Protein Crystallization Hits

Protein Sample Preparation: Shoot for > 20 ODml

I learned this simple trick from Larry Miercke @ UCSF: rather than measuring OD280 and computing mg/ml using an extinction coefficient, go straight with ODs.

Here's how this works:
Measure the OD280 (that's the Optical Density at 280 nm; using a proper reference buffer) of the protein solution at hand. For simplicity reason let's say you've got 50 ml of pooled eluate fractions and you're using a 1 cm path length cuvette. If your OD280 reading is 0.5 you'd multiply 0.5 OD by 50 ml and say that you've got "25 ODml worth of protein".

Note that as a rule of thumb for most proteins:
1. you can assume that an 2 ODml correspond to 1 mg, and an OD of 2 corresponds to ca. 1 mg/ml;
2. concentrate the protein solution to higher than 20 OD to set it up in a protein crystallization trial.

So, when you're done with a protein prep you want to keep concentrating until you've got an OD exceeding 20. Say your solution is now concentrated to an OD of 22 in a single ml - that's 22 ODml. Let's snap freeze 400 ul and use the remaining 600 ul to set up 6 x 96 1 ul+1ul crystallization experiments (Wizard I, II, III and IV and Cryo I and II of course ;) for incubation at RT and at 4C.

What do you do if someone asks for the protein concentration? Using the rule of thumb (OD of 2 corresponds to ca. 1 mg of protein) you can do the quick math: knowing that your purification yielded a total of 22 ODml, divided by 2 equals: 11 mg of protein! And since the concentration procedure resulted in a ml of solution the protein concentration you used to set up the crystallization experiment is 22 ODml / 1 ml - that's 11 mg/ml.

Pretty simple, hm? No need to employ Beer Lambert.

At first sight there's no fundamental difference between accounting in terms of ODml or using properly calculated mg/ml. There are several advantages of the ODml system though:
1. you're dealing with a tangible parameter that's easy to assay anywhere in the purification protocol,
2. It's easy to monitor the slight decrease of the ODml during concentration, if the ODml increase there's a problem,
3. you're not obligated to use a parameter that you know is associated with some margin of error since the calculated extinction coefficient is not perfectly accurate.
I've also seen how extinction coefficients for a particular protein have changed as I progressed with my project. Other than mundane OD280 errors from Raleigh Scatter and buffer or sample contamination there are actual reasons why the extinction coefficient may change during the course of purification: This may be due to the
4. Cys redox state that you're starting to better manage at one point or
5. ligands that dissociate during purification.
Going back in your notebook you may be wondering, 'which extinction coefficient did I use to compute this protein concentration'? You're not bothered by such questions when using OD and ODml.
BTW, a similar argument may be said about the molecular weight (MW) of the target: you *think* you're working with a 46 kDa protein as calculated from sequence until you've seen the Mass Spec analysis. Since the MW is part of your Beer Lambert conversion you can avoid this systematic error.

In short: ODml are for real and are simple to use.

Nevertheless, for all of you wishing to get an estimate of the protein concentration in mg/ml using amino-acid based estimated extinction coefficients, here's how to use the ProtParam tool for this purpose:

1. Paste in your amino acid sequence into the sequence box (I'm using the sequence of chicken lysozyme)

 

ProtParam accepts amino acid sequences, counts Tyr, Trp, Cys, adds up the corresponding extinction coefficients and outputs this number as an overall Extinction coefficient  (and other data) to the user.

