Single-Cell Proteomics Tackles Vaccine Development

Sponsored content brought to you by

Vaccine development can be daunting. The overarching goal to create long-lasting immune protection requires an in-depth understanding of the immune system’s response to an infectious agent, immune monitoring to determine and predict the building of the protective response, as well as cytokine-level monitoring for potential toxicities related to cytokine storms.

GEN spoke to five leading researchers to hear their views and strategies on infectious disease research and vaccine development.

 

GEN: How does understanding and detecting immune response and function affect developing therapies and vaccines for infectious diseases? How do single-cell tools play a role?

Moriya Tsuji: Before trying to develop vaccines or immunotherapies, we need to identify and characterize the immune response, particularly the protective response, against the infection. In human populations, a certain subset is susceptible and another resistant; the same is true in small animal models. It is important to compare the response of both populations at a single-cell level to understand what separates the two.

In small animal models, such as a mouse model, we can challenge both resistant and susceptible strains with a pathogen and monitor survival rates, as well as analyze single cells at different time points to determine which cytokines/chemokines dictate resistance or susceptibility. To determine the efficacy of vaccine candidates in vivo, animals can be immunized with each and challenged with the virus of interest. If one vaccine exerts more protective efficacy than others, you can then evaluate the differences in the immune response.

Using a cutting-edge, single-cell assay, you can also harvest specific tissues/organs from the vaccinated animals and measure the cytokines, T cells, B cells, macrophages, dendritic cells, etc. to see which cell types producing particular cytokines are most correlated with the protection. With this information, you can go back to the drawing board and design a vaccine that would induce such cell types and cytokine responses.

Jim Heath: The immune system has heterogeneous cell types that work at different times in different roles. In a cascade of coordinated events, one expects that if something disrupts the process, the patient outcome is much poorer. The ability to monitor cellular activity over time to quantitatively map out an immune trajectory in an individual patient is key to understanding how similar patients are likely to respond and can inform on timing of therapeutic interventions. At one point you want your T cells to aggressively pursue the virus, but at another the inflammation can spiral out of control and you might want to suppress it.

In a normal, well-validated, double-blind clinical trial everyone begins at the same time, and, hopefully the ones who got the drug cross the finish line faster. In the COVID-19 pandemic, patients are randomly distributed around the track and you have no idea when some finish or if the drug had an effect. This confounds how to interpret patient response; you need detail.

If you just consider cytokines in the blood, you lose that whole concert of immune cell behavior, and you cannot interpret. Single-cell tools begin to parse that out. In this circumstance, you need to treat every individual patient as their own trial and you can only resolve that with single-cell tools.

Stanley Perlman: For understanding immunizations, measuring antibody responses is very critical because you want to learn how people respond. The ultimate question is: if they are exposed to the pathogen in question again, will they get the disease? Single-cell sequencing and other measurements give you a fine-tuning of what is going on. Although they may not be the most critical thing in the beginning, if we have that information it can only help us in the end. In the beginning we just want to know what protection is and have we reached it with our vaccine or prior infections. Those are things we do not know yet with COVID-19.

Rong Fan: Years ago, a review article by Mario Roederer demonstrated that using low-plex flow cytometry, you could show a handful of effector proteins/cytokines/chemokines produced by single T cells that determine the immune function. The ability to simultaneously produce different effector proteins correlated with the potency of immune cells against the infectious pathogens or the possibility to clear dysfunctional tissues, such as tumors. Single-plex T-cell activation assays remain the standard in the vaccine industry; however, they are not sufficient.

That led me to develop single-cell cytokine secretion profiling in a microfluidic device, which IsoPlexis commercialized; the primary purpose then was to use it to evaluate T-cell vaccines for AIDS, supported by Bill and Melinda Gates Foundation, which was the very first grant I received after joining Yale. With single-cell multiplex cytokine profiling tools, scientists are in a better position to develop the most efficacious treatments. For example, in COVID-19, some patients’ T cells very likely become activated early on, overreact, and quickly go into a full life cycle and apoptosis. These patients exhibit a significant reduction in leukocyte counts. How to accurately measure the activation status of different immune cells upon exposure to the coronavirus is important to understand how to mitigate.

