Bacterial Polysaccharides and chemically defined synthetic vaccines. Synthesis and characterization of "protective epitopes": a key approach (L. Mulard)
The development of a vaccine against Shigella flexneri, a Gram negative bacterium responsible for the endemic form of bacillary dysentery, is one of the WHO priorities in its program for enteric vaccine development. The O-antigen (O-Ag) component of the surface lipopolysaccharides (LPS) is the main target of the protective immune response elicited by the host following infection. For that reason, bacterial vaccines are under development based on the use of these surface antigens. Recently, the notion of "protective carbohydrate epitopes" has been defined. As an alternative strategy to the standard "polysaccharide:protein" approach, we wish to develop optimal chemically defined constructs exposing, in a multivalent fashion, a combination of "protective B epitopes" (oligosaccharides) specific for the major Shigella strains, and appropriate T epitopes required for the memory response. To demonstrate the feasability of such an approach, serotypes 5a and 2a of S. flexneri are currently under study.
In order to characterize the structural, as well as the conformational carbohydrate epitopes for these two serotypes, di-, tri-, tetra-, and pentasaccharides representative of frame-shifted fragments of the O-Ag of S. flexneri serotypes 2a and 5a, which repeating units are two branched pentasaccharide regioisomers, have been obtained by chemical synthesis as their methyl glycosides. In the S. flexneri 5a series, data derived from the NMR conformational analysis of the synthetic oligosaccharides now allow the modelling of larger fragments of the O-Ag, which may help further understanding of the protective antibodies:O-Ag complementarity. In both series, study of the binding of the synthetic oligosaccharides to protective monoclonal antibodies directed to S. flexneri LPS were performed by inhibition ELISA. In the case of secretory IgAs specific for S. flexneri serotype 5a, the pentasaccharides were best recognized although with a rather low IC50 (20 mM). In both series, higher IC50 (1 to 100 µM) were obtained for seric IgGs. For serotype 5a, the branched point was found critical for recognition whereas for serotype 2a, the minimal recognized structure was a linear trisaccharide. For such systems, further structural studies of the oligosaccharide:antibody complexes are ongoing. Besides, a strategy to potential synthetic vaccines against S. flexneri was designed based on the binding data. The synthesis of such constructs has started.
Peptidic antigen presentation (F. Baleux)
Preliminary results of mice immunisation with peptides mimicking the O antigen of Shigella flexeneri LPS show that it is possible to induce antibodies that recognise the Shigella flexeneri LPS. These peptides are presented via the MAP system (Multiple Antigen Peptide) or coupled to BSA. The aim of this project is to develop a synthetic vaccine against Shigellose.
The MAP system enabled us to identify a new B-cell epitope in the Pf72/Hsp70-1 of Plasmodium falciparum not detected by Elisa with the linear peptide. This result suggests that the MAP strategy is also particularly useful as a means of obtaining suitable synthetic models for conformation-dependent epitopes.
Glycosylated epitopes secreted by Mycobacterium tuberculosis (L. Mulard, F. Baleux)
Recent data point out the crucial importance of secreted mannosylated proteins as immunodominant antigens upon infection by M. tuberculosis. Glycopeptide dominant epitopes are anticipated. However, although, the amino acid sequence of potential targets are known, the biochemical approach failed to reveal their precise glycoforms. Taking advantage of the flexibility of both carbohydrate chemistry and peptide solid phase synthesis, the synthesis of a panel of potentially critical glycopeptides is in progress. The required glycosylated amino acid building blocks have been synthesized.
Synthetic glycopeptides for anti-tumor immunotherapy (S. Bay)
Using a rational and specific approach of anti-tumor immunotherapy, we are developing synthetic vaccines based on carbohydrate tumor markers. We have synthesized a new immunogen : the MAG (Multiple Antigenic Glycopeptide) which carries the carbohydrate epitope (i.e. the tumor marker) associated with a T CD4+ peptide epitope. Such synthetic conjugates are particularly attractive for both their purity and accurate chemical definition, which are essential features for a safe vaccine.
We showed that the MAG system is an efficient strategy for inducing high levels of carbohydrate-specific antibodies and for increasing the survival of tumor-bearing mice, after prophylactic and therapeutic vaccinations (collaboration with R. Lo-Man et C. Leclerc).
We are currently elaborating new generations of MAG vaccines more powerful and potentially active in humans, by introducing CD8+ T cell epitopes as well as universal CD4+ T cell epitopes. The immunogenicity of first linear model compounds are under investigation in transgenic mice for the HLA molecules. This approach will be next extended to the MAG system using a fragment ligation strategy.
Another part of our program concerns the synthesis of glycopeptides i) as immunogens for inducing specific CTL responses, ii) as substrates for studying various glycosyl transferases specificity, iii) for Tn-binding proteins specificity studies (lectins and antibodies).
