Study Suggests That Moderna and Pfizer Vaccine Selection Triggered Disease Enhancement in Delta and Its Spread
Scientists are careful to hedge on the political ramifications of their findings, but publish anyway
Researchers at Boston University have reported that individuals vaccinated with the Pfizer or Moderna vaccine have no immunity to more recent variants; in fact, v2 and v3 of the Beta variant had escaped humoral immunity from these vaccines (Full text). This escape no doubt helped Beta to dominate early via a process called vaccine selection. Beta has been replaced by a succession of newer variants, the latest being the rise of Omicron, which may be replacing Delta.
How Delta Rose to Prominence May Also Involve Vaccine Selection
Using an index called the Transmissibility Index, a different research team in France led by Jacques Fantini analyzed millions of SARS-CoV-2 sequences. They have found that while Delta arose before mass vaccination, the combined features of the virus - affinity, kinetics, surface potential, makes it stand out as a variant that very likely acquired new characteristics that led to facilitated cell entry both with the help of facilitating antibodies and, remarkably, without.
The researchers knew that enhancing antibodies reinforce the binding of the spike trimer to the host cell membrane by clamping the N-terminal domain (NTD) to lipid raft microdomains. The opening of the spike protein demasks the receptor binding domain; once this is accomplished, electropotential makes cell fusion inevitable. The more negative the viral surface potential, the stronger the electrochemical attraction to the cell surface.
The NTD is also targeted by neutralizing antibodies, and therefore the data from the Fantini team point to a balance between neutralizing and facilitating antibodies in vaccinated individuals that would favor of neutralization for the original Wuhan/D614G strain.
Importantly, the viral characteristics involve a known disease-enhancing epitope - recognized from the days of SARS-CoV-1. That disease-enhancement epitope, found in SARS-CoV-2 spike protein amino acids 597 to 603, is LYQDVNC".
The same epitope was found in SARS-CoV-2, and the researchers point to published antibodies that develops against the SARS-CoV-2 spike protein, antibody 1052.
Fantini and his colleagues studied variants from the original Wuhan virus, Alpha, Beta, Gamma, Delta, Mu, and Lambda by aligning the consensus sequences and found that the ADE epitope was well conserved in all variants.
In the same analysis, they also examined the neutralizing epitope of the N-terminal domain (NTD), and saw succession of changes from the Alpha variant to Lamba. These changes included substitution and many deletions, especially for Delta.
The original infection in humans started with the SARS-CoV-1 version of the facilitating epitope, and in the first six months, one mutation dominated: D614G.
Their findings point to the fact that while the neutralizing epitopes have been evolving, the disease-enhancement, or “facilitating” epitope has mostly been conserved, meaning that selection maintained viruses with the overall facilitating epitope.
Comparing the SARS-CoV-2 epitope to those found in SARS-CoV-2 variants, the researchers found that rather than being completely conserved, nearly 100% of the SARS-CoV-2 variants had an amino-acid altering mutation in the middle of the epitope.
That mutation was the D614G mutation.
The researchers knew that when when the virus when left Asia, where there were SARS-CoV-1 antibodies, there was no D614G, but when it arrived to Europe, where there were no SARS-CoV-1 antibodies, D614G, the facilitating epitope, spread. D614G traces back to Alpha, which had 50% increased transmissibility compared to the Wuhan
Looking at how the mutation would influence viral fusion and entry, the research found that D614G is involved in the loss of hydrogen bond between distant amino acids that, prior to 2020, kept the spike protein closed.
This means the variant lineages had evolved an improved cellular entry mechanism and spread faster relative to other variants as a result.
In fact, it is known that SARS-CoV-2 neutralizing antibodies have a decreased affinity for the Delta variant spike protein, whereas the team’s research showed that facilitating antibodies had a strikingly increased affinity.
The full name of ADE is antibody dependent disease enhancement, but by virtue of sloppy language curation, it evolved to “antibody dependent enhancement”. In ADE, the disease enhancement occurs due to an interaction between antibodies produced by the host and the cellular entry apparatus that makes cell entry easier.
The research team found that due to changes in affinity for ACE2, changes in electropotential (becoming more negative), and becoming less stable (the spike legs more easily separated), upon arriving in Europe, Delta had a massive selective advantage over all other variants.
The research concluded that ADE may be a concern for people who received vaccines formulated on the original Wuhan strain spike sequence, and called for the production of vaccines that lack the structurally-conserved ADE-related epitopes.
Since we cannot find the ADE-enhancing motif in the Pfizer mRNA encoded protein, if Dr. Fantini is correct the other dynamics he discussed may prove sufficient to induce ADE. We clearly need more data on this right away.
I wrote to Dr. Jacques Fantini inquire on the accuracy of my representation of his findings, and to let him know that while his molecular/biochemistry-based Transmissibility Index predicts that Omicron will not replace Delta, the population-level dynamics of illness and death will cause Delta to be replaced - if Omicron is, indeed, milder.
He replied that my account was accurate and agreed we need to consider all parameters.
Good, hello everyone.
We are excited to welcome back Professor Jacques Fantini today, who will tell us about the structural approach to research on SARS-CoV-2 and the impact of its structure on sensitivity and pathogenicity in particular.
