Treacherous ancestry
An extraordinary hunt for the ghosts of SARS-CoV-2
Prelude: The Ongoing Confusions
Why are US scientists dragged in front of Congress in Benghazi-style show trials? How come highly rated research programs get quietly shut down? Is it justified to treat virologists with suspicion, consider them guilty of recklessness or dishonesty, even leak their subpoenaed private conversations to professional smear artists for cherry-picking?
The origin controversy is political and polarized. Myths that COVID-19 was somehow a manmade pandemic are still impactful, whether they are true or not. Polls have shown that 2 out of 3 US citizens believe that SARS-CoV-2, the virus that started the COVID-19 pandemic, came out of a laboratory rather than nature.
Scientists worldwide vehemently disagree. The emerging scientific consensus among domain experts is that SARS-CoV-2 is a natural virus that entered humanity via zoonotic spillover (more importantly, there is a consensus stemming from the body of evidence that is entirely unequivocal).
Yet fewer and fewer citizens seem to hear or care for their evidence-based voices. Domain experts are distrusted, but how much of that is self-inflicted, and how much is sown by anti-science actors?
“We keep changing our minds about things based on data — that is science, not fraud.” — Robert Garry
We can observe scientists are increasingly bullied out of public discourse by information combatants or targeted by harassers on social media. Most simply have their knowledge drowned out by an information ecosystem that favors emotional falsehoods over complicated facts. It is fair to say that science has come under pressure.
On top of that, various conspiracy myths about a “lab leak” continue to garner attention in the media, international forums, and the halls of US Congress, all while elected representatives, senators, anti-science activists, and propagandists seek to gain popularity, profit, or power by instigating a “gain-of-function research” moral panic*.
How many concerns about GoF research are justified, and how many just make for a good story? How many officials and agencies are allegedly entangled in a cover-up? Media manipulators are skilled at fabricating false uncertainty and shaping sensationalist narratives out of whole cloth. When it comes to supporting evidence, the ‘gain-of-function’ emperor remains embarrassingly naked. But it makes for a fantastic story.
“It takes more effort to debunk a conspiracy theory than to create one. It took only one podcast to make me think the lab leak theory was possible, but it took months of research to understand why it’s not” — Peter Miller, who recently won $100.000 in an 18h origin debate against a lab leak proponent
All scientists can hope for is that eventually, citizens will get tired enough of the spectacle and make-belief to take a closer look at what we already know about the topic.
Outside the heated spotlight, multiple teams of international scientists have quietly honed in on a set of new clues to the origin of COVID-19
Clues that might not only hold the definitive answer to whether SARS-CoV-2 came from nature or the lab, but also how to best anticipate and mitigate the risk of another SARS-related coronavirus pandemic.
So where is the birthplace of SARS-CoV-2, and how exactly did this chimeric virus come about? Which trajectory did it take before the first outbreak emerged at the Huanan market in Wuhan? Will we ever find out?
Given current geopolitics, deliberate sabotage of scientific processes, and Beijing’s obfuscation of the matter, many say the true origins of COVID-19 will never be solved. Some governments explicitly root for that outcome, that we will be stuck in perpetual uncertainty forever.
I would not take that bet against scientific inquiry — and after reading this deep dive on the origin controversy — I believe neither will you.
Scientific literacy is a superpower in the (dis)information age. It enables citizens to cut through the noise, intelligently engage with a topic of global controversy, and come out wiser. Scientifically literate citizens however need relevant evidence and reasoning of scientists to be made accessible to form their opinion. On top of that, they will probably need months to familiarize themselves with the complexity of the origin discussion, and honestly, who has time for that? Having done the (sometimes painful) legwork, here is where I believe I can offer a helping hand. If you gift me your curiosity and time, this article will be a guide to understanding some of the most compelling evidence for the natural origin of this virus.
Tune out the politics, the broken media, and the bad actors. Let’s return to the process of scientific inquiry.
It is time to put the false “gain-of-function research” myth to bed once and for all, and learn about the cutting-edge science on the hunt for the true origins of SARS-CoV-2.
Buckle up, or fetch a large cup of tea, this is the origin story told through the lens of viral recombination.
High-quality extended pdf version of this article:
An Extraordinary Scientific Hunt
“If you gave me a billion dollars to find the origins, I`d probably spent 90% of that outside of China in South East Asia”
— Bat immunologist Linfa Wang, Duke-NUS Singapore
Chapter 1: Unraveling a Mysterious Chimera
I) An uncanny genome
When SARS-CoV-2 emerged, it was a confusing virus for a lot of reasons. First, it was very infectious to humans for a novel virus, spreading effectively between them via the respiratory route. Second, it did not cause severe disease in many patients, was sometimes asymptomatic, and subsequently hard to track. Third, side-by-side comparisons to known SARS-related viruses seemed to show that the novel virus was a chimera. It had parts of very high genetic similarity, and other parts of low genetic similarity, on top of a few other oddities that gave researchers initially a hard time to wrap their heads around. Even seasoned virologists were quoted wondering how “this gets accomplished in nature” in early February 2020.
As the pandemic turned into full swing, multiple “man-made” theories of varied quality were advanced on how the SARS-COV-2 — and its odd genome — came to be. From bioweapon development to gain-of-function research, from de-novo genetic engineering to the alleged introduction of HIV sequences, from serial passage through human cells or humanized mice to arcane vaccine experiments, many asserted that some type of human manipulation was necessary to explain how this dangerous patchwork virus of high and low sequence similarities to other coronaviruses came about.
Only coronavirologists with decades of experience with that particular viral sub-family would disagree, they certainly saw nothing unheard of in the genome. But the biggest problem in debunking the plethora of false notions was the lack of reference points; namely viral cousins of SARS-CoV-2. Those only gradually came trickling in once the severity of the outbreak turned into a global pandemic and jolted more and more scientists into urgent action. Rather than speculate on inconclusive data, many researchers set out to find related coronaviruses, either by discovering neglected genomes in large biomedical databases or by direct sampling of bats in nature.
They quickly realized that SARS-CoV-2-related viruses all looked a bit weird and stitched together; for example, a pangolin virus had an ACE2 receptor binding domain very closely related and able to bind and infect human cells. Another bat virus discovered in Mengla country in China had an insertion reminiscent of the furin cleavage site at the S1/S2 boundary, as well as what looked like an ancestral genome to SC2, at least for about the first 2/3rds of its full span. How could these diverse animal viruses seem so closely related to SARS-CoV-2 in one part, and so distant in another?
II) On recombination versus artificial assembly
Almost from day one, coronavirus veterans had been quick to educate their collaborators that CoV genomes might just seem unintuitive because they are shaped by a process called recombination.
Recombination is a mechanism for genetic exchange between two different parental viruses that creates a new viral genome sharing genetic information of both. It requires two (usually distinct) viruses to be present in the same host cell.
As a useful but imprecise abstraction, one might think of recombination as a form of “virus sex” that produces unique offspring sharing a mix of parental genomic regions.
Offspring that came about by recombination is in many cases unproductive, meaning it can either not fulfill all essential functions necessary for the virus to replicate and infect new hosts; or it can do so but worse than the parental lineages, thus getting quickly outcompeted by them and vanish into nothingness. A virus needs to constantly spread and adapt to persist, after all.
However, sometimes this “virus sex” brings forth recombinant offspring that is in some aspect better — meaning more fit in its current or new environment — than the parental lineages, ergo it will spread and gain ground in the host population or particular environmental niche, possibly establishing itself for years to come.
Recombination frequencies vary dramatically between viral families, from the promiscuous to the prudent. For example, recombination does not seem to play an important role in some Flaviviridae (Zika) and Filoviridae (Ebola) and Paramyxoviridae (Hendra, Nipah). At the same time, segmented viruses such as influenza use a different mechanism of genetic exchange altogether called “re-assortment”.
The coronavirus subfamily of positive-strand RNA viruses reportedly recombines frequently. Recombination in CoVs is thought to be facilitated by a molecular mechanism called RNA-dependent RNA polymerase (RdRp) template switching leading to copy-choice recombination (see figure below).
What this means is that the RNA production machinery (RdRp) can jump between instruction templates at any point if more than one is available, producing a hybrid sequence (chimeric genome).
I sketched out the mechanistic interplay between replication and viral gene transcription (see below) to illustrate how template switching is not a freak event but an essential step in the viral life cycle.
“Replication of the coronavirus genome requires continuous RNA synthesis, whereas transcription is a discontinuous process unique among RNA viruses. Transcription includes a template switch during the synthesis of subgenomic negative-strand RNAs to add a copy of the leader sequence.” — Sola I. et al., Annu Rev Virol., 2015
An RdRp that needs to jump back and forth between genetic elements constantly might easily jump to another different genome when thrown into the mix, thus providing ample mechanistic opportunity for recombination. (Where CoVs get physical opportunity to have so much virus sex and what we can learn from it we will look at in Chapter 2)
This is of course not a secret to experts in the field. That coronaviruses fuck around — sorry recombine constantly — has been scientifically established for many years before COVID-19.
For now, what is important is that the promiscuous sarbecovirus (SARS-related beta-coronavirus) sub-genus gets a lot of recombinant progeny; and over time, established offspring lineages can themselves engage in more virus sex (= undergo recombination) with other circulating viruses. From long-estranged cousins and viral strangers to incestuous siblings and even their own offspring, anything seemingly goes (hey, not our place to judge!).
Again, thinking about this promiscuous mingling is merely a useful abstraction, because in reality, each individual virus genome is just one of many millions copies in a single infected person. Out of this complexity, recombination can result in ever new chimeric offspring that — if successful — will carry parental genetic segments forth from the generations that came before it. However, in the early days with less information, SARS-CoV-2 did not seem very recombinant. But maybe something changed with more bat cousins being found since?
If SARS-CoV-2 and related sarbecoviruses indeed came about by frequent recombination, we would expect a colorful mix (a mosaic) of genome segments shared between them and their closest cousins.
Here is how that looks for SARS-CoV-2 and some of it’s closest relatives:
For SARS-CoV-2, the above figure shows that it shares regions of high sequence similarity with a bunch of different bat viruses, namely RmYN02, Banal-103, RpYN06, Banal-52, and RaTG13. Close cousins. However, for each SARS-CoV-2 to bat virus comparison individually, there might be some regions of very low sequence similarity (usually around the spike) where they are not similar at all.
