This interview — with immunologist Dr. Akiko Iwasaki — seems to these GEITP pages to be a very worthwhile update on COVID-19, possible mechanisms of its infection, possible vaccines, and gender/age/genetic differences in response to the SARS-CoV-2 virus. Therefore (warning: it is quite a long read), let’s share this informative interview with everyone. 😊
COVID Immune Responses Explained
Eric J. Topol, MD; Abraham Verghese, MD; Akiko Iwasaki, PhD
August 21, 2020
Eric J. Topol, MD: Hello. I’m Eric Topol and I’m with Abraham Verghese for a new edition of Medicine and the Machine. Today it’s a special treat to have a chance to talk with Akiko Iwasaki, professor of immunobiology at Yale. Welcome, Akiko. Great to have you with us.
Akiko Iwasaki, PhD: Thank you very much, Eric. I’m delighted to be here.
Topol: Akiko’s background is quite fascinating. She came from Japan to Canada, finished her baccalaureate and PhD at the University of Toronto, was a post-doc at the US National Institutes of Health, and she’s been at Yale for 20 years. She is a Howard Hughes Medical Institute scholar and National Academy of Sciences electee. She has become my go-to immunobiologist, through Twitter, her writings, and her videos. In fact, toward the end of July, she posted an Immunology 101 on YouTube that is a must-see.
Before COVID, the most precious talent-people were the data scientists. Now it’s become the immunologists. Akiko, I know you spend a lot of effort on innate immunity and the interferons (IFNs), but can you give us all a broad view of the two main components of the immune response to SARS-CoV-2?
Iwasaki: You are right. Immunology has become quite relevant with this pandemic. And I’m trying to not only do research on immunity to COVID, but also to communicate to and educate the public about the immune system and how it works.
Our immune system consists of two different layers. The first layer is the innate immune system, which acts within minutes of infection to provide kind of a rapid response. This doesn’t require any specificity; it is engaged after any kind of infection. But this innate activation is important to triggering the second layer of the immune system, which is the adaptive immune system. In the adaptive immune system, the key players are B cells and T cells; they eventually acquire specificity and memory, which are the basis of the vaccinations. Having those T- and B lymphocytes that are specific to a particular pathogen and provide a memory response in the long term is very important.
Topol: One thing that came up recently was this kind of scare whereby some said that the antibody response that was tested in a few different series of patients who recovered from COVID seemed to be abating over time. That may have been a false alarm. Maybe you can put that in context.
Iwasaki: I don’t blame people for becoming worried about this because the longevity of the immune response is what we are counting on for protection of the whole population. That’s what vaccines are supposed to do. If you follow COVID-infected patients over time, their antibody levels do seem to wane to some degree within 2-3 months. But that is not a cause for alarm because that’s what happens when you get infected or when you become immunized for the first time. The antibody levels peak within the first couple of weeks and then eventually come down over a few months.
That’s okay because you still have memory B cells specific to that antigen as well as a T-cell immune response to the viral antigen. So the second time you’re exposed to the same virus, you can mount a rapid, specific, and robust immune response. It’s likely that you won’t feel anything the second time you’re infected. It will be a pretty mild or asymptomatic infection.
I want to make that clear, that waning antibody levels alone is not a huge cause for concern. The second point I want to make is that with vaccines, we usually give booster vaccines, which stimulate a much more robust and long-lasting immunity. That immunity should be sustained for years. So again, having a waning antibody response to natural infection shouldn’t be a huge concern for going forward with vaccination.
Topol: That’s very reassuring. There are so many different dimensions where immunology comes into play here. One of the areas that’s especially interesting — and it’s somewhat related to your recent paper in Nature where you characterize the different responses in people with moderate vs severe disease — is this deficient IFN response in some people, and whether or not that can be boosted early with some type of IFN. It’s complex, though, because there’s not just one type of IFN, and the timing is kind of like Goldilocks — you have to get it just right. Where do we stand? I know several clinical trials are ongoing with type-I IFN.
