Scientist Stories: Anuradha Ray, Asthma & New Insights into its Complex Immune Landscape

All right. Well, thank you, everybody, for attending. I'll keep this short because I know we have a lot of asthma enthusiasts in the building. But it's my pleasure to introduce Anuradha Ray. She's the endowed chair in lung immunology at the Department of Medicine at the University of Pittsburgh. I'm not going to take you through all the great stuff that I was looking at in your CV. I will notice that just this year she got the Scientific Achievement Award at the American Thoracic Society, which seems like a big deal. And although she's got, you know, over 100 publications and almost a dozen ongoing grants at the moment, the— what I really appreciated was almost 2 pages of trainees. So she seems like a person who like really cares about not only the science, but having the trainees train well and do well in their own careers. So with that, I'll just say thanks for coming and we're glad to have you.

Thank you, Sam. It's so wonderful to be here and to see people, young people in the audience who are, you know, escaping the gorgeous weather. We always appreciate coming from the Northeast that, and you know how good it is here. But thank you for the invitation, Sam, and to the community as well. As the title is shown here, I'm going to actually take you through our current understanding of asthma as a whole, as a disease, especially severe asthma. And I'll tell you why it's called severe asthma and why it's so important for us and I think for the community also to know more about. Although it's an old disease, but we still are learning from the new tools that we are beginning to use to understand the disease. So with that, I don't have any disclosures except to say that we have an agreement with PRIS Pharmaceuticals to do some work with them. Okay. So, the outline is as follows. I have one slide which really depicts a little history of asthma. And it'll be fun to talk about a few things there. And why is it that we are still studying asthma when I mentioned that it's such an old disease? Everyone knows the term asthma.. But it's a very complex heterogeneous disease which is recalcitrant to therapy in some. Although corticosteroids, they are still the cornerstone of asthma therapy. And they manage disease fairly well in compliant patients in mild, but not many severe asthmatics, even when the biomarkers suggest a type 2 immune profile. And I'll touch on this type 2 Th2 immune profile a few times because that's how immunologists were introduced to study the disease. But I I think it's too simplistic, and I'll tell you why. However, it is kind of— and we don't understand this, I've just raised it as a question— is that I told you that a subset of severe asthmatics poorly respond to corticosteroids even when given in extremely high doses, both inhaled and orally or intravenously. But yet they do respond to T2-directed biologics. So that creates a conundrum. And we don't completely— we have hunches, we have hypotheses, but we We don't know exactly why. And what is the basis of this heterogeneity then? Because if it is heterogeneous, you cannot have a kind of one-size-fits-all treatment. So we have some new insights using a high-dimensional multi-omics approach and machine learning, but it is an evolving story. And very recently, I think we have landed on a potential novel target in severe asthma. That we believe bridges different types of immune responses, especially Type 1 and Type 2 that can go hand in hand in this disease and can be present in the most severe patients. What do we know about this T1 dominance? And we are trying to understand more what causes this very peculiar— that's seen in autoimmune disease and we are finding in severe asthma. And what is the consequence of this? Does it cause tissue senescence? Not only senescence of the cells themselves, a very pro-inflammatory phenotype, but that might actually relate to inflammaging and impact not just asthma but many other disease types. So, many of us who have been studying asthma for the longest time may not be aware that it was almost more than 300 years ago that an Englishman, Sir John Floyer, first introduced asthma in the English language in A Treatise of the Asthma. And this is apparently the only picture of him that you can find anywhere in online especially. And then in 1941, Dr. Francis Rackermann, he introduced— he split asthma into two kinds that many people were aware of. And, you know, you should not just take away from this that asthma was first understood in 1698. It has been known for thousands of years and has been treated in many different ways. But Rackermann introduced the terms extrinsic or allergic asthma versus intrinsic of unknown etiology. And he has been considered the greatest expert in studies of allergies and asthma in the 20th century. So, it's an ancient disease and it affects almost 5% to 10% of the population in developed countries with an estimated US burden of almost 25 million cases. And it's increasing. But about 5% to 10%, and I think it's more like towards 10% these days, of all asthma is classified as severe and is poorly managed by corticosteroids. And that's where one of our colleagues that I have been working with and I have had the good fortune of working with Dr. Sally Wenzel. She was one of the few initial investigators in this field who were treating these patients, realized that this is a very different type of asthma and needed to be treated differently. All they could do initially was to add more steroids and bronchodilators, of course, along with it. But that does not help. It creates more side effects. And because of frequent healthcare utilization and also the types of medications they use, it inflicts a disproportionate use on the healthcare cost for all asthma. Although it's just 10%, which amounts to close to 20, then half of it, $28 billion to manage the disease in this country. And you can appreciate in this table how the cost per person for mild versus severe asthma is very, very different. So this is nothing new to many people like Vijay, my good friend who's in the audience, Dr. Poirier, who is also has been studying asthma for many years, and many of you others in the audience. We began our investigations into asthma not too long after this was associated with Th2-like cells. And this too, Th2 cells and asthma were linked together in this very famous New