IN EARLY DECEMBER, cases of COVID-19 soared in Kent, England—and scientists wanted to know why. For clues, Nick Loman, who is part of the COVID-19 Genomics Consortium U.K., and his colleagues examined how the coronavirus was mutating. By looking at this zoo of slightly different viruses, they could roughly track the outbreak’s spread through the community.
For SARS-CoV-2, these mutations—the small errors made naturally when genomes are copied—develop at a steady pace of one or two each month, says Loman, a professor of microbial genomics and bioinformatics at the University of Birmingham. Yet among the Kent cases, scientists found a large cluster that was remarkably different, with a total of 23 mutations arising without prior notice and faster than anyone expected.
“That’s how many mutations you have to go back to get to anything we’d seen before,” he says. “That’s a very striking and unusual finding.”
This discovery is part of what led British officials to sound the alarm last week. A follow-up investigation by Public Health England showed that the variant, known as B.1.1.7 or 501Y.V1, began to thrive at a time when cases were spiking in Kent and other parts of southeastern England. Retroactive tracing through a database of samples tied B.1.1.7 to patients as early as September 20. But by mid-November, the variant made up between 20 and 30 percent of cases in London and a region east of the city. Three weeks later, it was roughly 60 percent. And on December 23, U.K. scientists announced a separate SARS-CoV-2 variant reported last week in South Africa had now been spotted in two people in England.
Scientists are still uncertain about how the cluster of mutations arose or what they mean long-term for the virus’s transmission. One possible hypothesis for their origin involves chronically ill patients treated with experimental therapies like convalescent plasma donated by recovered COVID-19 patients. In such lengthy illnesses, the virus has more opportunities to replicate, increasing the odds for mutations. The consistent use of the therapies, meanwhile, may put more pressure on the germ to evolve.
“Some of these people who are chronically infected have some quite big shifts in the virus,” says Ravindra Gupta, a virologist at the University of Cambridge. “Some are immune-suppressed. Some of them have had convalescent plasma. Some of them have had [the antiviral] remdesivir.”
If this suspected origin story does prove to be the case, it could have implications for treatment, says Muge Cevik, a clinical lecturer in infectious diseases at the University of St. Andrews. Earlier in the pandemic, the best path for helping patients was unclear. That led hospitals to give patients a buffet of therapies, with the hope that some combination might work. But if new-wave medicines like antivirals and antibody therapy contributed to the development of viral variants, it will be “a reminder for all the medical community that we need to use these treatment options carefully.”
The many mutations
While mutations edit genetic code, they don’t always lead to outward changes in a germ or organism. That’s why these newfound variants have garnered so much concern. It’s as if the virus entered a dressing room and came out with a new outfit, rather than the normal circumstances where it would only change its hat.
Of the 23 mutations in the United Kingdom’s variant, 17 are at positions in the genome that alter the building blocks that make up the virus’s proteins, as described in a recent COVID-19 Genomics Consortium report from Loman and his colleagues. The consortium stated such a large shift is so far “unprecedented” for the COVID-19 pandemic. Eight of those changes lie in the region that encodes for the spike protein—the key that SARS-CoV-2 uses to enter cells.
While there is no direct evidence that this collection of mutations influences the severity of disease, modeling and prior laboratory work hints at the possibility that it could make the virus more contagious. A greater abundance of cases could mean more hospitalizations and deaths.
For example, lab experiments suggest that one of the observed deletions that eliminates two building blocks in the spike—dubbed H69 and V70—may double the viral infectiousness, according to a recent preprint. Other research hints that another mutation—N501Y—increases the spike protein’s binding capability. This one also independently arose in the South African variant (501Y.V2), which was first detected in October. But more work is necessary to determine if and how these changes might translate to differences in human transmission, Cevik says.
On December 18, the U.K.’s advisory panel for emerging respiratory outbreaks—which includes Cevik and some members of the COVID-19 Genomics Consortium—released a preliminary assessment of the variant. Their modeling hints that the variant might account for up to 70 percent greater transmission, but this takeaway is uncertain, Cevik cautions. Some of the spread could originate from human behavior, for instance more group gathering under loosened restrictions. More time, more research, and the country’s latest round of stay-at-home orders may yield clarity on the matter.
