Protein Crystallization Hits

PDBsum rocks for crystallization protocol

This week I came across a message from PDBsum letting us know that certain figures and captions of a paper we had published in 2008 (Gerdts et al. (2008). Acta Crystallogr D Biol Crystallogr, 64, 1116-1122.) had been included into PDBsum. I had not visited the PDBsum site before and was at first intrigued and then positively surprised about the wealth of information that was presented on a protein structure (3cxk). This is a high-density, information rich way to get a quick impression on a protein structure and other accessory information.

Figure: Screenshot of the PDBsum entry for 3cxk. The crystallization experiment is nicely referenced in PDB-sum.

What I liked in particular about this presentation is that the crystallization experiment becomes part of the story. Our paper described an earlier, beta version of the MPCS - the plug-based nanovolume microcapillary protein crystallization system. Since the publication of the paper in 2008 the technology has matured substantially (check out the New Product Award 2010 that the PlugMaker has received last week).

Any context that goes beyond just reporting the precipitation reagent helps. Having such exquisite detail available when trying to reproduce protein crystallization experiments is often necessary to build on published research.

Way to go EMBL-EBI!
Peter

Seeding with protein 'crap': Microseed Matrix Screening

What's the new thing that people are tying out these days? Lots of new methodologies, ranging from low-volume plug-based crystallization (of course) to new crystallization screening matrices for membrane proteins. I've noticed that there's a 'new' seeding method around that has come up in several conversations I've had with protein crystallizers over the past 3 years or so. It's called Microseed Matrix Seeding. Judging from people who try and stick with it, it is my impression is that this seems to be working rather well.

What is Microseed Matrix Seeding in practical terms?

You start with a 'failed' primary protein crystallization tray and:

1. Harvest some or all the precipitated drops, pool them (yes!) and call this seed stock.
2. Spike each drop of a new, secondary crystallization trial with a portion of the seed stock.
3. Obtain crystals from a protein/formulation combination that is different from that you used to create the seed stock with.

The Matrix Microseeding method and its application to yeast cytosine deaminase was first described by Gregory Ireton and Barry Stoddard in:

G. Ireton & B. Stoddard

Microseed matrix screening to improve crystals of yeast cytosine deaminase

Acta Cryst D60 (2004), 601-605.

Then Alan D'Arcy picked up on this new method and initiated a robotic application for this new protein crystallization seeding method:

A. D'Arcy, F. Villarda, M.Marsh

An automated microseed matrix-screening method for protein crystallization

Acta Cryst D63 (2007), 550-554.

Seeding with 'crap'? (mind me - not my own words, but I've heard this very question more than once)

Maybe not. What you see as precipitated material may be not properly characterized crystalline material. For all I know, there could be sub-micrometer sized microcyrstalline protein material mixed in the precipitate. And there's just no way for you to see that. Alternatively, the precipitated protein material itself may form a heterogenous nuclation surface in similar ways that seaweed or horse hair can serve as nuclei for protein crystallization.

A case for Microseed Matrix Screening(?) 

If you you've got many drops with precipitation in it (B), and no crystallization leads whatsoever - why not try give it a try?

Thanks,

Peter

Best Protein Crystallization Paper 2009

Guess what - in my opinion there was not a single paper published in 2009, covering the topic of protein crystallization review-style,  that comes close to Naomi & Emmanuel's collected wisdom published in 2008:

As much as I prefer open-source publications, this is one of the best. Nature Methods, Vol. 5 No.2, February 2008. Naomi Chayen and Emmanuel Saridakis.

Why?
Because it covers all the basics: Setting up initial protein crystallization trials and classical screening methods, deals with the question how many trials should be set up, what ideal volumes are and what to do if protein crystallization trials fail. In addition, Naomi Chayen and Emmanuel Saridakis lucidly explain the choices of available crytallization formats and go into a fairly detailed description of practical ways to influence the crystallization process to grow protien crystals that are sufficient for X-ray diffraction . Protein crysallization optimization is explained in a hands-on fashion, extensively referencing the body of published literature. A joy to read.

I've saved this paper in my archive using the file name Naomis_bible_paper.pdf

Happy new 2010!