The observation of this mosaic genome makes clear that SARS-CoV-2’s overall genome could not have been designed or derived from any known single progenitor virus (for example RaTG13 or a BANAL-like virus, as lab leak activists like to argue), but rather that it is a genetic chimera containing bits and pieces from multiple related viruses. We know this technically because no engineering or culturing approach could have magically created the “ancestral” sequence in hundreds of positions any time a cousin diverged from SARS-COV-2.
Given this reality slowly breaking through, lab leak activists (not prone to hold any coherent hypotheses on this issue anyway) have since moved on to fantasize that SARS-CoV-2 must then have been “stitched together” in a lab from a set of (multiple known and unknown) progenitor viruses secretly sampled by the Wuhan Institute of Virology. (“What can be asserted without evidence can also be discarded without evidence”, I would say about the never-ending epicycles of conspiratorial ideation)
In reality, all of the proposed “stitched together” speculations are of course not only contradicted by evidence and epistemically risky at face value (I classified them as fraudulent), but they also fail to account for the fact that all other SARS-CoV-2 relatives found in the wild look equally stitched together from multiple viruses.
In other words, if the SARS-CoV-2 chimeric genome is supposedly “stitched together” artificially, how come every random viral cousin found in the wild looks equally “stitched together”?
There is no question that recombination is the only scientific explanation that can account for the observation of these natural chimeras. (There are many more technical details as to why scientists know that SARS-CoV-2’s mosaic genome was not assembled and stitched together in a lab from a sequence, but you get the main point. Virus sex, baby.)
On top of that, the observed recombination patterns left behind by past “sexual” encounters between parental lineages can not be produced, simulated, or faked in a laboratory. Pause for a second and read that again.
The relevant “sexual history” that played out over the last five decades or so needed to be lived in the wild, it could only have happened in nature where all these viruses meet.
All scientific evidence and prudence suggest therefore that a naturally evolved, immediate bat ancestor to SARS-CoV-2 must have existed. It did not spring from a computer sequence, was not dreamt up by a mad scientist, or recklessly assembled from disparate parts. SARS-CoV-2’s bat ancestor came about as naturally as all the other recombinant children of parental sarbecoviruses that we have since discovered in bats.
This is what scientists mean when they say the “backbone” of SARS-CoV-2 is natural and was not stitched together.
On the larger genetic makeup, this ancestral bat version of SARS-CoV-2 already looked very much like the one that first surfaced in Wuhan. But this alone is not sufficient to exclude that such a recombinant bat ancestor was found by researchers and then “tinkered with” in the lab. Recombination has a low resolution, it can not track targeted single mutations. So genetic tinkering remains possible and plausible, right?
How else can we explain that a bat virus seems so damn good at infecting human cells?
III) Recombination patterns are not random
Another common line of erroneous argumentation by lab leak proponents concerns the idea that because SARS-CoV-2 was capable of infecting humans very well, it must have been somehow pre-adapted or optimized to do so via serial passaging in human cells, or been given this remarkable ability by thoughtful engineers. How else could a bat virus circulating in bats develop this human affinity just by chance?
For me, this argumentation from supposed “optimization” always has a bit of a fallacious character — “How can something marvelous such as the eye have come about by chance alone?“ — seen many times in creationist arguments. Evolution is just not the same as chance. We humans tend to underestimate the diversity of nature and the power of evolutionary selection at our peril.
So how do we know that nature has optimized the human affinity of SARS-CoV-2 all by itself?
I am glad you asked. CoV virologists have studied over the years how the S gene — encoding the spike protein — seems to be one of the biggest factors influencing cellular and species tropism (the ability to successfully infect different hosts). When scientists plot the frequency of recombination events along the Sarbecovirus genome, they found that there was not an even distribution, but rather “hot spots” and “cold spots” for genetic exchanges, with the spike region shining bright red.
What does this mean? Well, let us remember that recombination almost exclusively causes fucked up offspring genomes that are unable to procreate or establish themselves. All scientists ever get to observe are the survivors, those one-in-a-billion viral children who came out fitter than their parents.
A recombination event that adds to niche fitness gets to stay, and whatever impairs fitness gets selected out.
[…] non-random and mostly conserved recombination patterns that we and others have detected in various coronavirus subgenera are likely shaped both by evolutionarily conserved variations in the mechanistic predispositions of different genome regions to recombination and by shared selective processes disfavouring the survival of recombinants that express improperly folded proteins.”
The figure above basically shows that within and surrounding the spike and some accessory protein genes (orange), observing new productive recombinant segments (and with it, a higher sequence diversity) is much more likely than anywhere else along sarbecoviruses’ genomes.
So finding that SARS-CoV-2 has an “unusual” spike gene sequence compared to many of its relatives is actually not unusual at all. It is the norm among members sarbecovirus sub-genus, they all have unusually diverse sequences there. The question is why?
In general, high sequence diversity enables viruses to stumble upon new functionalities, from evading immune systems to changing their tropism (what type of cells they can infect), facilitating more efficient cellular entry, or infecting new host organisms.
However, developing and maintaining sequence diversity over a viral population is not always easy to come by.
Single point mutations arising from replication errors rarely get the job done, and face obstacles such as purifying selection.
On top of that, RNA viruses with large genomes require a strong proofreading ability to correct replication errors that give rise to deleterious mutations.
That is why for some RNA viruses, such as HIV, it is thought that sequence diversity more frequently arises through recombination than point mutation.
In coronaviruses, recombination is also assumed as an efficient mechanism for how these long, proofreading RNA viruses can rapidly create genetic diversity; presumably to cope with quickly changing niche environmental conditions.
“[…] recombination also has the potential to act as an evolutionary “fast-forward” by quickly shuffling genetic material between vastly different viruses.
For the same result to be produced by mutation alone, long spans of time would be needed for selective forces to shape such extensive nucleotide changes, especially considering the high proofreading capacity of coronaviruses.”— Wells H. et al., Cell Host Microbe, 2023
When we think of general genetic diversity as a “potential for new functionalities” and recombination as a way to “shuffle varied genome segments” around, some alarm bells should be ringing.
With chimeric sarbecoviruses genomes, we are in essence observing direct evidence of countless potential “gain-of-function” experiments conducted on a scale hardly imaginable that must have happened in the past. (More on that in Chapter 2)
The existence of untold numbers of such recombinant cousins in the wild implies that chimeric sarbecoviruses must have been birthed in some vast, natural “gain-of-function” laboratory.
A laboratory where gain-of-function experiments with promiscuous parental lineages and distinct genetic functionalities relentlessly produce chimeric offspring that might have tricks from both parents.
Over time, the constant combinatoric mixing of varied elements and functionalities will on occasion produce some very elaborate traits, such as the ability to infect multiple different hosts (broad species tropism), including us humans.
I think it is worth looking at that in detail.
IV) Of locks, keys, and the door to human infection
Broken down to a structural level, the major determinant of CoVs to infect various host species has to do with the makeup of the receptor-binding domain (RBD), a part of the 3D organization of the larger viral spike protein that has to fit the 3D scaffold of host receptors (ACE2 in humans) exposed on host cells. This is a bit of the lock-and-key principle.
The higher the genetic diversity of the RBM (the genetic sequences coding for the RBD in the spike protein), the bigger the repertoire of potential keys available to CoVs.
“Sarbecoviruses exhibit extensive genetic diversity in RBM, likely arising from frequent recombination and the high selective pressure associated with inter-species host jumping. “— Si J. et al., biorxiv, 2024
Viral recombination in CoV spike genes is considered a major evolutionary mechanism that drives new adaptation processes, such as viral host switching. (If you already have a key factory within the family, and promiscuous virus sex where successful keys get passed around like a hot potato, all new chimeras need is the physical opportunity to try them on previously locked doors)
Here is an interesting little sidebar:
In any case, we do not need to speculate about whether it is possible or what potential ambitions genome engineers might have had or not. Since 2021, researchers have not only found more cousins of SARS-CoV-2, but they also discovered that some of these cousins had keys very much identical to it, proving that nature is and remains the ultimate key master for this particular set of keys, so to speak.
A very recent illuminating preprint from Chinese scientists — including Zhengli Shi, the unjustly blamed “Batwoman” herself— took a deep dive to understand how exactly RBM motifs (key shapes) relate to broad ACE2 tropism.
Experimentally, they produced 56 individual cell lines expressing ACE2 orthologues (locks) from bats and selected mammals and then tried 14 different RBDs (keys) from Sarbecoviruses.
What they discovered was that some keys could open almost all locks, whereas others could open only one, or even none of the locks presented.
Much of this is not unique, but based on specific deletions in specific positions of the RBM (see below).
“We revealed that most sarbecoviruses with longer RBMs (type-I), present broad ACE2 tropism, whereas viruses with single deletions in Region 1 (type-II) or Region 2 (type-III) generally exhibit narrow ACE2 tropism, typically favoring their hosts’ ACE2. Sarbecoviruses with double region deletions (type-IV) exhibit a complete loss of ACE2 usage “
The RBD in SARS-CoV-2 has had a very broad binding affinity from the start, which is why it is not only great at infecting humans, but minks, deer, house- and zoo animals as well.
Why Sarbecoviruses would need either very broad or very specialized keys we will look deeper into in chapter 2; it has to do with evolutionary and selection pressures in niche host environments.
But just to sum up the implications of our deep dive into spike recombination, genetic diversity, and broad ACE2 tropism:
The supposedly “uncanny” ability of SARS-CoV-2 to infect human cells is neither uncanny nor especially unique. Some bat CoVs can just infect humans from the get-go. No pre-adaption, no magic human hand, no designer, no arcane lab experiments needed.
On top of this, unique recombination patterns in the spike gene and the discovery of identical RBMs in nature make it unequivocally clear that SARS-CoV-2’s RBD was not “designed”, “created” or “swapped in from an unknown virus” by researchers in any kind of “gain-of-function” setup.
It acquired this receptor-binding domain in nature.
V) Defusing the myth of the Furin cleavage site
But what about the furin cleavage site? Project DEFUSE?
You might hear lab leak proponents lament. Recombination does not have the granularity to rule out that this small sequence motif was artificially introduced. The idea gained traction after a rejected research proposal from 2018 was played up by credulous amplifiers in the press.
In the “man-made” mythology, the furin cleavage site (FCS) — a polybasic cleavage site recognized by furin-like proteases — is a dramatic functional element and allegedly the secret sauce and trigger whose artificial insertion turned an ordinary bat virus into the pandemic blight pathogen we have today.