Iwasaki: As you say, type-I IFN has to be delivered at the right time with the right dose and the right type. What’s coming out, including from our own study, is that a prolonged level of type-I IFN, especially late during disease, may be associated with a worse disease outcome. So giving COVID patients recombinant IFN late in disease is probably not going to be a good idea. Whereas using recombinant IFN as a prophylaxis against infection, or if you can catch the patients very early during the infection and give enough IFN to shut down viral replication, those strategies hold promise. I’m waiting to see what happens in the clinical trials. But I think the early timing and giving enough dose to block the virus replication is going to be key going forward.
Topol: A paper just came out in Nature Reviews Immunology that reviewed four small ongoing trials and that multicenter study from China. All of them look quite promising. If you were to predict the future, do you envision that we all could have an inhaler with an IFN that we would take at first diagnosis or exposure? Do you think that’s a possibility?
Iwasaki: As long as we can give the right dose without toxicity, that may be the future, especially in a preventive setting. For example, if your family member was diagnosed with a viral infection, you could potentially treat the rest of the family members with prophylactic IFN, and healthcare workers or people who are exposed to high-dose virus on a daily basis. That’s what happened in China; they gave an inhaled IFN to healthcare workers and none of them were infected. So this may be a good thing to do in the future.
Verghese: Akiko, you have a lovely way of explaining very complicated things in a simple fashion. I’m an infectious disease physician, but I must say, the way cytokines and the cytokine storm are described is so bewildering. Can you help make sense of what we should be paying attention to in all the various cytokines and why?
Iwasaki: I hate to add to this bewilderment, but the typical cytokine storm that has been reported includes IL-1, IL-6, tumor necrosis factor, these sort of acute innate cytokines. But in our recent report, we also found cytokines that belong to completely different kinds of immune responses also coming up, like those that are dedicated to fungal response or helminth response — type III and type II immune responses. So in severe cases, the immune system kind of looks confused and disoriented, generating all types of cytokines that are also causing some sort of storm. I think COVID disease may start with the typical cytokine storm but then extend to different types of tornados and hurricanes and all kinds of misdirected immune responses.
Topol: Everything gets dysregulated. It seems as though people, for the most part — unless they have other hits such as a pulmonary embolism or something else — the cytokine storm is the principal cause of death in terms of the actual mechanism. So far, the only thing we have to block it is dexamethasone. But that’s not having exceptional efficacy. We need to do better on that whole process.
Another big topic right now is “long COVID.” You’ve taken a systematic approach to the whole spectrum of the disease, from mild to very severe or critical; long COVID seems like the next chapter. Very little immunologic work has been published to date. But with the joint pains, profound fatigue, and many other signs and symptoms, it looks like an immunologic condition. What are your thoughts on this?
Iwasaki: I absolutely agree. Long COVID is quite mysterious in terms of what’s driving such long-term disease. Also, the symptoms in these patients appear to shift during the course of post-exposure to COVID. I have three hypotheses to explain it, but it could also be a combination of these. The first hypothesis is that a reservoir of virus is hiding somewhere that’s activating and reactivating periodically to cause these types of responses. The nasopharyngeal swabs that we currently use to test for the virus are unable to pick up those kinds of reservoirs.
The second hypothesis is related to this. Perhaps it is not reservoirs of infectious virus but bits and pieces of viral RNA or protein that are being retained somewhere in the body that are activating an immune response against the virus and causing these shifting and prolonged symptoms. The third hypothesis is that the infection generates an autoimmune disease. Perhaps the virus is mimicking self-antigens or virus infection, being so inflammatory in this case, and is eliciting autoreactive T and B cells. We are trying to understand which of these possibilities is true. But perhaps all of these things are happening at the same time.
Topol: The autoimmunity may not be so difficult to address, but how would you get at a virus reservoir or the bits of RNA that are somehow creating or sustaining the hit?
Iwasaki: Some of the insights are the result of autopsy reports of patients who’ve had COVID and passed away. We don’t have a lot of insights from the “long-haulers” in terms of autopsies. But at least the investigation of those autopsies after acute infection is revealing infection in many places in the body, including the lung, obviously, but also the gut and many other places. Autopsy results will reveal whether remnants of virus or potentially infectious virus are hiding in these organs.
