Dangerous Original Antigenic Sin Anticipated in SARS-CoV-2
Disclaimer: This is an opinion and a prediction that includes as many sources and the best logic possible. Do not base your personal health decisions on it.
Plain Language Introduction
SARS-CoV-2 is a novel virus with its closest relatives in bats, emerging in the human population at some point in late 2019. As a newly introduced pathogen, there was only one variant at first. Now we have thousands of strains that have been identified, with countless individual virus particles with countless mutations. The genetic diversity of SARS-CoV-2 is much greater than scientists thought it would be by this point in time.
That's because we failed to notice that SARS-CoV-2 relies heavily on recombination for evolution. With recombination, in a host that is infected with two strains of the virus, the virus can effectively create a child from two parent virions(a virion is a single virus) through a process called strand switching. When creating a new virion, it starts reading from one virion's RNA, and then it abruptly starts reading from another virion's RNA. Strand switching can occur multiple times in the creation of a child. This makes a new hybrid virus that includes mutations from both parents, so it's like sex in living creatures.
For the virus to recombine, a person has to be co-infected by different strains of the virus. This will happen more frequently as the virus continues to evolve, because the adaptive immunity developed by the body in response to infection or vaccination will be less effective against other strains of the virus: they will be more genetically distant. It becomes easier for a second variant to infect someone who's already infected with an unrelated strain. This means the probability of recombination will increase over time, increasing the evolutionary rate of SARS-CoV-2 exponentially.
SARS-CoV-2 recombines frequently. This allows SARS-CoV-2 to rapidly try new combinations of mutations and effective combinations to spread. The more mutations there are in parent virions, which accumulate randomly at a rate depending mostly on how many cases there are, the more possible recombinants can be produced. This additional boost to recombination is layered on top of the growing likelihood of co-infection.
Most of the variants of concern like Delta and Alpha were created this way, as the only other plausible mechanism to explain the observed rapid evolutionary rate of SARS-CoV-2 is long-term infection in single hosts, which is unlikely given how frequently this has happened.
Vaccinated and Recovered People
When someone is infected by or vaccinated against a virus that their body doesn't recognize, it generates a new adaptive immune response tailored to that virus. This process takes some time, so vaccination allows it to happen in advance.
Highly effective vaccines were developed against the initial virus. These vaccines have a single target: the S protein of the original virus. They worked very well, but our use of them was based on the idea that we would be able to produce and distribute them quickly enough to keep up with the virus' evolution. We have since learned that the virus evolves much faster and better than we expected, and we've seen the effectiveness of the vaccines drop significantly against new variants. With further evolution, we'll see further drops, and this evolution is inevitable without eradication.
What does that mean for people who have been vaccinated or infected before? The strong memory of the immune system mostly protects you, but are three possibilities: the immune system might be able to continue to use its existing immunity to protect you, it might consider a variant to be different enough that it needs to make a new immune response, or it might continue to use its old response, but that existing immunity could be useless against the new variant -- or even harmful.
The whole concept is known as original antigenic sin. I believe that SARS-CoV-2 will exhibit dangerous original antigenic sin, like many other viruses do. We've just haven't seen it clearly yet because, as a new disease, there wasn't enough genetic variability in the virus for it to be an issue. With the rapid and exponentially increasing evolutionary rate of SARS-CoV-2, I believe it's about to be a big issue.
Original Antigenic Sin(OAS)
The classic definition of OAS is the failure to produce neutralizing antibodies or protective cells in response to a virus that is too similar to another virus that infected you earlier, but vaccination can be thought of as infection in this context. Once the body thinks it recognizes the virus, it will rely on existing, tested immune responses rather than slowly generating a new one.
The lack of neutralizing antibodies is the biggest problem because they offer the best protection against infection. Although it's rarely discussed, the same phenomenon happens with T cells and other immune cells.
The immune response by people who were infected rather than vaccinated is, with time, both broader and more durable than the immune response of vaccinees. This is likely due to memory B cell effects, with much less somatic hypermutation and affinity maturation exhibited in infection-naive people than in vaccinees, where the germinal center processes seemingly never really get kicked off. In essence, vaccinee antibodies are less diverse and prolific.
