GPCRs as Drug Targets: From Classic Pharmacology to Next-Generation Hit Discovery

G protein-coupled receptors (GPCRs) are one of the most successful families of drug targets in medicine. More than a third of all approved drugs act on a GPCR: a 2025 analysis counted 516 approved drugs, around 36% of the total, acting on 121 receptors [18]. Beta-blockers for heart disease, antihistamines for allergy, opioids for pain, antipsychotics for schizophrenia and the GLP-1 receptor agonists used for diabetes and obesity all work by binding to members of this one receptor family.

The human genome encodes around 800 GPCRs [3], and only around 120 of them are targeted by approved drugs [18]. Hundreds of receptors with strong links to disease remain undrugged, including many so-called orphan GPCRs whose natural ligands are still unknown. The reasons are partly biological and partly technical: GPCRs are embedded in the cell membrane, they constantly shift between conformations, and much of their signalling machinery is built from flexible, intrinsically disordered regions that conventional structure-based methods struggle to address.

What follows covers GPCR biology and pharmacology, the technical reasons so many receptors remain undrugged, and where DNA-encoded library (DEL) screening fits in, particularly when the selection is run inside living cells.

All GPCRs share a common architecture of seven transmembrane alpha-helices (7TM) that span the lipid bilayer. The helices are connected by three extracellular loops and three intracellular loops. An extracellular N-terminus, which ranges from a few dozen residues to several hundred, often contributes to ligand recognition. An intracellular C-terminal tail interacts with signalling and regulatory proteins [4].

Human GPCRs are usually grouped into families based on sequence and structural similarity. The GRAFS system divides them into Glutamate, Rhodopsin, Adhesion, Frizzled/Taste2 and Secretin families [3], while pharmacologists often use the closely related classes A, B, C and F.

Receptor numbers are approximate and based on [3] and the GPCRdb resource [6].

When an agonist binds, the receptor undergoes a conformational change. The most prominent feature is an outward movement of transmembrane helix 6, which opens a cavity on the intracellular side [5]. A heterotrimeric G protein engages the activated receptor, which promotes the exchange of GDP for GTP on the Gα subunit. The G protein then dissociates into Gα and Gβγ, which regulate effectors such as adenylyl cyclase, phospholipase C and ion channels. The four main G protein families, Gs, Gi/o, Gq/11 and G12/13, give each receptor its characteristic downstream signature [5].

Signalling is switched off, and in some cases redirected, by G protein-coupled receptor kinases (GRKs) and arrestins. GRKs phosphorylate the receptor’s intracellular loops and C-terminal tail. Beta-arrestins then bind the phosphorylated receptor, block further G protein coupling, drive internalisation and can initiate their own signalling cascades [5]. As discussed below, the receptor regions that make these contacts are often flexible and partly disordered, which has important consequences for drug discovery.

For researchers, the GPCRdb database collects sequences, structures, mutations and ligand data for the whole receptor family and is a standard starting point for target assessment [6].

GPCRs are druggable because their binding sites are reachable from outside the cell, many have pockets that fit small molecules, and their response can be amplified and finely tuned:

• Accessibility. Most GPCRs present their ligand-binding site on the cell surface, so drugs do not need to cross the plasma membrane to reach it.
• Defined binding pockets. Many class A receptors have a deep orthosteric pocket within the transmembrane bundle that evolved to bind small molecules such as amines and lipids, an ideal fit for oral drugs [4].
• Signal amplification. A single activated receptor can activate many G proteins, so modest receptor occupancy can produce a strong physiological effect.
• Tissue-specific expression. Many GPCRs are expressed in restricted cell types, offering a route to selective effects.
• Tunable pharmacology. Ligands can be designed as full agonists, partial agonists, antagonists, inverse agonists or allosteric modulators, giving chemists fine control over the response [4].

Against these advantages sits a large gap. Of the roughly 400 non-olfactory human GPCRs, around 120 are targeted by approved drugs [18]. The remainder include more than 100 orphan receptors whose endogenous ligands have not been identified or confirmed [7]. Many of these orphans are associated with diseases of high unmet need through genetic or expression data, but without a known ligand they lack the tool compounds needed to validate them and the starting points needed to develop drugs against them.