He came a few months ago, and at that time he went a little bit further in terms of structural research.
Over to you, Jacques.
Thank you Philippe, and thank you for inviting me again to speak at the IHU.
So, actually, the reason for this second conference is that since April when the Delta variant was rearing its head, a lot of things have happened, there have been many variants. Delta remains the majority one, so it raises questions.
These are questions which I am trying to answer through biochemistry, and a lot of simulation, of molecular biology, but mostly in silico.
Not really manipulation, but a fairly detailed structural analysis of the spike protein of different variants.
So, just to give you an idea,
I was struck by this magnificent article in PNAS.
Now obviously, they rounded up the numbers, but it's always useful to know what we're dealing with.
We're dealing with a pandemic, everyone knows that.
But above all, we're dealing with very different viruses which can be quantified.
So, the quantification was quite interesting:
for every infected person, there are a hundred billion viruses produced every day at peak infection,
which is very little in terms of mass.
It's 100 micrograms maximum.
And as a result, we can estimate each day multiplying this mass by individual, by extrapolating the number of people infected, we can calculate the approximate mass of SARS-CoV-2 which is currently circulating around the world.
They calculated it as between one hundred grams and ten kilos.
So it sounds crazy, because it's biologically... concentrated, but that's how it is.
Now, this virus mutates.
I think we need to really take that into account.
You’ll see that I haven’t failed to do that.
So the virus mutates. Plus, the virus mutates easily.
Therefore in an infected individual - still referring to the same PNAS paper - when you factor in 3 to 7 replication cycles during the course of infection, you get, roughly, 0.5 to 1 mutation when the virus comes out of the individual.
When you have inter-host contamination, there is an estimated mutation rate of three mutations per month. And then, finally, you end up with viruses with 20 to 40 mutations.
It can be less, it can be more.
These figures are quite convincing.
This one doesn't mean much because it really depends on the geographical area, the virus and a whole lot of parameters that are not taken into account.
However, what is important is to know that the genome of this RNA virus is a very large genome: with 30,000 base pairs.
It's almost the maximum possible in such a small compartment.
Thank you Philippe Colson for this magnificent representation.
I like to represent the different proteins in this way, with the genes corresponding to the proteins and mutations.
And we see that ultimately, distribution is not even, that is, there are still mutation hotspots.
And it's not surprising to find that the spike protein, the one involved in entry into the virus, concentrates a lot of these mutations.
However, there are mutations in other proteins, and if we really want to understand not only infectivity,
But also the pathogenic effect of this virus, we also have to look at the other proteins.
If you don't look at the other proteins, you only get a partial view that is actually focused at the level of infection.
So, this entry mechanism,
I’ve broken it down, very arbitrarily, but nonetheless instructively, into four phases.
The first phase is the attraction phase, when the virus has to reach the plasma membrane, the surface of the cell that it will infect.
And not all viruses, if different, have the same chances.
Some viruses are faster. Some viruses are slower.
The second phase is ACE-2 attachment.
Once the virus is on the cell, it will attach to the ACE-2.
And this requires - as I said in my first conference - this requires lipid rafts, that align the viral particle and the ACE-2 receptor.
And it’s in this way, through a set of chance - and then cooperative - meetings, that the virus will actually find the ACE-2.
The third phase is cleavage, by furin-type proteases, that will actually cut the spike protein in two
to enable release of the fusion mechanism.
So now we see that the protein has been cut in two.
And what happens next? I think it's the most phenomenal,
And this goes into the realm of thermodynamics, because when you're at that point,
the spike protein sends arms towards the cell.
And then several phenomena occur.
Which, in the end, leads to fusion of the virus envelope with the cell membrane.
And that mechanism is thermodynamic, and we need to understand that ultimately, from the moment where we have separate stages, we're going to have areas of spike protein, which is huge, that will actually assume these different functions.
So we're going to have different selection pressures that are exercised on different parts of the protein.
So these four areas, I summarized them here.
Fusion, over in this area here, the cleavage site, affinity for ACE-2 and the Receptor Binding Domain, RBD, and here for attraction, therefore kinetics, the speed at which the virus reaches the cell, this is the NTD domain.
So we map the spike protein and, as a reminder, the NTD will interact with the lipid raft.
The RBD will interact with ACE-2.
What is interesting is that interaction forces are roughly the same, that is, the first approach allows the virus to attach to a raft.
That lets it wait to find the ACE-2, and when the virus finds the ACE-2, the interaction is practically the same as the initial interaction.
However, this involves affinity because we have a receptor for an RBD domain.
Here, avidity is a factor, because the raft includes a large number of ganglioside molecules, mainly GM1.
So, to study these phenomena, you have to obtain files from the PDB database, to obtain the three-dimensional structure.
The problem is that there is no whole spike protein.
Why? Because whatever method is used to get to the structure, there are gaps, by which I mean areas that are not defined, and areas that are not defined, leads to this.
You see here, in this first area, you pass directly from histidine to aspartic acid, from H to D, and this is missing.
If you have a mutation that falls here, how are you going to analyse the impact of this mutation on the structure of the protein?
You have no other solution than to reconstruct the protein, to use the PDB file and to add the missing parts.
