Three COVID-19 infections diagnosed in Washington in October were caused by virus with a mutation that might boost the respiratory bug’s ability to dodge immune defenses.

The mutation, called E484K, is also present in two of the worrisome new viral variants spreading around the globe — those that originated in South Africa and Brazil. But the virus detected in Washington did not have any of the other mutations that characterize those variants, said researchers at the UW Medicine Virology Lab.

No other infections with the mutation have been detected since October, though surveillance is limited in the state.

“Based on what we have right now, it hasn’t taken off,” said computational biologist Pavitra Roychoudhury, part of a team that sequenced the three genomes. “We definitely want to keep an eye on it.”

The mutation has been spotted sporadically in the U.S. since spring, said Trevor Bedford, a computational biologist at the Fred Hutchinson Cancer Research Center who has been tracking genetic changes in the virus since the start of the pandemic.

Those isolated sightings haven’t sparked major outbreaks. “It appears that just having the (E484K) mutation isn’t enough to make a huge difference to the virus,” he wrote in an email. However, in combination with the 10 or more other mutations in the South Africa and Brazil variants, it is spreading rapidly.


The mutation is located at a specific site on what’s called the “receptor binding domain” — the part of the viral spike protein that latches on to human cells.

A research team at The Hutch recently identified that site as the most concerning, because changes there seem to make the virus more difficult for some people’s antibodies to neutralize.

However, laboratory experiments showed that while the mutations might dampen immunity, they don’t wipe it out, said Fred Hutch virologist Jesse Bloom, who collaborated with graduate student Allison Greaney and others on the work.

Genetic changes in the virus are raising alarms because several new variants have emerged recently around the world, some of which spread more easily. None of the new variants have yet been detected in Washington, though the variant now dominant in the U.K. has popped up in several other states.

One of the biggest worries is that the virus will morph to the point where it can evade the body’s immune response, either natural or vaccine-induced. At least one person in Brazil who had recovered from COVID-19 was reinfected by the variant with the E484K mutation.

It’s probably inevitable that the virus will someday outwit vaccines, Bloom said. But that will take several years, giving pharmaceutical companies time to tweak their formulas, as they do every year for the flu vaccine.


“It’s not something that the average person needs to get really worried about every time they hear about another of these mutations,” he said. “It’s not going to cause catastrophic failure of immunity.”

But more contagious forms of the virus, like the U.K. variant, are likely to have major impacts in terms of surging numbers of infection and deaths.

To understand how viral evolution can affect immunity, Bloom and his team examined all 4,000 possible mutations in the receptor binding domain. In lab experiments, they tested each mutation to see how it affected the virus’ ability to bind to human cells. They also exposed the mutated spike proteins to blood serum from 11 people who had been infected by the virus and recovered.

In most cases, the mix of antibodies in the patients’ blood was able to attack and neutralize the mutants. The main exception was E484K and related mutations, which reduced the potency of some patients’ antibodies by a factor of 10. But other patients’ sera had no trouble fighting the mutant form, Bloom said.

The results were posted recently but have not yet been peer-reviewed or published in a scientific journal. Bloom and his team are now working to evaluate the mutations’ impact on vaccine-induced immunity.

The three Washington cases with the E484K mutation were discovered during routine genome sequencing as part of a surveillance system for new variants, Roychoudhury explained. The sequences were submitted to a repository called GISAID, where a researcher flagged the mutations.


The team at the UW Virology Medicine Lab works with randomly selected leftover specimens from positive coronavirus tests. All of the identifying information is stripped away, so they don’t know anything about the cases, including where they occurred, how they may have been related, or how the patients fared, Roychoudhury said.

The absence of related cases suggests the infections may have been dead ends, said Alex Greninger, assistant director of the virology lab.

“These viruses are not omnipotent,” he said. “If someone has a variant virus, but they’re not interacting with other people, it’s not going to transmit.”