Imagine inhaling a sensor that could monitor lung disease patients’ response to therapy, emitting a signal when they breathe out. Like a breathalyzer that recognizes alcohol, such a device could sniff out compounds released only by specific illnesses to gauge how well treatment is working.
Biomedical engineers at the Massachusetts Institute of Technology have developed a synthetic biosensor using specialized nanoparticles to detect and then report the presence of molecules indicating bacterial pneumonia or the genetic disease alpha-1 antitrypsin deficiency. Leslie Chan, a senior postdoctoral fellow in Sangeeta Bhatia’s lab at MIT who is now headed for a faculty appointment at Georgia Tech, led a team that tested this device in mice, demonstrating a proof of principle they hope to bring to people.
The Bhatia lab has already engineered biosensors to test how well medicines are treating a patient’s condition, gleaning signals in urine. Bhatia is also a co-founder of Glympse Bio, a startup developing nanoparticlebased
biosensors to be injected into the blood as an alternative to traditional biopsies. These biosensors carry nanoparticles coated with peptides that are not naturally produced by the body; they are released when they are cut by proteins called proteases linked to a variety of diseases.
Chan and Bhatia talked with STAT about their paper published this week in Nature Nanotechnology and about the future they envision for this technology. This interview has been condensed and edited for clarity.
Why did you focus on the breath as a way to diagnose and monitor disease?
Chan: We were interested in breath because it’s been underutilized clinically, especially when you compare it to blood and urine. Currently in the breath research field, there’s a lot of interest in mining breath to identify volatile compounds that would make great breath biomarkers. Within breath, there are hundreds of thousands of these trace
volatile compounds and there are a lot of clinical trials to look for these breath biomarkers. There are some challenges with going the route of breath biomarker identification.
In our approach, instead of relying on naturally occurring breath volatiles, we wanted to be able to engineer the breath signal that we could use to monitor lung disease.
What were the challenges?
Bhatia: One of the things about the biomarker field in general is you’re basically sampling the body to see what you can find out. In blood tests, that’s whatever is circulating in the blood. You look at your cholesterol levels or your PSA levels for prostate cancer and so on. And the problem with all of these naturally occurring biomarkers is that they vary a lot. They vary with age, they vary with disease states, they vary with genetics.
Obviously, the volatiles in your breath change with what you eat. And so you’re looking for a needle in a haystack. The technology that Leslie developed is really unique because she picked a volatile that does not naturally occur. You inhale the sensor and what comes out is something that is unique to the exact protein that you were looking for, and there is zero background.
You zero in on protease activity. Why?
Chan: We have over 500 different proteases — proteins that exist naturally in our body. Their function is to cleave other proteins and help regulate and maintain homeostasis in the body. When there’s dysregulation of these proteases, that oftentimes can lead to disease. We can see dysregulated activity as a hallmark of disease, so the idea behind these nanosensors is they’re able to take advantage of this aberrant activity from these proteases in order to generate a noninvasive readout for disease.
Previously in the lab, we had developed these nanosensors that we could administer into the body where the sensors would interact with the active proteases. And then during this interaction, there’s some cleavage event that leads to the release of a reporter. And so for our existing diagnostic platform, we had this reporter release that would result in a urinary readout.
For this newer work, we were interested in seeing if we could modify our existing platform. Instead of having a reporter that is released and cleared in the urine, have it released and be exhaled and measured in breath. It’s easier to collect in point-of-care diagnostics, and that could potentially speed up testing time.
How did you get from your idea to your solution?
Chan: We teamed up with our collaborators at MIT Lincoln Laboratories. They’re a really interesting group because they specialize in chemical detection.
Bhatia: In a demo that they showed us, they can even detect caffeine molecules in the airspace above coffee — decaf or not. They work on bomb detection and train dogs to sniff bombs using their machine to calibrate levels, so they had a super-sensitive detector.
How did you make it work for your experiments with mice?
Chan: The key was being able to engineer the chemistry in such a way that once the protease has cleaved a protein, it releases the gaseous molecule that we use as our reporter. And so the challenge was in being able to attach our reporter and have it be released back into its original form so that it would become a gas again to be detected.
Bhatia: An analogy might be thinking about, like, if you have a balloon you have to let it go without a weight on it. You cut that rope for it to be able to fly away. And so what Leslie engineered was basically like the rope connection to the weight so when it was cut by the protease it would fly away.
Bhatia: Imagine if you could have a smart mirror in your bathroom mirror, where you breathe on the mirror and it would work like a car breathalyzer. We’re working on that.