Cellular location, function in physiology, and relevance in disease, are among the many factors that make GPCRs druggable targets.
In Depth: GPCR Analysis
Compounds targeting G-protein-coupled receptors (GPCR) comprise almost 27% of all current FDA-approved drugs (Overington et al. (2006) Nature Reviews Drug Discovery, 5, 993). But what makes these transmembrane receptors so druggable? Why are many of the well-known GPCR targets located in the central nervous system (CNS)?
Drug Discovery & Development magazine conducted a roundtable discussion with industry scientists to answer these questions and more. To review the full discussion, see all related articles.
James Netterwald: GPCRs seem to be particularly popular targets for CNS indications. Is the CNS a primary area of GPCR research, or are there other, competing areas?
Suresh Poda: There are other disease areas of GPCR research because they are expressed across [all of] the tissues in the body. So, while Lundbeck is interested in [the] CNS area, I am sure that other companies are targeting many other disease areas as well.
Charles Lunn: Many classic papers defining receptor pharmacology used the brain for ligand-binding experiments. Modern research labs continue to study receptor pharmacology in brain tissue. There may be a preponderance of work on GPCRs in CNS because of this history, but there are lots of opportunities to pro-gress GPCR therapeutics in many other areas.
Richard Eglen: I would agree with that. There are certainly many programs looking at the role of GPCRs in obesity, the role of GPCRs in pain states, and in controlling [tumorigenesis]. [But] clearly there are a lot of CNS diseases that [involve] GPCRs, ranging from depression to neurodegenerative diseases such as Alzheimer’s disease. So they are looked at in both the peripheral and the central nervous system.
Netterwald: What makes a GPCR a druggable target, in general?
Lunn: Their [very] nature makes GPCRs particularly good targets for therapeutics. The job [of the GPCR] is to sense the environment and to communicate some specific aspect of that extracellular environment to intracellular machinery. The topology of the GPCR means that an effector compound does not have to traverse the cell membrane barrier in order to reach its site of action. When targeting intracellular enzyme targets, significant medicinal chemistry effort is spent trying to get compounds into the cells—a real headache. So eliminating this problem makes the process of designing compounds to progress into clinical trials a bit simpler.
Poda: GPCRs are large proteins with seven-transmembrane, extracellular, and intracellular domains. As Charles mentioned, the most important thing is these are expressed at the cell surface area, where they are involved in a number of key physiological processes controlling cell functions. So by targeting these cell membrane proteins, you can control a number of cellular functions.
Lunn: We have a fair understanding now about the pocket within GPCRs where ligands can bind. These pockets have profound effects on the conformation of a GPCR and, therefore, can have all sorts of intracellular effects that we would like to target [with] our therapeutics. Because ligands vary, there are more differences between the GPCR-binding pockets, more than is observed in the ATP-binding site in kinases. Some kinase inhibitors have a very broad specificity, inhibiting a number of kinases. It complicates drug discovery efforts when a desired inhibitor is not as specific as we seek with GPCR ligands.
Eglen: GPCRs are very open to technologies that one can use to do screening against them. As Charles said, many of them have good knowledge of the kinds of ligands that may be looked at to the extent that many companies now have GPCR-rich compound libraries for screening. But I think the other point is, one, they have a proven success rate, and they are often cited as one of the targets for which high throughput screening has been conspicuously successful in terms of generating novel compounds. The fact that one can get very high-affinity, selective ligands [and] test their druggability. And, even in a receptor class such as the muscarinic receptor class, the Lilly group has recently reported very high selectivity by looking at allosteric regulators. So I think, for all those reasons, they are viewed very highly as important drug targets, as they are very druggable.
Netterwald: So on the flip side, some of you have worked with targets that are not of the GPCR class. How would you compare those targets to the GPCR targets that you’re studying now? How do they compare in terms of, let’s say, ease of druggability? And what are some of the challenges when readying the targets for an assay?
Poda: Let’s take ion channel drug targets. These are more complicated than other drug targets because we don’t have technologies like [we do with] GPCRs. Also, the multiple subunits for ion channels make it more complicated to develop an in vitro assay. Most importantly, the variation between the rat and the human sequence is a big challenge for preclinical studies. If the difference is significant, the compounds developed in preclinical studies may not be efficacious in humans.