Topol: That is something we need to prevent. It seems to be not uncommon. To your point about the retention of virus, many of the cases I’ve reviewed remain PCR positive weeks and months later. So there’s something that’s consistent with that hypothesis for sure. Somewhat related is the multisystem inflammatory syndrome in children (MIS-C). It is obviously a variant of Kawasaki’s, but it’s different with respect to somewhat older kids, perhaps more heart and other organ involvement. And that does appear to be an inflammatory, autoimmune type story. What are your thoughts about that?
Iwasaki: In order to understand whether autoimmunity is involved, we need to identify autoantigens. I know that a lot of groups are working on it, but currently we don’t really know the relationship between COVID exposure and the development of autoantibodies or even autoreactive T cells that may cause these types of inflammatory diseases. So identification of the culprit antigen is going to be key.
Verghese: You have an elegant paper on a mouse model for COVID, which would be a wonderful way to tease out the separate elements of this disease. Talk, if you would, about developing that model and what you see as its future applications.
Iwasaki: Thank you for bringing that up. The mouse model we’ve created is a very easy-to-use and versatile model where we transduced the mice with adeno-associated virus (AAV)-encoding human ACE2. Its versatility comes from the fact that you can use any kind of background animal — whether it’s knockout, transgenic, or reporter mice. You simply have to transduce the mice with AAV for 14 days and until the expression ACE2 becomes fully developed. Then you can infect these animals with the human SARS-CoV-2.
That publication reflects the beginning of what we have done. We are looking at all kinds of immune response players, different cell types, different cytokines, and their relationship to protection against this virus, and also pathology that results from immune activation. In that first paper, we were able to examine the role of the type I IFNs, going back to Eric’s question, and showed that, at least in the mouse model, type I IFN appears to be incompetent for blocking the virus replication. If you look at the viral titer of animals that are either wild type or IFN-receptor knockout, the titers are not very different on different days of infection. What is different is the pathology we see in the line. Type I IFN induces a lot of chemokines that attract leukocytes into the lung. Unfortunately, it’s sort of fueling the fire by recruiting these leukocytes into the lung, and at the same time, not being very competent in blocking the replication of the virus.
This sort of mimics what we found in human patients, which was this long-term, smoldering type-I IFN response being associated with mortality and length of hospital stay. In the animal, we’re seeing the same thing: endogenous IFN is not very competent to block the virus replication and instead it’s leading to pathologic effects. This is the first insight we can see in parallel, from human to mouse and mouse back to human. Now we’re looking at other types of immune responses and whether they’re also contributing to pathology vs protection.
Verghese: Do you think this is a problem with the level of IFN or is it a qualitative defect in IFN’s ability to handle this particular virus?
Iwasaki: It is probably a mixture. We’re probably not getting a robust IFN-induced response because SARS-CoV-2 encodes numerous evasion mechanisms to block the induction of IFN and even IFN receptor signaling altogether. That’s why when we knock out the IFN receptor from the host, we don’t see much of a difference in bioreplication. It speaks to the ability of this virus to evade and sort of suppress the IFN response. That’s why having a recombinant IFN treatment early might make sense.
Topol: As long as it isn’t overridden somehow by this very elusive, difficult virus.
One of the things you’ve been uniquely championing is that we’re not paying enough attention to the mucosal immunity, the immunoglobulin A (IgA). Almost all of these 200 vaccine programs are working on spike protein shots or some component of the virus to provide the antibody and T-cell response. But there’s a whole other path, which would be to get to mucosal immunity. Maybe you could tell us more about the IgA response and why you think that’s not being given enough priority.