England Journal Robinson Barry Case paper. Not too long after Th1, Th2, this concept was introduced in the mid-1980s. And a few years down the road, we then linked GATA3, the transcription factor, to be an essential regulator of IL-5, which is expressed by only Th2 cells, not by Th1 or Th17 or other cell types. And then our lab and Richard Flavelle's lab identified GATA3 as the master regulator of Th2 cells. Many people don't like the word master regulator, but I would— I still use it because that's how GATA3 works. And then, of course, subsequently, TBET was linked to Th1 and RORγT to Th17. And then we linked GATA3 GATA3 to human asthma in collaborative work. We also made a mouse with a dominant negative GATA3 and showed how important this transcription factor is in inducing allergic inflammation in a mouse model. Subsequently, other cell types have been described, like ILC2s, which are also studied at this institution, that express GATA3, and then they also produce type 2 cytokines. So, mast cells, eosinophils, they all express— whichever cell types thus far has been shown to produce type 2 cytokines expresses the transcription factor GATA3. So very gratifying for us that a few years down the road in 2015, our German colleagues, they published this paper. They targeted GATA3 using a DNA enzyme. And in allergic— allergen-induced asthma, they found that if you subdued GATA3, then you could relieve the symptoms. However, life is not so simple. We love the simple things in life, but unfortunately asthma is not so simple as a disease to treat. Why is it? And as I said, it's very complex and it covers the whole spectrum and the symptoms are almost the same. It can be more frequent. It can be many, you know, worse in severity-wise in terms of decline in lung function, everything else, wheezing, chest tightness, shortness of breath, is reversible airway obstruction. But it can cover the whole gamut from type 2 high to type 2 low. And see the transition in use of the word Th2 to type 2. Because as I said, many cell types express these cytokines. So, Th2 is not so much used anymore in the asthma literature. It's more type 2. And it— atopy or allergy-wise, it also covers the whole range. Even in mild asthma, that's important to understand that many people people don't quite appreciate. The age of onset can be different, early to late. The cells that come in and infiltrate the lung can be eosinophils, not just eosinophils, but also neutrophils or a mix of them. And then in mild disease, thankfully, it is still corticosteroid responsive. It's inhaled steroids. And thanks to those inhaled versions of steroids that they don't dissipate out into the periphery, And but however, even in mild disease, only 50% or so are allergic. It could be due to other reasons. And then it's no longer— many people are not treated with steroids or don't need steroids. But at least the disease is mild and it can be controlled. The other response I just mentioned, interferon gamma, that I'll get— this is a story that really I'll be talking about— is really present in a subset of these severe asthma patients. And again, severe asthma by definition is poorly responsive to corticosteroids, whether inhaled or oral. Just this morning, we were talking about genetics of many diseases that many of you know Vijay's group has been working on and published so nicely over the years. And the locus that has been most commonly and reproducibly by many different groups associated with asthma is the 17q12.21 locus. And within this locus, there are many genes. Very well-known ones are this ORMDL3, gastronomin A and B. But on top of genetics is really the etiology of the disease lies in gene-environment interactions that many, all of you know. And exposures, the interactions are with allergens, but also with infectious agents such as bacteria, viruses, and also fungus. And many other covariates are there, many other factors like obesity also impinge on the disease. But ultimately, you end up with what? You end up with immune memory. And while many people in the audience at this institution study good memory where you need in vaccinations, in response to infections, you want the memory to be recalled to really clear the pathogen. Not so in asthma or in autoimmune diseases where this memory, if rekindled, it can cause aggravation and exacerbation, as we say. And the infectious agents are mostly the culprits are these viruses. Rhinovirus, and sometimes even RSV in the elderly, other viruses too. But ultimately, you can end up with stable lung function which resolves, but it can resolve and recur again in many patients. But then in many others, there is lung function decline. It's not that mild progresses to severe. There's debate there. People start off as mostly as severe with severe disease, and then it can get more severe. It can persist or then further worsen with age. And many things impinge on this that I don't have time to get into. But many, many reviews have been written on the topic. In 2023, the asthma umbrella— and I've adopted this from Sally Wenzel, the way she talks about it as umbrella. She wrote a very nice review in Nature Medicine. We have three small parasols, if you will. The type 2 classic inflammation, which is more Th2 in mild disease where it occurs. And you have the reproducible biomarkers that are your eosinophils, exhale nitric oxide. Then you also have low or no type 2 inflammation. And that has been published by the UCSF group, where you absolutely don't see any of those signatures. You see some IL-6, and obesity has been linked to that. But then what we have been studying along with the type 2 is this mixed type 1, type 2, where there is interferon gamma-dominated airway inflammation. But behind also, if you just go by biomarkers, you will see those same biomarkers coming up in this group as well. And people People have been misled for the longest time. In fact, when we were trying to publish these studies, many people thought this is wrong. I mean, you— this is either from drug treatment or from other causes. But I'm very happy that independently now every center, every— and there's a major manuscript now being written by the whole SART group where this interferon gamma signature is very much there. But behind the curtain of T2 lies this complexity that we have been trying to understand, that it is not just— if went by the biomarkers, we will be interpreting things in a very wrong way