For lab leak believers, the DEFUSE proposal mentioning the possibility of introducing cleavage sites coupled with the “suspicious” occurrence of an FCS in SARS-CoV-2 is all but proof that the virus was engineered or tinkered with. In other words, researchers wanted to create a pandemic virus, or had reckless disregard if their work would do so.
A quick sidebar:
Despite the DEFUSE proposal being irrelevant, I think it is worth discussing the two associated insinuations based on scientific ignorance a bit deeper.
First, there is the common confusion that the introduction of a single genetic element has the power to make a pandemic pathogen.
Second, there is deliberate deception about how likely it is that nature or engineering came up with the FCS insertion in SARS-CoV-2.
The FCS is probably one of the most misunderstood elements in the history of the SARS-CoV-2 origin controversy.
So let us first talk about why the Spike protein needs to be cleaved, where it needs to be cleaved, and how the FCS might help. (The next points are going to be highly technical, so I made a visual summary to help below)
The Spike protein in SARS-COV-2 needs to be cut during viral processing, first in the S1/S2 region, and second at the S2' site
TMPRSS2 extracellular cleavage facilitates entry at or near the cell surface (cell fusion), as opposed to viral entry through the endosome pathway and late cleavage by cathepsins
Extracellular cleavage allows the virus to avoid the potent endosomal/ endolysosomal restriction factors — the IFITM proteins — which inhibit viral membrane fusion and can stall virus replication
TMPRSS2-expression is particularly high for cells in the upper respiratory tract, so viruses that can use this route more efficiently will have a selective advantage in respiratory transmission
Furin pre-cleavage aids TMPRSS2-mediated cell fusion
The FCS motif allows spike proteins to be pre-cleaved before they leave their host cell; but also makes them less stable
Spike protein egress that was pre-cleaved by the furin-like proteases in S1/S2 seem to take away some of the processing work of TMPRSS2, bind better to ACE2, and thus make viral entry via this membrane-fusion route more efficient for the virus
Studies have shown that TMPRSS2-mediated entry was particularly potent for virus particles that had FCS-containing spike compared with the non-furin-cleaved mutants
In ferrets it was shown that virions with FCS-containing SC2 spike protein could spread to new hosts, but did not observe this for virions with FCS-mutated spikes
Conservation of the FCS since pandemic start argues for its selective advantage in human-to-human transmission as well
Therefore, we know that the FCS in SARS-CoV-2 is critically involved in transmission and pathogenicity. (More about this and other respiratory adaptations in Chapter 2)
That an FCS acquisition can aid a respiratory virus is not exactly unobserved in nature either; for example, the transition from low pathogenic influenza strains towards high pathogenic versions often happens by the acquisition of a polybasic cleavage site which can be recognized by furin-like proteases.
However, context matters:
It was shown that even furin-deficient cells can still cleave SARS-CoV-2 spike protein at S1/S2
The furin-cleavage site in SARS-COV-2 is quickly lost by normal cell culture techniques because of slower kinetics, so it impedes the virus in this laboratory “serial passage” context
The requirement of FCS and basic residues at S2′ for S-mediated cell fusion is entirely cell type dependent
an FCS does not make (or necessarily break) an pandemic virus, the whole viral-host context matters
For a subset of highly pathogenic viruses infecting humans, an FCS might be a necessary, but not sufficient element to efficiently infect humans or sustain respiratory transmissions.
However, merely adding an FCS can not magically turn any bat sarbecoviruses into a pandemic pathogen
For example, MERS-CoV and SARS-CoV-1 both caused epidemics, one with and one without an FCS. In the case of SARS, some experiments have shown that the artificial addition of an FCS does not increase infectivity or pathogenicity. Additionally, SARS-CoV-1 without an FCS is no less infectious in ferrets compared to FCS-containing SARS-CoV-2. Some of our endemic human CoVs have an FCS (hCoV-HKU1) but do not use ACE2 as an entry receptor, others have no FCS but still use ACE2 (hCoV-NL63).
In nature, no genetic element acts alone.
For a pandemic virus, the right genetic elements, host and environmental conditions have to fall together to create a perfect storm. We also would expect for any natural virus capable to start a pandemic to have some unusual genetic tricks up its sleeve. Otherwise, every ordinary virus would cause a pandemic.
So why are many suspicious about SARS-CoV-2 having an FCS?
The main reason why the short FCS sequence in SARS-CoV-2 (12 nucleotide insertion) has been hotly discussed is because it stood out like a sore thumb in side-by-side sequence comparisons to known coronaviruses in early 2020. (Since then, scientists have found plenty of naturally occurring FCS in the wider CoV sub-family, but not in the sarbecovirus sub-genus specifically)
Just because something is rare or even unique among currently known sarbecovirus members is however not good evidence for artificial introduction given the high sequence diversity at S1/S2 (see below). On top of that, there is some evidence to suggest the FCS region itself underwent recombination with a related sarbecovirus. Again, we underestimate the diversity of nature at our own peril.
But let’s discard for a second our experience with vast and crafty nature — and the naïve presumption that an FCS can turn any virus into a pandemic threat — and take the idea of a “man-made” FCS introduction as a null hypothesis.
Is there some evidence that could help us confirm or reject this hypothesis?
For an engineered FCS, the observed sequence motif in SARS-CoV-2 is odd, to say the least (see figure below for a visual summary)
All previous FCS created by researchers (for example here in 2006, or here in 2009, all published years before the “secret” DEFUSE proposal suggesting similar work got played up as a smoking gun) have avoided using an “insertion”, and rather chose to “substitute” — rewrite — existing nucleotides in this position to change the encoded amino acids to create an FCS. This is necessary and basic biology because the addition of new amino acids to any spike protein will likely disrupt its structure and break the whole thing.
Curiously, an additional, unnecessary proline residue is part of the insertion. Proline is not part of any FCS motif, however, the amino acid is considered a helix-breaker; something that will mess up your protein structure with a high probability. Why risk introducing that?
Even more curious, the whole sequence insertion is “out-of-frame”, meaning the nucleotides were not introduced neatly but rather staggered for no reason, again something that scientists would not do ordinarily (or at all)
However, the FCS insertion is right next to an RNA sequence motif prone for copy-choice recombination, arguing for this natural mechanism bringing it about (and copy-choice recombination is reading-frame independent and favors out-of-frame insertions 2:1)
The GC content of the insertion is 83%, very unusual for artificial introductions
The coded amino acids themselves make an odd, non-canonical, and weak recognition motif (RRXR) for furin-like proteases; a motif that might ordinarily not work compared to strong motifs such as RRSRR (which would usually be used by virologists).
The FCS cleavage at S1/S2 destabilizes the original (Wuhan-Hu-1) spike protein, which only got compensated partially when the D614G mutation emerged quickly in the early pandemic, arguing for a recent acquisition, not design or cell culture
Furthermore, potentially because the FCS in the original SARS-COV-2 is not that stable, it is lost very quickly in Vero cells and other typical cell culture systems. Selection pressures seem to run against it in these research-associated amplification systems; so how exactly did an FCS-containing virus that first has to be amplified in cell culture before it can infect a researcher escape from a lab and do so with an FCS?
From an engineering perspective, none of these odd, suboptimal, and self-defeating FCS design choices make any sense and certainly contradict the assumption of “rational design” as a whole. They also serve to rule out engineering or mimicry of existing furin cleavage sites.
But that alone is not proof. Who knows what weird and arcane experiments those Chinese researchers would cook up in a lab, right? (Some lab leak truthers with no molecular biology experience would say. For actual genetic engineers, the situation is already clear as day) Fortunately, we do not need to rely on experience or speculations about any motivations or experimental procedures Chinese scientists might have followed.
This is because nature just has more secrets hidden in the FCS that humans could not have come up with in the first place.
After intensive mechanistic studies of the FCS in SARS-CoV-2, researchers discovered a hitherto unknown synergistic interplay between the odd, suboptimal FCS and the genetic backbone of the virus (which we by now know is natural because of recombination).
We have learned that pre-cleavage of SARS-CoV-2 S1/S2 by furin is what boost TMPRSS2-mediated cell fusion and thus impacts cellular tropism and transmission dynamics (ergo infectivity)
But it turns out, that was not the full mechanistic story. In nature, no genetic element acts alone
The viral QTQTN amino acid motif is an uncommon natural sequence in several sarbecovirus spike proteins directly upstream of the FCS sequence
In cell culture experiments, QTQTN is not stable and repeatedly lost, similar to the FCS. Loss results in impaired viral replication.
QTQTN determines how tightly the loop harboring the FCS is bound to the spike protein, thus it regulates spike protein stability and shapes how well the Golgi-bound furin-like proteases have access to recognize the FCS motif and pre-cleave at S1/S2
The QTQTN motif is also glycosylated and loss of that glycosylation impairs viral replication as well.
Researchers showed beautifully that there is an intricate functional interplay between the FCS, the loop length, and glycosylation: These separate but co-dependent elements work together synergistically and all elements are ultimately required for efficient FCS pre-cleavage (and with it, impact viral replication and pathogenesis). In nature, no genetic element acts alone.
Co-dependency and synergy between genetic elements are hallmarks of evolutionary selection; these type of complex interactions are extremely hard to design for even with perfect knowledge of all structural components, which nobody had in 2019. In the case of SARS-CoV-2’s FCS, there is no conceivable way how engineers could have designed the observed — but hitherto undiscovered — mechanistic synergy purposefully.
Some might still contest that engineers could have stumbled upon this interaction by sheer “lucky accident”, like winning the lottery.
But I think even that special pleading for a “miracle accident” evaporates in comparison to what we know about nature’s relentless trying to bring forth such elements. Mechanistic details and probabilistic observations are key to wrapping our heads around the FCS in SARS-COV-2. If we understand that:
an FCS alone can not turn a random bat virus into a weapon of mass disruption
the FCS has many hallmarks speaking against design or engineering but strongly arguing in favor of evolutionary mechanisms
a previous unknown synergistic interaction (with an uncommon natural sequence contained in a 100% natural backbone that evolved in nature) could not have come about or stumbled upon by any known laboratory experimental setup
…only then can we fully appreciate the scope and intricacies of gain-of-function experiments nature runs every day.
They are of a sophistication and breadth human engineers yet can not hope to match, nor fully fathom.
So this point in our genetic deep dive, there is only one thing to do:
VI) Putting the “gain-of-function research” narrative to bed
Speculations that the “odd and unusual” genetic features of SARS-CoV-2 came about by gain-of-function research, genetic engineering, design, or any other type of human tinkering — while hard to disprove in 2020 — are today contradicted by available evidence and published scientific literature. Science did not stand still in the last four years, researchers learned an incredible amount about the intricacies of the virus that disrupted the world.
