Stuart Neil Profile picture
Apr 21 21 tweets 4 min read Twitter logo Read on Twitter
1. Contrary to popular belief, most virologists I know are not cowboys when it comes to safety. They prefer to use far more tractable surrogate systems when studying dangerous viruses. A case study to highlight the nuances of viral manipulation, GoF etc.
2. Imagine you have related 2 bat viruses. One is a transmissible deadly human pathogen and must be studied at BSL4, the other is antigenically distinct but has caused no disease in the handful of people known to have been infected or in experimentally infected mice.
3. Both viruses replicate fine in some human cell lines, but only the pathogenic virus replicates in primary human macrophages. The attenuated virus gets into these cells but fails to replicate because the inflammatory/interferon response it induces kills viral replication.
4. Previous studies had identified a protein encoded by the pathogenic virus, VPn, that binds to the human TBK1 protein and inactivates it, thereby inhibiting the interferon response. This interaction is thus a potential drug target for therapeutics.
5. When lining up the VPn protein you see clear differences between the sequence of these 2 proteins and you wonder if thus underlies why the ‘attenuated virus’ is non-pathogenic. How do you answer this question responsibly and cost-effectively -remember containment is expensive
6. The first easy set of experiments to do is to express both VPns on their own in cells and see if they can inhibit TBK1-dependent interferon responses. And you find that VPn of the attenuated virus is unable suppress TBK1 activity to the extent of the path virus – paper Fig1
7. You then do binding studies between the two viral proteins and TBK1 and find that the reason for the lack of activity in the attenuated virus is because its VPn can’t interact with VPn – paper Fig2
8. Now you turn your attention to why that is. You look at the differences between the 2 proteins and make individual reciprocal mutations and find one single amino acid change accounts for the difference in phenotype – paper fig 3
9. All the above experiments have involved no infectious virus, safe, but have included both ‘gain’ and ‘loss’ of function expts. Now you want to know if this one difference is what accounts for the path discrepancy between the viruses – you need to know this for drug development
10. Now you must do BSL4 experimentation because you can’t answer this question with non-replicative systems. What are the most important experiments to do to answer the question, and which should you do and not do?
11. First you need a ‘reverse genetic system’ for one or both viruses. This is the virus in a form that you can make mutations in the lab and then grow them to see how they behave. Making this is neither GoFRoC/ePPP or LoF, but it must be contained as you will generate the virus
12. Now in which virus should you make the mutation, and does it matter?
13. I would argue that you take the pathogenic virus and change the amino acid to that of the attenuated one and then infect human macrophages and mice, thus showing whether it is sufficient to attenuate the virus – i.e. Loss of function. fig 4 – write up and send to Nature.
1. The reviews come back. Revs are not satisfied. The single amino acid change attenuates the virus but not as much as the non-pathogenic virus. They want you to do the reciprocal experiment. Mutate VPn in the attenuated virus to confer virulence. Should you?
15. So this is clearly a GoFRoC experiment. Is it justified, is it important, can it be done another way?
16. Some will argue, it’s contained in a properly run BSL4 anyway, there really isn’t an issue and it is important for drug development. Others will point out that the virus will be antigenically distinct now with TBK1 evasive properties so you shouldn’t even think about it
17. Others further will argue that simply by identifying and publishing the initial safe ‘GoF’ lab experiments in figures 1 and 2 above you’ve just given a blueprint to terrorists and bad actors to generate a bioweapon.
18. You push back at reviewers 1 and 2 and say, no we aren’t going to do that experiment, but we can show that LoF mutant is fully pathogenic in TBK1 knock-out mice (new fig 5). Should the editor accept your paper?
19. Lastly imagine all the above is the same but now we are talking about two viruses that are already spreading in humans. Does that change anything.
20. These interlocking questions of the importance of the question, the timing and rationale, the capacity to perform the experiments safely, the downstream applications of the results, and whether there are alternatives are all important considerations.
21. These are the sort of nuanced discussion we need on this subject. Much more than ‘GoF bad/working on viruses bad’ which we get from some of the loudest voices.

• • •

Missing some Tweet in this thread? You can try to force a refresh
 

Keep Current with Stuart Neil

Stuart Neil Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @stuartjdneil

Apr 20
1. Something I haven’t seen much discussion of at #PathogensProject is about surrogate experimental systems for dangerous pathogens.
2. Predominantly from the retrovirus field, the establishment of lentiviral pseudotype systems (PLVs) for studying enveloped viral glycoproteins, as well as replication competent Rhabdovirus-based systems are cheap, easy and fast.
3. While there are limitations on some of the cell biology they can tell you about the dangerous virus of choice, they can give you most of the info you need to know including receptor, host range tropism, sensitivity to antibodies, vaccine escape etc.
Read 11 tweets
Apr 18
Opening statements (paraphrased)

Ratcliffe - Biden and those awful virologists are blocking the CIA agreeing with me
Feith - GoF is bad; and 3 sick WIV workers..
Lowenthal - intelliegence is diffcult and we may never get a straight answer.

So this is going to be revealing 🤦‍♂️
More paraphrasing

Feith - all scientists that dont agree about a lab leak are by definition conflicted.
Lowethal - intelligence services just dont have enough expertise in biomedical science and public health
Read 7 tweets
Apr 15
1. The complexity of bat sarbecovirus-S/ Rhinolophus ACE2 interactions. Something you see stated all the time from LLers is that SARS2 is 'human-adapted'.
2. This notion, in part, is based on the SC2 spike receptor binding domain (RBD) binding and using the ACE2 of the ‘presumed’ bat reservoir (R affinis) much less efficiently than human ACE2 to enter cells.
3. What is demonstrated by @SystemsVirology and co here is that two of the BANAL viruses (B52 and B236 – the former has an RBD that is the same as SC2) have quite distinct tropisms for ACE2 orthologues from different SE Asian Rhinolophus sp.
Read 19 tweets
Mar 8
1. Revisting this famous section from the DEFUSE grant proposal because I think we will be hearing a lot about it again. And let’s analyze what it actually says and compare it to the mantra of ‘they propsed to insert FCSs in bat CoVs”
2.Firstly thevirology because even 3 yrs, the authors of high profile books don’t get it (or don’t want to). For cell entry CoV spike proteins undergo at least 2 proteolytic cleavage steps by host cell protease enzymes. One is at the S1/S2 boundary, the other at the S2’ site.
3.These processing events allow the S2 fusion peptide to be exposed and promote fusion of the viral and cellular membranes.
Read 33 tweets
Feb 28
1: The mark of any decent hypothesis is that it needs to account for all the existing evidence without exception.
2.That means accounting for the early cases (both the HSM-linked and unlinked cases) clustering with very high statistical significance with the market.
3.It needs to account for why both lineages (A and B) both have the same HSM association, and why the evolutionarily older lineage A appeared LATER than the younger lineage B with no transitional genomes that stand up to scrutiny.
Read 12 tweets

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Don't want to be a Premium member but still want to support us?

Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!

Ethereum

0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy

Bitcoin

3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us on Twitter!

:(