Scientist Stories: Tony Wyss-Coray, Brain Aging and Neurodegeneration

Welcome everybody. It's a great pleasure to introduce Tony Wiskorek from Stanford to give our seminar this morning. Tony obtained a PhD from University of Bern in 1992. And did work in a number of places, including UCSF, before moving to Stanford, where he has been located for, I guess, almost 2 decades. And he has positions at Stanford and at the VA. And Tony has done terrific work on several themes, looking at the roles of immune and injury responses in neurodegeneration, importantly, how aging to many of these factors and also to Alzheimer's disease. And he's used a number of different pioneering approaches, including animal models, genetic approaches, and large-scale omics, including proteomics. He's received many honors and awards. If I go through all of them, I'll be cutting into his time for talking. So, um, With that, it's a great pleasure then to invite Tony to bring us up to date about some of the aging and regulation of degeneration.

Thank you, Doc. Thank you all for having me. It's a great pleasure to be in San Diego again. I I came as a postdoc to Scripps, so San Diego is very dear to me. One of our daughters was born there, but I know it's only virtual. So disclosures. Our lab is interested in understanding how the brain becomes susceptible to cognitive dysfunction, neurodegeneration, and diseases such as Alzheimer's disease with aging. And the idea that age is a key driver of neurodegeneration is, of course, not new. And this slide here visualizes that aging is a key risk factor for actually a large part of diseases in our society. And it's estimated that at least half of over-65-year-olds have 2 or more conditions. That are clearly driven by the aging process itself. And you can see here how each one of these diseases sort of increases the percentage in the population that has 1, 2, 3, or up to 8 different age-related diseases. And the approach to deal with this problem has been to develop individual treatments for diseases. Such as heart disease, diabetes, and so forth. But a growing sort of drive in the research field that comes really from the process of trying to understand aging has proposed to try to understand the aging process and actually target itself in a therapeutic way. And the idea would be to slow down or reverse aspects of aging by understanding the molecular processes that underlie aging in what is often called rejuvenation. And this is, of course, not a new idea. This has been a dream of humans since recorded history, probably. This is just one example of a painting from Lucas Cranach the Elder in the 16th century who shows here a fountain of youth. That would rejuvenate people. So they enter old and frail on one side and emerge young and rejuvenated on the other. Of course, we know this is not reality. It's not close to within reach, I think, for humans. But what is really amazing is that there are a growing number of interventions in animal models, and most excitingly, I think, in mouse models, in complex mammalian organisms that show that interventions can— that target the aging process can indeed slow down or reverse it and rejuvenate these animals at multiple levels from molecules to cells, tissues, and most importantly, function. And I just list a few here. Metabolic interventions such as caloric restriction, exercise, small molecules that target specific processes of aging, removal of senescent cells. One of the hottest topics may be transient epigenetic reprogramming of aged cells, so whereby you activate the Yamanaka transcription factors that allow you to transform an adult cell into induced pluripotent stem cells by activating these transcription this only for a short period of time and turn them basically on and off, what people discovered is that this does not put the cell back to a stem cell, but rather makes it younger and erases some of these age-related epigenetic marks that people have described. And then treatment with young blood is something that we have been working on, but just sort of highlighting the the hype, if you will, in this field. And this is actually Juan Carlos Belmonte, who is at the Salk, has been really a driver in this process. Sorry, is that UCSD? Or no, Salk. He has been really a driver of this field and of epigenetic transitory programming. And This has contributed, I think, in large part to this absolutely incredible investment of $3 billion into Altos Labs to try to harness this concept and see what this can be applied to the treatment of age-related diseases and even slow down the aging process to increase healthspan. Now, Juan Carlos was actually quoted that within 2 decades we will be able to prevent aging. I don't think that's what he really said, but it just shows sort of the excitement that exists in this field. Now, we have been working on, as I said, on trying to understand the role of the systemic environment in regulating the aging process. And we did this by using a model called parabiosis, and this was really inspired by work from Tom Randall, who in recent years revived these models to study stem cell aging. So, Tom was interested in aging of the muscle, and with age, the muscle does not rejuvenate as a young muscle does if you extend it, exert it excessively, or if you injure it.. And what Tom was able to show is that if these old stem cells in the muscle are exposed to a young environment, a systemic circulatory environment from a young mouse, then the muscle would regenerate and basically almost histologically look like a young muscle. He also saw similar effects in the liver and noted that in the brain there was increased proliferation of cells in the hippocampus, and that was the basis of a collaborative project with our lab over several years that showed that indeed young blood can slow down and reverse aspects of aging, as I'll show you in a minute. Others have shown similar effects in an increasing number of tissues in the mouse, and this has been