I think these headlines, possibly purposefully, overlook the possibility that vaccine antibodies eased acquisition of infection induced antibodies to the point of being subclinical events.
That is, a vaccinated person could be exposed to a variant for which they are partially protected, not notice clinical symptoms, while the person's immune system still remembers the details of the variant.
As an iterative process, this is akin to software updates (for a running immune system).
So to say that variant A.30 isn't covered by the mRNA induced spike protein antibodies is a very different statement than to say it's able to evade an immune system primed by mRNA antibodies which then re-entered daily life and thereby was able to continuously adapt to variants.
It's also worth noting that the most people still had antibodies effective against A.30-- just not as effective. So the existing antibody response would be expected to slow infection and then the body is primed to broaden the variety and quantity of antibodies after infection.
It's only a small portion of those receiving the AZ vaccine whose serum had no significant activity against A.30-- all of those with the mRNA vaccine did and most of those receiving AZ did.
> It's only a small portion of those receiving the AZ vaccine whose serum had no significant activity against A.30-- all of those with the mRNA vaccine did and most of those receiving AZ did.
How about those that received no vaccine, but have previously recovered from Covid? Is it an insignificant number because I rarely hear it discussed.
I have an open question that I've never been able to find the answer to:
What are the limits to the memory of our immune system? Surely there are limits, to the extent that our blood can only carry N number of Memory B cells, and those cells that still exist carry some degree of senescence, and we need M of them to mount an timely response to a recurrence of infection, how many protein antigens can our immune system actually remember and respond to over long periods of time?
Except that data from the UK shows that’s precisely what’s not happening. Those who were immunologically naive when receiving the vaccine continue to not have N protein antibodies after subsequent infection. That is they did not gain new immunity that would have been gained in the reverse order of events.
It depends if the person who got vaccinated was exposed to the virus before the vaccination. If not, it is very likely that these variants will escape the vaccine induced immunity due to original antigenic sin, especially considering that mRNA vaccines only stimulate the production of spike ABs and not other ABs such as nucleocapsid ABs which target proteins that do not mutate as much as spike.
Antibodies is also just one part of the immune response, memory T cells are just as important.
We know the vaccines are not sterilizing so you can still catch and spread it. They do reduce the likelihood of hospitalization and that seems to be about that.
They do a lot more than that. A lot of the symptoms (coughing, sneezing, etc.) increase infectivity. The vaccines reduce not only the likelihood you'll get symptoms at all but if you do they reduce both the longevity and severity.
Saying "well you're just less likely to go to the hospital" underplays their importance and gives unnecessary credence to the completely false notion that if you don't get vaccinated the only person you're potentially hurting is yourself.
> We know the vaccines are not sterilizing so you can still catch and spread it
Unfortunately US media and government messaging at the beginning of this year touted the vaccines effectiveness as we historically understood vaccines to be, which was relatively sterilizing and a preventative. When Delta started infecting the fully vaccinated, it underscored the skepticism around these vaccines.
Ironically, right now, I know more people in my social circle that have had breakthrough cases then people who had Covid prior to vaccine’s release.
A.30 is basically dead, well outcompeted by Delta etc. The UK don't report on it now.
"VUI-21FEB-01 (A.23.1 with E484K), VOC-21FEB-02 (B.1.1.7 with E484K), VUI-21MAR-01, (B.1.324.1 with E484K), A.30, B.1.633, B.1.214.2 and B.1.1.7 with S494P have not been observed in the UK or within the international GISAID dataset within the last 12 weeks. These variants are no longer included in the data update."
The biggest pressure on variant mutation right now is vaccine-evasion - so the reason the Delta variant has out-competed it is likely because the Delta variant does an even better job of evading the vaccine.
In general, I'm disappointed that we are not rolling out updated boosters with variant protection. Flu vaccines are updated each year - and here we are almost two years from the initial outbreak, and we're still vaccinating people against the original form of the virus - which isn't even spreading anymore
That's not to say that it's not still lifesaving to get the current vaccine - but F - we can do better!
I've not seen much reported in the way of clinical trials / peer-reviewed publications for modified vaccines that target (the spike protein of) these variants of concern that are actively circulating.
