The P pilus system of uropathogenic Escherichia coli will be used as a model system to study the molecular basis of bacterial attachment and host tropism. The goals are to provide experimental evidence for a model that takes into account the interactive surfaces of bacterial adhesions, structural proteins and assembly proteins. The P pilus family contains a number of adhesion variants (G adhesions) that are part of highly homologous operons that bind to different subsets of Gal alpha(1-4)Gal containing glycolipids. The relationship between the binding surfaces of the G adhesions and the overlapping but unique surfaces on the glycolipid isoreceptors that are recognized will be correlated to the specific interactions that account for the tissue and host tropisms. The structural and functional properties of the G adhesions will be elucidated by X-ray crystallography and by biochemical analyses using high pressure liquid chromatography. The assembly of an adhesion into the pilus tip requires an interaction of the adhesion with the chaperone protein, PapD and with one or more tip associated pilus subunit proteins. The known three dimensional structure of PapD will be used to design site-directed point mutations to identify the surfaces on the chaperone protein important in its activity. The carboxyl terminus of the G adhesion may form part of the surface involved in the interaction with PapD based on a biochemical analysis of PapG-PapD complex formation using G carboxy-terminal deletion mutants. The hypothesis that all of the pilus components contain structurally conserved surfaces through which PapD interacts will be tested by measuring the ability of carboxy terminal peptides of PapG to inhibit the PapG-PapG interaction in vitro and to inhibit pili assembly in vivo. The protein- protein interactions responsible for initiating pilus assembly will be studied in vitro by purifying the three putative proteins involved in this process (PapF, PapK, and PapC) for use in an in vitro assembly system analyzing the sequence of events that lead to the association of pilus tip subunits to the adhesion and the dissociation of the PapD chaperone. Another pap protein, PapJ and other proteins encoded outside of pap will be investigated for their role in assisting protein folding and assembly in these models. The hypothesis that G adhesions have evolved in different species to recognize different surfaces on the same or similar isoreceptors will also be studied. E. coli isolates expressing G adhesions from a panel of species will be tested for their binding specificities against purified glycolipids isolated from the same species and immobilized on microtiter wells, thin-layer chromatograms or incorporated into artificial lipid films. The major glycolipid isoreceptor recognized in vitro by each G adhesion will be correlated to the in vivo binding characteristics of the strain as measured against erythrocytes and tissue culture cells. In this way, the molecular basis of tropism in the urinary tract (human vs. dog, for example) will be correlated to the accessibility of the binding epitope when present in the eucaryotic membrane. This application will establish a model system for the development of chemotherapeutic compounds that interrupt novel assembly interactions or binding specificities to epithelial surfaces.
