ulfate groups through intermolecular electrostatic interactions have also been postulated. We have also studied the interaction of Langerin with other GAGs, using competition approaches. Data obtained clearly showed a selectivity of the lectin for HS-like GAGs, although Langerin also bound to a much more modest level to CS/DS. Interestingly, we also observed binding selectivity amongst the CS/DS samples tested, Langerin exhibiting the highest binding to CS-C. Comparison of these data to GAG disaccharide analysis showed that binding to Langerin could not simply be attributed to a net charge effect and that specific saccharide features were most likely required. Our results suggest that C6 sulfation as well as Danoprevir iduronic acid strengthen the binding. Moreover, the affinity loss observed for heparin upon HSulf-2 treatment highlights the importance of the C6 sulfate present in the motif. We used the recent crystal structure of the langerin trimer to undertake molecular modeling analysis of Langerin interaction with heparin fragments. Combining the trimeric X-ray structure of a truncated ECD with the previously modeled neck region yielded a reasonably robust model to initiate the search for putative favourable heparin binding regions. Two main areas of interaction with heparin have been identified on the whole Langerin surface through MOLCAD electrostatic potential analysis and EADock DSS cavity detection and blind methylsulfate docking. Thanks to this preliminary dual approach, the more precise Autodock docking of heparin fragments was restricted to those specific areas. Outcome results of these molecular simulations yielded three main conclusions: i) neither methylsulfate nor heparin fragment docking pose interact with the calcium ions; ii) Heparin fragment-Langerin interactions are driven by direct polar forces; iii) the molecular recognition of heparin fragments depends upon more than one Langerin CRD: the most populated docking clusters occupy both CRDs characterizing the edge of the a-helix coiled-coil. Building on these clear modelling outputs, it was then possible to construct straightforwardly a heparin decamer in situ. In the model, the double sulfation of GlcNS residues appeared essential for the interaction, acting also as a bridge between both CRDs. Globally, the proposed model of heparin/Langerin molecular recognition is in full accordance with the biochemical results. However in order to get an accurate estimation of the free energy of binding and to go further towards a physically relevant description of molecular recognition, molecular dynamics studies would be considered as suitable to take into account the charged and flexible aminoacids coating the binding region. Moreover, the construction of the heparin chain was limited to ten monomers. Modelling of a more extended heparin chain could involve other areas of the protein, for instance at the top of the described region, toward the electropositive cavity involving Lys 299 and Lys 313 residues. Finally, characterization of the Langerin ECD structural organization in solution by SAXS will be conducted and will help to improve also the model of the protein itself. The multiple approaches of our work give convergent evidence for a novel binding mode of Langerin ligands. Remarkably, the binding is independent from the canonical Ca2+-site. Previously, the existence of two distinct binding sites within Langerin has been postulated on the basis of an X-ray structure of Langerin in c