Four Reasons to Measure GPCR Signaling Bias in Drug Discovery

Introduction

G protein-coupled receptors (GPCRs) don't simply switch on or off. Most are pleiotropically coupled to multiple intracellular pathways, and different ligands at the same receptor can produce fundamentally different cellular outcomes, a phenomenon known as GPCR signaling bias, or functional selectivity. Two molecules with identical binding affinity can diverge sharply in their downstream effects: one emphasizing G protein activation, another driving receptor internalization and β-arrestin recruitment, each producing a distinct pharmacological fingerprint.

Measuring GPCR signaling bias in drug discovery is critical, as it forms a selection criterion for agonism. Measuring and quantifying signaling bias reveals which candidates emphasize therapeutically beneficial pathways, and which may be falsely characterized as equivalent by single-pathway assays. Drug discovery programs that do not consider investigating for signaling bias often fail to fully understand the distinction and risk candidate molecules from advancing along the pipeline.

In this article, you'll learn:

• Four reasons why bias measurements improve candidate selection at every stage of the discovery pipeline

• How biased ligands can emphasize beneficial signaling pathways, and actively suppress harmful ones 

• How bias can expand pursuit of targets previously considered too toxic or undruggable 

• Why measuring signaling bias across multiple pathways gives a more accurate picture of true molecular selectivity

Obtaining a GPCR Pathway Pharmacological Fingerprint

To obtain a pharmacological fingerprint (or profile) for GPCR therapeutics, it is best to show how signaling bias and receptor selectivity can be quantified using cell-based assays.  These assays offer a simple approach to the comparison of agonist profiles across different pathways and biased signaling. For example, Figure 1 shows cell-based assay results from the cAMP G protein activation pathway and the β-arrestin recruitment pathway for several incretins (metabolic hormones and therapeutics) acting on GLP-1 and GIP receptors. Both assays provide unique profiles and a preferred receptor signaling bias for the therapeutics tested.  

Figure 1. Dose-response curves for incretin agonists on GLP-1 and GIP receptors. Comparison of the curves reveals the relative selectivity of retratutide for GLP-1 receptor over GIP receptor compared to tirzepatide and yields sufficient data to calculate the signaling bias of the agonists for β-arrestin over cAMP responses. This data is calculated from the estimated max and EC₅₀.

How GPCR Signaling Bias Strengthens Drug Discovery Programs

There are four distinct reasons why bias measurements improve drug discovery outcomes.

1. Bias can make better drugs by emphasizing beneficial signaling pathways and de-emphasizing harmful ones. Opioid receptors offer clear illustration of this: G protein-biased agonists at the μ-opioid receptor have been explored as a strategy to preserve analgesia, while reducing β-arrestin-dependent adverse effects such as respiratory depression and constipation (Raehal et al., 2005; DeWire et al., 2013). Another example is at the angiotensin AT1 receptor, biased ligands like TRV120027 block deleterious vasoconstriction, while engaging β-arrestin signals that provide beneficial effects in heart failure (Violin et al., 2010).

   
   

2. Bias measurements can identify structurally differentiated hits from high-throughput screens. Two hits from a primary screen may appear equivalent in a single-pathway assay but diverge significantly when tested in orthogonal functional assays. Counter-screening in biased assays distinguishes molecules that are genuinely different on a molecular level, and therefore more likely to produce distinct phenotypes in complex therapeutic models.

   
   

3. Bias reduces complex efficacy profiles to measurable, optimizable scales for medicinal chemistry. Efficacy (comprising of both quality and quantity attributes for different agonists) reveals how an expected overall cellular response can be achieved from cellular signals. The quality attribute can be captured by bias measurements allowing for the reduction of complex phenotypes to graded activation of signaling pathways. Once a favorable efficacy fingerprint is identified in therapeutic cells (through cell-based assays), medicinal chemists can work backward to amplify or tune the relevant bias.
   

4. Bias measurements are essential for accurate selectivity profiling. Quantifying bias offers a way to determine selectivity based on all known signaling for a given molecule. For instance, a compound that appears highly selective at a target receptor based on cAMP readouts alone may show far less selectivity when β-arrestin signaling is included. The β2-adrenoceptor bronchodilator clenbuterol, for example, exhibits 500-fold selectivity for β2 over β1 receptors in cyclic AMP assays, but exhibits a reduced selectivity (approximately 5.7-fold) when β-arrestin-mediated effects are measured (Casella et al., 2011). Without multi-pathway assays, selectivity assessments can be misleading.

Emphasizing Beneficial and De-emphasizing Harmful Signaling Pathways

One of the most powerful applications of GPCR biased signaling is the ability to separate therapeutic effects from adverse ones at the receptor level, not achieved with binary pharmacology. As different ligands stabilize different receptor conformations, it is possible to design molecules that selectively engage the beneficial pathways over harmful ones.

