receptor-ligand binding experiments?

[chef]

let's say I want to find out where ligand X binds to receptor Y. the sequences of both are known, and recombinant proteins are amply available.

what is the most reasonable approach to get the answer? site-directed mutagenesis analysis?

also, is it more useful to find out the binding domain on the receptor, or the ligand, if the goal is to eventually develop new drugs that mimic the endogenous ligand??

---

Members don't see this ad.

 

[Ursus Martimus]

Assuming your protein is not completely unrealated to any known protein/proteins out there, I would think the most reasonable approach from a structural approach. There are data bases of recepter/ligand co-cystallization studies available which can be overlaid with the predicted, which is the operative word, 3d confirmational of your receptor/ligand pair. Further defining the interfaces could involve mutagenesis studies followed by binding studies and/or cocrystallization of your mutants with the ligand/ligand mutants to assess your predictions. At any rate to be so precise is also so tedious - but it is in part the story behind rational drug design as such interfaces are thoroughly examined, peptidimimetics tested and finally and hopefully a small molecule with oral bioavailability constructed.

 

[leverp2000]

The other option for narrowing down the region to which the ligand binds is trying to attach a crosslinker to the ligand and try to map the region where this crosslinker is found by digestion of the protein. This will give you a general range. Cocrystallization or NMR studies are definitely the best way to go, but these will be very tedious and time consuming.

 

[AndyMD]

I'm not in this field, and I know precious little about the conventional methods applied in these situations, but I'm actually going to say mutagenesis analyses seems like the best route. Though a structural approach (eg crystallization) would give a definitive answer, I think it's probably prohibitively time-expensive and of course requires an x-ray source on the chance you're actually able to crystallize.

Mutageneses, on the other hand, are fairly straight-forward though also time-consuming. I can imagine that you could use bioinformatics tools to distinguish different domains of the receptor protein that would be likely to form a binding pocket or something for the ligand, and then you could begin to map the region of interaction by performing a targeted mutagenesis study involving just those likely domains in the receptor ....

You'll need some assay to provide the read-out for ligand-binding to your expressed mutant receptor proteins though -- I've heard of ppl being able to use radio-labelled ligand in competition with unlabelled ligand to determine binding affinities, which in your case would decrease when a mutation disrupts the region of ligand-receptor interaction. I think an analysis of the kinetics of this competition could confirm that the radio-labelled competitor (if it is structurally different than the wild-type ligand) was binding competitively, and therefore at the site of wild-type ligand interaction, rather than at an allosteric site, as a control for the validity of the results.

IF all this worked, then you could begin to figure out which part of the ligand was binding to the receptor by using peptide fragments of your ligand, to determine which domains are capable of binding in isolation.

As I said before though, I really have no idea what's normally done. Haha actually, your best bet would be just to search the literature for current methods

Anyways, good luck!

 

[spinman]

That depends:

1) How large is the receptor - is it a GPCR?
2) How well do transient trasnfections of this receptor express in tissue culture?
3) Is there a clearly defined assay or biomarker to detect the activity of this D/R interaction?

There are a few approaches that largely depend on the answers to these questions:

1) If you know what the recpetor is, i.e. you have a good idea of the structure due to homology or if the receptor itself has a known complete or partial structure - SDM (Site Directed Mutagenesis) can be a great start. Generate a library of scanning mutants with a rational approach where mutations are inserted only in regtions where a drug could possibly bind. Express these in cells, see if a particular mutant has a greatly altered dose response.

2) Depending upon the receptor, sometimes fragments of the receptor can be expressed in baculovirus, RRL (retic lysate), or even bacteria if post translational modifications are not of concern. In this case, you can run a gel mobility shift assay whereby the recpetor fragment can be radiolabeled. These gels can then be obsevred using storage phosphor screens. This method will only work if either the whole recpetor can be used (not always feasible if it is membrane bound) or if fragments of the recptor fold properly

3) structure - ditto what others have suggested. XRay has the potential to be the fastest, but homology modeling has some real potential if the structure or a closely related homolog/ortholog is known. If the structure is known, computer modeling can predict where the drug might be binding. if so, simple peptides or peptidomimetics can be generated to act as an antgonist.

There are other more complicated methods that will depend upon lab resources and expertise of the lab or collaborators. If the pharmacophore of the drug is known, it might be possible to generate a drug analog that could be used in tagging or tethering the drug to a support surface or for cross-linking to the peptide. In so doing, MS or other simple assays might work. This is a bit more of a challenge unless you have access to synthetic chemists who have the time to commit to this.

 

[chef]

wow, thanks for the insightful and helpful comments!! anybody know how difficult/long photoaffinity labeling experiments take??

 

[chef]

1) If you know what the recpetor is, i.e. you have a good idea of the structure due to homology or if the receptor itself has a known complete or partial structure - SDM (Site Directed Mutagenesis) can be a great start. Generate a library of scanning mutants with a rational approach where mutations are inserted only in regtions where a drug could possibly bind. Express these in cells, see if a particular mutant has a greatly altered dose response.

SPINMAN,

if the structure & sequence of the receptor is fully known, and if it's expected that the ligand binds to the extracellular domain of the receptor, how do you actually generate a library of mutants? are the mutations random? if you can explain this to me or show me the links to such protocols & explanations, that'd be SUPER. THANKS!!