Crystal growth: Tips and Tricks

Getting crystals

I tried Hampton Screen I & II with my protein and have no crystals. What can I do?

1. There are other screens. While it is proably true that when proteins do crystalize, they in many cases produce hits from Hampton screens I or II, we are dealing here with exception to that rule. Lists of screens are available from manufacturer websites:
   • Hampton Research
   • Qiagen (Nextal)
   • Emerald Biosystems
   • Jena Bioscience
   • Other screens
2. Vary protein concentration. This is rather obvious suggestion when all your drops are clear or all have precipitate, in which case your overall concentration is too low (too high). More elaborate approach is to modify concentrations based on the results of your first screen, i.e. double it for the conditions which produced clear drops and cut it in half for those which produced precipiate. To avoid concentrating/diluting protein, you can simply try to mix your original protein stock with reservoir solutions in different proportions. For instance, mixing 2ul of protein with 1ul of reservoir solution is (to some degree) equivalent of doubling protein concentration.
3. Set your screens at different temperature. 25 and 4 degrees are the most popular options. If you completely lost hope for your original screens, just transfer them to the cold room. Protein solubility will change and you may get crystals.
4. Add a purification step. Purity can be critical to protein crystallization and the purity judged from SDS-PAGE or even gel filtration can be misleading.
5. Modify your protein.

Improving crystals

Reducing the mosaicity

Things to try in the order crystallization-related / cryo-related / geometry-related:

1. change precipitant, pH, temperature; try additives
2. try co-crystallization with glycerol (2%, 5% or 10%).
3. incubate your crystal longer in your cryo/stabilization solution.
4. optimize the cryo conditions
5. anneal (freeze/thaw) the crystal - DISCLAIMER: your mosaicity can go up as well as down with annealing
6. the mosaicity may be anisotropic, so try to orient the crystal in the loop such that you can shoot along a different axis
7. try smaller crystals (often growth faults lead to end of crystal growth)
8. use a synchrotron beam (as it is less divergent than the usual home source, and the apparent mosaicity is the sum of the crystal mosaicity and the crossfire)
9. shoot at parts of the crystal to find if there are good and bad parts (this requires a small beam)
10. look at Elspeth Garman's papers, and/or check her talks at RapiData workshops (anybody with a link??)