The Amazing Genome and Evolution of SARS-CoV-2

SARS-CoV-2 caught most experts off guard with its evolutionary alacrity, but it always should have been expected to be very good at evolution despite a relatively low mutation rate. The progenitor virus(or, at least, the majority of its genome) came from bats. Bats have amazing immune systems due to their colony lifestyle in cramped quarters and their high metabolic rate.

Further, coronaviruses have the longest genomes of any known +ssRNA viral family, and viruses are under constant evolutionary pressure to keep their genomes small. The genome of SARS-CoV-2 is about 30kb in length. The value of the genome was evidently great enough to compensate for that pressure. There is more highly conserved genetic material in coronaviruses than others, which in SARS-CoV-2's case means a lot of finely tunable accessory genes that screw up the immune system badly and even vicious microRNA's outside the coding regions of actual genes.

SARS-CoV-2 has absolutely ingenious replication machinery. RNA-dependent RNA polymerase(RdRp)/nsp12 is extremely error-prone in its transcription and can swap strands between viruses of different strains in co-infected people, creating a chimaeric new strain, a process known as recombination. To compensate for this sloppy copying, after helicase processing, an enzyme known as exoribonuclease(ExoN)/nsp14 proofreads the genome. However, it can only proofread that which is there, making deletions and insertions particularly common.

ExoN is so important for SARS-CoV-2 that, unlike for some coronaviruses, knocking out the gene encoding it results in such wild mutation that SARS-CoV-2 fails to produce viable progeny. This and the early number of large, conserved deletions should have been red flags that the virus' evolution was led by recombination, not SNP's, and that it would eventually evolve very rapidly.

https://pubmed.ncbi.nlm.nih.gov/32938769/

Recombination is a great evolutionary strategy. Beyond the possibility of rapid acquisition of beneficial mutations from different parent genomes, recombination is an extremely effective secondary strategy for eliminating deleterious indels and keeping beneficial ones. The ability expands the evolutionary space and increases the pace of SARS-CoV-2 evolution dramatically. RdRp's higher error rate in SARS-CoV-2 was a strong indication that recombination is key to SARS-CoV-2 evolution, which has now been empirically shown.

https://www.biorxiv.org/content/10.1101/2021.08.03.454981v2.full

Furthermore, because RdRp and ExoN are encoded by the virus itself, it can modulate its mutation rate and style. It will tend towards high-fidelity replication when it is transmitting easily and tend towards more mutation when it runs into roadblocks.

For this reason, our vaccines, targeting a highly mutable region such as S, were a poor idea. It's similar to why we can't develop a universal flu vaccine or an HIV vaccine easily: evolution simply happens too rapidly. We were never likely to achieve a vaccine for SARS-CoV-2 S that retained high efficacy for a long period of time, especially with such a high R0.

The targets for vaccines, natural immunity, and therapies like monoclonal antibodies(mAbs) are frequently the same, so introducing selective pressure by vaccination is gradually robbing us of epitopes that are greatly needed for mAb and other treatment. You can think of it like spraying an inadequate quantity of an antibiotic over the whole population. We have hoped it would be adequate, but instead, it's producing the viral equivalent of a lot of antibiotic-resistant bacteria.

How many progressively more challenging passages through immune responses were required to produce a virus completely resistant to the best serum found in 20 convalescent patients? 13, and a total of 3 mutations. That should have been another shot across the bow, and we got it early.

https://www.pnas.org/content/118/36/e2103154118

We have effectively reconstructed this sort of evolutionary ladder with the wide variation in antibody protection by person. This would have happened anyway through natural infection, but slower and less dramatically. Only simultaneous global administration of extremely effective vaccination for all extant strains would have sufficed, and that would be politically and logistically impossible.

We needed to focus on reducing serial transmission through non-pharmaceutical intervention more intensely. Treatments also develop resistance, but they do not provide the chance for OAS, and this resistance is not passed on if the patient is well contained. Neither ADE nor OAS has been conclusively shown in clinical COVID-19, but they have each been shown in studies, and with SARS-CoV-2's big genome, it may be easy for SARS-CoV-2 to develop those clinical abilities in further evolution. Vaccination could enable OAS.

This is why we are so careful about prescribing antibiotics, often prescribe two at once, and ensure people take the full treatment. Treatment was always the more promising approach than vaccination, and vaccination makes development of treatments more challenging by promoting viral immune escape.