Why Some Brains Never Get Alzheimer’s | Dr. Sarah Marzi

What my bigger interest is, how do different known environmental causes and toxicants that increase your risk for Parkinson's disease, how do they act on different cells? And do some of their effects, despite having different primary mechanisms or being different chemicals, do maybe some of their effects on our cells converge? And if so, those would be really interesting pathways to look at. Wow. And target as therapeutic or potentially even in prevention.

Hey everybody, we'll get right back to the podcast, but I do have some very, very exciting news I want to share with you. And that is that my new book, Brain Defenders, that we've been talking about on the podcast, is now available for purchase. It is going to be published in August of 2026, but it is available now. If you want to learn more about Brain Defenders, go to the website, and that is, oddly enough, braindefenders.com. This is really empowering information as it relates to charting your own brain's destiny. Now let's get right back to the podcast. Hey everybody, Dr. David Perlmutter here on The Empowering Neurologist. We've got quite a program today. I'm joined by Dr. Sara Marzy. She is one of the world's leading scientists redefining how we think about neurodegenerative conditions, things like Alzheimer's and Parkinson's, through the lens of what we call epigenetics and gene regulation, the epigenetic control over the expression of our DNA. Dr. Margi is a senior lecturer in neuroscience at King's College in London and a principal investigator within the UK Dementia Research Institute, where she leads a lab that's focused on how genetic risk and environmental exposures converge at the epigenetic level to ultimately shape brain disease. She completed her doctoral training in neuroscience and has since built an internationally recognized research program that integrates human postmortem brain tissue evaluation, advanced epigenomic profiling, and cutting-edge disease models. Her work has been especially influential in uncovering how microglia, the brain's immune cells, translate genetic risk into disease activity. And this is, in fact, the fundamental subject of my new book, Brain Defenders. Dr. Marzi has shown how risk genes, like what we've talked about on the program, the APOE risk genes or risk genetic polymorphisms, reshape microglial behavior in Alzheimer's disease, and how things like environmental toxins such as pesticides can reprogram gene regulation in Parkinson's disease as well. What makes her work so incredibly powerful and certainly so relevant to all of us is that it moves us upstream. It helps explain why risk is expressed not just in the parts of the brain where the disease appears, but how disease risk is expressed ultimately throughout the body. Today, we'll explore how epigenetic regulation links genes, environment, metabolism, and immune signaling, and what this means for prevention, resilience, and for future therapies. I'm very excited to get this podcast underway. Let's go. Dr. Sarah Marzi, it is a great honor to be able to spend time with you this morning, or afternoon for you. How are you doing?

I'm great. And thanks so much for inviting me to your podcast, David.

Well, and I'll probably say this many times throughout our time together today, you're doing amazing, amazing work. And, you know, we're going to just jump right in. I think clearly the focus is the neurologic, neurodegenerative conditions that we are so familiar with, Alzheimer's and Parkinson's, but you're doing the real deep in the weeds kind of work. And I'd like to start off by having you unpack this notion of, you know, people talk about Alzheimer's-related genes, and we'll talk about APOE. Certainly that's an area that many of our listeners are familiar with. But, you know, there are well over 100 when we look at these genome-wide association studies. But that said, many people, I think, think about genes in terms of what they code for, which is reasonable, but you're upstream of that. You are looking at what initially turns that gene on, then allows it to do either good things or bad things. So, what drew you to this notion of genetic control, this sort of epigenetic control that you are so well known for?

Well, it's basically the causes of these neurodegenerative diseases and actually of many chronic complex diseases that we face today are really complicated. And, it's rarely ever one gene that's the smoking gun for these diseases. That happens in the minority of cases. Occasionally people will get a disease from just one mutation and one gene. But the vast majority of people that get, say, Alzheimer's disease are caused by a combination of many, many hundreds potentially, or thousands of little genetic risk factors that probably interact with the environments they're exposed to, And the lifestyle they lead and what they accumulate over their life course.

Well, you mentioned environment, and it will certainly, I think that the relationship is well-defined in Parkinson's. We'll get there. But let me let you continue. I think I interrupted. Yeah.

So what I try to understand is basically how these many of hundreds or thousands of little risk factors, what they actually do to the cells in our brain. and what then predisposes them to disease. And they often act in concert. They don't do 100 separate things. They often target maybe a handful of different processes in these cells where either they lose some protective functions that they normally fulfill, or maybe they start to do some things that are bad for our brain. And so we use something called epigenetics, which is the study of how genes are turned on and off, basically. So the biochemical switches that regulate gene expression to understand what genes are being regulated by these genetic variants that are associated with disease and what pathways come downstream of that. But also, what are— there are a number of different regulators that can orchestrate changes in many, many different genes all at once. And these are so-called transcription factors. And often we look for these so that we can understand what is really the switch that is driving a whole program in a cell that is putting it into a vulnerable state.

So again, people say, well, my dad had it, my mother had it, and, you know, are they really thinking in terms of, you know, straightforward Mendelian inheritance patterns? And I think that What you're saying here is that there's a lot involved in the actual expression of these genetic predisposition genes and that the regulatory factors of these histones and other regulatory factors are— there's a vast array of these. And perhaps it's fair to say that they must conspire. Many of them must conspire to ultimately manifest than the disease process. So it's not as if, like in Huntington's disease, for example, that we talk about a genetic mutation lead— and it can be inherited— leads to this manifestation. How does this open the door then, this idea of this multiplicity of inroads? How does it open the door to maybe first clinicians to have inroads for possible therapeutic intervention if there are so many factors that are involved?

Yeah, I think you touch upon kind of the crux of it. It's really complicated because all these factors sort of come together, and maybe you need to meet a certain threshold in a combination of environmental lifestyle and genetic factors that put you there and that then initiate the disease. And it can seem overwhelming, but I think my hope and what we're seeing is that maybe these many, many different factors act on a limited number of cellular processes. And if we understand what these processes are, and if we understand the upstream regulators, if you will, of these processes, then we can target those and reverse or prevent some of those changes that come from the genetic risk and the environmental risk.

So let's just talk about what some of these processes might be. I know, let's certainly, the augmentation of inflammatory cytokines is one.