2. Click the button 'Computer parameters' and a data rich report is generated, listing

• Number of amino acids
• Molecular weight
• Theoretical pI
• Amino acid composition
• Total number of negatively charged residues
• Total number of positively charged residues
• Atomic composition
• Formula
• Total number of atoms
• Extinction coefficients
• Estimated half life
• Instability index
• Aliphatic index
• Grand average of hydrophathicity (GRAVY)

That's a lot of data to digest. Let's look closely at the section 'Extinction Coefficient':

The extinction coefficients are served ready to get inserted into Beer Lambert's Law, where the protein concentration is calculated from:

         OD280 x MW (in g/mol)
-----------------------------------------------
Ext. coefficient (1/Mcm) x path length (cm)

3. Using the example above (OD of 22), the concentration becomes:

   22 x 16,238.6 g/mol
----------------------------- = 9.4 mg/ml
   37,970 /Mcm x 1 cm

That's a suitable concentration to go into protein crystallization trials with and is pretty close to the 11 mg/ml estimate using the OD rule of thumb.

Peter

Waiting the protein to concentrate: setting up your crystallization experiments for success

Protein samples need to be concentrated prior to setting them up in a protein crystallization experiment, ideally to above 10 mg/ml. Since this concentration business takes a lot of patience, it's a good idea to plan ahead for a successful protein crystallization trial:

• Know what your target concentration is in OD units rather than in mg/ml - no need to waste any time with Beer-Lambert.
• Workplace: bring out all your tools: crystallization plates, sample container, pipettors, tips, pens, stands, notebook, box for used tips, Wizard screening kits ;). Arrange everything on your bench so you're ready 'to do the robot'. Label your crystallization trays (not the lid).
• Prepare for a steady hand and some quiet time. I used to enjoy setting up crystallization trays all by myself, coffee abstinent for 4h (otherwise I'd get the jitters...). Less traffic in the lab means fewer distractions, helping to avoid pipetting errors.
• Using Crystallization Robots? It almost goes without saying that at this point you want to have had a training session in proper use of the machine. Do clean any liquid carrying pathways, check the waste containers and all the lines. Get all the accessories in place, you don't want to run out of those special pipetting tips or clog a needle. Running a single trial dispense doesn't hurt either.
• Prepare to shock-freeze in liquid nitrogen a sample of the concentrated protein, round up the dewar with liquid N2, safety goggles, vials and tongs. This is for the positive control and follow-up optimization trials. I used to work in a lab where getting liq. N2 was a pain, involving taking the elevator into the basement and then operating a scary piece of equipment making loud unpredictable noise. I learned the hard way that N2 levels were usually low towards the end of the week.

Now you're getting close to combine your perfectly formulated protein solution with sets of crystallization reagents.

Here's the pre-crystallization setup countdown:
t= -15 min: Centrifuge stops, sample out
t= -10 min: OD280 is within your target range.
t= -5 min: Shock freeze the portion of protein that you're not planning to set up. Freeze the solution you're not using with this crystallization trial later
t= 0 min: pipette!

 

 

While it's not rocket science, proper preparation for the crystallization experiment minimizes errors and sets you up for a successful protein crystallization trial.

No more "Huston, we have a problem".

Peter

P.S. Somewhat related to this post: Michael Sawaya has written up a nice intro on "What every crystallographer should know about a protein before beginning crystallization trials"

5 Rules of Crystal Cryopreservation for X-ray Diffraction

Your crystals are filled with water inside and they are wet on the outside. How do you cool a single crystal down to liquid nitrogen temperature without turning that water into hexagonal ice crystals? That's the prerequisite for macromolecular cryocrystallography: obtaining low-mosaicity X-ray diffraction data while avoiding the nasty 'ice rings' at 3.897, 3.669, 3.441, 2.671 and 2.249 Angstroms.
Here are my 5 Rules of Crystal Cryopreservation:

1. Small is beautiful. Larger crystals have lots of mass that takes time to cool. Smaller crystals cool faster. If you have enough X-ray flux, better go with the smaller crystals.

2. Soaking wet is bad. Try to wick away as much water from the crystal as possible. Avoid the big blob of water with crystal swimming around. Wicking away excess water by tapping on a dry surface has worked very well for me in several cases (e.g .tap on dry spot on glass cover slide next to the drop you're fishing the crystal out of). Dragging crystals through oil helped me a lot in a particularly stubborn case.

