As both a bodily defender and meticulous record-keeper, the immune system knows exactly which infectious threats it has already vanquished. Every cold, flu, and vaccine that has ever triggered an immune response leaves an indelible mark on the immune system. While much of this immunological memory resides in B and T cells, these cells are not the immune system’s sole chroniclers.
Using machine learning and systems biology, John Tsang, a systems immunologist at the National Institute of Allergy and Infectious Diseases (NIAID), and his colleagues discovered that multiple different kinds of immune cells become altered in response to an immunological trigger such as a vaccination (1). They found that in healthy individuals after vaccination, subsets of immune cells increase and decrease in number while also changing their gene expression.
A computer scientist turned systems immunologist, John Tsang studies how differences in individuals’ immune system baselines contribute to how well their immune system responds to vaccination.
Credit: John Tsang
“These cells were not measured in the antigen specific manner, meaning they're not specific to particular pathogens. They're just cells that circulate around,” said Tsang. “It's really a circuitry of cells that are being maintained in certain states by certain molecules, and these cells also interact with each other.”
Tsang and others refer to the record of every infection or vaccination a person has ever faced as an individual’s immune baseline. Due to someone’s personal genetics and environmental exposures, their immune baseline is unique to them. Leveraging the power of systems-level approaches, Tsang investigates how people’s immune baselines influence how their immune system responds to a new infection or vaccination. By identifying biomarkers of a healthy immune response, he hopes to develop quantitative models to better predict how individuals will respond to vaccination. These biomarkers will also help researchers develop ways to boost the immune response to vaccination in people who may be immunocompromised or otherwise unable to generate a strong immune response.
How did you first become interested in systems immunology?
Before I started in immunology, I studied computer engineering and computer science and then I trained in systems biology during graduate school. One thing that really caught my attention about immunology was the extensive heterogeneity of immune responses in the human population. When I first talked to immunologists about that, they saw that variation as a roadblock to understanding the immune system. They focus on animal models because they can control everything and make things very consistent. From my own quantitative way of looking at it, the heterogeneity of people’s immune responses is actually a huge advantage. Because everybody has different responses, if we give them a perturbation like a vaccine, we see a variable outcome. From these outcomes, we can start to infer the components, variables, and interactions that shape the immune response.
Does a person’s immune baseline change often?
No, it tends to be fairly stable within individuals, but the baselines can be very different from person to person. It's like a race; there are different cars, and they're tuned in different ways. When someone opens the gate, those differences are quite relevant to how individuals respond to immune perturbations. Based on more recent work, it's also quite clear that even though the immune baseline is temporally stable and quite distinct from person to person, it can be shaped by different perturbations and exposures over time as well.
Compared to vaccination, how does infection influence a person’s immune baseline?
Infection is likely an even bigger shaper of the immune baseline than vaccines are. My colleague Rachel Sparks and I recently designed a study to look at how a mild SARS-CoV-2 infection shapes these immune baselines (2). We compared people who had symptoms of COVID-19 and then recovered versus age and sex-matched people who never had COVID-19. When we did this work in the fall of 2020, none of the participants had received a COVID-19 vaccine or had come down with the flu. The idea was to see whether even a mild SARS-CoV-2 infection could change the immune baseline status. We also vaccinated all of the participants with the flu vaccine. Because the flu vaccine does not challenge the same immune memory that arose from a SARS-CoV-2 infection, we wanted to see if SARS-CoV-2 changed how people’s immune systems responded to a new vaccination.
How did a prior mild SARS-CoV-2 infection affect people’s immune baselines and their response to the flu vaccine?
In the group with a prior SARS-CoV-2 infection, we detected immune baseline changes in various cell types, including in the innate immune cells, monocytes, and certain T cells. In response to the flu vaccination, males but not females also had an altered innate and adaptive immune response. This was a nice example of how an infection, in this case a mild one, can generate a systemic difference even long after it’s resolved.