Is there ongoing work here I'm missing? What progress has been made? I'm somewhat surprised that almost a year on from the initial vaccination roll-out in the UK we're now dishing out boosters which are .. identical to the vaccinations given out initially.
That is a fantastic question. We are still vaccinating & boosting against the original variant. Would "the experts" please demonstrate that vaccines against variants are effective, given that original antigenic sin is a long known and well documented phenomenon?
They haven't found original antigenic sin so far. There's a couple links with some data (although I couldn't find a pre-print or published article right now) in this comment: https://news.ycombinator.com/item?id=29005187
I work at an immunology lab which is next door to most of the Oxford vaccine group.
I think the quick (and overly simplistic) answer to this question is that models (e.g. convolutional networks) that predict which chunks of viral (or human) proteins are displayed to immune cells (HLA presentation), along with other kinds of models (like those predicting crossreactivity of T cell receptors against mutated epitopes) are really primitive (mostly because of low quality training datasets, it's not really a hard problem like e.g. protein folding).
Hence, rushing a new vaccine is not easy as there is a lot of labwork to do. As a matter of fact, all COVID vaccine designs had essentially the same payload (the whole spike protein). Modern subunit vaccines would typically include only little chunks of the spike (to increase efficiency and avoid side effects). But this was not trivial to do quickly without good in silico models.
Why not just update the vaccine with the corresponding sequence from the Delta variant? Isn't the Delta variant today a better baseline than the ancestral virus?
Or do a mix - first shot ancestral, second shot Delta.
Moderna and Pfizer tested variant-specific (B1.351) boosters and found they weren't significantly more-effective than a third dose of the original vaccine.
https://www.nature.com/articles/s41591-021-01527-y
>found they weren't more-effective than a third dose of the original vaccine
This goes against the paper:
>A boost with mRNA-1273.351 appeared to be more
effective at neutralization of the B.1.351 virus than a boost with mRNA-1273, evidenced by the
higher mean GMT levels in the Part C cohort 1 participants (1400) than the GMT Part B
participants (864) against the B.1.351 virus. Additionally, the difference between the wild-type
and B.1.351 assays at day 1 dropped from 7.7-fold prior to the boost with mRNA-1273.351 to
2.6-fold at 15 days after the boost.
Thank you for posting this though -- I was looking for it earlier for my own comment and couldn't find it! Bookmarking now.
Novavax has also been working on a Beta variant booster (and I believe has moved on to a Delta booster that hasn't started NHP trials yet) -- here's some info from a presentation: https://www.novavax.com/sites/default/files/2021-05/NVAX-WVC... (It's slides 17 and 18)
Pfizer/BioNTech definitely has a Delta booster getting ready for NHP trials, but my Google Fu is lacking today, apparently.
Even if A.30 was "more contagious", like Alpha, it isn't as contagious as Delta and thus would get outcompeted.
There's a lot of variants of concern. Its important to keep an eye on new variants, but don't worry about them until they cross the 1% or 5% mark. At that point, you can better theorize that they might be outcompeting the dominant strain.
You _really_ don't know that something is outcompeting until it reaches 50%. But by then its too late. The 1% or 5% points are still a month or two in advance of the 50% crossing, in both Alpha and Delta.
So that's where I keep my attention: at the 1% and 5% points. A variant gotta get to 1% before it can reach 50% of the world ya know?
> Is it possible that the heavy mutations detected on the A.30 variant also make it less contagious?
That wouldn't surprise me. The spike is central to how SARS-CoV-2 is so infectious. If the virus mutates the spike too much, it might evade the vaccines, but at the cost of being less infectious.
What's surprising is this research says, in vitro, A.30 is more infectious.
Regardless, as you and others say, in real world conditions A.30 doesn't appear to out compete Delta.
>Is it possible that the heavy mutations detected on the A.30 variant also make it less contagious?
100%. This strain hasn't been seen since May. It has already died out because of Delta.