Now, the missing parts can be larger or smaller.
Here you have a protein which is somewhat more complete, but there are still missing parts. For example, on this one, you see what is missing, it's the cleavage site, in that you don't have it on the PDB file.
In any case, the proteins obtained from this database are all incomplete, so they need to be reconstructed, which involves a huge amount of modelling work because it is not enough to add them, you have to minimise, you have to do the modelling to get a protein that meets a certain number of constraints.
This takes work and, if it's not done, you analyse things that are incomplete and this can lead to false results.
So what we finally see, is that there are areas that are critical for the virus, and there are areas that are less critical.
And finally, we realise that the selection pressures work slightly counter to this entry mechanism.
What made it possible for the SARS-CoV-2 virus to infect humans?
Well, this is the infamous furin-like cleavage site, which appeared for whatever reason, we're not going to get into that debate.
But nevertheless, you see that this site was highly preserved.
Which means that from the moment the virus has reached this furin site, it can concentrate on other areas of the protein.
What are the other areas?
You have the RBD domain, which is responsible for recognising ACE-2, with mutations that directly influence the affinity of the spike protein for ACE-2.
For example here, the N501Y, which, by means of a stacking mechanism, i.e. stacking of aromatic cycles, enables better interaction with the ACE-2.
And then phase 1 is the kinetic effect. Which appears afterwards.
Why does it appear afterwards? When the cleavage issue has been resolved, when the issue of affinity for human ACE-2 has been resolved, then you get viruses whose primary selective advantage is faster than the others.
So faster, what does that mean?
Well, in two senses. But above all, I will draw your attention to the NTD domain, that is often overlooked with respect to the RBD domain, which is more noble because it recognizes ACE-2.
This NTD domain recognizes the raft, which is negatively charged because of the gangliosides.
And as you can see, everything that is red is negative.
Everything in blue is positive, in terms of the surface potential.
So you see there is a good match between the surface of the NTD, which is blue overall, and the surface of the raft, which is negative overall.
Nevertheless, you see that there are possibilities for action.
You see that when you have red opposite red,
It's not optimal for an interaction.
So, it means that, on those spots, it is possible to act.
It is possible to act within certain limits.
But in any case, to act according to a logic involving a race to the electropositive surface potential.
It's the same for ACE-2, which is, overall, electronegative on the surface.
And the RBD domain is somewhat mixed: positive on one side, negative on the other.
Clearly, if we modulate at this level on negative charges which are in red on the RBD, if we modulate the amino acid composition by targeted mutations, we’ll be able to get a better match between the ACE-2 and the RBD, and it will translate into better affinity.
So we understand that there is a mutational logic, that makes the NTD evolve with selection pressure for interaction with the raft.
The RBD will evolve with selection pressure for interaction with ACE-2, and of course, as these are separate areas, the two events will multiply.
If you have a virus that has both a mutation that confers better affinity for the RBD and a mutation in the NTD that confers faster kinetics for the NTD, you have a faster virus which will win the race compared to other viruses.
I'm just talking about infectivity.
Pathogenicity depends on other parameters, other viral genes, here we are talking about infectivity: which viruses are infecting (more infectious) compared to others?
So in April, I came here to do a conference and the Delta variant had appeared the week before.
And so, I wanted to talk about the Delta variant and at that time, I had the idea based on these parameters that I had already predetermined, of finding out whether we could evaluate, ultimately, the major selective advantage of a variant with respect to the circulating virus, considering the surface potential, affinity for ACE-2 that are ultimately the two key parameters that allow the virus to infect a cell.
And indeed, I found this way of evaluating the potential transmissibility of a variant by measuring the surface potential.
And this was between the original Wuhan strain and the English variant.
What we see very clearly, is when a variant imposes itself compared to another, regardless of all other parameters but at least with this parameter, we find that the imposing variant is the one with less red, more blue.
So, clearly, less electronegative potential, more electropositive potential.
This allowed me to make a prediction for the Delta variant.
Well, because by combining all these parameters, affinity, kinetics, surface potential, I developed an index.
I called it the T-index, for transmissibility index, and we saw very clearly, while with the Delta variant, there were only a few cases, we saw very clearly its selective advantage emerging, i.e. its index was a lot higher than the index for all the other viruses that were circulating at the time.
We had English, South African, Brazilian, because they were named like that.
But in any case, the Delta variant really had a selective advantage.
And this advantage is such that since then, as we will see, some variants have appeared.
None of them exceeded that value.
None of them imposed themselves over Delta.
So, again, this is one parameter.
It doesn't mean it's the only parameter to be taken into account, but it is a parameter that is important
And that has at least allowed us to... I don't like to say predict, but to analyse existing variants and say in a strict sense that when I look at all these variants, I see that there is one that stands out from others.
After that, it is up to the variants to decide.
So in 2019, we had Wuhan, the original strain.
In 2021, we had Delta.
That's the NTD.
With Delta, we saw immediately that we could say yes, it will evolve.
Why? We're going to have a Super-Delta, simply by replacing the red with the blue, and then we will have an even faster virus.
But that never happened.
That is to say that in the seven months that Delta has been circulating, it never happened.