and incorrect way rather. So academia and industry, both sides, I would say that both sides have been fixated on Th2-mediated disease or type 2-mediated disease. All the drugs that you see on this side, and many of you must have seen this commercial, you can do more with asthma with Dupixent. Dupixent, which is the alpha receptor, alpha-targeted biologic. And in fact, Sally Wenzel ran the first trial on Dupixent and it was successful. However, subsequently, she and others who treat patients with dupilumab are realizing that many patients, when you go by the 300 eosinophils per microliter above that or with the XLNO cut off and you treat the patients, you come out with worse disease. And this has been puzzling. I said, really, Sally? But didn't we say this before, that if you use DUPI in this group with the T2/T1 mix, then— and this is the immunology concept really, having worked in this field for many years— that if you treat that group, then it should not do well. And I had told her many, many times, even we wrote in our grant applications. So there is really nothing more. While it is is now being recognized that you have these other types of inflammation, people don't know what to do with it. And in fact, we do need to do more, and industry is coming around. I'm happy to say that. So this is a fun article that we wrote. And I was just mentioning to Sam this morning that Cell Reports Medicine invited us to write an article on the whole issue. If you are interested, it is in January on artificial intelligence in medicine. In. And we said, okay, we'll have probably a spot. They said no guarantees, we'll review everything, many diseases we have invited, and, you know, we'll see. So we wrote this in collaborating with Sally, of course, and Jishnu Das, who Sam also knows. He's a very talented systems immunology person, computational biologist. And the whole idea was that all of this is known. I have already told you that there is immune dysregulation, there are altered biological pathways, but all the groups all over the world, there's whole European consortium too, using multiomics approaches, mainly transcriptomics thus far, but beginning— have some have used— it's all detailed here, where the main studies that have gone into understanding asthma. Ultimately, we still are very poor in our knowledge of biomarkers. We have to link new biomarkers to these pathways such that the word endotype that's so loosely used to describe any disease, endotypes truly are where where a specific pathway, molecular pathway, that through multi-omic studies or other studies you have identified can be targeted in vivo in a patient. And you see that there is relief of symptoms, the primary endpoints are borne out, and then it's defined endotype. So that's where we need to be. And for some of the pathways, I mean, even for the alpha receptor, alpha or the alpha pathways, At least Sally refuses to call it an endotype for very different reasons, and we can have a discussion about that later. So, we began into around this time period a little before 2015 when we published this, a very focused study with the— let me see. Yeah. So, we established with all of this in the background around this 2015 period, we proposed to the— to NIAID that could we use the knowledge of SARP, Severe Asthma Research Program, that NHLBI has been running for so many years with 7 sites, and could we borrow some samples to write a program project? So NHLBI and NIH, there's a little bit of competition with the asthma thing because both of the institutes fund asthma. But graciously, because of Sally, I would say, and because maybe we had done some good immune work in the past, that NHLBI agreed to part with some samples so that we could use to generate preliminary data to write the program project. Sally's expertise is in epithelial dysfunction in asthma. Ours is on the immune side of things. And we said, look, they crosstalk, and I'll show you a beautiful example of how they crosstalk continuously. So let us— you study the epithelial, we study— we'll go back and forth between human samples as well as mouse models, some new ones that we have to establish because the ones that exist are not adequate. But again, we'll use the rich resource of SARP, which is the Severe Asthma Research Program, their microarray data and other things. But we will also establish our own cohort, Immune Mechanisms in Severe Asthma, or IMSA. And we were very happy to get funded. It was just renewed last year too. So initially around that time point when we were beginning to establish the program, we ran a very focused study because we didn't know anything how to go about understanding this very complex disease. We said we know T cells. Let's look at the T cells from the airways by lavage. And I was telling Sam how when some of these patients are lavaged, you have to understand, and not many centers are allowed to do the type of bronchoscopy that our IRB allows Sally to do because the lungs can collapse. And the lungs do collapse. She said this patient's lung was, and they still consent because they have so much faith in Dr. Wenzel. And I'm very, very happy about that. So with these kinds of precious samples, I told Sally, let's look at the overall immune response by ICS and also by the cytokines they— with a brief stimulation, what they produce. And we found that both by ICS as well as by measuring the secreted cytokine, that there's a group that expresses— produces a lot of interferon gamma. And we thought it will be all IL-17. Believe me, I've been working with Jay Coles, also collaborating on studying severe asthma using mouse models. And the world thought that that the problem of severe asthma is solved with CH17 pathway. So, we didn't see that signature very clearly. And many people said, well, you know, you didn't measure many. And we agreed. This was, you know, some 30 or 40 samples. And maybe we need more. We need to go across the centers. But these were all from different centers. But at the same time, all the industry was running trials on not only IL-17, but IL-17 receptor alpha, and they all failed. So this then became the— people were puzzled that IL-17 brings in neutrophils. Neutrophils are so bad. How come that the trials have failed? And we said, well, interferon gamma can also bring in neutrophils, and many people don't know. We have written a review on that as well. So what then we found that— yeah, yeah. Yeah. Right here.

