On the road towards complete antibody escape

The short version of the story is: We’re still seeing the SARS2 virus evolve towards complete escape from the neutralizing antibody response.

The immune landscape is still governed by the mass vaccination experiment. For this we can look at this study.

Here are the relevant excerpts:

…head-to-head comparison also showed that mRNA vaccine recipients retained higher IGHV3-53/66 antibody abundance and thus stronger imprinting than inactivated vaccine recipients. Consequently, compared to NB.1.8.1, XFG exhibits greater immune evasion in mRNA-vaccinated individuals but not in inactivated vaccine recipients. This explains the regional predominance of XFG in mRNA-vaccinated populations, while NB.1.8.1 prevails in inactivated vaccine-dominated countries.

[…]

Recipients of mRNA vaccines, primarily in Western countries, exhibit pronounced immune imprinting, persistently recalling ancestral Wuhan-Hu-1 (Wuhan) spike-reactive antibodies even upon repeated Omicron exposures, significantly limiting the development of Omicron-specific neutralizing responses. In contrast, individuals only receiving inactivated vaccines, predominantly in China, demonstrate greater adaptability in humoral responses, developing robust Omicron-specific antibodies upon repeated
infections or boosters, thus overriding Wuhan immune imprinting.

So the whole Western population is still stuck with an immune response that struggles to neutralize Omicron variants, persistently recalling the wrong antibodies, even after multiple Omicron exposures.

There’s a small subset of antibodies, that are continually being boosted by Omicron exposure, that prohibit new Omicron-directed antibodies from forming in the mRNA vaccinated:

Enabled by the IGHV3-53/66 germline genes, Wuhan priming induces a high-frequency pool of these potent “masking” antibodies, the abundance of which scales directly with the intensity of the priming.
Critically, a significant portion of this pool is not escaped by Omicron and remains cross-reactive, making it available for recall upon subsequent Omicron exposure. Therefore, during Omicron vaccination, this pre-existing and cross-reactive memory B cell population is preferentially reactivated and expanded. This potent, recalled response actively outcompetes and suppresses the de novo activation of B cells targeting novel, Omicron-specific epitopes by masking, providing a direct mechanistic basis for the strong immune imprinting observed in humanized mice.

And it matters, because the sort of immune pressure caused by vaccination is simply very different from that caused by natural immunity. A study was recently released that looks at antibody concentrations in the vaccinated in the 4th quarter of 2022, long after antibodies have had a chance to wane. They found that whereas the unvaccinated had a concentration of 75 binding antibody units, the hybrid immune and vaccinated had concentrations of 2550 and 1950 units respectively. So you´re looking at a rough 30-fold difference in antibody concentrations.

Vaccinated populations are thus exerting immense pressure on this virus to escape the antibody response.

We know it’s a dead-end strategy because that’s what the studies now say. Broad T-cell immunity now compensates for failing antibody neutralization. But the T cells have just one way to deal with a viral infection in the absence of antibody neutralization: The CD8+ T cells can kill infected cells. They’re not very good at this either, they often kill innocent neighboring cells instead, with lung injury as a result.

But this is really important.

Over here you see the antibodies that still work against BA.3.2. All the way to the right, you see Class 5. Class 5 is very unique, these antibodies target a region that normally doesn’t mutate. It’s conserved in 99.92% of SARS2 virus isolates and also works against other SARS-type viruses.

But not anymore. It looks like BA.3.2 is the first variant to go global that’s not affected by the class 5 antibodies. There are five classes of regions in the Receptor Binding Domain targeted by the antibodies. Some have proven to be easy to develop resistance against, while others are “conserved”, like class 5.

This means that after a while, the immune system begins to focus largely on class 5, as everything else just keeps changing. After multiple vaccines and Omicron infections, they found in India that most neutralizing antibodies in highly cross-reacting serum were against class 5. Note, neutralizing antibodies are antibodies that prevent the Spike protein from doing its job of binding to the ACE2 receptor.