Tina Wang: Immune response is important for vaccine development because if you can understand the immune factors, how they function and correlate with host protection, then you have a better idea of how to develop a vaccine. If you understand the immune factors involved in the pathogenesis, then you can use that as a target to develop a therapeutic approach for controlling active disease. Single-cell analysis can provide more information to understand immune cells on a single-cell basis and how the functional cells play a role in immune protection and pathogenesis. This helps to develop better vaccines and immunotherapies.

 

GEN: What models are available today to look at the innate and adaptive immune responses in relation to infectious diseases and cytokine storms?

Stanley Perlman: We do not necessarily need other models; there are so many cases of COVID-19 that we can use human specimens. If we can use our human infections well, we can learn a lot about what is going on during infection. For COVID-19 there are several animal models where animals do not get very sick. They may not be totally useful because much of what we care about is people getting sick and having pneumonia, and immune responses that contribute to that pneumonia. For mild disease, you can look at monkeys, ferrets, hamsters, and mice but if for severe disease, animal models are not there yet.

To have access to human samples, one needs informed consent and Institutional Review Board approval. Those are critical steps, but if one can use human samples, then you do not have a discussion about the relevancy of an animal model system.

Moriya Tsuji: The innate and adaptive immune response against an infection can be measured in various animal models, including mice, rats, ferrets, hamsters, and nonhuman primates. All models have pros and cons. In nonhuman primates, there are ethical, financial, genetic background, and reagent issues. Ferrets have been used for influenza studies, and hamsters have been shown to be very receptive to COVID-19. The downside is that there are almost no immunological reagents available for these two species. In this regard, the mouse model has many advantages because there are many such reagents available. Furthermore, you can harvest organs easily to measure the tissue-specific innate immune response after infection or vaccination at early time points and the tissue-specific adaptive immune response at later time points, for example at 2, 6, and 12 weeks.

There is a big difference between the immune response of mice and humans. More than five years ago, I began generating a humanized mouse model that mimics the human immune system by transducing genes that encode human leukocyte antigens and human hematopoietic cytokines in highly immunodeficient mice, followed by the engraftment of human hematopoietic stem cells. In my work with malaria vaccines, this meant I could administer a human vaccine and measure a human immune response. Humanized mice are not a perfect model, since endothelial, epithelial, and other cells are still mouse-derived, and as a result, human viruses cannot infect well in challenge studies. We know that the coronavirus (SARS-CoV-2) infects humans via the angiotensin-converting enzyme 2 (ACE2), particularly in lung alveolar cells. I am now designing a human immune system mouse which expresses human ACE2 in the lung to try and replicate human infections in a mouse model.

Tina Wang: It is better to study the innate and adaptive response using in vivo models. Within vitro models, it is hard to understand unless you identify adaptive immune cells to study. In vivo you can study kinetically because the innate response is boosted very early and the adaptive immune response develops later. You can use tools, like specific antigens, to determine if the response is pathogen specific, which is related to the adaptive response. For vaccine development, the adaptive response plays a very important role because you have a memory response to help the host prepare and prevent disease.

It depends on the pathogen, but generally in preclinical studies we try to develop an animal model that can either partially or fully mimic the human disease so we can study immunity. This is the best way, but sometimes there is no feasible animal model—or there may be a model but a lack of critical reagents to study the immune response.

In those circumstances people also study in vitro, and in recent years organoid systems have become popular, which are a better way to mimic human disease than traditional cell culture. They provide a better system to understand infection and immune response. Groups used organoid systems to study the Zika virus.

When bacteria, viruses or parasites infect cells or animals, the immune system is boosted and you can measure innate cytokines. It has been reported in COVID-19 that a cytokine storm plays an important role in the viral pathogenesis. The main cytokines involved are the pro-inflammatory ones, such as IL-6, TNFα, and IL-1β.

Rong Fan: In terms of the adaptive response, when T cells and B cells see antigen from the virus processed by antigen-presenting cells and presented on the surface, they recognize it, become activated, and then quickly expand, and the individual begins to adapt immunity against the pathogen. This happens annually in seasonal flu. If you already have a vaccine, antigen-specific T cells and B cells have been induced by the vaccination process and some turn into memory cells in circulation, then quickly respond when infection occurs.

Once activated, the cells are in a different state and produce different effector cytokines to battle the viruses themselves or viral-infected cells. Some return to memory stage and produce different proteins. At any given time of the infection, T cells or B cells are always a dynamic heterogeneous population undergoing a complex differentiation process and display a wide range of effector functions that is difficult to dissect. Looking at single cells in their full-range functional states and performing a highly multiplexed cytokine evaluation are very important to figure out what actually constitute such heterogeneous populations.