Peter

 

Facing Protein Crystallizers' Remorse

The end of 2009 is near and it's time to clean up to make room for new and exciting projects in 2010. I'm just returning from the lab with a stack of protein crystallization trays to toss. Some of the crystallization experiments were prepared more than a year ago! There were two categories of crystallization trays that I dealt with:

1. Trays set up with protein that have never yielded any crystals at all
Should I keep them and hope that via slow desiccation or proteolytic cleavage protein crystals will eventually form? Nah! Everything that's older than 6 months must go. Gone they are. - I'm having second thoughts though, now that I'm writing this. I could open the crystallization chambers for a while and let the drops dry out just a little and then close them again. After all, protein crystallization by dehydration does work sometimes. Or move the trials to a different temperature? Or add chemotrypsin to the drops and create target fragments that crystallize.
Crystallizers' remorse setting in big time....

2. Trays set up with protein that have yielded crystals and structures
I could throw them out altogether. Crystals diffracted, structure is done. But, why not keep a few trays with crystals? You never know when a project 'comes back' - with the need to co-crystallize together with a small molecule ligand or protein partner. Even if the crystals don't diffract they may be useful and serve as seeds.

Pretty (Ammonium sulfate) crystals. These are the easy protein crystallization setups to give up on and throw out.

Phew! - that was easy, actually.

2010 here I come!
Peter

 

Happy Holidays: Plenty of slack to fight the curse of dimensionality in protein crystallization

In many high-throughput protein crystallization laboratories protein crystallization is dealt with as a standard process that usually goes like this:
1. Search for protein crystallization condition using a standard array of crystallization reagent matrices (BTW - many labs use Wizard I, II, III and IV for such a first pass ;). These experiments are of the trial-and-error type, sparse matrix screening that is akin to shooting in the dark. Protein crystallizaiton trials are set up with volumes as low as the liquid dispensing tools at hand allow and plates for sitting (and, less frequently) hanging drops in fairly standardized way (i.e. combining 300 uL protein solution + 300 uL precipitant). Plates are then stored in the dark at one or two temperatures and inspected after a day, several days and weeks. Once crystals or crystal-like objects are identified, the next step concerns
2. Optimizing the protein crystallization condition to grow better X-ray diffracting protein crystals. At this stage the repertoire of protein crystallization optimization procedures explodes, and one needs to consider available resources, experience and the protein at hand to apply them best. Popular crystallization optimization procedures include:
• Systematic grid screening
• Temperature variation
• Additive screening
• Seeding

The good news about these optimization procedures is that they are systematic and the results of the crystallization optimization experiments are often informative because they point to trends. For instance, you can identify a temperature that produces larger protein crystals. However, these optimization experiments need to be carried out in a fairly disciplined way, as one dimensional variations of a single parameter, keeping all other parameters constant. The alternative would be screening an astronomical number of crystallization conditions:

There are only a few points in the multi-dimensional protein crystallization space that one visits in typical protein crystallization experiments. Looks daunting, doesn't it?

And here's the problem with this scheme (not that I have a solution to it, but I'd like to point it out nevertheless):
By screening just one dimension at a time, large spaces of the multidimensional crystallization phase space remain unseen. This I call the Curse of Dimensionality in protein crystallization. I'm wondering if there are there any practical ways to getting around it. Any ideas?

There's no need to get too tripped up with this, though. The good news is that most proteins can be made to crystallize. In my mind this means that we have quite some slack in defining protein crystallization conditions.

Happy Holidays,
Peter

Seeking Protein Crystallization Guidance

This is somewhat of a response to an issue brought up by Sean  over at p212121. He left a remark, saying that there is only little to find in the technical literature discussing the practical side 'if this doesn't work, try this'. Having thought about it for a while, I think that the literature may not be the best place to look for such information.

Nevertheless, there's some 'tips-and-tricks'-type guidance on the topic of protein crystallization and crystallography as a whole. Protein crystallization in particular is a moving target since technologies develop and bottlenecks have shifted over time. Also, I've noticed that there's been a gradual shift in some publications picking up more of the detailed nitty-gritty. My all-time favorite of such publications is the "Methods in Enzymology" series that are a treasure troves when it comes to detailed methods descriptions. Thanks to Google you may find the relevant sections online for those that don't have online or library access to these books.