There's a good reason for that, it's that if a Super-Delta appears in that way, with too much electropositive charge, what’s the result of that?
It will make the virus stick to the surface, but then it'll stay stuck. It's too strong.
It's too strong, and it's kind of like Captain Haddock’s plaster.
It can’t get unstuck, so the virus can’t enter.
And we have a scenario in biology, in virology, that corresponds perfectly with that, it's the case of the PK blood group.
The PK blood group is a very rare blood group, one in a million, that is associated with overproduction on the surface of lymphocytes of a glycosphingolipid which is called GB3.
Electronegative surface potential.
And I did the experiment for that, in collaboration with Lingwood, in Canada,
When you take the virus and put it in contact with these lymphocytes, the virus sticks but doesn’t enter.
Whereas GB3, is a factor that is necessary for the virus to enter "normal" lymphocytes, in quotation marks.
So, that means we have an analogy, which I find striking, between the HIV situation, where we have a resistance blood group which means that the virus cannot enter because there is too much GB3, too much electronegative charge, so it's not the fault of the virus, it's the cell's fault.
Always keep in mind we are looking at virus-host interactions.
Well, this analogy is striking with the situation we would have with the Delta, where, indeed, you can always imagine a super virus all blue, very electropositive.
Perhaps by the way it did appear, but quite frankly, myself, I wouldn’t bet much money on it because it's a virus that, as I calculated, can't do anything but stick to the surface.
So for me, it's very reassuring. Why?
In April, my thinking and my position was that I thought the Delta would impose itself.
Will the Delta evolve to become Super-Delta?
For me, there is no chance, because of this phenomenon.
So that's how it operates. That's my analysis.
Here you have the kinetic parameter, or the speed at which the virus will move.
You may have different viruses that compete with each other and you can rank them depending on their surface potential.
Well, you can see that the Wuhan strain is very good.
It’s very well adapted to the human species.
But, when a slightly more electropositive variant appeared, it had a selective advantage. It was faster.
It didn’t have greater ACE-2 affinity. I think all the studies that tried to demonstrate that mutations, variants have given the virus increased ACE-2 affinity, when we look at the figures, we don’t see logs of 10.
We see 1.5, 1.8, 2 maximum, whereas the difference on the kinetic parameter is huge, represented by the T-index, where we see the Delta log is greater than ten.
You see, the Alpha has a slight selective advantage,
Beta a little more,
Gamma even more, Delta even more.
But look where we are.
As I see it, we're almost at the end, for this parameter.
For this particular parameter, if we imagine a Super-Delta, it’ll stick to the cell, It’ll almost enter, will incorporate into the membrane, but it can’t do much else for precisely that reason: it’s too much.
Which I think is pretty reassuring.
So now I want to talk to you about another aspect of these variants
Which, of course, all induce immune responses to a greater or lesser extent, be these responses to a vaccine or natural responses to an infection.
And the spike protein is huge.
It means that whether it's the vaccine with a code for the protein or a virus,
there will be a lot of antibodies against the Spike protein.
Lots of antibodies, which means that most antibodies won’t be doing anything.
In fact, they stick to the Spike protein. They're neutral.
They have no particular activity.
You have antibodies against well-targeted areas, and these well-targeted areas, inevitably, are the interface, i.e. the top of the NTD, the top of the RBD.
If you have an antibody there, on the NTD, the raft will not be reached and on the RBD, the ACE-2 will not be reached.
So these are the neutralization sites.
But regardless of that, there are also antibodies called “facilitators”.
So why facilitators?
Well, because they totally fail to recognize the virus-to-cell interface,
However, they do recognize areas that are somewhat hidden, and which, by means of different types of phenomena, will not block infection, but on the contrary, will stimulate infection, for example by attaching to the Fc fragment of Immunoglobulin and by helping macrophages to pump virus that will be released later.
So these antibodies were discovered on SARS-CoV-2, in patients.
Which means that they exist.
And there are two papers in Cell that came out one after the other, by two different groups, that have revealed, this one in August and this one at the end of June, that have revealed precisely such a facilitating antibody, directed against the NTD domain.
You have it here in green, so the same on the other side.
And then, in that study, you see that the other facilitating antibody recognizes in fact exactly the same region.
We have a facilitating epitope.
Hence the question I asked myself:
if we have a facilitating epitope, is this epitope preserved in variants?
Since now we have Delta, plus all the variants that have since appeared.
Well, what I discovered and published, is that this facilitating epitope,
which is recognized by an antibody, was of course characterised with the original Wuhan strain and the strain, which we will talk about, which is mutated, D614G, which is the derivative of Wuhan.
But when I did the modelling, I was surprised to see that not only does the antibody that is in green here
fully recognize the NTD of the Delta variant, of the main Delta variant and of all Delta variants, but it also recognises it with greater affinity than its recognition of the original Wuhan strain.
Clearly, the Delta has evolved in such a way, whether it is coincidental or not, that the affinity of the facilitating antibody for its spike protein, is greater than that which exists for Wuhan, whereas this is not the case for the Wuhan strain.
We did a great job, I think, with Bernard La Scola, in which, we review all these variants, the neutralising response of these variants,
And in the modelling, I was able to measure the affinity of the neutralising antibodies of the various variants.