All of these discussions are academic -- we really are only talking about if a Delta variant can gain these mutations and really wreck havoc. And the answer is "probably not" (for a while, anyways)
the results in the paper suggest, the mutations facilitate viral mechanisms of entry and antibody evasion.
efficacy of Oxford, Pfizer, and hetero-innoculation with both were examined, indications were given of reduced efficacy of either vaccine alone vs CoV A.30. The combination of both vaccines[hetero-inocculation] was associated with higher efficacy vs A.30
the mutations appear in vitro; to confer inhibition of vaccine induced antibody binding; enhanced cellular entry; enhanced pulmonary[lung] trophicity[targeting]
relative efficacies of moderna; Johnson&johnson; vs Oxford; Pfizer were not examined.
genuinely asking, how do we actually know the delta variant is real, and is the dominant variant causing the current cases?
do people actually get tested for specific variants, or is it just a generic covid test?
It depends where you live. The UK was an early[0] leader in sequencing a lot of their cases. They hit 600,000 sequenced cases in July[1].
That said, if you just want to know what proportion of the population has delta, you don't need to sequence that many cases. The point of sequencing a large fraction of cases is to catch emerging variants, that wouldn't be likely to be caught otherwise. But to know whether Delta is 30%, 50%, or 70% of cases, you just need a good random sample, not a particularly large one, and I imagine you can do that at the labs where the PCR tests are being run.
(at one point, I recall that the UK had a test that matched on three distinct sites on the virus, and they found that one of those sites would turn up negative for a particular variant, so they were able to use that as a proxy for the spread of the variant. I don't remember if that was Alpha or Delta though, and obviously it's no replacement for sequencing)
NZ had until recently sequenced 100% of its cases, and used that sequencing information to map person to person infection linkages. This included all of the people in the mandatory 2-week quarantine at the border. So in NZ, yes, every case was tested for specific variants. Delta does seem to be real.
Some small subset of COVID tests are sent to larger labs to perform a more detailed genetic test to determine specific variants. The prevalence of different variants within that set is extrapolated to the population of the region from which it originated.
If the lab finds that 87% of the samples from Adam's College are of the Omega Mu strain, then it's assumed 87% of total cases at Adam's College to be Omega Mu variant.
In Washington state, a percent of positive tests are sent to the dept of health for variant sequencing. You can see the published results by week. Delta has outcompeted every other variant for several months. All covid is essentially delta.
This is a speed-up that pre-dates the mRNA trick Moderna had been working on, a virus from chimps (so humans aren't immune to it because it wouldn't infect humans normally, yet chimps are similar enough that the virus can get into a human cell) but hollowed out to make whatever you want instead of more copies of itself. You put any payload inside it and it'll re-program the patient's cells to make that for a brief period until the immune system cleans up the mess. This vaccine uses the spike protein as payload to train your immune system.
Is it also the case that the human getting vaccinated build up immunity to the carrier virus and thus booster shots might become less effective? Would that be a reason to get a booster shot from something other than AstraZeneca?
IIUC, when COVID first gained notoriety, there was a large public effort to solve its molecular structure [0].
Articles like this seem to imply that they've solved the structure of numerous variants, making me think we (collectively) can do it much more quickly now.
If that's really true, what's changed? Does knowing the structure of other COVID variants give us a massive head start? Or has there been a massive investment in computing power for solving COVID-variant stuctures?
Famously, once the novel coronavirus was sequenced, Moderna had their mRNA sequence ready just a few days later. This was well-covered in the press.
All the time from then to general availability of their vaccine was spent on testing and scaling.
This is not exactly what you asked. But if the question is whether we have a speed advantage on vaccines because of greater understanding of viruses now, the answer is yes. It’s one reason the COVID-19 vaccine was the fastest in history.
Should it become necessary to separately vaccinate against coronavirus variants, the same advantage would confer. And vaccine makers have said they are monitoring variants and ready to go, if necessary.
The question is, will it be necessary? And how much testing will variant vaccines need to undergo?
No, there are not solved structures for these variants (in general). They're using computer modeling.
Even if there were, changes of a single amino acid that don't induce significant conformational (shape) change are the sort of thing that you'd need a very good structure to discern, since they're mostly on the surface of the spike, and probably spin around a lot.
