Thanks for the opportunity to talk about some of the engineering that went into these vaccines.
Dr. Novella is technically correct that a full-length spike protein is used in the Pfizer vaccine, but he clearly (at least to me) thinks that the protein is still functional, and he is very wrong about that. I chose "fully functional" on purpose (and not just as a Star Trek: TNG reference), because while it is full-length, it is non-functional. Here's why:
The mRNA vaccines all use a prefusion-stabilized construct. The Pfizer vaccine does use a full-length spike, but it is locked into its prefusion conformation by the introduction of 2 key proline amino acids to prevent the conformational change that allows the spike protein to function. This modification is described in detail here: https://cen.acs.org/pharmaceuticals/vaccines/tiny-tweak-behind-COVID-19/98/i38
The prefusion-stabilized spike protein is functionally incompetent and this is very important. While it is possible that the functions observed in the pseudovirus paper that I shared are preserved in the construct, this is highly unlikely. The conformational change is necessary for the fusion of virus particles to cells, and the spike protein is pretty useless without it. See more on the cell entry mechanisms here: https://journals.plos.org/plospathogens/article?id=10.1371/journal.ppat.1009225
I just don't really see how a prefusion-stabilized S protein could have these kinds of impacts on a cell or its levels of ACE2. Since the paper didn't study the prefusion-stabilized construct, I am definitely making a call here based on a gut feeling, but I feel pretty confident about it. Still, I'd be curious what a coronavirus entry expert would say about this.
Anyway another dimension to all of this that I didn't mention--and should have yesterday--is that none of the vaccines contain the machinery needed for SARS-CoV-2 budding. None of the vaccines express the virus membrane (M) or envelope (E) proteins, which are both important in the budding process. So, none* of the vaccines can produce even replication-incompetent virus particles that could somehow migrate to the lung. The only mechanism by which any of the prefusion-stabilized spike could even get to the lungs would be if input vaccine material was carried there, something that I really don't believe happens. And even then, it isn't a functional spike protein.
I should also note that the AstraZeneca vaccine DOES use a competent spike protein that is full-length. I still think that, injected intramuscularly, it would not get to the lungs, and it remains that no lung pathology has been observed in vaccinees. If Dr. Novella had leveled his points at that vaccine, though, rather than the Pfizer vaccine, I would think he had a better point.
*Except the weird, apparently replication-competent mutant of Sputnik V, but I'm not really considering that to be a vaccine--it certainly isn't the Sputnik V product as-designed.
I'd be happy to share my thoughts with him, but maybe better to do so over email. If you can put me in touch I'd be happy to chat with him.
Bottom line is I'd like to see the group that did that paper, or maybe Sean's group which is also working with pseudoviruses, directly compare the prefusion-stabilized construct with the functional construct. I am basing my feelings here on judgment regarding the functional consequences of prefusion stabilization but I'd rather have some experimental data in a similar experimental system to back up my opinion here.
Since the original paper here wasn't asking whether the vaccine constructs could have this effect, we don't really have the experiment, so I am mostly defaulting to "well, we don't see anything like this with any of the vaccines" as justification for my thinking.
This could be directly relevant to the development of intranasal/inhaled vaccines, so I really do want to see the experiment where the stabilized construct is compared with the natural version.
Steve Novella disagrees with you about the mRNA vaccines, specifically the Pfizer vaccine. He says it does in fact contain a complete and functional spike protein. (He doesn't say "functional," but I think he implies it.) https://sciencebasedmedicine.org/spike-proteins-covid-19-and-vaccines/?utm_source=rss&utm_medium=rss&utm_campaign=spike-proteins-covid-19-and-vaccines
Thanks for the opportunity to talk about some of the engineering that went into these vaccines.
Dr. Novella is technically correct that a full-length spike protein is used in the Pfizer vaccine, but he clearly (at least to me) thinks that the protein is still functional, and he is very wrong about that. I chose "fully functional" on purpose (and not just as a Star Trek: TNG reference), because while it is full-length, it is non-functional. Here's why:
The mRNA vaccines all use a prefusion-stabilized construct. The Pfizer vaccine does use a full-length spike, but it is locked into its prefusion conformation by the introduction of 2 key proline amino acids to prevent the conformational change that allows the spike protein to function. This modification is described in detail here: https://cen.acs.org/pharmaceuticals/vaccines/tiny-tweak-behind-COVID-19/98/i38
The prefusion-stabilized spike protein is functionally incompetent and this is very important. While it is possible that the functions observed in the pseudovirus paper that I shared are preserved in the construct, this is highly unlikely. The conformational change is necessary for the fusion of virus particles to cells, and the spike protein is pretty useless without it. See more on the cell entry mechanisms here: https://journals.plos.org/plospathogens/article?id=10.1371/journal.ppat.1009225
I just don't really see how a prefusion-stabilized S protein could have these kinds of impacts on a cell or its levels of ACE2. Since the paper didn't study the prefusion-stabilized construct, I am definitely making a call here based on a gut feeling, but I feel pretty confident about it. Still, I'd be curious what a coronavirus entry expert would say about this.
Anyway another dimension to all of this that I didn't mention--and should have yesterday--is that none of the vaccines contain the machinery needed for SARS-CoV-2 budding. None of the vaccines express the virus membrane (M) or envelope (E) proteins, which are both important in the budding process. So, none* of the vaccines can produce even replication-incompetent virus particles that could somehow migrate to the lung. The only mechanism by which any of the prefusion-stabilized spike could even get to the lungs would be if input vaccine material was carried there, something that I really don't believe happens. And even then, it isn't a functional spike protein.
I should also note that the AstraZeneca vaccine DOES use a competent spike protein that is full-length. I still think that, injected intramuscularly, it would not get to the lungs, and it remains that no lung pathology has been observed in vaccinees. If Dr. Novella had leveled his points at that vaccine, though, rather than the Pfizer vaccine, I would think he had a better point.
*Except the weird, apparently replication-competent mutant of Sputnik V, but I'm not really considering that to be a vaccine--it certainly isn't the Sputnik V product as-designed.
You might tell Steve that, or if you prefer I can. Steve was a guest at several cons I helped with and one that I chaired.
I'd be happy to share my thoughts with him, but maybe better to do so over email. If you can put me in touch I'd be happy to chat with him.
Bottom line is I'd like to see the group that did that paper, or maybe Sean's group which is also working with pseudoviruses, directly compare the prefusion-stabilized construct with the functional construct. I am basing my feelings here on judgment regarding the functional consequences of prefusion stabilization but I'd rather have some experimental data in a similar experimental system to back up my opinion here.
Since the original paper here wasn't asking whether the vaccine constructs could have this effect, we don't really have the experiment, so I am mostly defaulting to "well, we don't see anything like this with any of the vaccines" as justification for my thinking.
This could be directly relevant to the development of intranasal/inhaled vaccines, so I really do want to see the experiment where the stabilized construct is compared with the natural version.
Ha! Before seeing this comment, I posted one to Steve's post directing him to John's. Great minds...