And with regard to the Delta variant, what is clear, is that the neutralizing antibody, that is supposed to attach here, well its epitope is very poorly preserved, so there is a significant loss of affinity while at the same time, there is an increase in affinity for the facilitating antibody.
So what does that mean?
It means that in the case of the Delta variant, the balance between facilitation and neutralization
- ADE means Antibody-dependent enhancement - therefore increased infection, well the balance for the Delta variant leans very clearly in favour of facilitation.
So, less neutralization, and facilitation.
Whereas for the original strain, that is not at all the case.
In the original strain, antibodies are totally neutralising, and although the facilitating antibody exists, it has an affinity that is not the one found in the Delta.
So for the original strain, including D614G, the balance tilts in favour of neutralisation, which means that this balance between facilitation/neutralisation, will depend on the variant under consideration.
And we can analyse this using molecular modelling techniques, since all of the PDB files are available.
That is modelling from a PDB file reconstituted to fill in the gaps, but it’s not modelling from scratch.
So, to give you an idea of what we did with Bernard La Scola and the whole team, Rita Jaafar in particular, and more that I will not cite in detail, is - I'm not talking about their very fine seroneutralization experiments - I am talking about the immune escape index, i.e. loss of antibody affinity.
Well, we see that we have made a logical progression from the original strain to the Delta variants.
We see that gradually, and it’s proven true as and when variants appear, there is a decrease in the affinity of neutralising antibodies.
Which is not surprising in itself. but at the same time, there is an increase in the affinity of the facilitating antibodies.
So, people often counter and object on social networks, which are very dynamic, Linkedin in particular, is very dynamic in that regard, People counter with: "yes, OK, you've demonstrated, you’ve published that this ADE, could theoretically exist”.
It's true, I have not conducted any ADE experimentation.
So they say to me: “but for SARS-CoV-2, it doesn’t exist”, and we can list the arguments.
And I say in response:
“OK, let’s look at things calmly.”
The ADE phenomenon has been demonstrated for all viruses: dengue virus, and there was a therapeutic trial that had to be... stopped under disastrous circumstances because with this facilitation phenomenon,
there were some deaths.
Ebola, Zika, HIV, influenza, coronavirus.
Then animal coronavirus, thank you, Patrick Guérin.
It’s a known phenomenon.
So, it's known, but people reply:
"But ADE was never detected for SARS-CoV-2 after mass vaccination.” “
It is true but where is the perspective?
As you know, studies are published months later.
Studies today published on these ideas of antibody efficacy are February/March and not encompassing the Delta variant.
Proved by one of the Cell papers that studies the possibility of ADE stating we take a monkey, we infect it, we throw in the facilitating antibody.
There is no ADE.
There are minor problems with inflammation but admittedly no ADE.
Which strain was used?
It was the Wuhan strain.
They didn't evaluate the current variant.
So we can ask this:
Does it appear with other variants?
Because for now, this phenomenon, has not been studied for variants.
Accepting ADE does not exist for SARS-CoV-2
It means that due to biology, this virus is free of the phenomenon.
It's possible, but we still need arguments that support this thesis.
For now, I haven't seen any.
The only factual argument that is given, is that it's never been seen before, so it doesn't exist.
So, what is the point of looking at these phenomena with the variants that exist?
The point and the option we have is to task databases, sequencing databases. Some people work with JZ, Nouara and I, we like to work with Los Alamos.
Anyway, the sequences are ratified in both cases.
So Nouara analysed a million sequences over the last four months.
So she took a million sequences and we took the neutralizing epitopes, the facilitating epitopes.
And it's not complicated: we align and then we have a certain percentage each time.
So what are the results?
Well, before giving you the results, I inserted this little slide to create a bit of suspense, but mainly because it highlights a quite particular phenomenon which is that we're not starting from scratch.
Pay attention to this factor, which was suggested to me again by Patrick.
We are not starting from scratch with ADE.
We have the experience of SARS-CoV-1.
In SARS-CoV-1, there is a facilitating antibody that has been characterized.
What does facilitation mean?
In infection tests, you have a percentage of inhibition.
Neutralising antibodies inhibit infection.
Then, in some cases, not only do they not inhibit, but they facilitate.
So we have facilitation, i.e. negative inhibition.
Well, the epitope recognized by these anti-SARS-CoV-1 antibodies, which is still human coronavirus.
Amino acids 597 to 603, LYQDVNC sequence.
Immediate question: does this epitope exist in SARS-CoV-2?
But what do the people who published in 2016 on this facilitating antibody say?
They say this: SARS-CoV vaccines, therefore SARS-CoV-1, could be developed while avoiding ADE.
How? By deleting this epitope from the sequence.
So, that's what I recommend.
If we have to modify vaccine formulations, the first idea that occurs to me, is to look, to look seriously at this ADE phenomenon.
And this is a concrete proposal, that is, if you rework a formulation which, for example, may be better adapted for variants- since this is the purpose of these messenger RNA vaccines to be able to adapt the evolution of an epidemic - if there are variants, we modify the sequence, so there’s the possibility of modifying, we could also modify the epitope, and see if there’s a protein that’s also immunogenic, etc.