Yes. Yeah. Thanks. What— so what are you—

your ex vivo— you're stimulating here with? We don't have these specific antigens. What we do is a cocktail of anti-CD3, CD20, and CD2 microbeads. But it's a very brief stimulation just to have the cytokines being— you know how T cells, unless you stimulate them ex vivo, you'll never get a signal.

Sure. Sure.

Yeah. So have you ever tried to stimulate with like house dust mite or—

Or something like that. With what? Oh, house dust mite? No, for the very reason because many of them are not allergic. So it's not that they are all, as I said in the beginning, just like in mild asthma, many of them are not really allergic to anything. It's more like an intrinsic disease of unknown. It could be an autoimmune and an autoantigen, believe me, because of things that we are learning about. Many of these patients and some of these things will be written up by fellows working with Sally have especially lupus and rheumatoid arthritis. So we really don't know. And yet they have the severe asthma phenotype. So all we do is any T cell that was— and this is for really the— it's a great question that you ask for the trainees too, that T cells, no matter how hard you try and try to get them to produce cytokines without any stimulation ex vivo, refuse to produce the cytokine, express the protein unless briefly stimulated. And sometimes they do. Sometimes some BAL cells we have found, but it's complete. And this is kind of the dirty secret, just like co-stimulation is a dirty secret. But the most important thing is you don't cause de novo differentiation of any cell, that you keep it long enough that it would cause that. So, a brief stimulation just to reawaken the open chromatin, that is a stimulate to egg the T cells to produce the cytokine. And again, you know, if you have questions, please, you know, I don't have a zillion slides. I can always be interrupted. And so one important point was that in the previous slide, all these patients by definition I've told you, severe asthmatics cannot be taken off steroids. I mean, it's not allowed, not in this country. So they are on high doses of steroids. But the mild asthmatics— and a reviewer asked a very astute question that some of these mild asthmatics probably are not on steroids. And we had already written that in our methods. And yes, some were not. Then say, how is their cytokine? And the ones that were treated with steroids, are they responding in your hands to steroids? And that, that was such a nice question. I love that. And indeed, those who were on steroids, their cytokine levels were actually down. So, this really creates the biggest, all the time, difference between mild and severe asthma, that steroids just don't work for many, for people. And that's the definition of severe asthma. Asthma, that they have to be at this level of steroids and they have so many exacerbations. So that's how they are categorized. And indeed, in our hands too, the milds do respond to steroids. So then we said, okay, if we have this interferon gamma response, how do we model this? And we had no model where we needed the model to not only express— you can treat with LPS and HTM, many people have that model with house dust and get a type 1 immune response. But each time we tried an adjuvant, so to speak, it could be suppressed pretty well by steroids. So it didn't quite fit that description that it would be only partially responsive to steroids until after 6 months of failure, a grad student was going to really give up. And we said, well, one thing, this molecule cyclic di-GMP or di-AMP is produced by bacteria as innately by the bacteria themselves. And it can induce both type 1 and type 2 interferons. Type 1 interferon, of course, induced by viruses associated with asthma. But bacteria, I mean, there are papers in Science, Nature that you can dig up, very good papers where many types of bacteria, especially 4 kinds, have been associated with asthma, severe asthma. And they induce type 2 interferons, of course. So, this cycle cyclic dinucleotide compound could perhaps be combined with— we do need an antigen here to, again, reactivate the T cells. And so we said, okay, HTM, and combined that with cyclic digMP without steroids. And the long and short of that was that this gave us a model. It's a model. It's not— a model is always a model. At least it gave us a phenotype where we had had a dominant Type 1 immune response, some Type 17, but also Type 2, all 3 that could be partially suppressed by steroids. But importantly, when we took interferon gamma knockout mice, then the whole airway hyperreactivity, a cardinal feature of asthma, collapsed. And but it did not— luckily, because Jay was part of the team, he had an IL-17 receptor alpha knockout on a BALC background, and we used that mouse and nothing happened. So this again told us that IL-17, not only that we could not measure IL-17 in many of these samples, but also in our model, when we knocked out that pathway, it did not impact really the airway hyperreactivity. Then Bruce Levy's group published this paper as was in 2015, in 2017, and showed they didn't quite do the intracellular cytokine. They went by chemokine receptors, assigned this group of chemokine except for CCR4 they used and CCR6 for Th1 versus Th2 and showed that it was the Th1 cells in these human, again, asthma that showed negative correlation with lung function and not the Th17 cells. Then subsequently, groups who studied pediatric severe asthma also published two papers showing association of interferon gamma with very severe steroid-resistant asthma. So we did a lot more work and we published some other papers where we went upstream to understand what really triggers from the dendritic cell side of things. The other day, Miriam Merad was speaking to us just last week. And she told us that with all the immunotherapies they are doing in cancer, I just love that, that at the end of the day, she's finding in situ with every technique they are employing, that it's a dendritic cell which is controlling wave feel this whole PD-1 response, where it works, non-responders versus responders. So dendritic cells, as you all know, play a very huge role not only in initiating T cell differentiation, but also maintaining them in many different ways. And we found that this transcription factor IRF-5 triggers IL-12 from the dendritic cells to induce Type 1 differentiation with human samples and our model. And then also we found downstream, and this is where the story has become very interesting to us, that the Interleukin gamma, as you know, through the JAK-STAT pathway— Sam studies that— can induce many chemokines like CXCL9, 10, 11, all belong to a family. And they work on their cognate receptor CXCR3. That then causes production of these chemokines like CXCL10 that then brings more Th1 cells. So it creates this vicious loop. And for many years, in my early years, we have been studying— I've been interested in how steroids steroids exert their anti-inflammatory functions. And we identified for the first time NF-kappaB as a target, as one of the central mechanisms by which the glucocorticoid receptor really shuts off inflammation for IL-6 and many other genes. But then many— we found here it's the opposite. We in fact used NF-kappaB LPS as a control in this work and found that if anything, the steroids are actually allowing STAT1 downstream of gamma to work even better. So it was cooperating to induce more CXCL10. And I said to Sally that we have to now— and she said, well, we know that, but what do we have on our other than steroids? That steroids are causing more harm than good in this group of patients looking at the data. So this is the type of thing we identified that it's a whole upstream downstream, this intimate pathways that then gets induced that cannot be covered by steroids.

Corticosteroids.

