 This is the proposal that you're investigating. It's known as Project Diffuse. And it's essentially a proposal to DARPA, the Defense Agency, to collect and manipulate bat viruses ostensibly for the purpose of predicting what kind of viruses might jump from bats to humans and cause future pandemics. And you summarize one of their proposals this way, that the scientists sought to insert furan cleavage sites at the S1, S2 junction of the spike protein to assemble synthetic viruses in six segments to identify coronaviruses up to 25% different from SARS and to select for receptor binding domains adept at infecting human receptors. And then you go with these bullet points and you list kind of one by one the characteristics that SARS-CoV-2, the ways in which it matches the viruses that are described in this proposal. So SARS-CoV-2 has a furan cleavage site, which has been discussed at length. And we can talk about a little bit more later. SARS-CoV-2 can be divided into six contiguous genomic pieces by the restriction enzymes BSAL and BSMBL, and orders for one of those enzymes, BSML, can be found in the documents. And the receptor binding domain appeared finally tuned for human ACE2 receptor. And then the genome has the genetic differences from SARS. So here's where I want to bring in Alex to help explain the significance of their plans to assemble the viruses specifically in those six segments using some very specific enzymes. Could you explain that for us a little bit more deeply, Alex? Absolutely. And one kind of interesting backstory is that the Diffuse Grant was proposed to the DARPA preempt call in 2018. And I had submitted or helped write in ACE7 as well. So this call is very intimately familiar to me. And in that context, what was distinct about Diffuse was their search for this motif that had never been seen in the SARS coronavirus before the fear and cleavage site, and their proposal to insert this thing that had never been seen before in the SARS coronavirus. And so the proposal to insert a fear and cleavage site instantly puts them down a particular laboratory path. So how do you insert a fear and cleavage site into a SARS coronavirus? The SARS coronavirus has this RNA genome, which you can think of as a really flimsy film of a genetic strand of RNA. And it breaks very easily. It's very hard to insert motifs inside of an RNA genome. So if you want to insert a fear and cleavage site, you have to build this much sturdier DNA copy of the virus. And in order to build a sturdy DNA copy of the virus, you have to build it one block at a time. We looked into the pre-COVID, well, I guess not we, but Valentin Britell and Tony Van Dongen these are bioengineers, Valentin and University of Pittsburgh and Tony at Duke University. They're bioengineers. They build these DNA blocks from scratch. So they looked into the methods for how people did this in coronaviruses pre-COVID and found that one of the most common methods for building these full-length DNA clones of a virus was to order these blocks and nip them at the end with these restriction sites that allow you to attach them one at a time until eventually you have that full-length DNA clone. And one of the diffuse PI's really popularized this technique. Ralph Beric wrote a paper on efficient reverse genetic systems. Well, that's a lot of words. It just means reverse genetics as you have this DNA clone that you're tinkering with to make these different RNA viruses. And in the process experimenting with the role between these genes and the manipulations of these genes like if you're in cleavage site on the function of the virus. So that was kind of the context we looked into this diffuse proposal and asked, okay, if they're inserting a fear in cleavage site they need a full-length DNA clone. If you're building a full-length DNA clone they're probably using methodologies like those that Ralph Beric popularized. Is there any evidence for this in the SARS-2 genome? We studied, we ran a meta analysis of all of these infectious clones and reverse genetic systems made before COVID and found that they all had this characteristic manner of manipulation. The researchers would look at the genome of the virus on the computer screen, look at where these cutting sites are and they move them around a little bit to make it easier to assemble these blocks. And they would move these cutting sites around with these mutations called silent mutations that change the genetic sequence without changing the resulting virus. So researchers before COVID would make these full-length DNA clones by using silent mutations to move these cutting sites that form these little notches you can use to attach chunks of DNA together to build that full-length clone. We looked into that and found that SARS-2 had this very unusual pattern of evenly spaced cutting sites that was consistent with how people built reverse genetic systems before COVID and that when you look closer at these cutting sites, yeah, this is the figure here, those cutting sites BSMB1 and BSA1 typically are randomly spaced throughout the genome. When you have randomly spaced things they don't often come with fragments of this similar size. But when you look at this visually, Valentin and Tonya just screamed at them, this is unusual and they have a lot of experience seeing what you see in the other, so SARS-CoV-2 is the top row here. All the other viruses from vanilla 20, 247 to SARS-1, their BSMB1 and BSA1 sites are randomly scattered and that's what you typically see when you upload a genome into this software that helps you see the cutting sites. Valentin and Tonya uploaded SARS-2, saw this unusual pattern and thought, wow, this is like an Ikea virus. This is something you could just order right from scratch and get this Allen wrench of BSMB1 and this screwdriver of BSA1 and stitch it together. So the question then was, what are the odds of this occurring in nature? And that's where I came in as a mathematical biologist to study the evolution of coronaviruses and try to get an estimate on the odds of this appearing in nature given that it is so consistent with a reverse genetic system. We found those odds were really low. So when you, the way we estimated it was looking at this longest fragment length, so this top figure here, the y-axis is the length of the longest fragment from cutting up a virus with any one or any pair of these enzymes. And then the x-axis is the number of fragments. When Ralph Berick wrote his paper on efficient reverse genetic systems, he proposed that red box is this idealized range for an efficient reverse genetic system. So more between five to eight fragments and ideally less than 8,000 base pairs long for the longest fragment because when you order these chunks of DNA online, 8,000 base pairs is a common cutoff. Now that said, the longer the fragment is, the more likely it is either contain a toxic element that prevents your ability to replicate it and clone it and make a lot of it. And it also reduces the odds of faithful assembly when you have a really long wiggly piece of DNA, it may not attach as faithfully or at high of rate as shorter fragments. So there's a preference to have smaller fragments in general, and that's what really was the lurking lab protocol constraints that led to