The most relevant thing to know is what a mutation / mutations looks like in the context of antibody binding, but that's even harder to get a good structure. It's not something that can be done quickly.
when the focus is on individual molecules it is about a statistical measure of bond angle, distance, resonanceforms, charge distribution, and rotational freedom.
this ends up looking like the quantum physics conundrum where there is a central tendency rather than certainty, until you pin it down in some way.
this is why we have a vaccine to begin with, the realization that there are conformational frequencies to the S protien, and a particular locus in the sequence is associated with the conformational change that sequesters the binding domain of S protien. thus the 2proline version of S was used in subsequent vaccines, allowing the ACE2 binding domain to remain exposed to the PAMP recognition machinery.
it takes months or weeks depending on how well the supervising scientist receives a request, it used to take years and decades in many cases.
Computational docking (figuring out what drug molecules bind strongly to) is still not solved. In order to find drugs that 1. work well and 2. aren't toxic we need robust docking estimators that work well over a wide variety of proteins. My understanding of the space (which comes from undergrad comp bio, so please someone feel free to correct me) is that just having a molecule that binds well to your target doesn't mean it's going to work well as a drug, because there are delivery and toxicity concerns.
at least for identifying variants, you don’t necessarily need to know the full (highly complex) structure of each variant, just the relevant diffs from the original variant, and computer modeling to tease out whether the diffs are significant enough to warrant a ‘variant’ label (along with epidemiological data, of course). once identified and labeled, pcr can give you enough information (via the differing prevalences of short sequences of dna/rna that act like unique fingerprints) to tell the variants apart without needing to know anything about the structure.
yes to all. the early days of biochemistry were heavily dependent on x-ray crystallography, meaning the subject structure was in crystalline form, meaning there was a bias toward understanding of structures that facilitated crystalization. sequencing and structural studies[folding] were slow tedious wetlab efforts until mid 90s when computing power was of caliber, and shared across the internet, we now have such computation cheaper smaller, and locally.
like any jigsaw puzzle, the more pieces correctly placed the easier it becomes to progress.
Readit News
That is, a vaccinated person could be exposed to a variant for which they are partially protected, not notice clinical symptoms, while the person's immune system still remembers the details of the variant.
As an iterative process, this is akin to software updates (for a running immune system).
So to say that variant A.30 isn't covered by the mRNA induced spike protein antibodies is a very different statement than to say it's able to evade an immune system primed by mRNA antibodies which then re-entered daily life and thereby was able to continuously adapt to variants.
It's only a small portion of those receiving the AZ vaccine whose serum had no significant activity against A.30-- all of those with the mRNA vaccine did and most of those receiving AZ did.
How about those that received no vaccine, but have previously recovered from Covid? Is it an insignificant number because I rarely hear it discussed.
What are the limits to the memory of our immune system? Surely there are limits, to the extent that our blood can only carry N number of Memory B cells, and those cells that still exist carry some degree of senescence, and we need M of them to mount an timely response to a recurrence of infection, how many protein antigens can our immune system actually remember and respond to over long periods of time?
https://en.m.wikipedia.org/wiki/Original_antigenic_sin
That is blatant fear mongering. :)
We know the vaccines are not sterilizing so you can still catch and spread it. They do reduce the likelihood of hospitalization and that seems to be about that.
Saying "well you're just less likely to go to the hospital" underplays their importance and gives unnecessary credence to the completely false notion that if you don't get vaccinated the only person you're potentially hurting is yourself.
That’s a huge difference, and not just reducing the likelihood of hospitalisation.
Unfortunately US media and government messaging at the beginning of this year touted the vaccines effectiveness as we historically understood vaccines to be, which was relatively sterilizing and a preventative. When Delta started infecting the fully vaccinated, it underscored the skepticism around these vaccines.
Ironically, right now, I know more people in my social circle that have had breakthrough cases then people who had Covid prior to vaccine’s release.
"VUI-21FEB-01 (A.23.1 with E484K), VOC-21FEB-02 (B.1.1.7 with E484K), VUI-21MAR-01, (B.1.324.1 with E484K), A.30, B.1.633, B.1.214.2 and B.1.1.7 with S494P have not been observed in the UK or within the international GISAID dataset within the last 12 weeks. These variants are no longer included in the data update."
In general, I'm disappointed that we are not rolling out updated boosters with variant protection. Flu vaccines are updated each year - and here we are almost two years from the initial outbreak, and we're still vaccinating people against the original form of the virus - which isn't even spreading anymore
That's not to say that it's not still lifesaving to get the current vaccine - but F - we can do better!
