Lightnews — Scholar-powered news

LightNews

Ryan Hisner

@ryanhisner.bsky.social

8.3K followers 360 following 810 posts

Teacher. Learner. Investigating mysteries of SARS-CoV-2 evolution. LongDesertTrain on another platform.

Posts Media Videos Starter Packs

Ryan Hisner @ryanhisner.bsky.social · 13h

BA.3.2 originally added several new glycans, so perhaps removal of the N122 glycan (as well as the posited T678 O-linked glycan with T678I) is something of a counterbalancing act.

Hard to know. Could also just be a dead-end singlet, albeit an fascinating one. 3/3
bsky.app/profile/ryan...

Ryan Hisner @ryanhisner.bsky.social · Mar 12

The new BA.3 has added 3-4 new glycans in spike at: N101 (via I101T), N185 (K187T), N354 (K356T—in 1/3 sequences), & N529 (K529N). Plus it lacks T19I, so it retains the N17 glycan that we’ve not seen since BA.1 (& which Delta also lacked due to T19R).
4/13
threadreaderapp.com/thread/17470...

Ryan Hisner @ryanhisner.bsky.social · 13h

It's not clear what effect this glycan removal would have.

BA.3.2 deleted large chunks of its NTD (1st part of spike, from ~S:1-305) & eliminated the C15-C136 disulfide bond, so its NTD environment is novel.

Old thread on the subject of NTD deletions below
threadreaderapp.com/thread/15690...
2/3

Thread by @LongDesertTrain on Thread Reader App

@LongDesertTrain: BA.2, driven to near-zero levels by BA.5, still haunt us, spawning monstrous viruses that, after vanishing for months, burst forth, gnarled & hideous, in novel antibody armor. The la...

threadreaderapp.com

Ryan Hisner @ryanhisner.bsky.social · 13h

I realized later today that I'd given the S:T124I mutation in this new sequence short shrift.

T->I amino acid changes may be the most common type, yet it's very rare (<0.01% of sequences). Why?

Because it removes a conserved glycan from N122! (Which is also conserved in SARS-CoV-1). 1/3

Eight of the 22 sites were determined to be occupied predominantly with high-mannose-type structures by Watanabe et al. (N61, N122, N234, N603, N709, N717, N801, and N1074).29 However, Shajahan et al., Walls et al., and Zhang et al. noted an abundance of complex-type glycans (Figure 1.1).30,48 Specifically, the preserved glycosites between SARS-CoV-1 and SARS-CoV-2 (N61, N122, N603, N709, N801, and N1074), which had previously been characterized as containing only high-mannose-type glycans, have since been identified as containing complex-type glycans.24,27,29,30 Sites N165, N331, and N343 express complex-type N-glycans containing both bi- and tri-antennary
structures.

Ryan Hisner @ryanhisner.bsky.social · 22h

In circulating lineages, I think so, though you'd have to ask someone like @benjmurrell.bsky.social or Trevor Bedford to know for sure.

We haven't had any saltation variants take over in a long time, which is the biggest difference between the past ~2 years and the first ~4 years of the pandemic.

Ryan Hisner @ryanhisner.bsky.social · 22h

BA.3.2 continues to surprise. It's foothold in both Australia and South Africa is solid, so it's not going to disappear anytime soon. It's also not growing especially quickly or spreading rapidly internationally.

But given enough time to explore, it may do both before long. We'll see. 6/6

Ryan Hisner @ryanhisner.bsky.social · 22h

In addition to the spike mutations, this sequence also has ORF3a:A110V, ORF9b:A11V, ORF1b:P1727S, & ORF1b:G2662D.

One paper found that ORF1b:P1727L, which is in the proofreading NSP14 protein (NSP14:P203L), doubled mutation rates in hamsters. Would P203S (ORF1b:P1727S) have the same effect? 5/6

Ryan Hisner @ryanhisner.bsky.social · 22h

There's been some mixed results on this, but the FCS-adjacent T678 is thought by many to be O-glycosylated, i.e. to have a sugary molecule attached to it. T678I would eliminate that glycan, though what effect it would have is uncertain (as far as I know).

This is a busy region in BA.3.2. 4/6

Ryan Hisner @ryanhisner.bsky.social · 22h

@siamosolocani.bsky.social can correct me if I've forgotten any recent 478 mutations. It's apparently a very flexible residue, & this diversifying selection suggests it's on the run from antibodies.

S:339 might be the only site we've seen more variety in, but not quite at this same scale. 3/6

Ryan Hisner @ryanhisner.bsky.social · 22h

The spike mutations are T124I, N478T, & T678I.

