Protein Crystallization Hits

Risky business or, how much protein sample do you need?

How much protein do you need? - Classic first question when a protein biochemist starts a collaboration with a crystallographer. How many milligrams do you need to grow protein crystals and determine the crystallographic structure of my target protein? A simple question that does not have a simple answer, unfortunately. Other than protein-specific reasons, the mg estimate depends in large part on the equipment at one's disposal. For example, the volume of individual crystallization experiments, the number of screens to test, minimum required crystal size, etc. There doesn't seem an end to screening possibilities. If you happen to work in a high-throughput protein crystallization environment it is straight forward to give an estimate for protein requirements for a first screening pass since the protocols and procedures are usually standardized. The high-throughput screening lab at the Hauptman-Woodward Institute for instance, asks for 600 microliters of protein sample at a concentration of ca. 10 mg/ml for an exploratory crystallization screen. That's 6 mg. Depending on your tools you may need 2-5 x that quantity to run a similar screen. But you may not even run that large of a trial and require less than 6 mg. Even in the best case where you obtain crystals from the first crystallization plate it may still be a long way - and many protein preps - to an experimental model on the computer screen.
It is the accumulation of the uncertainties in protein expression, purification and crystallization success that make risk assessment very hard for a single protein structure determination project. This is not at all like converting sheep wool into yarn and the yarn into a pullover. When you think about it, what we're doing in crystallization is akin to throwing darts. Fortunately there's one good rule to go by: the more darts you throw, the higher our chances to hit the bull's eye.