For the innate response, it is more complex. Scientific evidence shows that innate cells may have certain memory or can be trained to develop certain memory; although not so antigen specific, they can be trained to respond to certain infectious pathogens in a more effective and rapid manner, representing a new avenue to develop vaccines. To identify, characterize, and quantitate the cells involved and how they remember and respond to the different pathogens, you need a single-cell resolution and highly informative analytic tool.

Jim Heath: You can harvest innate and adaptive immune cells from blood and look at their genetic regulatory network and surface markers. The adaptive immune response generally recognizes something very specific about the foreign entity through T-cell and B-cell receptors, which is why it takes time to develop that recognition. The genes for the receptor repertoire are built by genetic shuffling and are different for every T and B cell. This gives tremendous diversity from a small number of genes, but it makes sequencing and analysis difficult.

Using different assays, we can understand what specific fragments of the coronavirus (SARS-CoV-2) the immune cells are seeing. The coronavirus spike protein binds to human ACE2 receptors, part of it cleaves off and then another part flips around, and the virus injects the RNA through the cone. This protein is by and large most of what your acquired immune system is going to be able detect—the whole protein that B cells evolve to see or an antigen fragment that T cells see.

Some fragments will lead to protective responses, and some will just exhaust your immune system if the T cells are activated against fragments nonessential to the virus. Single-cell analysis of patient samples shows many exhausted T cells. At the heart of vaccine strategies is how to resolve what is protective versus immune dominant. The innate and adaptive models are the result of years of research. With single-cell technology, we can put these models to an extremely severe test at high resolution.

Cytokines are general molecules that help the immune system communicate. Cytokine storm is a symptom that many patients exhibit; if you can trace these cytokines, you can determine the source. It will not be the same for everyone, but it will have some common characteristics and is controllable. COVID-19 is complicated. We have the tools to parse through the complexity, but those tools are confounded by the very large heterogeneity in the patient population.

 

GEN: What are some challenges in preclinical vaccine and therapeutic development that can be overcome by having potency tools and better ways to characterize human immune response? Which cellular analysis tools will reveal this potent response?

Stanley Perlman: Generally preclinical means experimental animals. As a community, when we were doing these studies methodically and carefully, we wanted to make sure that whatever we used worked in small animal models, then worked in nonhuman primates and was safe without side effects. Typically,  we would want to monitor experimental animals clinically and for immunopathology with blood samples or lung dissection, and we would conduct challenge studies and measure antibody and T-cell responses. Then we would move to people and do very methodical testing. A lot of that is circumvented or ignored by the urgency of the COVID-19 situation.

Now we are doing the same things, but we are doing them simultaneously to get vaccines up and running for human populations. Antivirals are different; there is no question that you have to show efficacy, and that has not always been done well. If you use a well-established vaccine platform and show safety, then you can move along a little faster. In other words, if we know that a particular vaccine platform works well, and then if we know it works well for a protein that is important for another human virus, and if all we did was use the same vaccine strategy but replace the known protein with a SARS-CoV-2 protein, there would be less worry that something bad would result.

When vaccines started in the 1950s, we knew little. We did the best we could and were lucky that the polio vaccine killed so well.

Tina Wang: In order to better understand and characterize human immune response, we want to have a feasible animal model so we can study the infection in vivo to understand the immune response and also to determine whether the immune response plays an important role in protection or pathogenesis. For that reason, you want to know more functional detail since different cells may play different protective or pathogenic roles. You want to determine the multifunctional cells, and if you have a tool to do that it would assist a better understanding. Single-cell analysis allows you to understand the function on a single-cell basis, along with more dynamic features about the immune function.

The challenge for the preclinical study is a lack of tools to fully understand the immune cell function. If you have tools to understand, like microarray proteomics, you can find more information on single cells and the multifunctional aspects that can help you understand their role. This also depends on what pathogen you are studying. In general, the more potent the tool you have, the better the understanding of immune function.

You can also do experiments to determine how the immune system correlates with host function, for example, you can deplete the cells using antibodies, or use genetic marker or transgenic mice models to determine that particular immune cell function correlation with host cell protection or pathogenesis.

