As the world grapples with how to safely reopen society in the midst of the coronavirus pandemic, scientists have been racing to understand whether COVID-19 infection confers immunity—and how long such immunity might last. A lot hangs in the balance: A strong immune response could mean people who have already been infected would be able to safely return to work. And it would also bode well for vaccine development efforts.
A small but suggestive new study finds that individuals who have had COVID-19 produce a robust response in immune cells called T cells. The adaptive immune system contains several main components: antibody-creating B cells, helper T cells and killer T cells. The latter two are important for recognizing and destroying a particular virus, respectively. Alessandro Sette and Shane Crotty, both professors at the La Jolla Institute for Immunology, and their colleagues found that of a group of 20 people who had recovered from COVID-19, 70 percent had killer T cells and 100 percent had helper T cells that were specific to the SARS-CoV-2 virus, which causes COVID-19. Importantly, the researchers observed a strong T cell response to the “spike” protein the virus uses to bind to and infect cells (and which most vaccine candidates target). They additionally detected a helper T cell response to SARS-CoV-2 in about half of blood samples they examined that had been drawn before the virus began circulating. This observation, they say, hints that exposure to seasonal common cold coronaviruses may confer some protection against the new pathogen.
The findings build on earlier studies showing that infection with the novel coronavirus produces protective, or “neutralizing,” antibodies. Taken together, these results suggest that people who have had COVID-19 possess at least some immunity—an encouraging sign for the dozens of vaccines under development. Separately, this week the company Moderna announced early results from a trial of its coronavirus vaccine candidate: eight individuals who have received the vaccine produced antibodies to the virus at levels similar to those of people who had the disease.
Scientific American spoke with Sette and Crotty about what their study means for immunity to COVID-19, possible protection from seasonal cold infections and the prospects for a SARS-CoV-2 vaccine.
[An edited transcript of the interview follows.]
What do we know so far about immunity to COVID-19? And why is it so important?
CROTTY: There’s just been a huge amount of uncertainty about immunity to COVID-19. And that question about immunity has two major implications: one, for just understanding the disease itself, and two, for vaccine development. It’s clearly been a world-on-fire type of situation. And so it has made sense for these 100-odd different vaccine programs to get going and just try moving things forward. The normal way you would try and make a successful vaccine would be to look at what gets you good protective immunity to that disease and copy that. A disease such as COVID-19 is normally an acute infection, and most people control and clear it without a lot of problems. That’s a good sign that indicates that the human immune system normally makes a good response to that virus and controls it.
But the immune system is a big, complicated place, with lots of different cell types with lots of different functions. And some are useful or important in one context versus another. For a vaccine, you’d want to know which components of the immune response are the important ones for protection against this disease. And without that information, you can very much go totally in the wrong direction with a vaccine program, either in terms of the type of immune response you’re trying to get or the [vaccine’s molecular] target. And both of those have been things that worried [Sette] and me and other people about these ongoing vaccine efforts. We really wanted to generate information that would help [us] understand the disease itself—and also generate information about which vaccine strategies are likely to be better or worse ones and whether people are getting the right [molecular target] or not. Our goal was to look at essentially average cases of COVID-19—ones where people definitely are making successful immune responses—and ask, “Okay, what does that immune response look like?”
Can you describe the different parts of the immune response and how they work?
CROTTY:Quite a few labs around the world have looked at antibody responses. Those are generally easier to measure and look at. But really, there are three parts of the adaptive immune system: you’ve basically got antibodies, you’ve got helper T cells, and you’ve got killer T cells. The T cells are tougher to measure, but they do really important things. You’ve got to have the helper T cells to get an antibody response. For example, in animal models, [helper T cells] are important for protecting against [severe acute respiratory syndrome (SARS)]. And the killer T cells are important for most viral infections. You don’t want to go forward without understanding anything about the T cell response. And [Sette] is the world’s expert in predicting and identifying T cell [targets], particularly in humans. So we collaborated to get COVID-19 patient blood samples as quickly as possible and to try and get information about those questions. We mostly have concluded it’s good news: things have largely looked the ways we would expect.
Do we know how long the immune response to the new virus lasts?
SETTE:What we certainly can say is that the infection induces a robust immune response, and this is in people that successfully deal with the virus and don’t get very sick. The question [of] how long this response lasts obviously takes time, because we have been dealing with this virus for only a few months, and we cannot possibly know what is going to happen a year down the line. But what we’ve seen thus far is encouraging, because these T cells look healthy, look happy. They are not exhausted, and they don’t express some of the molecular features that are associated with cells that are about to die.
In general, immunologic memory is like any other memory in the sense that the intensity of the event dictates how strong the memory is. Pretty much like any event in your life: if it was a life-threatening situation—for example, you almost got run over by a truck—you remember. If it was instead what kind of socks you wore, you might not remember. It’s the same for the immune system, in the sense that a very strong infection with a microbe that reproduces to high levels generates a strong level of immune response, which then creates a long-lasting impression. I would speculate that the memory generated by SARS or [Middle East respiratory syndrome (MERS)] could be somewhat different from one generated from a common cold, which is fairly adapted to not cause much trouble in the human host.
You also saw some T cell responses, or “cross-reactivity,” to the new coronavirus in blood from people who were never exposed to it, correct?
SETTE: We looked at the COVID-19 patients, and then we looked at a control group. We purposely went after blood donations that were obtained in 2015 to 2018—before any SARS-CoV-2 was around. Surprisingly, in about half of these people, we could see some T cell reactivity. And we looked at the data hard from the left and from the right and convinced ourselves that this was real. We do not know, at this point, exactly what this cross-reactivity means, but it’s reasonable to assume that it is the result of people having been exposed to common cold coronaviruses that are different from SARS-CoV-2 but have some similarity [to it]. This potentially has very strong implications, because one of the things that is unknown and everybody wants to gain more information about is why there is such a spectrum of different COVID-19 outcomes: some people are totally asymptomatic, whereas other people die. Of course, age and other health issues are factors, but one element could be immunological: If someone has some T cells that can cross-react to SARS-CoV-2, their immune system has an advantage. They can get going to generate antibody responses faster, maybe, and that could give a better outcome. In the context of vaccination, this is also very important, because imagine that you have a group of people, and half of them have this coronavirus cross–reactivity, and half of them don’t. Now you give these people a vaccine. It could be that the people that have the common cold cross-reactivity will respond a lot faster and a lot better to the vaccine, compared with the other ones.
One piece of data that is encouraging, in speculating that some preexisting immunity may be beneficial, is data from the 2009 H1N1 “swine flu” pandemic. You might recall that in that case, older people did better than younger people. And in fact, it turned out that the age of the people that did better correlated with when another H1N1 strain, a cousin of the swine flu pandemic strain, had circulated—so that the people that had been exposed in the 1950s to this other strain, their immune system still remembered a bit. Not that the people didn’t get sick, of course, but they got less sick. And they fared better than people who were totally naive and had never seen this particular subtype of influenza.
Do we know whether people that had asymptomatic COVID-19 infections might be less protected against reinfection than people who had a very severe case?
CROTTY: We did this study with people who didn’t have bad diseases—sort of average cases who definitely got well. Asymptomatic cases are definitely a big unknown. We have no idea [if they will be protected against reinfection].
Can you comment on the Moderna results from the phase I trial of its coronavirus vaccine candidate and the prospects for a vaccine in general?
CROTTY: There are actually three human vaccine candidates that have been tested in monkeys that gave what seems to be pretty good protection: one’s an inactivated-virus vaccine; another is a chimpanzee adenovirus vector [a type of double-stranded DNA virus used to harmlessly deliver genetic material to a host]; and a third one is a DNA vaccine [a DNA sequence that stimulates the host to produce part of the virus and mount an immune response against it]. And then there's the Moderna vaccine, which hasn't been tested in monkeys but has been tested in a mouse model and in humans to measure their immune response. So those are the three examples of interesting vaccine candidate data that are available as of today. And I think if we combine those with the data from our paper showing that the T cell responses generally look good and data from a number of papers about people making neutralizing antibody responses overall, I would say those vaccine studies—particularly the two that were done in monkeys—suggest, so far, that it’s not that hard to protect against this virus. I'm certainly encouraged, based on the magnitude of the immune responses to the vaccines—and the magnitude of immune responses we’re measuring in people who actually have disease and what happens in protection models. So far the available data are positive here.
SETTE: One encouraging thing is that there are so many different vaccines that are being developed. So our hope is that there is not going to be a winner, but there are going to be many different winners. The one thing that is important from our study is that the vast majority of these different vaccine concepts rely on one particular protein, which is the spike protein. And we saw very good responses, both in terms of killer and helper cells against the spike, which is really good news, because this was not a given. In this particular case, it so happens that it’s a good target for all three different types of immune response—which bodes well for people that are developing the spike-based vaccines. At the same time, our data found that there were responses also against other pieces of the virus, which opens the way to thinking that maybe these other pieces could also be included to further fortify a vaccine concept.