Taking together various recombination analyses, sequence diversity in nature, and hallmarks of evolution in the backbone, RBM and FCS leave only one parsimonious conclusion:
SARS-CoV-2 is a fully natural virus, its specific genomic features could not have come about by any known or unknown laboratory experiment. Period.
Note: A natural virus in itself does not disprove a lab leak, nor all conceivable research-related origin scenarios. But it certainly disproves all notions of “gain-of-function” research causing the pandemic.
Currently, it can not be disproven that an extremely rare pandemic-ready SARS-CoV-2 ancestor circulating in the wild could have been found by researchers, brought to Wuhan secretly, being cultured secretly (and quickly) before it lost the FCS but with high enough titers to somehow infect a researcher. The virus then beat the odds and escaped from a lab unnoticed, then beat the odds again to cause an outbreak, then beat the odds again to start and spread at a wildlife market far away from the lab but nowhere else in a megacity. And then a week later, the exact same unlikely chain of events happened with a slightly different SARS-CoV-2 strain called lineage A; because evidence shows that both lineage B and A are centered around the Huanan market. On top of that, everything that could point to the occurrence of such unlikely chains of events was covered up perfectly not only by Chinese scientists, but by an international cabal including independent scientists, governments, and international organizations.
These or similar convoluted scenarios might remain theoretically possible forever, but are they plausible?
Science works with evidence, so it is often impossible to prove a negative — that something did not happen — scientifically, as there would be no evidence of something “not occurring”. Conspiracy myths are adaptive and built to sustain the belief that the evidence-based “mainstream” explanation is wrong. This often makes their scientific refutations difficult especially when the goalposts of how an event supposedly happened are constantly shifting, including central elements such as which lab is supposed to be the culprit can shift at a moment’s notice.
All I am cautioning about is that such cover-up conspiracy myths filled with extraordinary and magical assumptions are epistemologically irresponsible, contradicted by multiple other strong lines of scientific evidence (we can not go deeper into here, but see below), and can not even explain most of the body of evidence we observe.
I believe it is fair to say that the continued push for a “gain-of-function lab accident” narrative in politics and media is not driven by evidence, not justified by remaining scientific uncertainties, and can be safely discarded. We have seen again and again that no amount of hard evidence for any scientific theory has the power to prevent certain activists, politicians, influencers, and other manipulators from trying to sell citizens a different — and usually emotionally more satisfying or engaging — story. Climate change denial, anti-vaccine activism, homeopathy, various alleged alternative medicines or wonderdrugs; none have any scientific leg to stand on, but are profitable narratives to use for popularity, persuasion, or power. Those who have something to gain from pushing false myths will continue to mislead no matter the evidence. We just do not have to buy into their debunked falsehoods.
The constant performative pearl-clutching and fearmongering about gain-of-function research or lab biosafety in the media should not distract citizens from the factual reality that SARS-CoV-2 is a natural virus.
Another important side note about this here:
My larger point is that biosafety is an important, complex, and omnipresent topic that virologists take seriously. But do its most fervent advocates in politics, media, and other halls of power?
If you observe these activists, politicians, contrarians, or influencers who claim gain-of-function research — or any type of human genetic engineering technology — created the virus, you’d be wise to question not only their motifs but their leadership.
Unfortunately, getting the origin of SARS-COV-2 right is not just an academic exercise, or avoiding being fooled. It is about the actions we collectively fail to take when falsehoods dominate the discourse.
The relentless mythmaking about a “gain-of-function” virus and alleged lab biosafety breaches have exposed a dangerous blind spot in its most fervent purveyors:
The inability to appreciate the vastness of viral diversity and the “gain-of-function” ingenuity of nature, both of which continue to pose a predictable and outsized threat to human health and prosperity.
False narratives and beliefs in a “gain-of-function lab accident” are not without consequences for society, and neither is scientific illiteracy or dishonesty in elected representatives and social elites. Are we truly okay with too many leaders chasing false myths while sleepwalking into the next natural pandemic they could not be bothered to understand? Whatever your risk-to-benefit perception about gain-of-function research in labs, ignorance about the scope and predictable danger of gain-of-function dynamics in the wild is a surefire recipe for disaster.
I hope citizens can at least agree that if you care about “gain-of-function viruses”, you should care about them anywhere in the world.
Any biosafety advocacy that magically stops at the lab door — that does not consider nor care for the quantifiable orders-of-magnitude larger natural risks — is not a good basis for collective decision-making in an interconnected world. The threat we neglect will be the one we will keep facing.
So let me help focus our attention and understanding of where the true origins of this chimeric gain-of-function virus — and its future unwelcome pandemic cousins — actually are to be found.
Chapter 2: Chasing the ghosts of SARS-CoV-2
Prelude: A river in the sky
A river of black drew its line in the darkening sky. Above bat ornaments on crimson roofs from the pagodas at Wat Khao Chong Pran, the river flow turned southeast tonight, towards the crop fields. Rationally, I knew that the cave housed around two and a half million horseshoe bats, but observing a seemingly never-ending flood of hectic creatures — the fly-out of 2,5 million bats took more than forty-five minutes — I realized that I never truly appreciated how many bats share the world with us.
There are around 1500 described bat species that have emerged from their last common ancestor over 60 million years ago. Because they are the only flying mammals, we conceptualize them all together under an umbrella term called “bats” like we do with “fish” in the sea. But based on genetic diversity, that simplification is rarely adequate. It feels equivalent to lumping together giraffes and cows with dolphins and whales, all of which diverged from a shared common Artiodactyla — even-toed ungulate — about fifty million years ago. Hardly justified we think of them as the same, so why do we do it for bats?
It is no exaggeration to claim that bats come in almost all sizes, shapes, and forms; from the thumb-sized Craseonycteris thonglongyai — also known as bumblebee bat weighing just 1,5 grams and holding the title of smallest mammal on earth — to various majestic flying foxes with wing spans of over 6 feet, close to two meters. Some bats look almost like dog puppies you’d want to cuddle and take home, others might appear like they’ve escaped from a horror-movie production. Maybe we humans just tend to be afraid of what we do not understand.
The horseshoe bats flying here were small insectivores — insect-eating bats — who got their name from the weird horseshoe-shaped disfigurement where their nose should be. Intuitively, we humans find them rather ugly — I was no exception — at least at first. I think this is partly because we assume their faces are weirdly deformed, like a fully cleft palate, rather than what they really are: optimized.
Horseshoe bats belong to a group of bats that echolocate — sending out and receiving sonar waves — primarily through their nostrils. Other bats mostly use their mouths. These varied shapes and forms in the middle of their faces however have intricate functionality for shaping their calls, impacting not only orientation but for their feeding and social lives too.
Given their enormous diversity, bats exist in almost all variations of social structures, from eremites that don’t want to bother with others, to small family groups, to villages, to multicultural megacities. Some like to mingle with other bat species, others are territorial and of the “get off my lawn” persuasion, with threatening grunts and fletching teeth and all.
Some horseshoe bat species are not only considered hypersocial cosmopolitans, they are also the most prominent host reservoir to SARS-related bat coronaviruses; close viral cousins of both SARS-COV-1 and SARS-CoV-2 that have since caused havoc in our human world.
I believe to truly understand where these dangerous viruses come from, we have to first take a look into the lively homes of their hosts.
I) The Cave(s) Where It Happens
“No one else was in the room where it happened. The room where it happened. No one really knows how the game is played, the art of the trade, how the sausage gets made. We just assume that it happens, but no one else is in the room where it happens” — Hamilton (Musical), the Room Where It Happens.
Physiologically, bats are extraordinary, they can speed up their metabolism 16 times, creating immense heat that would denature our proteins and fry our cells. A bat’s heart can beat up to 1,000 beats per minute and its body temperature can rise to 42 degrees Celsius [107.6 F] during flight at night. Even if we could withstand the initial stress of flight, our inflammatory response after would render us sick. That is not true for bats. Many species have unique immune systems that do not overreact to stressors; including viral infections. There are also no reliable markers of aging after they reach adulthood, bats seem young until they are dead. Some species tend to live up to 40 years in vast, dense, and diverse colonies. All of the above seem to make bats uniquely suited as reservoirs to almost all viral families that befall mammals. But for Sarbecoviruses specifically, that is not the whole story.
Virus outbreaks are always social phenomena as well, and that rarely is limited to humans.
There is emergent evidence that the lives of their primary hosts, the Rhinolophids — horseshoe bats with odd nostrils — are not only a lot more crowded and cosmopolitan than the average bat, but also a lot more socially intricate. Horseshoe bats vary in nose shapes because that allows them to specialize in echolocation at very wide ranges of frequencies. While not fully understood today, researchers believe that bat echolocation is not only for orientation and hunting but also for communication and vocal learning. Some bat’s extensive vocal repertoire is needed to create specialized social calls used either for parent–offspring reunions, territorial defense, or maintaining group integration. Some speculate that the evolution and differentiation of vocal frequencies within a species leads to mating preferences and exclusions, driving niche formation and ultimate species separation into species complexes.
A species complex is interesting from the perspective of viruses. The genetic stratification of their host reservoir means that some viruses will specialize on a small niche or subgroup of the species complex, whereas others might try to develop or maintain broad affinity to multiple species that might still inhabit the same spaces, but not mate with each other.
Let’s recall an observation from before: the RBDs of sarbecoviruses can be very broad (even opening the door to non-bat ACE2 receptors), or very niche.
It is reasonable to assume that the intricate population structures of Rhinolophids create evolutionary pressures for viruses to constantly adapt to new niches. Recombination as an “evolutionary fast-forward” is the mechanism of how sarbecoviruses survive in such an environment, which is why this is such a prominent and critical feature in their evolution.
Given that the shuffling of genetic material has a very high probability of producing genomes that are no longer functional, the selective pressure for the underlying mechanism of recombination to be evolutionarily maintained must presumably be very strong. […]
Recombinant viruses that are most likely to succeed under competitive circumstances with small population sizes are those for which the recombination event has conferred a strong selective advantage over both parental viruses. — Wells H. et al., Cell Host Microbe, 2023
While various horseshoe bat species were identified as the most prominent hosts and natural reservoirs of sarbecoviruses in the wild, they are certainly not the only bat family involved in creating overall sarbecoviral diversity.
