Skape Bio Unlocks Generalizable GPCR Drugs Using AI Protein Design

The year was 2022. Chris Norn, PhD, was wrapping up his time as a postdoctoral researcher at the Institute for Protein Design (IPD) at University of Washington (UW). AlphaFold was taking the field by storm, while a new generation of deep learning tools was rapidly advancing de novo, or from-scratch, protein design with unprecedented success rates validated at atomic resolution. Within just a few years, these AI breakthroughs, widespread applications across pharmaceuticals, nanomaterials, biosensors, and more, would help earn Norn’s mentor, David Baker, PhD, the Nobel Prize in Chemistry.  

“There’s so much dark space in biology. The precision of protein design was becoming incredible.” said Norn in an interview GEN. “Designing function from scratch is going to be incredibly impactful for treating diseases.” 

Norn’s research investigated the subtle structural differences that caused G-protein-coupled receptors (GPCRs) to change conformation from a healthy state to disease driver. These integral membrane proteins are the largest protein family encoded by the human genome and represent approximately one-third of drug targets, across cancer, metabolic disease, and neurological disorders. Yet, they are traditionally difficult to hit because their accessible regions barely protrude from the cell membrane. 

Today, Norn is co-founder and CEO of Skape Bio, a Copenhagen-based AI protein design company building a generalizable platform to target underexplored GPCRs and treat diseases once deemed undruggable. The team has published a new study in Nature demonstrating the design of functional miniproteins that target 11 GPCRs across a diversity of receptor families implicated in itch and pain, cancer, metabolic disorders, and migraine, with examples that penetrate deeply into hard-to-reach GPCR pockets. Notably, agonists were validated against three targets. 

In a key example, the study designed a chemokine receptor antagonist that mobilizes hematopoietic stem and progenitor cells in a mouse model at a level comparable to a clinically used drug, with fewer side effects.  

At the core of Skape Bio’s technology stack is a proprietary high-throughput platform that screens GPCRs directly within their native membrane environment, enabling accurate measurement of how conformational changes influence cell signaling and function. The approach represents a significant advance over traditional screening methods, which remove GPCRs from their membrane-embedded context and can fail to capture native structural dynamics. Over 100,000 miniprotein designs can be screened per target on a single-platform campaign. 

Edin Muratspahić, PhD, postdoctoral research scholar at UW and co-corresponding author of the Nature study, highlights that the rise of de novo models, such as Baker lab’s RFdiffusion, has fueled the growing momentum for protein-based GPCR drugs. Compared to small molecules, protein therapeutics offer high selectivity, protease stability, and extended half-life. Notably, the small size of miniproteins allows better tissue penetration compared to antibodies. 

“Many GPCRs remain underexplored because we didn’t have the tools to look at their pharmacology,” Muratspahić told GEN. “We’re excited to illuminate new biology beneficial to developing better and safer protein-based therapeutics.” 

The post Skape Bio Unlocks Generalizable GPCR Drugs Using AI Protein Design appeared first on GEN - Genetic Engineering and Biotechnology News.