this pattern of unusually even spacing in these restriction sites. So we looked at, we qualified the odds of seeing something as or more extremist SARS too just in the pattern of the spacing of these sites. And that was about a 1 in 1400 event. So one out of 14,000 is actually the most extreme case of a coronavirus that we've seen for these enzymes that can be used for the assembly of these reverse genetic systems. Then we look further at the silent mutations. That's where we found hotspots like glowing genetic dust right on these specific sites. And wherever they were moved around, they were moved around exclusively with the mutations bioengineers use. And that was our most significant finding. And so this is what the title of your paper is endonucleus fingerprint indicates a synthetic origin of SARS-CoV-2. So the fingerprint that you're referring to there is essentially the fact that there are, let me pull up that figure one more time, these rather evenly spaced like somewhat regularly length segments as opposed to in a naturally occurring virus, you would get much more variation in the length. You'd get some really short ones and some really long ones. And so it's just mathematically unusual for that, or maybe even mathematically impossible or highly improbable that that would occur in nature. I want to give credit to DARPA. They invented the internet and they rejected the fuse. So this is a diffused proposal of interest. The fuse proposal reveals the intentions of these researchers who previously had conducted a lot of gain-of-function research of concern. They would find these bat coronaviruses lurking in nature and they would swap around these parts of the bat coronavirus making these chimeras to then ask, oh, which chimera is the most infectious in people? And that question is modifying things found in nature to improve their transmissibility and the intention their search was for something more transmissible. So you would expect this research program to result in something more transmissible because that's what they were trying to make. And the fear and cleavage site specifically was well known before COVID that fear and cleavage sites are found in other viruses. And when they're found in other viruses, they're kind of this master key that allow the virus to unlock all sorts of different host cells and the enter into bat cells and human cells and dogs and whatever. So the fear and cleavage site was this master key. Why would you give that to a SARS coronavirus? That's anticipated to improve the transmissibility of this virus possibly improve the virulence as well if the virus is able to infect more tissues within the human body. So that the assertion of the fear and cleavage site was the gain of function research of concern. And that's- And just to be really clear for our listeners who may not have been following every detail of this, the fear and cleavage site, the reason that there's so much attention on it is because A, it is this SARS-CoV-2 is the only virus in its category of these SARS-like coronaviruses that has it. And B, that makes it particular humans, particularly susceptible to it. It eases the entry into human cells. Is that more or less the reason why the fear and cleavage site was like the first kind of red flag for people who thought that this might have originated in a lab? That's exactly right. The fear and cleavage site had never been observed before on a SARS coronavirus. And when we build the evolutionary tree of the SARS coronaviruses before COVID, we had 1,000 years of branches in this tree. And there wasn't a single indication that the SARS coronavirus existed anywhere in that tree. Yet, they were very prominent in the minds of virologists. And this is where I can speak as someone who's in that DARPA preempt community of accepted grants that we all knew that the biggest crux for these jump-capable quasi-species, the biggest crux for host switching is binding onto the receptors of a new host. Because new hosts have very different funky receptors in their cells. So how can a virus latch onto that and enter into the cell? The fear and cleavage site assists with that process and it was known to assist with that process. So virologists everywhere were looking for these things because they really wanted to find them. If you found a fear and cleavage site, in for instance, an influenza virus that passed through a chicken farm, which they did once in one chicken farm, out of all the chicken farms that had been infected with this, that was documented and there was explosive news because virologists understood the context. So the fear and cleavage site was a motif, this master key that virologists knew about what nature didn't or nature knew about, it would stumble upon it randomly but had not stumbled upon it in the 1,000 years of SARS coronavirus evolution that we'd seen. Yet then we find a SARS coronavirus with this master key in Wuhan, exactly where researchers proposed to give it to a SARS coronavirus. So that's the fear and cleavage site. That's the gain of function research of concern component and our work was focused more on the laboratory protocols that they had said they wanted to find all these wildlife coronaviruses. So we usually go out and you catch a bat, you sample its poop or its mouth or whatever and then you send that into a lab and you sequence it. When you sequence it, you get this long genetic code, 30,000 base pairs on a computer. Now, how can you go from that genetic code to then saying, this virus is more likely to infect people or less likely to infect people. For that, use all these laboratory techniques to construct parts of the virus or entire viruses. And our work was focused on that, both the rescuing of viruses from a genetic sequence using these reverse genetic systems or infectious clones. And that's a really remarkable thing to think about is that most viruses go from a virus to cell to a virus to cell to virus to cell and you can trace that back in time indefinitely to the origin of the virus, but infectious clones, they go from a virus to a cell to a poop sample in a bat, to a genome on a computer screen and then from that genome on a computer screen, you order these blocks and stitch them together and then we make from that DNA clone, this RNA, we electrocute a cell to form these holes in it an electro paraded cell and shove that RNA inside the cell, it starts making a virus. It's this sort of immaculate conception of modern biotechnology. And that was the most common way of rescuing coronaviruses from these wildlife samples where you could have the genome, but you may not have cultured the virus, the virus may have died while it was being transported to a lab, but you can still get the genome, you can still get the virus, you can still study it and that methodology was also in diffuse, but it was a bit more technical and subtle to find these fingerprints in the genome, but that's what we found. Hey, thanks for watching that clip from our new show, Just Asking Questions. You can watch another clip here or the full episode here. New episodes drop every week, so subscribe to Reason TV's YouTube channel to get notified when that happens or to the Just Asking Questions podcast on Apple, Spotify or any other podcatcher. See you next week.