3. Cool fast. Minimize the time to go from drop to liquid nitrogen. It turns out that the last milliseconds before the crystal feels the liquid nitrogen are crucial. That's why Robert Thorne recommends puffing away that thick layer of insulating room temperature nitrogen gas and plunge the crystal quickly into the liquid nitrogen. He calls this hyperquenching.

4. Test, test, test. Test different cryo-reagents and procedures. In most cases you'll dip the crystal into a cryo-solution before cooling the crystal in liquid nitrogen. Testing different cryosolutions and methods will likely result in an optimal procedure for crypreservation. For inspiration check out Artem Evdokimov's nice simple and thorough recipe for cryoprotection of delicate crystals.

5. Laissez-faire. If you only want to check if the crystal you've got is a protein crystal (i.e. has many spots in patterns) and not something else: just go for it! Dip the crystal into the liquid nitrogen without any further ado and don't worry about the ice rings, just get the crystal into the beam and optimize cryo-conditions later.

Example for a hexagonal ice crystal. Not what you want to see when cryo-cooling protein crystals.

And here's my shameless plug: the smart way to pre-empt any of the above is to include the Emerald BioSystems' Cryo-screens into your primary crystallization screen repertoire. I've heard many crystallizers praise these screens. Any protein crystallization hit in Cryo I or Cryo II will cool in liq. nitrogen without creating any of the dreaded hexagonal ice diffraction patterns.

As always: wear your gloves and safety goggles when handling liquid nitrogen,
Peter

Protein Crystallization by dehydration

I deeply appreciate admitting mistakes. This is how we learn. Try, fail, try again with new spin and succeed. More power to crystallizers like Miriam L. Sharpe, who manages to get a paper out of a perceived failure. She didn't get any crystals in her initial "....attempts at crystallizing the protein, including screening 681 different conditions, were unsuccessful. Initial screens included Crystal Screens I and II (Hampton Research), a systematic PEG-pH screen (Kingston et al., 1994 ), a PEG/Ion screen (Hampton Research), Footprint Screen No. 1 and the PEG Footprint Screen...". Check out her complete story here.

So much for that. But guess what - Miriam sets up an entire tray with 100 nl protein drops each, and admittedly by mistake forgets to add any precipitant solution (neither in the well, nor in the drop). And what happens? Get this: seven months later she finds crystals in the dehydrated drops. She then goes on, dilutes the viscous matrix around the crystals, fishes them out before they dissolve, flash freezes, collects X-ray diffraction datasets and determines the 2.1A structure.

Crystals grown by incubation under 'carefully adjusted' dehydrating conditions, or: forgetting to add precipitation reagent and letting the crystallization tray sit on the shelf for 7 months.

Congratulations to the structure of Hupoxic response protein I, Miriam Sharpe! And thank you so much for this great crystallization story!

Peter

More plugs: simpler membrane protein crystallization with PLI

Good news for membrane protein crystallizers: check out this paper by Liang Li et al (full disclosure: I'm part of the et al.). Bringing down the volume for a single crystallization experiment to below 1 nano liter is a technical feat of course, and is an important step towards easening the sample volume requirements and ultimately the cost for new membrane protein X-ray structures.
But what I'm most excited about is PLI: post LCP-formation incorporation. What it means is that for LCP-based membrane protein crystallization setups the membrane protein does not need to be reconstituted into the LCP in a first (and somewhat tedious) step. In fact, it's fine to just add the detergent-solubilized membrane protein sample to a portion of monoolein-based Lipidic Cubic Phase (LCP; FAQ on LCP-based xtallization here) and add a precipitation reagent as a third step. We're not quite sure about what exactly is going on on a microscopic scale, but it's another methodological step in the right direction.

Get it? You can crystallize membrane proteins by just having some LCP floating around in the crystallization drop.

Crystallization plug prepared within a microfluidic device holding a crystal of Photosynthetic Reaction Center surrounded by an aqueous and an oily phase. 

