Jan 21, 2021
This month on Episode 20 of the Discover CircRes podcast, host Cindy St. Hilaire highlights four featured articles from the January 8 and January 22 issue of Circulation Research. This episode features an in-depth conversation with Drs Stefanie Dimmeler and Wesley Abplanalp from Goethe University in Frankfurt, Germany, regarding their study titled Clonal Hematopoiesis-Driver DNMT3A Mutations Alter Immune Cells in Heart Failure.
Article highlights:
Li, et al. FA Scaffold Genes Are Novel TAA Genes
Z Perestrelo, et al. ECM Structure and Mechanics in Heart Failure
Castranova, et al. Zebrafish Intracranial Lymphatics
Rogers, et al. Computational Phenotypes for VT/VF Risk
Cindy St. Hilaire: Hi, and welcome to Discover CircRes, the podcast with the American Heart Association's journal, Circulation Research. I'm your host, Dr Cindy St. Hilaire, from the Vascular Medicine Institute at the University of Pittsburgh. Today I'll be highlighting four articles selected from the January 8’th and January 22’nd issues of Circ Res. After the highlights, Dr Stephanie Dimmeler and Wesley Abplanalp at Goethe University in Frankfurt, Germany will join me to discuss their study, Clonal Hematopoiesis-Driver DNMT3A Mutations Alter Immune Cells in Heart Failure.
Cindy St. Hilaire: The first article I want to share is Variants of Focal Adhesion Scaffold Genes Cause Thoracic Aortic Aneurysm. The first authors are Yang Li and Shijuan Gao, and the corresponding authors are Jie Du and Yulin Li from Beijing Institute of Heart, Blood and Lung Vessel Disease in Beijing, China. Thoracic aortic aneurysm is the localized expansion of the blood vessel. This expansion causes weakening of the vessel wall, causing it to rupture, which is a life-threatening emergency.
Although there are several genetic mutations that lead to thoracic aortic aneurysms, more often thoracic aortic aneurysms occur as an isolated event with no known cause or family history. To gain greater insight into the genetic underpinnings of isolated thoracic aortic aneurysms, Li and Gao and colleagues performed whole exome sequencing of DNA from 551 patients and 1070 healthy controls. They found that five percent of the patients screened harbored mutations in previously identified genes associated with TAA. Importantly, they identified a number of novel candidate gene variants in the remaining 95%. In four patients they discovered mutations in a gene named Testin. Testin is a scaffold protein found at focal adhesions, which are the points of connection between the extracellular matrix and the cell’s intracellular cytoskeletal framework. Mice that lacked Testin, or carried a mutant version of it, had dilated aortas and impaired contractility of vascular smooth muscle cells. Moreover, the team found additional focal adhesion gene variants present in the patient cohort, suggesting weakening or dysfunction of the structural elements may be a driving force in thoracic aortic aneurysm pathology.
Cindy St. Hilaire: The second article I want to share is titled Multi-scale Analysis of Extracellular Matrix Remodeling in the Failing Heart. The first author is Ana Rubina Perestrelo, and the corresponding author is Giancarlo Forte, and they're from St. Anne's University Hospital in the Czech Republic. After a myocardial infarction, when a lack of blood supply causes injury to the cardiac muscle, the damaged muscle tissue is patched by proliferating fibroblasts and the remodeling of the extracellular matrix. This is a process that is called fibrosis. However, this fibrotic process often continues after the initial repair and itself causes progressive loss of cardiac function. To better understand how cardiac fibrosis progresses, Perestrelo and colleagues examined cardiac extracellular matrix and fibroblasts from patients with and without heart failure. Using both microscopy and mass spectrometry, they found that the extracellular matrix from heart failure patients had a larger content of collagen and other extracellular matrix proteins, as well as more compact and less elastic fibers than non-heart failure controls.