And then we published this paper last year where we have seen that interferon gamma, and we can discuss this more, is also found in many mild asthma patients. But the prevalence is more in severe asthma. And it's not just type 1. And I'll get to that in the last part of the talk, which is the most interesting concept I think that is beginning to evolve. But also there is increased oral corticosteroid use in these interferon gamma high patients. More in the severe asthma group— I mean, more in the interferon gamma high group than in the low group. So you have two things. You have more interferon gamma high in severe asthma, but you also have within that group more oral corticosteroid use. So at this point, another fellow had joined, Matt Camiolo, and he had some bioinformatic background. And we were invited by the Stanford group, Kari Nadeau's group, that Anuradha, you're studying all of these. Can we collaborate? These are precious samples. Because we know how hard it is to get these samples, to use site-of that we have and then study together. I say absolutely. So as I said that the classification of severe asthma is really a guideline-based diagnosis related to medication use and has no— you don't get any idea about the underlying immune and molecular dysfunction. Patients don't respond. You have to give them more steroids. They have more exacerbations. All these features, their clinical features are taken into account to give them the label of severe asthma. You don't get any other clue from that. So, we needed a different approach to really understand the immune milieu, needed an unsupervised method. And so, then Matt said, well, we'll use— great, we can use mass cytometry to analyze these lavage cells, which apparently had not been done before. And then we can integrate the data with RNA-seq data. We didn't quite have the funds to do single-cell RNA-seq at that time. But now, we used our own cohort. We were not dependent on the SARP cohort because the SARP cohort didn't have the BAL cells anymore. We needed the BAL cells to be preserved in very precise ways so that they could be revived with almost 95% revival survival after being frozen. So we collected all the samples. This time we used IMSA. And using 33 cell surface markers, we then employed mass cytometry and then, of course, the flow SOM to cluster those cells. And we identified 25 clusters overall here. And when we, based on the BAL cell count, when we clustered the patients, we ended up with 3 clusters. All the healthy controls fell into group 1, but the severe asthmatics fell into 2 and 3. And the miles were distributed between 2 and 3, more in 3 than in 2. But what was remarkable, what we had seen before is is that PG2 and PG3, while they clustered so differently here, you can see, but if you went by their blood eosinophils and their excelled, I know they were indistinguishable. So I have actually the data that I'll show you. We published that too. So see, if you had just gone by that, you would treat everyone with dupilumab. And Sally has been doing that because we have no other way to differentiate. What we noticed is, and we also, Matt did a very fine job in associating these cell populations lung function. And what we found is that some of these cells with 16, 22, or 15, and then, you know, some of the others 13, they all expressed CCR5, this chemokine receptor. What was so remarkable about CCR5, but then when you know about these cell types, you know why it's so important to be aware of expression of CCR5 on these cell types. So Why? And I'll come to that. And that's really where I'll finish off the work. In this study, we also showed that cytokine-wise, they had very distinct signatures. As we had found in SARP, we found that PG3 was the T-cell-dominated group. This group expressed a lot of interferon gamma, but also type 2 cytokines. And the healthy controls expressed more IL-10 coming from macrophages. And the other The other group, very interestingly, we have identified all these Fc epsilon R1 alpha. We, for the longest time, we thought these are mast cells. But these are really not mast cells, especially a cell type here that the reviewers argued with us, and we showed the mast cells all aligned as before we had shown with the T cell group. This is a very different cell type that produces a lot of IL-4. And we really don't understand. We have a mouse now to understand with the knock-in Cre that we have created in this locus. And we are trying to understand. But they express IL-4, they express IL-5. It is my belief that it is this group that responds best to steroids, to dupilumab, because interferon gamma is very low. So we then went on to associate, because we didn't do single-cell RNA-seq, what are the gene modules, what is the genetic makeup of these? And he established a new algorithm called ICV, ICELITE. Yeah, ICELITE is actually a Pittsburgh beer, and which is the acronym for Immune Cell Linkage Through Expiratory Matrices. And that allowed us to deconvolute the bulk RNA-seq data that many of you do all the time, to associate gene modules with these specific 25 immune cells. And we got about 25 to 26 such gene modules. And this algorithm was also compared to WG-CNA that many of you also know. It's very well established, and we got very similar report back from WG-CNA. So what were the distinct gene signatures? What we found very briefly is that with the innate group, innate-rich group, we found Fc epsilon R, Fc gamma R, and IL signaling to be these gene modules going with that patient group. But adaptive immune response, T-cell co-stimulation, interferon gamma response, all of these are here. This is all published. You can see. And very interesting molecule molecules that some other groups like Vijay Kucheru's group and others linking SLAM F1, SLAM molecules, the DPP molecules associating with this high interferon gamma, which was very gratifying, you know, nice to see that. Okay, but these days you cannot just show in one cohort. What if it is that Pittsburgh cohort that has this immune signature? So we had to validate in a different cohort, but we were a little stuck. So the only other cohort, of course, is the SARP cohort accessible to us. They had their, thankfully, dial microarray data, not only their epithelial, but they had no immune profile from this cohort with the SART cohort. So what Matt did is— we discussed this— he first— a gene scoring method was applied to IMSA RNA-seq data to train a model that yielded the same PG1 to PG3 classification that was obtained by clustering CyTOF data. So he used the— it was a laborious process back back and forth where he had to train the RNA-seq data to get that exact site of classification. That was then applied totally unsupervised to the SAR-BAL microarray data. Remember, this is coming from 7 different asthma sites, and we got the exact same classification. This the reviewers really appreciated, and we appreciated it a lot. Of course, with a little variation, we have some mild actually 1 or 2 severe asthma in Group 1, which we didn't have in our EMSA cohort. But by and large, and most importantly, when we correlated the lymphocytes, we had the manual differential counts, we found that it associated with Group 3. And also the gene modules could be associated with personal lymphocytes, the modules that were linked to T cells in PG3, the same modules linked to lymphocytes in the SARP cohort. So all of this showed us that this was not just applicable to not only to our cohort, but also to the rest of the asthma samples that we had access to. But I would like to also mention and really emphasize the point that why initially we didn't also understand that what is the meaning, biological significance of this mixed T1/T2 phenotype. Does the T2 phenotype, which is giving similar eosinophils, exhaled nitric oxide as the pure T2 more or less, does that matter? And what is impact on disease? So T2 cytokines can be made by many cells. And we found that in the T cells— this is in our supplemental data— that not only gamma, which is very bright red, but they made some type 2 cytokines also. But