Watching for SARS-CoV2 hypermutability

One of my perennial concerns in this pandemic has been that the SARS-CoV2 virus would evolve a hypermutability phenotype, mutate rapidly, and become much more difficult to control. I’m going to briefly cover 1) why this would be bad 2) why this would be possible and, in a rare bit of optimism 3) why I no longer consider this very likely.


Right now, mutation in SARS-CoV2 is the rate-limiting step in adaptation. Most new variants are lost (~98%) and almost all of the remaining ones are not beneficial. So SARS-CoV2 adaptation happens in steps: a long period of waiting, followed by the emergence of an adaptive mutation which sweeps to fixation. Appearance of new adaptive mutations depends on the intrinsic mutation rate of the virus, the population size of new infections available to mutate, and the likelihood that a new adaptive mutation is lost before its lineage can grow large.

We have seen in the past year that adaptive changes become more rapid as the population grows (i.e., Fisherian acceleration). We had ~6 months before the first adaptation, D614G, but multiple variants under positive selection occur in the last few months. The intrinsic mutation rate of SARS-CoV2 has been relatively static, with a few exceptions: 1-2 new mutations every month.

The exception is the red and orange dots there (the new 501Y variants of concern), which seem to have had a large jump in mutations right before emerging, then settled down to the normal rate.

As long as most time is spent waiting for the emergence of a new adaptive mutation, increasing the mutation rate would increase the rate of adaptive mutation. In the context of an emerging pandemic, rate of adaptation is bad, and we want to keep it as low as possible. Richard Lenski’s experiments in adaptive evolution are worth looking into. He has, for 35 years, been growing flasks of bacteria and watching them evolve over time. In three of those lines, he saw the emergence of a hypermutable phenotype: the strain suddenly greatly increased its mutation rate enormously, and that strain then took over the environment.

It looks as though hypermutability isn’t intrinsically more adaptive to the bacteria; its emergence involved breaking error-correcting enzymes that are there for a reason. Instead, a strain with hypermutability will occasionally by chance develop a beneficial mutation soon enough after that that beneficial mutation will carry it to fixation. And precisely because they are hypermutable, they are much more likely to catch a beneficial mutation by chance than a random lineage. Once fixed, it will then evolve significantly faster. In a local optimum, hypermutability would be bad for the bacterium; new beneficial mutations would be absent and you would see an increase in genetic load with no offsetting benefits. Far from genetic optima, in a new environment, hypermutability would allow much more rapid adaptation to the new environment.

Breaking things

SARS-CoV2, like MERS and SARS-CoV, is an RNA virus with a naturally high mutation rate. RNA is much less stable than DNA, and RNA viruses normally have a fairly high mutation rate. Because these coronaviruses are so large, they have developed a mechanism to reduce the mutation rate. One of the proteins they produce through cleavage of the Orf1ab polyprotein is an exonuclease (nsp14) that proofreads the RNA polymerase that copies the virus.

Previous work with the original SARS-CoV virus showed that knocking out the nsp14 exonuclease with a pair of mutations in the enzymatic active site increased mutation rates 10-20 fold, causing a hypermutable phenotype. The obvious concern is that a single random mutation in the SARS-CoV2 exonuclease would do the same thing, causing much more rapid adaptive selection.

SARS-CoV2 exonuclease is essential

It turns out, this experiment has been done by a Dutch group already researching the very similar MERS virus, and applying the same techniques to SARS-CoV2. They found that simple knockouts of exonuclease in both MERS and SARS-CoV2 were lethal rather than causing a hypermutability phenotype as in SARS-CoV1. They suspect that the exonuclease has been so thoroughly integrated into the replication complex that replication is no longer possible with the enzyme broken.

In short:

  • Hypermutability would be bad, causing much faster adaptive evolution of COVID
  • There’s a proofreading enzyme that, when broken in the SARS1 virus, causes hypermutability
  • That enzyme is present in the SARS2 virus, but is so tightly integrated that breaking it kills the virus
  • So hypermutability phenotypes are probably fairly hard to produce in SARS2
  • We can expect COVID to continue to evolve at about the same rate

Join the Conversation

1 Comment

Leave a comment

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Create your website with
Get started
%d bloggers like this: