Immune system vs. virus: Why omicron had experts worried from the start

Right from omicron's first description, researchers were concerned about its variant of the SARS-CoV-2 virus. Looking over the list of mutations it carried, scientists could identify a number that would likely make the variant more infectious. Other mutations were even more worrying, as they would likely interfere with the immune system's ability to recognize the virus, allowing it to pose a risk to those who had been vaccinated or suffered from previous infections.

 

Buried in the subtext of these worries was a clear implication: scientists could simply look at the sequence of amino acids in the spike protein of a coronavirus and get a sense of how well the immune system would respond to it.

 

That knowledge is based on years of studying how the immune system operates, combined with a lot of specific information regarding its interactions with SARS-CoV-2. What follows is a description of these interactions, along with their implications for viral evolution and present and future variants.

Ts and Bs

To understand the immune system's function, it's easiest to break its responses into categories. To begin with, there's the innate immune response, which tends to recognize general features of pathogens rather than specific properties of individual bacteria or viruses. The innate response doesn't get fine-tuned by vaccination or prior exposure to a virus, so it's not really relevant to the discussion of variants.

 

What we're interested in is the adaptive immune response, which recognizes specific features in pathogens and generates a memory that produces a rapid and robust response if the same pathogen is ever seen again. It's the adaptive immune response that we're stimulating with vaccines.

 

The adaptive response can also be broken into categories. In terms of the relevant immune responses, we care most about those mediated by antibody-producing B cells. The other major part of adaptive immunity, the T cell, uses a completely different mechanism for identifying pathogens. We know a lot less about the T cell response to SARS-CoV-2, but we'll come back to that later. For now, we'll focus on antibodies.

 

Antibodies are large (in molecular terms) assemblies of four proteins. Most of the proteins are the same in all antibodies, which allows immune cells to interact with them. But each of the four proteins has a variable region that is different in every B cell produced. Many of the variable regions are useless, and others recognize the body's own proteins and get eliminated. But by chance, some antibodies have variable regions that recognize a segment of a protein made by a pathogen.

 

The portion of the pathogen's protein that the antibody recognizes is called an epitope. Epitopes vary from protein to protein, but they share some features. They have to be on the exterior of the protein, rather than buried in its interior, for the antibody to bump into it in the first place. And they often have amino acids that are polar or have a charge, since these form stronger interactions with the antibody.

 

You can't simply look at the amino acids in an antibody and tell what it's going to bind to. But if you have sufficient quantities of a specific antibody, it's possible to do what's called "epitope mapping," which involves figuring out precisely where on a protein the antibody is binding. In some cases, this can include a precise list of the amino acids that the antibody recognizes.

 

In general, having antibodies stuck to a pathogen in the bloodstream makes it easier for the pathogen to be spotted and disposed of by specialized immune cells—for this function, it really doesn't matter where the antibody sticks. But there are also specific interactions that can inactivate a virus in some cases, as we'll see below.

Antibodies, SARS-CoV-2, and vaccines

Antibodies generated against SARS-CoV-2 recognize proteins that are on the exterior of the virus's coat. In general, that means the spike protein, which allows the virus to latch on to cells and merge with a cell's membrane, depositing the virus's genome inside. There's only one other protein present on the virus's exterior in significant amounts; it simply maintains the virus's structure and doesn't play a specific role in infection. Most vaccines currently in use simply expose the immune system to the spike protein and leave out any additional viral proteins.

 

The parts of the spike that are commonly recognized by antibodies have been mapped, and they follow a typical pattern. There are hotspots that tend to be targeted by multiple antibodies because they stick out of their surroundings and have lots of charged or polar residues.

 

A number of these hotspots contain epitopes that are recognized by what are called neutralizing antibodies. These antibodies can interfere with the virus's ability to infect new cells. They work because they bind key regions needed for the spike to function—or simply because antibodies are bulky and can get in the way of the spike's interactions with the proteins it attaches to.

 

These neutralizing antibodies appear to be critical to the immune system's ability to limit the virus. While the immune response has a lot of features beyond the antibody response, a variety of studies have indicated that the levels of neutralizing antibodies present in an individual tend to correlate with protection against infection and severe symptoms. This doesn't mean other antibodies are not important—they can still aid with the clearance of the virus and may limit the severity of the disease. Neutralizing antibodies simply seem to have a more direct relationship with protection.

 

The immune system does not remain static once it starts producing antibodies. As infections are cleared, the number of antibody-producing cells drops, and many of them get converted to memory cells that can easily be mobilized should the pathogen reappear. This remobilization may not occur quickly enough to prevent reinfection once sufficient time has passed since the initial infection, especially if the immune system is challenged by a different variant of SARS-CoV-2. But it still appears to be enough to prevent serious disease in most cases, and boosters cause a remobilization that is highly effective.

 

One last note on antibodies before we move on. The genes for a number of neutralizing antibodies have been isolated, and they've been used to create cocktails of neutralizing antibodies that are somewhat helpful to COVID-19 patients if treatment begins early enough. The therapy made by Regeneron is an example of this approach.

