The following continues from Part 1:
Instead of the second and third “myths” which were previously presented in the following definitely not-copyright-infringing graphic, Part 2 will tackle a secret add-on bonus myth!
ii. There was a coronavirus pandemic in 1889.
It is a coronavirus, not a flu.
Apologies for springing this stunning revelation at the reader without warning. SARS-CoV-2 is a coronavirus.
In Part 1, we looked at why flu’s genetic dynamism is at least partly a myth. Nonetheless, the widespread equivalence between this “novel” coronavirus and past and present flus which was smuggled into the collective unconsciousness in early 2020 also demands we ask: Was there ever any reason to think the two should behave similarly?
Coronaviruses were not known to cause pandemics.
Given that viruses were, before the era of electron microscopes and cell culture, simply a placeholder concept for invisible “somethings” that could be observed to transmit disease from one animal to another, it is very difficult to assert a negative in terms of previous pandemics. I can only wage the case that the evidence for a coronavirus pandemic in human history doesn’t exist.
First, it is necessary to reassert my definition of pandemic — this is a naturalistic word used to describe “everyone, everywhere” becoming sick in a short time. What requires reference to health care statistics to validate is not a pandemic. A self-limiting outbreak is not a pandemic. Therefor, no pandemics have in fact occurred in the 21st Century. There has simply been an outbreak of calling limited outbreaks pandemics. At best I would concede that Omicron has behaved like a pandemic; but not the versions of SARS-CoV-2 that preceded it. They were slow, ephemeral, and often self-limiting (outside of high-population-density areas that did not lock down).
Clarifying the 1889 “Coronavirus Pandemic” Retcon
In 2004, a paper was published which attempted to fix the date when coronavirus OC43 crossed over from a cow coronavirus.1 Essentially, the two viruses are still close to identical. Here is the bovine (white) and 1960s human OC43 (red) spike protein, focusing on the junction with the, ahem, furin cleavage site.
Since the human version has been mutating at a "certain rate" since first successfully being recovered in cell culture in the 1960’s, one can align the added difference between then and today and guess what year the virus started infecting humans. It lands in the end of the 19th Century:
Astoundingly, this work was rehashed just in time to greet our newly-declared Pandemic™ in 2020.
A Danish TV program is potentially responsible for giving the retcon enough energy to have partly seeped into the wider consciousness since then (the show is justified by “recent work” from two Danish researchers; but I can’t find anything newly published by them on the topic).
There are, of course, several problems with the theory. As my mark-up of the graph above shows, the virus’s rate of evolution appears to be accelerating in the final years before the study. What could cause a human virus to mutate faster in the 1990’s than in the 1960’s?
Population growth = more infections = quicker accumulation of viral generations
If the average copy of the genome became “1 copy older per year” when the population of the earth was 3 billion, it might become “2 copies older per year” when it is double that. So an adjustment for the acceleration effect is necessary. But at the same time, OC43 might have mutated more quickly when first adjusting to its new host.
But, the reality is that Vijgen, et al.’s work probably needs to be re-analyzed from the ground up. Sequencing of a differently-passaged version of the 1960’s OC43, conducted the same year, showed only 6 nucleotide changes with a current isolate; vs. 30+ in the Vijgen sequence. This suggests either that the latter was potentially heavily mutated by cell passage, which renders it unfit for use in molecular clocking, or that the former was a contaminate.2 Moreover, current analysis of newly found genotypes of OC43 (usually regionally clustered in Asia, where no one had been looking until recent years) suggests that all current clades share a common ancestor from the 1950s; so this could be the actual decade of emergence.3
It’s fair to say that the best we can do with the molecular evidence is just guess. Which reality do you like?
As it turns out, however, it is plausible that some kind of weird, non-flu illness spread around in the modeled year of OC43’s emergence — but it wasn’t a “pandemic.”
First, it is important to clarify that 1889 was clearly a flu pandemic. Half of people got sick with flu-like symptoms in a given area really quickly, all over the world. Decades later, it was clear from the antibodies that were still in people alive in that era that some type of new flu appeared around 1889, either with an H2 or H3 spike (HA) glycoprotein. So the idea that a coronavirus really “was the 1889 flu” is ridiculous.
But then in 1891/2, things got weird. As Edwin O Jordan would recount (emphasis added):4
The initial influenza epidemic of 1889-1890 in England and Wales, compared with the respiratory epidemics of the succeeding years, is seen to have a character of its own, just as has been shown to be the case in the United States. The epidemic of 1891, as contrasted with that of 1890, was said by Dixey (1892) to be “distinguished by its much greater severity, by its comparatively slow rise, its protracted period of high intensity, and its rapid and uniform fall.” The later outbreaks were also characterized by a different age distribution of mortality as compared with the 1890 epidemic, and showed a much greater fatality at advanced periods of life. The official tabulations of standardized mortality from influenza in England and Wales, given by the Registrar General (1920) for the years following 1891, show without exception a proportionally lower mortality than in 1890 for the ages 0-15, and 15-35, and a higher mortality for the ages above 55.