Deorphanisation remains one of the most intractable problems in GPCR drug discovery [7]. Without a reference ligand there is often no functional assay to build a screen around, and no chemical starting point for medicinal chemistry. Binding-based screening methods, which do not require a functional readout, are therefore well suited to this group of receptors.

For decades, GPCR drug discovery meant finding compounds that compete with the natural ligand at its binding site. Over the past fifteen years, structural biology and new pharmacological concepts have broadened that view.

Orthosteric ligands bind the same site as the endogenous agonist. Because this pocket is often conserved across closely related receptors, such as the five muscarinic acetylcholine receptor subtypes, orthosteric drugs can be hard to make selective.

Allosteric modulators bind elsewhere on the receptor and change how it responds to its natural ligand. Positive allosteric modulators (PAMs) enhance the response, negative allosteric modulators (NAMs) reduce it, and silent allosteric modulators occupy the site without effect. Because allosteric sites tend to be less conserved, they can offer greater subtype selectivity, and their effect depends on the presence of the endogenous ligand, which can preserve the natural timing and location of signalling. Structural studies have mapped allosteric sites across the receptor surface: in the extracellular vestibule, on the lipid-facing outside of the transmembrane bundle, and on the intracellular face where G proteins and arrestins bind [8]. Approved examples include cinacalcet, a PAM of the calcium-sensing receptor, and maraviroc, a NAM of the chemokine receptor CCR5.

A GPCR does not simply switch between “on” and “off”. Different ligands can stabilise different active conformations, each favouring a particular set of downstream partners. A biased agonist activates one pathway, for example G protein signalling, while activating another, such as beta-arrestin recruitment, much less [9]. The hope is to separate therapeutic effects from side effects carried by a different pathway, an idea tested most intensively at the mu-opioid receptor, where G protein-biased agonists were pursued as analgesics with reduced respiratory depression. The clinical value of bias has proven more complex than first hoped, but it remains a central concept in modern GPCR pharmacology [9].

The number of experimentally determined GPCR structures has grown from a handful to well over a thousand, driven first by receptor engineering and lipidic cubic phase crystallography and more recently by cryo-electron microscopy (cryo-EM). Cryo-EM is particularly suited to GPCRs in complex with G proteins, arrestins and peptide ligands, which are difficult to crystallise, and these structures have directly informed the design of clinical candidates [8], [11].

Structures also enable virtual screening at very large scale, with docking campaigns now routinely covering hundreds of millions of make-on-demand compounds. Structure-based methods still depend on a good model of the relevant receptor state, however. They perform less well where the binding site is shallow, highly dynamic or formed by flexible regions that are poorly resolved in structures.

The glucagon-like peptide-1 receptor (GLP-1R), a class B1 receptor, is a striking recent example. Its activation stimulates glucose-dependent insulin secretion, slows gastric emptying and reduces appetite. Peptide agonists such as liraglutide and semaglutide were developed first for type 2 diabetes. In the STEP 1 trial, weekly semaglutide produced a mean weight loss of about 15% in adults with obesity [10], and the dual GIP/GLP-1 receptor agonist tirzepatide has since produced larger reductions still.

The next step was orally available, non-peptide small-molecule agonists. The cryo-EM structure of GLP-1R bound to LY3502970 (orforglipron), reported in 2020, showed how this already characterised compound engages the receptor: it occupies a binding pocket in the upper helical bundle, quite different from the site used by the natural peptide [11]. Orforglipron has since been approved by the FDA as a once-daily oral treatment for adults with obesity, or overweight with weight-related comorbidities [20]. That a class B receptor, long regarded as a poor fit for small molecules, can be activated by an oral small molecule suggests that some “difficult” GPCRs are waiting for the right screening approach rather than being undruggable.

Despite the family’s track record, GPCRs present real technical obstacles. They explain why so many receptors remain undrugged, and which screening strategies are most likely to succeed.

GPCRs are integral membrane proteins, typically expressed at low levels and unstable once removed from the lipid bilayer. Purifying them requires detergents, nanodiscs or other membrane mimetics, and often extensive engineering: thermostabilising point mutations, fusion partners inserted into intracellular loops, truncation of flexible termini, and stabilising antibodies or nanobodies. These tools have made structural biology possible, but each modification risks shifting the receptor’s conformational balance. A ligand discovered against an engineered, detergent-solubilised receptor may not behave the same way against the native protein in a cell.