Viral recombination needs the physical opportunity for other viruses to infect the same cell; parental viruses can not create offspring if they never meet, but their offspring might also not be as diverse if they never meet very distinct others. So how do our hypersocial rhinolophids do on that “meeting new people” front?
“Some of those different species will roost together, especially when it is cooler or during hibernation. So when we have done work in various cooler caves […], we will see clusters. Rhinolophus [horseshoe bats] cuddled up to Miniopterus [long-winged bats], and the next one is Myotis [mouse-eared bats]. Now these are lineages that diverged 50 million years ago.” — Bat ecologist Prof. Alice Hughes
Beyond mere cuddling with strangers, rhinolophids are known to switch roosts constantly. There are maternity-only roosts, summer and winter locations, and roosts that are calm or have special geographic features. Some bats like to stay as permanent residents in a single cave, others cycle outside but come back periodically over the year — maybe kids going to college would be a social analogy, if only we could tell bat age, that is — and many bats do food tourism with the seasons. All of this leads to constant turnover, and various bat species mixing and mingling with each other.
So we have a colorful and complex social, seasonal, and geographic mixing of various bat species even when just studying single locations, like a specific cave or a forest.
Some of this movement and mingling was always a natural part of bat lives. More recently, however, human encroachment on bat territory — urbanization, deforestation, hunting, tourism, mining, etc — is considered the biggest factor in displacing bats from their traditional roost sites and forcing species together that might not have ordinarily met.
If some bat species that diverged often millions of years ago tend to huddle together — and we already learned that the ACE2 tropism of various sarbecovirus RBDs can be very broad — do the diverse viruses they carry infect each other and influence recombination patterns as well?
This is one question researchers were able to answer recently with a resounding yes.
By using a meta-transcriptomics approach — taking a sample and sequencing every single piece of genetic information in it — they discovered that many of the bats they sampled carried more than one virus. On top of that, the study found that some of the discovered viruses were shared between different bat species, suggesting frequent spillover events distributed the virome over multiple hosts. Dramatic opportunity for virus sex (recombination) and diversifying the viral gene pool with new genetic elements.
“The frequent virus spillover among phylogenetically related or spatially co-located bats provides an opportunity for viromes of different bat species to exchange, further expanding genetic diversity of circulating viruses”
Whatever innate mechanistic ability for recombination sarbecovirus genomes might have, it is largely the social lives of their host reservoir that create the opportunity and evolutionary pressures for recombination to shape sarbecovirus genomes.
These caves are nature’s vast “gain-of-function” laboratory, all while our human encroachment on bat territory is stirring the genetic cauldron ever faster.
So now that we know what “gain-of-function labs” we are looking for, do we know where to look? Where exactly is the birthplace of SARS-CoV-2?
II) Undersampled Karst in the Wild-Wild East
Karst is a topography formed by the erosion of soluble carbonate rocks, such as limestone or dolomite. With a total area of 500.000 square kilometers in China and 400.000 square kilometers spread over various nations in Southeast Asia, these countries house some of the largest and most biodiverse Karst regions in the world. But it is truly the power of water over eons that has shaped the landscape to include spiky spires, enormous sinkholes, underground rivers, and intricate cave systems below the old-growth subtropical forest and rainforests. This is bat country, and each valley and limestone formation can not only house millions of bats but is a microcosm in itself.
“Each isolated limestone hill can host more than 12 unique species found nowhere else on earth, with up to 100 microsnails, endemic begonias, orchids, and geckos, and yet an estimated 90% of cave-dependent species are undescribed. “— Prof. Alice Hughes, University of Hong Kong
The Karst region spanning Southern China and Southeast Asia forms a unique and biodiverse ecosystem that in considerable parts is still relatively untouched by humans, albeit that has been changing especially with deforestation, slash-and-burn, and cash crop agriculture leading to the degradation of the land. Its importance to the global climate and conservation is probably second behind the Amazon rainforest, with projections suggesting around 40% loss until the end of the century if we do not dramatically change course.
The shifting landscape use in turn is one of the major drivers of zoonotic spillover. Some studies suggest that around 65.000 spillovers of CoVs happen each year, most of them from viruses that are yet undiscovered and that do not cause an outbreak for a variety of reasons.
“The risk for zoonotic viral disease presence and emergence in humans also increases in geographic areas with higher mammal diversity, where previously pristine forests have been recently deforested. This makes the areas of Southeast Asia, which support large, intact natural habitats and have ongoing ecosystem fragmentation, at high risk for disease emergence “
The majority of viruses that spill over might just not replicate well in the new host cell, or get shut down by our innate human host defense, such as the interferon pathway, without humans ever being the wiser. Even most replicating viruses might be unable to transmit between humans, or unable to spread effectively enough. Outbreaks are also always social phenomena, even a perfectly capable pandemic like SARS-CoV-2 would have died out with a 99,5% likelihood if it spilled over in a remote village, rather than a megacity.
Given most of these sarbecoviruses die out and will never be known, narrowing down where exactly SARS-CoV-2 came from within an often inaccessible, remote Karst region 2,5 times bigger than the size of Germany is a daunting challenge.
Yet this background of viral emergence and constant spillovers gives researchers three lines of attack to at least narrow down where the bat ancestor of SARS-CoV-2 likely originated.
discover the immunological footprint of related viruses through serology in humans and farm animals
discover the immunological footprint of related viruses through serology in wild bats
discover related viruses through sequencing bat samples
Evidence for SARS-CoV-2-related sarbecoviruses has been found in China, Laos, Vietnam, Cambodia, Thailand, Myanmar, and Malaysia.
Most direct evidence comes from sequenced bat viruses in China (before sampling in Yunnan became forbidden), where researchers found a pool of SC2-related viruses in Yunnan province. (China has been more sampled than other nations)
Another trove of informative viruses was found by French and Laotian researchers in Vientiane province in Laos. Here the closest match to SARS-CoV-2 broad-affinity RBD was found, proving nature unlocked that particular door to human infection; not human engineering.
More distant SC2-related viruses have also been found in Cambodia and Thailand; which in themselves can house informative genetic elements such as insertions at the S1/S2 site reminiscent of the polybasic cleavage motif in SC2
Some have summarized the possible origin place at very low resolution (below) based on the sampling location and overall genetic similarity of the various SC2-related viruses.
Yet such efforts are bound to be highly dependent on sampling biases and offer a low granularity that is almost meaningless. The map has also not considered some other relevant findings such as:
SC2-related viruses have been found in pangolins smuggled over the Chinese and Thai borders
Especially the karst regions bordering Laos, Northern Vietnam, and Myanmar seemed very rich in these viruses, with Myanmar lacking any bat sampling efforts there so far
However, serological evidence from Myanmar further South found a very high prevalence of undiscovered SC2-related viruses in humans that had contact with wildlife. More than 1 out of 5 had antibodies and cross-reactivity to a panel of these viruses (most prominently RaTG13), arguing for high prevalence and spillover potential in this region, but the concrete viruses remain undiscovered
Bat sampling in Japan, Southern Vietnam, and Malaysia shows more ancestral and distantly related sarbecoviruses
Sampling in Northern Thailand has been politically difficult to publish (personal communication) but might be insightful as well
Here is a rough summary of scientist’s efforts so far:
Honestly, while a lot of valuable information has been gathered in the last few years, it serves merely as a snapshot of the scope of our ongoing ignorance.
The reality is that there is still much we do not understand about our diverse world. According to Prof. Alice Hughes, it is estimated that 90% of caves in Southeast Asia alone remain scientifically undescribed and uninventoried, while some estimate around 60% of bat species remain to be discovered, and a staggering many of the bat species we know about have never been sequenced. We still need to learn much more about viral and bat geography, how human drivers from deforestation to urbanization to tourism interface with bat immunity, viral-host interactions, bat-human contacts, and propensity for viral shedding.
We are working with a map where many spots are and will likely remain blank because any research on bats has been so highly politicized by the distorted origins debate.
There is a need to study these complex interactions between viruses, bats, and humans. However, many Southeast Asian nations, as well as China, have little interest in allowing researchers to sample more bats, discover new viruses, or even publish ones they have already found (personal communication from multiple bat hunters).
Especially sarbecoviruses are a sensitive subject almost everywhere. Governments are currently not interested in discovering where these dangerous chimeras are circulating in the wild or what drives their emergence.
Some governments fear more scientific discoveries will lead to them being blamed by a world that has not made peace with natural pandemic risks.
A world that asks for reparations, culprits, or impossible insurances, rather than offering a helping hand on a collective problem. A world where leaders seek power with geopolitical grandstanding, or worse, gain popularity by exerting vengeance on inconvenient scientists rather than heeding their warnings.
I believe collectively, we need to do better on that front.
Scientists are doing their part, despite these artificial limitations and political constraints. Some of them have been tinkering with new methods and ingenious ideas to learn more about the origin history of SARS-CoV-2, and there are some exciting findings to report.
III) Ghost hunting with phylogenetic inference
Gaining knowledge from sparse data is difficult, but not impossible.
To date, only ~200 sarbecoviruses have been found. 139 of them are closely related to SARS-CoV-1, and 26 are SARS-CoV-2-like. Not much to work with to pinpoint a geographic location, and individual genomes can also not be used to reconstruct a faithful evolutionary history because all of these are chimeras, so molecular clock estimates based on mutational divergence are being misled by the “evolutionary fast-forwards” of recombination.
Evolutionary rate estimation can be profoundly affected by the presence of recombination — Boni MF. et al., Nature Microbiology, 2020
The number of sarbecovirus relatives discovered to date is however just good enough to identify recombination breakpoints, places where genetic similarity from one parental strain stops and similarity to a different parental strain starts. A way to think about recombination breakpoints is as a minimum number of sexual acts that must have happened in the past that can explain the segmented shape of the chimeric genome today.
The genomic segments in between recombination breakpoints are called non-recombinant regions (NRRs), basically intact genomic segments from a parental strain. On these segments, molecular clock estimates work normally and can thus be informative. Virologists will just have to individually work with as many clocks and evolutionary histories as there are NRRs, which is 27 in the case of SARS-CoV-2, and 31 for SARS-CoV-1. (see figure below)
“If you align all viruses in the same place, and then chop up the genome into segments — what we call non-recombinant regions (NRRs) — then each NRR will have their evolutionary history” — Dr. Spyros Lytras, an evolutionary virologist at the University of Tokyo (exclusive video interview video at the end of article)
Imagine NRRs not as pieces of larger viruses, but as floating genetic elements that can shuttle in and out of viral backbones. Why should they not be analyzed individually?