“Models are informative but not concrete,” Cevik says. She noted that the answer will likely come from a combination of evidence from a variety of fields—epidemiology, virology, genomics, as well as modeling. “Everything together tells us a story.”
Therapeutic sledge hammers
Where the story of B.1.1.7 begins is also a tantalizing mystery. Many scientists are pointing to the possibility of extra opportunities for viral evolution in people with compromised immune systems. These patients tend to suffer from chronic infections, during which the coronavirus can linger for weeks or even months.
Such a situation presents additional chances to replicate and accrue random mutations. One case report of a 45-year-old immunocompromised man, who was infected for nearly five months before succumbing, documented “accelerated viral evolution.” Most of the mutations occurred in the spike protein, including changes present in both variants under scrutiny in the U.K. and South Africa.
“The virus has just got a chance to stretch its legs a little bit,” Loman explains. After a year of closely tracking these mutations, scientists know that most don’t do anything noteworthy. Some are even harmful to the virus’s ability to multiply. For example, one mutation widely touted this year—D614G—increases coronavirus replication and infectivity while also making the germ more vulnerable to neutralization by antibodies.
But pressure from partially effective therapies for chronic patients could be part of the evolutionary push that allows some mutations beneficial for the virus to thrive. The idea is similar to HIV patients developing resistances to treatment after taking incomplete drug courses, says Gupta, who has spent a decade studying HIV resistance. “If you take a sledgehammer to a nut you can always crack it,” he says. But scientists don’t yet have a sledgehammer-like therapy for chronic COVID-19 patients.
For example, convalescent plasma has large differences in potency between doses thanks to natural variation between the array of antibodies produced by the immune systems of the donors, he says.
A few case studies hint at this kind of rapid evolution for chronically ill SARS-CoV-2 patients. In a recent preprint article, Gupta and his colleagues document mutation of the virus after a patient received three treatments of convalescent plasma starting at day 63 of their illness. Two of the viral mutations developed in genes that code for the spike protein. Something similar happened inside a 65-year-old cancer patient who survived after 105 days with the virus. And one of the mutations recently spotted in the South African variant—N439K—may allow the virus to bypass monoclonal antibody drugs, according to a preprint released by the COVID-19 Genomics Consortium in November.
Similar rapid viral evolution has also occurred in patients with influenza, says Emma Hodcroft, co-developer of Nextstrain, a global repository that tracks pathogen evolution in real time. Still, she cautions that there could be other origin stories. “There’s almost never just one way to do something in biology,” she says.
For example, a 2017 mathematical model of the stomach-churning norovirus predicted that immunocompromised individuals are too rare to generate variants that spread widely. Alternatively, uncontrolled spread may just offer enough chances for the virus to mutate in regular people with short bouts with the disease, says Mark Tanaka, a mathematical and computational biologist at the University of New South Wales Sydney, who is an author of the study. Both Pfizer and Moderna are checking if the variants can escape the antibodies generated by their respective vaccines, but pioneering scientists behind these drugs believe this scenario to be unlikely.
Hodcroft stresses that the discovery of the U.K. variant emphasizes the importance of genomic sequencing in COVID-19 surveillance. Out of the U.K.’s 2.1 million cases, the COVID-19 Genomics Consortium has sequenced 137,000 SARS-CoV-2 genomes total, which is about half of all those sequenced worldwide. Contrast that with the United States, which has sequenced only about 51,000 of its 18 million cases.
Such enhanced genomic testing explains why scientists in the U.K. were able to act so fast and warn others as B.1.1.7 slowly started to trickle into Denmark, The Netherlands, Italy, Belgium, Hong Kong, and Australia—and how they just discovered two cases of the variant linked to South Africa.
“If we can try to put the effort in right now to try to identify who these people might be, get them tested, get them into quarantine, it might be that we can actually stamp [it] out,” says Hodcroft, adding that there is no time to waste. “We might be able to do that with 100 people, it’ll be much much harder if we had thousands.”-National Geographic