Thread by @alexbolze

Our latest study is live on @medrxivpreprint


We asked why some (unvaccinated) individuals remain asymptomatic after SARS-CoV-2 infection, using 3 large cohorts.

And learned a lot on HLA & role in chronic phase of infection vs. acute phase

Check 🧵medrxiv.org/content/10.110…

Not everyone responds the same to SARS-CoV-2 infection, even before vaccination.

One of the clearest example is the UK SARS-CoV-2 Human Challenge Characterisation study.

Only 1 of 18 infected was asymptomatic. Severity of infection is like a quantitative trait.

So why do some get develop a life-threatening pneumonia while others remain asymptomatic?

Older age was the 1st identified risk factor.

Based on genetic studies of other infectious diseases, we and others hypothesized that genetic variants may play a role incl. IEIs & HLA.

In a remarkable way, the human genetic field quickly identified many genetic associations with severe and critical COVID-19.

These discoveries point first to the innate immune system and type I IFN production & signaling.

1 review on the topic: annualreviews.org/doi/abs/10.114…

But it did not explain why ~5% of individuals (adults) unvaccinated still remain asymptomatic after SARS-CoV-2 infection.

This is an important question. Identifying genetic variants may help identify drug targets.

In July 2023, a study showed that a common HLA allele, HLA-B*15:01 was associated with asymptomatic SARS-CoV-2 infection.



Mechanism ?
Prior infections with seasonal coronaviruses (common cold) may protect those with HLA_B*15:01 against SARS-CoV-2nature.com/articles/s4158…

It would make sense immunologically if seasonal coronavirus-specific and HLA-restricted T cells cross react with SARS-CoV-2 antigens presented by the same molecules.

We therefore tested the hypothesis that HLA alleles are associated with silent SARS-CoV-2 infection

We studied 3 diverse cohorts:
- A prospective cohort from USA where participants self-reported symptoms. @my_helix & Healthy Nevada Project
- A large international consortium where physicians refer patients
- the UK SARS-CoV-2 Human Challenge Studycovidhge.com

Study design

The hardest part was how to define asymptomatic, especially for the US prospective cohort. See next tweet.

Overall we did 3 HLA-WAS for each cohort. And then 3 meta-analyses.

Asymptomatic should be: no symptoms at all. Easy.

Issue is N.
So we considered 2 other definitions:
- allowing 1 symptom for 1 day.
- allowed symptoms that lasted ≤2 days (similar to what they did in the B*1501 paper, see their Figure S10).

In the US prospective cohort, we used HIBAG to call HLA alleles in 7 genes.
We tested 105 alleles for association, those present in at least 1% of our cohort.

There were no association between any HLA allele and asymptomatic SARS-CoV-2 (using any of the 3 definitions) in the US prospective cohort.

Similarly, there was no association between any HLA allele and asymptomatic SARS-CoV-2 in the COVID HGE cohort.

This analysis was led by Astrid Marchal and @AurelieCobat

Doing the meta-analysis did not identify any statistically significant associations either.

This includes B*15:01. We did not find any association, and no enrichment at all, of B*15:01 allele in the asymptomatic groups in any of the 2 cohorts.

I was surprised by this result because the US prospective cohort in our study is similar to the cohort in the @Nature study:
- from US
- volunteers
- self-reported data (same symptoms assessed)
- more women

and we had sample size and power to replicate their results in theory

To be sure, we collaborated with @SanchoShimizu & @ChrisChiuLab to look in the UK Human SARS-CoV-2 challenge study.

3 of the 18 infected had a B*15:01 allele. All had symptoms. AND there is evidence that all were previously infected with OC43-CoV and HKU1-CoV.

Why this difference between our studies?

I am not sure. It might be linked to population stratification. HLA-B*15:01 is a common allele, and its frequencies varies a lot between populations in the US, but also within Europe.
=> important to really use genetic PCs in the test

Going back to HLA alleles and infectious diseases.

HLA is the main (or 1 of the main) association in studies that looked at chronic infections.

But it is essential for the acute phase. Example: there were no HLA associations in the first large GWAS for severe COVID-19.

I hope our work helps understand and clarify the role of HLA and memory T-cell immunity in the acute phase of a new infectious disease.

This was a huge team effort with many large consortiums involved. See all authors 👇 and link to the preprint: medrxiv.org/content/10.110…

Special thanks to Astrid Marchal for doing most of the work.
To @ETCirulli @kellyschiabor @my_helix @grzymski & Iva
To @AurelieCobat, @LaurentAbel4, @casanova_lab
To @SanchoShimizu @ChrisChiuLab
To @jacquesfellay (for teaching us everything on HLA and infectious diseases)