The work plan was divided into three main areas; (1) identify and characterise a selection of catch bonding protein complexes of interest, (2) develop a mild, site-specific methodology to functionalise polymer scaffolds with complexes of interest, and (3) to characterise the resulting hydrogels and their adaptability to mechanical stress and environmental changes.
Serine-aspartate repeat protein G (sdrG) from S. epidermidis and its complex with the beta strand of the extracellular matrix (ECM) protein fibrinogen (FgB), as well as the complex between pilus-related adhesin A (RrgA) from S. pneumoniae and the ECM protein fibronectin (Fn) were initially chosen as catch bonding (CB) complexes of interest, representing one of the strongest (sdrG-FgB) and one of the weakest (RrgA-Fn) bacterial CB complexes known at the time. However, initial experiments identified issues with RrgA-Fn as a model complex due to the weakness of the rupture forces and the similarity of the low and high force rupture regimes, which were more dramatically different in the sdrG-FgB complex. This was deemed to be more useful for eventual tuning of macroscale materials. Thus, focus was turned to exploring the limits of adaptability of sdrG-FgB and the effect of geometric and biochemical modifications to the sdrG structure, environmental and mechanical stress using single molecule force spectroscopy (SMFS). Here, we identified a strong effect of calcium concentration, geometric arrangement and point mutations, as well as the type of substrate (if we isolated the beta strand or used the full Fg protein as the complement) on the behaviour of the complex, while the expected dependence on pulling speed was remarkably small. This encouraged us to look more closely at cooperativity between the protein complexes because cells expressing CB proteins do have the ability to stick stronger as a function of applied force, but why not the individual proteins to such a great extent?
Instead of moving directly to the macroscale, we displayed the sdrG and its various mutants on the surface of yeast cells and used a technique called spinning disk microscopy (SDM), to apply shear stress to the cells after incubation on an Fg or FgB substrate and look at the cell distribution on the glass substrate as a function of shear force to determine the adhesive behaviour of the individual complexes. Here we saw the shear dependent behaviour we were hoping for, but this indicated that perhaps the sticking ability of pathogenic bacteria is a cooperative rather than a single molecule effect, something we hadn’t previously expected! We followed this up with particle imaging velocimetry and numerical modelling using computational fluid dynamics to further develop this theory, which further confirmed the importance of cooperativity but downplayed the importance of geometry at the microscale.
Finally to the macroscale. We successfully developed a mild methodology to ensure site-specific immobilisation of the proteins on aminated polysaccharides (primarily chitosan, but we were also able to aminate other polysaccharides and achieve functionalisation of the surface with the proteins) with reasonable reproducibility (issues with reproducibility were more linked to the variations observed in polymers of natural origin). We also explored embedding yeast cells with displayed proteins or protein-functionalised microspheres in inert polysaccharide matrices of alginate or gellan gum. This gave us detailed insight into the importance of the polysaccharide backbone; a weaker or more flexible polysaccharide made it easier to see the effect of the CB cross-links but gave us a material that was difficult to manipulate, while a more robust backbone gave better handleability but dampened the effect of the catch bond. A lot of time was spent optimising the perfect CatchGel until it was clear there was no perfect option, just the optimum for a specific application. Mass photometry, rheology, and AFM-microrheology were all used to come to this conclusion.
In summary, by employing single molecule force spectroscopy, spinning disk microscopy, rheology and a number of ancillary techniques to verify our observations, alongside numerical modelling we were able to investigate complex multi-lengthscale interactions and elucidate design parameters for the exploitation of mechanosensitive protein complexes on the macroscale.