Yes.

increased reactive oxygen species. Let's list a few of what these processes are that get the brain in trouble.

Well, there's a lot of different things. One of the features that neurodegenerative diseases have in common is the aggregation of different types of proteins. So proteins that build clumps that they shouldn't build, and they build them either in the neurons that then eventually die or in the space between the neurons. And it can be different types of proteins and different diseases. But virtually all neurodegenerative diseases have some of these protein clumps. And so the mechanisms that really clear up these clumps and should prevent them from happening, so these are things like autophagy or the lysosomal pathway, or even the function of immune cells in your brain, like the microglia, that should basically eat up and digest these protein clumps. And if any of these mechanisms don't work as efficiently as they should, then you're at a higher risk for developing a neurodegenerative disease. Yeah.

No, I was going to say, traditionally, research has focused on the neuron.

Exactly.

Even as it relates to these genetic predispositions that have been well elucidated. And I think that what your work makes clear is, you know, the neuron's interesting, but it seems to be at, dare I say, at the mercy of adequate function of the microglial cells. And maybe you can expand on that in terms of where these genes are represented or quite active, where they're being activated in comparison to neuron versus microglial cell.

Yeah, absolutely. And I think for me, this is one of the most exciting but also one of the most reproducible insights into Alzheimer's disease specifically that we've had in the last 10 to 15 years. And that's basically when you look at the genetics of Alzheimer's disease, so not the one gene that causes it, but these many hundreds or thousands of little risk factors that we talked about, they quite overwhelmingly, they fall into regions of the genome that are active and that are regulating expression of genes only in microglia and not in any of the other brain cell types. So not in neurons, for example. And so that's really suggests to us that they must have an effect in microglia. And therefore, the genetic risk that determines whether one person is more likely to get Alzheimer's disease than another, that really influences how that person's microglia function. So if you have certain genetic variants in your microglia and a combination of those, that can either predispose you or protect you from Alzheimer's disease.

Well, you, you had mentioned a moment ago that one of the defects common across certain neurodegenerative conditions, uh, is this idea of misfolded proteins and the ability that our brains have via the microglia to phagocytize or eat up these damaged proteins. So that gene regulation of the microglia then through this pathway might be something very important.

Exactly. And this is one of the pathways that really might be influenced by the genetic variants that predispose you to Alzheimer's disease. We think that it's probably more than one pathway. There's a number of different functions that are influenced, some of which are loss of function. So for example, this eating up or phagocytosing of proteins is something that we think is, or that we see is impaired actually in people with Alzheimer's disease. And then, as you mentioned earlier, these microglia also tend to produce a lot of pro-inflammatory cytokines, causing inflammation in your brain. And over time, that's something that's really bad for the neurons as well. So if they have these heavily pro-inflammatory signaling microglia around them, that again makes them vulnerable and can cause them to degenerate potentially longer term as well. So you have both A loss of things that are good and protective in these microglia, but also you gain certain functions like this inflammation that's bad for your brain.

You know, you mentioned again the aggregation of proteins, beta amyloid being certainly what is talked about most. Well, phosphorylated tau as well, but in, at least in terms of therapeutics for Alzheimer's disease. But this is downstream of the regulatory pathways that you're talking about, isn't it? So what you're exploring is upstream of the ability of the brain to appropriately get rid of things like beta amyloid?

That's a really good question. So I mean, I think the answer to what is the first event in Alzheimer's disease is still not settled definitively. A lot of people think it is the aggregation of beta amyloid. And then certainly after the aggregation or around the time of aggregation, you get cellular effects as well. So where cells around the aggregates start to misbehave or start to act in ways that aren't normal, like the inflammation in the microglia. Which one precedes which other? That's a good question. It could be that just by chance and through aging, you will start to produce more amyloid beta, and that eventually, just by chance, it will aggregate in, in some way and if you have more protective microglia, those aggregates will be consumed or broken down and will not have as detrimental an effect. Whereas if your microglia don't work as well, maybe you're more likely to get Alzheimer's. So I wouldn't say we certainly know that the microglia come upstream of the beta amyloid, but certainly the microglia are genetically the main reason why some people get the disease and others don't get it. So it's either how they lead to the aggregation, and they do seem to play a role in that as well, or how they deal with the aggregation.

Well, I definitely want to unpack more what's going on with microglia, both in terms of your— the ApoE work that you've done, as well as one study that I'd like to talk about that I sent to you that I'm certain you're aware of. But let me just— let me probe a little deeper into this beta amyloid story, because Dr. Rudolph Tanzi at Harvard, has talked about amyloid representing what we call an antimicrobial peptide, that the amyloid is produced in response to perhaps microbial infection in the brain, chlamydia, herpes simplex type 1, and that in many ways, the original formation of beta amyloid is a way of protecting the brain so that the idea of offloading all this beta amyloid might be a little bit misguided. So what have you thought about? whether you've studied it or not, but what have you thought about in terms of amyloid representing a response in the brain?

Yeah, I mean, that could well be. And I think this association between infections, maybe even peripheral infections, but certainly central nervous system infections, so infections that get into your neurons and that get into your brain, and herpes simplex is one of, yeah, the most well-known ones, and the risk for Alzheimer's disease, I think, is a really important field of study, and I think we still have a lot to learn from epidemiology there. So I, my guess would be infections play a big role in increasing our risk for Alzheimer's disease.

As I recall, oh, I'm sorry, go ahead.

And partly that could be through triggering mechanisms like the amyloid beta as a viral defense mechanism, but partly I also think they, activate and train our microglia and thereby make them more likely in the future to become more pro-inflammatory, causing this neuroinflammation that is detrimental eventually.