Iwasaki: Because of the nature of this pandemic, people are trying to develop a vaccine as soon as possible, which is important for us to return to regular society. So I think the first generations of vaccines will be the ones that elicit the most robust neutralizing antibody in circulation. That has a huge importance for preventing further spread of this virus. But given a little time and more effort, people are also starting to look at mucosal vaccines and potentially delivering the same kind of vaccine through the nose instead of into the muscle. When those vaccines are tested side by side, I believe we’ll see a difference in providing a sterilizing immunity in the mucosa vs inhibiting disease after exposure. I’m not discouraged by the fact that these vaccine companies are developing a systemic vaccination. But in the future, I think we should be looking more into mucosal vaccines.
Topol: Do you think they would be complementary? The vaccines that are in rapid development, phase 3 trials, are not going to prevent the infection or achieve sterilization but will modulate the response, whereas the IgA antibodies could actually block the infection in the first place. Could they be paired at some point?
Iwasaki: I think so. But if the systemic vaccines are providing a robust blocking of disease, that would be a great first start. By the way, sterilizing immunity is not the goal of most vaccines.
Topol: Some of the vaccines in trials are claiming it in their papers, in non-human primate models. It doesn’t need to be achieved. But it’s interesting how it’s being claimed right now.
The geneticist George Church recently took a mucosal vaccine via nasal spray. And it was criticized because it is like the Russian vaccine — there’s no safety profile. But George is quite a pioneer. It would be interesting to get actual results from the crew of people who are testing that vaccine.
One of the controversial issues right now is what’s been called effective herd immunity, a natural immunity that is being debated. Most experts agree that you need to get 70% or 80% of the people to be immune to establish herd immunity. But recently we have seen cases in the United States — particularly in states that were hard hit, such as Texas, Arizona, and Florida — where, for some peculiar reason, it looks like their infection numbers are going down, and not just because there’s less testing, which is another confounding issue. That has led to some people theorizing that maybe this 20% of people who are infected is providing a lower amount of spread. What is your sense about this? Is it an explanation for reduction in cases?
Iwasaki: First, I have to say that I’m not a mathematical modeler. I can only provide my insights as an immunologist. It’s a little dangerous to rely on herd immunity at this point to open up society because herd immunity requires a significant proportion of the population to be immune to the virus. Even if there is a community in which 20% of the people have immunity, we don’t know how long this lasts because they acquired it through a natural infection. We just don’t know enough about the protective level of antibodies and T cells in people who recover from this infection to know how long such a protective immunity will last. Also, every individual may be slightly different. I believe it’s premature and dangerous to depend on those numbers without a vaccine that can be distributed throughout the population.
Verghese: You know, before we leave innate immunity, I have what may seem like a naive question, but I know you’re an expert in mucosal immunity. Early on, some kitchen-sink wisdom suggested gargling and using nasal rinses as a preventive, because if the virus is going to attach, that’s the first step. Is there any logic to that? Might that be a strategy before we actually have immunity at a local level? The strategy would be to somehow block virus attachment to receptors in the nasopharynx.
Iwasaki: I’m not sure how much virus saline nasal rinses would actually get rid of. However, I am a proponent of humidity. We’ve published a mouse model looking at the role of humidity in our respiratory tract and immunity to influenza virus. We saw that a 40%-60% relative humidity helped the animal to be able to remove the virus from the respiratory tract through mucociliary movement. Whereas if you kept the mice in 20% humidity, which you will find indoors during the winter months, those animals did poorly because they were not able to clear the virus. So there is a natural mechanism for removing the virus, which we can take advantage of, by maintaining the relative humidity at certain levels.
Topol: Getting back to the vaccine front, there’s been a lot of consternation about what the goal should be. Some vaccine candidates produce relatively little T-cell response, particularly CD8 T cells, whereas others do. What would you envision as the ideal response, if you were to look at a vaccine and try to project that it’s going to achieve a durable protection? What’s your sense about that?
