I have been working for several years on a rare inherited disease due to a defective intestinal zinc transporter. In order to better understand the impact of mutations on the protein structure and to potentially improve the molecular diagnosis proposed in my laboratory, I wish I could know more about the 3D structure of the metal transporter.
I have tried many structure prediction softwares, yet I couldn’t even get close to a likely structure, because of a lack of references for membrane transporters in the Protein DataBase. I might be wrong, but I thought that crystallization of my transporter of interest could help me to move forward.
As a matter of fact, I surely have false ideas about crystallography, and I would rather say that I’m almost ignorant about what it involves. I would therefore be grateful if anyone could give me information about field of research. Any advice, articles (even kind of “crystallography for dummies”), links or contacts are welcome.
I was told that crystallography of a protein is very difficult to get, but I’d like to determine to what extent it is feasible.
Thanks for your posts, Sébastien.
The short answer: get ~10 mg of pure protein, make a well-ordered crystal of that protein (at least 20x20x20 nm? ) , collect diffraction data, solve structure, do the write-up.
Each of these steps can be anywhere between piece-of-cake easy and mind-boggling hard.
Getting help from experienced people is probably the best first step - I wouldn't advise starting such an adventure based on reading literature only.
Here is a some randomly picked "interesting links" to get you started
http://isbcgranada.org/
http://px.cryst.bbk.ac.uk/
http://www.jic.ac.uk/staff/david-lawson/xtallog/summary.htm
Recombinant expression and crystallization are especially difficult for membrane proteins. Your best step would be to seek a collaboration with a group that specialises in this work. It sounds like you have something interesting to bring to the collaboration, if you can say how different mutations affect the phenotype.
Yes, you’re all definitively right to say that I shouldn’t try crystallization by myself. I don’t really mean it, as it’s a specialty totally different from all my current and past activities. Yet, I’d like to know more about it before searching potential collaborations. I assume that the quality of the exchange would be better, and I would have a better idea of what I could expect. I figure that it might contribute to improve the diagnostic service which we propose right now, through a better genotype/phenotype correlation. Yet, I’m not sure about it. I don’t even know if it’s worth it, as I don’t have any concrete example.
And thanks, Klaas, for the links.
Don't be afraid either. Assuming you have enough soluble protein (?!), setting up a crystallization screen is not that hard. It's a matter of he right labware, a few gadgets and a keen eye. They will love you if you bring crystals to the meeting!
Sebastien,
We are using fussion to GFP as a tool for screening conditions of expression and purification of MP. See Nature Protocols 3, - 784 - 798 (2008)
GFP-based optimization scheme for the overexpression and purification of eukaryotic membrane proteins in Saccharomyces cerevisiae.
Very shortly,
1. Device a method to purify your protein using different chromatographic techniques for example.
2. Introduce your protein to different crystallization conditions using different screens (JCSG+, Nucleix and so on) available commercially. This means you can set up the sitting drops in Corning plates and observe the behavior of your pure protein in different conditions.
3. Based on those observations, you find out which condition gives you a 'hit' and you optimize that further to obtain well diffracting crystals.
As some of them have suggested, it'd be smarter to collaborate with experts in this field. Obtaining good crystals is very tricky and the whole process takes time. One must check hundreds of different conditions before choosing one condition to further optimize.
If you get crystal structure data of your metal transporter, there is no doubt that it would help clarify many assumptions about the inherited disease. You must definitely go for it.
All the best for your research,
Best Regards,
Sagar Sridhara
Hello Séb.!
For a memb protein it is a trial and error mostly in my opinion. Check papers on pubmed there are plenty.. way too many by Drew or Iwata or Stroud and too many authors.
In general you need mg quantities of the transporter. For membrane proteins the probability is almost 100 times less to get a structure in hight throughput projects. You might need many homologues. You need mg quantities, obviously functional protein (liposomes or other in vitro/purified protein assays), and lots of optimizations (expression yield, stability, SEC profile, detergents, additives, lipids, crystallization methods (vapor diffusion or lipidic cubic phase, etc...), lots of screening of crystal diffraction (too many crystals), lots of data collection... then you move to the step where you might have difficulties solving the phase problem of your diffraction..
This is when you have maybe to repurify the protein with Se-Met labeling, or soak the native crystals with heavy atoms.. Try different anomalous diffraction techniques..
This is necessary to solve the phase since apparently you said no homologue or a good structure are out there.
Crystals generally have weak diffraction and die fast.
If you aren't lucky, this is too much work and might take years. Unique memb protein structures solved so far in bit more than 25 years are like 350 so far.
As mentionned, if it's not not your field and you don't have the appropriate equipment you should try to find collaboraters to run crystallograghy for you.
Anyway, I think you could find 'plateformes' to which you can send your purified protein and they try to crystallize it for you (let me know if you are interested, I think we can do that in my laboratory...).
Concerning your protein, crystallography of membrane proteins is not straitforward (very few solved structures!). However, depending on your biological system, another alternative could be to try crystallography with the ectodomain, you can get many valuable information anyway.
Thanks Toufic. This is not really optimistic, but probably realistic.
Obviously, I can't spend years for crystallization. It's too bad, because I think it would be very useful to rely on the 3D structure to propose a more acute molecular diagnosis. This is true for the rare disease I'm focused on, but it would also be true for many other diseases due to defective ion transporters. I will have a look at the papers you mentioned, though.
Hi Philippe. Thanks for the proposal. It would be very intersting discussing this with you. I bet there exist tricks to bypass the dfficulties related to membrane proteins.
I've seen genescript proposes such a service (http://www.genscript.com/structural_biology_services.html). I may request a quotation, as I'm curious to know what's the turnaround time and how much they charge. Not that my lab can afford it, it would be just to get an idea about the whole process. Actually, I would prefer a collaboration, as I feel it more intersting and it could be a great experience.
Thank you so much Timothy for this very comprehensive answer. It's a very long process indeed. I will check the references you gave and come back to you if I have any questions.
Sébastien
Basically, if you have a few mgs of pure protein, then you are in buisness. Using the screens and robotics, a few weeks and you should either have crystals or not...then the fun starts.
Its not rocket science to solve a structure if you have good diffracting crystals.
- Badges
- Science method