An Oral History of DNA Barcoding

It’s January 2013, and the UK and Ireland are both gripped by the same scandal. News has broken that several food products sold in both countries (all frozen burgers or ready meals) contain horse meat. All products claimed to contain beef, but analysis carried out by the Food Safety Authority of Ireland found horse DNA in over one-third of the beefburger samples and pig DNA in 85% of the samples. Follow-up tests in the UK revealed that some of the “beef” in frozen lasagne products made by the French manufacturer Comigel consisted of up to 100% horse.

Health concerns are raised. The very integrity of the European food supply chain is called into question. By February, the EU has intervened, announcing a three-month coordinated control plan of DNA testing of processed meat across the union. In May, the first arrest of a meat wholesaler is made.

Over 10 years later, the 2013 horsemeat scandal is considered one of the most egregious cases of food fraud on the continent, and it was all uncovered by one simple technique: DNA barcoding.

Yet this technique wasn’t developed for probing the accuracy of frozen food labels, but for mapping biodiversity.

Fittingly, though, it was first envisioned in a grocery store.

Source code

“I was walking down an aisle and looking at the short strings of numbers that we use, retail barcodes, to tell all the items in a grocery store apart, and I thought, ‘My God, a small string of numbers like this, why don’t we view DNA as a small string of numbers? And why wouldn’t it tell species apart?’” Dr. Paul Hebert recalled.

“And the answer was it seemed like it should, because every third position in a DNA strand can’t be held constant by selection because of the degeneracy of the genetic code, which means that every third position should be varying and, in principle, should vary independently in different species,” the professor of biodiversity genomics at the University of Guelph told Technology Networks.  

Inspired by some early RNA sequencing work on microbial communities in the 1990s, Hebert tested his idea by sequencing the DNA of insects caught in his own backyard.

“I turned on some ultraviolet lights in my backyard and collected 200 species of one particular group of insects and was actually surprised; every one of them could be distinguished,” he said.

Instead of RNA, Hebert relied on a tiny chunk of mitochondrial DNA called cytochrome C oxidase subunit one (CO1) to tell his insects apart. The method worked so effectively he wondered whether it could be applied to all life.

“And I thought, ‘Oh, OK, if it works in my backyard then surely we can do this on a global level and tell apart all the species on the planet.’”

But not everyone in the world of taxonomy was as confident in the method’s universality.

“A lot of people were pretty certain that that tiny pilot study was insufficient,” he said. “And I will agree that it was a bit of bravado, but the Moore Foundation in California provided me with three million dollars to prove that barcoding wouldn’t work. And, actually, we proved that it did work. People sent us lots of specimens. And we found it was highly effective in many different taxonomic groups of animals, and then it got extended into plants and fungi and protists.”

A barcode for every species 

Ever since that first moth experiment, Hebert and his team have been adding more and more species to their growing catalogue. In 2010 they founded a consortium called the International Barcode of Life with the aim of recording the CO1 DNA signatures of half a million species.  

“The Linnaean Enterprise has been putting names on things for 267 years, and it has about 1.7 million animal species today. And we wanted to barcode half a million animal species in five years, and we did it.”

“That was good,” Hebert said, “but it wasn’t inexpensive; it cost about $120 million. And if you really wanted to complete the inventory of life on the planet, it was going to cost many billions of dollars.”

To achieve his dream of cataloguing all of life, Hebert realized he’d need to make a few costs savings.

“We sat back, spent three years developing new methods to slash the costs for analysis, and then in 2019 launched our second project, BIOSCAN, which is $180 million project, and it will raise the registration count up to at least 2 million species,” he told Technology Networks. “So, we’ve collapsed the cost for analysis in that project down from about $20–30 to $1 today with the help of some new technology out of the UK, from Oxford Nanopore Technologies.

Armed with this more affordable technology, Hebert is confident he and his team at BIOSCAN should be able to sequence every species sample they receive and personally collect during international trips.

“I will be in Peru in one month setting up a large program in the western Amazon,” Hebert disclosed. “In the summer, I’ll be up in the high Canadian Arctic. In between, I’ll be in some other cool places.”

“But this last week we got two giant shipping crates, filled up with samples from Madagascar,” he added. “This week we got samples from Ghana. Next week, we’re getting samples from Namibia and Guinea and Cote d’Ivoire. So, yeah, we have people collecting around the world.”

With this level of species turnover, Hebert says the BIOSCAN project should hit its target of barcoding every species on the planet by 2045.

“We will have exceeded the Linnaean species count and this long quest to register every species on the planet will be done.”

From phylum to food fraud

So, Hebert and his team have a lot of species to process between now and 2045. But in between expeditions to Peru and coordinating sample deliveries from partners around the world, Hebert says he still gets made aware of the increasing use of his pioneered barcoding method in the field of food fraud.

Because the 2013 horsemeat scandal was only the beginning. Since then, a growing number of research groups have used Hebert’s CO1 DNA method to uncover more and more deceptive dishes. Deer tendon products were found to be water buffalo tendon. Fish dishes were found to contain meat from endangered sharks. “Halal” food was found out to be anything but.

Even food products that don’t contain animal or plant products, but are derived from animals and plant, like honey, can now be tested with DNA barcoding methods.

“We have identified DNA barcodes which are short sequences of DNA indicative of a plant species,” Dr. Maria Anastasiadi, a lecturer in bioinformatics at Cranfield University and pioneer of honey adulteration testing, told Technology Networks last year. “Then we perform amplification of this sequence using qPCR. When the presence of these barcodes exceeds certain thresholds, the honey is suspicious for adulteration.”

“This is a very sensitive method able to detect adulteration with sugar syrups even at 1% level,” Anastasiadi emphasized.

What does Hebert make of his method’s second life as a food fraud finder?

“I think it’s great,” he said. “We were not aiming to address food substitution problems, but it happened very quickly, with fish in particular. Of course, it even happened in Europe with some of your horse meat substitution for beef.”

Does he collaborate much with such research efforts? Not as much as he used to, he says. He’s just happy to see them flourish.

“We jumped in on some of the early fish papers, the substitution papers there; we jumped in on the honey substitution. We’ve done projects like that, but we tend to jump in and then jump out and see if other people are interested and let them pick up with it and run,” he said.

“There’s a lot of competent people now using DNA for tracking food adulteration and monitoring, and we let them do that for the most part. We’re just happy to see the method applied because it’s always cool when basic science helps society in unexpected ways.”