Cindy St. Hilaire: RNA analysis of fibroblasts from patient and control hearts revealed heart failure patients had increased transcription of genes involved in assembling both the extracellular matrix and focal adhesions, which, as we just learned, are the points of connection between extracellular matrix and the cell's cytoskeleton. One such gene encoded the transcription factor yes-associated protein, or YAP. In cardiac fibroblasts, high levels of YAP drove expression of extracellular matrix factors. Similarly, extracellular matrix material from heart failure patients was particularly potent in triggering YAP activity. In highlighting this positive feedback of extracellular matrix remodeling, the work suggests that blocking this YAP-driven process may be an effective strategy for slowing heart failure pathogenesis.
Cindy St. Hilaire: The third article I want to share is titled Live Imaging of Intracranial Lymphatics in the Zebrafish. The first author is Daniel Castronova and the corresponding author is Brant Weinstein from the National Institutes of Health in Bethesda, Maryland
Until recently it was believed that the mammalian central nervous system lacked a classical lymphatic system. However, that belief was overturned a few years ago when canonical lymphatic vessels were discovered in the mouse brain. The discovery has implications for the understanding of the brains inflammatory and protein clearance processes, as well as disorders associated with these processes, such as in Alzheimer's disease. Because in vivo analysis of the mammalian brain lymph system is hindered by the thickness of the skull, Castronova and colleagues turned to a recently engineered zebrafish that has practically transparent tissues. Visualizing the fish brain through the top of the animal’s skull, the team found a complex network of lymphatic vessels covering much of the brain, particularly the cerebellum and the optical areas. The team confirmed the identity of the vessels with a series of lymph markers and showed that the vessels both carried out tissue drainage and contained trafficking neutrophils. The work introduces the fish as a valuable model for studying intercranial lymphatics in both health and disease states.
Cindy St. Hilaire: The last article I want to share before we switch to our interview with Dr Dimmeler is titled Machine Learned Cellular Phenotypes Predict Outcome in Ischemic Cardiomyopathy. The first authors are Albert Rogers and Anojan Selvalingam. The corresponding author is Sanjiv Narayan from Stanford University in Palo Alto, California. Sudden cardiac arrest affects over 300,000 people per year in the US alone. Individuals with reduced left ventricular ejection fraction are at an elevated risk for sudden cardiac arrest. Many of these patients qualify for implantable cardiac defibrillators. However, in the first year of implantation, these devices are rarely needed to deliver life-saving therapy, and identifying means to further risk stratify these patients has been elusive. The authors of this study hypothesized that the morphology of individual ventricular monophasic action potentials in patients with ischemic cardiomyopathy could possibly identify tissue or cellular electrophysiological phenotypes that can be identified by machine learning and then be used to predict long-term outcomes for patients.
Cindy St. Hilaire: Using 42 patients with coronary artery disease, the team recorded 5,706 ventricular monophasic action potentials and left ventricular ejection fraction during steady state pacing. Patients were then randomly allocated to independent training and testing cohorts. Support vector machines and convolutional neural networks were trained to two end points. The first, sustained ventricular arrhythmia, and the second, mortality at three years. Patient level predictions in independent test cohorts yielded a strong concordance statistic and were the most significant multivariate predictors. This machine learning of action potential recordings in patients revealed novel phenotypes for long term outcomes in ischemic cardiomyopathy. These computational phenotypes may reveal cellular mechanisms for clinical outcomes and could be applied to other conditions.
Cindy St. Hilaire: Today, Dr Stephanie Dimmeler and her postdoctoral fellow Wesley Abplanalp, from the Goethe University in Frankfurt, Germany, are here to discuss their study, Clonal Hematopoiesis-Driver DNMT3A Mutations Alter Immune Cells in Heart Failure. This article is in our January 22nd issue, the second issue of 2021. Thank you both very much for joining me here today. I know it's the evening where you are, so I appreciate you taking the time to sit with me.
Wesley Abplanalp: Of course. Thank you.
