G protein-coupled receptors, or GPCRs, are the body’s most essential communication network. They sit in the cell membrane, acting as molecular antennae that detect everything from light and smells to hormones, adrenaline, insulin, and medicines. They regulate nearly every physiological process in the human body.
Given this central role, it’s no surprise that about a third of all FDA-approved drugs target GPCRs. Yet despite their importance, precisely controlling these receptors, turning them on or off with surgical accuracy, has remained one of drug discovery’s most persistent challenges.
Now, a breakthrough study published in Nature demonstrates that artificial intelligence can finally meet this challenge. Researchers at the University of Washington Medicine Institute for Protein Design and the biotech company Skape Bio have used AI to create custom-designed “miniproteins” that can both activate and block GPCRs with remarkable precision.
This achievement opens the door to new treatments for diseases that currently lack effective medicines, from metabolic disorders to cancer and neurological conditions.
What Are GPCRs and Why Are They So Hard to Target?
GPCRs are transmembrane proteins, meaning they span the cell membrane, with a portion inside the cell and a portion outside. When a signaling molecule binds to the extracellular portion, the receptor changes shape, triggering a cascade of biochemical events inside the cell.
The challenge lies in the receptor’s structure. The signaling switch of GPCRs sits in deep, flexible pockets that change shape depending on whether the receptor is active or inactive. This conformational flexibility makes them especially difficult to target with conventional drugs.
Traditional small molecules can bind to GPCRs but often lack selectivity, affecting related family members and causing off-target side effects. Antibodies, on the other hand, are generally too large to access the deep binding pockets where GPCR signaling occurs. As a result, generating antibodies that act as GPCR agonists, molecules that activate the receptor, has proven nearly impossible.
“The deeper problem is that conventional screening methods weren’t built for GPCRs,” explains Dr. Christoffer Norn, co-founder and CEO of Skape Bio. “GPCRs are highly dynamic, membrane-dependent proteins whose conformational equilibria are strongly influenced by their lipid environment.”
The AI Solution: Designing Proteins from Scratch
The research team, led by Nobel laureate Dr. David Baker, director of the Institute for Protein Design, took a fundamentally different approach. Rather than screening existing molecules, they used AI to design entirely new proteins that had never existed in nature.
“Protein design takes our understanding of how proteins fold and reverses it, asking if we can envision, with the aid of AI computing, a new protein that sticks to a target in a purpose-built way,” Baker explains.
The team developed specialized design strategies to build miniproteins, compact proteins of fewer than 100 amino acids, that could slip into the hard-to-access pockets of GPCRs. By targeting specific active or inactive receptor states, they designed molecules that could either activate or block signaling with high affinity, potency, and selectivity.
This represents a crucial advance. Designing agonists, molecules that activate receptors, is considerably more complicated than designing antagonists, which simply block them. Agonists must promote a specific structural change that activates the receptor, effectively stabilizing its active form.
“GPCRs are very dynamic molecules, sometimes they assume the active state even though there’s no molecule there,” Norn explains. “So when you then come in with your miniprotein that is made just to fit the active state, you can lock that into place.”
A New Screening Platform: Testing in Living Cells
Alongside the protein design breakthrough, the researchers also invented a new high-throughput screening system that can test tens of thousands of designed proteins directly in living human cells.
Traditional screening for GPCR-targeting drugs is notoriously difficult because many methods require purifying or stabilizing the receptors, steps that can distort their natural signaling behavior and fail to capture native structural dynamics. By keeping receptors in their native membrane environment, the new “receptor diversion” screening system preserves biological relevance while dramatically accelerating discovery.
The platform, known as OPS-RD, can screen up to 100,000 miniprotein designs per target per campaign. Each cell expresses one design alongside the target receptor; binding causes the receptor to be diverted from its normal trafficking pattern, generating a detectable signal. In situ sequencing then reads a DNA barcode inside each cell, linking each signal back to its specific design across millions of cells.
For Dr. Edin Muratspahić, a postdoctoral research scholar at the Institute for Protein Design and first author of the study, the moment of validation was profound.
“Existing drugs such as antibodies bind to but often fail to activate or block GPCR signaling,” he says. “Seeing computationally designed miniproteins not only bind but actually control GPCR signaling in living cells was a defining moment for me.”
Validating the Approach: Structure, Function, and In Vivo Success
The team didn’t stop at design and screening. They validated their miniproteins across multiple levels of biological complexity.