GPCRs exist as ensembles of inactive, intermediate and active conformations, and ligands work by shifting the balance among them [5]. A purified receptor locked into one state presents only a subset of the binding sites available in a living cell, where the receptor interacts with lipids, G proteins, arrestins, kinases and scaffolding proteins. Sites that exist only in particular states, or only at the interface with a partner protein, are easy to miss.

Classical GPCR high-throughput screening (HTS) relies on functional cell assays measuring calcium flux, cAMP levels, reporter gene expression or arrestin recruitment. These assays are powerful but come with constraints. They need a known agonist or a measurable baseline, which orphan receptors often lack. They are biased towards a single signalling readout. They are also typically limited to corporate libraries of one to a few million compounds, a tiny sample of drug-like chemical space. For receptors where chemical starting points are rare, that sample is often too small.

GPCRs are also not entirely folded proteins. Intrinsically disordered proteins and regions (IDPs and IDRs) lack a single stable three-dimensional structure and are widespread in signalling proteins, where their flexibility lets them bind multiple partners and be finely regulated by post-translational modifications [12]. Analyses of the GPCR fold have shown that while the 7TM core is well structured, the N-terminus, the C-terminal tail and the third intracellular loop (ICL3) are frequently long and disordered [13].

The C-terminal tail and ICL3 carry the serine and threonine residues phosphorylated by GRKs. The pattern of phosphorylation, often described as a phosphorylation “barcode”, determines how beta-arrestin binds and which functional consequences follow [14]. ICL3 and the C-terminal tail also bind scaffolding proteins, PDZ-domain proteins and other regulators that steer receptor trafficking and signalling.

Disordered regions are usually removed or replaced to make receptors crystallise, and they are often invisible in cryo-EM maps. They are therefore largely absent from the models used in structure-based design, and conventional docking cannot target a binding site that has no fixed shape. Yet intracellular sites can be highly selective and pharmacologically powerful. Crystal structures of the chemokine receptor CCR9 revealed an antagonist binding to an intracellular pocket that overlaps with the G protein binding site [15]. Small molecules that bind the intracellular face of a receptor, or the flexible regions that recruit G proteins and arrestins, could in principle block or redirect signalling in ways orthosteric ligands cannot. Finding them requires screening methods that do not depend on a pre-defined, rigid binding pocket.

A DNA-encoded library is a mixture of small molecules in which every compound carries a unique DNA barcode recording its synthetic history. Because the library is pooled rather than arrayed, hundreds of millions to billions of compounds can be screened in a single tube. Binders are separated from non-binders, their barcodes are amplified and read by next-generation sequencing, and the enriched structures are decoded and resynthesised for validation [16].

The approach has moved into mainstream drug discovery. DEL-derived compounds have entered clinical development, and reviews of the field now describe DEL as one of the principal modern routes to hit identification, complementing HTS, fragment-based approaches and virtual screening [16].

Yes, although membrane proteins present additional technical challenges. Most early DEL campaigns used purified soluble proteins immobilised on beads or tagged for capture, which suits enzymes and soluble receptor domains. Membrane proteins were harder, because immobilising a GPCR usually means detergent extraction and engineering, with the attendant risk of losing the native conformation.

Several groups have nevertheless reported successful DEL screens against membrane targets, using nanodiscs, virus-like particles, membrane preparations or whole cells as the presentation format [16]. DEL screening has also been applied to a GPCR directly: a selection against the purified beta2-adrenergic receptor identified a negative allosteric modulator, an allosteric “beta-blocker” that acts outside the orthosteric pocket [19]. Solution-phase and cell-based selection methods are particularly relevant, because they avoid immobilisation artefacts and let the receptor stay in a lipid environment. Our article on receptors as drug targets covers ligand discovery across receptor classes in more detail.