But anyway, there you have it, that was known, and so we have to wonder why it doesn't exist.
So anyway, I won’t overdo the suspense,
I will talk to you, though about this sequence alignment.
Above you have the ADE epitope, for SARS-CoV-2, recognized by antibody 1052. It’s one of the antibodies published in Cell.
So it recognizes a multiple epitope, and therefore several sub-epitopes,
Which in this case are: 27-32, 64-69, etc.
You have all that. Right.
When aligning the consensus sequences, for now, of variants, well, we realise that, overall, and in any case, this is the case for the Delta here, well, this ADE epitope, is generally very well conserved
in all variants.
Then we look at the neutralizing epitope, the canonical neutralizing epitope of the NTD.
These aren’t non-existent results.
It’s the result we see in sequencing databases.
It’s the characterization of these viruses.
Anyway, we see that, starting from the English Alpha variant, we start to get changes.
A lot of deletions.
The Delta variants are deletions. Many deletions in these two areas.
We even have the Colombian Mu variant, with an insertion.
In any case, the findings are very clear: on all these variants, and we tested Alpha, Beta, Gamma, Delta, Mu, Lambda, and I will talk about another one shortly, we have significant variability amongst the neutralising epitopes, whereas we have conservation of the principal facilitating epitope.
So now we’re going to look at the initial epitope of SARS-CoV.
This one is really exciting,
Because I was expecting that it would be conserved.
And then, when Nouara did the alignments, she said: “no, it is not conserved at all since almost 100% of the sequences have a mutation right in the middle.”
I said: “Oh, that's weird.”
And then she realized that the mutation was the D614G mutation.
So that means that when the virus arrived in Europe, when it left Asia, there was no D614G,
when it arrived in Europe, there was D614G.
Which makes you wonder. Why?
Well, because this is the initial curve,
For the first half of 2020, of this D614G.
We see that initially, we started with the facilitating epitope.
So we can say that the first people who were infected with SARS-CoV-2,
had facilitation with pre-existing antibodies, since the same antibody recognizes SARS-CoV and SARS-CoV-2.
And then the pandemic set in with one modification.
There weren’t half a dozen mutations during those six months, there was only one, in the spike, the D614G. That’s the one we got.
So you have to return to the fundamental mechanism of ADE and D614G.
D614, which is that of the initial virus, actually, it's involved in a hydrogen bond that keeps the spike protein closed.
You see this hydrogen bond.
With D614G, of course, you lose the possibility of this hydrogen bond and one of the obvious mechanisms, is precisely that separation is going to be much easier.
But this separation, you will see this here with the facilitating antibody.
This is what the facilitating antibody does.
And if there is no facilitating antibody, you will see it through the mutation.
Clearly, when the pandemic took hold, I am convinced that the mutation appeared to enable people who had this antibody, but who could no longer benefit from it, to be able to infect as many people as possible.
Clearly... I will finish and then answer your question.
In fact, this mutation, it can infect without facilitation.
You have a question, Christian?
Do you have any idea where this facilitating antibody came from, since there was no SARS-CoV-1 in Europe?
Exactly! In Europe, there is no need for the antibody.
The antibody is in Asia.
In Asia, you have the antibody since they had the infection.
So infection is facilitated.
But in Europe, it is not.
So how do you get separation? You get separation without needing the antibody, because look what the antibody does.
Look at what the antibody does: that's the facilitating antibody that will help you to separate the two.
So you have the same mechanism.
You have exactly the same mechanism.
Here, the mechanism, if it is closed, if locked by the hydrogen bond, you need a facilitating antibody.
So this is the Asian situation. The virus spreads.
In Europe, there is no facilitating antibody, because they have not been infected
Huge selection pressure.
Remember, it’s way before interaction with the cell, it's almost at the level of the fusion mechanism.
This is what will unmask the RBD for interaction with the ACE-2.
And, if you need a facilitating antibody, it's closed by the hydrogen bond.
On opening, you no longer need a facilitating antibody.
Actually, to my mind,
this is one of the aspects to consider. The significance of this D614G, it’s a phenomenal, exponential increase in infection in Europe.
Automatic, in the absence of a facilitating antibody.
And in the beginning you need a facilitating antibody to separate the trimers.
So that probably explains it.
So now let’s return to our analysis of a million sequences.
Because the Los Alamos database is absolutely fantastic.
Very specific questions can be asked and we can ask for the number of mutations
on each epitope.
So, just think about that.
The number of mutations classified by epitope.
So already you see that here you have... 1052. It's the facilitating epitope, which is actually an epitope that has several sub-component domains.
And you see there are zero mutations here in this segment, 99% of the sequences.
On that segment 90%, 99, 99, 99.
You have variability which, globally is minimal worldwide, out of one million sequences.
Now, what happens with the neutralizing antibody?
Well, for the neutralizing antibody, you have two sites.
You have a first site that doesn’t vary yet.
It’s shaky, but 99% of it remains conserved.
But look at this site.
Zero mutation. There is no more mutation.
It means that the site recognized by this neutralizing antibody, has already mutated.
One mutation, only 4% of sequences.
Two mutations - can you understand the significance of that? Two mutations in the neutralization epitope - 92% of sequences.
