Diagnosing some diseases could be as easy as breathing into a tube. MIT engineers have developed a test to detect disease-related compounds in a patient's breath. The new test could provide a faster way to diagnose pneumonia and other lung conditions. Rather than sit for a chest X-ray or wait hours for a lab result, a patient may one day take a breath test and get a diagnosis within minutes.
The new breath test is a portable, chip-scale sensor that traps and detects synthetic compounds, or "biomarkers," of disease, which are initially attached to inhalable nanoparticles. The biomarkers serve as tiny tags that can only be unlocked and detached from the nanoparticle by a very particular key, such as a disease-related enzyme.
The idea is that a person would first breathe in the nanoparticles, similar to inhaling asthma medicine. If the person is healthy, the nanoparticles would eventually circulate out of the body intact. If a disease such as pneumonia is present, however, enzymes produced as a result of the infection would snip off the nanoparticles' biomarkers. These untethered biomarkers would be exhaled and measured, confirming the presence of the disease.
Until now, detecting such exhaled biomarkers required laboratory-grade instruments that are not available in most doctor's offices. The MIT team has now shown they can detect exhaled biomarkers of pneumonia at extremely low concentrations using the new portable, chip-scale breath test, which they've dubbed "PlasmoSniff."
They plan to incorporate the new sensor into a handheld instrument that could be used in clinical or at-home settings to quickly diagnose pneumonia and other diseases.
"In practice, we envision that a patient would inhale nanoparticles and, within about 10 minutes, exhale a synthetic biomarker that reports on lung status," says Aditya Garg, a postdoc in MIT's Department of Mechanical Engineering. "Our new PlasmoSniff technology would enable detection of these exhaled biomarkers within minutes at the point of care."
Garg is the first author of a study that details the team's new sensor design. The study appears online in the journal Nano Letters . MIT co-authors include Marissa Morales, Aashini Shah, Daniel Kim, Ming Lei, Jia Dong, Seleem Badawy, Sahil Patel, Sangeeta Bhatia, and Loza Tadesse.
Tailored tags
PlasmoSniff is a project led by Loza Tadesse, an assistant professor of mechanical engineering at MIT. Tadesse's group builds diagnostic devices that can be used directly in doctor's office and other point-of-care settings. Her work specializes in spectroscopy, using light to identify key fingerprints in a chemical or molecule.
Several years ago, Tadesse teamed up with Sangeeta Bhatia, the John and Dorothy Wilson Professor of Health Sciences and Technology and of Electrical Engineering and Computer Science at MIT. Bhatia's group focuses in part on developing nanoparticle sensors - tiny particles that can be tagged with a synthetic biomarker. Bhatia can tailor these biomarkers to cleave from their nanoparticle only in the presence of specific "protease" enzymes that are produced by certain diseases.
In work that was reported in 2020 , Bhatia's group demonstrated they could detect cleaved biomarkers of pneumonia from the breath of infected mice. The biomarkers were exhaled at extremely low concentrations, of about 10 parts per billion. Nevertheless, the researchers were able to detect the compounds using mass spectrometry - a technology that is highly sensitive but requires bulky and expensive instrumentation that is not widely available in clinical settings.
"We thought, 'How can we achieve that same sensitivity, in a way that's accessible, at the point of need, and in a chip format that can be scalable in terms of cost?'" Tadesse says.
A fingerprint trap
For their new study, Tadesse's group looked to design a sensitive, portable breath test to quickly detect Bhatia's biomarkers. Their new design centers on "plasmonics" - the study and manipulation of light and how it interacts with matter at the nanoscale.
The researchers noted that molecules exhibit characteristic vibrational modes, corresponding to the motions of atoms within their chemical bonds. These vibrations can be detected using Raman spectroscopy, an optical technique in which molecules are illuminated with light. A small fraction of the scattered light shifts in energy due to interactions with a molecule's vibrations. By measuring these energy shifts, researchers can identify molecules based on their distinctive vibrational fingerprints.
To detect Bhatia's biomarkers, however, they would need to isolate the comparatively few molecules from the dense cloud of many other exhaled molecules. They would also need to boost the biomarker's vibrational signal, as the Raman-scattered light by an individual molecule is inherently extremely small.
"This is a needle-in-a-haystack problem," Tadesse says. "Our method detects that needle that would otherwise be embedded in the noise."
The team's new sensor is designed to trap target biomarkers and boost their vibrational signal. The core of the sensor is made from a thin gold film, above which the researchers suspended a layer of gold nanoparticles. The gold nanoparticles are coated with a porous silica shell, generating a 5-nanometer-wide gap between the gold nanoparticles and the gold film. The silica is modified to strongly bond with molecules of water. The hydrogen in water can in turn bond with the target biomarkers. If any biomarkers pass through the sensor's gap, they stick to the water molecules like Velcro.
The sensor's gap is engineered to strongly amplify light due to plasmonic resonance, where electrons in the nearby gold structures collectively oscillate in response to incoming light, concentrating the electromagnetic field into the gap. Biomarkers trapped in these gaps experience a greatly enhanced electromagnetic field, which amplifies their Raman scattering signal. The researchers can then measure the Raman scattered light, and compare the pattern to the biomarker's known "fingerprint," to confirm its presence.
The team worked with Daniel Kim, a graduate student in Bhatia's lab, and tested the sensor's performance on samples of lung fluid that they obtained from healthy mice. They spiked these samples with biomarkers of pneumonia that Bhatia's group previously designed. They then placed the spiked fluid in a vial and heated it to evaporate the fluid, to simulate exhaled breath. They placed the new sensor on the underside of the vial's cap and used a Raman spectrometer to measure the scattered light as the fluid vapor passed through the sensor.
Through these experiments, they showed the sensor quickly detected biomarkers of pneumonia at extremely low, clinically relevant concentrations.
"Our next goal is to have a breath collection system, like a mask you can breathe into," Garg says. "A patient would first use something like an asthma inhaler to inhale the nanoparticles. They could then breathe through the mask sensor for five minutes. We could then integrate a handheld Raman spectrometer to detect whatever biomarker is breathed out, within minutes."
Breath tests for disease, sometimes referred to as disease breathalyzers, are an emerging technology. Most designs are still in the experimental stage, and take different approaches to detect various conditions such as certain cancers, intestinal infections, and viruses such as Covid-19. The MIT team notes that its design can be used to detect diseases beyond pneumonia, as well as biomarkers that are not related to disease, as long as the biomarker of interest has a known vibrational "fingerprint."
"It's not just limited to these biomarkers or even diagnostic applications," Tadesse says. "It can sniff out industrial chemicals or airborne pollutants as well. If a molecule can form hydrogen bonds with water, we can use its vibrational fingerprint to detect it. It's a pretty universal platform."
This work was supported, in part, by funding from Open Philanthropy (now Coefficient Giving). Several characterization and fabrication steps were conducted at MIT.nano.