Some viruses twist the immune system's adaptive response strategy to their advantage by encouraging production of a stale immune response that will be ineffective or even beneficial for them. The best example is dengue virus, but this is reasonably common: it happens with influenza too, the virus where this was first discovered. It has not been observed definitively in SARS-CoV-2 yet, but I believe that is because there was insufficient genetic diversity, and that is now changing quickly.
All in all, vaccination can range from highly protective, to unprotective, as it was for people trying to get Gardasil-9 after getting Gardasil, to actively dangerous. Dengvaxia protected against 1 of 4 kinds of dengue fever, but left vaccinated people more vulnerable to severe disease from the other kinds. Hundreds of thousands of children were gravely endangered, and researchers nearly went to jail. Which category current COVID-19 vaccines will fall into ultimately depends mostly on the virus and its evolution, which is hard to predict.
If OAS by current vaccines is problematic, it would not have been revealed in clinical trials and will only emerge after enough time and evolution.
Detailed Scientific Analysis with References
Ra4991, since renamed RaTG13, is the closest known natural relative to SARS-CoV-2. It is endemic in bats and was discovered after an outbreak at a mine in Mojiang, Yunnan Province, China. Its emergence was precisely what would be expected of a zoonotic spillover: a handful of people were exposed to extremely high levels of inoculum and infected with very severe disease, but there was no onward transmission.
The presumptive path from RaTG13 to SARS-CoV-2 is controversial and immaterial to this article. The important points are that SARS-CoV-2 came as:
- A single genome that originated in bats and was ~29kb of +ssRNA, which is extremely large. Viruses are under constant evolutionary pressure to keep their genomes small, so this genome, among the largest +ssRNA genomes known, must include a great deal of genetic material worth keeping.
- An extremely contagious virus with an estimated R0 ranging from 2.5 to 5.7 in humans at its introduction, despite having a genome generally designed for bats.
SARS-CoV-2 is a sarbecovirus, and other viruses in its clade such as SARS-CoV and MERS-CoV exhibit profound pathogenicity in humans. However, none of them has ever had a chance to transmit for an extended duration in humans, so we don't know what the optimal level of pathogenicity of a generic sarbecovirus is. This understandably drew immediate alarm from the scientific community. Variants to date have not demonstrated wide variability in pathogenicity, and studies disagree as to whether they are trending towards greater virulence.
A Matter of Time
A race to develop vaccines began, focused on the original S protein from the isolate Wuhan-Hu-1. Subsequent research has revealed that this is not the original virus, differing by a few nucleotides, but it's close enough.
Vaccines were developed with amazing alacrity and efficacy against the ancestral strain of SARS-CoV-2, and evolution appeared to be very slow. This turned out to be an illusion created by the low genetic diversity present at the emergence of the pathogen. Since SARS-CoV-2 relies extensively on recombination for creation of new strains and evolution, mutations needed to gradually be introduced by SNP's and be selected for before recombination could kick in, a slow process. The vaccines would appear effective for a long time at first.
The short-term efficacy of the vaccines, particularly in prevention of severe disease, was not in question: they worked well. They create an enormous initial surge of anti-Wuhan-Hu-1-S and subunit antibodies, memory B cells, CD4+ and CD8+ T cells, and durable Tfh cells.
The efficacy is sharply time-bounded, however. This implies that the bulk of the protection conferred by the vaccines comes from rapidly waning antibodies, which is consistent with studies of correlates of protection, which have found anti-S and anti-RBD antibodies to be key.
Limited attention has been paid to the longer-term effects and efficacy of vaccination, as they cannot yet be observed, and prediction is a risky practice for a researcher. Vaccine formulation is always a balancing act. Booster shots are already being administered, but there are safety concerns about increasing vaccine immunogenicity, and simply increasing immunogenicity will not be able to address loss of efficacy in the long term due to evasive mutations. Finally, because the vaccines are leaky, continued evolution of SARS-CoV-2 is virtually guaranteed and may result in a very pathogenic virus.
Cellular Immunity and SARS-CoV-2
The strong correlation between neutralizing antibody titers and vaccine efficacy draws attention to the extremely complex relationship between T cells and SARS-CoV-2. On net, T cells appear to be protective, particularly early in infection. However, high numbers of exhausted or otherwise dysfunctional T cells are often found in severe infections and appear to lead to pathogenesis.