Cindy St. Hilaire: I think I want to start with the definition of clonal hematopoiesis. Just to get every listener on the same page. Our bone marrow produces billions of blood cells every day and the traditional view is that maybe 10 to 20,000 of these hematopoietic STEM and progenitor cells create all the progeny blood cells. This idea of multiple hematopoietic and progenitor cells, or HSPCs, is in contrast to this phenomenon called clonal hematopoiesis. That is where a sizable portion of the differentiated blood cells at a given time in a human has been derived from a single, kind of dominant HSPC. This idea of clonal hematopoiesis can really best, I think, be conceptualized when we think about cancers, like leukemia, but a lot of clonal hematopoiesis has been linked to what is called clonal hematopoiesis of indeterminate potential, or CHIP. I was wondering if you could kind of give us a little bit of details about what is known regarding the drivers of CHIP, this clonal hematopoiesis of indeterminate potential, and what's the genesis of exploring the role of CHIP and how it affects cardiovascular health, and specific to your study, heart failure?
Stephanie Dimmeler: Well, thank you very much. Maybe I start with the more general question, actually the CHIP refers to the occurrence of mutations in hematopoietic STEM cells, which leads to the extension of these mutated cells. Initially it was thought that this is correctly linked to cancer and the development of the leukemia, but it turned out that the occurrence of such mutation is not exclusively seen in patients with leukemia, but actually also healthy persons can acquire such mutation, and has such mutations of blood. So it's an age-dependent phenomenon and with increasing age, up to 20% of the people have such mutations.
Cindy St. Hilaire: Wow, 20%.
Stephanie Dimmeler: If they are old enough. There are a few of such mutations, particularly in these enzymes, which we are also studying in MTCA, or type two, this doesn’t lead to leukemia, but still subject to such mutations die earlier. So they have a poor prognosis. As said it was not linked to leukemia or cancer, but it was shown by Eisner and colleagues and Dr Libby also that such mutation have a higher risk of dying from coronary artery disease. We have added to this information, and our group is actually working together with the hematology department, that also heart failure patients with such mutations have a very poor prognosis.
Cindy St. Hilaire: Interesting. Maybe when we're talking about clonal hematopoiesis I know in your paper you mentioned, I think it was three or four commonly found mutants. Is there something shared between these genes that are mutated? I know in this study we're focusing on DNMT3A, but what does that do normally and what are some of these other drivers of clonal hematopoiesis, and is there a similar theme to the mutations?
Wesley Abplanalp: Well, I guess I could jump in a little bit. We're looking at DNMT3A, and the other very commonly found gene that's often mutated is TED2, and I think what's really interesting about these two genes is that they can both epigenetically control gene regulation, of course. So one is a DNA methyltransferase and of course the other has the opposite effect. This is not true necessarily for all CHIP-associated mutations, but I do think it is quite interesting that these tend to be the two most abundantly found mutated genes, especially in the context of CHIP, and especially within these heart failure cohorts.
Cindy St. Hilaire: So how common are these mutations? I guess we can specifically talk about DNMT3A. How common is that in the general population as a whole? Do we know that yet?
Stephanie Dimmeler: Yeah, some studies, it's a clear age dependent phenomenon. It depends on the age. In young subjects only very few subjects have such mutations, but with increased age, like 80, for example, you have a significant number and it's even higher if you look at heart failure patients, who have up to 40% of heart failure patients with high age having such mutations. Also, of course I have to say it depends a bit how you count the mutations. It depends where you set the cut off. We, for example, set the cutoff at two percent of mutations carrying DNA in the blood, and it depends with the numbers, depending with how low or how high you set the cut off.
Cindy St. Hilaire: Wow. So really it could be quite high. I guess I didn't realize it was that high. So could you maybe walk us through the study? What did you start with and what analysis did you perform?