reproduced by many different labs. So, again, here the concept is that through this surgical model called parabiosis, there's a shared blood system, circulatory system, a capillary network that allows blood from the young animal to enter the old organism and vice versa. We were also able to show for the first time that you actually don't need the surgical model, but you can collect plasma from young animals and repeatedly infuse it into old mice and reproduce many of the effects that you see with parabiosis. And this was in the lab really pioneered by Saul Villeda, who's now an associate professor at UCSF. And Saul could show at the time that either with this model of parabiosis or by repeatedly infusing plasma— so we actually need only the liquid fraction of the blood— repeatedly infusing this over 3 weeks, every 3 days into old mice, that this increases neural stem cell activity, increases synaptic plasticity, and I think very importantly reduces brain inflammation throughout the brain, but specifically also in the vasculature. And it led to, most importantly, to an improvement in memory. And again, many other labs have been able to replicate these findings and also show even in vitro, Tom Sudhoff showed that iPS-derived neurons benefit from exposure to young plasma factors as compared to old plasma. What's also very interesting is that if you sort of do the opposite and you take plasma from old mice into young, you can actually accelerate the aging process and get the opposite effects that you get with young blood. You reduce neurogenesis, synaptic plasticity, you impair memory function, and you increase inflammation in microglia, astrocytes, and again, prominently also in the vasculature. And people have identified a number of different proteins that mediate some of these effects listed here, positive factors or negative factors that can accelerate aging. But I think the field accepts sort of that there is not a single factor and that there are probably many different different types of processes, and I'll get to that a little bit in a minute. Because plasma is— has been used in humans as a treatment, whether it's during surgery with— if you have blood loss, but also different fractions of plasma have been used in the treatment of different types of deficiencies, or antibodies are purified from plasma for people with immunodeficiency transplants, it became sort of obvious to try whether this would work in humans. It's a safe treatment. And I'll just list here two studies, one from the company that we've started, Alkahest, where we treated patients with mild to moderate Alzheimer's disease. This was a dose range finding study. It didn't have a placebo arm, but we got encouraging results functionally in patients if we compared this to historic placebo controls. Of course, this is not, you know, a final result, but it's, again, it's encouraging. And we have also proteomic data that suggests that there is indeed a rejuvenating effect to some extent that we also see in mice. And then this study here that was done by Grifols, which acquired Alkaes recently. And Grifols has been working on this idea for a number of years, initially with the idea to remove amyloid from the periphery, from this concept of the peripheral sink that amyloid accumulates in the blood, and if you removed it, it would be beneficial for patients. They did a study where patients with Alzheimer's disease are treated initially with what's called apheresis, where, so again, you remove aged plasma from these patients, and then they give them back an albumin fraction which is rich in growth factors. It contains several hundred other proteins but is 99% albumin. And what they could show is in a a double-blind placebo-controlled trial, which was published a couple of years ago, that these patients show significant clinical benefits on disease progression, memory, daily functioning, and suggesting that there is indeed something in this plasma fraction, this albumin fraction, that is beneficial. I should say that the plasma from which this that from the donors that provide this plasma or donate this plasma, they are typically an average age of 35 years. So, this is really young plasma. So, essentially what this study did is remove aged plasma and give a young plasma fraction. And again, this suggests that there are benefits that can be translated from mice to humans. Now, as a scientist, of course, you want to know how does this work. This seems to be too good to be true. So, how does young blood potentially rejuvenate the organism? What are the factors? Where do they come from? Which cell types does young blood actually rejuvenate? How do the factors reach the brain? And that got us sort of thinking more globally about the aging process. Do tissues actually age in synchrony? Do cells age in synchrony, or are some cells and tissues age faster in, in a given individual than other cells? And very much related to that, if we ever want to apply a treatment that is based on the aging process, we need to be able to measure aging. Like, if you want to treat hypertension, you have to be able to measure blood pressure. We want to treat aging, we have to be able to measure it in a reliable way. And because tissues and cells age in different sort of trajectories, we need to be able to measure the age of a specific tissue. So one of the questions that we wanted to know is how do cells and tissues age across the organism in an attempt to understand where would these factors that are potentially rejuvenating or age-promoting, where are they being produced? Where do they come from? And this was a project from graduate student Nick Schaum at the time, who took sort of initially a somewhat naive approach, looking back, where we thought if we just measured the transcriptome of every tissue in mice, we will be able to figure out where these factors are produced. Now, a lot of factors are produced in many different tissues, and so it turned out to be much more challenging, but