Honeybees, credited with the best olfactory abilities, are now trained to sniff lung cancer. These little insects can even detect the most minute of the disease's scent from a patient's breath, providing new promise for early cancer detection. The scientists used the bees' incredible sense of smell in an experiment where their brains were hooked up to electrodes, all sorts of scents were passed under their antennae, and brain signals were recorded.
"The response patterns, or the neural fingerprint to different odours, are so clearly distinguishable, somewhat like day and night—whether the bee's antennae are responding to a chemical or not," says Debajit Saha, a neural engineer at Michigan State University. Different smells produced different brain activity patterns—neural fingerprints if you will. The group published these results on June 4th in the journal Biosensors and Bioelectronics. The hope is that one day, honeybees might be used in clinics as living sensors for early disease detection. While there are already some electronic noses and other mechanical devices that can sense odours, they are nowhere near as good as nature's versions. "Biology can differentiate between very, very similar mixtures, which no other engineered sensors can do," Saha explains.
The scent is everything regarding the communication of many insect species—it is their language.
The idea of animals being able to smell out diseases is not particularly new. Doctors reported that a border collie and a Doberman sniffed out their owner's melanoma back in 1989. Lately, however, scientists have shown that dogs can detect COVID-19 by smelling people's sweat. Many insects probably have similar disease-detecting abilities. For example, ants can be trained to detect the smell of cancer cells grown in a lab dish. The abilities of bees, however, had not been quite in evidence until today.
By directly accessing the bees' neurons, scientists can bypass the need for behavioural training. Instead of weeks training a dog to signal when something smells suspicious, researchers can immediately obtain their answer from the bee's brain.
Then, following this delicate brain surgery of workers, where wires are attached to the bees' odour-sensitive regions, bees are restrained in 3-D printed plastic harnesses with a bit of wax. In place, a device delivers puffs of air to the bees' antennae, similar to a salesperson spritzing perfume.
Such puffs may have been a mix of odours made to be characteristic of healthy people's breath or that of lung cancer patients. To these odours, bees' brains responded differently, and it was easy for scientists to correctly point to which of the two sets of the breath samples was what in at least 93% of cases. In another experiment, air above cultured lung cells was collected. Bees were trained to recognize the difference in this air collected above healthy cells and that above cancerous cells, which also included small cell lung cancer and non-small lung cancer.
His lab has also discovered that bees can sense the presence of other trace odours, such as those given off by notoriously hard-to-detect PFAS (perfluoroalkyl and poly-fluoroalkyl substances) in their environment. "That blew my mind," Saha says. "PFAS in the environment is very, very hard to detect."
As the team members explained, the apian scent sensor that they came up with would be tried on the breath of live cancer patients. However, its only disadvantage is that the shelf-life of the device itself is very short: after several hours, the bees' brain health is compromised, and it responds unpredictably. In addition, it works very fast: it gives results in an instant. According to the team, a single bee brain could process samples of more than 100 individuals, meaning that this method bears the potential to improve the early detection of diseases.
References:
M. Parnas et al. Precision
detection of select human lung cancer biomarkers and cell lines using
honeybee olfactory neural circuitry as a novel gas sensor. Biosensors and Bioelectronics. Published online June 4, 2024. doi: 10.1016/j.bios.2024.116466.