And there’s more.
There are already viruses appearing with five mutations, the hotspots that are represented here.
But you see that you have over-representation of variability on this neutralizing epitope, whereas it is non-existent on this facilitating epitope.
In passing, look at our D614G, the epitope: a D614G mutation, 99%.
Once acquired, it no longer varies.
It's too big for the virus.
So you see that through these analyses of sequences, we can really conclude that variability among variants of SARS-CoV-2, between neutralizing epitopes and facilitating epitopes, is not the same at all.
It's the opposite.
We have conservation of the facilitating epitopes, and we have phenomenal variability of the neutralizing epitopes.
So, very often,
that's what happens, that is to say we discover that there is a new variant.
So when people discover that there's a new variant there is panic.
So, each time I come to give a seminar at the IHU, there is the variant that appeared the week before.
So this one is the variant...
It appeared in Brittany, but we're told it’s from Congo.
And this is what we can read about it.
And believe me, I read this with my students because it's really very informative.
This is the title:
“The coronavirus variant discovered in Congo is to be followed closely, here’s why”
So we expect some explanations, and that's what we find.
"We said that Delta had become hegemonic and that the next variant would be one of its descendants.
This is another very different branch, a lineage that seems to have been living life under the radar.”
As for a virus that lives its life...
I will not go there.
Next you’ll have the scene where its arms fall off.
I didn't include the name, but it's an interview with a scientist.
“It is possible that part of this diversity is linked to human or technical error when sequencing these viruses.”
So, in actual fact, we are presented with a series of mutations, and we are not sure they belong to a virus because there is a possibility.
And the end reads:
"We are unable now to say whether this variant will spread or develop.”
Okay. Let me tell you this.
Still using the same method,
Surface potential analysis.
Starting from Delta, because it's in relation to Delta that this is going to be seen.
If we started with Wuhan, it would be different.
We’re starting with Delta. What do we observe?
We observe red central zone, red central zone.
Then maybe it's a bit more red here, a bit more blue there, but it's compensated for here.
Therefore, the overall surface potential of the domain is the same between Delta and B.1.640, for the NTD.
Now, regarding the RBD, the area that interacts with the receptor, this is where Delta came in, and this is where the new variant is.
There, actually, there is no photo, the selective advantage is for the Delta.
So, with these same parameters that allowed me to estimate that the Delta variant was going to spread because it has a huge selective advantage, I now calculate that, with respect to Delta, this new variant does not have any particular selective advantage.
Which means that, for these parameters, it has no reason to develop at the expense of Delta.
And I may be wrong, because once again I’m only using one parameter.
If I was wrong, I would be very surprised.
Anyway, in science, you learn from your mistakes, and, of course, it would help me find out why this variant did not obey the law that I thought was universal, allowing us to make further progress to understand some aspects of this virus that have escaped us up to now.
But in any case, the analysis such as it was carried out for Delta absolutely fails to justify the commotion we are hearing regarding this variant, the presence of which is very, very slight and very little represented.
I don't think that it was particularly... dangerous, not in terms of pathology, but dangerous in terms of competition with the Delta variant.
So I added it to the list, and I looked at the epitopes with Nouara.
Same as before: facilitating epitope, in any case, it was conserved, neutralizing epitope, yes, but that's the case with all variants.
So, it means that ultimately, regardless of the variant, there is a neutralization problem.
So what does that leave?
Surface potential and transmissibility index, at this stage, to explain why the Delta variant persists compared to other variants.
It is this variant that, until now, has the best selective advantage for surface potential combined with ACE-2 affinity, etc.
So in the meantime, I tested the Delta Plus,
I tested a whole load of variants that appeared and we were told that...
There isn’t a single one with a transmissibility index above five, be it the Delta Plus, or the Delta subvariant now appearing.
At the moment, it's really the two forms of Delta that are circulating the majority ones, and personally, they will continue to be for quite a long time.
So, the future aim of this work, as you can understand, is to organize an observatory of variants,
precisely in order to be responsive.
Out of curiosity, to find out if a variant will develop or not.
Nevertheless, it is a colossal job, because when you look at the Los Alamos database, you have lots of cards and distributions.
And then, when we say "the variant", That’s not correct.
Overall, a variant is within a lineage, the nomenclature for this virus is starting to be a total headache.
From the times of HIV, we had subtypes: subtype B, subtype E.
We knew what it meant. We knew what to refer to.
Also, it varies all the time.
So it's very difficult even going back in the files.
When I look at files that were compiled in the study with Bernard La Scola on the viruses that we studied, it was India...
Then, to go back, you have to look at the sequences again, it's rather a hassle.
So that's what we're dealing with, a whole load of sequences.
The case of Singapore is very interesting.
You really have to look closely at it. Why?
Because it's in a vacuum, that is, Singapore is very protected and they’ve vaccinated almost 100%.
And with 100% vaccination, they have a Delta epidemic.
So already, we're sure of one thing, only for transmission, the vaccine does not protect against Delta because Delta has spread.
And then, which is very interesting, in the final analysis, the Delta variant existed before mass vaccination, and so the question arises of vaccine selection of a particular strain in a geographically confined territory.
