At EPFL's Laboratory of Therapeutic Proteins and Peptides in Lausanne, Christian Heinis and his team have solved a puzzle that has long frustrated drug developers: how to slip molecules through cell membranes to block the protein interactions that drive disease. The breakthrough centers on membrane-permeable cyclic peptides—tiny, ring-shaped chains of amino acids that can sneak past the cell's protective barrier and disrupt disease-causing interactions from inside.
The challenge they tackled is profound. Many diseases are driven by proteins interacting with each other inside cells, but conventional drugs struggle to block these interactions. Small-molecule drugs are often too small to grip the broad, flat surfaces where proteins meet. Peptides, by contrast, can cover larger surfaces, and cyclic peptides—whose ends are chemically linked into a ring—are especially promising because their compact structure lets them bind tightly to difficult targets. But there's a catch: most peptides cannot cross cell membranes, which means many peptide-based drugs must be injected because they simply cannot navigate the biological barriers that surround cells.
The EPFL team took a different approach. Rather than starting from known binding patterns, they developed a way to discover membrane-permeable cyclic peptides from scratch. Using high-throughput synthesis methods, lead researcher Xinjian Ji synthesized and screened a library of 15,360 fully random cyclic peptides, all designed to be small, compact, and relatively nonpolar—properties associated with the ability to slip through cell membranes. As Heinis puts it, the work is "a bit like searching for the needle in a haystack."
The researchers focused on a specific protein interaction: the connection between Keap1 and Nrf2, which is linked to inflammation, oxidative stress, neurodegeneration, and cancer. Through repeated design, synthesis, and testing cycles, they refined the most promising candidates and solved X-ray crystal structures showing how their synthetic peptides bound differently from the natural Nrf2 protein. The result was peptide 30—a compound that balanced strong target binding with the ability to cross cell membranes. The peptide weighed just 890.6 daltons, small enough for membrane permeability yet large enough to bind and inhibit the challenging interaction. Remarkably, compared with the natural Nrf2 sequence, peptide 30 had no electrical charge, far fewer hydrogen bond donors, and a much lower polar surface area—features that allowed it to slip through membranes while maintaining its grip on the target.
When tested in living cells, peptide 30 successfully inhibited the Keap1–Nrf2 interaction in a dose-dependent manner, proving the method works not just in test tubes but in actual cellular environments. The significance of this achievement extends far beyond this single interaction. The approach opens access to intracellular targets previously considered difficult to drug, potentially broadening the range of diseases that can be treated with peptide-based therapies. It could also pave the way for orally available peptide drugs—a long-standing goal since molecules taken by mouth must cross the membranes of intestinal cells before reaching the bloodstream.
Heinis's group has patented the method and founded the spin-off company Orbis Medicines. The team is now advancing the technology to synthesize and screen even larger libraries of membrane-permeable cyclic peptides, with their sights set on some of the most challenging cancer targets: KRAS, b-catenin, and c-Myc. For patients facing diseases driven by protein interactions long considered beyond the reach of drugs, this breakthrough represents a genuine shift in what medicine can reach.