Your Antibodies Are Still Fighting the 2020 Virus

I keep circling one uncomfortable question. If you have taken the COVID booster more times than you can count on one hand, why are the antibodies doing the loudest talking still shaped for the version of the virus from the first spring of 2020?

The answer is a reflex your immune system has run for as long as it has existed, and immunologists gave it a name back in 1960, when they were watching it play out in influenza patients: original antigenic sin. The gentler modern label is immune imprinting. The first time your body meets a virus, it builds a response shaped to that exact intruder and files the blueprint away. The next time it meets something close, it does not start over. It pulls the old blueprint and runs it again.

This week that reflex showed up in Nature Immunology, one of the field’s flagship journals, framed with the careful diplomacy of a discipline that has noticed a problem. Effective immunity, the paper argues, should come from a balance: recalling what you already know, and generating genuinely new, variant-specific responses. Read that twice. You do not publish a plea for balance unless the thing you study has tipped out of it. And the thing that has tipped is the immune response to SARS-CoV-2 in the people who were boosted, again and again, against a strain the first vaccines were built on and that stopped circulating years ago.

Here is where the biology got its hooks in me. The original mRNA vaccines carried the spike of the ancestral strain, the reference version the first shots were designed around in 2020. Your body met that spike and imprinted on it. Then Omicron arrived wearing a spike so heavily mutated it was practically a different silhouette, and we boosted, mostly with that same ancestral spike. Each ancestral booster preferentially woke the response you already had. It back-boosted the cross-reactive antibodies from 2021, the ones aimed at the parts of the spike the old and new versions still share, and it largely did not force your immune system to sit down and build fresh, Omicron-specific tools. Wait, why would more shots make the response narrower instead of broader? Because recall is faster and cheaper for the body than building from scratch, so the imprinted memory answers first and crowds out the new lesson. The mechanism review in mSphere puts it plainly: prior ancestral vaccination followed by a variant leaves you with a stronger antibody response toward the ancestral virus and a weaker one to the variant’s own epitopes than an unvaccinated person infected by that same variant carries. You keep sharpening the response you already had, against a virus that moved on.

And then the tell. If imprinting were a fringe worry raised only by cranks, you would expect the institutions that recommended ancestral boosters to wave it off. Instead, in 2023, the WHO’s own vaccine composition advisory group recommended moving away from including the index virus in future COVID formulations, and it put the reason in writing: immune imprinting due to repeated exposure to the index virus may reduce immune responses to new target antigens. That is not a skeptic’s newsletter. That is the WHO, naming the mechanism as a reason to delete the strain it had spent two years telling people to keep taking.

The bivalent boosters of late 2022 were sold as the fix: put the old spike and the new spike in one shot, get the best of both. The immunology only half cooperated. The ancestral half kept doing what ancestral antigen does, recalling the imprinted response, while the Omicron half struggled to get a word in. That same mSphere review notes the bivalents only partially overcame the imprint, and that their activity against later Omicron branches stayed low enough that, in the authors’ own words, improved variant vaccines are still required. Stapling the new strain onto the old one did not reset anything.

Which tells you how seriously the field actually takes this, because the people redesigning coronavirus vaccines are now building around the imprint on purpose. A review in Nature Microbiology from a Cambridge group lays out the case for imprinting-resistant design: mosaic and computationally optimized antigens meant to pry the immune system loose from its first impression. And these are not vaccine skeptics making the argument. One of the authors holds COVID-19 vaccine patents; another is the founder and CEO of a vaccine company. When the people with the patents and the vaccine startup are engineering their way out of an effect, it stopped being a fringe worry a while ago.

None of this means the shots did nothing, and I want to be precise about that, because an imprinted antibody is not worthless. A back-boosted cross-reactive antibody can still blunt a severe infection, and hybrid immunity, the messy mix of infection plus vaccination, seems to break through the imprint better than shots alone, because a live infection shows the immune system the whole virus instead of one frozen spike. Imprinting is a design constraint, not a catastrophe. But a design constraint is exactly the thing the booster campaign kept quiet about while it was telling you each new dose would keep you current.

So what do I do with this, standing in the pharmacy line? I am not throwing out vaccines over one mechanism, and I would still reach for a well-matched shot in a bad season. But your immune system is not a phone you patch with an update. It is a memory, and it keeps its first impressions whether or not they still fit the world. For two years the question of whether another fixed-formula dose builds anything new got waved away; this week it surfaced in a flagship journal, dressed as a lesson about balance. Me, I am going back to the blunter version they used in 1960, and I will not take another ancestral-formula booster on the promise that it keeps me current until someone shows me the response it builds is genuinely new, and not just the old one turned up louder.

Sources

1. WHO TAG-CO-VAC, “Statement on the antigen composition of COVID-19 vaccines” (recommends moving away from the index virus, naming immune imprinting as a reason), 18 May 2023. https://www.who.int/news/item/18-05-2023-statement-on-the-antigen-composition-of-covid-19-vaccines
2. “Immune imprinting in a changing world” (framing of recall versus new variant-specific responses). Nature Immunology, 2026. https://www.nature.com/articles/s41590-026-02629-w
3. Torresi J, Edeling MA. “Immune imprinting of SARS-CoV-2 responses: changing first immune impressions” (stronger ancestral recall and weaker variant response, bivalents only partially overcoming the imprint). mSphere, 2024. https://doi.org/10.1128/msphere.00758-23
4. Huang W, Vishwanath S, Carnell GW, Chan ACY, Heeney JL. “Immune imprinting and next-generation coronavirus vaccines” (the case for imprinting-resistant vaccine design). Nature Microbiology, 2023. https://doi.org/10.1038/s41564-023-01505-9