Thank you all for joining our session exploring the differences between the pediatric and adult response to critical illness. I'm going to focus on age-related differences in the immune response to viral respiratory infections. We've known for a long time that young children and older adults are more likely to develop severe influenza-like illnesses than other age groups. Here's a chart depicting hospitalization rates in Canada over a 10-year time period. And as you can see, kids less than 2 and adults over 65 are much more likely to be hospitalized with influenza-like illnesses. When you add RSV into the mix, you can see that kids less than 2 are much, much more likely to develop severe RSV infection than other age groups. In contrast, and for unclear reasons, SARS-CoV-2 does not have the same age distribution as other common respiratory viruses like influenza and RSV. So instead of young children and older adults being most susceptible, SARS-CoV-2 is really much more likely to be severe in adults over 65 than it is in other age groups and in children less than 4 appear to be relatively protected. Now these numbers are from hospitalizations in 2020 and of course they're much higher now but this age distribution has remained the same. There are a variety of changes that happen in the immune system over childhood that might impact the antiviral response and there's growing recognition that this is not exclusive to the adaptive immune system, that there are multiple changes that happen in the innate immune system, both in innate immune cells and in non-immune cells like epithelial cells that might impact the response to viral infections. And this includes changes in antiviral interferon signaling. There's also of course the acquisition of adaptive immunity throughout childhood, including the increase in the antibody repertoire and a decrease in clonal diversity among T-cells. To add to the complexity, there are both short and long-term consequences of environmental and pathogen exposure that impact the subsequent response to challenge, some of which enhance the immune response and some of which diminish it. Due to these changes in innate immunity over time and the acquisition of adaptive immunity, one might expect that the consequence on viral infection in childhood is that children would be less able to control viral replication and less efficient at viral clearance. However, multiple studies in influenza and growing studies in other viruses do not support this to be the case. This is data from influenza, both seasonal and pandemic, and in red are children and in gray are adults. And as you can see, viral load in both children and adults is highest closest to symptom onset and clears at a similar rate in both age groups. And here is one of now many analyses in SARS-CoV-2 looking at viral load in children and adults, demonstrating that there is no difference in the amount of virus recovered from the noses of children and adults infected with SARS-CoV-2, suggesting that differences in viral load alone cannot explain differences in disease outcomes. In addition, there's no clear data that viral load is associated with severity of illness. Here was a study in children with influenza showing that there was no clear association with viral load here on the left and total symptom score. And similar analyses have been performed for other viruses. However, in this particular study of influenza, there was an association with symptom score and type 1 interferon in the nose. So here in the middle, interferon alpha, and on the left, the chemokine secretion MCP3. Our group has long been interested in why children are more susceptible to influenza virus than adults and why they have an increased risk of hospitalization for severe disease. And these data associating type 1 interferon with outcomes and inflammatory cytokines with severity of illness made us ask the question, does the innate immune response to influenza predispose children to severe disease? In order to explore this, we developed a mouse model of juvenile influenza infection. And here we used 28-day-old juvenile mice because their lung development is similar to that of a child between the ages of 6 months and 8 years. We compared their mortality in response to influenza. We infected the mice intratracheally. And regardless of how much influenza we gave them, we saw increased mortality in juvenile mice compared to adult mice. Importantly, when we removed the lungs and homogenized them, we did not see a difference in viral titer between the juveniles and adults. So although the juveniles were dying at a much higher rate than adults, this was not due to a difference in viral clearance. And not surprisingly, given the higher rate of mortality in juvenile mice, juvenile mice have much more severe lung injury following influenza infection than adult mice as shown here on histology, with the adult mice on the top and the juvenile mice on the bottom. Despite not seeing a difference in the amount of virus recovered from the lungs in juveniles and adults, we did see an increase in the secretion of type 1 interferons. So when we measured interferon beta in the BAL fluid from juvenile mice, we saw an increased amount of interferon beta throughout in the late stages of infection. So despite a falling viral titer, we saw increasing antiviral interferon. And downstream from that, we saw an increase in the production of the chemokine responsible for recruiting monocytes to the lungs, MCP1. We then measured all the CD45 positive cells in the lungs throughout the course of influenza infection in juvenile and adult mice. And here I'm just showing you the neutrophils, the monocyte-derived alveolar macrophages, and the tissue-resident alveolar macrophages. But none of the other cell types were different between juveniles and