Hey everybody, we're going to get right back to this podcast in just a moment, but I wanted to tell you something really exciting about 40 Hz light exposure. We have actually done a podcast on this. We know that when you're exposed to 40 Hz stimulation, This is linked to what we call gamma— increasing gamma activity in the brain. In other words, balance in the brain, which is really fundamental for brain health and brain functionality. In fact, in early clinical research using what's called 40-hertz stimulation, researchers have reported signals that are consistent with a slower decline in cognitive function in some participants. So the company that we are talking about, who is in fact sponsoring our podcast today, is OptoCeutics. Their Evy— we'll call it Evy Lite— is designed to bring 40 Hz light and sound into a simple at-home routine that you can use while you're reading or watching television. I use it actually while I'm working on the computer. Optosutics makes it easier to try, so you can get 90 days of risk-free trial with support from their care advocate, somebody who's going to work with you And if it's not creating value, then you're able to return it to Optocytix for a full refund. And how can you lose? They're going to pay the shipping both ways. Quick note, Evi is a wellness device and it's not intended to diagnose, treat, or cure or prevent any disease. And this episode is, as I mentioned, sponsored by the Optocytix company. So if you'd like to try it, go to optocytix.com. And use the code PERLMUTTER26 and get $200 off your order. I think you're going to find this device to be really quite amazing. Let's get right back to our podcast. Well, you mentioned something I think that's really very important, and that is that infections can activate this whole pathway. And those infections don't necessarily need to even be in the brain. I mean, that the inflammatory cytokines can reduce be produced anywhere in the body, make their way to the brain. We know that in Alzheimer's and as we age, the blood-brain barrier is certainly less competent, less effective, and that these inflammatory cytokines having their origin anywhere can help to polarize the microglia away from being supportive to being in a different phenotype that is less able to digest away things like beta amyloid or other things that it must do.

That's what we think. And we don't quite know whether it's the cytokines themselves that get through the blood-brain barrier or whether through the cells that compose the blood-brain barrier, there's a secondhand signaling that then eventually gets to the brain and communicates with the microglia.

Yeah. I'm just thinking about our HSV-1, herpes simplex virus type 1 conversation. That original research was from your country, from England. Ruth— That's right. Yeah, and, uh, that was not widely, uh, accepted.

Not for a long time, no.

Yeah, and it's so unfortunate, you know. I mean, people, I think, tend to be down on what they're not up on. So let's go back then to the notion of ApoE. I mean, many people are getting their genome sequence, and they know their ApoE 2/2, 2/3, 3/3, 3/4, whatever their combination may be, and they're either feeling good about it or they're very concerned about it. And I think, you know, in Western cultures, what is it, 25%, 28% of people carry at least one APOE4 allele. So it's certainly the most well-recognized genetic risk factor. And you decided to do a deep dive. And let's tell our audience, if you will, how you constructed the rodent research because it's so fascinating. That's how I originally learned about your work.

Yeah, so basically APOE is the single strongest common risk factor for Alzheimer's disease genetically, and it's very different from all the other genetic variants in how big and how large the effect actually is. So if you have 2 copies of the APOE4, the risk allele, that can increase your risk for Alzheimer's by up to 12-fold. So That's a huge increase in risk. But APOE, so it's been studied for many years, but it's a very complex gene and protein, has different functions in different cells. And we thought, because we've known from our and other research that the microglia are so critical in Alzheimer's disease, we really wanted to dissect what does it do in microglia and just in microglia and not APOE communicating from one cell to another, what effect does it have in microglia? Now, microglia, uh, well, so with the diseases we study, uh, our one big challenge is we often can't study them in humans because you can't just go into the brain and study some brain cells in a living person's brain. We can, and we do, look at brains from people that pass away with diseases, and we can learn a lot about the disease there. The problem is by the time you collect those brains, it's a very late stage of disease. So it's very difficult to disentangle what might be a cause of the disease here, what processes have been going on over time, and what might be the outcome. So that's a challenging thing. We do have nice models for cells that have been developed in the last 15 years or so where we can basically reprogram cells into any cell type that we're interested in. by a technique called induced pluripotent stem cells. So the idea is you take a skin cell, for example, or skin cell biopsy, you reprogram these cells to become stem cells, and then you can turn them into any cell type that you're interested in studying, including microglia-like cells. However, the problem is when you put these microglia then in a dish, they don't really look and behave like normal microglia do. And that's because microglia are essentially extremely sensing and communicative cells. They need the other cells around them to function and look normally like microglia. So if they're on their own in a dish, how good is that model and how much are you really representing microglia? So we teamed up actually with 2 researchers, Bart de Stroeper and Renzo Mancuso, from VIB in Belgium, who had developed an amazing model where you can take these human pluripotent stem cell-derived microglia, and you put them into a brain of a mouse. And to me, it still kind of sounds a bit like science fiction, but it actually works amazingly well. So first, you deplete the mouse's own microglia with an enzyme, and then you inject the human microglia in a few different places, and they really spread out throughout the mouse brain. And when you image them, or when you look at Their function and gene expression, they look pretty much like normal human microglia. They integrate well and they interact with the mouse cells. So this for us was an amazing model where we could then study the effect of this gene APOE in the microglia. So we took basically genetically engineered microglia that only differed in their APOE genotype. So APOE2, APOE3, ApoE4, and we injected it into a mouse brain. Now, because we were particularly interested in Alzheimer's disease and the role of ApoE in Alzheimer's disease, we used an Alzheimer's disease model to do this. So this is a mouse that has some genetic mutations that cause it to develop these amyloid beta aggregates that we talked about that are one of the hallmarks of Alzheimer's disease. Now, there isn't a perfect model of Alzheimer's disease in the mouse, And this one is an amyloid model, so it only has these amyloid aggregates. In humans, you would also get hyperphosphorylated tau tangles in the neurons. This model doesn't have it, but at least it lets us study how these different ApoE microglia interact with the amyloid, how well they clear it, and really look at their gene expression and epigenetic programs to understand what might be making these ApoE4 microglia harmful. But then we also had the APOE2, and APOE2 is actually a strong protective factor. So people with APOE2, one copy or even 2 copies, very rarely get Alzheimer's disease, meaning there must be something that's really protecting them from that disease. Yeah, and so we took these 3 types of microglia and put it into this Alzheimer's mouse model and actually found a lot of striking differences in their transcriptional and in their epigenetic programs. That should suggest, in fact, that the ApoE4 microglia, as we've been talking about before, they're pro-inflammatory. They release a lot of pro-inflammatory cytokines. They're less good at clearing up proteins and other debris. So this ability to phagocytose is impaired. A third feature that we found was they seem to be less good at cellular motility and cellular proliferation. proliferation. So normally when microglia are called into action to clear up something, they need to be motile, they need to proliferate, and they need to get to the site of injury or aggregation to clear up the proteins. And based on our gene expression results, we really see that genes that promote these kinds of processes are downregulated in the E4 microglia and actually conversely

