46295-46303)

Figure 3. Superimposition of 40, 80, 120, 160 and 200 picosecond snapshots of a 200-ps dynamics trajectory for wild-type TEM-1 and mutants Trp (W) and Asn (N), which form a ‘wall’ to the active site in presence of benzylpenicillin (BZ), compared to Gln (Q) which allows significant displacement of the substrate over time. Structures are represented as sticks. For simplification, the backbone direction of residues 104-108 is represented by a yellow ribbon. Only the substrate molecule (BZ) and side-chains of residues 105 and 107 are shown.