Mike, This is a complicated question :) The floppy bits can contribute to packing in several non-obvious ways.
I may be able to help you get a better crystal form, if you're willing to share the specifics of the protein (which will be kept confidential of course). I've developed new procedures for rational protein crystal engineering and now need more Guinea pigs to test them. Artem > Greetings, > > I've taken a protein (structure known) with some disordered termini and > created a new truncated construct that lacks these disordered bits > (approx. > 5% by mass of the total protein). > > To my delight, the new construct crystallizes much more readily than the > full-length one (more hits in screening, nicer looking crystals, etc.) > However, while the crystallization is more robust and the crystals look > much > "nicer", they don't diffract as well as the full-length crystals (go from > around 2.7 angs for full-length to 3.2 or worse for truncated construct). > > While the truncated form gives many more crystallization hits than > full-length, optimization always gives the same crystal form (same space > group and unit cell dimensions), no matter which conditions are used. > > Given that I've removed floppy bits from my protein, I would have expected > improved diffraction. But since I keep getting the same unit cell, I > wonder > if I've just increased the solvent content in my truncated construct > crystals (less mass per same unit volume). > > Does this rationale make sense? Is it possible that the disordered parts > I > removed were in some way restraining the ordered parts of the structure? > Should I keep searching for a new crystal form with improved lattice > packing? btw, I have tried several different approaches to dehydrate the > crystals and none has successfully improved diffraction. > > Thanks for your help. > > -Mike >
