Protein crystallization by intelligent design?
If they awarded a prize for best seminar title, Zygmunt Derewenda would win it, hand?
If they awarded a prize for best seminar title, Zygmunt Derewenda would win it, hand?s down. According to the abstract book for the Keystone Symposium on Structural Genomics, his seminar was to be entitled "Protein Crystallization: From Art to Science." But the University of Virginia researcher decided that was a bit too provocative, so he opted for a more "neutral" title: "Protein Crystallization by Intelligent Design."
Derewenda's point, of course, is that crystallization isn?t random but rather stems from a series of procedural decisions, and as this symposium demonstrates, those decisions are getting harder and harder for the average structural biologist to make. Some proteins crystallize easily under standard conditions. But many do not. Over the past two days we?ve heard discussion of a number of strategies for dealing with such proteins, including adding a nucleant, making N- and/or C-terminal truncations, and Derewenda?s strategy, surface engineering.
linkurl:Surface engineering;http://nihserver.mbi.ucla.edu/SER/ is an approach in which patches of lysine, glutamic acid, and glutamine on a protein?s surface are changed by site-directed mutagenesis to another residue ?- typically alanine, though tyrosine, threonine, and histidine also work. Derewenda calls this strategy "shaving the surface" of the protein, in that the targeted residues all have long, floppy side chains that could inhibit crystallization.
Derewenda claims a success rate of between 60% and 80%. But of course, advocates of competing techniques can put up equally impressive numbers of their own. So faced with a recalcitrant protein, what?s a structural biologist to do?
Right now, they must rely on their best judgment. But perhaps not for long: One theme emerging at this meeting involves the need for benchmarking of these various procedures. That would allow the scientists to base their decisions on hard data, instead of on anecdotes and hype.
**Follow-up note**: A few weeks ago linkurl:I blogged;http://www.the-scientist.com/blog/display/22940/ about Naomi Chayen?s development of a possible universal nucleant. She presented those results this evening, and noted that the company commercializing this technology has now developed a crystallization plate with the nucleant already added. That should pave the way for adoption of the technique into structural genomics pipelines.