Protein Crystallization Hits

All drops clear? Back to Old School protein crystallization

Here's a somewhat unusual crystallization situation that I was asked to give advice on during the ACA meeting: All drops are clear, after multiple crystallization trials were set up with popular sparse matrix crystallization screens. Not a single hit was obtained. No precipitation, only clear drops. I have to say that this is quite a perplexing situation and I've never experienced such a case myself.

So, here are my thoughts. After double checking that the protein concentration was indeed high (>100mg/ml, feels viscous), this looks like the protein molecule in questions does not tend to aggregate with itself. So, what's there to do if there are only clear drops?

I suppose this is a demonstration where conventional sparse matrix crystallization approach fails. Going back to 'Old-School' crystallization may help. Not that I've seen those times myself ;) but I've heard - and tried myself - that before the advent of commercial sparse matrix screens, protein crystallization trials were set up in a rather systematic way. The primary goal was to first characterize sample precipitation behavior and then work towards conditions that convert non-productive aggregation to assembly of a crystal. The practicalities run against today's common practice: select a small set of precipitation reagents and then optimize, optimize, optimize. Very high concentrations of good precipitation agents are a good start. Salts: Ammonium sulfate, Ammonium Phosphate, Sodium Chloride, Sodium Acetate, Sodium Citrate. Organics: MPD, Isopropanol, PEG 3350.
Typically you'd prepare saturated solutions of the salts, keeping un-dissolved salt in the bottom of the bottle. If you've never done this before, be prepared for a shocker, there's *a lot* of Ammonium sulfate that can be dissolved in water. More than 75 g in 100 ml of water. This table for salt solubility may come in handy where there's no Merck's Index around. You'd then take liquid out of that saturated solution, dilute to the desired % saturated solution (e.g. 50%, 75%, 100% saturated) and set up a batch-type crystallization experiment (no reservoir solution to equilibrate against). To keep the work load at bay I'd just set up three salts and MPD, each at 3 concentrations.
One of these highly concentrated solutions ought to precipitate the protein (if not I'm at my wits end). Carefully inspect the precipitate - any birefringence there? Sure, that these are no salt crystals? Once protein precipitation (or oiling out, for that matter) is established, you can pull out all your optimization tools at hand. Optimizing by adding components, varying pH, temperature, adding back precipitate to all the clear drops for seeding and so on.

Another approach would be changing the sample buffer to a more unfavorable one. The batch crystallization experiments may give a hint at the composition of the sample buffer to transfer the protein into before setting up a conventional crystallization trail.

Does this help?
Peter