Wesley Abplanalp: So with this study, we enrolled subjects with chronic ischemic heart failure. We were beginning there because we'd already found that the patients who have chronic ischemic heart failure and harbored these mutations have a worse prognosis. So our question is clearly there's something in the blood that's happening that's maybe facilitating this. We wanted to know more about this. It was already understood that these cells might be associated with inflammation, but the real question is we wanted to know what the transcriptional signatures would be in these patients. We enrolled six heart failure patients with a DNMT3A mutation.
Wesley Abplanalp: We had already screened through subjects before, and another four subjects with heart failure, with no known CHIP associated mutation. So we screened for 56 mutations that are associated with CHIP or other hematological malignancies. Then this is how we began our cohort. From this, then we took the peripheral blood from these subjects and use the peripheral blood mononuclear cells. So basically it'd be immune cells, which are circulating from these subjects and then performed a droplet single cell RNA sequencing analysis on just these circulating cells. We didn't necessarily enrich any cell type. We were trying to take an unbiased approach to really capture what was happening in the landscape. Then from here, we really dove down into some of the most abundant cell types that are there. For example, the monocytes and the T cells.
Cindy St. Hilaire: That's great. I think I read you found there were no significant changes in the actual types of cells. So both mutant and control populations had similar numbers of different types of cells, but what you did find that was significantly different, was the gene expression profiles within the cells of the mutant versus the controls. Can you talk a little bit about these changes? What was the same and what was different between your groups?
Wesley Abplanalp: Right. So one thing that I think that was really striking, like you're saying, or important, we're not seeing the change in the relative shifts in the abundances of cells. So therefore we can ask what's really happening within the cell. That's where the strength of the technology really has its full effect, I guess. What we're seeing is we were able to kind of add and confirm that the hallmark inflammatory cytokines, like IL-Beta and IL-6, IL-8, for example, were upregulating, this gives us great insights for potential interventions for these subjects, for example. These were quite different. We were also seeing, so we found this in the monocytes, we were also seeing an increase in resistin. So this was at this point, unknown. In resisitin, I think is a really interesting molecule because this is a secreted protein. It's been shown that when endothelial cells are exposed to this, that they become activated.
Wesley Abplanalp: We could also take this in vitro and silence DNTM3A in monocytes, and then we could add the supernatant. So what's secreted from these monocytes and add them to otherwise naive endothelial cells. We could see indices of endothelial cells becoming activated through increases in IL-Beta and BK-1.. We are additionally showing increasing interactions between endothelial cells and monocytes, which had otherwise not been shown before. We are also kind of showing these novel interactions between monocytes and T-cells, which I think is really cool because then you wind up having this capacity for a small number of cells enriching the impact on the greater blood and population. Through the interactions with T-cells and endothelial cells, we wind up seeing strong evidence, for example, for a potential bystander effect. So that for a few rogue cells to really have a much broader impact on these cells in the greater milieu.
Cindy St. Hilaire: Yeah. I found that graphic, your graphical abstract, it was something just really neat to think about in terms of there's this clonal hematopoiesis, but it's not a hundred percent, right? Is that correct? These mutations aren't in every single one of these circulating cells. I was wondering, could you find evidence that the inflammatory, maybe this would be an in vitro study, the inflammatory activation or cytokine release or activation of endothelial cells is better, worse, the same if the monocyte has it versus the T-cell has it, versus both? Are they equal contributors? What, I guess what, in terms of a stepwise progression, where do you think these mutants are more potent?
Wesley Abplanalp: Oh, this is a good question.
Cindy St. Hilaire: If you can speculate. Maybe you can't speculate yet.
Wesley Abplanalp: Yeah, I think at this point, I think it's a lot of speculation. With the monocytes we wind up seeing, I think some of the biggest changes we wind up seeing are just in these cytokine, they turn into kind of a cytokine factory. They just really push out these cytokines. There's a much more mixed response, I think from the T-cells, but inherently the T-cells are a much more diverse population. So it begins to become a little bit difficult to compare because they all have different roles and different functions. Unfortunately, I think it's finding a good T-cell model in vitro, for example, is a little more difficult than recapitulating some of this for monocytes, for example.