Moriya Tsuji: One of the best methods available so far is IsoPlexis’ multiplexed technology, which measures more than 40 cytokines/chemokines at the single-cell level. This is very powerful. You can pair this technology with other cell-based assays, such as multicolor flow cytometric analyses. Flow cytometry is well established and accessible, and with recent improvements you can determine dozens of surface markers on the same single lymphocyte.

In the case of COVID-19, sick patients exhibit lung pathology and secretion of IL-6, IL-10, TNFα, and possibly IL-1 in severe cases. These are most likely produced by lung alveolar macrophages, and the pro-inflammatory nature of these cytokines may be one of the causes of mortality. At the same time, the CD8+ and CD4+ T cells that secrete interferon gamma are reported to be depleted, most likely due to exhaustion. You need multiplexing tools to study these mechanisms in-depth.

Using multiplexed technology, you can determine the inflammation caused by the innate immune response. Knowing these details, you may be able to create a vaccine, for example, which elicits immune cells that would inhibit IL-6 in response to the infection in advance, and thus induces a protective response to combat the pathogenic one in a preventative fashion.

Jim Heath: If we know which immune responses are protective versus noisy, that can help us design a vaccine. Typically, multiple vaccines are developed simultaneously; there are a host of COVID-19 vaccines under development. Some are quite novel technologies. Now you can use the genetic material to make the neutered pathogen and not the neutered pathogen itself like the smallpox virus. This looks more like a natural course of infection and may possibly lead to a better immune response.

You can chop up the virus without the RNA, but we are not sure what is immuno-protective or noise. Coronaviruses change over time. If a vaccine worked against SARS-CoV-2 there is no guarantee that it will work in the future. If you really understood the details of the virus in terms of genetic mutations, you could design a vaccine that has a much longer protective mechanism. For COVID-19, the strategies and patient population are complicated.

Rong Fan: The test of vaccine-induced antibodies in animals was often conducted by immunization of the animal against the viral component. Then, a simple binding assay allows us to see if the induced IgG antibodies can bind the viral proteins. To evaluate the vaccine-induced T cells, you can spike the viral peptide antigen in dendritic cells that present the antigen on the surface. A similar binding assay allows you to see whether the induced T cells can recognize the viral peptide antigen-specific signal.

But you do not know if the binding provides a functional consequence. It may occur but not induce the required immune responses to, for example, recruit other partner cells, mount antiviral activity, perform cytolytic function, and so on, in order to completely clear the infection; the T cells might just recognize the viral component but cannot do their job properly. Eventually you need an assay to confirm the functional efficacy. This is the most challenging and time-consuming step.

Characterizing the functional outcome allows determination of a signature that correlates with potency and durability to predict the most efficacious T- or B-cell responses and the corresponding vaccine candidate before you complete a six-month animal test. For example, using the coronavirus (SARS-CoV-2) spike protein, you immunize animals to develop the antigen-specific T cells. Then perform monoclonal expansions of the polyclonal population, and do detailed analysis of those T-cell functions. Next you sequence the T-cell or B-cell repertoire to determine which clone is the right therapeutic T cell. This is time consuming and should be carefully monitored at every single step as you cannot do many repetitions of the trial-and-error process given the urgency of this pandemic.

If single-cell functional signatures can predict the outcome in animal or patient, it would be faster to characterize the induced T cells without monoclonal expansion and then determine the functional profiles along with TCR sequencing to come up with the optimal design without a lengthy screening process to speed up the vaccine development. Single-cell tools give you that diverse characterization within several days.

 

GEN: What role do you think assessing and understanding a proteomic cytokine response from immune cells and more systematically in bulk will play in developing better vaccines and therapies for diseases such as COVID 19?

Moriya Tsuji: I want to emphasize the importance of measuring the cytokine response. Recent data from China demonstrated that 10 out of 175 patients who recovered from COVID-19 were seronegative, which means that they had no antibodies. Nine out of ten of those seronegative patients were under 40 years old. This study suggests that other than antibodies, factors such as T cells and the cytokine response may have contributed to protect these seronegative patients.