Vipergen’s screening technologies are designed to reduce the dependence on immobilised, purified target protein. Our Binder Trap Enrichment (BTE) technology performs the selection in solution rather than on a solid support. The library is incubated with a DNA-tagged target, and the mixture is then diluted by a large factor so that complexes begin to dissociate. After a defined dissociation time, the sample is emulsified into water-in-oil droplets. A library member still bound to the target ends up in the same droplet as the target’s DNA, where the two DNA strands are ligated. Unbound library members share a droplet with a target only by chance, so after amplification and sequencing, candidate binders are identified through enrichment of their DNA barcodes relative to background. Because there is no immobilisation and no washing step, the method avoids the surface artefacts of bead-based selections, and the dissociation time gives direct control over the stringency of the selection.

The same principle makes cell-based DEL screening possible, with the selection performed inside living cells. Cellular Binder Trap Enrichment (cBTE) was the first published method for screening a multi-million-member DEL inside a living cell, using Xenopus laevis oocytes whose large volume allows direct injection of the library. In that study, a 194-million-member DEL was screened against three pharmaceutically relevant targets—p38α, ACSS2 and DOCK5—without requiring highly purified active target proteins [17]. The target is expressed in the oocyte from injected mRNA as a fusion with a prey protein. A small-molecule bait attached to DNA, which binds the prey specifically, is microinjected together with the library, so that the target becomes labelled with DNA. After binding inside the cell, the oocytes are lysed, diluted and emulsified, and the library DNA of bound compounds is ligated to the bait DNA in the droplets, as in BTE. Performing the selection in a living cell removes the need to purify the target, and the cellular environment can preserve membrane localisation, relevant conformational states and interactions that may be lost during purification. Xenopus oocytes are a heterologous expression system, however, so not every native partner or modification of a human receptor can be assumed to be present. Because the library is delivered directly into the cytoplasm, cBTE may enable screening of accessible intracellular regions of GPCRs without requiring the DNA-encoded compounds to cross the plasma membrane.

The potential advantages for GPCR programmes are:

• No receptor purification required. This removes a major bottleneck for membrane-target hit finding.
• Cellular conformational states. The receptor may access physiologically relevant states that could be altered or lost during purification.
• No functional assay required for initial selection. Binding-based screening may enable hit identification for orphan receptors without a known agonist, provided a suitable target presentation can be established.
• Large chemical coverage. Hundreds of millions of compounds can be tested in a single experiment; the published cBTE study screened a 194-million-member library [17], far beyond what plate-based HTS can reach.
• Potential access to intracellular and interface sites. This may include flexible regions involved in G protein and arrestin coupling, which must be validated experimentally for each target.

Intrinsically disordered proteins have long been described as undruggable, because they have no fixed pocket to dock into. That reputation is changing: small molecules have been found that bind disordered regions, often by engaging transient conformations or by binding at the interface formed when a disordered region folds upon binding a partner [12].

Because DEL selection reports on binding to whatever conformational ensemble the target presents, rather than on docking into a predefined pocket, it is in principle well suited to this class of target. Screening in a cellular context may strengthen that advantage, because the disordered regions are present in a cellular setting rather than truncated or engineered away. For GPCRs specifically, this could open a route to compounds acting on the C-terminal tail, the third intracellular loop and the receptor–transducer interface, regions that are central to signalling and almost entirely untouched by current drugs.

A DEL campaign against a GPCR follows a defined sequence. Timelines depend on target complexity, screening format, hit validation and follow-up chemistry.

1. Target definition and format selection. Decide how the receptor will be presented: in cells, in membranes, in nanodiscs or as a stabilised construct. Agree which receptor state matters, for example agonist-bound, apo or in complex with a transducer.
2. Selection. The library is incubated with the target under conditions that preserve native structure. Comparative selections, with and without a known ligand, with a mutant receptor or against a control cell line, help establish where hits bind.
3. Sequencing and data analysis. Barcodes are read by next-generation sequencing and enrichment is evaluated statistically. Structure–activity relationships often emerge directly from the sequencing data, because related compounds in the library enrich to different degrees.
4. Off-DNA resynthesis. Representative compounds are synthesised without their DNA tag.
5. Hit validation. Binding is confirmed by orthogonal biophysical methods such as surface plasmon resonance or thermal shift, and by radioligand or functional assays where a readout exists.
6. Hit expansion and lead optimisation. Validated chemotypes feed into medicinal chemistry, supported by structural work where structures are obtainable.