Chemokines/HIV (F. Baleux)
Interest in the discovery of HIV entry co-receptors (CCR5 and CXCR4 mainly) was reinforced by the observation that chemokines, their natural ligands, inhibit the early replicatives stages. CC chemokines RANTES, MIP-1a and MIP-1ß (CCR5 agonists) inhibit viral entry of R5 strains isolated during the early stage of the disease. CXC chemokine SDF-1, the only identified CXCR4 ligand, inhibits cell fusion and infection by X4 virus strains isolated during later stage of the disease. Moreover, the biological activities of chemokines are thought to be influence by their association with proteoglycans. We have already identified the major heparin-binding site of SDF as a cluster of 3 basic amino acids. Looking closer into this interaction, we have demonstrated that even single point mutation of one of this basic residue strongly decrease SDF/HS affinity and a model of SDF/heparin has been proposed.
Chemokines are small proteins (70 AAs) readily accessible to chemical synthesis with good yields (10-50 mg) and great purity. More interestingly, selective labelling by biotin or fluorescent compounds could be achieved by chemical synthesis, leading to still active chemokines. Non natural amino acids could also be incorporated in the chemokines. The N-terminal lysine residue plays a crucial role for SDF biological activity. Labelling this chemokine by classical methods could lead to an inactive compound. Using our non-binding heparin mutants in a cell fusion test and FRET between GFP-CXCR4 and Texas Red C-terminal labelled SDF, we found that optimal inhibition of X4 HIV strains by SDF requires interaction with cell-surface heparan sulfates proteoglycans. In order to cocrystallize SDF/CXCR4 and CCR5/MIP1ß complexes, synthetic C-terminal biotinylated chemokines are used for the purification of properly folded receptors.
Proteolytic regulation of SDF and CXCR4 is also under study.
Synthesis and biological activity of heterodox nucleosides (A. Kaminski, S. Pochet)
The goal of our research group is to increase the repertoire of monomers that can be incorporated into DNA as well as polymers replicatable in vitro and/or in vivo. Two kinds of DNA alterations are currently investigated : the first modification concerns the heterocyclic moiety. A combinatorial approach aiming to produce of a great number of nucleosides triphosphates in one step was developped in order to find monomers having highly specific or ambiguous base-pairing schemes or, new inhibitors to replication. The second modification concerns the sugar conterpart. HNAs are conformationally constrained furanose mimicking nucleic acids in which the sugar part is replaced by 1,5-anhydrohexitol. The capacity of such polymers to support genetic information transfer was investigated in vitro and in vivo. In order to enlarge the repertoire of unnatural nucleosides/nucleotides available by enzymatic process, nucléosides phosphorylases and N-désoxyribosyltransférases are evolved by combining mutagenesis and in vivo. A project aiming to elaborate novel inhibitory compounds against kinases which are essential for the bacterial growth in M. Tuberculosis started.
Synthesis of nucleoside and nucleotide analogs as potential antivirals (L. Mulard)
The design of new potent antivirals, based on the study of two key steps of the viral replication inhibition, is our main objective. Recent data demonstrated that a-borano triphosphate derivatives of known antivirals were better inhibitors of viral replication than their triphosphate analogues. Using d4T as a model, the vectorisation of such analogues, or precursors, is under study either as encapsulated material or as lipophilic prodrugs.
In order to develop new antivirals targetting the Dengue virus, Ribavirin was chosen as a lead compound. Several analogs specifically modified at position 2 and 3 of the ribose moiety were synthesized.
Oligonucleotides (T. Huynh-Dinh)
The synthesis of DNA or RNA fragments was carried out for the structural studies of triple helices GT (parallel and antiparallel) and in the loop-loop interactions of the genomic HIV-Lai RNA associated with the nucleocapside NCp7. In another study aiming to the understanding of the cross linking of trans-diamino dichloroplatin, different methods of chemical (in solution or in solid support) and enzymatic ligation were compared for their potential ability to produce large amounts of circularized oligonucleotides.
In a collaborative work in the field of DNA micro arrays, we have tested different methods of coupling activated glass slides to oligonucleotides at the 3' or 5' ends.
Directed enzyme evolution (JL Jestin)
To direct evolution of enzymes, we develop a chemistry of Inovirus to select proteins for catalytic activity. We aim to establish a strategy for the isolation of enzymes for given chemical reactions by in vitro selection.
Our in vitro selection of proteins for catalytic activity is an affinity selection for the reaction product crosslinked to the phage-enzyme that catalysed the reaction from substrate to product.. Enzymes are expressed on the surface of filamentous phage to link genotype and phenotype.
DNA-dependent DNA-polymerases have been expressed on the surface of bacteriophage. Using the phagemid system, only one phage in thousand displays a polymerase. A signal sequence library was designed and submitted to in vitro selection for the optimisation of polymerase on phage. Expression levels of 1 in 5 have been achieved suggesting that isolation of an appropriate signal sequence might be a general means of improving the display on filamentous phage.
Currently, we aim to generalise the in vitro selection for catalytic activity to other enzymes and to isolate catalysts for given chemical reactions from synthetic or natural protein libraries.