Before testing his theory, Kobayashi's team added a targeting agent to the fluorescent compound, to help it home in on cancer cells. The researchers chose to attach the cancer drug Herceptin, an antibody that binds to HER2 receptors found on certain early-stage breast-cancer cells. The idea is that once the compound binds to HER2 receptors, the cancer cell absorbs the compound, which then comes in contact with the acidic lysosome, which activates the fluorescent agent.

The team injected the compound into mice with HER2-positive tumors, and used an endoscope to capture images of lit areas. Researchers found that the pH-sensitive probe detected tumors with 99 percent accuracy, lighting up just the tumors, compared with an "always on" fluorescent probe, which only detected 85 percent of the tumors, and also lit up nontumor tissue in the background.

Kobayashi found that a pH-sensitive probe could have other advantages. Cancer cells that are weakened or dead cannot maintain cellular processes, so the lysosomes often become less acidic. Therefore, a probe sensitive to pH levels may be able to tell a live cell from a dead one.

The researchers took biopsies from rats treated with the pH-sensitive probe and the "always on" probe, then washed the tissues with alcohol, killing the cells. In the aftermath, they found that tissue with the pH-sensitive probe had gone dark, but they found little change with the "always on" probe.

"Once the cell is dead, the fluorescence is gone," says Kobayashi. "So we can potentially monitor cancer therapy in real time."

Kobayashi acknowledges that the system has its limitations. The pH-sensitive probe is based on green light, which doesn't penetrate very deep into the body, meaning it cannot be used for noninvasive imaging. A more practical application may be imaging during surgery, when an endoscope could come in close contact with tissue. For example, the fluorescent probe could be injected into a patient presurgery. After a surgeon has taken out large tumors, an endoscope could be threaded through the area to illuminate residual cancer cells.

In the future, Kobayashi says, the pH-sensitive probe could be tailored to target other cancers, by attaching agents that bind to molecules that signal specific tumor types. He believes that such a targeted imaging strategy could be used in humans within the next five years.

Scott Hilderbrand, a researcher at Massachusetts General Hospital's Center for Molecular Imaging Research, is also developing fluorescent probes using molecular targeting strategies. He says that in the future, Kobayashi's team may want to consider working with probes in the near-infrared spectrum, which are able to illuminate deeper into the body.

"This specific system may have limited application, especially in regards to taking this forward into the clinic," says Hilderbrand. "But the concept and approach are definitely very relevant towards the area of enhanced tumor detection at the molecular level."