Crystal harvesting out of native crystallization drops (hanging- and/or sitting-drop geometries) can sometimes be severely hampered by the propensity of crystals to stick (sometimes very tightly) to their crystallization surface. Here are some tips on how to approach the problem:
OPERATE ON THE PLASTIC RATHER THAN ON THE CRYSTAL!
Take a sturdy needle (like one of the microneedles from a Hampton kit, or a very thin syringe needle, or an accupuncture needle) and stick the needle into the plastic a bit away from the crystal. Push hard. If you’re using polarizers, you may be abe to visualize the stress forces in the plastic by the shifting of the colors. The basic idea is to stress the plastic under the crystal without touching the crystal in any way. By digging and twisting, one can generate little movements across the plastic which can be enough to free the crystal. Sometimes you have to push quite hard, and to wiggle the needle a bit. Beware, however, as the needle can slip and ruin the crystallization drop.
COAT THE CRYSTALLIZATION SURFACE WITH A THIN LAYER OF GREASE
If working with the 'sitting-drop' geometry, one can coat the sitting drop depressions with a thin layer of vacuum grease. You only need a very thin layer of the grease (i.e. keep wiping off with a KimWipe until the grease is almost completely gone). Upon crystal harvesting, the crystals will have the tendency to slide right off the grease. One may opt for Vaseline or petroleum jelly rather than silicone-based grease.
COAT THE CRYSTALLIZATION SURFACE WITH SILICON
You can try various siliconizing fluids such as AquaSil (Hampton Research). Such a product will not melt the plastic of the crystallization plate as opposed to e.g. Repelcote.
CHANGE CRYSTALLIZATION PLATE
One may want to try plates made from COC (cyclic olefins), such as the "UVP" plates made by SwissCi. They are less sticky than polystyrene plates. COC is halfway between polystyrene and polypropylene. Polypropylene is even less sticky than COC but is not rigid, therefore not recommended for crystallization robots. You can get plates made of "clarified polypropylene" from Emerald, and you can also get polypropylene "bridges" that you place in Linbro wells. I think Hampton still sells them.
If you have good and bad crystals in the same drop, one may try to push crummy crystals into a good crystal, thus creating the necessary leverage to release the useful crystals from their growth position.
CRYSTALLIZATION WITH AGAROSE AS AN ADDITIVE
Crystals grown in the presence of 0.1-0.2 % (w/v) agarose will grow inside the soft agarose gel. Therefore, they are mechanically protected and will not settle to the bottom of the sitting-drop well. When you harvest a crystal cut generously around it with a microtool, pick it up (e.g. using a nylon loop) and do not mind if some agarose comes with it.
Reference: Biertmpfel, C.; Basquin, J.; Suck, D. & Sauter, C. Crystallization of biological macromolecules using agarose gel. Acta Crystallogr D Biol Crystallogr, 2002, 58, 1657-9 PMID: 12351881
'FLOATING-DROP' CRYSTALLIZATION METHOD
References: -Application of a two-liquid system to sitting-drop vapour-diffusion protein crystallization. -Adachi, H. et al, Acta Cryst. (2003) D59, 194-196 Promotion of large protein crystal growth with stirring solution. Adachi, H. et al. Jpn. J. Appl. Phys. Vol. 41 (2002) pp.1025-1027 -Two-liquid hanging-drop vapour-diffusion technique of protein crystallization. Hiroaki Adachi et al. Japanese Journal of Applied Physics. Vol. 43, No. 1A/B, 2004, pp.L79-L81.
THe necessary reagents for applying this method can be found at: http://hamptonresearch.com/product_detail.aspx?cid=4&sid=185&pid=32 HR2-797 100% Fluorinert FC-70 Fluid 100 ml
This method should allow better control of nucleation events in case the crystallization surface.
Put a small piece of dry ice on the opposite side of the plastic from the crystal. Perhaps the difference in thermal expansive coefficient will let the crystal(s) break away. Don't overdo it though.
‘IN-PLATE’ DIFFRACTION EXPERIMENTS
Beamline FIP-BM30A at the ESRF (contact Jean-Luc Ferrer) has a setup that allows crystal testing, even data collection, while the crystal remains in the crystallization drop.
The MiTeGen microtools kit: http://mitegen.com/products/microtools/microtools_kit1.shtml comes with a "MicroSaw", which is a 10-micron thick kapton saw that is intended for this purpose. That is, you don't pry the crystal off the surface, but rather rest this saw against the surface, bring it over to the edge of the stuck crystal and then work it back and forth until you have cut underneath the crystal.
Add a drop of 10ul mother liquor to the crystallization drop. Sometimes, the crystals attached to the plastic would float to the surface. If not, take another 10ul and pipet it in and out the drop as close as possible to the bottom. This might dislodge additional crystals from the plastic.
One may want to sonicate crystals off its crystallization surface; apparently this was the standard approach for papain when Jan Drenth determined the structure a long time ago.