Research Activities

RESEARCH ACTIVITIES

The objective of our laboratory is to develop new methods to increase the potential of the directed evolution of enzymes and to modify existing enzymes in order to confer them new desirable properties. As the handling of large libraries require highly efficient selection methods, we are investing a significant effort in the development of in vitro selection techniques using phage-displayed enzymes. These methods are used in attempts to modify the properties of existing enzymes: change of specificity, increase or change of catalytic activity, creation of allosteric binding sites or of chemically reactive sites.

Evolution between phylogenetically related activities

Following the approach of "directed evolution", our paradigm in regard of this project is the attempt to evolve a DD-peptidase into a beta-lactamase. DD-peptidases, enzymes involved in the construction of bacterial walls, are blocked by beta-lactam antibiotics in the form stable acyl-enzymes, hence their qualification as PBPs (Penicillin Binding Proteins). Beta-Lactamases form similar but very unstable acyl-enzymes with these antibiotics, they are thought to have evolved from DD-peptidases by dramatically increasing the rate of deacylation.

In the early nineties, we had functionally displayed a beta-lactamase on phage and demonstrated the possibility of selecting the most active mutants from mixtures containing phage-enzymes of varying activity.

                          

We recently discovered a family of cyanobacterial genes encoding DD-peptidases that are the closest homologous to class A beta-lactamases known today. These proteins feature a significant sequence difference in the omega loop which normally carries a glutamic acid essential for the beta-lactamase activity since it is shorter by six residues and the glutamic acid is not conserved. From this family, the thermostable PBP-A from Thermosynechococcus elongatus was chosen as the target enzyme for directed evolution. In a first attempt to generate beta-lactamase activity, a 90 fold increase in deacylation rate was obtained by introducing a glutamate in the shorter W-loop (L158E mutation). This activity is still around five orders of magnitude lower than a natural β-lactamase. This reflects an important puzzling paradigm in enzyme structure-function relationships since the essential residues of both active sites are almost perfectly superimposed. 

                                                   

Superposition of the active site residues of PBP-A (pink), PBP-A L158E (blue) and TEM-1 beta-lactamase (green). Spheres represent water molecules. The green one plays a crucial role in the mechanism of beta-lactamase and finds an equivalent in the improved PBP-A mutant.

Two libraries of phage-displayed mutants were then created, one by introducing random mutations along the entire L158E mutant, and a loop-library by insertion of degenerated full-length W-loops in place of the shorter one. Although in vivo selections for ampicillin resistance was unsuccessful, phage display selections for acylation using a biotinylated penicillin allowed the identification of mutants featuring at least a 10 fold increase in the acylation rates, with the best mutants found in the loop-library. 

Innovative in vitro selection protocols for beta-lactamase activity (improved acylation and deacylation), validated by kinetic modelling, are also tested on these libraries.

Key publications:

  • SOUMILLION P, JESPERS L, BOUCHET M, MARCHAND J, WINTER G, FASTREZ J, Selection of bêta-lactamase on filamentous bacteriophage by catalytic activity, J. Mol. Biol., 237, 1994, 415-422.
  • VANWETSWINKEL S, MARCHAND-BRYNAERT J, FASTREZ J, Selection of the most active enzymes from a mixture of phage-displayed beta-lactamase mutants. Bioorg. Med. Chem. Let., 7, 1997, 239.
  • LABARBE C, Vers l'évolution d'une DD-peptidase en beta-lactamase, Thèse de doctorat, 2006.

 

Creation of artificial allosteric regulation sites

We have previously shown that the TEM-1 beta-lactamase could accommodate, in some loops surrounding its active site, insertions of random sequences without compromising its activity. From the libraries of insertants that were generated, active enzymes binding with high affinities to different monoclonal antibodies could be isolated. Antibody binding was shown to markedly affect the enzyme activity, which implies that these hybrid enzymes can behave as signalling molecules in homogeneous assay. In a second series of selection experiments, clones endowed with affinity for streptavidin, horse spleen ferritin and beta-galactosidase could be isolated. Affinity maturation of a clone binding to ferritin allowed obtaining beta-lactamases with affinities comprised between 10 and 20 nM (Kd).

We have further extended this work by creating more complex libraries in which insertions of random peptides or random mutagenesis was performed in three contiguous loops. The purpose was to generate allosteric cavities with a potential affinity for small molecules. This design was inspired by the structure of camelid antibodies, which feature only a heavy chain with three hypervariable loops (CDRs). The libraries were created hierarchically by insertion or randomisation, selection for activity, combination of active libraries to afford a final library of 3 x 108 clones. Interestingly, 2% of them were quite active (5% of the wild type) and in more than 20% of the final active library, the three loops featured mutations and/or insertions. From these, we have been able to select β-lactamase variants endowed with affinities for transition metal ions (Kds between 1 and 100 µM) and an aminoglycoside (Kd around 10 µM). In all cases, ligand binding affects the activity upward or downward.

                                  

Key publications:

  • LEGENDRE D, SOUMILLION P, FASTREZ J, Engineering a regulatable enzyme for homogeneous immunoassays, Nat. Biotechnol., 17, 1, 1999, 67-72.
  • LEGENDRE D, VUCIC B, HOUGARDY V, GIRBOUX A-L, HENRIOUL C, VAN HAUTE J, SOUMILLION P, FASTREZ J, TEM-1 beta-lactamase as a scaffold for protein recognition and assay, Protein Sci., 11, 2002, 1506-1518.
  • MATHONET P, BARRIOS-CAMACHO H, SOUMILLION P, FASTREZ J, Selection of allosteric beta-lactamase mutants featuring an activity regulation by transition metal ions, Protein Sci., 15, 2006, 2335-2343.
  • MATHONET P, DEHERVE J, SOUMILLION P, FASTREZ J, Active TEM-1 beta-lactamase mutants with random peptides inserted in three contiguous surface loops, Protein Sci., 15, 2006, 2323-2334.

 

| 24/04/2008 |