Signaling Bias as a Framework for Therapeutically Relevant Orthosteric Agonists: The β₂ Adrenergic Receptor Paradigm

Abstract

The β₂ adrenergic receptor (β₂AR) is a prototype receptor for which the agonists isoetharine, salbutamol, the naturally occurring agonist epinephrine (adrenaline), salmeterol, and formoterol can activate distinct pathways through selective engagement with Gs, Gi, or β-arr/GRK proteins. The selectivity profile arises from the engagement of different transducers with distinct receptor conformations. Understanding the transient conformations involved in preferred pathways relative to deleterious ones is critical for drug design and development, and current pharmacological research is heavily focused on exploring this concept. A recent study by Sykes, Veprintsev, Grether, Rufer, and collaborators in 2025 highlighted that the use of BRET-based technology with suitable G protein and β-arrestin biosensors could analyze β₂AR ligand efficacy across these two distinct signaling pathways. This was measured for β₂AR ligands, including short- and long-acting β₂AR agonists (SABAs and LABAs), which are effective therapies for asthma and COPD, respectively. SABAs, such as the older isoetharine, showed a minor G protein bias and were β-arrestin-biased ligands. However, more recent drugs, such as salbutamol, demonstrate significant G protein bias. Indeed, over the last 50 years, there has been a move toward a G protein bias, reflecting efforts to enhance asthma therapies. Among LABAs, salmeterol and the SABA salbutamol show greater G protein vs β-arrestin bias compared to that of balanced ligands, such as the LABA formoterol or the SABA epinephrine (Figure ).

1.

Biased signaling of β₂AR induced by natural and synthetic agonists of therapeutic relevance.

Salmeterol, a long-chain analog of salbutamol and the only FDA-approved bitopic drug (Figure ), is a biased agonist toward Gs due to temporal bias (“kinetic bias”) for coupling to β-arrestin (Figure ). This suggests that salmeterol may induce slow phosphorylation of β₂AR and β-arrestin coupling. By comparison, the difference in binding mode between salmeterol and epinephrine can provide structural insight into these effects (Figure ).

2.

Binding of (Α) epinephrine (PDB ID 4LDO) and (Β) salmeterol (PDB ID 6MXT ) inside the orthosteric binding site of β₂AR. Protein is illustrated using an indigo cartoon representation; some amino acids, including those in the orthosteric binding sites, are shown as sticks; the carbon atoms of epinephrine are shown in green sticks, and those of salmeterol in orange sticks. Hydrogen bonds are indicated with yellow dashes, while π–π interactions are shown in light blue dashes.

Indeed, structure with PDB ID 6MXT which corresponds to salmeterol−β₂AR–Nb71 complex (Nb71 acts as a G protein mimicking nanobody stabilizing the fully activated conformation), reported by Kobilka, Zhang, and collaborators in 2018, can be compared to the X-ray crystal structure (PDB ID 4LDO) of the full agonist epinephrine−β₂AR–Nb80 (Nb90 acts as a G protein-mimicking nanobody), reported by Kobilka, Garcia, and collaborators in 2013. In more detail, salmeterol forms π–π interactions with F194^ECL2, Y308^7.35, and H296^6.58, and, like epinephrine, hydrogen-bonding interactions between its alkylamino and β-hydroxyl groups with D113^3.32 and N312^7.39, and between the saligenin moiety and S203^5.42 and S207^5.46. Furthermore, salmeterol is stabilized in the exosite through a critical hydrogen bond formed between its ether oxygen and F193^ECL2. Thus, in contrast to epinephrine and salbutamol, salmeterol binds through its long aryloxy alkyl chain to an exosite located between ECL2 and ECL3, as well as with the extracellular ends of TM6 and TM7, interacting with many residues. Such binding resulted in a longer residence time of salmeterol within the receptor, thereby producing a longer-lasting effect than epinephrine and salbutamol.

Additionally, epinephrine with a m-hydroxyl group in the catechol moiety forms strong hydrogen bonding interactions with N293^6.55, which in turn forms a hydrogen-bonded network with S204^5.43. In contrast, in salmeterol, which has in the saligenin moiety a hydroxymethyl group at the m-position instead of a hydroxyl group in epinephrine, the hydrogen bond N293^6.55 is missing (Figure ), and this was also shown in the MD simulations in ref . In combination with mutagenesis experiments, the results showed that salmeterol’s less extensive polar contacts with N293^6.55 may contribute to its decreased efficacy in activating Gs and the near lack (or much slower rate) of β-arrestin recruitment, presumably due to poor coupling caused by diminished GRK phosphorylation. It remains unclear, however, how these binding differences are precisely transferred from the orthosteric binding site to the cytoplasmic end, thereby enabling engagement of the transducer. For example, the weaker hydrogen bonding interaction of salmeterol with N293^6.55 can stabilize less TM5 in the conformation required for β-arrestin binding. Isoetharine and other β-arrestin biased ligands have been shown to stabilize for prolong time an active TM7 conformation for β-arrestin engagement, as reported, for example, by Milllar and collaborators using smTIRF microscopy in 2020.

Salmeterol is a full agonist at Gi recruitment while being a partial agonist at Gs. Temporal bias is also an important factor in the Gs-to-Gi transition of β₂AR coupling in the healthy heart, and understanding differences in the structural and binding profiles of biased agonists for Gs versus Gi proteins is critical. Indeed, Gi-mediated signaling is important for the development of therapeutics against β₂AR. Thus, in contrast to salmeterol, which favors Gi over Gs coupling, the fully biased-Gi agonist LM189 was recently identified using medicinal chemistry and structural biology techniques, as reported by Kobilka, Lerch, Gmeiner, and collaborators in 2025. In the cryo-EM structure of the agonist LM189−β₂AR–Gi complex (PDB ID 9BUY ) residue F139^34.51 of ICL2 inserts into the hydrophobic pocket of Gi. In LM189, developed in ref , or in epinephrine, a catechol ring replaces the saligenin ring in salmeterol (Figure ), while the lengthy chain is retained. As a result, LM189, like epinephrine and in contrast to salmeterol, forms stable hydrogen bonds with S203^5.42 and S207^5.46, as well as with N293^6.55 (Figures B, ). This major difference in hydrogen bonding with N293^6.55 may induce a transient conformation of the β₂AR involving the transmembrane domain and the receptor’s intracellular domain, thereby favoring LM189’s specificity for Gi. The orthosteric binding site of β₂AR in PDB ID 9BUY and the position of Cα5 are like in the X-ray structures (where applicable) in the complexes, salmeterol−β₂AR–Nb71 and epinephrine−β₂AR–Nb80 or super agonist BI-167107−β₂AR-Gs (correspondingly, PDB IDs 6MXT, 4LDO, or 3SN6). The most important changes were seen in conformations of TM5, TM6, and ICL2. The elongation of the TM5 helix was smaller in the Gi structure than in the inactive structure of β₂AR. The CW-EPR showed partial unwinding of ICL2 in the Gi compared to the Gs structures. The DEER measurements also showed a difference in rotation of the intracellular end of TM6 and a smaller by ∼3 Å TM6 opening in the LM189−β₂AR when Gi binds, compared to Gs, indicating slightly different conformations. Overall, when connected to Gi rather than Gs, the complex of LM189 stabilized a specific cytosolic TM6 structure and enhanced ICL2 dynamics.

3.

(A) Βinding of LM189 and (B) binding of the 3-(5-F-pyrridinyl) analog of salbutamol inside the orthosteric binding site β₂AR. The docking pose in (A) corresponds to PDB ID 9BUY while in (B) was generated from our docking calculations. Protein is illustrated as in Figure , carbon atoms of LM189 are shown in pink sticks, and carbon atoms of the F-ligand are shown in gray sticks.

Insufficient research has been performed on the conformations of GPCRs for binding to GRKs and their therapeutic relevance as drug targets. Due to β-arrestin-dependent desensitization and Gs/cAMP-induced cardiac side effects, β₂AR agonists are regarded as poor clinical options for the treatment of glycaemia. Recently, β₂AR agonists that favor GRK coupling via ligand-based virtual screening and chemical optimization were identified by Bengtsson, Wright, Lauschke, and collaborators. Using BRET-based assays, the presence of agonist-driven β₂AR conformations that facilitate GRK2 binding was shown. The structure–activity relationships revealed that compounds that do not form hydrogen bonds with S203^5.42 and/or S207^5.46 in TM5 are more efficient for coupling to GRK2 than Gs. These compounds, however, maintain hydrogen bonding interactions with D113^3.32 and N312^7.39. The treatment of metabolic disorders like type 2 diabetes and obesity may be possible with such β₂AR pathway-selective agonists, like the 3-(5-F-pyrridinyl) analog of salbutamol (Figure ) that favors GRK coupling.

Overall, we presented here structural models derived from experimental work for the development of biased ligands to β₂AR in the catecholamine series. It was shown that the coupling to different transducer proteins depends on subtle changes in ligand structure and binding interactions. The structure-based design and development of biased GPCR ligands is a recent advancement and under investigation. Additionally, the signaling bias is strongly associated with the use of reliable technology, a prerequisite for a proper understanding of the activity of such drug molecules. At the same time, biased ligands are important in modern drug design targeting GPCRs. However, the therapeutic benefits of biased agonists, including efficacy and safety, warrant further investigation.

The open access publishing of this article is financially supported by HEAL-Link.

The authors declare no competing financial interest.

Published as part of Biochemistry special issue “Biased Signaling”.