When Lauren Cowley arrived in Guinea in June 2015, the country was still trying to contain the record-breaking Ebola epidemic that had begun in early 2014. With several new cases cropping up every week, epidemiologists had to work frantically to track the skeins of the virus as they threaded from one person to another.
Cowley was there to help. By sequencing Ebola genomes from newly diagnosed patients, she could help her colleagues chart the source of new infections. Ebola mutates at a fairly constant rate, so if two people have identical strains, it's likely that one infected the other, or at least that they are nearby links on the same transmission chain. By tying fresh cases to existing clusters, Cowley could help local health workers to nail down the routes through which the virus was spreading—and develop effective strategies for stopping it.
Speed was essential and, until recently, out of the question. Even last year, the only way to carry out such work was to send samples of Ebola's genetic material to specialist labs with expensive sequencing machines, which took weeks or months to spit out the results. But by the time Cowley arrived, she could do that work herself in an ersatz desktop laboratory within 48 hours, thanks to a revolutionary sequencer called the MinION.
Unlike rival sequencers, which are as big as microwaves or fridges, the MinION is the size of a chocolate bar. Cowley had three, and she could clutch them all in a single fist. These devices quite literally bring the power of modern genomics to the palm of your hand. And at a cost of just $1,000, they herald a new era where sequencing moves away from well-equipped institutions and into places where it is most needed, from hospitals to epidemic-afflicted hot zones. Rather than sending samples from outbreak sites to special labs, scientists like Cowley will be able to take the labs to the outbreaks.
The MinION uses a technique called nanopore sequencing, which involves a donut-shaped protein whose hole is just a billionth of a meter wide—a nanopore. When the pore is unblocked, ions can flow freely through it, creating a measurable electric current. But if something gets in the way—say, a strand of DNA—that current collapses. The four bases of DNA—A, C, G, and T—each change the current through the nanopore in different ways. So, as a DNA strand threads its way through the pore, the rising and falling current reveals its sequence.
Nick Loman, a genomicist from the University of Birmingham who works on infectious diseases, was one of the hundreds who signed up. Last June, his team used the MinION to study a suspected Salmonella outbreak, which had spread through the local Heartlands Hospital. Within two hours of receiving samples from the hospital, the team had completely sequenced the bacterium, confirmed that it was indeed Salmonella, determined its strain, and showed that the various cases were all part of the same cluster. Loman was so impressed that he began looking for a bigger challenge.
By then, West Africa's Ebola outbreak had become a serious emergency. Deaths were mounting and the virus had spread from Guinea to neighboring Sierra Leone and Liberia. Observing from the UK, Loman was shocked to learn that despite thousands of cases and months of intensive control efforts, researchers had only sequenced a hundred or so Ebola genomes. “I thought: Why aren't we generating more genome data?” he says. “That should almost be a basic part of surveillance now.”
So, in April 2015, Loman's student Joshua Quick travelled to Guinea with three MinIONs that the team called Ribz, Chicken, and Brisket, for reasons best known only to them. Quick also packed three laptops, some chemical reagents, a miniature centrifuge, and a small machine for amplifying genetic material. He fitted this fully-functioning sequencing lab into just two suitcases, which he unpacked onto two small desks in a pop-up diagnostic laboratory, run by the European Mobile Laboratory Project. Over the next 12 days, he sequenced Ebola genomes from 14 patients. When Quick flew home, he left the porta-lab behind. His colleagues, including Sophie Duraffour, Lauren Cowley, and Raymond Koundouno (pictured above), continued the work. Together, the team have so far sequenced around 130 Ebola genomes.
“Ebola sequencing in the field is an amazing feat,” says Mark Akeson from the University of California, Santa Cruz, who is one of the pioneers of nanopore sequencing. “None of us thought the technology could have advanced this rapidly.” Loman's team have also confirmed that the MinIONs are now 90 percent accurate—a significant improvement over their performance at launch. Loman adds that this statistic underestimates the machines. Each nanopore sequences one strand of DNA at a time, so while each readout may have a 10 percent error rate, these mistakes average out once you combine the results from thousands of reads.
In fact, the team found that the sequencing was the easiest part of the work. The difficult bit was... well, everything else, according to Cowley, who spoke about the experience at the Genome Science 2015 conference. When she took collections of new samples, which had often travelled for hours over road-less terrain, she had to run them through several chemical reactions to prepare them for sequencing. Simple enough, except that insects would repeatedly land on her face and, worse, in her reagents.
The electricity supply to the lab would also shut off erratically so Cowley frequently had to work by the gleam of a head torch; the MinIONs were fine since they can run off a laptop via a USB connection. The heat was stifling. The humidity loosened the tiny magnets on one of her test-tube racks, which would launch themselves across the room at anything metal. “I thought I was more likely to die from flying magnets than from Ebola,” she recalls. And perhaps the biggest challenge was finding an internet connection fast and stable enough to send sequences to the UK for Loman to analyze. (He also uploaded the data to a public site called Nextflu so that other scientists could view Ebola’s evolution in real-time.)
These sequences helped the team and their colleagues to work out how each patient became sick, and how best to react. For example, if they treated a man who fell ill after burying his mother, and found that both carry genetically identical viruses, they could reasonably deduce that he became infected through contact with her—case closed. If they learned about a woman who became sick even though all her friends and family were healthy, they could look for other places in Guinea that had genetically similar viruses; perhaps that might pinpoint a source like, say a particular food vendor. And when those sources were found, health workers could set up quarantine centers, or monitor people for symptoms.
“I saw, first-hand, epidemiologists being able to accurately track transmission routes in real time and then intercept the chain to prevent further transmission of the virus,” says Cowley.
At the peak of the outbreak, there were too many transmission chains to keep track of, but that number has fallen. As of this month, only one chain remains in Guinea—a single persistent dynasty of Ebola that keeps rearing its head. “I think it’ll be not too long before we go down to zero,” says Cowley, who plans to carry on her sequencing work until the outbreak is completely stopped. “We need zero cases in [Guinea, Sierra Leone, and Liberia] for 42 days, before we can breathe easy.”
To Loman, the success of the portable MinION lab is a taste of things to come. “When it's a doddle to have sequencing anywhere you are, whether it's an a sewage works, or a doctor's surgery, or a hospital ... and when you can detect links between patients, or between patients and the environment ... and when you can get other information about pathogens like antibiotic resistance ... if you can do all that in one cheap, available assay, that's clearly going to be the future,” he says.
“The MinION has proven that you can squeeze genome sequencing down to this incredibly tiny size,” he adds. “If we can continue to miniaturize, you can have genuinely handheld diagnostic devices, or even sensors like in a water plant—a biological Internet of things.”