The molecule adenosine triphosphate (ATP) is the basic unit of biochemical energy that fuels the activities of all cells. Now, a team led by researchers at Weill Cornell Medicine and the Howard Hughes Medical Institute (HHMI) Janelia Research Campus has developed and tested a high-resolution sensor for tracking the real-time dynamics of ATP levels in cells and within subcellular compartments.

The new tool represents a major advance over prior ATP sensor technology, and the researchers expect it to accelerate many areas of biological research.

The researchers, who published their work on May 15 in the Proceedings of the National Academy of Sciences, developed the sensor by modifying an ATP-binding bacterial enzyme and combining it with GFP, a natural fluorescent protein. Because the sensor is made of proteins, it can be encoded in DNA and produced within targeted cells. The researchers demonstrated the utility of their new ATP sensor in several experiments, including one that illuminated, in unprecedented detail, the dynamics of ATP use at individual synapses on a single nerve fiber.

“This sensor will enable us to view basic processes of biology at much higher-resolution,” said study co-corresponding author Timothy Ryan, M.S. ’86, Ph.D. ’89, tri-institutional professor of biochemistry at Weill Cornell Medicine.

The other co-corresponding author is Jonathan Marvin ’95, a senior scientist at HHMI/Janelia.

In recent decades, advances in molecular biology and fluorescent imaging technology have enabled the development of various experimental ATP sensors. However, none has had the attributes needed for widespread adoption. Instead, ATP monitoring is still done mostly with sensors that detect proxy measures in cells such as changes in oxygen levels. Such sensors have their own major limitations, including the fact that they can measure ATP use only in bulk groupings of cells.

The new sensor, iATPSnFR2, is a next-generation version of a direct ATP sensor that Marvin and colleagues developed several years ago. When it binds to ATP, its shape changes in a way that enhances the fluorescence of the GFP. This increase in fluorescence intensity with ATP binding is much larger in the new version, giving the sensor high sensitivity across a very wide range of ATP concentrations.

The sensor also can incorporate a second fluorescent molecule, with a different emission wavelength, that doesn’t change intensity with ATP binding; this provides a reference signal for the concentration of the sensor molecules, enabling a more accurate readout of ATP levels and their dynamics.

In demonstration experiments, the team showed that the sensor can be used to track ATP levels not just in single cells but in distinct subcellular compartments, and on time scales of just hundreds of milliseconds. In one set of experiments, they used the sensor to track ATP levels within individual mitochondria, the tiny biochemical reactors that burn oxygen to efficiently make ATP for cells. They also looked at the dynamics of ATP levels at individual synapses on a nerve fiber, finding that each synapse has its own distinct pattern of ATP consumption and depletion.

“We did these experiments mainly to showcase the basic utility of the method, so we didn’t dig very deeply,” said Ryan, who is also a professor of biochemistry in anesthesiology at Weill Cornell Medicine. “But since then, we’ve been busy using it to dig deeper, and it has turned out to be very powerful for us as a tool for making new scientific discoveries.”

Jim Schnabel is a freelance writer for Weill Cornell Medicine.