To understand the stability of the immune baseline in people who recovered from SARS-CoV-2 infection, we tested whether the immune baseline parameters correlated with the time since the initial infection, but most of the parameters didn’t resolve over time. Rather, these parameters get set and reset. These “set points” become fairly stable within individuals and can influence a future response to an exposure, in this case to a flu vaccine. This gave us a proof of principle of how these initial baseline states can be established.
Is it possible to predict a person’s potential immune response to a vaccine based on their unique immune baseline?
We can start to do that now. One can propose several parameters to measure that would give association results, but there are a lot of other variables that may interact to give rise to more accurate predictions. For example, age and sex play a role, but if we’re talking about the immune response to a flu vaccine, other more flu-specific information may also help predict a person’s response to it.
The concept of vaccine adjuvants can help us interpret that kind of data. Adjuvants enhance the innate immune response to the vaccine. What's very interesting is that some individuals, because of prior infections or vaccinations, have immune baselines that look a lot like typical immune responses after vaccination. In other words, these individuals are already naturally adjuvanted, so when we give them a vaccine, they immediately respond better. What gives rise to that natural adjuvant? What maintains that state? That's something that we're trying to understand. We can even think about how we might give immunocompromised individuals something to boost their immune baselines and improve their immune responses to vaccination. We could modulate and then vaccinate, rather than designing different adjuvants for every vaccine.
What do you find most exciting about studying the immune system at a systems level?
The immune system is a perfect arena for studying the basic principles of how a complex biological system operates and how interactions among cells and molecules give rise to emergent outcomes. It's really complex, but it's a really cool model system. I also find it fascinating that the immune system seems to be the anchor across health and disease. If we look from cardiovascular disease, metabolic syndromes, to cancer, we can always find the immune system playing an important role. Often, we find that immune cells are signatures for predicting outcomes or are associated with disease severity. Internally, the immune system plays a crucial role in keeping us healthy. That's why it's so exciting; we study something that touches multiple aspects of health.
What is one of the biggest challenges of studying human immunology at a systems level?
Looking at human populations, it's quite easy to get information from blood, and we can draw from the same individual over time. On the other hand, blood is not where things happen. Let's take an infection that typically happens in the upper airways. The immune response itself occurs in lymphoid tissues. In humans, it's not easy and is sometimes impossible to even get those kinds of samples, so we have to make inferences based on blood alone. It's less limiting in the sense of finding biomarkers because there may still be quite a bit of information that we can find in the blood that reflects the events and states in tissues, but from a more mechanistic and quantitative understanding standpoint, it is a major challenge. We need live, deep tissue analyses where we can take a peek at where immune cells are and what their states are in the body.
What are some of your long-term goals for your research?
I’m really interested in what I call “bottom-up” systems immunology, which is to look at all these cell interactions and molecular networks and learn how they work together to lead to complex outcomes. For example, if we wanted to triple the dose of a vaccine — let’s say we went from a Pfizer COVID-19 vaccine to a Moderna COVID-19 vaccine, which is about three times the dose — without doing a clinical trial, can we predict what the differences in the immune response would be for an average person? Nobody can answer that question right now. We don't have a good quantitative model for how the immune system works. That's the Holy Grail. I hope that in the coming years, we can develop new approaches so that for the simple question I just posed, we wouldn’t actually need to look at 10,000 people to get an answer. We could predict what might happen and then do a small-scale experiment to see if we’re right or not.
I also want to apply these systems-level approaches and concepts to better understand health maintenance. How do humans as individuals deviate from homeostasis? How can we restore immune homeostasis quickly so that we can intervene or treat with something well before someone gets a disease? I’m very excited about understanding, intervening, or even modulating a person’s immune system before they get infected with a pathogen or develop a disease.
This interview has been condensed and edited for clarity.
References
1. Tsang, J.S. et al. Global Analyses of Human Immune Variation Reveal Baseline Predictors of Postvaccination Responses. Cell 157, 499-513 (2014).
2. Sparks, R. et al. Influenza vaccination reveals and partly reverses sex dimorphic immune imprints associated with prior mild COVID-19. Preprint at: https://www.medrxiv.org/content/10.1101/2022.02.17.22271138v1