I also found the "Current Topics in...." series useful and the recently started journal "Nature Methods". See for instance Naomi's review on protein crystallization (rocks!).

I have been glad to see several - freely available - online supplemental sections in both Science and Nature magazines, giving detailed experimental background that was inaccessible or hidden in more obscure journals years ago.

Really, the best source of info for your particular protein crystallization case may come out of a discussion with a collegue and be more appropriately discussed at a scientific conference, over a beer (or both).

A month in the lab can save you a quick trip to the library.

Cheers,
Peter

The single most useful practice to reduce effort in protein crystallization

Taking a wrong turn while hiking the great outdoors can be the start of a big adventure, but in the protein crystallization lab it usually means a waste of time, especially when dealing with optimizing protein crystallization hits. The critical point is a typical go/no-go decision where the question you need to answer is this: is it worthwhile to follow up on this particular crystallization experiment? The answer is "yes" if there are no other, potentially better crystallization hits and if you're sure that the objects you're looking at are protein crystals, though not yet useful for X-ray diffraction experiments. The result of a positive decision will be committing to an often tedious protein crystal optimization project, possibly weeks of re-purification and preparing numerous small variations of that protein crystallization setup.

At this point there's one very simple observation that you can make to avoid taking a wrong turn: check the reservoir solution for crystals. Any crystals there resemble those in the drop?
Since many crystallization cocktails are formulated with high salt concentrations, just a little dehydration can cause salt crystals to form.
So, if there are crystals in the reservoir you're very likely dealing with salt crystals in your drop and not the sought-after protein crystals. Don't take this turn and keep on searching for a better protein crystallization hit.

Good reasons to curb your enthusiasm: crystals in the reservoir solution of (A) hanging drop protein crystallization experiments or in (B) sitting drop protein crystallization setups are a harbinger of bad news: salt crystals.

Bon voyage,
Peter

Growing monster-sized protein crystals

Let's say you have plenty of protein sample available and you want to grow really large crystals. How large? No-microscope-necessary-to-see-them large. Here's how to do it:

Small-scale batch technique

Ivan Rayment describes in his paper

Rayment I. 2002. Small-scale batch crystallization of proteins revisited: An underutilized way to grow large protein crystals. Structure 10: 147-151

that large protein crystals can be grown (for ca. 70% of proteins) in this way:

1. Forget vapor diffusion (no sitting drop, no hanging drop)

2. Know your initial crystallization condition (this method is not about screening for a crystallization hit)

3. Learn the slow mixing technique and prepare setups as small-scale batches

4. Identify the seeding sweet-spot (sub-nucleation conditions) 

5. Reproduce by scaling up in larger volume 

At first you'll need to 'translate' a given crystal growth condition to small batch crystallization. This is done by slowly mixing 60-80% of the precipitant with an equal volume of protein sample and placing it into a dimple glass plate with a seal. With a gradual variation of the precipitant concentration you'll determine the conditions that produce a supersaturated solution but no productive nuclei. The key is slow, homogeous mixing that Ivan describes like this: start with 5 ul of protein in the bottom of an Eppendorf vial, then vortex slowly (just 5-10 rotations per second) and add the precipitant solution s-l-o-w-l-y. For example over a period of 2 to 5 seconds. Don't mix by aspiration! The goal is to get into a homogenous mixing regime. The mixture is then transferred to the sealed container.
In the next step you want to identify the solution that is supersaturated but has not produced nuclei - a clear drop that produces crystals when seeded. This can be tested by streak seeding. Practically this is done by touching a crystal with a clean probe and streaking it through a clear drop. The drop that grows crystals after this procedure is suitable for the next step:

Scale up.

At this point you know how to produce a supersaturated solution of your protein and how to induce crystallization. Adjustments of precipitation concentration, pH, salt, temperature etc (sic!) may be necessary, but ideally you would just repeat the experiment with larger volumes, let's say 100 uL. If you can, use a test tube as the vessel to slowly add precipitant to the protein solution. On the vortexer. V-e-r-y   s-l-o-w-l-y.

