Hi Thierry, We have worked on a number of proteins where the oligomerisation state differs from that found in solution. A recent example is a project we worked on involving the HC fragment of tetanus toxin (see Qazi et al., J Mol Biol. 2007 Jan 5;365(1):123-34) which crystallises as a monomer. The discovery of it multimerising was somewhat serendipitous in that the protein (like so many these days) was purified using a his-tag and then crystallised. We planned to use SAXS to study proposed conformational changes and looked at the protein on a native gel as part of our standard preparation for these experiments. The native gel showed the protein multimerised and the combination of the SAXS and excellent crystal structure from the Isaacs group led to new ideas about the function of the toxin.
A few points: 1. Crystallisation conditions can be very selective for different oligomerisation states of protein and other parameters ( e.g., pH, ionic strength, oxidation state, exogenous ligands,...) besides concentration can affect the equilibria controlling whether a protein appears to be a monomer or not as the case may be. 2. It is always worth looking carefully at native gels or other sizing data for any protein which is crystallised. With the great expression systems out there using affinity tags this sometimes gets forgotten. 3. Be aware that truncated forms of proteins (even small deletions) may affect oligomerisation states (we have also seen this with a membrane bound receptor we are working with now and I think there are other similar examples in the literature) best wishes, Kate