Separating the histories of NRRs allows researchers to gain some critical information. For example, while some NRRs might be very old and just serve as reminders of sexual conduct decades ago, others might be very recent additions; the final sexual (recombination) events that gave rise to the chimera.
Indeed, researchers found that the youngest recombinant additions for SARS-CoV-1 and SARS-CoV-2 to their closest-inferred bat virus ancestor happened not decades ago, but were in 2001 (SARS-CoV-1 spilled over in 2002) and in 2014 (SARS-CoV-2 spilled over in 2019), respectively. This is of course not the final word, as the researchers noticed the more cousins are discovered, the closer the time estimate moves up towards toward the emergence date.
“The virus pieces that were most similar circulated in bats very recently” — Dr. Jonathan Pekar, US San Diego
Having identified the 27 disparate recombinant pieces constituting SARS-CoV-2 also allows researchers to do something very interesting, which is constructing the closest common recombinant ancestor (recCA) genome, basically looking at which viral cousin of SC2 has the highest similarity to SARS-CoV-2 for each NRR.
The recCA genome is an aggregate, built from the ghosts of closest-inferred ancestors for each NRR that must have existed at one point
Tellingly, researchers have found that the more bat virus relatives got discovered in nature, the closer the genome sequence of the recCA resembled the human-infecting SARS-CoV-2 that emerged.
In other words, we have evidence to infer that a 98.8% identical to SC2 bat virus ancestor circulated just a few years before emergence; with temporal and genetic granularity still increasing as more viral cousins are discovered.
But what can we do with that knowledge?
IV) Reconstructing the phylogeographic histories
As we have observed in the karst region, sarbecoviruses, just like their bat hosts, show a degree of geographic structuring; with the border region of Laos, Myanmar, Northern Vietnam, and Southern China seemingly quite rich in sarbecoviruses closely related to SARS-CoV-2.
But does that mean SARS-CoV-2 came from that region, or is it just a mixture of undersampling and coincidence?
By using these phylogeographic data, a phylogeny (family tree) for each NRR can be constructed that is spatially mapped to sampling locations.
“When we build those family trees, we want to calibrate them with time” — Dr. Jonathan Pekar, US San Diego.
Combining geographic information from sampling locations of close ancestors with phylogeny scaled with units of time of their closest recombinant ancestors creates a geo-temporal record for the various genetic elements constituting the SARS-CoV-2 viral genome.
A record of their ancestor’s whereabouts throughout history, and where and when they had time to meet each other to mingle.
These geospatial and temporal efforts also allow us to narrow down a dispersion zone for these “floating genetic elements”, and with it, the viruses that arose by their (re)combination.
In other words, the most likely geographic birthplaces of direct bat ancestors to SARS-CoV-1 and SARS-CoV-2 (see below).
“The phylogeographic patterns seemed to be spatially structured, with phylogenetically similar viruses being generally sampled in geographically similar regions and almost no back-and-forth travel of viral lineages across relatively large regions of the study area.” — Pekar et al., biorxiv, 2023
This last result gained from studying sarbecovirus ancestry unearthed one more essential clue to how SARS-CoV-1 and SARS-CoV-2 emerged.
It starts with a suspicious conclusion:
“Direct ancestors of the SARS-CoVs likely could not have reached sites of emergence via the bat reservoir alone”
SARS-COV-1 emerged in Guangzhou, around 1000km away from its suspected home; and SARS-COV-2 emerged in Wuhan, around 1500 km away from where the direct bat ancestor existed.
So how exactly did SARS-CoV-2 find its long way from the sarbecovirus “gain-of-function” heart of the Karst region to a wet market in Wuhan?
V) Smoking guns pointing at the wildlife industry
An outbreak rarely divulges all of its mysteries. How the virus made it to Wuhan still leaves room for uncertainty and speculation. Lab leak advocates might be quick to smell an opportunity to allege that maybe a researcher from Wuhan has gotten infected and brought the virus to the city. Weren’t Zhengli’s group and some other teams sampling in Yunnan after all?
These insinuations are unfortunately quite shallow, scientifically naive, and ultimately false because the leftover uncertainties are much smaller than one might appreciate.
I could spill much more digital ink on arguing how a priori unlikely it is for bat researchers and virus hunters to ever stumble upon a pandemic-ready chimeric virus from the small bat sampling efforts researchers ever get to conduct. One might find genetic puzzle pieces, but how nature put them together only comes to light after selection created those million-to-one lottery winners that spill over into other mammals. If Chinese researchers had collected and stored hundreds of those spillover viruses, maybe suspicions had a ground to stand on. The reality is that Zhengli Shi’s team only ever found one SARS-CoV-2-related virus, even non-pandemic viruses are hard to find. One could also contend that there is no evidence for any of these speculations. One might also list the eerie circumstances surrounding SARS-CoV-2 that look an awful lot like SARS-CoV-1; from the fact that both started in November (suggesting seasonality) to the finding that sampled wildlife farms around Wuhan had SARS-CoV-1 lineages potentially ancestral to the outbreak that happened in Guangzhou, to the fact that contemporary pangolin serology in 2003 found 1/7 pangolins testing positive for an unknown SARSr virus, and now smuggled pangolins turn up with a SARS-COV-2 related… but okay, I am spilling digital ink again, so I will stop. (See note in the extended pdf version)
Again, it is 2024 and scientists have uncovered a lot more evidence that completely exonerates Chinese and other bat researchers who ever sampled in Yunnan or various places in South East Asia.
Let us just focus on two puzzle pieces that I think are illuminating because no type of research-related accident can explain them.
The first piece
The first one is — what I feel — one of the biggest oversights in the discussion, and it has to do with the damned furin cleavage site again.
In Chapter 1, we spend considerable time explaining why we know that the FCS does not have an artificial origin. We also know from looking at sarbecoviruses that while these FCS motifs can easily be created, they do not seem to be maintained in bat viral lineages. They are even unstable in many cell culture systems. However — within certain real-life transmission contexts — the FCS is strongly preserved as a respiratory adaption.
That extends beyond humans; minks, deers, and other animals that got infected with SARS-CoV-2 via reverse spillover and the respiratory route from humans have been shown to preserve the FCS as well. Bats do not appear to do so, for reasons one can speculate (e.g. bat sarbecoviruses are gut-adapted, not respiratory) but are certainly host-context dependent. No genetic element acts alone.
This observation however means that one has to assume that the direct bat ancestor of SARS-COV-2 almost certainly did not have (or maintain) an FCS. But we also know that the FCS does not have an artificial origin.
If neither bats nor humans created the FCS, where does it come from?
By just following these two evidence-based assessments to their logical conclusion, it becomes clear that an intermediate host had to be involved.
Somewhen between as early as 2014 (likely later) when a recombination event allowed the bat ancestor to reach its final chimeric form, but before it emerged in 2019 at the Huanan wildlife market, the virus must have acquired and/or sustained it’s FCS in a non-bat host (complimentary evidence like D614G stabilization also argue for a very recent FCS addition).
“The Furin cleavage site gave me the strongest hint that the progenitor virus of SARS-CoV-2 is not in bats”, Prof. Linfa Wang told me. He is by far not the only virologist who believes this is the case. “It’s in pangolins, raccoon dogs, civets, badgers, or whatever, maybe another small mammal we don’t know”
There is a recent precedent. Evidence for such FCS acquisition in intermediate host species has been found in 2023 with trafficked pangolins. A bat HKU4-related merbecovirus (bat ancestors to MERS also do not contain an FCS) has acquired a minimal polybasic cleavage motif experimentally shown to be cleaved by furin. It seems that for CoV spillovers, intermediate animal populations tend to bring these polybasic cleavage sites forth and can maintain them.
An identical argument for another such respiratory adaptation that gets maintained could also be made for the T372A mutation that alters the 3D confirmation of the trimeric S glycoprotein to a more open and respiratory infectious form. Bat sarbecoviruses do not have this, humans could not have come up with it. So where did it come from? Same story really.
A priori, it is very unlikely that bats directly infect bat researchers, and even extremely unlikely to ever find a pandemic pathogen from the pithy sampling efforts bat researchers ever get to conduct. But add to this the likely requirement of a host-context switch and respiratory adaptation in an intermediate animal to maintain an FCS/T372A? Pretty much impossible.
I trust this should finally dispel the myth that bat-sampling researchers brought SARS-CoV-2 to Wuhan.
There has been a lot of hysteria and myth-making surrounding the furin-cleavage site in the media over the last four years.
To now understand that the one genetic element many have falsely called the “smoking gun for engineering” contributes much more to bat researchers’ exoneration than to their incrimination is certainly a fitting and long-overdue twist of fate.
Talking about non-bat intermediate hosts though...
The second piece
Something no research-related accident can explain is the fact that the multiple independent lines of evidence point to the Huanan market as the epicenter of the pandemic.
Researchers had already established in 2022 that SARS-CoV-2 susceptible wildlife had been sold at the market — despite denials and obfuscations from Chinese authorities continuing to this day — and that the spreading pattern of two lineages centering from the market is best explained by a multi-spillover scenario from an infected pool of animals.
Much more could be said here, but let’s go to the cutting edge right away:
In 2023, sequencing data of environmental samples from the market were finally released by Chinese scientists which provided a trove of genetic information to sift through for new possible clues.
And new clues were found.
Genetic diversity of SARS-CoV-2 samples is consistent with viral emergence at the market; also contradicts the market as a mere “amplifier” or “superspreading” event
Environmental samples that tested positive for SARS-CoV-2 contained lots of DNA/RNA from various SARS-CoV-2 susceptible wildlife species
Animal DNA/RNA correlated spatially with stalls that housed these animals, but not other places in the market
Environmental samples from these wildlife stalls contained sequencing reads for wildlife-specific animal viruses and SARS-CoV-2, but not other human viruses (arguing for sick animals and against human contamination of these samples)
single-nucleotide variation (SNV) analysis if raccoon dog reads found that the raccoon dogs at the market did not belong to a commonly breed species used for fur production in large operations in the north, but rather likely a wild-caught variety
These market data make it clear that no infected lab worker just walked into Huanan to start the pandemic, nor sneezed on the animals to make it look like they were involved. On top of that, the metatranscriptomics data provides multiple direct links to the wildlife trade and industry; from the species that were identified to the types of viruses that plagued those animals; and even to the geographic region where they might have come from.
These findings are consistent with the WHO report that also suggested the involvement of the wildlife industry in Yunnan and beyond.