Most of SARS-CoV-2's accessory proteins directly interfere with the immune system's ability to function properly. Some of them specifically dysregulate T cell responses.
An early strain of SARS-CoV-2 that was somewhat successful in Europe, B.1.177, directly altered a CD8+ T cell epitope with the mutation S:P272L, leading to escape. However, this mutation has appeared only sporadically and never become fixed in any other identified lineage, indicating that T cell escape is an inferior strategy.
Instead, it appears much more beneficial to directly manipulate T cells. ORF8, for example, downregulates MHC-I, making it much harder for CD8+ T cells to recognize and kill infected cells. Earlier strains of SARS-CoV-2 often included premature stop codons in ORF8, presumably to limit pathogenicity and enhance spread, while current VOCs bear multiple mutations in the gene. This is a unique ability for a respiratory virus and it should be a warning sign.
Children actually have lower CD4+ and CD8+ T cell responses to SARS-CoV-2 than adults and they exhibit vastly less vulnerability to COVID-19. Conversely, sera taken from severe cases in adults exhibits fewer T cells than sera from mild cases during the acute phase of disease.
The low level of circulating T cells in such severe cases is partly a relic of localization, as T cells are drawn to infected tissue alongside resident T cells. SARS-CoV-2-specific T cells resident in the lungs are very different from circulating T cells, and COVID-19 induces apoptosis in these critical cells.
SARS-CoV-2 T cells recognize a median number of 35 epitopes, making it much more difficult for the virus to evade them. It is highly unlikely that a sufficient number of mutations could be made that would not be detrimental to the virus. Even then, the epitopes recognized by the T cells differ by the HLA genotype of the host.
As a result, I do not expect a dominant strain of SARS-CoV-2 to mutate epitopes to evade recognition by T cells. I expect one to develop its other strategies to evade and abuse T cells for transmission, and soon.
T cells resulting from SARS-CoV infection persist for decades. We should expect similar durability from T cells generated by vaccination against or infection by SARS-CoV-2 and plan accordingly.
ADE in COVID-19
ADE is exhibited by some diseases when the level of non-neutralizing cross-reactive antibodies is greater than the level of neutralizing antibodies. Likely because most antibodies remain neutralizing for current strains and RBD antibodies are more neutralizing, it has not been observed clinically in COVID-19 yet.
Antigenic drift, falling antibody titers, and the evolution of other mechanisms to shield or alter immunodominant epitopes used by antibodies will gradually lead to the ratio changing. This will approach and then tip over a point to where ADE manifests itself progressively, and probably at little cost to the virus.
A variant of B.1.617.2 was created in vitro which evaded all neutralizing antibodies from current vaccines. It demonstrated enhanced infectivity even in highly concentrated sera. COVID-19 results in the production of antibodies that enhance infection, some of which work by promoting the open conformation of S independently of Fc.
Positive selection for antibody escape has been pronounced, with dominant strains being less neutralized by sera than the original SARS-CoV-2. However, not a single VOI or VOC contains a mutation in S:167-180, the most common and highly immunodominant Tfh epitope, despite mutation in the NTD being extremely common. This shows that SARS-CoV-2 wants to appear like it's the same virus even when it is a mutant.
The NTD and the RBD of S are the two major antibody targets. The full purpose of the NTD in SARS-CoV-2 is unknown even though it is highly immunogenic and conserved as a subunit. Tellingly, it is used for sugar binding in most coronaviruses. Based on the immunogenicity, infection-enhancing NTD antibodies, and its interaction with biliverdin, I propose that immunogenicity for immune dysregulation is a primary function of the SARS-CoV-2 S NTD. The genetic localization of the NTD immediately proximate to the RBD further supports this assertion, as does its fully folded structure and physical location in all conformations of S.
The NTD may serve as a stalking horse for the RBD, triggering antibody generation against it that is much less potently neutralizing. Because it contains many antibody and T cell epitopes, it would also make OAS easier to achieve without sacrificing viral function. Delta has already completely escaped from NTD antibody neutralization by vaccine sera.
More antibodies are produced against the NTD than the RBD by infected people, but more are produced against the RBD than NTD by vaccines, though it varies by vaccine. Antibodies from vaccines are more likely to be non-neutralizing overall. Additionally, as earlier noted, the memory B cells that result from vaccination don't proceed through germinal centers nearly as well as those produced by infection.SARS-CoV-2 also performs ADE through the same mechanism as dengue virus, afucosylation. This afucosylation leads to at least three detrimental effects: thrombosis, clotting, and heightened inflammation.