Because here, really, it's a textbook case.
We were able to go back to the first Delta sequence and we matched it with the start of vaccination.
And also, there is a threshold, there is a threshold of 30% vaccination which is evidence of an outbreak.
So again, I’m not saying that vaccination selects variants.
Let's be clear, I’m saying it is a problem that arises - and which can be studied logically,
- Which vaccine? Scientifically. It's Pfizer. Pfizer, Moderna, that's what they used.
So it's something which can be studied logically, scientifically and I would add, calmly.
We're not engaged in a religious war, we are engaged a scientific analysis.
There is data, we analyse them, we interpret them, because when I study the evolution of variants,
I need to have all the parameters.
I want to know.
Vaccination, yes but vaccination plus masks,
It's not the same thing.
So we can't judge like that.
We have to have all the parameters and sometimes it's complicated.
In cases that are geographically restricted,
it's easier because the policy was stricter.
So we know exactly what happened and we can go back.
So, this is an analysis that, in my opinion,
could yield quite interesting elements.
So, vaccine strategy.
So, I'm not a vaccine manufacturer, but if I were a vaccine manufacturer, I can give them some information, if it is relayed to them at some point, which is that the famous ADE epitope, which is here, which is conserved in all variants, you can delete it by acting on two positions, H69 and D215.
So you delete both those positions, you replace them.
The antibody will no longer recognize it.
So you take the spike of a variant, for example Delta, you have a vaccine which is specific to a variant, you take out these two amino acids and problem solved, even if it is theoretical, even if theoretical, of this potential ADE.
So, I think that's kind of the rationale of my approach.
It's not just to criticize a situation.
I have no criticisms to make, I’m trying to understand, and in trying to understand, I realize I'm arriving at facts that appear relevant to me.
And it is my duty,as a university professor, to communicate these facts and say whether there must be a modification of the vaccines at any given time.
It I were them, they’ll do what they want, but if I were them, I would use the sequence of a variant because the Wuhan sequence, has been out of circulation for over a year and a half.
And also, I would focus in particular on that epitope since it is conserved in all variants.
And since it's not complicated.
We look at whether the protein is immunogenic, etc.
But there you go.
Now, herd immunity. For this, actually... We can take the example of smallpox.
Yes, but... Smallpox? DNA virus, no animal reservoir.
Here we have an RNA virus, which mutates with animal reservoirs.
You can vaccinate 200% of the world population, there will never truly be...
I would put my hand in the fire, there will not be, there cannot be herd immunity.
It's a myth, it's almost a myth even in virology, it's something which has been discussed for a very long time.
But in my view anyway, I don't think it's realistic claiming to want to achieve herd immunity.
It's not just me saying that, some highly eminent scientists are saying “no”, “it’s unattainable, we’ve given up on it.”
But it highlights a study that we initiated with Christian, on minks.
There are some things to note in any case.
What is an animal reservoir?
It's a reservoir with receptors bearing an incredibly close resemblance to the receptors found in humans.
So that is a superposition of the three ACE-2 receptors:
human ACE-2, European mink ACE-2, American mink ACE-2.
Actually, the modifications are tenuous.
Very little difference, so it's relevant to explain certain things, but still, there's a very close resemblance.
So it's not surprising there are animal reservoirs.
Blame the ACE-2.
I would like to say one last word, if you will let me continue, on mutation analysis.
This is the analysis of mutations...
The significance of mutations, because we are hearing anything and everything.
First, a silent mutation, a silent mutation means there is no modification of amino acids.
Does that mean there is no change in the protein?
The answer is no.
It's counter-intuitive. Those of you here, you know that, but the speed at which the protein is synthesized in the ribosomes depends on the availability of the transfer RNAs.
Some codons are rarer than others.
Some transfer RNAs are rarer, at a lower concentration, and therefore, you can have a messenger RNA that has a silent mutation where there is no change in the amino acids, but because the speed of translation is not the same, you end up with a different protein.
Then, in other respects, for mutations, keep in mind: just because there is a mutation doesn’t mean its a catastrophe.
Very often, mutations offset one another.
Look at this, you have positive charge arginine, negative charge aspartic acid, in a protein that forms a loop here.
So you need a positive and a negative
If you switch arginine with lysine, and aspartic acid with glutamic acid, you can have roughly the same result.
But worse than that: you can also have the same result with totally unrelated amino acids.
A tryptophan and a valine instead arginine and aspartic acid.
So, it means that mutation analysis, is an analysis that must be integrated into a context.
So we should not think that a mutation is disastrous.
Mutations are something to be studied.
In what context?
It’s not easy.
So, the team's publications.
The team is:
Henri Chahinian, who is here, Nouara Yahi, who couldn’t come, Coralie Di Scala, who is in Finland,
Fodil Azzaz, who is a doctoral student.
That’s the core of the team.
And we have some very interesting collaborations, with the IHU, with Philippe, with Bernard, with Rita, and more exotic collaborations from the west of France with Patrick, where we are starting to disseminate our ideas and discuss these variants collectively.
So that's what I wanted to say today on variants.
Most importantly, thanks for listening.
One last thing: frankly, the hand-washing message,
That’s the most important! For me, that's the most important thing.