Structural studies using cryo-electron microscopy (cryo-EM) showed that five designed miniproteins closely matched their computational models, confirming the accuracy of the AI design pipeline.
The team successfully generated functional agonists and antagonists across 11 GPCR targets, including receptors involved in pain, metabolism, migraine, itch, and cancer. This diversity demonstrates that the approach is generalizable rather than limited to a single receptor type.
Perhaps most importantly, the researchers validated their approach in living organisms. In one mouse study, a designed chemokine-receptor antagonist mobilized hematopoietic stem and progenitor cells at levels comparable to a clinically used drug, but with fewer side effects.
“This paper showcases how we can do this repeatedly for different GPCRs in ways that capitalize on their dynamic motion to either activate or inactivate them,” Baker says. “The result is a generalized approach to targeting biologically critical receptors.”
Why Miniproteins Are a New Therapeutic Modality
Miniproteins occupy a unique space in the therapeutic landscape, positioned between small molecules and antibodies.
“Miniproteins are a distinct therapeutic modality that sit between small molecules and antibodies,” Norn explains. “They are typically 40-70 amino acids long, completely de novo designed, and do not rely on a naturally occurring protein scaffold.”
This positioning offers several advantages:
Accessibility: Their small size allows them to reach binding pockets that are often inaccessible to antibodies. In the GPCR programs, some designs penetrate deep into receptor binding pockets below the membrane surface.
Selectivity: Unlike small molecules, which often bind related family members, miniproteins can be designed for high selectivity, reducing off-target effects.
Designability: “We can design these proteins from scratch for a specific biological function,” Norn notes. “Rather than starting with a naturally occurring molecule and optimizing it, we can design a miniprotein to stabilize a receptor in a desired state and then engineer properties such as stability, half-life, and manufacturability around that function.”
Tunable Pharmacokinetics: The properties of miniproteins can be tuned for different therapeutic contexts. An unmodified miniprotein is cleared rapidly from circulation, while an Fc-fused version extends exposure by more than 200-fold.
From Discovery to Therapeutics: The Skape Bio Platform
The study was a collaboration between the Institute for Protein Design and Skape Bio, a Copenhagen-based company co-founded by Norn and Baker to translate these discoveries into therapies.
Skape Bio’s platform unifies GPCR-tailored AI design, a proprietary native-receptor screening system in human cells, protein production, and pharmacology into a single integrated system. The company’s goal is to create drug candidates for metabolic, inflammatory, and neurologic diseases by targeting GPCRs that have remained largely inaccessible to conventional drug discovery.
“The methods we are sharing in this new study form the roadmap for achieving all-computational design of protein ligands for any GPCR,” Norn says.
The company is already exploring a pipeline of GPCR targets across multiple disease areas, with a careful balance of agonist and antagonist mechanisms. According to Norn, the goal is to demonstrate that miniproteins can consistently deliver precise control of receptor function across different biological contexts, not just in one program.
What Remains to Be Done?
While the breakthrough is significant, challenges remain. The initial study focused on GPCRs with protein ligands and relatively accessible binding pockets. GPCRs with more occluded or smaller binding pockets remain challenging.
Additionally, the field is still early relative to antibodies or peptide therapeutics. Over the next five years, miniproteins will need to move from early validation into the clinic and begin to establish themselves as a complementary modality alongside existing approaches.
Nevertheless, the study represents a fundamental shift in how drugs can be discovered and designed.
“Many GPCRs remain underexplored because we do not have ligands or tools to explore their function,” Muratspahić notes. “With these computational methods, now we can explore other GPCRs in more detail.”
Ztec100 Conclusion
The AI-driven design of miniproteins capable of precisely controlling GPCR signaling represents a major milestone in drug discovery. By combining advances in computational protein design with a high-throughput screening platform that preserves the native membrane environment of receptors, researchers have created a generalizable approach to targeting one of medicine’s most important yet challenging receptor families.
“This signals a shift toward the rational design of biologics-based therapies for GPCRs, a target class historically dominated by small molecules but underrepresented by biologics,” Norn explains.
For patients with diseases that have long lacked effective treatments, this breakthrough offers new hope and a glimpse of a future where medicines are not discovered by chance but designed with purpose, one protein at a time.
Reference: Muratspahić E, Feldman D, Kim DE, et al. De novo design of miniproteins targeting GPCRs. Nature. 2026. DOI: 10.1038/s41586-026-10656-8
by JOHN DEVIN