Could this have been OC43? Sure. But was it a “pandemic”? No, it was a slowly developing, quickly disappearing, sporadic illness mostly affecting some of the very young or very old. While the death toll exceeded the 1889/90 flu, neither was dramatic in absolute terms.
Plenty of other years in the 1800’s feature unclear, not-flu-like waves of recorded higher respiratory deaths in various records. The only thing 1891/92 has “going for it” is adjacency to the preceding flu pandemic (which, though it caused fewer deaths, was a true pandemic in terms of regionally simultaneous experiences of illness).
But really, OC43 could plausibly have crossed from cows to humans at any time in the century(s) before it was first isolated; or it could have done so just a few years before, in the 50s. It is not a precedent for a “coronavirus pandemic.”
Ok, ok — but if coronaviruses weren’t known to cause pandemics, they were still known (like flu) to adapt to evade human immunity, right?
No one knows if human coronaviruses adapt to evade immunity (!)
The first analysis attempting to even probe the topic may have been in 2015. To quote the authors (emphasis added):5
[L]ittle is known about how HCoV-OC43 genotypes persist in human populations. It is assumed that the continuous adaption of viral antigenic gene is required for the persistence of OC43 genotypes. However, this hypothesis has not been carefully examined by precise evolutionary pattern analysis.
You know what they say when you assume…
At all events, what did the authors find when they looked for evidence that the recently busy-in-China OC43 clade, D, was under any sort of selection pressure?
Ren, L. et al. (2015) findings:
Some sites in the N-terminal domain (this is where the OC43 spike protein binds 9-O-acetylated sialic acid, meaning it binds with its “forearm” rather than its “hand”) had high rates of “change the amino acid” vs. “don’t change the amino acid” mutation, suggesting selection pressure
But so did sites downstream of the NTD, including in the normally highly conserved S2 region (the spike protein’s upper arm).6
The Spike gene of the D clade mutated at a rate of .883× 10−3 substitutions/site/year. About double the rate of less active C and E clades; which again could simply be a reflection of infection rates (more virus generations added per year).
And so it could be that OC43-D, like flu, mutated some of the amino acids adjacent to its sialic acid binding site between 2004 and 2012 due to immune pressure. Obviously, more work would need to be done — like trying to passage OC43 through immune animals to see if an NTD-mutant pops out.
On the other hand, the overall rate of mutation of the OC43-D spike protein during the study period is still markedly unhurried when compared to the (semi-mythical) exemplar of immune escape, influenza A:
And in general, coronaviruses tend not to score aberrantly high for mutation rate in the overall roster of sequenced RNA viruses. From Holmes’s Evolution and Emergence of RNA Viruses:7
Given that, once again, accumulation of viral generations per year is a factor in the rate of mutation per year, the rate for coronavirus in the graph above is potentially further inflated by how fast coronaviruses reproduce in other animals, where infections may be more frequent thanks to modern industrial farming.
In sum, what was anyone even thinking?!
There is little to no evidence that human coronaviruses either mutate quickly in general, nor that they mutate in response to human immune pressure.
So why has it been taken and passed-off as a “given,” for the last 3-odd years, that SARS-CoV-2 would do either one of those things?
Again, I will quote the same review of the virus’s early evolutionary behavior from 2020:8
Viral diversity has challenged vaccine development efforts for other viruses such as HIV-1, influenza, or Dengue, but these viruses each constitute a more diverse population than SARS-CoV-2 viruses.
Finally, in Part 3, we will look at just how little SARS-CoV-2 proved to evolve in wild transmission (at least before the Omicrons). Spoiler: It was half the speed of OC43-D.
If you derived value from this post, please drop a few coins in your fact-barista’s tip jar.
Vijgen, L. et al. “Complete genomic sequence of human coronavirus OC43: molecular clock analysis suggests a relatively recent zoonotic coronavirus transmission event.” J Virol. 2005 Feb;79(3):1595-604.
St-Jean, J. et al. (2004.) “Human Respiratory Coronavirus OC43: Genetic Stability and Neuroinvasion.” J Virol. 2004 Aug; 78(16): 8824–8834.
Recently, L. Vijgen and coworkers have submitted a complete sequence of the HCoV-OC43 genome to GenBank (NC_005147). The virus strain used for this sequencing is described as corresponding to the virus strain that was used in our laboratory (VR-759). However, comparison of our sequence with theirs show that they differed at 33 positions, 29 mutations being located in the S gene, including two mutations in the S2 subunit. Of the four other differences, one is located at the beginning of the genome sequence, where a guanine is added with respect to our sequence, whereas the other three are scattered throughout ORF1a. Despite these differences, the availability of the complete genome sequence from a clinical isolate reinforces the validity of our sequence, since the HCoV-OC43 ATCC and Paris sequences only differ by 6 nt. Therefore, this observation suggests that the viral strain used by Vijgen and collaborators could have been adapted in cell culture, given the differences observed in the S gene, which is known to be associated with viral adaptation (27). No differences were noticed among ORF1b sequences between HCoV-OC43 ATCC, Paris, and the one from Vijgen and coworkers.