So you set about to answer the question, why are APOE4 people more at risk for Alzheimer's in comparison to APOE3, and certainly in comparison to APOE2? You took a mouse model, and did you— what did you do to the mouse's microglial cells prior to infusing The manipulated microglia. Did you get rid—

So you get rid of them exactly. You use an enzyme that basically wipes out the mouse's own microglia. And then on the slightly more technical side, normally if you inject human cells into a mouse, the mouse's immune system wouldn't tolerate this. So this is a mouse model that also has a slightly altered immune system that makes it accept rather than reject these human microglia.

So you create what are going to be microglial cells that express ApoE4, ApoE3, ApoE2. And was there not a null set as well?

There was a knockout one, exactly.

Then you put these microglial cells into the brain of an Alzheimer's mouse, right?

Correct.

And the results in the ApoE4 correlate with what we observe in humans in terms of higher levels of beta amyloid. I'm being simplistic here, in the APOE4 versus APOE3, and much certainly positive results on the APOE2, which confirms exactly what we see epidemiologically as it relates to humans who carry APOE2, 3, or 4. And what about inflammatory markers? What did that reveal?

Yeah, so we both, in terms of expression of already known pro-inflammatory cytokines. We see increased expression of these in the ApoE4 and reduced expression in the E2. And then we also tend to use some fancy statistical models where we look not at individual genes but at whole groups of genes that covary together or that build these biological networks. And again, we identify one network that is upregulated in E4, downregulated in E2, and very much contains genes that are pro-inflammatory And so suggesting that E4 mounts this stronger pro-inflammatory immune response, whereas E2 might be more anti-inflammatory with a reduced immune response.

So now— oh, go ahead. Now I'm just going to say, so as I was reading your study, you know, I was trying to get to the point where, as one would expect, okay, so what can I tell people is important here? And, you know, aside from the fact that other aspects of our lifestyle tend to, literally inflame these pathways. You studied vitamin D receptor functionality, and I think there may be some traction, some translatable traction.

Agreed.

With respect to that. So let's unpack that a little bit.

Yeah. So that's, for me, the most exciting result of the study because we started to ask at that point, we see all these big gene expression changes, are there certain drivers, triggers, These transcription factors that I mentioned earlier in the podcast that are driving these different gene expression programs, either protective programs in the E2 or harmful programs in the E4. And one of the strongest results that we got for these enrichments is that in the E2, in the protective microglia, there seems to be enhanced signaling via binding to the DNA of the vitamin D receptor. Now, vitamin D receptor is really, really interesting. It's a transcription factor, which means it's a protein that binds DNA and thereby regulates the expression of a big group, whole arsenal of different genes. And it's activated, as the name suggests, by vitamin D, which in humans we produce through sun exposure, or we can also supplement in our diet. And it's doubly interesting because it's well known that vitamin D deficiency, so if you don't have enough vitamin D, is a powerful risk factor for Alzheimer's disease. And in fact, even once you already have Alzheimer's disease, vitamin D deficiency can make your course of disease faster and your progression worse. So clearly vitamin D epidemiologically is really important in Alzheimer's disease. So we were really excited to find this really strong molecular link That suggests vitamin D binding and vitamin D receptor activity is driving genetic gene expression programs that might be protecting vitamin D, FOE2 carriers from Alzheimer's disease. And now if we look at the function, so what, where, which kind of genes are regulated by vitamin D receptor? These tend to be, well, they also, it has effects in bones and bone density, but in immune cells like the microglia, it really clearly has effect on immune functioning, and in particular it has anti-inflammatory effects. So most of the genes that are downstream and that are being regulated by vitamin D receptor will drive an anti-inflammatory phenotype in microglia or other immune cells. So that really gives us a piece of the puzzle where anti-inflammatory signaling is something that occurs in ApoE2 microglia, and that we think is probably protective. And at least some of it, maybe a lot of it might be driven by the enhanced binding of vitamin D receptor. So that, as you said, gives us something really tangible, something that we might be targeting both through vitamin D supplementation, for which the trials in the past have not been entirely conclusive, but also by understanding what is required for the signaling of vitamin D receptor signaling. So yes, it needs to bind vitamin D. So you need to have the vitamin D available. And to me that means, yes, in winter you should probably be taking some vitamin D supplements. I certainly do. But the question is, is that enough or does it— are there other biological steps required? Because sometimes it's not as simple as giving one molecule and that will solve everything. And so for example, vitamin D receptor, heterodimerizes with the retinoid X receptor before it binds the DNA? Or are there— we're investigating that functionally now, actually. Are there other factors that affect how well it can bind the DNA in presence of having enough vitamin D already?

