G-protein-coupled receptors (GPCRs) form the largest and most diverse family of membrane receptors in the human genome play critical roles in physiological processes. They are implicated in cancer, cardiovascular and metabolic diseases, and neurological disorders, making them central targets for drug discovery. Biologics such as antibodies, nanobodies and peptides have become important GPCR-directed tools, but designing biologics that modulate GPCR signalling has remained an outstanding challenge because these receptors are integral membrane proteins and conformationally dynamic. 

Generating GPCR agonists in particular has required extensive antibody and receptor engineering, and existing computational design methods struggled with the flexible, recessed binding pockets typical of GPCRs. The researchers aimed to develop de novo design methods together with high-throughput screening to generate GPCR-binding miniproteins with high affinity, potency and selectivity. They designed agonists for receptors involved in itch and pain, as well as antagonists for receptors implicated in cancer, metabolic disorders such as diabetes and obesity, and migraines.  To engage deeply recessed binding pockets, the team implemented a motif-guided RFdiffusion approach alongside MetaGen, which uses structurally diverse scaffolds from the AlphaFold-predicted structural metaproteome. These scaffolds contain protruding elements that facilitate engagement with recessed epitopes. Following backbone design, ProteinMPNN was used for sequence design. Designs were screened using receptor diversion (RD), a purification-free microscopy-based assay performed in human cells, deployed at scale via optical pooled screening (OPS-RD) which enables screening of up to 100,000 designs. Although its maximum capacity is lower than that of display technologies, OPS-RD enables screening against receptors in their native membrane environment without requiring receptor solubilization. Results were benchmarked against yeast-surface display and surface plasmon resonance (SPR), five receptor-bound designs were structurally characterized by cryo-electron microscopy (cryo-EM), and a designed chemokine receptor antagonist was tested in vivo in human CD46 transgenic mice.

The OPS-RD binding signal was highly predictive of SPR-validated hits, with an area under the curve (AUC) of 0.92. Agonists against MRGPRX1 reached EC50 values as low as 427 nM after optimization, and additional agonists were identified against NK1R. The CXCR4 antagonist dCX1_001 blocked CXCL12-mediated signalling with a pA2 value of 7.60 ± 0.3, corresponding to 25 nM. Its cryo-EM structure, determined at a resolution of 3.28 Å, closely matched the designed fold, with a Cα RMSD of 0.6 Å across the binder. Miniprotein antagonists targeting GLP1R, GIPR, PTH1R and CGRPR exhibited picomolar-to-nanomolar potencies, while binding miniproteins were also identified against GCGR. In mice, dCX1_001 mobilized haematopoietic stem and progenitor cells (HSPCs) comparably to the approved drug AMD3100 (plerixafor), while inducing significantly lower circulating leukocyte levels and no adverse haematological or inflammatory cytokine effects.

The authors report that these results establish de novo design as a viable strategy for engineering GPCR-targeting miniproteins across 11 receptor targets. However, the study focused on peptide- and protein-binding GPCRs with relatively open and shallow orthosteric binding pockets, rather than the more challenging receptors that naturally bind small-molecule ligands. OPS-RD enables screening of difficult targets directly in human cells without requiring soluble receptor preparation.  They caution that targeting active or inactive receptor structures alone is not sufficient to guarantee intended agonism or antagonism due to GPCR signalling complexity, noting some designs may act as allosteric modulators, and suggest future work could extend this framework to GPCRs with small-molecule endogenous ligands.

 

 

Author: Melissa Öner

Editor: Elinsu Ak 

 

 

 

Reference: Muratspahić, E., Feldman, D., Kim, D. E., Qu, X., Bratovianu, A.-M., Rivera-Sánchez, P., Voss, J. H., Hertz, E. P. T., Jeppesen, M., Dimitri, F., Sakamoto, K., Nallathambi, A., Peceli, P., Cao, J., Cary, B. P., Belousoff, M. J., Keov, P., Trinh, P. N. H., Chen, Q., . . . Baker, D. (2026). De novo design of miniproteins targeting GPCRs. Nature, 656, 1044–1053. https://doi.org/10.1038/s41586-026-10656-8  

Share This

Share

Share this post for the scientific community