most importantly, there were other cell types in this group, like mast cells, a very strong signature. They also produced T2 cytokines. And an earlier study by Brian Woodruff Alina, who worked with Sally, had shown that just from my transcriptomic data that the most severe patients showed a combined signature of type 1, type 2. Initially, we thought it was just high levels of gamma. He said, no, actually high levels of gamma, many mild asthmatics also express, those who don't have that obvious and atopic signature. But that alone is not important. T2 is always important. He always maintained that. And you can appreciate how in these two groups, if you look at eosinophils in the blood or the exhaled NO parts per billion, you do not really see that big a difference at all. So, then what we have to grapple with this T1/T2 mixed information. T1 and T2 cross-regulate each other. This is one of the basic tenets of immunology. And we see that all the time. In fact, I was at Yale for 10 years before moving to Pittsburgh. And that's where we had identified GATA3 and all that. And my colleague, Lauren Cohn, who trained with Kim Bodrumli, she published many papers. And in fact, interferon gamma was used by Bob Kaufman in a different setting to treat asthma in general, not severe asthma, because gamma cross-regulates and dampens type 2. I would maintain that's still true in mild asthma, that if you have some interferon gamma response you can induce, then yes, that would actually regulate the disease, but not so in severe disease. And it's this co-presence of this has been now being noticed by us, by other groups. I told you about the already the published work. But I will also say that while T2 direct targeted biologics are highly effective in some, and these may be the PG2 with the more innate cells producing just type 2 cytokines, some of these T2-high asthma patients actually show worse response. And they have to be taken off midstream, the biologic. And they continue to struggle with exacerbations.. And this may be due to a heightened gamma response. In fact, we should publish ourselves, but this was published in 2020. And you can see that side effects of T2-directed biologic, immune deviation to Type 1. So, this group showed antagonism of alpha receptor. Alpha promotes Th1 signaling among T cells from patients with atopic dermatitis after stimulation. And then dupilumab-induced peripheral neuropathy in a patient with severe asthma. So— and other studies are also Whoever treats with Dupuytren knows that this comes up all the time. So in addition to what else then governs this T1/T2 mixed phenotype? In addition to this group of chemokines, CXCL9, 10, 11 working through CXCR3, CCL5 or RANTES can signal through CCR5, this very same chemokine receptor that was present on all those immune cell types that associated with negative lung function in our published study. And that has been shown to promote also Th1 chemotaxis. CCL5, CCR5 also promotes eosinophil chemotaxis. But the eosinophils can produce this chemokine in turn that can further recruit interferon gamma-positive T cells. So this became very relevant to us very recently, actually, not that long ago, because we have been struggling to understand the mixed phenotype. We went back to our SARP-BAL microarray data, and we saw— and Mark Gauthier, another— he's assistant professor now, has taken a lead with Sagar Kale, who's a very kind of experienced postdoc and now going to be a staff scientist. And they worked together to— and they found a strong correlation between CXCR3 and CCR5 in the BAL from the patients. And early studies have identified these dual CXCR3 CCR3-positive, CCR5-positive cells at tissue sites of T1 inflammation, mainly studying autoimmune diseases. So, we went back to a mouse model. And remember those immune cell types that express CCR5? And we have identified these TRM cells, both CD4 and CD8 expressing CCR5, but also CXCR3. So, they are dual expressors as well, 80% or 75%. And then we have also identified this very peculiar CD161-positive CCR5-positive cells. If some of you study autoimmune disease, this cell type has been noticed in colitis by other people. The reason we came across this or identified this is because of literature in other diseases where it's not that frequent. It's not that abundant in the population. And yet, it is extremely resistant to corticosteroids described in the literature, extremely resistant to immune suppression by Tregs. And they can become— they are are considered like ex-TH17 cells. You know how many people have worked on these peculiar cells that extinguish TH17 to then produce gamma. And these are considered— and some even are dual producers and can be very pathogenic. So, these are those kinds of cells. And we have identified in our model too, but we haven't really studied in any detail. But then Mark came up with an idea, and that's why it's so great to have worked with physicians who are aware of these drugs. He said, look, we can maybe repurpose this drug which targets CCR5, maraviroc, which has been used to— it was used, no longer used to treat HIV patients, and see whether if CCR5 is so important, go back to the model and see whether it does anything to airway hyperreactivity. And it did. Just targeting, although CXCR3, we have a lot more detail. We have done the knockouts and the individuals and all that although you have CXCR3 and CCR5 dual producers, but if you just block CCR5, it was able to inhibit airway hyperreactivity and also it dampened inflammation. So, this was really exciting for us. So, now— Is it a small molecule or an antibody? It's actually a small molecule. Yeah. And so, and it can be given and it's very easy to use, given in drinking water, which we did. So then we said, well, we have tied the chemokine receptor. How about its ligand, CCL5? And this is the latest paper we have submitted— I mean published, partly published, partly not submitted yet. So then we looked at in SARP3 now. SARP has entered its SARP4 phase, back to Bell. They go through epithelial, to BAL-SARP to sputum SARP. And SARP3 is working with all sputum from the patients. Of course, Sally is— we are one of the sites, so she has access to all the samples and the data. So SARP3 very graciously— this study is being led by UCSF, the whole group. We were surprised they let us use their data to publish this paper because they have similar data too, but they are writing a pretty big manuscript right now with some of us. What we found is that while high CCL5 expression did not discriminate if you just look at median expression between healthy, mild, moderate, or severe, if you go a standard deviation above the healthy control median right here, then you begin to identify some asthmatics who have high CCL5 expression. And this CCL5 low and high, if you look at them, then the high group associated associates with sputum lymphocyte numbers. So there is— it suggests that the T cells, which we signature we saw in our paper we published in Cell Reports as well, that the T cells produce CCL5, they have a signature, appears to be one source that at least appears to be the T lymphocytes themselves. And what we have shown is that using both cohorts now, that it is this high group— CCL5 low is blue and red is high— then that associates with lower lung lung function. And with the T1 high group of patients in the IMSA cohort, remember PG3 is all T-cell response in our own cohort, and this is where CCL5 expression kind of lands. And therefore, again, suggesting CCL T-cells to be very intimately involved here. So then we again went back to this compound maraviroc and said, what happens to TRMs and the cytokines they secrete? We didn't find any anything number-wise that the TRMs are perturbed or in any decrease. But we found that gamma and 17 are actually decreased from the cells. So, the cytokine production appears to be decreased in response to this CCR5 antagonist. And so, CCR5 signaling then in some way does control interferon gamma and IL-17 expression in the cells that we don't know yet how. And seeing is believing. This is one patient. We