Cindy St. Hilaire: I'm not an immunologist, I guess, being a vascular biologist, you're a little bit immunologist always, but what are our abilities to model this in a mouse system? Are these cell types very easily translatable between mouse and human? I know there's different cytokine profiles when we're talking macrophages and things like that, is it similar for T-cells also? Or is that more translatable?
Stephanie Dimmeler: The proton signature have been observed in type two hetero zygote STEM cell transplants in mice as well. Also, DNTM3A editing has been shown in actually, a very nice Circulation Research paper by Ken Walsh to have an effect also on heart failure. These studies suggest that at least the pro-inflammatory signature in monocytes can be recapitulated as well, can be the phenotypes that grows its development, as shown by [Ken Walsh, and the heart failure phenotypes. I think Cynthia, you've touched upon a very interesting question. Mainly, to what extent is the transcription we are seeing related directly to the mutation of the cell? To what extent can you explain and understand that so many of the cells have this changed signature? I think our data clearly suggests, at least for the monocytes, that it's not only the mutated cell alone.
Stephanie Dimmeler: They have also a channel effect on the other, non-mutated cell, because our percentage of cells which has the mutation is between maybe 2% and 10% to 30% to 40%. We have many more cells which have the inflammatory signature. In the T-cells it's a bit different maybe, but also we have excessive activation. I think we don't know yet the reason. It could be many, but we are really doing, what we are currently doing is to try to target which cell is mutated and which cell is not mutated in humans. Then we can distinguish biochemically between the mutation carriers and the biotype cells, and then we can tell more what happens directly, and what happens secondarily, because of course, if you have an inflammatory cell, this inflammation can influence the neighboring cell as well. In the STEM cell niche, the cells are also in the same environment, so] how much your environment in STEM cell niche, this may also affect the neighboring other hematopoietic STEM cells. This is what I think is the next step to do.
Cindy St. Hilaire: Yeah, that's one thing I was actually thinking about is obviously the clonal hematopoiesis aspect means there is a STEM cell harboring a mutation that is selecting or allowing that population to grow much more efficiently or faster than the non mutant cells. What is that doing to the other STEM cells in that area? Is it inflammation just in the periphery once these cells are differentiated in little packets of cytokine releasing cells, but yeah, what's happening at the level of the niche in terms of these mutations?
Stephanie Dimmeler: A very interesting question. I think it's so far understudied, at least we have no insights, but what is known is that for some heart failure, and acute myocardial infarction also had an impact on the STEM cell niche, so there is activation of the osteogenic niche, there is a change in the vascular niche. So this are maybe effects, which may link also hematopoiesis and the cardiovascular and heart disease on some level, but to know what is he and egg need some more studies to do.
Cindy St. Hilaire: Yeah, yeah. As always, I have that same problem with calcification stuff. What is the cause? What is the consequence? I guess it's ripe for funding and studies. The schematic, your graphical summary is really focusing in on this monocyte T-cell interaction on the endothelium, which is obviously not the heart tissue itself. So how do you envision, this is kind of easy to conceptualize or picture when we're talking about an atherosclerotic plaque, right? It's right at that interface where these blood cells are touching the vasculature, but what do you think is happening to exacerbate or drive the heart failure component regarding clonal hematopoiesis and specifically this mutation?
Wesley Abplanalp: So one thing that we are seeing, and I think that's particularly interesting, is perhaps this interaction between the endothelium, like you were saying, and monocytes. Ken Walsh, and I think others have shown, but Ken Walsh I think had shown that with a DNMT3A loss of function study, that they could see an increase in the extra visation, or these monocytes coming in to the myocardial tissue. That there seemed to be some kind of indices of heart failure that then accompanied this. How or why this was happening I think people didn't really know, but I think there are many ways in which this could happen. If, for example, if these cells are secreting resistin, and there's increased adhesion molecules, and of course there's increased chances for extra visation where these cells can leave the blood and then go into the heart.
