In February 2020, Coutard and colleagues published a short and careful observation. The spike protein of the new coronavirus had a furin-like cleavage site at the S1/S2 junction, and coronaviruses of the same clade did not. That sentence was scoped precisely. Everything that happened next involved widening it.
Two arguments grew out of that observation. They usually appear in the same paragraph, they are usually treated as one point, and they have almost nothing in common.
The first is a design inference. It says the site itself carries marks of human hands. Look at the codons, the reading frame, the restriction enzyme site, the choice of sequence. This is a claim about a specific stretch of RNA, and it can be argued to a conclusion.
The second is an absence claim. It says no related coronavirus has a furin cleavage site, so there was nothing for SARS-CoV-2 to acquire one from. This is not a claim about RNA. It is a claim about a database. Its truth depends entirely on how many bats have been swabbed.
Keeping them apart is the whole exercise.
| Design inference | Absence claim | |
|---|---|---|
| Says | The site looks made | There was nothing to inherit it from |
| Is a claim about | A sequence | A sample |
| New data | Can settle it either way | Can only weaken it |
| Failure mode | Being wrong | Staying true while meaning less |
Double CGG. Quay and Muller in the Wall Street Journal, June 2021: the two arginines are encoded CGG-CGG, a combination they said had never been found naturally in this class of coronavirus, and one a laboratory would plausibly choose. Rebuttal: CGG-CGG does occur elsewhere in the family, including MERS isolates, and codon choice in a synthesized insert is a design decision that cuts both ways rather than a signature.
The out-of-frame insertion. Segreto and Deigin argued the 12 nucleotide insert sits oddly relative to the reading frame. The frame is in fact preserved, since the insert is a multiple of three and splits a serine codon. The argument survives only as a claim about which alignment you prefer.
The FauI restriction site. Argued as a convenient screening handle for a molecular biologist checking whether an unstable insert is still present. Tyshkovskiy and Panchin ran the arithmetic across hundreds of enzymes and found roughly a 99.5% chance that any given 12 nucleotide stretch contains some restriction site. A near-certainty is not a fingerprint.
The site is optimal for furin. This one runs backward. RRAR is a functional site but not a maximally efficient one, and a designer optimizing for cleavage had better options available. The argument from design has to explain why the designer stopped short.
Where it standsNone of these has been shown impossible. Each has been shown unremarkable, which is a different and more corrosive outcome. An argument that depends on a feature being surprising does not survive the feature becoming ordinary.
Stated tightly, as Coutard stated it, the claim has not been falsified. No sarbecovirus has been shown to carry a validated polybasic furin cleavage site at S1/S2. Everything found since sits outside that boundary in one way or another.
And the boundary keeps moving inwardEach new find has required the claim to be restated more narrowly to stay true. That is not refutation. It is erosion, and it is harder to notice.
| When | What turned up | What it cost the claim |
|---|---|---|
| 2020 | RmYN02, a Yunnan bat virus reported with an insertion at the same junction | Contested from both sides. Deigin and Segreto argued it is a deletion, not an insertion. Chan and Zhan, writing from the lab-origin side, agreed the alignment is one of many possible. No cost, but it put alignment ambiguity on the table. |
| 2020 | Wu and Zhao map furin motifs across the coronavirus spike phylogeny | Establishes independent multiple origins family-wide. Forces the claim to narrow from "coronaviruses" to "SARS-related coronaviruses." |
| 2021 | Khosta-1, a Russian sarbecovirus, flagged with a predicted site | The first sarbecovirus hit. It is a prediction, not an experiment, so the claim narrows to "validated." |
| 2022 | Sander and colleagues, European horseshoe bat sarbecoviruses | Purine-rich sequence and RNA structure at S1/S2 resembling what lets avian influenza acquire polybasic sites, and two viruses one nucleotide from a furin site. The mechanism now has a described route in the right host. |
| 2023 | Bat CoV CD35, Hainan, polybasic furin-like site, basal to the SARS lineage | Polybasic and close to the lineage, though hibecovirus rather than sarbecovirus. The claim narrows to the subgenus. |
| 2026 | BRZ batCoV, Brazil, experimentally validated functional site | Validated at last, but in a proposed new subgenus, from the wrong hemisphere, and the motif is not polybasic. The claim survives on three qualifiers at once. |
Six years ago the sentence was "no related coronavirus has one." Today the sentence that remains true is "no sarbecovirus has been shown experimentally to carry a polybasic furin cleavage site at the S1/S2 junction." Both are accurate. Only the first one persuades anybody.
