Â鶹ÒùÔº

Cellular membrane proteins play many important roles throughout the body, including transporting substances in and out of the cell, transmitting signals, speeding up reactions and helping neighboring cells stick together. When they malfunction, it can cause serious diseases including cancer, making them attractive drug targets. But understanding how membrane proteins behave and function can be challenging because their position within the cell's lipid membrane—a tightly-packed double layer of fat-like molecules—makes them difficult to study.

Now, Scripps Research scientists have developed a new computer-driven strategy to understand how these proteins work at the atomic level. Published on October 7, 2025, in , the team designed synthetic membrane proteins that are easier to study in the lab—while also revealing the structural basis for how some maintain their shape. Scientists can use this method to design new drugs, biotechnology and therapeutics that target membrane proteins directly.

"Billions and billions of dollars a year are going into making molecules that target membrane proteins to alter their behavior and combat disease, but in order to modulate these proteins, it helps to first understand how they work," says senior author Marco Mravic, an assistant professor in the Department of Integrative Structural and Computational Biology at Scripps Research. "Our study uncovered some new rules of sequence and atomic arrangements inside membrane proteins that are essential for them to function."

Membrane proteins consist of multiple helices that are folded and packed tightly together, similar to the small, intertwined strands in a rope. To maintain their complex architecture and function correctly, different parts of the protein must bind more tightly to each other than to the lipid membrane in which they're embedded.

Mravic's team wanted to understand the role of a common pattern or "motif" that occurs in many different types of membrane proteins: a small amino acid that repeats every seven amino acids in protein chains as they traverse the cell's lipid membrane. This pattern means that these small amino acids are present in the same position on every second turn of a given helix.

They hypothesized that these motifs represent potential "sticky" spots that help membrane protein helices bind to each other and organize within their membrane folds. To understand why this motif is so conserved and how the atoms create stability, the researchers used a computer program to design what they thought to be idealized versions of the motif to study in the lab.

"It's usually really hard to study how membrane proteins behave within our bodies, because as soon as we break them out of the cell, they want to fall apart," says Mravic. "Our approach is unique in that we design new synthetic proteins from scratch with computer programs to approximate the behaviors and atomic structures of membrane proteins from nature. We can use these designer proteins as models to ask questions and clarify rules underlying many of the complex processes happening within cell membranes that we could not see or study otherwise."

First author Kiana Golden wrote a software program to identify amino acid sequences containing this motif and used this information to design optimized synthetic membrane proteins with enhanced stability. When the researchers produced these synthetic proteins in the lab, the proteins folded as predicted, supporting the hypothesis that these motifs create "sticky spots" between adjacent helices that hold the membrane proteins together in lipid.

Likewise, Golden showed that when the motifs were given the most optimal sequences, this led to synthetic proteins which were extremely stable—and even remained intact under boiling conditions.

"We found that the motif's stability was driven by an unusual type of hydrogen bond that's typically very weak, but when the motif is repeated, these weak hydrogen bonds all add up to make a very stable interaction," says Golden, who worked on the project as a UCSD undergraduate and is now a graduate student at Princeton University. "This type of hydrogen bond is rare in the natural world, so it was really surprising that this is largely what's driving this motif to occur, and that biology has evolved to use it within specific motifs and structures across nature."

Now that they have shown how this motif contributes to membrane protein structure, the researchers say that this information will help scientists and doctors identify and understand genetic mutations that could contribute to disease. Since their team proved their new software can build very strong protein complexes accurately in lipid, they are now working to design molecules to directly target membrane proteins within the cell.

"Our approach vastly accelerates what we can discover about the inner workings of membrane proteins and how to make better therapies," says Mravic.

In addition to Mravic and Golden, authors of the study, "Design principles of the common Gly-X6-Gly membrane protein building block," are Catalina Avarvarei, Charlie T. Anderson, Matthew Holcomb, Weiyi Tang, Xiaoping Dai Minghao Zhang, Colleen A. Mailie, Brittany B. Sanchez, Jason S. Chen, and Stefano Forli of Scripps Research.