OAS Enhances ADE and its Cellular Version
SARS-CoV-2 is also capable of infecting macrophages and monocytes in the same manner as SARS-CoV, by abusing antibodies, and turning the cells into immune-system destroying machines. It screws up dendritic cells and T cells for fun. OAS would enhance the ability of SARS-CoV-2 to ride antibodies into monocytes.
The antibodies produced by the body several months after vaccination largely originate from memory B cells, which are long-lived, but less so than Tfh cells in SARS-CoV. Memory B cells undergo somatic hypermutation in germinal centers where they are then challenged against cognate antigens presented by Tfh cells, which are also long-lived. The memory B cells deemed most efficacious are released from the germinal centers to produce antibodies primarily using plasmablasts. The virus will endeavor to make these cross-reactive, but non-neutralizing, probably mostly through antibody epitope alteration.
Long-lived CD4+ and CD8+ T cells that recognize an antigen will rapidly clonally expand in response to infection, and it has been demonstrated above that SARS-CoV-2 can abuse these cells to enhance disease. Additionally, there should be intense selective pressure to conserve particular CD4+ epitopes so that the Tfh cells in germinal centers continue to release stale memory B cells that encode antibodies for epitopes that have been escaped or defended, allowing for superior infection and ADE, consistent with observations.
The Likelihood of Significant OAS in COVID-19
The ability of SARS-CoV-2 to perform ADE and manipulate cellular immunity indicates that OAS would provide a strong evolutionary advantage to the virus.
Clinical trials are underway on a Beta-specific spike protein booster shot administered to previously vaccinated people, which will elucidate the degree to which OAS is an issue with current vaccines. The vaccines are not as immunogenic as they could be, reflected by the higher doses in Moderna's clinical trial, and they can be made polyvalent(mRNA-1273.211). This gives us some room to maneuver.
However, a booster shot or a modified version could simply mostly trigger a stale immune response rather than adding protection if the immunodominant epitopes of the protein(s) included in the modified version are sufficiently homologous. There is considerable cross-reactivity between coronavirus antibodies, and exposure to other betacoronaviruses has been found to reduce the ability of the body to generate a novel immune response to SARS-CoV-2, indicating some degree of OAS. The same is true of viruses as distantly related as influenza.
Reinfections tend to be much more severe than initial infections, which may reflect lingering pathology from the initial infection, OAS, or both. Reinfection by Delta was much more severe than initial infection by Alpha in a fully vaccinated health care worker.
Thus, I expect deleterious original antigenic sin will come to haunt us, and soon. Very few CD4+ and CD8+ epitopes have been escaped by SARS-CoV-2 variants to date, indicating conservational selective pressure is in action, and it's just a matter of time for enough antigenic drift to occur.
This should take far less than a year and a half, which is roughly the time it took B.1.617.2 and B.1.621, whose sera appear to very poorly cross-neutralize, to emerge from a landscape of no genetic diversity and fewer cases. Antigenic drift should proceed much more quickly now and escape the vaccines faster, as there is less distance that needs to be created and more mutations to do it with.
So What Does This Mean for Vaccination?
The conclusion that I draw from these findings is that vaccination is protective in the short term, but it may have profound detrimental effects in the medium term. We may be protecting ourselves from an immediate threat but exposing ourselves to severe future risk.
Vaccination has temporarily slowed viral evolution modestly. The above processes would have slowly occurred with natural infection alone, but much more slowly and with superior resulting immunity in the population. We are preventing morbidity today, which is wonderful, but at the probable expense of more morbidity later, which is not.
Vaccination also may be impossible to continue indefinitely even if OAS is not an issue because BNT162b2 also reprograms the innate immune system, reducing its ability to respond to viruses and bacteria. It is unclear whether this is cumulative, but experience with BCG would suggest it is.
While global Reff is currently below 1, we have clearly already lost the race to produce, cap, fill, and distribute vaccines faster than the virus can evolve around them, in the process losing valuable epitopes for treatments. Vaccination isn't a good idea and treatments and NPI's should have been the foundation of our defense.
Comments
Post a Comment