it led to, I think, a very interesting and useful tool that we developed together with Steve Quake, who became the co-director of the Chan Zuckerberg Initiative and had these resources available to really make this a major project. We recruited about 18 faculty at Stanford with different tissue knowledge, and we made these atlases where at the single-cell level, we basically profiled every major tissue and measured the transcriptome of these cells that led to the single-cell transcriptomic atlas of the mouse. And then we looked at how these tissues, how the transcriptomes at the cellular level changes with age. And that really provided a lot of very exciting and interesting insight. I think one of the principal ones is that aging is not a linear process. It's also not a synchronized process across the organism. So, you can see this here, for example, in gonadal adipose tissue, You have your mice from 1 month to 27 months of age, and these are hundreds or sometimes thousands of genes that we were able to measure. This is bulk RNA sequencing, so the whole tissue, but we have this also at the individual cell level for all these— for over 100 cell types. And you can see that the process, how these genes change with age, is not linear, and it starts very early on.. And you can, of course, debate whether this is aging or still development. But clearly, with age, there are dramatic changes, and they're different in different tissues. So here we look at downregulated and upregulated genes, again, from young to old in all these different tissues. And you can appreciate that some tissues show relatively few changes with age, and others show just massive changes. With genes going up several-fold from young to old, fat tissues, interestingly, the skin showing major changes with age. And again, they don't change all in synchrony, leading the field and from many different organisms over the years to really conclude that aging is asynchronous at the tissue and cellular level. Trajectories of aging are not linear, and they are to some extent stochastic, but some of this is probably also genetically encoded. And so, in just a simple cartoon form, you can, of course, divide lifespan of an organism into these principal timeframes of development, reproductive phase, and post-reproductive phase. And I think while initially people thought sort of aging is something that happens mostly in the post-reproductive phase, we really see that the aging processes continue to take place throughout a lifespan. And again, different tissues undergo much more dramatic changes at different times during life. So, which cell types does young blood rejuvenate? And this is some— this is sort of the third project that we took out of this Tabula Moris effort where we actually took down the mice all at the same time now 5 years ago for single-cell analysis of just young mice, of aged mice, and then of this parabiosis model. What we typically use here is we use normal aging sort of as a ground truth, young and old mice. And then we have in the parabiosis model, we have pairs of young mice, pairs of old mice, they're called isochronic. And then we have these heterochronic pairs where young mouse is exposed to old, and we call this accelerated aging paradigm. So, how does the young how do old factors influence and potentially accelerate the aging process in young organisms? And then the old mouse exposed to young, looking at rejuvenation, what are the genes that are changed in a rejuvenating direction, so reversing the trajectory of aging? And just a few highlights from this study. When you just look at normal aging first and sort sort of by one measure that you can use, how many genes change in given cell types, you can sort them and you see that most cells show dramatic changes over sort of until medium age of 18 months in mice, which would be an equivalent of about 65 years in people. And they really all change quite dramatically with age. If we look at the heterochronic mice, either the young mouse exposed to old blood, we see many genes that change. Again, you see thousands of genes sometimes changing, and now we get a different sort of different ranking, if you will, of cells that are most affected by this aged environment. Interesting, many types of endothelial cells in different tissues are at the top here, consistent with the idea that blood is the main driver of this, and endothelial cells would be the first cells to be exposed to an aged environment. And so many of these genes are actually consistent with the normal aging process. And then When we look at which genes are most— which cells are most affected by young environment, we see interestingly stem cell types being most dramatically affected, such as hematopoietic stem cells, stromal cells, suggesting that these cells are indeed susceptible to the environment, to the circulatory environment. And when we look specifically at how many genes are changed in the opposite direction in this rejuvenation paradigm, we see that, that again, hepatocytes, for example, and stem cells are the ones that show clearly changes that reverse the trajectory of aging. So what we look at here is the percentage of genes that show the same change as you see with normal aging. And so this would be a consistent aging process. So we really mimic aging here in these type of cells. And then here we reverse the trajectory of genes that would normally go up, we reverse them so they go down again. Or vice versa. And here again, hepatocytes are the most susceptible cells, but then also a number of different stem cells. We wanted to know what are the key pathways that are responsible or are driving these processes. And what we see is that in the accelerated aging paradigm, for example, in fat stromal cells, that changes in extracellular matrix, upregulation of matrix genes involved in matrix formation is one of the key hallmarks of this accelerated aging paradigm. And this is consistent with what people have postulated to be a major