From illegal trafficking of rare animals over the border, hunting and trapping them in the wild for later sale or sustenance, or supplementing breeding in wildlife farms, the porous industry has many facets. Some activities are illegal but lack enforcement, some are traditional and culturally valued, while other activities such as wildlife farms were explicitly promoted before the pandemic.
“Local officials trumpeted the wildlife trade as a way to close the rural-urban divide and to meet ambitious national targets to alleviate poverty.”
— Emily Fang, reporting for NPR
So that is it? This is how the virus made it to Wuhan?
Outbreaks rarely divulge all their secrets, and insights have to be pried from nature and history through the often painstaking work of scientists, journalists, and other truthseekers.
Science is a process of approaching ever more likely explanations of reality by rooting out false hypotheses and narrowing down existing uncertainties.
The remaining scientific uncertainties of the origin puzzle surround politically sensitive topics such as a rampant regional wildlife industry — estimated to be over USD 70 Billion in China and around USD 10 Billion in South East Asia just for trafficking — as well as more complex patterns of how socioeconomic behavior and ecosystemic disruption influences spillover risk factors and pandemic prevention policies at various virus/animal/human interfaces. That is why this type of “pandemic origin” research must continue, despite knowing this virus did not come from any lab. There is still much more to learn on how to prevent SARS-CoV-3. But that is a topic for a different time.
For now, I think we finally reached a (hopefully satisfying) end of this epic journey towards the cutting edge of origins research. So what is my personal takeaway?
Conclusion: Don’t bet against scientific inquiry
Finding the origin of a new virus is a lot more difficult than finding the needle in a haystack, because viruses evolve, adapt, branch out, and almost always die out without us ever being the wiser. It’s like finding a particular grain of sand on a beach, always in danger of being washed away by the next wave. The search for any “bat patient zero” almost four years later would be a hopeless effort even before the deliberate sabotage of scientific processes by governments and malicious actors.
Yet in some twisted sense, we were lucky that the genome and evolution of sarbecoviruses are so intricate and reliant on recombination; it means that SARS-CoV-2's evolutionary history contains many overlapping stories. Stories that can provide a richer picture of what, when, where, why, and how it happened, akin to a set of contemporary witnesses with limited knowledge, or a set of partial fingerprints all over a crime scene.
Puzzling these genetic and phylogeographic fingerprints together, documenting what places they touched at what time, and listening to what story they have to tell allows researchers to draw a much larger picture from the mosaic knowledge of our ignorance.
The reality is that the observed viral ancestry of SARS-CoV-2 betrays any conceited notions we might have held about virologists — or gain-of-function research — bringing about the deadly chimeric pathogen. Scientists have the time and place of its birth and emergence narrowed down to a sufficient granularity to exclude any lab-related incidents. On top of that, we learned about some viral features that only natural selection, not human engineering, could bring forth.
Yet our understandable fears and suspicions of a “man-made” pandemic — albeit misplaced into too simple narratives or wrongfully projected onto convenient scapegoats in suspicious white coats— are also not entirely baseless.
As often with popular myths and sentiments, there is a grain of larger truths embedded in them.
Today, there is plenty of scientific evidence to make the case that collectively, we humans are not entirely innocent. Neither when it comes to fueling the forces that create these dangerous viruses, nor the reasons that make them spill over into us. In nature, no element acts alone. That includes us and the circumstances we create. As a now publicly smeared and unjustly character-assassinated zoologist once told me:
It is what we humans currently do to the planet and every other species on it that largely sets the conditions for zoonotic spillover events, and we drive them at an ever-accelerating pace.
What we can and should do about this sobering reality needs urgent societal discussion.
For that to happen, I believe that the era where false “gain-of-function” origin myths are wielded as weapons for persuasion, profit, or power has to end. Not only is this deeply unethical and scientifically irresponsible in 2024, but it has thus far proven counterproductive for scientific research, pandemic prevention efforts, and the public good.
In this article, I focused our attention on an interconnected set of scientific knowledge related to viral recombination, and it is important to mention that this — while very comprehensive — is just a mere facette of the larger origin picture. There are other comprehensive lines of investigation that this article could not go into, like epidemiology, which also points to viral emergence in the Huanan market and contradicts any research-related origin scenario. No single article can cover it all.
Understanding the origin story through the lens of viral recombination is incredibly powerful because it exposes the shallow ideas and trivial falsehoods about the virus thrown into our faces every day. It also shows us that nature is much more complex, nuanced, and vast than we commonly give it credit for. It would be wise to exert more caution and compassion before trying to shape every last ecosystem to our human whims.
A comprehensive body of scientific evidence has shown us that the immediate bat ancestor to SARS-CoV-2 came from one of the countless natural “gain-of-function labs” spanning the vast biodiverse Karst region from Yunnan in Southern China towards Myanmar, Laos, Thailand, Cambodia, Vietnam, and maybe even Malaysia in Southeast Asia. The lingering and promiscuous endemic viral elements in that enormous geographic region constantly mix and bring forth new chimeric combinations within their socially intricate reservoir hosts; while human activities and encroachment on bat territories stir the genetic cauldron ever faster.
Once a particularly combustible set of genetic elements produced a potential pandemic pathogen with broad host tropism, the legal and illegal mammalian wildlife industry likely became the maturing vessels through which the virus we now know as SARS-CoV-2 reached its final explosive form. From there, it was dragged in front of hundreds of immune-naïve future hosts visiting the largest wet market of one particular Chinese megacity well connected with the entire world.
The rest is history, one which we are currently bound to repeat.
Maybe after four years of political myth-making and societal inaction, it is time to face scientific reality. I certainly believe we’d be better off fighting for solutions rather than for who is to blame.
Because no matter what we want to believe, these natural chimeras will keep haunting us if science and society don’t come together to stop them.
[end]
Additional material:
Listen to an exclusive interview with Spyros Lytras and Jonathan Pekar about their groundbreaking SARS-CoV-1 and SARS-CoV-2 recombinant ancestry study:
Booklet version online:
Treacherous ancestry : simplebooklet.com
A word of caution about science communication versus peer-reviewed research and scientific reviews
In this article, I explained how multiple lines of scientific inquiry strongly exclude the possibility of gain-of-function research leading to COVID-19. Furthermore, the article highlights a body of literature warning about the clear and persistent danger of another zoonotic sarbecovirus spilling over into the human population.
Please consider that no single article can summarize all the valuable scientific contributions that scientists have made in support of investigating the origin question. It is a large and still-growing body of evidence spanning hundreds of papers that are mutually consistent with each other, the extant evidence, and the reality we live in.
I want deeply thank Prof. Alice Hughes, Prof. David Robertson, Prof. Linfa Wang, Dr. Alex Crits-Christoph and Dr. Jonathan Pekar for reviewing relevant sections of this article.
My humble contribution to this topic was to focus on explaining recent key scientific arguments through the lens of viral recombination for non-experts and also accurately present the conclusions scientists reached in their work by considering a large, detailed body of evidence.
While I did go to great lengths to avoid misrepresentations — such as having hours of expert interviews and article reviews by domain experts before I post — I cannot always control how my words will be interpreted. If there are uncertainties arising from my simplifications, omissions of brevity, or bad analogies, I advise you to first consult the primary literature for clarification rather than presume there is an obvious mistake in the science or reasoning of scientists.
At this point, the scientific consensus arising from a large body of evidence is pretty unequivocal on key issues surrounding the origins of SC2. However, this does not mean that there are no more questions to be asked or no more lessons to be learned on how to prevent SARS-CoV-3. That is why origin research continues, despite the “lab leak” question being mostly settled in the scientific literature.
Whenever there is a (fake or real) controversy about a scientific topic, I strongly believe that only the fruits of scientific inquiry — not a toxic combination of politics, media, and activists — will produce reliable knowledge and factual insights to act upon.
All that is missing currently is for some citizen defenders of an evidence-based worldview to do the legwork of closing the rift between science and society, for example by building bridges of accessibility where there are none. There is no infrastructure and no money in science communication, and science desks were the first to go in a crumbling newspaper economy. That is why I have to do this work without any financial aid or compensation (see motivation statement below); and against a lot of headwinds from emotionally engaged conspiratorial communities.
To be completely honest, the lack of accessible information and resources on this topic is not only a monetary issue, nor a coincidence in a broken information ecosystem. Many virologists have been generous and great communicators trying to compensate for public knowledge gaps before they were publicly smeared and character assassinated into disengagement (or even defamed) by motivated actors. Anti-science pressure organizations like US-right-to-know (USRTK) target outspoken virologists explicitly for that reason.
“They extract a high cost for free speech, they coerce the informed into silence”
— Nature Magazine editorial about the anti-science pressure group USRTK
I believe nobody in society should condone anti-science activism and abusive behavior. It would surely help if citizens stood up for scientists unfairly targeted by politicians once in a while. That virologists are bitterly disappointed in institutions, media, and politics that facilitate such attacks is an understatement. In the end, science and democracy can not function without a society that supports them.
So I encourage you and other defenders of a “weight-of-evidence”-based worldview to talk about and speak up for the scientific method, use it to dispel popular myths, or simply enjoy having understood something important about a dramatic world event so you won’t be a sucker for the next manipulative charlatan that comes along.
Motivation statement and copyright
As always, my hope and goals are to educate and equip citizens with conceptual tools and new perspectives to make sense of the world we inhabit.
This article took a lot of time and effort to conceptualize, research, and produce, actually almost irresponsibly so given that I do not monetize my scicomm here; and neither do the many scientists that so graciously gifted their time and expertise to help me and others understand their field of expertise.
I see this work as a public good that I send out into the void of the internet in hopes others will get inspired to act
You are also invited to deepen this work or just derive satisfaction from understanding our chaotic modern world a bit better.
So feel free to use, share, or build on top of this work, I just ask you to properly attribute (Creative Commons CC-BY-NC 4.0).
Cite this work:
Markolin P., “Treacherous ancestry. A phylogeographic hunt for the ghosts of SARS-CoV-2”, April 12, 2024. Free direct access link:
https://www.protagonist-science.com/p/treacherous-ancestry
Also, since this topic is close to my heart, I’d be happy to hear your thoughts
References:
Below I listed some of the key publications referenced in this article, with a short comment on what they are about for easy navigation and further reading.
