Wednesday 14 May 2008

Gabriel Waksman on bacterial secretion systems

Gabriel Waksman, head of the School of Crystallography at Birkbeck, has given a seminar in Birkbeck's School of Biological and Chemical Sciences describing his group's work bacterial secretion systems. Gabriel studied at the University of Paris and then spent the first part of his academic career in Washington University in St. Louis, Missouri, before joining Birkbeck in 2003. He now combines his position here with a chair in Biochemistry at University College London and directs the new Institute of Structural Molecular Biology which links the research in the two colleges.

Gabriel's research group works mainly on elucidating the structure and function of bacterial secretion systems, which produce hair-like appendages, or pili, on the surfaces of Gram negative pathogens. These are very important for infection as they allow bacteria to recognise, and then attach to, host cells. Mutant bacteria in which these proteins are not expressed are not pathogenic.

Two of the best understood of these systems are found in variants of E. coli which infect the human urinary tract. The P pilus recognises and binds to kidney cells, causing the kidney infection pyelonephritis, and the type I pilus recognises bladder epithelial cells and causes urinary infections. Both pili can be found on the same bacterium. The structures of both pili are similar, consisting of many protein subunits; it is the tip subunit that recognises its target cell types by binding to different cell surface sugar residues.

All subunits associated with a particular pilus are encoded by genes within a gene cluster and named accordingly: the P pilus genes are Pap genes and the type I genes Fim genes. Structures of at least one representative of each type of subunit have now been solved, many by Gabriel's group and its collaborators. Knowing these structures has enabled the group to understand the mechanism through which the pili are formed.

The pilus subunits that polymerise to form the main part of the fibrous structure all have similar structures. They are immunoglobulin-like, mainly-beta structures in which one sheet is lacking a central beta strand, and they are therefore unstable independently unless they are bound to a chaperone protein. A strand from the chaperone fits into the gap, forming regular hydrogen bonds with the neighbouring strands, and this stabilises the chaperone-subunit complex. This is then transported to the growing pilus, where the N terminal peptide from a subunit already in the structure replaces the chaperone strand in the new subunit, adding it to the polymer via a mechanism called "donor strand exchange". The resulting fibre therefore consists of a string of similar subunits, with the N terminal peptide of one subunit forming a strand in the central beta sheet of the previous subunit in the assembly.

The Waksman group's most recent structural studies concern the protein through which the pilus is assembled, known as the usher. This is a mainly beta membrane protein (link is to material in PPS section 11, which will be released next week) which is embedded in the E. coli outer membrane. The structure of the E. coli P pilus usher, solved by X-ray crystallography, shows the beta barrel and a middle or plug domain which interrupts the main beta sheet of the barrel. With 24 strands, it is the largest outer membrane beta barrel protein structure to be elucidated so far. In its inactive form, the plug domain fits inside the barrel, completely blocking it. They also used cryo-electron microscopy to isolate the structure of a type I pilus complex during pilus assembly. The usher forms a dimer within the cell membrane but, interestingly, the EM studies show that a pilus is secreted through only one monomer of the dimer.

This is very complex work which can only be touched on in a blog post. If you would like to know more, have a look at a few of these papers (links to abstracts in PubMed):