In terms of the antibody (or humoral) response, many vaccines against the Dengue virus have been shown to produce antibodies that cause a phenomenon known as antibody-dependent enhancement (ADE). The induction of ADE increases virus infection, and as a result, these vaccines actually make the disease worse. This is why some scientists are hesitant to make a vaccine against COVID-19; there is some indication that some people may have antibodies that cause ADE. Therefore, you would have to design a vaccine very carefully in order to induce antibody response that mediates protection, but not ADE.

In view of this, multiplexed proteomic assessment of the cytokine response becomes quite important. The response could be from macrophages, T cells, or other cells, separate from the antibody response. This bulk and cutting-edge assessment could lead to the identification of an indicator of a protective or pathogenic immune response that could lead to morbidity and mortality. Such data may therefore have predictive value, and is particularly important for COVID-19 due to the urgent need for therapeutic or preventive vaccines.

Stanley Perlman: Knowing more about the fine points of the immune response will help us make better vaccines. Until recently we did not have the potential to evaluate the cytokines/chemokines, etc., but now that really could help. You can measure single cells from people and measure the cytokines they make, do RNA analyses, flow cytometry—these tools are so sophisticated nowadays that on a single cell you can do fifty measurements, with RNA a thousand measurements. There are a lot of things we can do; a challenge is interpreting everything. We get a lot of data and you want to make sure you can make sense out of it.

Jim Heath: The advantage of looking at immune cells at the single-cell level is that cytokines inform about function. You can tell the different roles of the immune cells and the different subroles of T cells by the functional cytokine signature. That is really important. The cytokines give you a real feeling of not only what the immune system is doing but also where it is headed near term. That is important for anticipating how the immune system is reacting to a vaccine or for diagnosing a cytokine storm.

If you take a drop of blood and look for certain types of T cells, like CD8+ T cells, you might have a million but probably about 1–10,000 are really dominating. If your T cells are going on the hunt against an infection, the queen bees controlling the hunt are the important ones to look at and understand. You can only see them if you do a single-cell analysis on a lot of cells to capture that small percentage; otherwise, it is very hard it is hard to resolve any of the hard questions.

Tina Wang: COVID-19 is caused by a novel coronavirus that has a high homology with the SARS-CoV-1 virus but also a lot of difference in terms of virulence, transmission, and permissive cell type. A lot of things need to be known. A potent tool to provide more proteomic information would be very helpful to understand how immunity plays a role in the disease. We know in general that cytokine storm could be correlated with disease severity and ADE, and we need to understand the underlying mechanisms. A tool to analyze cytokines and other immune factors systematically would be helpful.

People are racing to work on vaccine development or immunotherapy. There are some antiviral and immunomodulatory candidates already, but almost none have been tried clinically. Using proteomic cytokine tools to analyze the data from patients who receive the trial therapies would be very helpful. Current reports are somewhat conflicting about the effects of antiviral agents because some were being used very urgently. There was no clinical trial specific for COVID-19; very limited data from COVID-19 patients are available.

If there is ADE that information is helpful to determine the design of the vaccine. Currently scientists are using different approaches for vaccine development until there is an effective one. There are many vaccines, such as inactivated virus, RNA, DNA, or recombinant proteins, and they induce different types of immune response. In vivo studies in animal models will test if ADE is induced during challenge studies.

In vaccine development, the minimum time is four or six months for the preclinical and clinical stages but could be much longer, especially if you want it for a particular strain of SARS-CoV-2. SARS-CoV-1 virus has already shown the ADE phenomenon, so you have to be careful to assess safety. Even if it is okay in an animal model, you still have to be careful when you go to clinical trial.

Rong Fan: Functional characterization of single immune cells is very important to better help you identify the top candidates during vaccine development. Eventually you need to look at the B cells and the antibodies circulating in blood, and whether there are enough of those and which cytokines are presented in blood to understand the systemic response. If there are adverse effects, you need to monitor what is going on in preclinical development as well as during clinical trials.

My research includes profiling of blood from COVID-19 patients, and we have seen many cytokines that are critical to monitor, especially in the lung. The early and later stages of infection are different opportunities. In the early stage during viral expansion, you can control and suppress expansion.

In the later stages, the virus is not the problem rather a systemic cytokine storm and organ failure. Antiviral therapies would no longer help. We need to determine how to analyze the system pathology in those patients and specifically tackle the problem by suppressing pro-inflammatory response. The response appears to be different across multiple patients and calls for precision medicine.

 

To download IsoPlexis’ Infectious Disease eBook, please visit IsoPlexis.com.