Conformational Heterogeneity: One of the things that can inhibit crystallization is conformational heterogeneity – where the sample is pure (as shown either on a gel or by mass spec), but the individual molecules in the sample are not in the same conformation. In particular, parts of the protein chain may be more or less well ordered.
What is Proteolysis? The idea behind limited proteolysis is to find a fragment (hopefully a large fragment) of your protein construct that is stable to proteolysis. We make the assumption that the resistance to cleavage is the result of being well folded – interior proteolysis sites are sterically unavailable to the protease.
What might get cut are the ‘floppy bits’ – termini and badly ordered loops which contribute to the samples conformational heterogeneity. The impetus for trying limited proteolysis comes from the success of the method in rescuing a number of protein samples, popularised by the Structural Genomics Consortium high throughput laboratory in Canada.
Which Protease? There are lots of different proteases one could use, and lots of different ways of doing this. Ideally, one would find a protease that gives a large, stable fragment on overnight protease exposure, then use some analytical tools to work out what part of the protein chain you have; and make that construct.
A much quicker way of testing the method is available in C3 – we have lyophilised 10 µg of protease into our standard PCR tubes. When we come to run a sample, we add it to a pre-treated PCR tube, and set up as usual (simply type this into the comments section of the batchload wizard). The lyophilised protease re-dissolves, and slowly chews away at floppy bits of the sample while waiting for crystallization. We have three different proteases available:
- Chymotrypsin (the classic)
- Trypsin (a little more aggressive, generally)
- Thrombin (often useful for cleaving off purification tags in-situ)
Reference: Dong, Aiping, et al. In situ proteolysis for protein crystallization and structure determination. Nature Methods 4.12 (2007): 1019-1021.