Iwasaki: If a vaccine can elicit very high levels of neutralizing antibody, that would pretty much block the spread of the virus in the person. T-cell immunity may not be needed at all, meaning CD8 T cells. Of course, to achieve that kind of antibody level, you needed to have CD4 T-cell help. So by default, such a vaccine would have elicited very good T–helper cell immunity. But whether you need robust CD8 responses and neutralizing antibodies to confer protection is an open question. I’m not sure we need both of them. If you only saw the CD8 T-cell response, you would not achieve a very rapid clearance of the virus, because antibodies are needed to really block the spread of the virus. CD8 T cells are great at picking off the virus-producing factories, but they’re not going to prevent the infection altogether. In terms of importance, it’s the high neutralizing antibody titer, and if there is a robust CD8 response, that’s sort of icing on the cake.
Topol: Most of the vaccine candidates have quite a good profile for neutralizing antibody response, so that’s encouraging. Can you make any inferences from the SARS epidemic and what worked then, since there’s a similar structure? Certain people who had SARS apparently still have signs of immunity to SARS now.
Iwasaki: Several vaccines were studied during the SARS-CoV-1 outbreak. Some vaccines worked really well and others, unfortunately, induced an inadvertent disease enhancement. For example, the double-inactivated SARS-CoV-1 virus vaccine elicited an antibody-dependent enhancement (ADE) type of response. But I’m encouraged that even these kinds of inactivated vaccines against SARS-CoV-2 are inducing pretty robust neutralizing antibody responses with no evidence of ADE. Thus far, none of the vaccine candidates out there have reported any major adverse effects or ADE type of effects.
Topol: Antibody enhancement is a paramount concern. Do you believe that, or an immune complex disease. won’t be a big issue going forward, or at least only on a very rare basis?
Iwasaki: That is my hope. The first two clinical trial phases haven’t reported any of these effects; however, it’s really during the phase 3 trials that we find out if there are adverse effects in rare cases and why. That’s why it’s so important not to rush that process.
Topol: We’ve noted a big gap between male vs female risk for COVID-19. Can you tell us what you think is causing that difference?
Iwasaki: We’ve been actively looking at sex differences in immunity to COVID-19. So far, we’ve found that male patients who develop severe COVID tend to have very low T-cell activation, whereas female patients who develop severe COVID have elevated innate immune cytokines. Thus, there seems to be a different way in which women and men respond to the virus. The lack of T-cell immunity in men is interesting, because when we look at age on the X axis and T-cell activation on the Y axis, we see an age-dependent decline in T-cell activation in men, but not in women. So this may have something to do with why there is more severe disease in men who are infected.
Topol: There seems to be a preponderance of women in the long-haulers with COVID. Have you made any connections with that?
Iwasaki: It’s tempting to speculate about a link between autoimmunity and these long-haulers, because the vast majority of autoimmune diseases have a preponderance in women vs men. Of the three hypotheses I listed, the autoimmune disease could be occurring in women and that may be contributing to the long-hauler disease. But without any data, I don’t really want to speculate.
Topol: I guess we’ve asked you to speculate a lot, which is fun. It’s great to hear your views.
I want to ask you about kids. We talked about MIS-C but we didn’t talk about this whole controversy now with schools reopening and that children are less likely to manifest disease. Perhaps their transmission is less. There’s been a lot of confusion about all of this. What are your thoughts about the unknowns in children as compared with adults? And is there a difference between younger and older kids?
Iwasaki: There definitely seems to be an age gradient of symptoms associated with exposures to SARS-CoV-2. But the symptoms do not relate well to whether they’re infectious. Children can have a high titer of replicating virus in their noses even without symptoms. Whether they become spreaders without knowing they are infected is a real possibility. I have two children and they are dying to go back to school; they’re so sick of spending the entire summer with their parents. So I totally get it. And it’s important for their mental health for them to be with other children. But I do worry about their ability to transmit the virus, even if they don’t have symptoms.
Topol: That brings up the question of their prior exposure to common cold coronaviruses and that children may have preexisting T-cell immunity. Perhaps that’s happening more in children than adults, in part, because there’s the temporal gap between when adults may have been exposed to the common cold before coronaviruses and when children were exposed. You’re in New Haven, Connecticut. Things are pretty quiet there, in terms of spread. Do you feel better about kids going to school and a lack of a transmission chain in a place like Connecticut?
