Computer-based molecular modelling was combined with biochemical and
biophysical methods to determine the structural features of peptides that are
important for their recognition by peptide transporters. Information gained from this study advances the understanding of ligand recognition in general as well as by peptide transporters in particular.
A stochastic search protocol, using implicit solvation, was used to generate
conformer sets for di-, tri- and oligopeptides. Dedicated programs were written to analyse these conformer populations and the results displayed using novel 3D plots. Contrary to current perception our results show that small peptides adopt a limited number of backbone conformations in solution. Combining information on these predominant backbone forms with results from peptide transport assays on individual peptides defined the molecular recognition templates (MRTs) for Dpp and Tpp. The relationship between the proportion of MRT conformers and the relative affinity of a peptide for a transporter showed Dpp and Tpp to be kinetically-driven. The ability of tripeptides to act as substrates for these transporters was explained by the presence of 'folded' forms that matched the dipeptide MRT. Extended tripeptides and oligopeptides comprised a subset of backbone forms that are substrates for Opp, highlighting the complementarity of the three archetypal bacterial peptide transporters. Information upon the MRTs of peptide transporters will greatly assist
the rational design of smugglin type drugs designed to exploit these transport
systems for their delivery.
Methodology devised here was also applied to the study of the recognition of
cell-wall peptides and β-lactam antibiotics. The information gained on the
conformations adopted by each of these compounds suggested features important for their recognition by transpeptidases and gave further insight into the inhibitory action of β-lactam antibiotics. Cell-wall peptide analogues, found in most vancomycin resistant strains, were also modelled and shown to have a different distribution of conformational types that may be an important factor in resistance to glycopeptide antibiotics. Knowledge of the MRTs of transpeptidases and vancomycin will aid the development of new antibacterial compounds and further our understanding of antibiotic resistance mechanisms.