adults. In fact, the only cell type that was different over the course of infection in juveniles compared to adults were the monocyte-derived alveolar macrophages. So consistent with the increased secretion of MCP1, the chemokine that attracts monocytes to the lungs, there was an increase in monocyte-derived alveolar macrophages in the lungs of juvenile mice late in infection and despite falling viral titer. The NLRP3 inflammasome is an important inflammatory response that's known to be prominent in influenza and potentially detrimental. So we wanted to compare how activation of the NLRP3 inflammasome differed between juveniles and adults. What we found was that mature caspase-1, so a marker of inflammasome activation in the BAL fluid, was initially up in adult mice compared to juvenile mice on day three post-infection. However, that early inflammasome activation quickly resolved as the adult mice recovered from infection. In contrast, the juvenile mice had a late and increasing activation of the NLRP3 inflammasome. And remember, this is as the virus is being cleared from the lungs. So despite viral clearance, there is ongoing and increasing activation of the inflammasome. Activated caspase-1 cleaves the proforms of both IL-1 beta and IL-18. So then we measured IL-18 in the BAL fluid. And consistent with caspase-1 activation, we saw increasing IL-18 late in infection in the juvenile mice. Using flow cytometry, we then looked to see where the inflammasome activation was most prominent. And we found that it was, in fact, in the recruited monocytes in the juvenile mice that there was increased caspase-1 and increased NLRP3 in the juvenile mice compared to the adult mice and in comparison to the other cell types in the lungs. So not only were there more monocyte-derived alveolar macrophages in the lungs of juvenile compared to adult mice, but these monocyte-derived alveolar macrophages had an increased inflammatory phenotype. So we wanted to know, were these juvenile macrophages intrinsically inflammatory, or was that phenotype promoted by the juvenile lungs? In order to do this, we took an in vitro approach where we infected precision-cut lung slices from either juvenile or adult mice. And then we harvested the supernatant, inactivated the virus, and then put that supernatant onto bone marrow-derived macrophages from either juvenile or adult mice. And looking at MCP1 secretion, we noticed that there was no significant difference in MCP1 secretion from the precision-cut lung slices from either juvenile or adult mice. However, when we put the juvenile supernatant from the precision-cut lung slices onto either adult or juvenile macrophages over here, we saw that there was an increase in MCP1 secretion from the monocyte-derived, from the bone marrow-derived macrophages. And if you used adult supernatant from adult lung slices, you did not have the same increased inflammatory response from either the adult or juvenile macrophages, suggesting that the juvenile lung was in part responsible for this increased inflammatory phenotype seen in the recruited macrophages. Due to the increased numbers of monocytes recruited to the lungs in juvenile mice and the increased activation of the NLRP3 inflammasome in those monocyte-derived alveolar macrophages, we wondered whether or not preventing recruitment of monocytes to the lungs in juvenile influenza infection would improve lung injury and prevent death. In order to do this, we gave an antibody against the receptor for MCP1, the CCR2 receptor, and we gave this anti-CCR2 antibody two days following infection with influenza. And we found that preventing monocyte recruitment improved mortality in juvenile mice by about 50%. And that with that anti-CCR2 antibody, we were very successfully able to prevent recruitment of monocytes to the lungs. Importantly, prevention of monocyte recruitment did not impact viral titer. So when we inhibited monocyte recruitment, juvenile mice were still able to control the virus and clear it from their lungs. Supporting our previous findings that monocyte-derived alveolar macrophages were a primary source of NLRP3 inflammasome activation and IL-18 secretion, prevention of monocyte recruitment during juvenile influenza infection decreased IL-18 in the bronchoalveolar lavage fluid. Somewhat surprising, we also saw that inhibition of monocyte recruitment also decreased type 1 interferon production late in infection. We then wanted to know if we could more specifically just inhibit the NLRP3 inflammasome and have a similar improvement in mortality in juvenile influenza infection as we did with preventing monocyte recruitment. So here we gave a small molecule inhibitor of the NLRP3 inflammasome, MCC950, three days following infection, and we found that we could in fact improve survival in juvenile influenza infection, and that this was associated with a decrease in IL-18 secretion into the BAL fluid. Since we had found more type 1 interferon in the juvenile mice compared to the adult mice late in infection, and that type 1 interferon response was abrogated when monocytes were prevented from entering juvenile lungs, we wondered if that late type 1 interferon response was harmful in juvenile influenza. To test this, we gave an anti-interferon receptor antibody, both either intertracheally or systemically, to juvenile mice four days following infection. When we gave it intertracheally, we found that we did not protect juvenile mice from influenza. In fact, we worsened their mortality. However, when we gave the anti-interferon receptor antibody retro-orbitally to the systemic circulation, we improved survival in juvenile influenza infection. So instead of all of the mice dying, about 60% of them survived. This suggests that giving an antibody against the interferon receptor and blocking type 1 interferon