hallmark of aging. And then equally exciting, if we look at the rejuvenation paradigm in hepatocytes, again, these are one of the most susceptible cell types to a young systemic environment with many genes going back to an earlier age, if you will. And here, mitochondrial electron transport respiration is one of the key pathways, again, consistent with a lot of studies by many different labs over the past decades that suggests that mitochondria are one of the key— sort of mitochondrial aging is one of the key hallmarks of the aging process. And so, it's really reassuring to see that with this intervention, with exposure to young blood, to young plasma, we can reverse some of these hallmarks of mitochondrial aging across actually many different cell types. So, if we look at which pathways are most dramatically changed across the whole organism, it's mitochondrial electron transport. You can also look at this in a sort of overview of all the major cell types that we have looked at in the different cells and asked, what are the key changes that you can see with aging, for example? And so here in this graph that we call the Picasso plot. We show with the size of the dots, the number of genes that change with age in each cell type. So the larger the dot, the more genes change with age. And the lines that collect the cell types show that the similarity in the top genes that change with age. So we see across the organism in individual cell types common signatures of aging. So some cells really share the same transcriptional process as they age. And what's really interesting is if you look at this across different cell types and tissues, you see, for example, that stromal cells or endothelial cells across many different tissues show a concerted sort of aging signature, suggesting that this is driven by the cell type itself. Whereas in other cells, you see that across different cell types, they show common signatures of aging in a given tissue such as the skin or the pancreas, suggesting that here there may be an environment that promotes the aging process and triggers sort of similar genes that change. That's, of course, only one interpretation, but it's interesting that we see these sort of signatures that allow you potentially to find out what are the key pathways that make up these signatures and can we intervene with therapeutic approaches potentially to slow down or reverse these processes. You can also look at this in the accelerated aging or rejuvenation paradigm. And you see that some of these age-related common signatures persist, and some of them form new sort of common signatures of aging or rejuvenation. Another study that recently took a similar approach looked at caloric restriction in the rat. And they looked specifically at young and old rats that had ad libitum food, so they could eat as much as they want, and the rats become slightly obese as they get old. And then the caloric restriction ones. And they looked also at the single-cell level in 6 different tissues, whether they can find sort of common signatures. And they could, again, like we did find that cells really respond in individual ways to these environments. They could show accumulation of pro-inflammatory cells in various tissues, and that caloric restriction attenuates aging-associated cell-type-specific gene expression changes. And it will be really interesting, I think, to move forward to look what are sort of common signatures of accelerated aging or more importantly here, comparing it to caloric restriction, of rejuvenation. So, if we looked at rejuvenation with caloric restriction, exposure to young blood, transient epigenetic reprogramming, or exercise, can we find common signatures of rejuvenation, if you will, and can these become sort of key new targets to intervene in the aging process and again increase healthspan and prevent some of these devastating age-related diseases. So one of the important questions in trying to understand the aging process is, of course, what are the factors? And then, as I said earlier, how can we measure the aging process? So again, if you think about this, why would this potentially work? The blood is really the tissue that connects all the other organs in our bodies and is not just a carrier of cells, of proteins, the immune system and of gases like oxygen and CO2, but it's really a carrier of information, if you will, between information that allows exchange between tissues. So for example, we talked about exercise. Exercise is beneficial for many different tissues, is beneficial for the brain. So, there are now studies that show from, from Saul Villeda's lab and from, from our lab, but also from many others earlier, that if you exercise, for example, mice, you can take their blood, you can inject it, infuse it into a non-exercise mouse, and you can reproduce some of the benefits of exercise, suggesting that the exercise effect, or at least part of it, is indeed in the blood. There must be factors in the blood that can confer this effect. And what interestingly both Saul's lab and our lab found is that we identified each one protein that mimics sort of some of these effects in plasma. And both of these proteins are liver-derived, suggesting that the muscle first talks to the liver and then the liver produces a factor into the blood that is beneficial for the brain. This is just one example of, I think, of showing us how tissues really communicate through the blood. And we can now measure thousands or even 10,000 different proteins in the blood. And start to quantify this physiological or pathophysiological communication between tissues. And so, we can ask, how does the blood change with age at a molecular level? What is the composition of the blood as an organism changes? And this is sort of an extension, of course, of what we use in clinical medicine. If you go to the doctor, they take a blood sample and measure all kinds of metabolites or proteins that tell us whether your organism is healthy or