Researchers at the Max Delbrück Center and Freie Universität Berlin have revealed how a protein complex guards the entrance to channels in mitochondrial membranes called cristae. The work, published in “ Nature Communications ,” provides a blueprint for modeling human proteins in action.
Twenty-four hours a day, seven days a week, mitochondria in our cells churn out energy from the food we eat, which keeps our bodies functioning. Much of this work happens along cristae — deep pockets in the mitochondria’s inner membrane. To work properly, cristae must maintain their shape, while carefully controlling which molecules enter and exit. At their narrow entrances sits a protein complex called MICOS, which helps stabilize these pockets and acts as a molecular filter. When this architecture is disrupted, the consequences can include neurodegenerative diseases and cancer. But exactly how MICOS works has remained unclear.
A team led by Evangelia Nathanail, a doctoral student in the Structural Biology of Membrane-Associated Processes lab of Dr. Oliver Daumke at the Max Delbrück Center, and Edoardo Rolando, also a doctoral student in the lab of Dr. Cecilia Clementi at Freie Universität Berlin, has now modeled the human Mic60-Mic19 subcomplex, a central part of MICOS. Their simulations, which were published in “ Nature Communications, ” show how its flexible structure spans the cristae entrance, allowing smaller molecules to pass through while blocking larger proteins.
“Our study not only reveals the molecular architecture of an essential cellular machine, but it also shows how one can model highly dynamic protein complexes that escape traditional structural biology methods,” says Daumke, co-senior author of the study.
MICOS is made up of multiple copies of several proteins. Its largest component, the Mic60-Mic19 subcomplex contains a long, disordered region — a flexible stretch with no fixed shape. This flexibility is one reason it has been difficult to image or model.
To build their model, the team combined several approaches. They first captured the structure of a section of Mic60 found only in animals using X-ray crystallography. They combined this structure with fungal structures and AI predictions to build a virtual model of the human version. The model, however, had a big weakness: it was static, so it did not show the subcomplex’s flexibility. With help from the Clementi Lab, the team then simulated computationally the Mic60-Mic19 subcomplex’s movements.
To test how well their model replicated the actual human subcomplex, they compared it to structural data from human mitochondria supplied by the lab of Dr. Fan Liu at the Leibniz-Forschungsinstitut für Molekulare Pharmakologie (FMP). They found 97% correspondence, suggesting that their model closely reflects the structure of the actual human subcomplex.
Finally, the team tested the subcomplex’s gatekeeping function by adding spheres to simulate the passing of different sized proteins. They found that the subcomplex blocked spheres with a radius larger than two nanometers. Its disordered regions were key to this function: “With one static structure, we might not have seen exactly how it swats all of those spheres away,” says Nathanail. “That's exactly why we needed an approach to see it in action.”
“Computational methods often produce impressive images and videos of molecular machines in action, but they are difficult to prove experimentally,” adds Daumke. “Our study now provides a way to do this.”
Mitochondria are essential in all cells, but tissues that need a lot of energy, such as muscle and the brain, are especially affected when they fail. The researchers found that a known mutation in patients with optic nerve damage and a developmental brain disorder alters the MICOS complex's core. Their model offers a possible explanation for how the mutation causes disease.
Confirming this will require observing the entire complex at work inside mitochondria. So far, the team has modeled only part of it. “We are now looking into the structure of the machinery inside cells, and how it changes during disease and aging,” says Daumke.
The Max Delbrück Center for Molecular Medicine in the Helmholtz Association lays the foundation for the medicine of tomorrow through today’s discoveries. At locations in Berlin-Buch, Berlin-Mitte, Heidelberg, and Mannheim, interdisciplinary teams investigate the complexity of disease at the systems level – from molecules and cells to organs and entire organisms. Together with academic, clinical, and industry partners, and as part of global networks, we turn biological insights into innovations for early detection, personalized therapies, and disease prevention. Founded in 1992, the Max Delbrück Center is home to a vibrant, international research community of around 1,800 people from over 70 countries. We are 90 percent funded by the German federal government and 10 percent by the state of Berlin.
Nature Communications
Experimental study
Not applicable
Integrative structural modelling reveals the human Mic60-Mic19 subcomplex as a diffusion barrier in mitochondria
26-Sep-2026