Hey everyone, we're going to get right back to this podcast, but first I wanted to share some information about olive oil. You know that I love olive oil. So here's a good tip. The most delicious olive oil that you can get is olive oil that comes fresh from the farm. This is when the olive oil is at its peak flavor and all the nutritional good things that we talk about, the polyphenols, etc., are really at their peak. When you buy your olive oil at the supermarket, you are getting an inferior product because they have been sitting on the shelf for months and they basically grow stale. And that's why I've been getting my olive oil direct from small, award-winning family farms really around the world, thanks to a fellow named TJ Robinson. I've done Instagram Live with him in the past. He's known as the Olive Oil Hunter. He's sort of the Indiana Jones of olive oil. He is able to source farm-fresh oils that are vibrant. They're healthful. They're grassy. They're incredibly delicious. And you can use them on whatever you want— on salads, fish, vegetables. Meat and even like I do on scrambled eggs. So if you want to taste the difference that freshness makes, TJ is going to send you a full-size bottle, which normally will cost you $39. He's going to send it to you for free if you pay them just a dollar. That will cover shipping. And that's how he then is able to introduce all of you to this Fresh Pressed Olive Oil Club, a club that we are members of. So there's no commitment. Just send him a dollar. He'll send you a full bottle of olive oil. You just want to go to getfresh389.com. That's getfresh389.com, and he'll make it happen for you. Let's get right back to our podcast. Well, you know, we do know that there's downregulation of the NLRP3 inflammasome with activation of the VDR. And so one wonders a couple of things. First of all, you know, people are getting their genome sequenced and might find that they have variants of VDR, a vitamin D receptor functionality based on their genome, looking at their genome. The other thing is one wonders if through the mechanism of reducing inflammation by vitamin D binding to its receptor, if we shouldn't be concerned. And you mentioned how the interventional trials with vitamin D were not that meaningful, to be kind. But one wonders if there couldn't be a convergence of other issues here in terms of inflammation that, because of our metabolic disarray these days based on the foods that we're eating, that sort of antagonizes or offsets any up— any benefit of the whole vitamin D pathway anyway. So, you know, and I think it gets back I had a thought when you were talking about APOE4 and beta amyloid vis-à-vis our discussion that amyloid may have been produced as a response to infectious agent, that perhaps because of— in our— that carrying the APOE4 allele, which persists in our modern world, there must be some advantage to it, or why would 28% of humans still carry the APOE4 allele. If it was across the board detrimental, then why did it— why did it lose its positivity over time? And one wonders if it's, uh, in juxtaposition to how our metabolic health has become so disrupted in, in our modern world. It's, it's just a thought. I think probably should leave it, leave it there. Um, I have to tell you one last thing, and that is— and then we'll move on. As you're describing this work, this recent research, what's going through my mind are the slides I made based upon your research from a presentation I gave a few weeks ago. And there's— and I mentioned this to you earlier before we went on the recording. I have this animation of the injection of these microglial cells into the brain of a cartoon character, so it's less distasteful, a cartoon character rodent. And I'll, you know, as I think about that, I'm visual, so that's how I understand your research. Yeah. Let's move on to some really exciting work that you've done with paraquat in the context of Parkinson's disease. First, what would motivate you to want to look at paraquat in that context, in Parkinson's?

So actually, it's been mostly rotenone so far, but we are starting some work with paraquat as well. So with Parkinson's, very generally, it's a much less heritable disease than say Alzheimer's or than some other common complex chronic diseases that we face. So if you, and you can study this by looking at twins. So you take identical twins and non-identical twins and you look at how often does a disease co-occur together in the identical versus the non-identical ones. And then you can put a number on it and basically quantify how much of it is genetic. And when you do that for Parkinson's, it's only about 20 to 30%, uh, genetics, meaning likely the environment that you're exposed to plays a much bigger role. And in fact, from epidemiological work, uh, we have a lot of good evidence for this, for very specific exposures that have happened. I know, uh, when looking through your podcast, you've had Ray Dorsey as a guest several times so far.

3 times. He holds the record. That's because he's doing— he puts out such New and exciting work all the time.

Exactly. And yeah, so I completely agree. And I think it's in the last 10 to 20 years that genetics has been a real big focus, both of research and of pharmaceutical companies, because genetic targets are more tractable or more likely to survive clinical trials. But I think for Parkinson's, that's not been a good thing because I think the environment plays such a big role. And so really, my— other than understanding genetic risk, as I've just talked about in Alzheimer's disease, we also want to understand environmental risk. And we want to understand what do these chronic systemic, often low-grade exposures over many years and sometimes even decades, what do they actually do to the relevant cells that make them vulnerable to neurodegenerative disease? And here again, these epigenetic mechanisms, these chemical switches of gene regulation, are potentially a really important mechanism because they can retain a memory of this exposure and they can basically be changed gradually over time, but cumulatively so that eventually harmful gene expression programs are there.

I want our viewers— I apologize for the interruption, but what you just said is so profound and so very, very important. Because it's very difficult to conceptualize that these events are slow and acting over decades. We don't see a smoking gun. It's not as if, well, you know, you— if people wear a helmet and are in a car, a bike accident, they're gonna have, you know, it's quick. You see, well, I was wearing a helmet and I survived. Yeah. In this case, and yet you're diving deep into the weeds on this and you've Made some incredible discoveries.

Yes. So basically, for a starting point, we focused on this pesticide called rotenone. And rotenone is, it's not that commonly used. It's in fact banned as a pesticide in the US and the European Union. It's still used in some places in the world. And there is strong epidemiological evidence that it increases Parkinson's risk quite substantially in humans. Um, but what's also for us important with rotenone is that we have good animal models for this. And so for this, we collaborated with Tim Greenemeyer and Emily Rocha at the University of Pittsburgh, who have a pretty amazing rotenone model. So this is a rat model where they inject, inject rats with low doses of rotenone over 2 to 3 weeks. Um, in our case, it was a 3, 3-week exposure. And then these rats basically develop Parkinsonism pretty comprehensively. So the same type of neurons that die in, in the people with Parkinson's die in this rat model. They get cellular dysfunctions that mimic what we see in Parkinson's. They get the same protein aggregates that people with Parkinson's get, and they develop the motor symptoms as well over time. And it's It's an amazing model because it's an entirely environmental model. So most animal models that we work in with in neurodegeneration, they use human genes that are put into a mouse or they need some form of genetic manipulation, whereas this is purely environmental. And so we teamed up with Tim and Emily to look at epigenetics and gene regulation in this rotenone model of Parkinson's disease. And we also wanted to understand why it so specifically affects one type of neuron and causes this type of neuron to die. So in Parkinson's, a type of neuron called dopaminergic neuron dies. It's called this because it primarily uses a neurotransmitter called dopamine. And these dopaminergic neurons that die in Parkinson's sit in a very specific area right in the center of our brain, in the midbrain, in an area called substantia nigra. So this is the neurons that die. You do see effects on functioning across the brain, and different regions are impaired. So for our rat study, we looked specifically at 2 different regions, the substantia nigra, which is where the neurons die, but also the motor cortex. Here you wouldn't get neurons dying, but you would see potentially signaling dysfunctions, things that affect your motor system, which is the primary symptom of Parkinson's, is impairment in your motor control. And so the striking thing was, yes, we see a number of different common things between the 2 brain regions. And a lot of those affect mitochondrial functioning. So mitochondria are kind of like the energy machines in our cells that give our cells energy. And rotenone, this pesticide, specifically targets and impairs mitochondria. So we would expect to see that basically in any cell that we would look at that's exposed to rotenone. But what was more striking is that there were actually profound differences between the substantia nigra and the motor cortex. And uniquely in substantia nigra, we actually see quite a striking upregulation of immune-related pathways, probably coming from the microglia, including an upregulation of an immune response called— so we see an upregulation of the complement cascade and particularly the C1q.