signaling in the lungs is harmful, probably because the epithelial cells rely on interferon signaling to control the virus. However, when you give an antibody against type 1 interferon signaling systemically, that is beneficial, either through systemic responses or through an impact on the cells that are going to be recruited to the lungs, like the monocytes. To test whether a lack of type 1 interferon signaling in recruited monocytes is beneficial in juvenile influenza infection, we first depleted macrophages from wild type mice, and then infected those mice with influenza. Two days following infection, we transferred bone marrow-derived macrophages retro-orbitally to the mice from either wild-type bone marrow-derived macrophages or bone marrow-derived macrophages lacking the interferon alpha receptor. And somewhat surprising to us, we had a dramatic improvement in survival where the mice that received the macrophages lacking the interferon alpha receptor all survived infection, whereas the mice given the wild-type bone marrow-derived macrophages had about a 60% mortality. So to summarize this portion of the talk, I have shown you that juvenile mice are more susceptible to influenza-induced lung injury and death than adult mice, that this is associated with an increase in type 1 interferon production, an increase in MCP1 secretion, and an increase in monocyte recruitment to the lungs of juvenile mice compared to adult mice. I have also showed you that there's increased NLRP3 inflammasome activation in the juvenile mice and that this activation is primarily happening in the monocyte-derived alveolar macrophage. We've also found that juvenile alveolar epithelial cells may promote an excess inflammatory response in recruited macrophages during influenza infection. And finally, late type 1 interferon signaling in recruited monocytes increases mortality in juvenile influenza infection. To assess how our findings in mice might compare to what is happening in children admitted to the ICU with RSV and influenza infections, a few years ago we started collecting human samples from the nose and from endotracheal aspirates to assess the interferon NLRP3 inflammasome response. When SARS-CoV-2 emerged, we also started to collect samples from patients admitted to our hospital with SARS-CoV-2 either found incidentally or as their reason for admission. And we concurrently started collecting samples from adults with SARS-CoV-2 so that we could compare the interferon and inflammasome responses between those two age groups. The way that we collected our samples was for bulk RNA sequencing. So the limits to that, of course, are that multiple cell types are collected at once and sequenced at the same time so you can't distinguish between the epithelial response for instance and an immune cell response. And that specific changes might be obscured in one cell population by another. However, it does have the advantage of not losing cell types through processing. So things like neutrophils that are very sensitive to processing for flow cytometry or single cell RNA sequencing are retained in this approach. And it's practical so that it can be used potentially for clinical prognostication. Like many others have now shown, when we assess the amount of SARS-CoV-2 viral reads in our samples, we found that there was no difference between children and adults. And we also looked at how the receptor for SARS-CoV-2, ACE2, compared between children and adults and did not see a difference. In addition, we looked at how ACE2 expression compared to SARS-CoV-2 reads and we did not see a correlation. So the amount of ACE2 in your nose did not correlate with how much SARS-CoV-2 we recovered. Because an increased interferon response had been reported in children with SARS-CoV-2 and was hypothesized to be protective in children, we looked at how the interferon response in the nose changed with age in our cohort. And we did not, in this group of children and adults with mild SARS-CoV-2 infection, find a correlation between the interferon response and age. In addition, we looked at children with influenza and RSV who were critically ill in the ICU and found that they had a similar or slightly increased interferon response compared to children with mild SARS-CoV-2. Suggesting that an elevated interferon response alone is not a marker of protection from viral infection generally. Instead, we found that the interferon response correlated quite nicely in SARS-CoV-2 and RSV with the amount of virus recovered from the lungs. This was not significant in influenza, which either means that there's no correlation between interferon response and viral reads in that cohort or we just had too few of samples to reach significance. Recently, using single-cell RNA-seq, Yoshida et al. found that children did have an increase in the interferon response signature at baseline in epithelial cells from the nose and immune cells from the blood. However, this increased interferon signature did not persist once in the children that were infected with SARS-CoV-2 in the nose. There was still an increase in type 1 interferon signaling in the monocytes in the blood from children, which we found interesting based on our work with influenza and mice and implicating those circulating monocytes in severe influenza infection. But the impact of this on SARS-CoV-2 is less clear. So while there was an increase in the type 1 interferon signature in healthy children compared to adults and in SARS-CoV-2-infected children compared to adults, at least in the immune cell compartment, when you broke up the SARS-CoV-2-infected participants by asymptomatic or mild disease, there no longer was a difference based on age. Instead, the asymptomatic and mild children and adults had the highest type 1 interferon response with the children and