XBB.1.9.1 with S:F456L(T22928C) and S:Q52H (G21718C) and 2 orf muts in Spain, Canada, Puerto rico, Finland, France, Ireland, England , Japan (18 sequences)

[BorisUitham]

Query: G21718C, T22928C, C25714T Collected: 2023-04-29 and 2023-05-12

Normally I wouldn't make an issue with only 2 sequences, but since this is pretty much the xbb.1.9.1 equivalent of xbb.1.9.2's eg.5.1 which does have a large growth advantage i'm tracking it here already, since I do expect it to grow similarly. If it doesn't grow even though its very similar, I will close this issue. Spikes are identical to eg.5.1, the difference is in a lot of the nucleotides before the spike. If you compare presence of accessory mutations, the difference with eg.5.1 is that this has: Orf3a:L108F Orf1a:I2010V

And eg.5.1* has Orf1a:a690v Orf1a:a3413v Orf1b:d54n (in 36% sequences, see my other proposal)

https://nextstrain.org/fetch/genome-test.gi.ucsc.edu/trash/ct/subtreeAuspice1_genome_test_63964_dc0270.json?c=gt-S_52

In the above image: top sequences are eg.5.1, bottom two are these xbb.1.9.1 sequences.

[carlottaolivero]

I don't know if this is the right place to write it but we have noticed that S:Q52H can derive both from G21718C and G21718T. We have seen both mutations in our wastewater samples.

[FedeGueli]

Many thx @carlottaolivero for noticing that ! i think G21718T is the S:Q52H of the fast EG.5.1 lineage so likely you are seeing that in your samples

[FedeGueli]

Still no move closing this for now

[BorisUitham]

There is a third one now in finland, but not much more than this so i'll leave it closed unless theres more activity

[BorisUitham]

5 more sequences, total 8 now. Do you think its worth reopening @FedeGueli ? 1 more spain, 2 canada, 2 puerto rico.

Total: 3 spain 2 canada 2 puerto rico 1 finland

[BorisUitham]

Interesting mutation: the latest one from england has s:l455w (only 82 seqs ever). But mutations there are getting more common so who knows

[oobb45729]

There are 2 sequences with S:L455W in England now. https://nextstrain.org/fetch/genome.ucsc.edu/trash/ct/subtreeAuspice1_genome_28b37_5272a0.json?c=gt-S_455&label=id:node_6606314

[oobb45729]

I suspect that L455W might be more beneficial than L455F (with F456L). The predictions from AlphaFold: Comparing to F455, W455 forms a new hydrogen bond with F490(S490), so the structure may be more stable. The distance between W455 and Y489 seems to be closer. Plus, L455W may escape antibodies that target this part better.

The reason we've seen less L455W than L455F probably is that U->G is less likely to happen than G->U. There might be a cation-pi interaction between W455 and R493 which may be the reason we saw less L455W in Q493 variants than R493 variants (BA.1 and BA.2). However, with many circulating lineages with F456L now, L455W may become favored again.

[BorisUitham]

@oobb45729 thanks for the analysis! An increase in stability might open up the path towards more mutations.

"In contrast, the mutation L455W which reduces ACE2 affinity in Wuhan-Hu-1 is affinityenhancing in Omicron RBDs. While L455W is accessible via single-nucleotide mutation from the position TTG codon present in Wuhan-Hu-1 and Omicron, mutations to site 455 have not yet been important in SARS-CoV-2 variant evolution (but see [29]). This may be because all such mutations previously decreased ACE2 affinity. But site 455 is variable across the broader evolution of SARS-related coronaviruses, e.g., Y455 in SARS-CoV-1 and W455 in RsSHC014 (SARS-CoV-2 numbering), and the epistatic shift that makes some mutations at this site favorable for ACE2 binding in Omicron suggests it could become relevant for future SARS-CoV-2 evolution." https://www.biorxiv.org/content/10.1101/2022.09.20.508745v1.full#ref-29

Also here a description of a couple ba.1.23 long term infections last year which eventually developed W455 L456 (this was between immunocompromised patients). In this case, w455 preceded l456. https://pubmed.ncbi.nlm.nih.gov/37270625/

[jbloom]

Yes, L455W is the most ACE2-affinity favorable mutation in the BA.2 RBD, which is currently the one closest to XBB for which we have deep mutational scanning data.

See image below, taken from here

[oobb45729]

I think Q493R might the reason that L455W increases ACE2-affinity in BA.1 and BA.2.

[tylernstarr]

@oobb45729 interesting theory, and it tracks nicely with new data in BQ.1.1 and XBB.1.5 (reverts back to Q493), wehre L455W affinity-increasing effect goes away again!

We haven't put the data out yet though our repository and the data are publicly accessible at this point. We have a bit more QC to do on the data, but the values shoudln't change too dramatically and adds additional context to this interesting observation.

[oobb45729]

Thank you so much for the new data!

1. So F456L is no longer detrimental for ACE2-binding in BQ.1.1 and it increases ACE2-affinity for XBB.1.5! That explains a lot. The R493Q probably is responsible for this too. For the difference between XBB.1.5 and BQ.1.1, maybe it has something to do with F490S?
2. L513F indeed is no longer highly detrimental for XBB.1.5 and BQ.1.1, but why?
3. H505Y does not increases ACE2-affinity for XBB.1.5 and BQ.1.1 any more.
4. R403K increases ACE2-affinity moderately, so does S494P.
5. Y453F, certain Q493Xs, and certain N417Xs increases ACE2-affinity significantly as always.
6. L425R seems to be still available for XBB.1.5 to gain.
7. A484K is beneficial for BQ.1.1.
8. There is a relative big RBD-expression loss for N450D in BQ.1.1 but not in XBB.1.5. Maybe K444T is the reason.
9. 348 and 356 do not change much.
10. 478 is still mysterious.

[BorisUitham]

Indeed, thanks for the new data! I think then, without 493q that f456l would be required to open up mutations at 455 to become possible in xbb+486p lineages. You can see this with f455, green line is f455 without l456 (looks like green is increasing but thats just lowered sequencing):

Whereas in the ba.1.23 it was possible for w455 to come first and then l456 to follow suit

And indeed i wonder why then 452 mutations are so rare in xbb lineages (edit: actually i do see a small rise recently) but perhaps that is discussion for another thread

[FedeGueli]

18 samples as today yesterday uploads from ireland and England , today one caught by US airport surveillance coming from Japan https://nextstrain.org/fetch/genome-test.gi.ucsc.edu/trash/ct/subtreeAuspice1_genome_test_11073_b05ca0.json?label=id:node_6636621

[FedeGueli]

Clearly growing and spreading please propose it.

[BorisUitham]

Thank you @FedeGueli , i will

[BorisUitham]

https://github.com/cov-lineages/pango-designation/issues/2072

[FedeGueli]

Designated FL.20