Ok, just to show that I am not biased against zoonosis, here is a much more plausible NATURAL way the FCS could have arisen in SARS2 than what virologists could muster so far — via recombination with host mRNA. Namely, pangolin mRNA. In fact, I found a 14-nt match to PRRA insert
in the GIABR pangolin dataset. Yes, that’s the same dataset that contains the SARS2 RBM and was the source for the MP789 pangolin CoV. I was checking it for the PRRA insert and found 3 reads in lung08 run (SRX6893154) that match a 14-nt stretch that fully covers the PRRA insert:
The first read actually mapped to a COMPLEMENT (minus strand) of pangolin mRNA, which is perfect for coronaviral RdRp recombination, as the CoV template strand is a minus strand.
The second read I couldn't really map to anything conclusive:
The third read again mapped to a complement of pangolin methyltransferase mRNA:
Oh and by the way another pangolin enzyme matched the 14-nt fragment, DUSP4:
While we haven't yet observed CoV recombination with host mRNA, it has actually been observed for other viruses. In fact, it was observed to increase cleavability in influenza:
So could the novel furin cleavage site have arisen naturally? Of course. Recombination with host (pangolin?) mRNA is much more plausible in my book than some mythical recombination with HKU9 as per Gallagher or "clade X" as per Lytras.
Does this rule out a lab leak? Not for me.
Why not? Even if FCS in SARS2 arose naturally, it could still have emerged in Wuhan via a lab leak - the recombination in a pangolin could well have happened in a lab, or a lab staffer could have been infected with a pangolin CoV with an FCS during field work.
Plus the possibility that the SARS2 FCS has been purposefully engineered (as per the EcoHealth and WIV DARPA proposal) is still on the table.
But all hypothesis must be evaluated based only on data, not preconceived notions or groupthink (relevant for both zoo and LL groups).
• • •
Missing some Tweet in this thread? You can try to
force a refresh
Latest SARS2 preprint provides additional evidence that not only was the novel FCS a key factor in its pandemic potential, but the way the FCS was acquired — via a 4 aa insertion — was also important. Re: "why would a genetic engineer use an insertion to create an FCS?"
I've hypothesized previously that the answers to "why use an insertion" and "why use a leading proline" lie in MERS, whose S1/S2 loop is 4 AAs longer than that of SARS-like CoVs, and also has the leading proline, making it more exposed to proteases:
So the desire to try a MERS-like FCS can explain both. Moreover, in the 3D structure of the spike protein, the S1/S2 loop in SARS2 with the insertion has its cleavage point ~3x closer to the MERS one than if the FCS in SARS2 arose w/o insertion:
Increasing neuroinvasiveness could indeed have been the research goal of a genetic engineer opting to use a “suboptimal” or “non-canonical” furin cleavage site like the one found in SARS2. And Baric did say that MHV and FIPV inspired the FCS interest of DEFUSE. Thread below:
Somewhat counterintuitively, in a human coronavirus, OC43, when it has the less efficient (“suboptimal”) RRSR furin cleavage site instead of the canonical RRSRR one, it is more lethal and more *neurovirulent* when tested in a mouse model:
But in a mouse hepatitis virus (MHV, one inspiration in DEFUSE) it is actually the JHM strain with the more efficient FCS, RRARR (reminiscent of RRAR in SARS2), that is *more* neuroinvasive and more lethal than the A59 strain with a suboptimal RRAHR FCS:
Excellent analysis of SARS2 cryptic lineages from persistent human infections (those nightmare scenarios happen when the virus colonizes your gut and remains there for months or years). Some implications for Covid origins there that I’d like to address in the thread below.
First an aside on extrapolating insertions found in SARS2 circulating in humans to the likelihood of insertions occurring (and getting fixed) in a SARS2 progenitor circulating in an intermediate host — SARS2 has now been circulating for years in hundreds of millions of humans, while in the wild it could not have had even a sliver of such a reservoir of non-bat hosts (and we all agree the FCS insertion could not have arisen in bats because it is detrimental to their preferred enteric tropism).
Of course, “a long time” for passaging in a lab could be just a few months, especially if we’re talking about in vitro passaging. Moreover, many novel mutations would arise quickly when the virus is put into cells from a novel host and organ system (respiratory vs. enteric).
🧵 Covid Origins: Lab leak critics claim the coincidence of a novel coronavirus emerging near the Wuhan Institute of Virology is dwarfed by the “counter-coincidence” of it first being noticed at a wildlife market. But the market was actually one of the most likely places for a lab-leaked virus to get noticed.
Here’s a clip from my podcast with @robertwrighter and a deep dive in the thread below:
Basically, if we imagine the map of Wuhan as a “probability landscape”, where each location has an inherent probability that a roaming lab-leaked virus would get noticed if it got there, the Huanan seafood market would dominate that landscape like Mount Everest.
Now, let’s walk through the thought experiment: How would a lab-leaked virus get noticed in Wuhan in late 2019?
Some critics of the Covid lab leak hypothesis say the SARS2 furin cleavage site is unlikely to have been engineered because in previous cases of creating novel FCSes in coronaviruses virologists have never inserted an FCS but rather created them in place via point mutations. But!
A close collaborator of Zhengli Shi and the Wuhan Institute of Virology, Shibo Jiang, in 2013 published a paper creating a furin cleavage site in a non-CoV synthetic vector via a 12-nt insertion (note the CGG codon for the leading arginine):
In 2016-17, WIV has created a novel reverse genetics system ("backbone") for the WIV1 virus, and Shibo Jiang subsequently coauthored a paper with Lei-Ping Zeng, the author of that backbone. (Lanying Du, another key collaborator of both WIV and Shibo Jiang, was also on the paper. She was also the editor for the 2013 paper inserting the RIRR FCS):
🚀 More amazing partial reprogramming news! This time from @davidasinclair — he mentioned some unpublished animal results for:
- *reversing* Alzheimer’s symptoms
- hearing loss
- ALS
- glaucoma (anticipating clinical trials in 2025!)
- rejuvenating skin, kidneys and liver
Great to hear that the FDA is ok with Tet-inducible (via rtTA3) partial reprogramming gene therapies, and that Life Bio is so close to the clinic. They presented encouraging results in monkeys in an eye stroke model (NIAON); I didn’t know they also tried a glaucoma monkey model.
PS: while Life Bio seems closest to human clinical trials among partial reprogramming companies, others aren’t far behind:
- Turn Bio with their skin therapy
- we at YouthBio with our Alzheimer’s therapy