N478T is a reversion to the ancestral AA, meaning it's gone from T-->K-->N-->T in this lineage.

There and back again.

S:478 has been by far the most active AA residue in recent months. We've seen K, T, I , E, R, N, L, M, and Q there of late. 2/6

Ryan Hisner @ryanhisner.bsky.social · 22h

There's a new BA.3.2.2 from South Africa today. For the most part, there's been little substantial change in BA.3.2 over the past few months—mostly synonymous mutations & very little happening in spike.

But this new one has 3 spike mutations & looks quite interesting. 1/6

Ryan Hisner @ryanhisner.bsky.social · 2d

Who will be blamed when it happens? I suppose the right will blame the very agencies they've defunded and destroyed.

Ryan Hisner @ryanhisner.bsky.social · 4d

If you have the nuc sequence, I wouldn't mind taking a crack at it. Alignment programs can make weird decisions. The Nextclade interpretation of the S:144-147 region of a recent chronic LB.1.3.3 was an "undeletion" at 144 + four nuc substitutions. Really it's just an AAC insertion—RdRp-stutter-type.

Ryan Hisner @ryanhisner.bsky.social · 5d

Definitely not a BA.3. I can't tell why it would be categorized that way. Looks like a humdrum low-coverage NB.1.8.1.

Ryan Hisner @ryanhisner.bsky.social · 5d

On its face, this seems hard to believe. The sky is big. Contrails are small. But 3% of flights contribute to nearly 2% of radiative forcing? Might be one of those things that's totally counterintuitive I suppose.

"[Contrails] contribute roughly 2% to the world’s effective radiative forcing."

Ryan Hisner @ryanhisner.bsky.social · 5d

Fascinating!

An R682L that showed up yesterday, but mutations that truly erase the FCS are very rare. R682X, R685X, or FCS deletion almost always turn out to be Vero-passaged samples.

@solidevidence.bsky.social can give more detail, but FCS is often badly ablated in Cryptic WW variants.

Ryan Hisner @ryanhisner.bsky.social · 6d

One last note: I never include S:K679N in these searches because the vast majority are artifactual reversions to wild-type. However, there's one workaround: G23599T is the reversion to WT, but G23599C also causes K679N. It was ~nonexistent before BA.2.86 & has ramped up dramatically recently.

Ryan Hisner @ryanhisner.bsky.social · 6d

Oh, god, please no. I hadn't seen this at all. Most of these recombinants are between extremely similar variants. The exception is XDV/NB.1.8.1, but that one's clearly on a slow, terminal decline. Frankenstein, ugh.

Reposted by Ryan Hisner

Chaoran Chen @chaoranchen.de · Aug 28

Finally, CoV-Spectrum is meeting trees 🌳 – this week, we released a new feature to browse the SARS-CoV-2 UShER tree on CoV-Spectrum. With the new feature, you can see a subtree of your variant of interest. 1/4

Ryan Hisner @ryanhisner.bsky.social · 6d

...as Delta was the most unstable and fusogenic variant of all, so mutations increasing stability and taming its fusion activity would likely be more tolerated than in a less fusogenic spike background.

Ryan Hisner @ryanhisner.bsky.social · 6d

All of these mutations were virtually absent prior to BA.2.86 (without ∆S31). The only very partial exception was in the Delta era, when there was a very small but very distinct increase in these mutation. I think this makes sense...

Ryan Hisner @ryanhisner.bsky.social · 6d

...I think the circumstantial is very convincing. Bulky or inflexible residues (F/Y/P) in & near the FCS should inhibit cleavage (S680F/Y/P). These are also highly correlated with R683Q/W/L mutations and to K679M/N, both of which are appearing far more frequently than ever before....

Ryan Hisner @ryanhisner.bsky.social · 6d

It's a very recent thing, so there's nothing in the literature about it. I should also say that this weakening isn't 100% certain since no one has actually tested these mutations in the lab. However...

Ryan Hisner @ryanhisner.bsky.social · 8d

Pretty rich for a guy trying to ram through extremely unpopular "reforms" against the will of the population to lecture others about democracy.

Ryan Hisner @ryanhisner.bsky.social · 9d

I'm still not sure if this is a T cell-evasion mut pattern (like another pattern involving 8 regions) or if there's something functional going on. The presence if K1795Q in so many of these (together with the fact that 6/7 are in NSP3) suggests to me there's a functional connection of some sort.

Ryan Hisner @ryanhisner.bsky.social · 9d

I haven't run through all of them, but at least two of these have the 7-region pattern that I've taken to calling "6-7a" (short for 6xNSP3 + ORF7a).

Here are a few of the many conventional chronic sequences with mutations in at least 6 of the 7 regions.