2019-nCoV is NOT engineered — Text of an interview — Q&A.
1. Would you be able to explain the reasoning behind the authors’ claim of an “uncanny similarity”? What exactly is the role of a spike glycoprotein? What is an “insertion”? And why is the fact that the insertion is conserved across the 2019 nCoV structures, but not in other coronaviruses, a big deal? Also, why do researchers study the structure of spike glycoproteins in the first place?
1.A) Would you be able to explain the reasoning behind the authors’ claim of an “uncanny similarity”?
The authors are clearly ignorant and did amateurish work, or else they intended to mislead. Their claim of uncanny similarity is just ridiculous, based on apparently not understanding HIV. HIV’s gp120 attachment protein has regions that mutate very fast, protecting the receptor attachment inside from antibodies. This is a major mechanism of HIV’s evasion of the immune system. So in that region, you are likely to find almost any sequence. HIV’s gp120 also opens up and exposes the conserved attachment region in order to latch on. This is different from what we have seen in bat coronaviruses. To make those coronaviruses work that way would require copying most of it, not just itty-bitty things.
1.B) What exactly is the role of a spike glycoprotein?
Below you can see the structure of corona viruses. This is SARS but they all have these features.
The spike protein (surface glycoprotein) is named that because it projects, like a spike. When viewed in electron microscopy, it looks like a corona around a center.
The spike is what attaches to a cell. This allows the virus to infect relatively easily.
1.C) Also, why do researchers study the structure of spike glycoproteins in the first place?
Researchers study them because they attach to specific receptors. If humans have a matching receptor, the virus can cross into us.
Coronaviruses are:
Monopartite — Means’s that it is all one piece. Influenza has 7 or 8 separate RNA strands)
Linear ssRNA(+) — Means single strand of RNA, so it’s not a double helix) genome
27–32kb in size (the largest of all RNA virus genomes)
Next I show you the Wuhan 2019-nCoV genome with annotation. You can see that it clearly follows the same structure as SARS.
You can also see that the Wuhan 2019-nCoV genome at 29,903 bases is an average size virus of its type. Nothing extraordinary.
1.D) What is an “insertion”?
Below you see SARS on top and Wuhan below it.
There are 3 insertions marked with black bars. The longest insertion I see in the genome (not shown) is 15 nucleotides. All of the insertions are in the spike protein.
1.E) And why is the fact that the insertion is conserved across the 2019 nCoV structures, but not in other coronaviruses, a big deal?
First of all, this is just wrong. Those are not entirely unique.
In the supplemental Fig S2, the author mentioned that the “Bat-SARS Like CoV in the last row shows that insert 1 and 4 is very unique to Wuhan 2019-nCoV”. In fact, the Bat-SARS Like CoV, discovered in 2013, contains all the four insertions. Insertion 2 and 3 of 2019-nCoV are identical to Bat-SARS Like CoV. — Song Yang, BioRxiv comment on the paper.
For those insertions that are unique, it doesn’t mean anything much. All that means is that the 2019-nCoV virus is a slightly differently evolved virus. Think back to the range of lengths of this family of viruses above. The longest coronavirus so far is in the 32,000 nucleotide bases range. So, 2019-nCOV, at 29,903 could be more than 2000 nucleotides longer. If you compared it with SARS it would have large insertions of a total length of around 2000. (Note that Rous sarcoma virus is only about 3500 nucleotides long and it is a functional virus.)
But, what we see is that 2019-nCoV is very close to SARS, which is 29751 nucleotide bases. So 2019-nCoV is only 152 bases longer. So this is pretty straightforward.
It is always pretty rare that you can define a short sequence that isn’t found somewhere else.
2. Was the similarity the authors saw between the 2019 nCoV proteins and the HIV glycoproteins possible by chance alone? Is the fact that compared the sequences only with other viral sequences a problem?
2.A) Was the similarity the authors saw between the 2019 nCoV proteins and the HIV glycoproteins possible by chance alone?
Yes. They say this could be chance in their pre-print paper, and it is made clear by commenters on the pre-print. https://www.biorxiv.org/content/10.1101/2020.01.30.927871v1
2.B) Is the fact that they compared the sequences only with other viral sequences a problem?
If they had compared to everything possible, they would see that the these sequences were found in a variety of organisms and that HIV was not even in the top 100 hits.
3. What do you think of the implied claim that the virus is engineered?
I think that implied claim is utterly irresponsible, possibly to the point of criminal. It is tabloid fodder and I suspect the authors know this. What they have done is to essentially accuse the Chinese government plus a specific Chinese scientist of what would be a kind of war crime and a treaty violation.
If you are going to make extraordinary charges, you need extraordinary evidence. All the evidence presented is that this is a completely normal and naturally evolved relative of SARS that is present in bats.
To date, nature is the greatest terrorist.
4. What exactly would you expect to see in an engineered virus? Would you see a perfect match with much longer protein sequences from HIV? Would you be able to talk about the example the Simian HIV which you mentioned in your comment on Biorxiv.
First, let me discuss my qualifications in this area which are somewhat unique. I worked on engineering viruses in grad school. I clarified a problem for Novartis which was using an alphavirus as the basis for creating a set of live virus vaccines that were not harmful to humans. They were expressing the protein on the surface. This has been done so many times it is pretty routine now.
I worked on biodefense and I developed concepts prior to that. I published on biodefense back in 2006 through the West Point Counterterrorism Center. Bioterrorism and Biodefense for Americas Public Spaces and Cities. This linked chapter was intended for lay people to read.
I published an epidemic simulation software also in 2006 as well. If you want a more current version, just ask. An object simulation model for modeling hypothetical disease epidemics — EpiFlex
I have a policy paper as well in the area of biodefense.
Security in a goldfish bowl: the NSABB’s exacerbation of the bioterrorism threat.
4.A) What would I expect to see?
I can only answer that in the most general way. I would expect to see an assembly made from parts that had some unusual effect. That effect would generally be to cause very high mortality, or effects that were terrifying, such as the hemorrhagic symptoms of Ebola. (Bleeding from eyes, nose, pores.)
If it was a bioweapon made by a nation, I would expect it to have features that made it not very contagious, but highly virulent once contracted. Anthrax is, for this reason, an archetypal bioweapon. The reason is obvious. Biological organisms that are contagious can come back and infect you, not just others.
I would expect to see things like optimization for human expression which would leave a signature — various things we don’t see.
To get more specific would not be responsible. We know that, for instance, Islamists have tried to learn how to culture certain viruses. So have other fanatic religious groups. Aum cult sent people to Zaire during an Ebola outbreak.
4.B) Would you see a perfect match with much longer protein sequences from HIV?
Something like that, yes. To make a virus work differently so that it is a real bioweapon requires more than just a few nucleotides inserted here and there. However, it is completely possible to create a non-matching nucleotide sequence and still produce the same protein. I won’t go further because it would be irresponsible to shout this into the ears of our enemies, (who are also your enemies).
4.C) Would you be able to talk about the example of the Simian HIV which you mentioned in your comment on Biorxiv.
This link shows you structure of lentiviruses in general which is the same as the HIV-1 virus first shown above. Viralzone doesn’t have SIV, the simian version there.
SIV is the simian virus. HIV is the human one. SHIV is the simian virus engineered to express proteins from the human version. If you were to compare them you would see clearly defined sections as shown below.
Simian immunodeficiency virus (SIV) proviral, complete genome M30931.1
Human immunodeficiency virus 1 subtype E DNA, complete genome, isolate:93JP-NH1
