Researchers at the University of California, Santa Barbara have developed a new protein-based sensor that could allow MRI machines to detect molecular activity inside cells. This innovation, detailed in a recent article in Science Advances, could help scientists study diseases such as cancer and neurodegeneration by providing insights at the molecular level.
MRI technology has been used since the 1970s to produce images of internal body structures without exposing patients to ionizing radiation. However, its ability to capture changes has been limited to anatomy rather than the underlying molecular processes. Arnab Mukherjee, associate professor of chemical engineering at UC Santa Barbara’s Robert Mehrabian College of Engineering, explained: “You can see the structures of your tissues — whether it’s the brain, the heart, the kidneys or the stomach — but you don’t get molecular information. So, the only time you can know that something is going wrong or something has changed is if you take another MRI, and the structure and morphology of the tissue changes.” He added that for many conditions, structural changes occur only after significant disease progression.
Mukherjee has worked on this problem since his postdoctoral research at Caltech. He said: “If we can see these molecular-level changes happening in real time, then we can ask questions like, ‘How do tumor cells metastasize?’ or ‘How does neurodegeneration progress at the molecular level as an animal ages?’ There’s currently no way to do that.”
The new sensor uses synthetic biology concepts and can be genetically engineered into cells so that MRIs can visualize specific molecular processes. The modular design allows researchers to swap out different proteins depending on what process they want to observe. The team describes this approach as having a LEGO-like architecture.
MRI scans work by aligning hydrogen atoms with a magnetic field and using radio waves to create images based on their response. While powerful for anatomical imaging, this method does not naturally provide information about molecules inside cells. Since the 1960s, scientists have used fluorescent proteins from jellyfish genes under microscopes to watch biological processes unfold in real time; however, similar capability did not exist for MRI until now.
Mukherjee’s research led him to focus on aquaporin—a protein channel in cell membranes that regulates water movement—as a potential basis for an MRI-detectable signal within cells. “Our water molecules are tiny, tiny magnets,” Mukherjee said. “If you can control or affect the rate at which water molecules move back and forth across the cell, you can make that magnetic signal specific to certain types of cells or biological processes.”
After joining UCSB in 2017, Mukherjee’s group began combining aquaporin with other proteins to create genetic circuits tailored for various cellular targets. Asish Ninan Chacko (Ph.D., ’26), who contributed as a graduate student in Mukherjee’s lab, noted: “This protein can be regulated using a lot of chemical signals… We can even replace this particular protease with another type of protease and use it to detect many different processes.”
Their system—called MAPPER (modular aquaporin-based protease-activatable probes for enhanced reporting)—enables tracking multiple chemical processes within laboratory animals using MRI technology. Chacko stated: “That’s a first in this paper because so far in scientific literature you’ve seen only four or five genetic sensors each used to detect a unique analyte… In this paper we describe close to ten systems we can detect with this one setup.”
The researchers believe MAPPER could reduce reliance on animal sacrifice during experiments by allowing continuous imaging over time within living subjects rather than taking single snapshots post-mortem. Chacko explained: “Our approach allows continuous imaging of the same animal over the course of a study giving a far more accurate picture of disease and biology.”
Because MAPPER is modular—allowing custom combinations tailored for specific needs—it eliminates redesigning sensors from scratch when monitoring new targets; instead components are swapped into place quickly. Mukherjee envisions creating training programs so others can use these tools effectively: “We want to take these sensors and put them in the hands of people who will actually use them… whether that’s neuroscientists… or developmental biologists.”