TL;DR: it's just a bit worse than the Beta variant.
[0] https://covariants.org/per-variant
Is there ongoing work here I'm missing? What progress has been made? I'm somewhat surprised that almost a year on from the initial vaccination roll-out in the UK we're now dishing out boosters which are .. identical to the vaccinations given out initially.
https://en.wikipedia.org/wiki/Original_antigenic_sin
https://www.jstor.org/stable/985534
There's plenty of evidence that current vaccines work against variants:
- https://www.cdc.gov/mmwr/volumes/70/wr/mm7034e4.htm
- https://www.nejm.org/doi/full/10.1056/NEJMoa2108891
- https://www.cell.com/cell/fulltext/S0092-8674(21)01057-6#%20
> given that original antigenic sin is a long known and well documented phenomenon?
This could apply both to vaccines and natural infections. Not sure what you are trying to say?
I think the quick (and overly simplistic) answer to this question is that models (e.g. convolutional networks) that predict which chunks of viral (or human) proteins are displayed to immune cells (HLA presentation), along with other kinds of models (like those predicting crossreactivity of T cell receptors against mutated epitopes) are really primitive (mostly because of low quality training datasets, it's not really a hard problem like e.g. protein folding).
Hence, rushing a new vaccine is not easy as there is a lot of labwork to do. As a matter of fact, all COVID vaccine designs had essentially the same payload (the whole spike protein). Modern subunit vaccines would typically include only little chunks of the spike (to increase efficiency and avoid side effects). But this was not trivial to do quickly without good in silico models.
Or do a mix - first shot ancestral, second shot Delta.
This may of course change with new variants.
This goes against the paper:
>A boost with mRNA-1273.351 appeared to be more effective at neutralization of the B.1.351 virus than a boost with mRNA-1273, evidenced by the higher mean GMT levels in the Part C cohort 1 participants (1400) than the GMT Part B participants (864) against the B.1.351 virus. Additionally, the difference between the wild-type and B.1.351 assays at day 1 dropped from 7.7-fold prior to the boost with mRNA-1273.351 to 2.6-fold at 15 days after the boost.
Thank you for posting this though -- I was looking for it earlier for my own comment and couldn't find it! Bookmarking now.
Novavax has also been working on a Beta variant booster (and I believe has moved on to a Delta booster that hasn't started NHP trials yet) -- here's some info from a presentation: https://www.novavax.com/sites/default/files/2021-05/NVAX-WVC... (It's slides 17 and 18)
Pfizer/BioNTech definitely has a Delta booster getting ready for NHP trials, but my Google Fu is lacking today, apparently.
Is it possible that the heavy mutations detected on the A.30 variant also make it less contagious?
There's a lot of variants of concern. Its important to keep an eye on new variants, but don't worry about them until they cross the 1% or 5% mark. At that point, you can better theorize that they might be outcompeting the dominant strain.
You _really_ don't know that something is outcompeting until it reaches 50%. But by then its too late. The 1% or 5% points are still a month or two in advance of the 50% crossing, in both Alpha and Delta.
So that's where I keep my attention: at the 1% and 5% points. A variant gotta get to 1% before it can reach 50% of the world ya know?
That wouldn't surprise me. The spike is central to how SARS-CoV-2 is so infectious. If the virus mutates the spike too much, it might evade the vaccines, but at the cost of being less infectious.
What's surprising is this research says, in vitro, A.30 is more infectious.
Regardless, as you and others say, in real world conditions A.30 doesn't appear to out compete Delta.
100%. This strain hasn't been seen since May. It has already died out because of Delta.
All of these discussions are academic -- we really are only talking about if a Delta variant can gain these mutations and really wreck havoc. And the answer is "probably not" (for a while, anyways)
efficacy of Oxford, Pfizer, and hetero-innoculation with both were examined, indications were given of reduced efficacy of either vaccine alone vs CoV A.30. The combination of both vaccines[hetero-inocculation] was associated with higher efficacy vs A.30
the mutations appear in vitro; to confer inhibition of vaccine induced antibody binding; enhanced cellular entry; enhanced pulmonary[lung] trophicity[targeting]
relative efficacies of moderna; Johnson&johnson; vs Oxford; Pfizer were not examined.