are doing it more. And you can now appreciate how these CD8-positive CD8 positive, granzyme B positive T cells are nestled in the airways of the severe asthma patient. This is a very well-known severe asthma patient to us who has very severe disease. And we have studied his cells for a long time. And we see how compared to healthy controls or even mild asthmatics, we have a ton of the CD8 T cells embedding the epithelium. Then why would less— The last 2 or 3 slides I have, why would high interferon gamma be pathogenic in severe asthma, you might ask? And of course, we were asked this question when we first published the first paper. What we showed is bioinformatically first that interferon gamma, we took AHR as an outcome measure. And we showed that using that as a measure, we associated interferon gamma inversely with the expression of these serine protease inhibitor Slippy, which is abundantly expressed in the airways by airway epithelial cells. And we showed how it correlated, epithelial Slippy, with interferon gamma in an inverse way. And what is this molecule Slippy? So, this is a tryptase and chymase— I mean, Slippy inhibits serine proteases made by mast cells, which express tryptase and chymase. Now, tryptase promotes AHR David Corey's group and others have shown how proteases are involved. He believes it's all protease-driven. And I believe too, to a large degree, where it's intrinsic asthma, I also am of the belief now that proteases play a big role. That tryptase promotes AHR by activating protease-activated receptor 2. Nice papers have been published. SLPPY also inhibits neutrophil elastase. So, it's really an important molecule to have around in the lung, in the airways. And low SLPPY activates TGF-beta. And of course, the TGF-beta, you know, McCroft's work, others here, TGF-beta is implicated in airway remodeling. So, if you have sleepy levels going down, that's not good. But also, another reason. And this is where we have found that in the literature, interferon gamma is kind of lost in the literature, can stabilize eosinophils. So, interferon gamma in a very peculiar way, another way has been shown to— eosinophils secrete CCL5. That has been known. T cells secrete CCL5. But eosinophils do also in response to interferon gamma. And that then creates this loop between eosinophils and T cells to bring in more T cells. But then this interferon gamma can promote the survival of eosinophils via JAK2 activations. So you see how they go hand in hand. Now, differences in sputum— now we went back. This is not published yet. Published yet, when we went back to the sputum cell differentials and we— between the mild and the severe, within the CCL5 high patients, we only see a difference. Remember that we also have some CCL5 high in mild to moderates. But the only cell type where we find a difference or something really standing out, and Mark is very excited, is in the numbers of eosinophils. So there is this collaboration. Why is T1/T2 important? This linkage importance. So, we think we have a target beyond T2 that links T1 and T2 is because if you have just high gamma and you don't have the T2 signature, you don't have basically the eosinophils. Interferon gamma might be causing too much of— practically, this is a highly inflammatory molecule. You don't want it around because interferon gamma has many other effects that we can get into. Of course, in autoimmune disease, in other context. And I told you already that many of these asthmatics also have features of autoimmune disease. They have mixed disease. These are really very sick patients. But the T1 inflammation molecularly through cooperation between STAT1GR, there could be other mechanisms. I mean, there are epigenetic mechanisms that we are interested in. And maybe we could do some studies together that we discussed this morning. And to be— T1 inflammation is very refractory to steroid therapy in these patients. But this linkage that CCL5, CCR5 might be providing, bridging things like T cells and eosinophils, might be something that we need to look at in more detail. So the company that makes maraviroc, I always get confused. I mean, they are very interested in running a clinical trial now, at least the phase 1. And we are hopefully, with participation of SARP because we won't have enough patients, we will be able to do something there. So the last one is that Rich Ramanell, who has joined our group too, very talented guy. And we are trying to understand through his own Parker B grant that he just received. We are ecstatic about that. It is a very nice lung award— I mean, award in lung research. And then my own R35 grant Grant to understand that how do you end up with this persistent high interferon gamma expression, and what is the consequence of that? We are interested in mitochondria because mitochondrial dysfunction has been associated with increased Th1 interferon gamma response. This group studied in TRMs, in fact, in CD4 T cells. And I don't know if it was TRMs, but in CD4 T cells, and associated with T cell and also tissue senescence. Supposedly, the interferon gamma, and not just this group, we had a nice NHLBI workshop where many people participated in talking about mitochondrial dysfunction in lung disease, can cause surrounding tissue senescence. That is something that we are very much interested in. And then not only do we have this then not replicative senescence that many people are studying in many different contexts and tying up with cancer, but premature stress-induced senescence in some of these cells. Cells causing inflammaging in some of the patients who are not age-wise that advanced in years, and yet their tissue might be completely destroyed, looking like really, you know, a lung from an older person. So, in these PG3 patients, we had already noticed when we published the paper, we have signs of oxygen transport, of hydrogen peroxide catabolism. So, we already have signs of mitochondrial activity being higher in the T cells of these patients. So as a whole, using imaging approaches, using all kinds of epigenetic studies in collaboration with Amanda Pohalek, who's very talented at our school, we are trying to understand the high gamma status of about 40% to 45% of the severe asthma patients who go on to develop very severe disease. I didn't have time to talk to you about this other cell type, which we are also we are also as interested in. It's an innate cell type and novel cell type that the reviewers forced us to go back and use the little number of cells that we had left. We didn't want to all use up, but they said, "You have shown us computationally, we are happy that it is a novel cell type, maybe, but can you do it at the wet bench and go back and do your high-dimensional flow cytometry and show you can identify?" So, we have that data. It does exist, and it exists in our different mouse model. We have now established a new mouse model to have this aggravated innate response with IL-4 and IL-5 expression. I would like to acknowledge the contribution of many past members of the team. Not everyone is shown here. This is Mahesh, who published that JCI paper, the first paper with the SARP cohort where we identified the interferon gamma phenotype at the protein level for the first time. And then, of course, acknowledge the participation of our cohorts, patients, and then all across this country in IMSA as well as SARP, and the 10, 12-year-old collaboration with Sally, without whom we could not have done this work. And I'm so thrilled that our program project was renewed, that we can continue at least for the next 5 years. My husband, Prabir Ray, and with him and his expertise in mouse genetics and establishment of the mouse models, and then environmental informaticians, and Amanda, and the younger group that has come and joined Pittsburgh. And we are working with them to understand the disease at a more granular level. And of course, many years of collaboration with Jay Hari and her team, and the, of course, the NIH funding and UPMC endowment. So, thank you. Questions do we have in the chat? I don't see questions in the chat.