You'll still need to prepare some LCP - but that's a piece of cake with the Cubic LCP kit.

Great work, Liang!

And thanks for such a stimulating collaboration with the Ismagilov lab!

Complete reference:

Li, L., Fu, Ql, Kors, C.A., Stewart, L., Nollert, P., Laible, P.D., Ismagilov, R.F.
A plug-based microfluidic system for dispensing lipidic cubic phase (LCP) material validated by crystallizing membrane proteins in lipidic mesophases
Microfluidics and Nanofluidics, DOI 10.1007/s10404-009-0512-8

Will my protein crystallize?

Let's assume for a minute you're dealing with a protein of a certain sequence and you're wondering if the protein is crystallizable. The tool to assess crystallizability is: XtalPred. This is an online server that accepts amino acid sequences and provides you with:
• a comparison of protein features with corresponding distribution from the TargetDB 
• summary features with crystallization problem areas  
• prediction of ligands and
• lists close homologs that are more likely to crystallize.
The algorithms used are described in this paper. There's also a related blog post about the biggest Secret in crystallography here.

How about taking XtalPred for a test drive? Let's make it simple and use Lysozyme as an example. There are many conditions under which Hen Egg white Lysozyme crystallizes - it's a great workhorse for test-crystallizations. The PDB gives you 1151 'Structure Hits' when searching for Lysozyme.

OK, let's take the amino acid sequence of Hen Egg white Lysozyme (P00698), remove the leader sequence and submit it to XtalPred.

XtalPred input page with Lysozyme sequence inserted.

Here's what you get:

XtalPred output page with results displayed in tabular form

OK - Xtal pred finds the target ID and lots of other info - even that there are 516 homologs available in the PDB. But it classifies Lysozyme into the red crystallization class: "Difficult"??? A poor grade of five on a scale of 1 (optimal) to 6 (very difficult).
Come on guys - Lysozyme, difficult?

Lysozyme cystallizes easily! Check out this paper from 1946, reporting on Lysozyme crystallization from egg white (check out the nice historical photographs of Lysozyme crystals).
Lysozyme crystals even grow when triggered with a flash of light. Lysozyme crystallization is indeed fool proof - I've successfully taught a physicist how to crystallize Lysozyme (a theoretical physisists who'd never touched a pipettor).

I guess that if you get an XtalPred result in the green = 'optimal' zone you've got crystals in your vial once you go back to the bench.

Peter

Protein Crystallizers' Hall of Fame

This week's announcement that Venki Ramakrishnan, Tom Steitz and Ada Yonath will be awarded the Noble prize for Chemistry made me think: how many other Nobels are based on protein crystallization and protein crystals? 

I've compiled a list with the pioneers, our heroes and role models - but keep in mind that these awardees stand on the shoulders of giants: often it is the post docs and PhD students that prepared proteins and grew the crystals to enable X-ray crystallographic structure determination.