I want to get into that. I want to talk about your refining of the C1q activation. We'll get back to the epigenetics in just a minute, but we know that when synapses are labeled with C1q, that microglia identify them as being what they want to digest, that this is an early life mechanism labeling synapses for the appropriate synaptic pruning of the brain. But in this case, we see that C1q labels these synapses, and then the microglia that are activated digest these away. So we get back to both this being a, if you will, synaptopathy, and also the fundamental role of microglia.

Exactly. Absolutely. So that's exactly right. So it's been shown in a number of different neurodegenerative diseases that microglia can and do phagocytose, eat up synapses, and that's probably not a good thing later in life. And that labeling happens through C1q. So there's some really exciting work on that from Beth Stevens and Soyeon Hong that have shown this as the signaling and phagocytosing pathway in synapses. So we see the C1q upregulated in bulk brain tissue. Where that exactly comes from, maybe some of it from the microglia themselves, maybe some of it from the cells that are being targeted, but it is extremely striking. So the difference is huge. It's not just a small change in the expression of this. And we did in fact also do some imaging. So looking in the tissue itself, and we in fact see that in the rotinone condition, you have activated phagocytosing microglia that are characterized by the upregulation of a marker called CD68. So this means these are reactive phagocytosing microglia. So you're exactly right. So they're probably phagocytosing different things. So it might be synapses, and it might also be the early stages of protein aggregation happening, because we know that does happen in the, in the rosinone model as well.

Hey everybody, we're gonna get right back to the podcast, but I have a very important message for you. You know, why do we exercise? Why do we pay attention to how much sleep we get and the quality of the sleep we get? Why are we looking at our diets? Because ultimately, we are sending information to our genome. We are affecting how our genes express themselves. I like to know about my genome. That's why I'm a real fan of the 3x4 Genetics Test. Why? So you can learn about your genetic predisposition. You know, in my world, as it relates to the brain, there are some genes that in fact are associated with increased risk for neurodegeneration. A lot of people know their APOE status, for example. But I want to be super clear that this is not a genetic determinant. Maybe it's a genetic predisposition, but what you're seeing here happening— these are our lifestyle choices that actually change our gene expression. You should know about your genome. You should get the 3x4 Genetics test. We'll talk about how to do that in just a moment. But there is a health program that they offer you as well when you get your genome sequenced through 3x4 Genetics, it's not just getting this information, but they're going to tell you what to do with that information. So, I think that's really very important. So, head on over to 3x4genetics.com/DrPerlmutter. That's D-R Perlmutter. They have a special offer for you and all of you podcast watchers. Let's get right back to the program.

Well, so I'm going to summarize then. We know that rotenone is a mitochondrial toxin and like paraquat, like MPTP, uh, and to a certain degree glyphosate. But what the supposition was, was that for whatever reason, uh, with a mitochondrial toxin on board, that the dopaminergic neurons in the substantia nigra were hypersensitive because they were— perhaps because they were so active. And that, uh, men might demand, because they're involved in more motor activity, I'm going off on a limb here, and therefore their substantia nigra are more active. That's why men are 2 to 1 more likely to get Parkinson's. I don't know the answer to that yet, but we should maybe kick that one around. But what you're telling us is it's more than a mitochondrial issue, because yes, the mitochondria in the motor cortex of the, of the rodent were involved, as were the mitochondria in the substantia nigra, but the epigenetic changes took place only in the substantia nigra. Yeah. Reason.

The immune-related changes. Yeah. And actually, we do see other interesting changes in the motor cortex as well that suggests that the functioning of synaptic proteins is a little bit awry, going in all sorts of directions. And that in some ways mimics what we know functionally about the disease, where it's more a functional impairment, a signaling impairment that's happening in the motor cortex, whereas the Death of neurons, that's something that really more exclusively happens in the substantia nigra. And it may relate to this immune upregulation of the microglia.

I just had chills because a thought came to mind. I've been puzzling for a long time why male— why the male-to-female predominance in Parkinson's, right? And why the female-to-male predominance in Alzheimer's?

In Alzheimer's.

Maybe we'll discuss that. One of the reasons that females— one discussion has centered on the C1q pathway, on the labeling of the synapses that happens suddenly with the decline of estradiol in the perimenopausal time. One wonders if those neurons are, uh, less protected in men for lack of adequate or a given level of estradiol, and therefore, uh, they are more susceptible. Don't know, but I'm gonna have to think about that.

I don't know either.

But, you know, uh, this is Lisa Mosconi's work with respect to the, uh, you know, when that— when is the timing of estradiol best. So I think we'll have to leave it for another time. Um, but let's, let's take a step back then and I did mention erroneously paraquat. Paraquat works in a— you've not done the epigenetic work, or maybe you have, but paraquat is also a— is still a widely used, at least in America, herbicide that is a mitochondrial toxin. So what are the implications in terms of your work as it relates to paraquat?

So paraquat's Affects different pathways as well. We think actually the primary mechanism of action is more through generation of reactive oxygen species and downstream DNA damage. Um, we are planning some paraquat experiments as well because actually, rather than just rotenone itself, what my bigger interest is, how do different known environmental causes and toxicants that increase your risk for Parkinson's disease how do they act on different cells in the brain? And do some of their effects, despite having different primary mechanisms or being different chemicals, do maybe some of their effects on our cells converge? And if so, those would be really interesting pathways to look at and target as a therapeutic or potentially even in prevention.