Cindy St. Hilaire: Has anyone ever looked at cardiac tissue from patients harboring these mutations and seen differences in either the cardiomyocytes or the cardiac fibroblasts or anything like that? Or is that what you're doing next perhaps, or-
Wesley Abplanalp: Stephanie's smiling.
Stephanie Dimmeler: We tried to, the problem is there's difficulties in these biopsies to get enough material to study the mutation.
Cindy St. Hilaire: Of course.
Stephanie Dimmeler: So far, the problem with the patients also incomparison to mice models that we have to clean off the circulating blood, which may also be stuck in the tissue. Therefore, we have some results, but not yet for publishing because the reviewers will make us pass time to make this more confusing. So, as soon as we can detect by single cell on a sequencing or nuclear sequencing, the mutated cells and the non-mutated cells, then we have a chance to get more insights. So far, we cannot distinguish really the cardiac monocytes versus the circulating monocytes, which makes our study a bit more difficult. Really I can add one more point to your previous question.
Cindy St. Hilaire: Sure.
Stephanie Dimmeler: I think one, an interesting message of our paper would also be that if you have proinflammatory monocytes coming from the circulation into the heart, and our study will claim that they are more likely to hold through the heart tissue and then invade, because endothelial affects the capacity. Then they may replace the cardiac macrophages, which would be seen as a more physiological and protected type of cell, which is involved, as we know in electrical continuation and so on. Then we have these bone marrow cells, which are increasing in holding and this of course could aggravate all the heart failure in addition to the cytokines they are producing. I think, finally, also one shouldn't neglect the T-cell. T-cells are known to play a role in heart failure, and if we have activation of T-cells, which we have to prove now in the cardiac tissue, but at least in the circulation here, I would assume that this is the case. This could have also consequences of course, and could link hematopoiesis to heart failure phenotypes as well.
Cindy St. Hilaire: So where do you think this will go in the future? How could this knowledge, and possibly even single cell sequencing technology, be leveraged for therapies in the future?
Stephanie Dimmeler: So my wish would be that our data would be leading to a type of guided therapy. If we understand better which mutation affects which pathways, or which genes, we may use specific anti-inflammatory treatments, not global anti-inflammatory treatments, but more specified or specific treatment strategies to target patients with mutations. I think single cell sequencing is a very good way to start it. Followed by proteomics and then other -omic technologies. This is actually what I would wish we could do with it, which other people hopefully start some pilot trials with patients and try some treatments, but we could maybe get even further insights by the single cell data. Then the responses also of the treatments, the single cell sequencing, you can see, but it can normalize the phenotype, which would be, of course, the wish.
Cindy St. Hilaire: That would be amazing because we tried in the Cantos Trial to just, let's block inflammation and see what happens. There were certain populations where it seemed to have a much greater effect than others, and maybe targeting clonal hematopoiesis could help tweak or tighter those therapies. This was a great study. I want to commend you both on this excellent story and thank you so much for joining me today.
Stephanie Dimmeler: Thank you very much. Nice to see you again.
Wesley Abplanalp: Thank you.
Cindy St. Hilaire: That's it for the highlights from the January 8th and January 22nd issues of Circulation Research. Thank you for listening. Please check out the CircRes Facebook page and follow us on Twitter and Instagram with the handle @CircRes and #DiscoverCircRes. Thank you to our guests, Doctors Stephanie Dimmeler and Wesley Abplanalp. This podcast is produced by Rebecca McTavish and Ashara Ratnayaka, edited by Melissa Stoner, and supported by the Editorial Team of Circulation Research. Some of the copy text for the highlighted articles is provided by Ruth Williams. I'm your host, Dr Cindy St. Hilaire, and this is Discover CircRes, your on-the-go source for the most exciting discoveries in basic cardiovascular research.