Lau, SKP. et al. (2011.) “Molecular Epidemiology of Human Coronavirus OC43 Reveals Evolution of Different Genotypes over Time and Recent Emergence of a Novel Genotype due to Natural Recombination.” J Virol. 2011 Nov;85(21):11325-37.
p. 144. Jordan, Edwin O. (1927.) “Epidemic Influenza: A Survey.” Archived online at https://quod.lib.umich.edu/f/flu/8580flu.0016.858
Ren, L. et al. (2015.) “Genetic drift of human coronavirus OC43 spike gene during adaptive evolution.” Sci Rep. 2015 Jun 22;5:11451. doi: 10.1038/srep11451.
Ren, L. et al.
Note, importantly, that the NTD actually contains the Receptor Binding Domain, not the residues labeled “RBD” (those were erroneously flagged by a computer program in a previous study). The binding-site-adjacent positive selection sites are 33 and 90, with 38 and 93 close behind.
For the correct sialic acid binding site, see Hulswit, RJG. (2019.) “Human coronaviruses OC43 and HKU1 bind to 9-O- acetylated sialic acids via a conserved receptor-binding site in spike protein domain A.” Proc Natl Acad Sci U S A. 2019 Feb 12; 116(7): 2681–2690.
Edward C. Holmes. The Evolution and Emergence of RNA Viruses (Oxford Series in Ecology and Evolution) (p. 54). Kindle Edition.
Dearlove, B. et al. “A SARS-CoV-2 vaccine candidate would likely match all currently circulating variants.” Proc Natl Acad Sci USA. 2020 Sep 22; 117(38): 23652–23662.
Coronaviruses tend not to score aberrantly high for mutation rate in the overall roster of sequenced RNA viruses, but I understood that was because they have a fairly effective copy-correction mechanism. The 4 common human coronaviruses - as I was told - have a fairly unique protein that modulates this protection. This in essence causes significant more 'bad' (infertile/broken) copies, but allows also for faster mutation. (COVID-19 does not have this protein BTW.) Hence, why the 4 common coronaviruses could reinfect humans over and over again.
At least that was the theory as I was told.
But if not true, then how can we get reinfected? Current doctrine suggests that immunity in the humoral parts of our body is for life. For flu we know it is, so I'm skeptical our memory CD4+ and CD8+ T cells for common corona are not effective for decades too.
Then if not mutations, then either cellular immunity in the throat/mucous has weaker memory that doesn't last as long as humoral immunity, or it has perhaps a too slow response there. After all somehow we get sick again.
I also do know that a Canadian study that followed some people for 20 years showed that pretty much everybody gets reinfected again with these common cousins of COVID-19. The frequencies ranged from 3 times in these 20 years to yearly and likely depend very much on your age and lifestyle. But still some of them had a yearly re-infection. So it seems that either the mutation rate must be fast enough to throw us off, or mucous immunity is indeed very weak.
That is important, as it is a predictor for COVID-19. Its spike doesn't seem to mutate very fast and even the help it got with Omicron's release did not materially change its antibody resistance. So, as it has no animal reservoir, it will become extinct if it has to rely on that.
But if it can rely on our throat/mucous cellular immunity memory being weak enough, it can become the 5th common corona yearly sneeze.
In that sense it is surprising, so little research has been done on throat/mucous cellular immunity vs humoral immunity.
(For flu BTW I was told its mutation freedom is the result of it not targeting a large protein and hence not having a receptor binding protein but targeting an amino-acid. And hence, our antibodies must target the nucleus giving flu more freedom to make its skateboard moves. Until as you pointed out in pt 1 it runs out and runs into the wall anyway.)
I don't know much about biology. Especially molecular biology. I've studied theories of evolution in kind of an outsiderish amateur way, and I know some things about dynamical systems and computer security in a more "serious" way. Squinting down through the fog from great heights of abstraction, a few rambling observations:
A virus is dependent on its host to reproduce. To even a greater extent than other (always ecologically situated) organisms, it has to co-evolve. With such a small genome, it can't hoard up a big bag of tricks allowing it to jump around on the fitness landscape, the way bacteria are known to do. Maybe a little bag of tricks. This landscape is determined by the hosts, and big lumbering multicellular guys like us change pretty slowly. On the other hand, our complex proteomes offer a vast "attack surface".
Viruses are definitely "swarms" in that they reproduce rapidly, in great numbers. Swarming doesn't, by itself, get you out of local fitness optima. You have to still do well enough in the sub-optimal valleys to survive to climb the next hill. A delicate little hopeful-mutant virus might not survive the immune onslaught.
Cultural factors (vitamin deficiencies, therapeutic fads, agricultural practices or other contacts with potential peer hosts) influence the overall landscape for human viruses. There are opportunities for "punctuated equilibria" to arise, at the decade-to-century scale. More so during periods of rapid change in material culture.
Some of this is just recapitulating your previous article. Did I get the gist? I'm enjoying these! Keep doubting the great doubts.