 

Taken from Ivan Rayment's paper 

All the best,

Peter

4 good reasons to grow protein crystals other than for X-ray diffraction

X-ray crystallography is only one of several disciplines requiring protein crystals as essential materials. Some of us X-ray centric structure biologists may actually be surprised that there's a use for protein crystals for anything else besides X-ray diffraction. Here's my run-down:

1. Neutron diffraction
You're right - this is a lame one, protein structure determination with neutrons instead of X-rays. To diffract a neutron beam you need protein crystals. Massive ones. Think milli meters, not micro meters. But you'll see hydrogen atoms and their 'gymnastics'. Here's Dean Myles arguing for a bright future of neutron diffraction.

2. Solid State Nuclear Magnetic Resonance
Protein crystals sizes under 1 um? No problem. The good news with ssNMR: nanocrystalline material is fine.

Martin RW, Zilm KW.
Preparation of protein nanocrystals and their characterization by solid state NMR.
J Magn Reson. 2003 Nov;165(1):162-74.

3. Protein Purification
Crystallization is a way to purify and enrich your target protein, remember?
Seriously, I predict that we'll be seeing a renaissance of protein crystallization as a purification method for bulk material processing.

Russell A. Judge, Michael R. Johns, Edward T. White
Protein purification by bulk crystallization: The recovery of ovalbumin
Biotechnology and Bioengineering, Volume 48 Issue 4, Pages 316 - 323 (2004)

Why? Columns are expensive, crystallization is cheap.

4. Protein Formulations
Many small molecule drugs are delivered as carefully formulated crystalline powders. The same can be done for protein therapeutics. See:

Alexey L. Margolin et al.
Stabilized protein crystals, formulations comprising them and methods of making them
US Patent 7,351,798 B2

This is a patent claiming dried protein crystals formulation as a means to provide slow release for biomedical delivery applications such as vaccines and modern biologics protein drugs.

And this may be an interesting primer if you're interested in the use of protein crystals for protein formulation design.

Anna J., Merkle, H.P.
Diamonds in the Rough: Protein Crystals from a Formulation Perspective (great title!)
Journal Pharmaceutical Research
Volume 18, Number 11 (2001)

Plenty of job opportunities out there for protein crystallizers!

Peter

Slam dunk protein crystallizations

Many crystallizers love to brag about the 'beast they tamed'. They tell stories describing all the different tools that were used and how in the end a pinch of luck was required to get a recalcitrant protein to crystallize and determine its structure.

How is it possible to judge that one protein is more difficult to deal with than another? Could it be that the standard 'round' approach taken didn't fit the square 'peg'? Could it be that you just followed a pretty but dead-end crystal form while ignoring an early ugly but productive one? Maybe you're not applying the tools properly? Do the fish you're trying to catch swim right through the large-size mesh? You see where I'm going with this: Struggling with a protein crystallization project may have many different reasons, you may be a poor crystallizer or the protein may have a very narrow crystallization slot, that's just very difficult to hit. I guess this could be at the heart of the 'crystallization is more of an art than a science' comments I don't think are useful. 

Some protein crystallization projects are just a 'slam dunk'

Regardless, it's a fact that some (not many) proteins express well, can be purified by standard two step chromatography to >95% purity, concentrate well and crystallize in 5% of all formulations of Wizard I, II, III & IV screens and in Cryo I - and diffract to 2A on a home source without any further cryoprotection. Some scientists call these proteins 'well behaved' (when you're a PhD student and your project involves determining a protein structure, my advise is to get several irons in the fire to identify such 'well behaved' protein). Nevertheless, the more proteins are put into structure determination pipelines and the more X-ray structures are determined, the higher the likelihood that you're confronted with targets that are not 'low hanging fruit' and that require a more sophisticated approach than that of a standard structural genomics pipeline. This is the challenging end of the spectrum and when you're done with limited proteolysis, engineering fragments and surface mutants and get a double mutant protein to form well-diffracting crystals you're welcome to exercise your bragging rights by publishing a paper titled "The taming of...".

This is exactly what Baranova et al. have done in this month's edition of Acta Cryst:

E. V. Baranova, S. Beelen, N. B. Gusev and S. V. Strelkov
The taming of small heat-shock proteins: crystallization of the [alpha]-crystallin domain from human Hsp27
Acta Cryst. (2009). F65, 1277-1281

You guys are awesome!
Peter