DEL screening identifies candidate binders. Where they bind, how selective they are and what they do to the receptor all have to be established afterwards, in orthogonal binding and functional assays.

• Orthosteric ligands, competing with the endogenous agonist, identified by competition selections against a known ligand where one exists, in formats that expose the extracellular side of the receptor, such as membrane preparations or nanodiscs.
• Allosteric modulators, binding extracellular vestibules, lipid-facing surfaces or intracellular sites, often with better subtype selectivity than orthosteric compounds.
• Ligands for orphan receptors, where a binding-based readout removes the need for a functional assay and can provide starting points for developing tool compounds.
• Interface and disordered-region binders, which could potentially modulate coupling to G proteins, GRKs or arrestins rather than blocking the agonist site.
• Novel chemotypes, since DEL chemical space is largely orthogonal to legacy HTS collections and has repeatedly produced scaffolds absent from existing screening decks [16].

Protein consumption is also modest. A single selection uses a small amount of target material, which matters for receptors that are hard to express. Combined with the fact that no functional assay is needed, this makes DEL a realistic option for receptors that would never justify the investment required to build a full HTS campaign.

GPCRs are among the best validated target families in medicine and, at the same time, one of the least exhausted. More than a third of approved drugs act on around 120 receptors, which leaves most of the remaining non-olfactory receptors, including more than a hundred orphans, with little or no chemical matter against them. The bottleneck is hit finding rather than biology: membrane proteins are hard to work with, receptors move between conformations, and some functionally critical regions have no fixed structure at all.

The technologies that close that gap are the ones that do not require a purified, rigid, assay-ready target. DNA-encoded library screening tests hundreds of millions of compounds in a single binding experiment, needs no functional readout, and consumes very little target material. BTE runs the selection in solution, without immobilising the target, and cBTE extends the same principle to living cells. For a difficult GPCR, screening in a cell may give access to receptor states and intracellular sites that purification puts out of reach. These are possibilities to test against a specific target, not results that hold for every receptor.

If you are working on a GPCR that has resisted conventional screening, an orphan receptor with no tool compounds, or an intracellular interface that structure-based design cannot reach, we would be glad to discuss whether a DEL campaign fits your target.

