>at least unless we get a vaccine-resistant strain.
...which will happen promptly. The selection pressures are perfect. You already have a variant with vaccine breakthrough, circulating in a partially vaccinated population.
I don’t know if you can make that bet - evolution of the virus needs to conserve the current mode of cellular entry and attachment and fitness of spike appears to be the main driver at present - and then spike mutation has driven drift from the original vaccine target leading to a degree of immunity ‘escape’.
So the question is, is the virus now in a local minima where it is highly adapted for spread (ie big increase in R0 from ‘original’ virus) and are those selective pressures in a situation where the virus can’t jump into a full immunity escape mode without climbing down off mount improbable and taking another face?
My bet would be that the selective pressures may continue to drive fitness for reproduction, but won’t lead to immunity escape because spike is too conserved
I'll echo that this is a reasonable assessment. In evolutionary hill climbing, the fitter a protein the fewer paths there are to climb higher [0]. Art Poon has some interesting work that effectively maps viral phenotype space by looking at compensatory mutations [1]. He induces deleterious mutations in page phage, grows them until they recover fitness and then assesses if they regained fitness by reversion or compensation. The worse a deleterious mutation, the more likely a compensatory mutation is to arise.
The more optimized the covid spike becomes, the harder it is to get better. What frightens me are polymerase mutations.
Mutations in the genetic replication machinery of the virus. All the world's worst and most untreatable diseases tend to have very fast, but very low quality polymerases.
This means they replicate quickly, and generate mutagenic variants quickly. HIV and Ebola have this in common.
Delta is already selecting for viral load over vaccine escape - if it keeps getting faster then it could become more dangerous, but it would also increase the rate it can explore it's genetic "phase space".
Rather than increasing efficiency, I'd be concerned about loss of function in the error checker. This would spike the mutational rate (and the mutational load).
Interesting, thank you. If I understand correctly, viral load mutations are the current and quickest path to spreading faster, currently. It hasn't selected as fast for vaccine escape, yet, which is not to say it isn't trending in that direction with delta?
Or, we're kinda f'd with regular transmissions at the moment, and vaccine escape is a possible accelerant?
The problem is closer to whether the spike protein is near an local maxima for infectivity since it's involved in binding to ACE receptors. Since mutations in COVID seem to conserve the spike protein region, it presently looks quite likely that it can't be changed very much without performing worse. Hence the viral replication optimum: spreading to more individuals makes the virus dominate the population quicker.
Evolution does whatever leads to more of the successful thing existing, so viral loading was a pretty obvious move: the question is whether there's any vaccine-evasion which doesn't compromise that advantageous (Delta suggests yes, but its possible that's as far as it can actually go).
I don't know or haven't seen any evidence that polymerase mutations are responsible for the increased R0 of Delta, although it's certainly possible.
What we do know is that the Spike protein is essential for COVID cellular binding.
We also know that spike has mutated between 'vanilla' COVID of 18 months ago and what all vaccines target, and Delta.
This makes sense, because the spike (or S) protein wasn't optimised for binding to Human ACE2 receptors; over time mutations in S (ie with Delta) have lead to increases in binding affinity with ACE2, the result being that Delta is more likely to infect a cell with the same viral load as 'Vanilla'. This would be a perfectly acceptable reason for the uptick in viral loads in Delta and infectivity and hence R0.
We are lucky/the vaccine researchers are smart in that S is quite antigenic (that is, triggers a good immune response, although in a 'natural' infection people will produce antibodies to many parts of the COVID virus), and because the selection pressures driving improvements in ACE2 binding affinity of S have not lead to such a great conformational change that antibody responses are useless (ie antibody-antigen affinity is decreased in Delta, but still effective), our vaccines against the earlier S protein are still effective.
Because of the way evolution generally works, pressures driving S evolution tend to end in a local minima (or maxima, depending on your way of looking at it - i tend to think in terms of conformational entropy so hence Xornot and my use of different terms for the same concept) - or as I mentioned earlier, 'Climbing mount improbable', a term coined by Dawkins in the book of the same name.
It is because it is very hard to get out of these local minima that the human body is full of such oddities, such as an eye that is designed with the nerves in front of the photoreceptors, resulting in a blind spot - it's easier to adapt around the minima to reach a new optimum then it is to reinvent the eye with the nerve in the position that would be logical from a functionality standpoint)
Therefore, it is very very unlikely that the S protein, or the virus, will be able to mutate to a S protein that is so radically different that it escapes antibody-antigen affinity from the vaccines we have today - the S protein would basically have to be completely redesigned, which is really unlikely to occur by the gradual process of evolution. Ie the chances of shaking the RNA coding for S into a new functional protein that works and is a completely new, novel shape, is infinitesimal.
So - basically, total vaccine escape is really really unlikely. Like, number of grains of sand on earth unlikely, if not more so. But polymerase mutations could increase the rate at which the grains of sand are sifted.
It is not something that I personally would lose much sleep over - we don't have any evidence from nature that I am aware of (ie flu virus etc) of highly conserved/selected proteins all of a sudden spitting out a completely new model - basically new diseases that we worry about come from species shifting (ie... let's just say SARS to avoid a debate about COVID origins)
Measles, polio, and smallpox beg to differ. The varicella vaccine is fairly new, but I haven’t heard of a vaccine-resistant strain evolving. Pertussis seems to be an odd duck, and its vaccine is not as effective as people would like.
History is mixed here, but most of the mandatory vaccines seem to be extremely effective and to work well for at least decades.
Those also have low breakthrough rates (assisted by high vaccination percentages), giving it relatively no opportunity to evolve. They also had millennia to reach (and then work to maintain) peak fitness, while COVID is still encountering beneficial mutations somewhat frequently.
Presumably also the human species had time to co-evolve some resistance to these illnesses. I shudder to think at the cost... hundreds of millions of lives for smallpox alone...
I'm not sure if it's inevitable given how crucial the spike protein is and how strongly it is being conserved. Even with the Delta variant, which is much more infectious and contagious, the vaccines are only modestly less effective and still highly effective at preventing severe illness and death.
...which will happen promptly. The selection pressures are perfect. You already have a variant with vaccine breakthrough, circulating in a partially vaccinated population.