Protein Crystallization Hits

Darren Begley recently put up an interesting post about "Grooming your fragment library: Checking the quality of commercially-sourced fragments". 

The point he's making is that tight QC measures are key to the success of fragment screening: 

"These results highlight the importance of implementing tight quality control measures for preliminary vetting of commercially-sourced materials, as well as maintaining and curating a fragment screening library. It also puts forth a statistical likelihood of around 10-15% failure, regardless of vendor. Most importantly, we have seen our methods reduce risks while accelerating drug discovery. "

This is certainly true for NMR-based experiments and of course also applies to crystallography-based fragment screening.

I was glad to see Derek Lowe pick up the discussion in this blog post: Not What It Says On the Label, Though

Several of the comments to this post support Darren's findings and drive this message home:  Buyer beware!

Peter

I have to admit, it is somewhat counterintuitive that crystallizers try to crystallize proteins by finding a milieu with low protein solubility by starting with a protein solution and then dilute it. How do you lower solubility by dilution - usually 50:50?. The trick here is of course, that the protein's solubility in the new medium (i.e. the crystallization cocktail that is used to dilute the protein with) is even lower than that in the starting solution. Hence, it seems reasonable to start with a protein buffer that allows to concentrate the protein as high as possible. One way to get to such a buffer is to enhance it for a specific target by increasing the solubility of that target protein. It turns out that this is indeed a practical way to increase the success of crystal screening, mainly due to enhanced nucleation.

Aude Izaac et al. describe a simple procedure to go about such customized buffer selection:

Izaac, A., C.A. Schall and T. C. Mueser (2006)
Assessment of a preliminary solubility screen to improve crystallization trials: uncoupling crystal condition searches
Acta Crystallographica Section D: Biological Crystallography D62, 833-842.

It works by first precipitating (ugh!) the target protein with PEG and then testing its resuspension in a variety of buffers and salts. Here's the protocol :
At first 150 ul of protein solution is precipitated by PEG 8000 (ad 190 ul 40% PEG 8000) at room temperature and spin in aliquots. Then 2 ul of 1 M salt or buffer solutions are added to 18 uL to test if the protein can be re-suspended. After centrifugation (20 ul total vol) the supernatent is tested for protein content. The higher the protein concentration in the sup, the more potent the resolubilization of the buffer for your specific target protein.
In a second step the best salt/buffer combos are determined. The salt and buffers that yielded the largest effect are combined at a standard 100 mM salt and 50 mM buffer concentration and compared to the standard chromatography buffer (50 mM Tris-HCl, pH 7.5, 100 mM NaCl). Then target protein samples are diluted in the new buffer and concentrated using a 10 kDa MWCO filter. The winner is the buffer that allows the protein to concentrate to the highest level without precipitation.

Fig: What goes up must go down.

Pretty nifty,
Peter

Concentrating the protein sample is the last, and sometimes nerve wrecking step prior to setting up crystallization trials. I've heard the outcries of desperation reverberate through labs and hallways when the final sample precipitated in the concentrator device after a week of purification work. Concentrating proteins to higher than 10 mg/ml is risky business but necessary to generate a supersaturated protein solution and coax it into the metastable region of its phase diagram.

Here's a way, I think, would avoid the concentration step altogether:

Employ a modified version of the Dilution Method (see here or here) that is based on this paper:

Dunlop, K.V. & Hazes, B. When Less is more: a more efficient vapour-diffusion protocol. Acta Cryst. (2003). D59, 1797-1800

In a traditional vapour diffusion crystallization experiment, equal volumes of protein and well solution are mixed (i.e. 1uL + 1 uL). In the subsequent equilibration period, the drop shrinks from 2 uL back to 1 uL due to the vapor transfer. This brings the protein back to its initial concentration but in a different physical chemistry environment. The hope is that this different environment has a lower solubility for the protein and hence crystal nucleation and growth ensues form this supersaturated solution.

Using the Modified Dilution Method one would dilute the precipitant solution by for example 5-fold with water and use it to set up a 1 uL + 1 uL hanging drop or sitting drop. Note that the protein would not be diluted as described in the paper. Let's say the concentration of the protein at hand is ca. 2 mg/ml in a low ion strength buffer (or approx. 4 OD worth of sample)

What happens during evaporation? The excess water in the drop evaporates and shrinks the drop from its initial 2 uL to ca. 0.2 uL (the precise final volume would of course also depend on the composition of the protein buffer, lower ionic strengths protein buffers would be preferred with this Modified Dilution Method). With this reduction in volume, the protein concentration increases from the initial 1 mg/ml in the combined drop to 10 mg/ml. 

'Back-of the envelope' sketch for the Modified Dilution Method for vapor diffusion protein crystallization experiments to avoid excessive protein concentration. Due to the initial dilution of the reservoir solution by a factor of 5 (prior to setting up the drop), the drop volume is reduced by vapour transfer until the reservoir solution is in equilibrium with the drop. In this case the drop shrinks by a factor of 10 from 2 uL to 0.2 uL, hence concentrating the protein from 1 mg/ml to 10 mg/ml. Note the use of Wizard I ;)

Since the crystallization regime of such crystallizations is different from conventional vapor diffusion based crystallization, such a dilution scheme could also be applied as a low-cost and simple way to optimize crystal growth.

Accurate manual nanoliter crystallizations prepared with realtively low initial protein sample concentration.

Try to beat that!

Peter

P.S. - let me know if it works.