On recombination in CoVs:
Sola I. et al., Annu Rev Virol., 2015 (recombination in CoVs is discontinuous with RdRp jumping around; deep mechanism of CoV transcription)
Patino-Galindo JA. et al., Molecular Biology and Evolution, 2021 (recombination patterns and frequency between viral families)
Boni MF. et al., Nature Microbiology, 2020 (recombination event between bats and pangolins not likely to have led to emergence of SARS-CoV-2, genetic elements of lineage have been circulating in bats for decades)
Klerk A. et al, Virus Evol., 2022 (Conserved recombination patterns across and within coronavirus subgenera)
Wells H. et al., Cell Host Microbe, 2023 (review article about CoV recombination, requirements and likely mechanisms)
Wang J. et al, Nature communications, 2023 (virus-sharing networks between bats highlight ample opportunity for CoV recombination in nature)
*recombination as “virus sex” is not accurate mechanistically: The exchange of genetic material between viruses is usually non-reciprocal, meaning the recipient of a genome portion does not act as donor of the replaced portion in the original source. In this respect, the term recombination does not have the same meaning in viruses that it does in diploid, sexually reproducing organisms wherein the exchange of genetic material between chromatids in the first meiosis division is reciprocal. (Perez-Losada M., Infect Genet Evol., 2015)
On the Furin-cleavage site
Andersen KG., Nature Medicine, 2020 (proximal origin noticing unusual FCS and structural predictions of changed glycosylation patterns)
Wu Y. et al, Stem Cell Research, 2021 (highlights FCS sites over wider CoV family)
Lavie M. et al., J Virol., 2022 (cleavage at S1/S2 and S2 viral processivity studies highlight the function of the FCS)
Jackson JB. et al., Nat Rev Mol Cell Biol., 2022 (FCS cleavage makes S1/S2 unstable, stabilizing mutation D614G. “This perplexing observation suggests that the acquisition of a furin-cleavage site by SARS-CoV-2 may have been a recent event.”)
Peacock T. et al., Nature Microbiology, 2022 (FCS cleavage of S1/S2 is essential in ferrets, pre-cleavage of the spike during viral egress enhances entry of progeny virions into TMPRSS2-expressing cells such as those abundant in respiratory tissue; and avoid endosomal IFITM proteins)
Fraser BJ. et al., Nature, 2022 (Spike processing by TMPRSS2; cuts extracellular spike in 3 places including S1/S2 if not pre-cleaved by furin to enable membrane fusion)
Garry R., PNAS, 2022 (FCS was not engineered and certainly not modeled after human EnAC)
Chaudhry MZ. et al., Virology, 2022 (FCS gets quickly lost in cell culture because of kinetics)
Vu MN. et al., PNAS, 2022 (synergistic interaction with the QTQTN motif proximal to the FCS plays a key role in infection and pathogenesis)
Sander AL. et al., Communications Biology, 2022 (high sequence diversity in S1/S2 and a proto-FCS site found in European SARS-related bat CoVs)
Alwine JC. et al., mSphere, 2023 (The FCS not lab adapted — initial SARS-CoV-2 isolates replicate poorly in traditional laboratory models — and not engineered — cleavage site loop length — best we can tell)
Neil SD. Cell, 2023 (Furin cleavage sites in intermediate host animals, but not bat reservoirs)
Steiner S. et al., Nature Reviews Microbiology, 2024 (recent review on viral entry and what is known about different mechanisms)
On the social lives of bats
Hughes AC. et al., Acta Chiropterologica, 2011 (Bat calls are destinct; rhinolophids use distinctive constant frequencies that can facilitate identification)
Vernes SC. et al., Phil. Trans. R. Soc. B., 2019 (vocalizations vary between species and include echolocation calls as well as social calls used either for parent–offspring reunions, territorial defence or maintaining group integration)
Irving A. et al., Nature, 2021 (bats are uniquely suited hosts and viral reservoirs)
Respicio JM. et al., Journal of Animal Ecology, 2024 (bat aggregation leads more negative and aggressive behavior)
Meyer M. et al., Nature Communications, 2024 (bat diversity and disease relationship in response to human encroachment and habitat change)
On Karst and caves
David Gillieson, Oxford Academic, 2005 (With a total area of about 400 000 km2, Southeast Asia contains some of the more extensive karst regions in the world. Many of these karst areas are of high relief with spectacular arrays of tower and cone karst. Many have now been inscribed on the World Heritage list in recognition of their unique geomorphology and biology)
https://en.wikipedia.org/wiki/South_China_Karst (500.000 km2)
On pre-pandemic Sarbeco seroprevalence in South-East Asia
Wang N. et al., Virol Sin, 2018 (October 2015 serum sampling from 218 residents in four villages in Jinning County, Yunnan province, China, 6 (2.7%) positive SARSr serology)
Li H. et al., Biosafety and Health, 2019 (pre-pandemic: 9 cases (0.6%) of SARS-CoV-1 positive serology from a snowball-sampled cohort of 1600 people in 3 provinces in Southern China between 2015–2017)
Sanchez CA et al., Nature communications, 2022 (65000 sarbeco-CoVs spillovers per year estimated)
Manning J. et al., Emerg. Inf Dis, 2022 (SC2 Elisa assays of pre-pandemic blood samples taken in Cambodia between 2005–2011 show 4–14% reactivity)
Evans TS. et al., Int J. Infect Dis, 2023 (neutralizing antibodies against various SC2r CoVs in Myanmar locals, high seroprevalence for rural (~20%) populations in 4 areas, 0% in city people, sampled pre-pandemic July 2017 to February 2020)
On sampling location, geographical distribution of SC2r genomes:
Li LL. et al., Emerg. Microb Infect. 2021 (RpPrC31, a viral recombinant ancestor, retrospectively discovered from bat samples in Yunnan)
Wacharaplusadee S. et al., Nature communications, 2021 (RacSC203, a cousin of SARS-2 discovered in Thailand containing S1/S2 insertion; also pangolin CoVs at wildlife checkpoint)
Delaune D. et al., Nature communications, 2021 (bat samples from 2010 cambodia find a close relative 92,6% of SARS-CoV-2 (RSHTT200), one bat was simultaneously co-infected with 4 viruses, 2 of them sarbecoviruses)
Lytras S. et al., Gen Evo., 2022 (recombination analysis and geographic distribution of Sarbecoviruses)
Temmam S. et al., Nature, 2022 (discovery of human-infectious SC2 relatives in Laos, “Banal” viruses and genome sequences)
Zhao S. et al., J Genet Genomics, 2022 (geogenomic distribution patterns)
Han Y. et al., Nature communications, 2023 (CoV sampling all over China)
Muylaert RL. et al., Nature communication, 2023 (landscape predictions and drivers of SC2 spillover)
Gilbert M. et al., Tropical Biomedicine, 2023 (SC2 related coronaviruses in Malaysia, no full genome sequencing yet but 99% similarity for a fragment)
Forero-Muñoz NR. et al., Virus evolution, 2024 (phylogeographic structure in bat hosts creates a landscape of selective pressure)
On moral panics
*moral panic: As a distinct species of collective behavior, moral panics represent contentious and intensely affective campaigns to police the parameters of public knowledge and morality. As such, they are necessarily dependent upon and constituted by claims-making, with interested parties historically seeking to actuate alarm by influencing the imagery and representations of the mainstream press. (Walsh JP, International Journal of Cultural Studies, 2020)
Recent preprints:
Hassanin A. et al., biorxiv, 2023 (phylogenetic divergence of close SC2 relatives between sub-tropical Northern Vietnam and tropical Southern Vietnam)
Crits-Christoph A. et al., biorxiv, 2023 (metatranscriptomics data analysis from Huanan market identifies wildlife species)
Pekar et al., biorxiv, 2023 (geospatial history and origins of SARS-CoV-1 and SARS-CoV-2, video interview here)
Si J. et al., biorxiv, 2024 (multi-species ACE2 adaptiveness, Sarbecos can infect humans from the get-go)
Video interviews
Video: Long-form discussion with bat ecologist Dr. Alice C. Hughes and the risk of SARS-CoV-3
Video: Long-form discussion with SAGO member Dr. Carlos Morel (SAGO is the WHO’s scientific advisory group on the origins of pandemics)
Bonus
Wow, I am surprised you scrolled all the end to the references. This curiosity (or due diligence) is of course rewarded with a little bonus information:
I have been working on an origins book with a twist that I believe will make some waves. It raises questions nobody else has yet dared to tackle: Where exactly does the “lab leak” myth come from, how did it move through society, and why were so many citizens susceptible to it? From the remote villages of the Lahu mountain tribes via chaotic conspiracy theorists and needy journalists to the halls of power and even the US presidency, THE LAB-LEAK MYTH will invite readers into my most challenging project yet; making sense of where the current rift between science and society really comes from.
Updates & Corrections
April 15:
An earlier draft misnamed Ebola as Flaviviridae, instead of Filoviridae. Also, SARS-CoV-2 has been incorrectly described as negative-strand RNA virus instead of positive. I want to thank Prof. Doug E Brackney for correcting the error.
April 12:
An earlier draft falsely read: “SARS-CoV-1 and MERS-CoV both caused an epidemic without an FCS” which has been corrected to “MERS-CoV and SARS-CoV-1 both caused epidemics, one with and one without an FCS”. I thank Twitter user CD8 for spotting the error.
An earlier version used the incorrect term “family” for CoVs and Sarbecoviruses; it has been corrected to “sub-family” and to “sub-genus” respectively. I want to thank Prof. Kristian Andersen for the corrections.
This is a great article, that I'll need to reread to really absorb. I hope you'll take "Zimmer-esque" (after Carl) as a compliment. The main lines I take away from a quick read are that
(1) Nature does its own gain-of-function research, it's called evolution. It has been doing so for a billion years, on a scale far beyond what we could imagine
(2) Even at the scale of bacteria and even viruses, there's a lot more genetic mixing than we thought even a few decades ago. Viruses have a lot of sex (metaphorically)!
(3) Human directed gain-of-function research isn't something unique or sinister, we're just trying to anticipate what nature might do next, at much smaller scale. We're not so smart that we can match the scale of the natural experiments.
(4) 1-3 make some people uncomfortable.
Great, invaluable in fact and much appreciated, I don't use the usual social media. I found you via a excellent commentator "BillyJoe on the excellent Neurologica blog Steven Novella runs.
There is a typo in the citations. "Bat calls are destinct" should be, of course, 'distinct'. That is in the "Social Lives of Bats" section. Like 911Truther grand conspiracies I expect this one to run for ever; I lost friends over that one and will on this one too. I am not going to enter into phony 'debates' mostly though.