It's hard to think of a more important class of drug targets than the G-protein coupled receptors (GPCRS). And back about fifteen years ago, I thought I had a reasonable understanding of how they worked. I was quite wrong, even given the standards of knowledge at the time, but since then the GPCR world has become gradually crazier and crazier.
The classic way of thinking about these receptors is that they live up on the cell surface, with part of the protein on the outside and part on the inside. The inside face is associated with various G-proteins, and the outside face has a binding site for some sort of signaling molecule. If the right molecule shows up and slots in the correct way into this binding cavity, the transmembrane helices of the protein rearrange, sliding around to change the shape and binding properties down there at the G-protein interface. This sets off some intracellular messaging - often by affecting levels of the messenger molecule cyclic-AMP. Thus is a signal from outside the cell relayed through the membrane to the inside.
Pretty nearly makes sense, doesn't it? Well, take a look at this new report from PLoS Biology. The authors rigged up living cells with a built-in fluorescent sensor system to monitor cAMP, and then studied the behavior of the thyroid-stimulating-hormone (TSH) receptor. That's a perfectly reasonable protein-ligand GPCR, but it turns out that it does things that are not (to us) perfectly reasonable.
This paper shows that when a TSH molecule binds, that the receptor gets taken back down through the membrane into the cell. That's certainly a known process (internalization), and was thought to be a regulatory process, a standard method for taking a specific GPCR out of the signaling business. Some receptors seem to do this right after they're used, and of those, some of them later resurface and some are broken up. (Other types hang around for many cycles until they're somehow worn out). But the ones that internalize quickly still set off their intracellular message before they get pulled back down. That's their purpose in life.
TSH does that. But the weird part is that the authors saw the receptor internalize along with its G-protein partners, and then continue signaling from inside the cell. Not only that, this extra signaling behavior set off somewhat different responses as compared to the first "normal" burst, and seems to be a necessary part of the usual TSH signaling pathway. It's a very odd thought, if you're used to thinking about GPCRs - it's like finding out that your cell phone works when it's turned off.
Now this sort of behavior has been demonstrated for a different class of signaling proteins (the tyrosine kinase receptors). And even GPCRs have been found, over the last few years, to be capable of setting off a different signaling regime (the MAP kinase pathway) after they've been internalized. (That's one of the weird findings of recent years that I mentioned in the introductory paragraph, and we still don't know what to do with that one as far as drug discovery goes). But everyone agreed that at least the good ol' cyclic AMP pathway worked the way we thought it did, through signaling at the cell surface, and thank goodness there was something you could still count on in this world.
Hah. Now we're going to have to see how many other GPCRs show this kind of behavior, and under what circumstances, and why. It may well turn out to be different for different cells or for different signaling ligands, or only occur under certain conditions. And we'll have to see how this relates to the other strange things that are being unraveled about GPCR behavior - they way that they can dimerize, with themselves or even other receptors, out on the cell surface, and the way that some of them seem to work in an opposite-sign signaling regime (always on, until something turns them off). Do these things still signal from beneath the waves, too?
Oh, this will keep the receptor folks busy, as if they weren't already. And, as usual when something like this shows up, it should serve as a reminder to anyone who thinks that we understand even the well-worked-out parts of cell biology. Hah!
1. Old signaling guy on August 20, 2009 10:14 AM writes...
The surprising thing is that people are continually surprised by the complexity of GPCR signaling.
G protein switching, natural inverse agonists, now this. To me, given the limited number of signals available, the spatial and temporal aspects of signaling that this study demonstrates implies, as you suggest, more work to be done.
The first part of the story was written by adrenergic and opioid receptor people and that became dogma. I gave up the dogma and became agnostic when agouti and AGRP were discovered.
Permalink to Comment2. Darjeeling on August 20, 2009 12:13 PM writes...
Sounds like commenter Mc Feader needs this kind of therapy: http://news.bbc.co.uk/2/hi/science/nature/8206280.stm
--Darjeeling?
Permalink to Comment3. Hap on August 20, 2009 12:13 PM writes...
Oh, look, our friend the drug spammer (soon-to-be-ex-post 2) is back. Whee!
Permalink to Comment4. NJBiologist on August 20, 2009 12:20 PM writes...
"It's a very odd thought, if you're used to thinking about GPCRs - it's like finding out that your cell phone works when it's turned off."
OK, I'm an in vivo guy, not a receptor person, so take this with a grain of salt. How about making the analogy be using your phone when the other person has put their handset down on their desk? The volume changes, and some frequencies get through better than others--but shouts will still get through.
All in all, though, this seems to back up some of the comments from the schematic notation post (#5, #12... and it looks like BamH1d has caught on to this before me).
Permalink to Comment5. barry on August 20, 2009 12:50 PM writes...
Since about half of GPCrs are olfactory, I'm tempted to try to understand them in this context. What good is a receptor that doesn't report that the stimulus has gone away? This report of persistent signalling doesn't comport with our experience of our own GPCrs in vivo.
Permalink to Comment6. Morten G on August 20, 2009 2:33 PM writes...
Try setting the alarm on your phone and then turning it off. Your phone is still on, ta-daa! They can also be monitored remotely while off, or at least they could in the old days (like 5 years ago). There was a warning from SÄPO about industrial espionage back then.
So, obviously, nothing about surprises me about GPCRs.
Permalink to Comment7. partial agonist on August 21, 2009 9:04 PM writes...
Ah,
In the good old days we knew about two flavors: agonists and antagonists.
Now we know about inverse agonists, allosteric antagonists, allosteric agonists, agonist potentiators, and partial agonists (my favorite, hence the name) of many different varieties that can each promote transduction of different parts of the signal and/or receptor internalization.
It used to be you wanted chocolate or vanilla. Now it's 31 flavors and it's hard to tell just what you want
Permalink to Comment8. partial agonist on August 21, 2009 9:05 PM writes...
Ah,
In the good old days we knew about two flavors: agonists and antagonists.
Now we know about inverse agonists, allosteric antagonists, allosteric agonists, agonist potentiators, and partial agonists (my favorite, hence the name) of many different varieties that can each promote transduction of different parts of the signal and/or receptor internalization.
Then they aggregate!
It used to be you wanted chocolate or vanilla. Now it's 31 flavors and it's hard to tell just what you want
Permalink to Comment9. hope springs eternal on August 22, 2009 7:58 AM writes...
More reasons to move back to a whole animal phenotypic screening approach to discovering new drugs: We'll never understand the biology as much as we think we do, and we miss opportunities to let the rat proteome give us unexpected efficacy with our current reductionist approach. Carlsson Research (NeuroSearch) currently uses in vivo SAR, I believe (discovered two Phase 3s with it?)
Does anyone know how the whole industry managed to do it --successfully -- decades ago, if it's so darned expensive to contemplate doing now, as Derek posted recently?
Permalink to Comment10. Pathi on August 23, 2009 8:53 AM writes...
"Neutral Antagonism" is yet another class of receptor modulators and very little research is being done to understand this new concept in pharmacology;Inverse agonism or neutral antagonism at G-protein coupled receptors: a medicinal chemistry challenge worth pursuing? Greasley PJ, Clapham JC. Eur J Pharmacol. 2006 Dec 28;553(1-3):1-9.
Permalink to Comment