The UW Medicine Institute for Protein Design and Skape Bio have led a new study demonstrating for the first time that artificial intelligence methods can be used to create computationally designed proteins to activate or block G protein-coupled receptors (GPCRs)
These receptors represent important but historically challenging drug targets.
The findings, published in the journal Nature, outline a generalised approach to targeting biologically critical receptors that communicate with nearly every physiological process in the body.
The challenge of targeting deep receptor pockets
GPCRs sit in the plasma membrane, which defines the boundary between the inside and outside of a living cell. They regulate a vast array of bodily functions, including the ability to see and smell, as well as the body’s sensing of adrenaline, insulin, nutrients, and medicines.
Developing molecules that can toggle GPCRs on and off in different contexts has been a long-standing challenge. The signalling switch of these receptors sits inside deep, flexible pockets, and their complex shapes make them highly difficult to target with conventional molecules. Existing drugs, such as antibodies, frequently bind to these areas but often fail to effectively activate or block the underlying GPCR signalling.
To address this, the research team developed specialised design strategies using AI computing to reverse-engineer how proteins fold. The team used these methods to design completely new miniproteins, which are proteins containing fewer than 100 amino acids. The miniature size of these molecules allows them to slip into the hard-to-access receptor pockets.
Credit
Edin Muratspahić/UW Medicine Institute for Protein Design
Testing and validation in living cells
By targeting specific active or inactive receptor states, the designed miniproteins demonstrated precise control over GPCR signalling in cell environments, turning the cellular signals on or shutting them down on demand. Subsequent structural studies confirmed that several of the manufactured proteins closely matched their computational design models. In a companion mouse study, one of the designed miniproteins performed comparably to an existing clinically used drug while demonstrating fewer side effects.
To accelerate the discovery process, the researchers also invented a native-receptor screening system that operates directly within living human cells. Traditional screening methods typically require GPCRs to be purified, stabilised, or otherwise altered, which can inadvertently change how they signal. The new system allows scientists to test tens of thousands of proteins against native GPCRs while keeping the receptors intact inside the cell membrane.
Moving innovation toward clinical applications
Skape Bio, a company founded by researchers from the Baker Lab and the BioInnovation Institute ecosystem in Copenhagen, aims to utilise this unified platform to develop drug candidates for metabolic, inflammatory, and neurologic diseases. The company’s platform integrates GPCR-tailored AI design, the human-cell screening system, protein production, and pharmacology to develop therapeutics for receptors that have remained largely inaccessible to conventional drug discovery.
The study was a collaborative effort involving multiple global institutions, including Monash University, the MRC Laboratory of Molecular Biology, Johns Hopkins University, the University of North Carolina, Novo Nordisk, Lundbeck, the University of Oregon, and the Indian Institute of Technology Kanpur. The published work reports selectivity profiling, pharmacokinetic optimisation, and in vivo data supporting the overall therapeutic potential of these designed modulators.
The post AI helps create miniprotein switches for drug targets appeared first on Open Access Government.