adults with more severe disease having a lower type 1 interferon response. And recently, this paper was just published and perhaps shed some light on these apparently contradictory findings of how interferon is either protective or harmful in viral infection. And what they found was that increased interferon lambda and increased interferon stimulated gene expression in the upper airways was associated with milder SARS-CoV-2 infection. However, an increase in interferon alpha beta and a decrease in interferon stimulated gene expression was found in critical SARS-CoV-2 infection. And that the interferon alpha beta and lambda expression was proportional to viral load in people less than 70 years old, but that relationship was lost in adults over 70. And I was excited to see this because this loss of proportion of type 1 interferon to viral load seems to be what we are seeing in our mice infected with influenza with that late secretion of type 1 interferon out of proportion to viral load being associated with increased lung injury and death in juvenile mice. And similarly, a different response is found in the lower airways where decreased interferon lambda and interferon stimulated gene expression is associated with more severe disease and increased interferon alpha beta and apoptosis is also associated with severe disease. So to summarize, the magnitude of the interferon response and its association with severity of disease is dependent on both the age of the host and the location being studied. Next, we assessed the inflammatory response in children compared to adults with mild SARS-CoV-2 infection. And here I'm just showing the differentially expressed genes from the nose of the participants in our cohort. And what we found were of the 441 upregulated genes in children compared to adults, they enriched for things like macrophage activation, response to TNF, and regulation of epithelial cell regeneration. In contrast, there were 296 genes that were upregulated in adults compared to children and those genes enriched for things involved in neutrophil activation and migration. In line with the macrophage activation we saw in children compared to adults, we saw an increase in IL-18 in children and its binding partner. And in line with the epithelial cell regeneration, we saw an increase in some keratins like keratin 17. In contrast for the adults, we saw an increase in the neutrophil chemo attractant IL-8 and other inflammatory genes related to neutrophils. Because cytotoxic CT cells have been associated with severity of viral infection and have been shown to increase with age, and because some specific literature in SARS-CoV-2 suggesting an inflammatory circuit between T cells and macrophages in the lungs leading to severe disease, we assessed T cell activation in the nose of children and adults. And we found that there was increased T cell receptor signaling in the adults compared to children and enrichment for genes involved in T cell activation. So our findings both in our juvenile mouse model of severe influenza infection and in our human cohort of mild SARS-CoV-2 infection are fitting into the larger body of literature that is growing regarding how children and adults might respond to viral respiratory infections differently. Clearly the timing of the infection and the type of virus matter and that the pediatric response to viral infection is not clearly always beneficial or always harmful as there is increased susceptibility in this population to some viruses like RSV and influenza but not others like SARS-CoV-2. How some of the pediatric responses are beneficial in one virus and harmful in another is not clear. However, we both in our influenza model and in our human samples have seen that there is consistent with this literature an increase in innate immune responses including macrophage activation in children compared to adults and not surprisingly a decrease in adaptive immunity in children compared to adults as seen by our decreased enrichment for T cell signaling in pediatric airways compared to adults. The impact of the local interferon response clearly depends on the timing and location and can be both protective and the right time at the right place in the epithelial cell compartment or the macrophage compartment but harmful at others when it perpetuates systemic inflammation, inflammatory macrophage recruitment and lung injury. The potential impact of our work and the growing body of literature surrounding age-related differences in the host response to viral infection is that the risk of severe viral respiratory infection is unlikely to be due to an inability to control viral replication. This has now been suggested in multiple viral respiratory infections that viral load is not clearly predictive of severity of illness. Our mouse data suggests that both the epithelial response to influenza and the inflammatory phenotype of recruited monocytes contribute to influenza-induced lung injury and that the microenvironment of the juvenile lung may predispose children to influenza-induced lung injury. Our influenza data and our SARS-CoV-2 data suggests that age-related differences in immune response exist but they may be protective or harmful depending on the virus and that the host response to viral infection can perpetuate lung injury and as such it will require precision medicine to modulate carefully in order to improve outcomes in viral respiratory infections in children. Many people have contributed to this work and need acknowledgement, especially Dr. Karen Ridge and members of the Ridge and Coates labs, all of my collaborators and advisors and my funding sources. Thank you for listening and hopefully we'll have a chance for questions.

pediatric

adult responses

immune response

SARS-CoV-2

innate immune system

viral load

precision medicine