	Enzymes can be crystallized  1946 Chemistry James Sumner (with John Northrop and Wendell Stanley) James Sumner is the father of protein crystallization. Check out James' acceptance speech where he actually goes through the Urease crystallization protocol. Phantastic story!
	Structures of globular proteins 1962 Chemistry John Kendrew & Max Perutz This one was for the crystal structure determination of sperm whale myoglobin.
	Determinations by X-ray techniques of the structures of important biochemical substances 1964 Chemistry Dorothy Hodgkin Dorothy determined X-ray crystallographic structures of cholesterol, vitamin B12 and penicillin - that's what this prize was for. Really, she should have received a second Nobel prize for determining the structure of insulin, a major breakthrough in 1969.
	Development of crystallographic electron microscopy and discovery of the structure of biologically important nucleic acid-protein complexes Aaron Klug 1982 Chemistry
	Determination of the three-dimensional structure of a photosynthetic reaction centre 1988 Chemistry Hartmut Michel (with Johan Deisenhofer and Robert Huber) Check out the crystallization story. First membrane protein structure ever - and what a beast it was! Although Hartmut Michel had grown crystals of bacteriorhodopsin earlier, the crystals of the photosynthetic reaction center diffracted much better (I've got a suspicion, why that is) and allowed useful X-ray diffraction experiments to be carried out.
	Elucidation of the enzymatic mechanism underlying the synthesis of adenosine triphosphate (ATP) and discovery of an ion-transporting enzyme 1997 Chemistry John Walker (with Paul Boyer and Jens Skou) Not only a membrane protein complex but a mechanical molecular machine. Very neat.
	Membrane channels: Potassium channels and water channels 2003 Chemistry Rod MacKinnon (with Peter Agre) Finally seeing the ions in the channel was s p e c t a c u l a r. Very impressive DIY approach by MacKinnon, Electrophysiologist turned protein biochemist and X-ray crystallographer. Very inspiring! I was thrilled that in mentioning of structural aspects of the aquaglyceroporins referencing Peter Agre refered to one of my own papers in his Nobel lecture. Membrane proteins
	Molecular basis of eukaryotic transcription 2006 Chemistry Roger Kornberg He developed of 2D crystallization on a lipid surface and authored the two 2001 Science papers on the yeast RNA polymerase complex (a molecular factory with DNA and RNA!) at 2.8A resolution.

 

And this year:  2009 Chemistry

Structure and function of the ribosome

Venki Ramakrishnan, Tom Steitz and Ada Yonath

How many atoms? Mindboggling. I remember that most protein structure visualization programs at the time could not even display that many atoms.
I believe Ada Yonath is to be honored to have worked out a lot on the crystallization aspects. But where is Harry Noller in this picture?

Newsflash: Nobel prize for the structure of the ribosome

C O N G R A T U L A T I O N  S   !  !  !  !  !

to: Venkatraman Ramakrishnan, Thomas A. Steitz, Ada E. Yonath

I'm glad to see years of work on ribosome purification, sample preparation, crystallization and endless optimization of both protein crystallization conditions and crystal handling (including crystal soaking and co-crystallization) being rewarded with the 2009 Nobel prize for Chemistry.

Congratulations,

Peter

How-to: Protein Crystallization in an ideal world

In an ideal world, this would be a fairly simple way to grow protein crystals for X-ray structure determination purposes:

Start out with ca. 400 ul of filtered target protein solution, in 20 mM Hepes buffer neutral pH and maybe 50 mM NaCl , freshly purified to ca. 95% purity and concentrated to 10 mg/ml.

1. Snap freeze four x 25 ul aliquots of protein samples in thin-walled PCR tubes and store at -80C. These sample will be used in a week for protein crystallization optimization.

2. Document the quality of the protein sample via SDS-PAGE and determine its concentration (ca. 5 ul). Measuring UV absorption (OD280) and calculating the protein concentration is good enough.  You'll need ca. 5 ul for such an OD measurement (depending on cuvette path length and dilution). Check this online tool to get an estimate for the extinction coefficient based on sequence.

3. Now let's get out the multichannel pipettor and set up all of the remaining 200 ul of sample solution as 1ul + 1 ul sitting drop vapor diffusion crystallization experiments. A great first pass would look like this:
Plate 1: Wizard 1 & 2 (96 drops)
Plate 2: Wizard 3 & 4 (96 drops)
By the way, Emerald offers these reagents as the Wizard Suite. This is a 192 point, non-overlapping, sparse matrix that has a proven track record to yield crystals in first-pass protein crystallization exploration trials.

And while we're at it: go with the Compact Jr. plates. These plates can be sealed with clear tape and drops form nicely on the hyrdophobic polypropylene surfaces.

4. Store the crystallization trial at room temperature, minimizing temperature fluctuations and vibration. Observe right away, after 1 day, 3 days and 1 week.