That said, if you just want to know what proportion of the population has delta, you don't need to sequence that many cases. The point of sequencing a large fraction of cases is to catch emerging variants, that wouldn't be likely to be caught otherwise. But to know whether Delta is 30%, 50%, or 70% of cases, you just need a good random sample, not a particularly large one, and I imagine you can do that at the labs where the PCR tests are being run.
(at one point, I recall that the UK had a test that matched on three distinct sites on the virus, and they found that one of those sites would turn up negative for a particular variant, so they were able to use that as a proxy for the spread of the variant. I don't remember if that was Alpha or Delta though, and obviously it's no replacement for sequencing)
[0] https://cen.acs.org/analytical-chemistry/sequencing/200000-c...
[1] https://www.gov.uk/government/news/uk-exceeds-600000-covid-1...
US does statistical sampling and also prevents patients from being told about variants, https://news.ycombinator.com/item?id=28419280
https://www.cdc.gov/coronavirus/2019-ncov/variants/variant-s...
https://www.cdc.gov/coronavirus/2019-ncov/variants/cdc-role-...
If the lab finds that 87% of the samples from Adam's College are of the Omega Mu strain, then it's assumed 87% of total cases at Adam's College to be Omega Mu variant.
See page [pdf] 5 on this: https://www.doh.wa.gov/Portals/1/Documents/1600/coronavirus/...
It is updated every Wednesday
I've been watching the delta variant take over the world over the past several months here: https://covariants.org
(dark green is delta)
There is also a way to differentiate on some RT-PCR tests (but not all), depending on how they are made.
Dead Comment
ChAdOx1 nCoV-19 is AstraZeneca
BNT162b2 is Pfizer/BioNTech
This is a speed-up that pre-dates the mRNA trick Moderna had been working on, a virus from chimps (so humans aren't immune to it because it wouldn't infect humans normally, yet chimps are similar enough that the virus can get into a human cell) but hollowed out to make whatever you want instead of more copies of itself. You put any payload inside it and it'll re-program the patient's cells to make that for a brief period until the immune system cleans up the mess. This vaccine uses the spike protein as payload to train your immune system.
https://youtu.be/n8FbMY-quW4
Deleted Comment
IIUC, when COVID first gained notoriety, there was a large public effort to solve its molecular structure [0].
Articles like this seem to imply that they've solved the structure of numerous variants, making me think we (collectively) can do it much more quickly now.
If that's really true, what's changed? Does knowing the structure of other COVID variants give us a massive head start? Or has there been a massive investment in computing power for solving COVID-variant stuctures?
[0] https://foldingathome.org/diseases/infectious-diseases/covid...
All the time from then to general availability of their vaccine was spent on testing and scaling.
This is not exactly what you asked. But if the question is whether we have a speed advantage on vaccines because of greater understanding of viruses now, the answer is yes. It’s one reason the COVID-19 vaccine was the fastest in history.
Should it become necessary to separately vaccinate against coronavirus variants, the same advantage would confer. And vaccine makers have said they are monitoring variants and ready to go, if necessary.
The question is, will it be necessary? And how much testing will variant vaccines need to undergo?
Even if there were, changes of a single amino acid that don't induce significant conformational (shape) change are the sort of thing that you'd need a very good structure to discern, since they're mostly on the surface of the spike, and probably spin around a lot.
The most relevant thing to know is what a mutation / mutations looks like in the context of antibody binding, but that's even harder to get a good structure. It's not something that can be done quickly.
this ends up looking like the quantum physics conundrum where there is a central tendency rather than certainty, until you pin it down in some way.
this is why we have a vaccine to begin with, the realization that there are conformational frequencies to the S protien, and a particular locus in the sequence is associated with the conformational change that sequesters the binding domain of S protien. thus the 2proline version of S was used in subsequent vaccines, allowing the ACE2 binding domain to remain exposed to the PAMP recognition machinery.
it takes months or weeks depending on how well the supervising scientist receives a request, it used to take years and decades in many cases.
Deleted Comment
like any jigsaw puzzle, the more pieces correctly placed the easier it becomes to progress.