Speeding up pathogenic bacteria testing

In an effort to prevent any sickness, produce such as lettuce and spinach is routinely tested for foodborne pathogenic bacteria like salmonella, listeria monocytogenes and pathogenic types of E. coli. Rapid testing of foods may occur, but it is slow in determining who is sick and from where the contaminated product originated. The current solution, often a recall, becomes damage control.

Researchers from the University of Delaware have set out to spot these bacteria before anyone consumes an affected product. As set out in an article published in the Journal of Food Safety, UD and Delaware-based startup Biospection aim to speed up testing. Faculty members Harsh Bais and Kali Kniel, alongside former graduate student Nick Johnson, teamed up with Andy Ragone of Biospection to detect foodborne pathogens in three to six hours.

“While the produce industry is working diligently to reduce risks associated with microbial contamination, tools like this have incredible potential to improve risk reduction strategies,” said Kniel, professor of microbial food safety. “Collaborations like ours between academics and biotechnology companies can enhance technology and impact food safety and public health.”

These pathogens easily find their way into plants, which are welcoming hosts. Plants use defence mechanisms to fight disease, but some human-borne pathogens have learned to push open a plant’s open-entry gates, called stomates — pores in the leaves or stem — and enter.

“Because these bacteria are not true pathogens for plants, you cannot physically see early signs that the plant is under stress,” said Bais, UD professor of plant biology. “Biospection’s technology allows us to say, very quickly, if the opportunistic human pathogen is present in the plant.”

The researchers combined their interdisciplinary expertise to create a multi-spectral imaging platform to look at plant sentinel response. A goal is to use this technique directly on a conveyor, scanning lettuce before it ever heads to the grocery store.

The technique scans leaves via multispectral imaging and deep UV sensing when the plant is attracting these pathogens. When the researchers looked at benign bacteria, they observed little change. But, with harmful, human-borne pathogens, the test can spot differences in the plant under attack.

“Using Listeria as an example, in three to six hours, we see a sharp drop of chlorophyll pigments,” Bais said. “That’s a strong signal that the plant is responding physiologically — a marker of unusual bacteria.”

The multi-spectral imaging technique is non-invasive and fast compared to current tests, where a lab scientist extracts a leaf, grinds it up, plates the bacteria and looks for disease. Biospection was awarded a National Science Foundation Small Business Innovation Research grant in 2022 to develop and commercialise its solution into a real-time imaging sensor.

“Harsh and Kali were certainly instrumental in the techniques that we developed with multi-spectral imaging and the use of deep ultraviolet fluorescence,” said Ragone, founder and chief technology officer of Biospection. “We built a portable instrument that could be commercialised.”

Vertical farming is an agricultural sector that can reap the benefits of this technology. Using less water and less space, vertical farms are a step towards more sustainable agriculture. They are, however, just as vulnerable to disease as traditional agriculture. One incidence of E. coli means a vertical farm must throw away an entire harvest.

Biospection is already working with agricultural companies to embed the imaging sensor into vertical farms’ shelves and, for outdoor farms, crop drones.

“Working with UD, we’ve laid the scientific foundation to create better instruments,” Ragone said. “We’re working toward an instrument that’s portable, automated and can give an answer in a matter of seconds.”

For future research, Bais has his eye on determining if this technology can differentiate between different microbes.

“If the sentinel response is different from one microbe to the other, that gives us the identity of the microbe based on plant sentinel response. We haven’t gone there yet, but that would be the ultimate achievement,” Bais said.

Image credit: iStock.com/wildpixel