Evolution and escape

Mutations in the epitopes that are recognized by anti-spike antibodies have the potential to eliminate their interactions. This can provide an evolutionary pressure that selects for changes specifically in these epitopes. If, because of past infection or vaccination, most people in a population have antibodies that recognize a specific epitope, then a change that alters the epitope and limits that recognition can allow the mutant virus to infect the population.

 

And there is clearly evidence that this has occurred. Many of the variants we've identified have mutations in the areas known to be recognized by neutralizing antibodies.

 

Fortunately, these mutations generally haven't eliminated the immune system's recognition of the virus entirely. The body's antibody response will typically recognize multiple epitopes of the virus. So while an individual mutation may eliminate the recognition of the virus by some antibodies, there are usually others being made that will continue to bind to other epitopes, including antibodies that don't neutralize the virus. The combination of gradually falling levels of antibodies plus reduced epitope recognition has led to breakthrough infections, but the immune response has been sufficient to prevent severe disease, especially when it has been boosted.

 

These changes, however, can cause problems for a number of therapies that are based on giving COVID-19 patients large numbers of two or three types of antibodies that are known to neutralize the virus. Here, a single change can potentially cut the efficacy of these therapies in half—and antibody therapies have not been especially effective to start with.

 

A more serious problem is that the virus can continue accumulating mutations in different epitopes. As these accumulate, they gradually reduce the number of antibodies that can effectively recognize the virus. There are potentially limits to this process. Some epitopes may also be essential to the protein's function, and thus difficult to change without damaging the spike protein's ability to mediate new infections. And a very large number of changes can potentially disrupt the basic structure of the protein, eliminating its function entirely.

 

One of the unfortunate lessons from omicron is that the spike protein can apparently tolerate a large number of mutations and continue to function effectively.

 

In fact, it was the past work in mapping common epitopes on the spike protein that caused experts to worry about the omicron variant when little more than its genome was understood. It was easy to look for where the mutations in that genome would change the spike protein and see that many mapped to epitope sites, including those recognized by neutralizing antibodies. This was a strong indication that the variant virus was more likely to escape neutralization by antibodies—something that has since been confirmed experimentally.

The wild card: T cells

Studies of immunity and SARS-CoV-2 have primarily focused on antibodies. This is partly motivated by the fact that levels of neutralizing antibodies correlate well with the degree of protection from severe infections. But it's also because it's relatively simple to measure antibody levels. By contrast, virus-specific T cells are very difficult to characterize. Yet, in part because they recognize the virus through a completely different mechanism, T cells could provide a degree of immunity even if antibody-based protection fails.

 

Instead of just recognizing only the proteins that are exposed to the blood stream, T cells can potentially recognize every protein made by a cell. That's because cells chop up a small number of the proteins they make and display some of the fragments on their surface. T cells are able to recognize when any of these fragments are unusual, either because the cell has picked up mutations that alter its proteins or because some of the proteins it is making are encoded by a pathogen.

 

There are several types of T cells that perform different specialized functions. The two major contributors to the immune response to SARS-CoV-2 are what are called helper and killer T cells. When helper T cells recognize a cell that is producing unusual proteins, it starts making lots of signaling molecules that activate other immune cells, including antibody-producing cells.

 

Killer T cells, by contrast, do exactly what their name implies. When they recognize a cell producing unusual proteins, they kill it. This keeps the cell from producing further pathogens, so it can limit the spread of infections after they start.

 

A key feature of T cell immunity is that it's much harder for viruses to evolve a way to escape it. Rather than being focused on just the few proteins that are on the surface of a virus, T cells can potentially recognize any protein the virus encodes. So instead of having mutations in just a few key epitopes, evading T cell immunity requires mutations scattered throughout the entire viral genome. It appears that, despite the many mutations in the spike protein of the omicron variant, omicron is still easily recognized by T cells.

 

The downside is that the most common vaccines only introduce the spike protein, so they don't provide the benefit of this more diversified immunity.

Where things stand

As mentioned above, immunity from severe disease appears to be linked to levels of neutralizing antibodies. Is there any evidence suggesting that immunity based on T cells is also useful? Perhaps the strongest piece of evidence is that some strains of the virus appear to be picking up mutations that interfere with T-cell-based immunity. There's also the unusual performance of the J&J vaccine, which appears to see a more gradual rise and decay of protection against severe disease. A few studies seem to suggest that this could be the result of a strong T cell response.

 

Further research is clearly needed to sort out the precise roles of the different arms of the immune system. In the meantime, we know enough about the immune system's response to omicron that we can clearly say a few things. The extensive number of changes in the variant's spike protein reduces the ability of the immune system to recognize it, both in people with prior infections and those who have been vaccinated. But the mutations do not eliminate that ability, and boosters restore extensive neutralizing activity.

 

What this has meant in practical terms is that the people who have been vaccinated over six months prior and have since seen their immune system switch over to memory cells are more likely to get a breakthrough infection when a distinct variant such as omicron appears. But they largely remain protected from experiencing a severe case. Booster doses re-establish an active immune response and provide an even greater degree of protection.

 

That's impressive given the number of relevant differences between omicron and the viral strains that the vaccine was designed to protect against. This suggests that it may take some time before enough mutations end up in a single variant that the effectiveness of the vaccines is severely compromised.