5. The initially clear drops with now contain precipitate, some clear drops and a few with clustered microcrystals or needles in them. If the corresponding well-solution does not contain similar crystals, chances are that you've grown protein crystals! Since these crystals are likely to be too small to diffract them using your home-source X-ray generator and detector system, you'll need to optimize crystallization conditions and grow larger crystals.

6. There are many different ways to optimize crystallization conditons, and depending on prior knowledge you may want to carry out seeding experiments, include additives or change the treatment of the protein sample (i.e. filtration). Here's a simple optimization schema: create a grid screen around all conditions that gave you crystals. This rational crystallization optimization schema works great since it separates the effect of pH, salt, precipitant and protein concentration). You should use the online ScreenBuilder design tool to create an optimization screen. Your fellow researchers at Emerald BioSystems  are happy to prepare and send such a customized optimization screen to you.

7. Set up 96 - follow-up optimization 1 ul + 1 ul crystallization experiments with the saved protein material that you thaw in your fingers. If everything goes according to plan, crystals of different sizes will grow.

8. Harvest a crystal, cryoprotect, diffract and determine its structure ;)

Enjoy,

Peter

Co-crystallization and crystal soaking for in-house SAD phasing: no magic necessary

I just saw this new paper published in Acta Cryst F  on the topic of heavy atom derivatization:"How to get the magic triangle and the MAD triangle into your protein Crystal" Acta Cryst. (2009). F65, 1068-1070 by T. Beck, C. E. da Cunha and G. M. Sheldrick.
To tell you the truth, had I reviewed the paper, the word 'magic' would not have made it into the paper (see may rant about a similar, delusional topic here). Otherwise, this is a really nice methodological paper, describing a new class of heavy atom compounds and optimized ways to incorporate these into protein crystals by either cocrystallization or soaking. The resulting derivatized protein crystals can be used to "...provide easy access to experimental phasing" via MAD or SAD techniques.

I'm particularly glad that the experimental procedures for co-crystallization and soaking into pre-grown protein crystals are described in such great detail. This facilitates protein crystallization attempts with your own target and the new compounds. The optimized protocols for cocrystallization or soaking using the compounds I3C or B3C look straight forward: add B3C from a stock solution at a concentration 10X that of the protein and set up crystallizations. I was very pleased to see that they employed the Emerald BioSystems Wizard Screen I.

But I cringed when I read that "Limited amounts of B3C may be obtained by contacting TB". Good luck with this in 30 years time. But the other compound, I3C is advertised as "commercially available as a powder from different chemical suppliers". Since I could not find the supplier info in the paper itself, I went back to my preferred chemistry search engine ChemSpider and was quite disappointed that the compound name spelled out (and used in my a copy-and-paste search) in the introduction of the paper (5-amino-2,4,6-triidoisophthalic acid) did not produce a hit (in retrospect I should have seen that there was a typo, it is 'triiodo').

What now? Pull out the ChemSpider drawing tool to create a 2D structural formula et voila: got the compound listed in its full glory as: 5-amino-2,4,6-triiodo-benzene-1,3-dicarboxylic acid. Curiously, the list of synonyms did contain my first search term - without the spelling error - amongst many others. Thankfully, the list of suppliers was also given and amongst them was
Sigma Aldrich - with a direct link to the proper product page in Sigma Aldrich website.

ChemSpider 'in draw a molecule'-mode and the 5-amino-2,4,6-triiodoisophthalic acid molecule with experimental 4sig anomalous electron density (in a Lysozyme molecule).

But I digress.

The breakthrough discovery in utilizing this new class of heavy atom containing compounds is that our toolbox for experimental phasing has been enriched in a substantial way. Not only do these compounds have the capacity for specific binding, they are inexpensive, have comparatively low toxicity and provide a strong anomalous signal enabling SAD phasing with data collected in-house. What else can you ask for?

Now that's what I call a breakthrough. Nothing magic about it.

Herzlichen Dank, Tobias!