I think it was probably you who gave the terminology that these toxins write them— they write themselves into our epigenome. Is that your language?

I'm not sure. Well, whatever. I may have written that some time ago.

It's a very interesting way, right?

I often use the terminology that they leave an epigenetic memory in our cells.

They insinuate. And I think then, you know, that explains why it's a long-term, slow— You know, slow sizzle that happens that ultimately has clinical manifestations. I want to go to your work on histone patterning, changes in histones in the Alzheimer's brain.

Is it?

And again, for the viewers, histones are gene— correct me if I'm wrong, how would I explain it— gene expression modulators. They determine which genes are turned on and off. And your work has identified an array of gene expression changes, histone changes seen in a part of the brain, the entorhinal cortex, that is characteristic of Alzheimer's. Correct?

Roughly right. So histones are very common proteins that everyone has in their cell nucleus, and it's what the DNA is wrapped up around in the nucleus of our cells. And there are 2 main benefits of this. One is that our genome and our total DNA in each cell is actually quite long. It's a huge number of molecules, and wrapping them around these histones makes it much more compact and makes it fit in the nucleus. But also, these histones can be chemically modified, so addition of different chemical groups to the histones. And these chemical modifications basically alter how tightly wrapped the DNA is around the histones. And simply, you can think about very tightly compacted, wrapped DNA is just inaccessible and therefore is quieted down and can't be transcribed because the mechanisms that produce gene expression and downstream proteins need access to the DNA. So if you completely wrap up the DNA very tightly, they don't have access to it and it can't be transcribed. And vice versa, if the DNA is not as tightly wrapped, then it's accessible and it can be transcribed. And so the histone modifications basically change how tightly

And that's important because whichever genes are turned on and/or turned off has physiologic impact.

Absolutely. Absolutely.

Might manifest as increased inflammation, pave the way for Alzheimer's, or might not. So it's really the gene regulation involving the histone switches that is the area of your interest that you've studied.

Exactly. And so we did a big study of this now 7, 8 years ago, 2028— 2018, sorry. It's 2026 now, just about. And we found that basically this histone acetylation, which was the one modification that we looked at, and it's interesting because it's dynamic and it responds to environment, but also it's changed by genetic risk. Factors. And we showed that there are really widespread changes in this histone modification in the brains of people with Alzheimer's, and specifically in the brain region that is first and most severely affected by the disease. And that was quite novel and critical, and we were surprised how strong and how widespread these changes were. And we could also link them to some of the known genetic risk factors. But back then we did the study on just bulk brain tissue, so meaning you mix up all the cells in a sample and you generate a profile. But really, as would hopefully shine through from our discussions here, it's really critical what's happening in different cells. And our research and other research has shown microglia are really important, but it's also really important to understand what are different cells doing Why are neurons degenerating? What are the microglia doing? Maybe what are some of the other cells that are around doing as well? And so we are basically, we've been doing 2 things to understand better what's happening in different cell types. One is just using clever computational methods where we take reference profiles. So we know what a standard microglia looks like epigenetically. We know what a standard neuron looks like. And so vice versa, different cell types. And we use that, so going back to our bulk brain samples to then basically estimate how much of each cell type was in each bulk sample based on what a standard cell looks like. And so when we do that, and we published this earlier this year as well, what we see is that actually in Alzheimer's disease, 2 cell types have really widespread changes epigenetically. And one is the microglia, as we would've expected based on the genetics and based on a lot of our other work. But what surprised us a little bit is actually that even more strikingly, oligodendrocytes show really vast changes in the epigenetics. And oligodendrocytes are basically the insulating cells. So they are quite fatty cells that wrap themselves around the axons, which are the signaling tracts of the neurons, and thereby insulate them. And they act very much like a cable insulation would act on an electrical cable. And—

It's interesting the oligodendrocytes are getting attention now. The past few years, it's been the microglia, right? Suddenly, everybody's interested in the microglia. But now, oligos. We didn't really— we, you know, they wrap themselves around the neuron to insulate it. That's pretty much all they do. Well, no, there's got to be more to this story. We, you know, we pigeonhole ourselves. It's—

I agree. And I think with— so, while the genetics, I would say, very clearly points to the microglia and not the oligodendrocytes, when you look at these Alzheimer's, late-stage Alzheimer's brains, there are a lot of— the most changes are happening in the oligodendrocytes. But we know that oligodendrocytes can functionally alter in Alzheimer's disease as well. So loss of myelination, loss of this insulation is something that is actually seen in Alzheimer's disease. And I think what's important is that with these complex diseases, probably the causes are more systemic and complex. And it would be surprising if it was just a single cell type and a single process that would being affected. I think it's a lot of things acting in concert. And my hunch is that likely if we want to treat them, we're going to have to basically administer a cocktail of things that treat different aspects, functional and causal aspects of disease.

So what you're saying is that to be able to modulate the activity of those histones, you would need a cocktail that specifically targets those ones that you, uh, that you and colleagues agree might be the ones most involved in a detrimental way and try to shut them down versus activate ones that are more salubrious?

Yeah, so my hope is that, um, again, we are not there yet, but that maybe we could identify, for example, transcription factor networks that are dysregulated in the oligodendrocytes as well, and that then in turn would give us potentially meaningful targets that drive big regulatory programs that we think are functionally connected. So that are causing these cellular malfunctions that are causing oligodendrocytes to demyelinate the axons, which is probably something that's really bad for the neurons and something you want to prevent.

So there has been work over the years before you made it very clear that there's an array of histones that are involved to consider the use of histone deacetylase inhibitors like valproate. to kind of globally affect histones in general. But that's not very precise based on what you're telling us.

I agree. So that's always surprised me in a way. That seems like quite a sledgehammer approach where you alter histone acetylation or histone modifications across the whole genome. I don't know, it seems like that could be risky because you're affecting many, many, many different genes. On the whole, these are pretty, they act quite ubiquitously and indiscriminately across the whole genome. Whereas I think it's more about specific locations in the genome, specific programs that drive very specific functions and processes in these cells. And yeah, my hypothesis is basically that they're regulated by a handful of relevant transcription factors. And that's probably where we have the best chances of going in pharmacologically and mechanistically in targeting and preventing disease.