Talk to our scientists about your GPCR target

1. Hauser, A.S., Attwood, M.M., Rask-Andersen, M., Schiöth, H.B., Gloriam, D.E. Trends in GPCR drug discovery: new agents, targets and indications. Nat Rev Drug Discov (2017), 16(12), 829-842. DOI: 10.1038/nrd.2017.178. https://doi.org/10.1038/nrd.2017.178
2. Sriram, K., Insel, P.A. G protein-coupled receptors as targets for approved drugs: how many targets and how many drugs? Mol Pharmacol (2018), 93(4), 251-258. DOI: 10.1124/mol.117.111062. https://doi.org/10.1124/mol.117.111062
3. Fredriksson, R., Lagerström, M.C., Lundin, L.-G., Schiöth, H.B. The G-protein-coupled receptors in the human genome form five main families. Phylogenetic analysis, paralogon groups, and fingerprints. Mol Pharmacol (2003), 63(6), 1256-1272. DOI: 10.1124/mol.63.6.1256. https://doi.org/10.1124/mol.63.6.1256
4. Rosenbaum, D.M., Rasmussen, S.G.F., Kobilka, B.K. The structure and function of G-protein-coupled receptors. Nature (2009), 459(7245), 356-363. DOI: 10.1038/nature08144. https://doi.org/10.1038/nature08144
5. Weis, W.I., Kobilka, B.K. The molecular basis of G protein-coupled receptor activation. Annu Rev Biochem (2018), 87, 897-919. DOI: 10.1146/annurev-biochem-060614-033910. https://doi.org/10.1146/annurev-biochem-060614-033910
6. Pándy-Szekeres, G., Caroli, J., Mamyrbekov, A., et al. GPCRdb in 2023: state-specific structure models using AlphaFold2 and new ligand resources. Nucleic Acids Res (2023), 51(D1), D395-D402. DOI: 10.1093/nar/gkac1013. https://doi.org/10.1093/nar/gkac1013
7. Laschet, C., Dupuis, N., Hanson, J. The G protein-coupled receptors deorphanization landscape. Biochem Pharmacol (2018), 153, 62-74. DOI: 10.1016/j.bcp.2018.02.016. https://doi.org/10.1016/j.bcp.2018.02.016
8. Thal, D.M., Glukhova, A., Sexton, P.M., Christopoulos, A. Structural insights into G-protein-coupled receptor allostery. Nature (2018), 559(7712), 45-53. DOI: 10.1038/s41586-018-0259-z. https://doi.org/10.1038/s41586-018-0259-z
9. Smith, J.S., Lefkowitz, R.J., Rajagopal, S. Biased signalling: from simple switches to allosteric microprocessors. Nat Rev Drug Discov (2018), 17(4), 243-260. DOI: 10.1038/nrd.2017.229. https://doi.org/10.1038/nrd.2017.229
10. Wilding, J.P.H., Batterham, R.L., Calanna, S., et al. Once-weekly semaglutide in adults with overweight or obesity. N Engl J Med (2021), 384(11), 989-1002. DOI: 10.1056/NEJMoa2032183. https://doi.org/10.1056/NEJMoa2032183
11. Kawai, T., Sun, B., Yoshino, H., et al. Structural basis for GLP-1 receptor activation by LY3502970, an orally active nonpeptide agonist. Proc Natl Acad Sci U S A (2020), 117(47), 29959-29967. DOI: 10.1073/pnas.2014879117. https://doi.org/10.1073/pnas.2014879117
12. Wright, P.E., Dyson, H.J. Intrinsically disordered proteins in cellular signalling and regulation. Nat Rev Mol Cell Biol (2015), 16(1), 18-29. DOI: 10.1038/nrm3920. https://doi.org/10.1038/nrm3920
13. Venkatakrishnan, A.J., Deupi, X., Lebon, G., et al. Structured and disordered facets of the GPCR fold. Curr Opin Struct Biol (2014), 27, 129-137. DOI: 10.1016/j.sbi.2014.08.002. https://doi.org/10.1016/j.sbi.2014.08.002
14. Nobles, K.N., Xiao, K., Ahn, S., et al. Distinct phosphorylation sites on the beta2-adrenergic receptor establish a barcode that encodes differential functions of beta-arrestin. Sci Signal (2011), 4(185), ra51. DOI: 10.1126/scisignal.2001707. https://doi.org/10.1126/scisignal.2001707
15. Oswald, C., Rappas, M., Kean, J., et al. Intracellular allosteric antagonism of the CCR9 receptor. Nature (2016), 540(7633), 462-465. DOI: 10.1038/nature20606. https://doi.org/10.1038/nature20606
16. Peterson, A.A., Liu, D.R. Small-molecule discovery through DNA-encoded libraries. Nat Rev Drug Discov (2023), 22(9), 699-722. DOI: 10.1038/s41573-023-00713-6. https://doi.org/10.1038/s41573-023-00713-6
17. Petersen, L.K., Christensen, A.B., Andersen, J., et al. Screening of DNA-encoded small molecule libraries inside a living cell. J Am Chem Soc (2021), 143(7), 2751-2756. DOI: 10.1021/jacs.0c09213. https://doi.org/10.1021/jacs.0c09213
18. Lorente, J.S., et al. GPCR drug discovery: new agents, targets and indications. Nat Rev Drug Discov (2025), 24(6), 458-479. DOI: 10.1038/s41573-025-01139-y. https://doi.org/10.1038/s41573-025-01139-y
19. Ahn, S., Kahsai, A.W., Pani, B., et al. Allosteric “beta-blocker” isolated from a DNA-encoded small molecule library. Proc Natl Acad Sci U S A (2017), 114(7), 1708-1713. DOI: 10.1073/pnas.1620645114. https://doi.org/10.1073/pnas.1620645114
20. Healio. FDA approves orforglipron, an oral GLP-1 for adults with obesity. Healio (1 April 2026). https://www.healio.com/news/endocrinology/20260401/fda-approves-orforglipron-an-oral-glp1-for-adults-with-obesity