Thank you Abdullah! Have you done crystal soaking before?

Hi Lyra,

As I know for fragment screening you had to have many crystals.

And you had to soak them in different fragment combinations.

Did those crystal grow from Bis-Tris buffer conditions?

If "yes" that means your compound had higher affinity than Bis-tris, and you should be able to get complex of your compound if you repeat crystallization experiment.

If you ran your fragment screening with crystals from different crystallization conditions, I would recommend to repeat those crystallization conditions for soaks.

When you soak your compound you have to take in account several things.

1. Is your binding pocket flexible?

If "Yes" than the best would be to soak not "native" (apo-) crystals, but crystals from co-crystallization experiments. Why that? Even if compound did not bind in co-crystallization experiment, it effects on binding pocket and forces it to go in "open" conformation. That will make easy to bind compound to the active site during soaking experiments.

2. If binding pocket is not flexible. I would recommend to soak your crystal with high concentration of your compound. If it binds to the pocket, it should replace bis-tris.

About Cryo solutions. There are several cryo compounds. If your crystallization conditions include PEG, then increasing to about 20-25% of the same kind of PEG, and that will work.

If your conditions do not have PEG. You can use PEG 400, MPD, Hexandiol. Finally, we often use 25% sucrose. We have 50% of sucrose stock solution and prepare another solution which will have twice the concentration of all components of the crystal screen conditions. Mix it 1:1 to 50% sucrose and get the cryo. Another way would be to calculate how much of sucrose powder you need to add to your screen solution to make 25% sucrose solution in your screen.

Usually, when you use cryo soak after your compound soaks, you will wash out some of your compound from crystals. To avoid that, you can add your compound to cryo solution. But first make sure it does not precipitate in your cryo solution.

If your compound is similar to substrate and protein turns it into product, the best would be to do several soaks in raw. Prepare several drops with your compound, and place your crystal in the first drop, than after few minutes move to fresh drop, and so on. This way you can get substrate in most part of your crystal. Otherwise you will have a mixture of the substrate and the product.

There are a lot of tricks and they are very specific to each case.

If you have question, let me know.

Good luck!

George.

Hi George,

Thank you so much for your suggestions! Very detailed and very helpful!

My fragments were identified through other screens (Alpha Screen & ThermoFluor). I am trying to get the co-crystals of the fragments with my protein by soaking my Apo crystal that is grown in BisTris & 30-35% PEG3350 by the fragment.

The problem is, the Bis-Tris is important to stabilize the protein structure for me to get good rectangular shape crystal that diffract at 1.2A, whereas other apo-crystals (with other conditions, no bis-tris) are pine needle shape and very unstable (easily melt by DMSO).

We found that Bis-Tris binds to a pocket that fragments potentially bind. Therefore, we want to soak the fragment into the apo-crystal and at the same time remove bis-tris from the binding pocket.

I don't know the best way to soak out bis-tris without damage the crystal at the same time. If you have any good strategies, please let me know.

I will try the cryoprotection methods you suggested. Yet if my PEG3350 is already at 30-35%, why wouldn't it be enough as cryoprotectant in my case?

Thank you again for helping me! I am very grateful!

Best regards,

Lyra

Hi Lyra,

Now I see that problem is bigger than I thought.

1. You have a very high resolution crystal structure of the complex of your protein with the bis-tris molecule. You can try to fit your fragments in the structure virtually and check if all residues involved in binding of bis-tris are necessary to bind a fragment. In other words, are any not active site residues involved in the binding of a bis-tris molecule? If there are residues not from active site involved in the bis-tris binding you can design a mutant which may have a lower affinity to bis-tris. Or you can screen the mutant and find different crystallization conditions.

I understand that since you have nice crystals it looks easier just to soak compounds and get crystal structures.

2. But you can find a proper fragment without crystal structure. You can run thermal shift assays of your protein with fragments. Then identify the candidates for crystallization. Incubate your fragments with your protein in different than bis-tris buffer and screen your complex in different crystal screens.

3. Since the bis-tris has so high affinity to your protein, you may use bis-tris as a scaffold for your drug design and add some specific groups and moieties to target other residues of the active site (if not all of them are involved in the bis-tris binding).

I am sure there are more strategies to overcome your problem.

George.

Hi George,

Sorry for my late reply!

In my case, I've already know which fragments I want to get co-crystal structures of and I have also purchase some homologs of the original fragment hit (from my thermoshift & alphascreen screen) to identify the best compound for co-crystal.

I like your suggestion of docking fragments on to the same site where bis-tris sits. I will use Autodock first and see if I get some new insights.

What I am going to do for this crystal shipping is soaking the fragment (at 40mM) into crystals in 5mM Bis-Tris buffer and see if my fragment can replace the Bis-Tris in my crystal.

Also, for cryoprotectant, I am going to try Sucrose as the powder can be added directly into the soaking solution. As another way for cryoprotection, I will select smaller crystals this time and use perfluoropolyether oil to remove water. Hopefully, with smaller crystal the oil can protect the crystal better.

We will ship out crystal next week and will let you know how it goes.

Thank you again & wish you have a great weekend!!

Lyra

p.s. A good paper about practical issues in crystallography is attached to share with anyone who is interested.

Hi Lyra,

Thanks a lot for manuscript.

I wish you good luck and collect good data set.

George.

Hi All,

Just want to share another useful info about fragment-protein co-crystal.

http://www.diamond.ac.uk/Beamlines/Mx/Fragment-Screening/Tips-and-Tricks.html

Best,

Lyra

Hi George,

I got the result back from ALS and this time I have much less icing problem, although the crystal mosaicity is a bit high. I obtained several datasets with one that reach high resolution. I will solve the structure this week & hopefully get ligand binding info out of it.

I found that lowering Bis-Tris concentration from 100mM to 5mM while maintaining the pH is still quite bad for the crystal, I can see cracks on the crystal with just 1 hour of soaking. So I didn't sent them.

I have tried thermofluor experiments to see if my compound can bind at the presence of 100mM of BisTris, and it turns out my compound can still significantly increase the Tm (~1.5 degree C). Therefore, it seems that BisTris may be binding to different site as my compound and the reason I don't get co-crystal structure after soaking might due to some operational mistakes I made earlier. I will know soon enough after solving the structures this week.

I meant to use Sucrose as cryoprotectant (at 25%) this time but I wrongly added 25% glucose instead. It seems to still work for me though since the icing problem is reduced.

Is there a reason to use sucrose instead of glucose as cryoprotectant? Does sucrose work better?

Also, my crystal drop tend to have thick membrane on it & make operation extra difficult. Do you know of any method to prevent membranes from forming?

Thanks again for your help! I think things have been looking up now.

Best Regards,

Lyra

P.s. A good drop optimization paper is attached to share with everyone with protein crystal problems.

A quick update for anyone who is interested.

I finally get the co-crystal structure of my compound with the target protein. I go for co-crystalization instead of soaking.

Moreover, I used 25% sucrose as cryoprotectant in another crystal shipping and it works even better. The icing issues are solved.

Thank you all for your help!

Lyra

Hi Chang,

How did you get co-crystal at last? was it just by increasing concentration of fragment without changing concentration of Bis-Tris?

I reduced the Bis-Tris concentration to 50mM and co-crystalize with my compound.

Best,

Lyra