Yes, yes, we do. I mean, nice to see that finally there's a long-Tage2 molecule you're testing.

Yes.

Do you think that the interferon gamma signature could be a biomarker of an underlying process? Yeah. How much have you characterized these CCL high patients as opposed to CCL low patients?

Yeah. So this is an ongoing exercise. And that's where we would like to identify the interferon gamma. I thought we could in the sputum samples because it's a non-invasive way of getting the samples. And surprisingly, you know, you are a pulmonologist, I mean, what the sputum does not yield a very high enriched the way lower airways doing actual BAL gives you the T cell enriched group. So, you can go by surrogate markers. I may have that slide, but we are seeing as we are using the chemokine receptors as surrogates. And we are seeing an association with the CCL5 high group in especially in the severe asthma patients. So, that's where our struggle is. But we are doing one thing more, Vijay, and that is a very important question. Question where Matt began to do it, but he has gone off to industry now. And we are working with someone else at Pitt. We are integrating epithelial, blood, and airway data altogether as integrating the data to see where the signature of each lies of these patients so that we could now begin to tease out, is it in the blood? We have made some headway with actually Matt. It's unfortunate that he left at the time he did because we are seeing some signatures of neutrophilic abundance in some of these patients with the T-cell high group in the blood. But, you know, it's premature for me to say, but that's where I push Mark, tell him go and find the biomarker. Who are these patients? And the industry people are asking us to do the same thing what you asked, that can you— you cannot so easily measure in sputum. They don't want in that. No one wants to identify in valve sample, but that's an invasive way of going, you know, oh, this is the— that's not the way to do that. That's why only a Pheno and, uh, used by industry. Uh, no one would go and, you know, in a company do bronch people.

Specifically follow up within your RNA-seqs that you have in these patients. If you took the CCL5 patient, yeah, is the RNA-seq giving unique signals?

Yeah, so the CCLR5 high patients, which are the— they— we have the— see, this is with the sputum data. The reason that we don't have yet is because UCSF wants to publish their work first with the sputum RNA-seq data, and that part, if we don't have access to yet. And so once they have that, it should be submitted soon, and I think Mark will be part of it. I'm not part of SARP. I can advise them, but Sally is, and, you know, each center is bound. There are many constraints there. So once they have the data, then we will go and see that. That's absolutely the thing to do.

Are there any likely— I only follow that question because I'm kind of naive in asthma field, but I have a vaguely concept that if you are very exposed to a type 1 immune context, let's say like bacterial infection, you should have a decreased incidence of allergic asthma. So like in your severe— because you also speak like more severe mouth model, you also use the HDM with the allergic asthma model. Do you think like there's a sequential consequence? Like if you have like asthma-like symptoms and after that you have an infection, it will maybe lead to more severe like asthma symptoms, not like the other one?

Yeah.

And also like I also wondering that What factor like induced the generation of the CD161 positive red cells generation?

Two very good questions. And I would say that, look, you can talk about exposures all the time, but there is an inherent genetic predisposition, you know, in these people. Some people are at increased risk for developing severe disease, whether it's that 17 locus are not what we don't know. But there is something there, and it was nice that you picked it up, that yes, I would agree. I myself said that, that for the longest time people felt that if you have interferon gamma, some infection giving you, or something giving you interferon gamma, that should be protective. Why is it not protective? And why do you have this unleashing of— and that cannot be controlled— of such a response where you generate memory and you keep getting that higher response. We don't know yet, but I think that's the last slide I showed you. And I feel that many things get into that from— and we actually do have some data there where we are doing ultrastructure studies of the mitochondria and everything, that there is something about a very high level of ROS production with the mitochondria looking very different from what you see in healthy controls are, you know, how mitochondria fuse, and then there's fusion-fission. And that has been linked to senescence or this SASP phenotype through epigenetic mechanisms. There are many things that connect there. And maybe they have to begin with in some defect in their SNPs have been noted at TFAM, the transcription factor that was in that Science paper that associated with with high gamma. We don't know what it is. I would not speculate. But there is something that has to do with an inherent risk for producing more gamma. But there has to be an insult. And what that actual insult— it may not be the same in different people. It could actually be their autoantigen that triggers that. And then you have the other allergen exposures that go hand in hand. So I cannot give you an answer at this time. But it's a Good question.

Yeah, I think for the sake of time, we'll just say thank you.

There is post-op lunch after this, please. Yeah, please. Yeah, I'd be happy to chat.