Skip to main content
European Commission logo print header

Structural Studies on the Interaction of Fibronectin and Collagen

Final Report Summary - FN-COLLAGEN (Structural Studies on the Interaction of Fibronectin and Collagen)

Final Report PIEF-GA-2009-235532

Fibronectin (FN) and collagen are two major components of the extracellular matrix. FN is constructed mainly from FnI, FnII and FnIII protein modules while collagen is made up of long triple helices assembled into fibres. FN/collagen interactions have been implicated in the control of fibroblast migration, collagen crosslinking and the removal of damaged tissue, but despite their clear biological importance, these interactions have not been well characterised. This project encompasses a thorough investigation of the previously elusive interaction between the GBD of FN and collagen I, the most common collagen type, on a molecular scale. We have identified four specific and conserved interaction sites that bind modules 8-9FnI of the collagen binding fragment (GBD) of FN (Diagram 1). We have found, in one case, that this interaction extends to include the second GBD sub-fragment, 6FnI1-2FnII7FnI, thus identifying, for the first time, a cooperative GBD binding site.

We have also studied the structure of the GBD and proposed a model based on the solution structure of 8-9FnI and the crystal structure of 6FnI1-2FnII7FnI (Erat et al., 2010). In contrast to a previously published GBD dimer under high [Zn2+ ] conditions (Graille et al. Structure 2010), we have good evidence that the GBD is a monomer under physiological conditions. The GBD structure has a 90° kink between 7FnI and 8FnI with restricted flexibility between 6FnI1-2FnII7FnI and 8-9FnI. Collagen binding stabilises this preformed, monomeric GBD conformation in solution. A local hydrophobic collapse, just C-terminal of the core 8-9FnI-binding collagen sequence, leads to a 90° change in peptide direction allowing the collagen peptide chain to interact snuggly with the 6FnI1-2FnII7FnI sub-fragment. We also observed that the weakest interaction site extends into 6FnI1-2FnII7FnI and forms a cooperative collagen interaction surface along the 2FnII7FnI interface. This extension of the interaction surface may have evolved in response to the weak binding at 8-9FnI to create a fourth more equipotent binding site.

Our results suggest that the GBD of FN is a partly elongated particle in solution with a preformed collagen interaction site, ready to influence the formation of collagen fibers in vivo. This provides further evidence to the view that fibronectin fragments fold to form distinct functional units that present unique coordination surfaces to their respective binding partners.

In our studies we have used a number of biophysical and structural biology techniques: The initial interaction sites were established by Nuclear Magnetic Resonance (NMR) chemical shift analysis. To measure dissociation constants for the different FN-collagen complexes we employed NMR as well as fluorescence polarisation anisotropy. The structures of the two GBD sub-fragments were solved by X-ray crystallography in the first half of this fellowship (Erat et al. PNAS, 2009, Erat et al. JBC 2010). Unfortunately, despite extensive trials the native conformation of the full GBD eluded crystallisation attempts. We therefore turned to small angle X-ray scattering (SAXS), a powerful method to study large biomolecules in solution. SAXS analysis revealed GBD to be a monomeric, relatively elongated, particle in solution. Quaternary structure modeling using these data and the crystal structures gave us insight into the molecular structure of the GBD in solution in its free state as well as bound to collagen. A manuscript with the conclusions of this work is currently in preparation.

In addition to my core work on the molecular interactions between collagen and FN, I established a fruitful collaboration with the laboratory of Katarzyna Barczyk at the University of Munster in Germany. Together we showed that the multi-ligand scavenger receptor CD163 promotes recognition, phagocytosis and killing of Staphylococcus aureus via binding of specific FN peptides (Kneidl et al. Cell. Microbiol. 2012). Furthermore, I was able to contribute to work in Jennifer R. Potts'laboratory on the interaction between the GBD and Streptococcus bacteria (Atkin et al, JBC 2010) and provide material and expertise to James Armstrong in the laboratory of Anthony Hollander at the University of Bristol, working on "New methods for introducing proteins to the surface of cells using gelatin binding protein". I also supported my host laboratory with expertise in X-ray crystallography and NMR (Kitagawa et al., Cell 2011/Mayer et al., JBC 2012). This enabled me to explore several potential future projects. Indeed the BBSRC recently awarded Ioannis Vakonakis (Investigator) and myself (Co-Investigator) a project grant on "Structural mechanisms of centriole assembly during cell duplication", and I started to work on this project at the end of my Marie Curie fellowship in October 2012.

I have presented my work in seminars at the laboratory and departmental level, as well as international conferences (NCCR Structural Biology Symposium, Zurich (Switzerland), BioNMR meeting Brno (Czech Republic)). I have also participated in several outreach activities: I am regularly involved in the Open Day for prospective students in our Department. Furthermore I have taught at the UNIQ Summer School, a free residential event that takes place each summer for Year 12 students currently studying at UK state schools, and for the Pathways Study Days at Lady Margaret Hall. These events have enabled me to share my enthusiasm for science with young students and, hopefully, to encourage them to take up a scientific career.

Apart from my salary costs, money from the fellowship was spent on consumables including NMR access charges and participation in workshops and meetings. In particular, I attended an EMBO Practical Course on'Solution Scattering from Biological Macromolecules'from October 25 – November 1, 2010, at the EMBL in Hamburg. My subsequent work on FN and collagen profited enormously from this course as described above. To carry out SAXS measurements I had to travel to the only available BioSAXS beamlines in Europe, at the ESRF in Grenoble and at DESY in Hamburg. A substantial amount of money was spent on fluorescently labeled collagen peptides that were essential for the determination of dissociation constants for the FN-collagen complexes. I was, fortunately, able to supplement these costs by applying for and receiving a grant from the Royal Society to cover the costs of peptide synthesis for 1 year in 2010. In October 2011 I gave birth to a son and took six months of paid maternity leave. Because the Euro was performing rather badly against the British Pound, my salary had to be reduced by ?171 per month from May2012. Otherwise there were no significant deviations in budgeted and actual spending