not, whether specific tissues are healthy or not. But you can expand that to a much broader— I think to obtain much broader insight into molecular processes, cellular processes of an organism. And so we started to do this probably 15 years ago or so with platforms that were not too reliable. They had too much background noise. We could not get really reproducible data, but over the past 5 to 10 years, a couple of platforms have really been able to measure thousands of proteins now very reliably and consistently, reproducibly. And one uses these modified oligonucleotides, they're called aptamers, and the company calls them somamers. Somalogic is the company, and these function like antibodies, and actually the molecular interface between a given protein and this aptamer has roughly the same footprint as an antibody interacting with an antigen. And so we used this in collaboration with Nir Barzil and Sophia Milman at Einstein College, and with data from the INTERVAL study, that is a healthy aging study in Europe. We were able to collect data from over 4,000 human plasma samples from people, healthy people aged 20 to almost 100. And overall, we had roughly 3,000 proteins that we were able to measure. And when you look at this sort of with a bird's-eye view, And you show on the x-axis here these 4,000 individuals. And then on the y-axis, the 3,000 proteins that we measured. So for each person, we list here the 3,000 proteins. And we use blue and yellow shades of color to show the changes in relative levels of these proteins. So blue would be low levels of a protein, yellow high level. And you can see again the dramatic changes in the composition of the blood in this cross-sectional study from young to old people. You can also appreciate that this is, again, it's not a linear process. Some proteins are young when you're— are low when you're young, they go up with age. Some go the other direction. But you also see changes that happen throughout lifespan. And you can actually, you can actually— excuse me, something's moving here— you can look at these changes that occur at different ages, and you can add up the number of proteins that change at each age. So how many proteins are significantly different, let's say, between a person age 30 younger than 30 compared to a person older than 30. And if you do that, you see sort of waves of aging. The most prominent, if you squint your eyes, is around here at age 35, then another one at 60, and then the most prominent one at age 85. And this is shown here sort of, again, summing up the proteins that change at each given age in people younger compared to older at that age, and we use 10-year intervals to calculate the number of significant proteins. And you see this very early wave of aging in the composition of the plasma proteome, another one at age 60, another one at around age 80. We can actually see this in multiple completely independent cohorts. With different platforms, with the Somalogy platform and another antibody-based platform called Olink. And you also see this at functional levels. So people have shown, for example, epigenetic changes that seem to peak around this age, around 60 and 80 in the blood, in immune cells in the blood, or even if you look at changes in function, cognitive changes, for example, some of them show similar trajectories. We can look at each of these waves or peaks. What are the proteins that change? And they are not the same. So we see different types of proteins. And it's interesting that again, some of these proteins are highly enriched in matrix-related functions and there's a lot of follow-up work to be done here to find out what exactly are these proteins doing and why do they change so dramatically. But I want to get now in the last few minutes to discuss unpublished data that try to really measure aging based on this insight that I've showed you before, the fact that you get these dramatic changes with age in the plasma proteome. Can you develop a model that correlates basically age with levels of these different proteins? And so people have built what oftentimes are called clocks with different types of molecular features that change with aging. And these include DNA methylation, facial features, telomere length, protein composition, metabolic changes, and so forth.. And maybe the most famous of all these clocks is the one from Steve Horvath that is based on changes in methylation pattern on the DNA. And this allowed Steve really to show for the first time that there are these changes with age that you can see in major tissues in humans, but also in any other organism. And based on these changes, you can then calculate in an unknown sample what the relative age is. And it has been suggested that the deviation from this predicted age has biological information. And so, we wanted to apply this to try to understand brain aging or aging of any tissue for that matter with the idea that if you find a protein in the blood and you can assign it to a specific tissue, you can potentially learn something about the physiology, the pathophysiology of this tissue, or the aging process itself. And this is work carried out by two graduate students in the lab now. And so here the concept again is that if you look at an organism, and we know now from sort of atlases of gene expression which genes are expressed in which cell type or tissue. You can use this information to assign the— here we have actually 5,000 proteins. You can assign these proteins to a specific tissue. Now most proteins, most genes are expressed ubiquitously, or at least a large number of them are, or they are expressed in many different tissues and cell types. So here we look specifically for those genes and proteins for which the expression is enriched in a particular tissue. And out of these 5,000 proteins that Somalarchic has on their platform, in this particular experiment, they will soon have 10,000 proteins they can measure. About 200 are highly enriched in the brain. So they're most