How did omicron mutate, and how could it evade antibodies? Fred Hutch scientist Trevor Bedford answers our questions

Little is known about omicron, the newest variant of the coronavirus that was first detected in southern Africa and has rapidly spread to other countries. But researchers across the globe are beginning to pin down theories about how the variant developed and how its mutations could threaten to prolong the pandemic.

The United States this week recorded its first known case of omicron in California, and Washington public health officials have said they wouldn’t be surprised if the variant has already landed in the state.

Trevor Bedford, a computational biologist at Seattle-based Fred Hutchinson Cancer Research Center, has been studying the coronavirus with his team for nearly two years, and became well-known early in the pandemic for detecting the outbreak in the Seattle area, the country’s first hot spot.

He’s long been analyzing outbreaks and developing surveillance networks, eventually helping launch Nextstrain, an open-source project that looks at pathogen genome data.

On Wednesday, Bedford answered our most pressing questions about omicron.

This conversation has been edited for clarity and brevity.

What are your initial impressions about omicron?

The two things to emphasize are: One, if you look at the genome, there’s a whole bunch of mutations in spike protein — 30-ish mutations in omicron, as opposed to eight or 10 in alpha, beta, gamma, delta, etc.

You worry that that’s enough differences there to make it so existing antibodies won’t work well against omicron and you’d worry just from that, that there’d be a fair risk of people who have been previously infected or vaccinated would be able to be susceptible to it.

The other side is that if you just look at what’s happening in South Africa, particularly in Gauteng, that you see what looks like rapid spread of this variant, faster than delta when it emerged. By looking particularly at case counts, you can see those were nicely leveled off after the delta wave completed, and now they’re shooting up.

So we have something that looks kind of threatening from an immune-escape perspective just by looking at the genome, and we see it looks like it’s also spreading rapidly.

Why is a variant’s high number of mutations significant?

The antibodies we have from getting vaccinated and being infected were developed by the immune system to bind to the original virus that’s existed in the past almost two years. And if you have 30 mutations in spike protein, it’s kind of all over the protein, and it changes its shape enough that your existing antibodies may very well not bind well to it.

That will make it so that you’ll still have your T-cell immunity that was engendered by vaccine or infection and it’s not like going back to square one, but it’s likely a significant drop in immunity.

It’s pretty equivalent to 10 years of flu evolution happening all at once. It’s a significant jump in just the number of mutations that exists on the protein.

How did that happen so quickly? Any theories on how omicron might have developed?

This is mostly an open question. People have different hypotheses. My favorite hypothesis is that this virus was incubating in an immunocompromised individual who had a chronic infection of perhaps a year or more, and during that time they were not able to clear the infection, but they had enough antibodies to drive evolution of the virus.

We’ve seen that in a handful of longitudinally sampled chronic infections that we know about. We don’t have a smoking gun if that were the case, but it fits with other pieces of evidence we know about. And if you look at the rest of the genome, there are mutations you’d expect for a year of evolution, but it’s not nearly as striking. And you can kind of see in the spike protein that there’s this large excess there.

What does omicron’s spread mean for the future of the pandemic?

Eight days ago, I was expecting that endemicity for SARS-CoV-2 would look a lot like influenza, where in the evolution there you have one mutation on top of another mutation on top of another mutation. You have a number of strains that are circulating that are a little different from each other and the one that happens to grab more or better mutations out-competes the others.

It’s a very continuous process, whereas if you can have this entirely different mechanism by which long-term chronic infections go off into left field and produce these wildly different viruses, that can create an overall dynamic that’s quite different from influenza.

So I don’t know if omicron is a one-off thing and we just got super unlucky with it, or if we’ll see another thing just as different (mutationally) appear in six months or a year’s time. And if it is a general feature, it would suggest that a pretty substantial fraction or contribution of SARS-CoV-2 evolution would be due to these kind of wild jumps that might be occurring.

How was the development of other variants different?

Alpha is still thought to have emerged from a chronic infection, mainly because it has some of these interesting signals — particularly mutational signals — that are seen in chronic infections. You can also find alpha’s common ancestor in the U.K. and then you find alpha in the U.K. and there’s a big gap.

The U.K. had amazing genomic surveillance in fall 2020, and so the best explanation for that gap is that it was evolving in one person during that time. But that was maybe a gap of three months, rather than a gap of over a year, like we see with omicron.

Delta is very different in that it had been more like this flu-like thing I was mentioning, where if you look from mid-2020, it’s been the continual evolution of mutations that seem to have made delta more and more transmissible.

Are future similar variants inevitable? What can be done?

Maybe to some degree inevitable. But if that is the source, we have ways to fight it. One would be if this is an actual connection with HIV in kind of reducing individuals’ immunity, you could be better at distributing HIV antivirals in much of southern Africa. We could be better at — as they’re coming to market — using SARS-CoV-2 antivirals to eliminate chronic infections. We could be better at making sure people that are immunocompromised are as vaccinated as possible to keep them from getting infected.

Basically anything that would reduce infections, and in particular, reduce infections or clear infections in immunocompromised individuals would behoove us.

What are some of the things you’re eager to learn about?

This was entirely expected. The main thing that I’m waiting for and will determine a lot of how bad this will be is if we look at neutralization titers (a measurement of immune response) — how much do they drop in people who are doubly vaccinated with the mRNA vaccines, how much they drop in people who have three doses, and if it’s possible that three doses will provide some decent protection from especially severe outcomes.

We have a range of scenarios going from a large epidemic that’s quite severe if it’s escaped immunity a lot, to a large epidemic that’s not severe at all because existing immunity might not prevent infection but does prevent severe disease. And I think that’s probably the main range I’d be thinking about.

When will we know more?

Part of that immune question will be resolved in about two weeks as we get the first neutralization assay results in.