The opioid system offers a highly studied example. Morphine provides effective analgesia but carries debilitating side effects including respiratory depression. Studies in β-arrestin knockout mice demonstrated that morphine produces significantly less respiratory depression in the absence of β-arrestin 2 signaling — pointing directly to a G protein-biased opioid agonist as a potentially superior analgesic, one that preserves pain relief while reducing a life-threatening adverse effect (Raehal et al., 2005; DeWire et al., 2013).

The angiotensin system illustrates a more nuanced application: not just de-emphasizing a harmful pathway but blocking it while simultaneously preserving a beneficial one. In congestive heart failure, elevated angiotensin signaling raises arterial pressure resulting in the failure of the myocardium. Standard angiotensin receptor blockers like losartan address this, but at the cost of eliminating beneficial β-arrestin-mediated signals. Biased ligands such as TRV120027 block the deleterious G protein-driven effects, while retaining the beneficial β-arrestin signals offering a meaningfully improved therapeutic profile (Violin et al., 2010).

Expanding the Druggable Target Space Through Bias

Beyond refining the pharmacology of established targets, biased signaling can rehabilitate entire target classes previously considered too toxic to pursue. The κ-opioid receptor illustrates this directly. κ-Opioid agonists carry genuine therapeutic potential in mood, cognition, and addiction but also produce serious dysphoria, which has historically precluded clinical development. Biased κ-opioid agonists that reduce dysphoric signaling, while preserving beneficial effects, offer a route into a target class that unbiased pharmacology cannot safely access (White et al., 2014). Rather than abandoning a target because of a harmful pathway, bias offers a different answer: design around the liability.

Conclusion

Bias measurements reveal that efficacy has quality as well as quantity attributes, and that quality can be engineered. A biased ligand can be designed to favor pathways.  For example, a pathway that build bones, relieve pain, or stabilize a failing heart, while avoiding pathways that cause respiratory depression, dysphoria, or dangerous arterial pressure. In the case of κ-opioid agonists, bias may be the only route by which an otherwise excluded target class becomes clinically viable at all.

Bias quantification increases the value of known lead compounds, sharpens selectivity assessments, and provides medicinal chemists with graded, optimizable scales to work from. Programs that characterize signaling bias early — across G protein, β-arrestin, and second messenger pathways through cell-based assay assessments — carry forward candidates whose vivo behavior can be fully understood providing a meaningful selection criteria and competitive advantage at every stage of drug discovery.

Eurofins DiscoverX provides the largest portfolio of GPCR assays and a unique service that utilizes state-of-the-art tools developed by Professor Terry Kenakin at the University of North Carolina School of Medicine for characterization of ligand bias. With the appropriate β-arrestin, internalization, and second messenger assays to quantify selective response and statistical tools to scale these effects, harnessing bias to produce selective ligands is now made simple.

For further reading on GPCR biased signaling and assay methodologies, explore the ‘Insights into GPCR Drug Discovery and Development’ eBook and  ‘GPCR Functional Cell-based Assays – Assessing Biased Signaling of Agonists’ White Paper by Kenakin, T. et al. (2025). 

Visit Eurofins DiscoverX GPCR Products and Solutions to explore the full portfolio of GPCR assays from Eurofins DiscoverX.

References

1. Raehal, K.M., Walker, J.K., & Bohn, L.M. (2005). Morphine side effects in beta-arrestin 2 knockout mice. Journal of Pharmacology and Experimental Therapeutics, 314(3), 1195–1201. https://doi.org/10.1124/jpet.105.087254

   
   

2. DeWire, S.M., Yamashita, D.S., Rominger, D.H., Liu, G., Cowan, C.L., Graczyk, T.M., … Violin, J.D. (2013). A G protein-biased ligand at the μ-opioid receptor is potently analgesic with reduced gastrointestinal and respiratory dysfunction compared with morphine. Journal of Pharmacology and Experimental Therapeutics, 344 (3), 708–717. https://doi.org/10.1124/jpet.112.201616
   

3. Violin, J.D., DeWire, S.M., Yamashita, D., Rominger, D.H., Nguyen, L., Schiller, K., … Lark, M.W. (2010). Selectively engaging β-arrestins at the angiotensin II type 1 receptor reduces blood pressure and increases cardiac performance. Journal of Pharmacology and Experimental Therapeutics, 335 (3), 572–579. https://doi.org/10.1124/jpet.110.173005
   

4. Casella, I., Ambrosio, C., Grò, M.C., Molinari, P., & Costa, T. (2011). Divergent agonist selectivity in activating β1- and β2-adrenoceptors for G protein and arrestin coupling. Biochemical Journal, 438 (1), 191–202. https://doi.org/10.1042/BJ20110374
   

5. Kenakin, T., Watson, C., Muniz-Medina, V., Christopoulos, A., & Novick, S. (2012). A simple method for quantifying functional selectivity and agonist bias. ACS Chemical Neuroscience, 3 (3), 193–203. https://doi.org/10.1021/cn200111m
   

6. Kenakin, T. (2019). Biased receptor agonism. Annual Review of Pharmacology and Toxicology, 59, 245–267. https://doi.org/10.1146/annurev-pharmtox-010818-021139