Crystallization Services

The MX core offers facilities and expertise for growing diffraction-quality single crystals of proteins, nucleic acids, small molecules, and complexes thereof. For all crystallization services a highly pure, monodisperse wet sample (protein, nucleic acid, etc.) or dry sample (small molecule) is required. The facility provides all the other necessary components and tools to conduct the experiments. The Protein Expression and Purification Core is co-localized in the same building and can assist you in generating a crystallography grade sample for structural biology studies on biological macromolecules.

Protein crystallization

We offer high throughput semi-automated crystallization screening for biological macromolecules.

• High-Throughput (HT) crystallization trials in a multi-sitting drop 96 well format setup with a Mosquito (SPT Labtech) liquid handler.
• Use as little as 11 ul sample per 96-reagent crystallization screen.
• Each 96 well tray takes about 3 minutes to set up.
• An Apricot liquid transfer robot (SPT Labtech) is available to facilitate rapid and accurate transfer of reagents from commercially available or prepared blocks into crystal screening plates.
• Recommended “Initial broad screening” volley:  Wizard 1-2, Wizard 3-4, Shotgun 1, JCSG+, BCS.
• Recommended “Further broad screening” group:  MCSG suite, JCSG core suite, PACT Premiere, Proplex, Index, PEG/Ion, PEG Rx, Salt Rx.
• More specialized crystal screens: AmSO4, Ligand-friendly (LFS), Berkeley, Precipitate Synergy 1 and 2, PEG/Ion 400.
• Deepest screening: Morpheus I, II, III, Slice pH.
• Contact us if you’d like to inquire about prices for having us perform crystallization screening on your sample(s).

Small molecule crystallization

We are actively developing small molecule crystallization as a service. It will have two pathways, one for water-soluble compounds and one for compounds soluble in moderately polar or nonpolar solvents. Please inquire if you are interested in this.

Crystal growth optimization

Temperature plays a very large role in sample crystallization. Our facility is equipped with numerous incubators to provide users with a lot of experimental options for optimizing crystal growth

• A walk in 4 Celsius room that is equipped with a Leica Stereo microscope and small work area.
• Three full sized temperature controlled incubators set permanently at 20, 16, and 12 Celsius.
• A half sized incubator that is meant for users who wish to freely experiment with temperature

UV fluorescence and white light imaging

We have a UVEX plate hotel (Jansi) with automated microscope that provides dual white light and UV fluorescence imaging of biological macromolecule crystallization experiments. It can distinguish between protein and salt crystals, helping you to prioritize the crystal hits more likely to lead to diffracting crystals. It can also help you identify protein crystals hidden in dense precipitates and decide if a crystal of poor morphology is worth using as seed material for second round screening. The UVEX is equipped with a CCD camera to acquire high quality digital images for presentation and publication.

The lab is also equipped with two Leica stereo microscopes, one mounted with a digital camera and polarizing lens. This represents another resource for triaging your crystal hits and acquiring quality images. The camera also provides 30 fps video that you can use in presentations or as a training tool when teaching lab members crystallography.

Crystallization FAQs

How much protein do I need?

– A typical experiment consists of 96 drops in an 8×12 array.
– Each drop consists of 150 nl protein mixed with 150 nl reagent.
– The spec of the instrument is 100 nl + 100 nl, but I don’t recommend going that low as it can miss drops.
– If you have more protein on hand, you can do larger volume drops. You can also do additional drops wherein you vary the protein:reagent mixture, such as a three drop setup: 200 protein + 100 reagent, 150+150, and 100+200 (each drop comes out to the same 300 nl volume).
– You need to allow around 25% loss with such minute volumes.
– So! How much protein needed: 150 nl * 12 columns * 8 rows * 1.25 = 18 ul.
– Typically we’ll set up crystallization screens in volleys of 4-8 such experiments to get good coverage of ‘crystallization space.’ Usually we’ll take the volume of protein you have on hand and decide how many screens we can get out of it.

How much protein, though?

– The general rule for an average 50-100ish kD protein is to have it at 10-15 mg/ml concentration. Lower for larger proteins (~5 mg/ml for ~400+ kD), higher for smaller proteins (~30 mg/ml for ~10- kD).

What’s the best buffer for my sample?

– The best buffer for your sample is minimalistic, just enough to keep the protein happy. E.g. 10 mM HEPES pH 7.5, 150 mM NaCl.
– Cofactors (ADP, GDP, NAD+), metals (Mg2+, Zn2+), and other necessities are ok, just use them as lightly as possible.
– Ligands for co-crystallization are ok, but their concentrations are highly dependent on binding affinity. You effectively have to work LeChatelier’s principle. KD around 100 nM, use 300-1000% (3-10-fold) excess. KD around 1 nM, use only 10-20% excess.

How do I co-crystallize a complex?

– Get each protein/macromolecule individually as pure as possible.
– If the proteins are the same MW, mix them together in a 1:1 stoichiometry. If they are different sizes, use an excess of the smaller one.
– How much of an excess depends on binding affinity. See above for that.
– Size exclusion on the mixed sample.
– If the proteins are about the same MW, SEC may not adequately separate the 1:1 complex from excess smaller component; but, if you make small fractions and a stringent selection, it could massively boost your chances of success.

Can I use a computational or threaded model in molecular replacement?

Using a computational model as the search in molecular replacement is fine. Any model bias (incorrectness) will be evident at that point, though it should be remedied immediately.