Anthony Blake, Chris Gell, David Brockwell, John Clarkson, Godfrey Beddard, John Trinick, Sheena Radford, Alastair Smith.

Recent advances in instrumentation have allowed single molecules to be both manipulated in a controlled fashion and their properties to be quantified. Laser tweezers and the atomic force microscope (AFM) have been used previously to unfold proteins by the application of force by the separation of their N and C termini. Upon returning to their original position the protein refolds to the native state. Mechanical unfolding therefore allows proteins’ response to mechanical stress to be studied at a single molecule level. This process has been used extensively to study various constructs of the giant muscle protein titin and other proteins which have been selected for the ability to withstand shear forces. This process should also be useful in the field of protein folding i.e. how an amino-acid chain selects a single native state over the millions of conformations available to it in space. The study on the single molecule level should enable details to be elucidated which are usually masked in ensemble studies, show rare events and give an idea of the range of structures that make up the denatured, native and intermediate (if any) states.

Using the FRET technique, single molecule fluorescence can be used as a ‘molecular ruler’ allowing the separation between two extrinsic fluorophores to be used as a reaction co-ordinate which can monitor the folding of a protein molecule in real time, and deviations from the native state.

Our aim is to study protein folding by using both mechanical unfolding by AFM and single molecule fluorescence and then ultimately to join the experiments.

Figure 1 Above: Cartoon showing the design of the concatamer, at the DNA level. Arrows show the positions of unique restriction sites which define the cassettes. The hexa-histidine tag allows one step purification. Two C-terminal cysteines allows for attachment of proteins to a gold surface.
Below: Cartoon showing stepwise single molecule unfolding of an I27 concatamer. Proteins are non-specifically adsorbed onto the cantilever (top); upon retraction from the gold surface (bottom) the cantilever bends (point A) and the force is measured until one domain unfolds (point B) and the sudden increase in length reduces the measured force (point C). Further retraction of the cantilever causes the unfolded domain to straighten causing the force to gradually increase (point D) up to a point where the next domain unfolds.

Mechanical unfolding

A ‘concatamer’ consisting of five copies of a single domain from titin (I27, a b-sandwich protein belonging to the Ig family) has been engineered (see figure 1). The cassette strategy used allows each copy to be easily substituted with a mutated I27 or even a completely unrelated protein. This modular approach therefore allows us to ask a range of different questions about proteins' response to mechanical stress. Firstly we have generated a concatamer consisting entirely of destabilised domains. By measuring the unfolding force dependence on cantilever retraction rate, information can be obtained on the unfolding kinetics and the unfolding transition state placement. Comparison of this data with the same parameters derived from traditional chemical denaturation studies will show whether the two techniques give information on the same unfolding pathway. The second phase of this study is to insert proteins into the concatamer which have not evolved to withstand mechanical stress. Repeating the same process as described above should reveal whether all proteins behave in a similar manner to I27.

The second area of research involves pulling elastomeric proteins such as elastin, byssus, dragline and the PEVK region from titin. These proteins store energy when deformed without rupture then recoil back to their original state. The usefulness of this approach is being assessed by initially studying the PEVK region from cardiac titin. This natively unfolded protein has been cloned into the concatamer and mechanical unfolding experiments in which the ionic strength of the buffer is changed are underway.

Finally, by attaching different protein domains at each end of the concatamer, which are differentially post-translationally modified in a unique manner, we hope to develop a surface attachment strategy that allows specific, directional tight binding of proteins to surfaces.

We thank the University of Leeds and the BBSRC for funding this research.