likely brain-derived. And you can do this for all different tissues. You have liver-specific proteins, blood-specific proteins, and so forth. And because some of these change with age, we can now try to model the aging process for each tissue. And again, the way this works is, you know, the actual age of a person or an organism, and then you use the information of the changes and the proteome, for example, and you build a model. This is typically a regression model, a linear expression model most of the time, eNet or LASSO, that allows you to then calculate the relative age. You try to make— basically model the age of a person And what you get is a, a curve like that where if you would absolutely specifically predict the age of a person, they would be falling on this line here. But for most people, you get some deviation from this prediction, what is called delta age or the age gap. So for example, this individual is predicted to be younger. Than their actual chronological age, while this person is predicted to be older than their actual age. And again, the idea is that this deviation contains biological information. So if the model says you're younger, then data— a growing number of data in model systems, in in animal models has shown that this is indeed associated with function or with molecular changes in these tissues. And we can do this not just at the organism level like most people have been doing, but because we have information again from these specific tissues, we can build clocks now for basically all major tissues as shown here. And so the way we did this is we have a healthy aging cohort. We model the aging process with all the proteins from a given tissue that we measure, or for all— with all the tissues, what we call an organismal clock. And then a brain clock, a kidney clock, a heart clock, and so forth. And then we test, we validate these clocks in independent dataset, one again from near Barcelona, Sofia Millman at Einstein College, and then samples from our Alzheimer's Research Center here at Stanford where we have a lot of clinical information, longitudinal follow-up. And we ask, can we use this information, this prediction of age, to infer functional states can we derive biological information. And so the way this looks like, so you get a correlation with— if you look all— if you use all the proteins, you get a very high correlation between age overall and biological predicted age. And this is lower with different tissues, but again, we're really interested in the variance between your actual chronological age and the predicted age. And so just as an example, first the heart. So here are these individuals where we predict this is their actual age and this is the predicted age and the deviation. You could be younger at heart or older at heart, if you will. We can extract which proteins make up this clock. So, the heart clock is made up of only 8 or so proteins. Interestingly, some of the key proteins that are already used to— in the clinic to look at the health of your heart, such as troponin and natriuretic peptide, are picked out of the bucket of all the proteins that we looked at supporting the biological validity of this model. But we have many new proteins that have— seem to have diagnostic properties here on heart aging. And if we look in these two different cohorts, we can show that people who are predicted to have an older heart are indeed found to have atrial fibrillation or pacemakers in the ADRC, or they had heart attacks or bypass in the Longevity Cohort. So we did not— we trained our algorithm on healthy aging people, yet the prediction that somebody has an older heart is then associated in independent cohorts indeed with heart disease. And if we look at this Specifically at the Longevity Cohort, we have actually a 17-fold increased risk if you're in the top quartile of heart age to develop heart disease 15 years down the line. And oh, this is actually shown here. So we have here follow-up. The sample, blood sample was obtained at time 0. And if we look at these people 10 or 15 years later, the bottom heart age. So, those predicted to have a young heart, they barely change. They have no heart-related problems. The ones who are predicted to have an older heart, the top 25%, they indeed have— sorry, a 15-fold increased risk to get heart disease. And lastly, similar studies with the brain. Here again, we can model brain aging. Look at the bottom quartile, so those with young brains are the top quartiles, those predicted to have older brains. When we look in our ADRC, out of 22 people with Alzheimer's disease, 18 are in the top brain ages and only 4 were predicted to have a young brain have Alzheimer's disease. And when we look in longitudinal follow-up, so blood was obtained at time 0, we follow these people now over 5 years. Those with young brains show relatively stable cognition, whereas those predicted to have older brains they show a 7-fold increase in cognitive decline here with CDR-SUM, 2-point change as sort of a measurement. And we can, of course, use many other ways to look at this, but consistent with increased susceptibility to cognitive decline. And what's also interesting, if you look at this sort of from an organismal perspective again, and we have the age of each tissue in a given person. On average, people without Alzheimer's disease, their tissues are mostly young, whereas people with Alzheimer's disease, they have not only an older brain, but they also show accelerated heart, muscle, and aging of other tissues. And it will be interesting to see what are the proteins here. Of course, the link between heart disease and cognitive decline, dementia is well established. And it will be interesting to see what are the proteins that are making up these models and are they potentially related. So with that, I would like to thank my lab, many collaborators who were just really tremendous help in carrying out some of these studies and then funding from all these sources. Thank you.