This doesn’t mean Omicron BA.3.2 is now completely antibody resistant. There are two regions where neutralizing antibodies bind, the receptor binding domain and the N-Terminal Domain. Most of the antibodies against the N-Terminal Domain target a single site that has been rearranged already in BA.3.2:

So, the hard stuff is out of the way. Antibodies against the N-Terminal Domain should no longer work, as the N3 loop has been largely deleted. Note by the way how I predicted this was going to happen. I wrote: “you can now expect deletions to emerge, around these regions: 14-26 (N1), 141-156 (N3) and 246-260 (N5).” You don’t need a crystal ball, you just need to assume that this virus steadily continues evolving until there’s no place left where neutralizing antibodies can bind. Class 5 antibodies are another example of hard stuff, these antibodies worked against every variant, until now.

But now it looks like the newest incarnation of the virus is left with neutralizing antibodies against regions of the Spike protein that have already been frequently mutated in previous variants and are thus “easy to change”, like the Class 1 antibodies.

We’ve seen those changes emerge before, within a few months new variants of the Spike protein tend to be covered with glycans, which would presumably force the immune system to focus on this conserved class 5 site.

You can find in the first study I mentioned that most antibodies in mRNA vaccinated populations are class 1 antibodies, resulting in the spread of the XFG variant. I quote:

Additionally, neutralization titres against BA.3.2 were significantly higher in the mRNA cohort (Fig.
119 1e). This variant carries unique mutations that have been shown to evade Class 1/4 antibodies prevalent in inactivated vaccinees but not the dominant Class 1 antibodies induced by mRNA
vaccines.

Note how this fits what we see in the other study: XFG, the mRNA vaccine favored variant, evolved to avoid the class 1 antibodies. That’s stuff that should now be easy to change in BA.3.2, as XFG already accomplished it. The hard thing to do, was avoiding class 5 (which has not been seen before until BA.3.2) and the NTD supersite.

But now it seems that you can expect BA.3.2 to be targeted at regions it can easily change or cover with glycans. This means we’re approaching the fireworks moment, the point where the virus has changed enough to avoid all neutralizing antibodies in at least some portion of the population.

You also have to keep in mind that not everyone is capable of making every type of antibody, due to exposure history and genetic diversity in the population. Class 1/4 for example is generally not produced by mRNA vaccinated people.

For what it’s worth, some of you will have been wondering what I meant with the “surprise” waiting for us in Australia. Well, have a look at this, it looks like BA.3.2 already has increased virulence compared to the predecessors.

It’s immediately obvious something’s off. You see far more BA.3.2 clinical samples than you see wastewater samples. Unless for whatever reason people just are not shedding this variant in their poop, this means it’s more virulent (sick-making).

Things in nature evolve and they take their time to do so. But unless you can find some other globally dominant variant that escapes the class 5 antibodies, you can’t deny that this virus is making progress towards complete antibody escape.

You’re not looking at cyclical evolution. You’re looking at progress. To illustrate that, have a look at the RBD immune escape score:

The RBD is growing more immune evasive with time. It makes steady progress. This is entirely opposite from what you would expect to see if the situation had stabilized.

In conclusion, it’s hard to avoid the impression that the virus is still moving towards complete escape of the neutralizing antibody response. The deletions in the N1, N3 and N5 loops of the BA.3.2 NTD and the new evasion of class 5 antibodies by BA.3.2 point in this direction.

So, to summarize, what have you learned today?

-Most of the population is still stuck with abnormally high antibody concentrations against Spike.

-Those antibodies no longer work very well due to the steady increase in antibody evasion of newer variants, now forcing the immune system to rely on T cell immunity as compensation, which works by killing infected cells and thereby causes lung injury.

-Antibodies that neutralized through the N-Terminal Domain no longer work, because the relevant regions have incurred deletions.

-Class 5 antibodies, which neutralized every variant and became responsible for most neutralization, no longer appear to work either.

-The immune system now relies on antibodies against regions of the receptor binding domain that are much easier to change.

The future

It seems the next step in the progress towards complete antibody evasion is already visible, as this is interesting:

RE.2.2 >> T17007C >> C10615T, C16134T, S:P479S (C22997T), A27195G
Query: T17007C, C5884T, A10864G
Samples: 10 France (3 Regions), Scotland, Spain
Query for further S:P479S : C10615T, C16134T, A27195G
Samples: 1 France (ARA)

This mutation puts a new glycan in the receptor binding domain. We know it’s a fit mutation, because it has already been spotted in 10 samples in 3 different countries and 3 different regions of France.