We have both continuous and continual SARS-CoV-2 variant evolution happening. What do I mean? 🧵
Viruses evolve at (at least) two levels: among hosts and within hosts. Among hosts, the main determinant of viral fitness (i.e., reproductive success and continued existence of that lineage) is getting into new hosts.
In very general terms, traits like transmissibility and infectiousness are under selection at the among-host (interhost) level. When there is some degree of immunity in the population, characteristics like immune escape also become very relevant to viral fitness.
Because there are huge numbers of hosts, because SARS-CoV-2 is *not* seasonal and is circulating at fairly high (or very high) levels year round, and because it is *not* running out of evolutionary space (see thread linked in next tweet), evolution at this level is continuous.
Continuous meaning ticking along pretty much non-stop at a fairly consistent rate with fitter mutations becoming more common over viral generations due to natural selection. That's how variant trackers have managed to predict with impressive accuracy which variants will take off.
This background, continuous, ticking-along-in-somewhat-predictable-ways is what happens within lineages that are evolving among hosts. 
In addition to continuous evolution at the interhost level, we have variants that evolve within individual hosts, either as they reproduce generation after generation during a persistent infection or by recombination between two variants that infect the same host simultaneously.
The traits that evolve within hosts may be different from what is under selection among hosts. Most obviously, getting into new hosts isn't under selection within a host. Things like being able to avoid that host's immunity or being able to infect multiple tissues might be.
(For evolutionary biology nerds, you'll note that intrahost evolution doesn't involve the massive transmission bottleneck that occurs every time a new host gets infected. So, less drift that way. But then there are differences of within-host vs. among-host/metapopulation sizes.)
Most of the variants that evolve within hosts never get into (m)any other hosts -- but some do. And most of those do not and manage to establish as a lineage at the host population level -- but some do. Those lineages then start evolving at the among-host level.
We might hope that variants derived from intrahost evolution or recombination would not compete well against variants that had been evolving among hosts the whole time, but we've seen several cases where only a few additional mutations made their descendants extremely successful.
This probably includes Omicron BA.1, and definitely includes both the XBB recombinant lineages and more recently the BA.2.86 (including its descendant JN.1*) lineage. Such variants end up being *very* divergent from, and go on to replace, what was there before.
Thus, Omicron BA.1* replaced Delta. The XBBs replaced the previously dominant BA.2* and BA.5* lineages. And now JN.1* has replaced the XBBs. You may notice also note that this has corresponded to all three times thus far that an updated vaccine was deemed necessary.
Speaking of vaccines, one of the issues is the failure to aim ahead of continuous evolution. The first booster targeted wild type plus either BA.1 or BA.4/BA.5. There was no BA.1 by then. The second targeted XBB.1.5. Same issue.
The next vaccine update will target JN.1, which is already mostly gone, or its descendant KP.2 (a "FLiRT" variant), which is likely to be replaced soon by variants with other mutations.
Again, that's just missing *continuous* evolution, which can be predicted fairly well.
Again, that's just missing *continuous* evolution, which can be predicted fairly well.
The arrival and takeover by new lineages is not the same thing as the continuous evolution involving the somewhat predictable accumulation of fit mutations within lineages. Instead, it involves one-off events that are rare but have major impacts on the variant landscape.
It's the difference between, say, evolution of beak shape in Darwin's finches on the Galapagos Islands vs. a meteor wiping out non-avian dinosaurs, which were ultimately replaced as the dominant terrestrial vertebrates by the evolutionary radiation of birds and mammals.
Major events happen repeatedly, but intermittently and less predictably. That is, they're continual rather than continuous. We don't usually see it happening day to day like we do with evolution within lineages among hosts, but when these divergent variants do show up it's major.
We can see the effects of continuous small-scale vs. continual large-scale evolution when we consider spike mutations specifically. Note the big jumps with the appearance of each new major variant lineage, followed by much smaller, more gradual change within those lineages.
Here's the thing. The more transmissions there are to new hosts, the more continuous interhost evolution there will be. The more hosts are infected, the more persistent infections and coinfections there will be, and so the more continual intrahost evolution there will be.
Less virus, less evolution of new variants. More virus, more evolution of new variants.
* And most of those do not manage to establish as a lineage at the host population level -- but some do.
I make typos on my phone. Like COVID, you'll have to learn to live with it. 🙃
• • •
Missing some Tweet in this thread? You can try to
force a refresh
