G-protein-coupled receptors (GPCRs) are the largest family of drug targets in the human genome, roughly one-third of all approved drugs work through them. But GPCRs do not signal through G proteins alone. When a receptor is activated, it is phosphorylated by GRK kinases, which recruits β-arrestin proteins. These β-arrestins then desensitize the receptor (physically blocking further G protein coupling), internalize it into the cell, and in some cases initiate their own signaling cascades.
For decades, researchers have wanted a way to modulate β-arrestin directly, to prevent receptor desensitization while preserving G protein signaling, or to block β-arrestin-dependent pathological signaling. They have had only genetic tools: knockout or knockdown. A study published June 24 in Nature from the laboratory of Robert Lefkowitz (Duke University) changes that, reporting the first small-molecule compounds that directly and selectively inhibit β-arrestins.
How the inhibitors work
The team, led by Sora Jin Kahsai, Natalia Pakharukova, HyeJin Kwon, and Vidisha Shah, screened 3,500 compounds from the NCI diversity set using differential scanning fluorimetry. They identified a series of compounds that bind to a previously unknown allosteric pocket on β-arrestin, the MCL site, formed by the middle loop (residues 129-140), the C-loop (241-249), and the lariat loop (274-300). This pocket is distinct from the N-domain phospho-recognition site where activated GPCRs normally bind.
The lead compound, Cmpd-5 (oridonin), stabilizes an inactive-like conformation of β-arrestin. Cryo-electron microscopy at 3.47 angstroms resolution shows it induces only an 8-degree interdomain twist, compared with the approximately 20-degree twist in the fully active state. The finger loop becomes flexible, physically occluding receptor tail binding. Key contacts include Y249 and C251 (hydrophobic interactions) and E134 and R285 (polar contacts). Two additional compounds, Cmpd-46 and Cmpd-64, show partial and β-arrestin-2-selective binding, respectively.
In vitro validation
In cell-based assays, the compounds produced clear effects. They inhibited β-arrestin recruitment to the β₂-adrenergic receptor by 77%, to the angiotensin AT₁ receptor by 82%, and to the GLP-1 receptor stimulated by semaglutide by 93%. Internalization of these receptors was reduced by 87-93%. Blocking β-arrestin recruitment meant G protein signaling was sustained: cAMP production through the β₂AR increased to 191-282% of control, and calcium mobilization through AT₁R rose to 279-490% of control. β-arrestin-dependent ERK signaling was attenuated, while EGF-induced ERK was unaffected, confirming specificity for the β-arrestin pathway.
Critically, the compounds did not affect G proteins (Gαs, Gαq, Gαi, Gα12), GRKs, or MAPK effectors (ERK2, SRC, p38α). They are selective for β-arrestins.
Physiological effects
In T cell migration assays, CCL19/CCR7-mediated migration was near-completely inhibited at 20 µM in CD4⁺, CD8⁺, and total T cells, suggesting applications in inflammatory and autoimmune conditions. In cardiomyocytes, the β-arrestin-biased AT₁R agonist TRV027 had its effects blocked by the modulators.
The therapeutic landscape
The implications span multiple therapeutic areas. In cardiovascular disease, blocking β-arrestin-mediated desensitization of the angiotensin receptor could preserve beneficial G protein signaling while preventing pathological β-arrestin effects in cardiomyocytes. In asthma, sustaining β₂-adrenergic receptor bronchodilation by preventing desensitization addresses a fundamental limitation of existing beta-agonists. In metabolic disease, the 93% inhibition of β-arrestin recruitment to the GLP-1 receptor suggests a potential strategy to extend the effectiveness of GLP-1 agonists like semaglutide. And in neurology and immunology, the T cell migration data point to inflammatory disease applications.
More broadly, the work establishes what the authors call transducer-targeted pharmacology: rather than designing drugs for each individual receptor, this approach targets the signaling protein that many receptors share. The MCL allosteric pocket, which was entirely unknown before this study, now provides a structural target for drug development across the GPCR superfamily.
Source: Kahsai, S.J., Pakharukova, N., Kwon, H. et al. Small-molecule modulation of β-arrestins. Nature (2026). DOI: 10.1038/s41586-026-10683-5