Each binding modules (ABM and TBM) and hybrid antibody recruiting glycodendrimers (ARGs composed of ABM and TBM parts) described in Annex 1 have been synthesized by click chemistry within the first weeks, then later in the project, including additional negative controls and labelled structures required to assess both their binding and biological properties. These fully original molecules have been carefully purified and characterized using RP-HPLC, NMR, HR-MS. Despite the structural complexity and the size of these molecules, no chemical problem was encounted at this stage.
As soon as the first hybride molecules were available, binding assays were conducted in our lab with LecA and LecB lectins, anti-rhamnose antibodies and human serum (ITC, microarrays), and in parallel with biological assays planed in the group of I. Attrée with P. aeruginosa strains that produce these lectins (viability and binding assays, biofilm inhibition). If binding assays between complete molecules and soluble lectins showed dissociation constant in the nanomolar range (measured by ITC), the biological effect of these molecules with bacteria was much more complicated for several reasons. The main problem was due to the fact that none of the planed assays are described in the litterature for this type of compound. Therefore, we had first to adapt the experimental set up (using technics such as FACS) to our objectives, ie the visualization of the lectin-mediated targeting of bacteria. If promissing interaction was confirmed with most of the tested compounds, unexpected binding was also observed in our hands with control molecules that could be attributed to unspecific binding of the antibody binding module with the bacteria membrane. At this stage and despite our efforts, we are not able to confirm that the binding with our hybrid molecules is due to the expression of specific lectins as it was expected. This crucial aspect to demonstrate the proof of concept is still under investigation with the group of I. Attrée and with close contact with other groups of experts in this field.
As a fallback solution, we have refined the design of the antibody binding modules by replacing rhamnose (ie ligands for serum antibodies) with peptides originally used to purify antibodies by affinity chromatography due to their bind properties to their Fc region. During the last six months, a series of peptides reported in the litterature was thus synthesized and immobilized on microarray in our group which has thus confirmed the binding with purified antibodies and from human serum. These experiments allowed us to identify sequences that are being conjugated to ABM to be tested with bacteria as soon as the experimental issues mentioned above will have been solved. This new series of molecules is fully promissing and innovative and a patent application is under preparation with our Technology Transfer Office (Floralis).
At the meantime, all the hybride molecules synthesized during the first months of PATHO-LEGO have been tested in killing assays using resistant bacteria and human serum. So far despite numbers of experiments, bacteria death was only observed in certain assays but not in a reproducible manner. This is certainly due to the unspecific binding of the antibody binding module with the bacteria membrane discussed above. This may thus prevent the formation of the ternary complex antibody-molecule-bacteria which is required to induce the immune response. We are optimistic in the fact that the new generation of hybride molecules with Fc-peptide will circumvent this problems. These new molecules are being synthesized and characterized and all the biological assays with lectins, antibodies an bacteria will be run in the near future.