Lunn: In terms of screening, biochemical assays for enzyme inhibitors are well developed. For example, kinase assays are really easy. Most of them use peptide substrates. The enzyme is very well behaved. You can purify the enzyme to homogeneity. The readouts are very simple, and you get all kinds of compounds. The problem is that, when you use a peptide as a ligand for a kinase, you are only interrogating a very small part of the active site of that kinase. The native ligands for kinases are complex polypeptides, for example, receptors or other proteins. By interrogating only a peptide-binding site, the investigator could be missing a lot of compounds. But you get so many compounds when you run a screen for a kinase that your therapy area colleagues are happy because they have this long list. [And] you are happy because the assay is easy to run, and you go on to the next campaign …
Eglen: Not all GPCRs are easy to look at. There are some that are very difficult to express. And there are some that are very difficult [for which] to discover a signaling pathway. So it’s not as though they were uniformly straightforward. But I think the chances of success with a GPCR ligand are probably higher than some of the other targets, particularly those involved in looking at protein interactions as well.
Lunn: We have been sort of talking around this subject, but there is the added complexity introduced by GPCR heterodimers. I think it’s an idea just about to burst in the industry. There’s increasing evidence that these heterodimers markedly affect the pharmacology of the ligands associated with it. A good example comes from work from the Portoghese and Whistler labs describing the kappa/delta opioid receptor heterodimers. Here you can find very specific compounds that target the heterodimer itself. And that affords a level of selectivity that most of the opioids just don’t have. So you find compounds [that] have the desired effects on pain regulation, but don’t have the unwanted side effects that you have if you target the individual receptors. The question is how do you identify those heterodimers that are therapeutically relevant? And, I think that is a very difficult and challenging problem that a lot of us are starting to think about.
Poda: Yes, I agree with Charles. Actually, this is more complex than we thought.
Lunn: … There are also examples of biased ligands that will target one GPCR conformation versus another, each of which could have unique therapeutic advantage downstream. There are a number of proteins that associate with these receptors—the GIRKs, the RAMPs, and so on. You would assume that such regulatory protein interactions will change the conformation of the receptor, and again, it could open an interesting new receptor conformation—a new targeting site within the GPCRs. The complexity seems almost overwhelming. The industry has had successes using recombinant receptors in an isolated non-native host cell, and has found compounds. But maybe we should start looking to reevaluate some of these receptors, using an appropriate regulatory factor or heterodimeric partner, if we can fathom what that is. There may be a whole new set of therapeutics out there that we have yet to discover.
|
Eglen: I think that’s exactly right. From the technology provider perspective, we often get many requests for technologies that would enable people to look at homo- and heterodimerization, not only because people are concerned about how they missed compounds in their initial screens, but also if the dimeric combination affects the pharmacology, does it do so on a tissue-specific basis. That is, is the dimer formed, for example, in the spinal cord different from the dimer that would be formed in non-neuronal tissue? That may open up the possibility for tissue-selective targeting of compounds, which in the eyes of many chemists would be a very laudable aim … I think the whole issue of dimerization has been historically quite controversial, in that, is it an artifact of the cell line or the expression systems that were used? But I think the evidence from the consensus is now moving to the fact that this may be a physiologically-relevant occurrence. And, as a result of that, one would need to look at technologies that will enable them to screen compounds against these partners that are forming.
Lunn: … It speaks a little to the profiling that we talked about earlier. If you can define a tissue type important to target for some therapeutic benefit and the receptors that are expressed in that [tissue], you have defined the candidates for a unique heterodimer within that tissue, should it exist. Now, if you could then design a strategy to screen rapidly through the paired possibilities, to see if you could identify an appropriate heterodimer within that tissue, I think you would really have something. But … I don’t think there’s a consensus on how to pursue this idea. I think as we go forward, that could be a very interesting line of research.
Published in Drug Discovery & Development magazine: Vol. 12, No. 2, February, 2009, pp.18-21.
Filed Under: Drug Discovery