Every one of those finds gets reported as support for natural origin. It is not, quite. Demonstrating that furin cleavage sites arise on their own in bats establishes that the feature is available in nature. It does not establish that this particular one arrived that way. A thing can be common and still, on a given occasion, have been put there.
So the scoring is asymmetric in an uncomfortable way. The finds retire an argument for engineering without supplying an argument against it. What they take away from one side they do not hand to the other. They just make the sequence less informative than everybody wanted it to be.
Which is the honest position, and almost nobody occupies it, because it requires saying that the single most examined stretch of RNA in modern history turned out not to be evidence of much.
An absence claim is only as strong as the search behind it, and the search behind this one covers a few thousand bats out of an order with roughly 1,400 species. The claim has never been refuted. It has been narrowed six times in six years, and each narrowing left it true and smaller.
Watch for the move in other arguments. A sentence that quietly acquires qualifiers while keeping its original rhetorical weight is the most durable kind of wrong, because at no point can anyone say it was false.
Sources
The original observation: Coutard B, Valle C, de Lamballerie X, Canard B, Seidah NG, Decroly E. The spike glycoprotein of the new coronavirus 2019-nCoV contains a furin-like cleavage site absent in CoV of the same clade. Antiviral Research 176:104742 (2020).
Design inference, primary: Segreto R, Deigin Y. The genetic structure of SARS-CoV-2 does not rule out a laboratory origin. BioEssays 43(3):e2000240 (2020). doi:10.1002/bies.202000240. Quay S, Muller R. The Science Suggests a Wuhan Lab Leak. Wall Street Journal opinion, 6 June 2021.
The rebuttals: Tyshkovskiy A, Panchin AY. BioEssays 43(5):e2000325 (2021) and 43(12):e2100194 (2021). doi:10.1002/bies.202100194.
Family-wide distribution: Wu Y, Zhao S. Furin cleavage sites naturally occur in coronaviruses. Stem Cell Research 50:102115 (2020). doi:10.1016/j.scr.2020.102115.
RmYN02, both sides: Zhou H, Chen X, Hu T, et al. A Novel Bat Coronavirus Closely Related to SARS-CoV-2 Contains Natural Insertions at the S1/S2 Cleavage Site of the Spike Protein. Current Biology 30 (2020). Deigin Y, Segreto R. arXiv 2012.00627 (2020), arguing the insertion is a 6 nucleotide deletion.
The sampling concession: Chan YA, Zhan SH. The Emergence of the Spike Furin Cleavage Site in SARS-CoV-2. Mol Biol Evol 39(1):msab327 (2022). doi:10.1093/molbev/msab327.
European bat sarbecoviruses: Sander AL, Moreira-Soto A, Yordanov S, et al. Communications Biology 5:491 (2022). doi:10.1038/s42003-022-03421-w.
Bat CoV CD35: Zhu W, Huang Y, Gong J, et al. Virologica Sinica 38:344-350 (2023). doi:10.1016/j.virs.2023.04.009.
The Brazilian virus: Takada K, Yamahoki N, Mifsud JCO, et al. Nature Communications, 5 September 2026. doi:10.1038/s41467-026-77599-6. Reviewed separately at wair.ajwein.com/bat-coronavirus-furin-2026-09-08-v1.