Iwasaki: As of now, we have a low number of cases, which is wonderful. But a lot of travel occurs within the country from states with high case numbers to those with low case numbers, even though there is a kind of quarantine. So it’s difficult to know when we would feel safe because spreading may be occurring without really knowing about it. It takes a couple of weeks for the numbers to actually come up. I’m definitely feeling better than I was in April, but at the same time, I still take as much caution as possible.
Topol: Will you have your kids in virtual classes or will they physically go to school?
Iwasaki: They would be very upset if I said they can only go virtual. Right now, the school is planning to open and they will have in-person classes. But it’s a fluid situation. I’m keeping my eyes on what’s going on right now.
Verghese: I just wanted to follow up on a personal note. Initially, in March, all the research labs were shut down and gradually they’re trying to open. But with all the challenges of having people in the same space, how have you managed your research lab? You’ve been incredibly productive. Tell us about the challenges of being at work.
Iwasaki: I personally haven’t been at my workplace for a long time (just like Dan Nebert), in order to provide enough space for members of my lab to work. We did shut down quite aggressively in mid-March. The only type of research that was allowed was COVID-related research, so even though the university shut down, my laboratory kept going. In fact, they were working harder than ever, trying to study immune response in real time. My lab has been working hard but it’s also following guidelines of physical distancing and de-densification. We couldn’t have all the members of the lab working at the same time; it had to be one person per bay, so it did slow us down in that way because we weren’t able to all be there. Fortunately, everyone practiced physical distancing and used personal protective equipment, so no one was exposed to the virus during that time.
Topol: Let’s leave the pandemic for a bit and learn more about you, Akiko. You’ve had an amazing career. Your father was a physicist, your mother was an activist standing up for women’s rights. How did they influence you in other things that happened in your career to get to where you are today?
Iwasaki: They had an enormous influence on me. Growing up in Japan, seeing how much my mother had to struggle to even keep working, really taught me the importance of speaking up and of believing in yourself as a woman, to be able to state the problem and to address it head-on. My mother is an activist. She is soft-spoken and the most gentle person you’ll ever meet. But even with that sort of personality, she exhibited this strength. I definitely was influenced by that and I tried to emulate that in my life. Sex-based discrimination or racial discrimination unfortunately happens everywhere. So she taught me to be proactive and vocal about it without having to be extremely loud.
Watching my father through my childhood was interesting because I told myself that I’m never going to be a scientist; he’s always reading journals and thinking about science, and what kind of life is that? Then eventually I became that person. I’m afraid that I’m deterring my daughters from pursuing science because I do the same as my father. Science is difficult, but it’s also the most exciting thing for me to do. I can’t imagine another job for which I get paid to do what I love thinking about.
Topol: Well, all the things that happened along the way created a phenom. You’re a great educator as well as a great scientist. We’re pleased to have had a chance to visit with you today.
Verghese: This has been a real eye-opener for me — a clear way to understand the immune system. Thank you for being with us, and good luck with your continued research.
Topol: I recommend to everyone that, if they’re on Twitter, they follow Akiko Iwasaki (@VirusesImmunity) as their number-one source for really useful information. And if you’re not on Twitter, you ought to get on it, because she’s got a lot to offer.
Akiko, thanks so much for taking the time to join us. We’ll follow your work and your group very closely, because I know you’re going to unravel and deconstruct a lot of the unknowns we have today.
Eric J. Topol, MD, is one of the top 10 most cited researchers in medicine and frequently writes about technology in healthcare, including in his latest book, Deep Medicine: How Artificial Intelligence Can Make Healthcare Human Again.
Abraham Verghese, MD, is a critically acclaimed best-selling author and a physician with an international reputation for his focus on healing in an era when technology often overwhelms the human side of medicine.
Akiko Iwasaki, PhD, is an immunobiologist at Yale University and the Howard Hughes Medical Institute. Her research focuses on immunity against viruses at the mucosal surfaces. She is particularly interested in educating the public about the immune system and how it works; she is also an advocate for improving the culture of science and for students in science.