So you're looking at whole brain homogenates, or, or at least homogenates from particular areas of the brain. Not exactly a diagnostic tool for a patient in your office. So that said, could you foresee in 5 years or 10 years the idea that using machine learning for evaluation of exosomes Could perhaps be correlated with what you have already discovered in terms of being able to fingerprint these illnesses?

You're saying basically in blood samples or cell—

Yeah. Blood, perhaps CSF, exosomes.

Yeah. I think there'll be many biomarkers, even peripherally, that we'll identify. So, we're getting better and better protein biomarkers already, protein biomarkers of aging, protein biomarkers of disease. Exosomes is another one. It all depends technically on the stability, how well you can collect, isolate them. There's been some really exciting novel work looking at cell-free DNA methylation. So this is another epigenetic mark. And basically as neurons die in your brain, some of their DNA gets released into the intercell, so basically just into the brain. And a lot of it is cleared up, but not all of it. And some of it makes its way into the bloodstream and can be picked up there. And because epigenetics is very cell type specific, if you pick up those specific epigenetic marks that are only found in neurons, you can say, yes, this piece of DNA came from a neuron. So that's quite an exciting prospect if we— and I certainly think in that context that machine learning, as you mentioned, is going to be really critical to identify from limited patchy information because you're not going to get a whole genome of cell-free DNA or a lot of exosomes in the blood. But from these limited sources, by having really good references and very good machine learning algorithms, I think we'll get better and better to actually be able to detect these diseases earlier in peripheral samples like the cerebrospinal fluid or blood.

As opposed to something more global like neurofilament light?

Yeah, so that's a good marker for neurons dying in general, but it doesn't tell you what disease is causing the dying of the neurons because it just is, it's found in samples of anyone with any neurodegenerative disease.

There's a lot of discussion these days about our lifestyle choices being able to influence our epigenetics as it relates to therefore controlling gene expression. How do, where would a guy like me go in a discussion of, based on Dr. Sara Marzi's work, How would lifestyle factors be considered as we look at this vast array of histones that are active or inactive as it relates to a disease like Alzheimer's? In other words, what is our ability to modify gene expression through this pathway based upon the choices we make each day?

So I don't think that's quantifiable yet, and I don't think we've done enough research to really pinpoint this intervention will do exactly this. There are increasing studies in that context. So epidemiologically, we of course know some of the lifestyle factors that might relate to Alzheimer's disease risk. So certainly exercise and activity is something that's quite protective. And I know people are now increasingly studying the effects of exercise on your body, more often on blood and muscle than on the brain. But, you know, we're slowly getting there. And that's something that you might be able— effects on the brain is something you might be able to study in a mouse model, for example. I think it'll take a long time to really mechanistically and in a fine-grained way dissect what exact epigenetics is being altered by which lifestyle factors, how does it exactly relate to disease. But I'm very hopeful and I'm excited about this kind of research and what we can learn about it, both in terms of I don't think you necessarily need the epigenetic evidence to say a lifestyle factor is protective or harmful. I think that we can get from the epidemiology. But once we understand the mechanisms and effects, that A, can give us better chances of finding pharmacological interventions, and B, can make us look for, do different factors that act on disease risk converge on a limited number of pathways? What are these pathways? And how, in terms of lifestyle factors and pharmacological intervention, can we target those pathways?

So I'm going to say that your research is what we call foundational. I mean, it really is so important for everything that we are thinking about as it relates to neurodegenerative disease. That said, what's next for you?

What's next? So I'm, as I mentioned before, really excited about the vitamin D finding. And in fact, we're doing quite a lot of work again with Bart de Stroep and Renzo Mancuso, but also just in our lab in dissecting that more functionally. So do, based on the gene expression results, do the microglia exhibit those different functions? Prove that a bit more in a controlled system, but also try and manipulate the vitamin D receptor genetically and pharmacologically to see ideally if we can, for example, rescue APOE4 microglia to make them more protective, like the APOE2 microglia as well. That's something we're really excited about. We continue to work in the Parkinson's space, and we've been doing work on human postmortem Parkinson's samples as well. So what I would be really interested in seeing is, do known environmental causes of Parkinson's like rotenone, do they cause patterns that overlap what we see in just sporadic Parkinson's cases where we wouldn't know what they've been exposed to, like different environmental and lifestyle factors over a whole life course. For normal people, this information is very, very difficult to collect or to quantify in any way. So if we can find things that converge between controlled models like the rotenone rat, or we can do this in cell systems as well, and also what we see in the brains of people with Parkinson's disease, that would give us more confidence that What we see is really disease relevant.

Wow. I know that Ray Dorsey, you mentioned him earlier. I mean, he's really focused also on trichloroethylene.

Yes.

And it's, it's a way that it made its way into the groundwater. It explains—

Absolutely.

Higher rates in— here we have in the States, uh, uh, Fort Lejeune, uh, where—

Exactly.

Very high rates. So I'm sure that's going to be studied. Um, I want to tell you that I thought quite a bit about our time together today prior to spending time together. What am I going to talk to you about? I wrote a list of questions which I didn't get to, unfortunately. But I think I really, again, want our viewers to know that it's your kind of work that allows us to really gain an understanding ultimately of the translational part, i.e., what do we tell people in the clinic? And it starts with this, as I mentioned earlier, foundational work that you're doing. And I'm very, very appreciative. And I know you're as busy as can be, and to take this time off to spend it with me and all of our viewers, I just want you to know I'm very grateful.

Oh, well, thank you so much for having me. I mean, it's been a really, really interesting discussion. And I'd be really interested in hearing your opinions on metabolic nutrition lifestyle factors that you would think about recommending to your patients. But maybe I can come back another time in the future.

That's a date. Consider it done. I'm going to tell you what I was thinking, and that is, um, because it's really— this is more about you, but the thought that went through my mind as we were having our discussion of the vitamin D receptor, um, and the, uh, and, and issues there related to inflammation via the NLRP3 inflammasome, that again, it's a convergence there, you know, the metabolic issues in the human body are influencing all that's going on at the level of microglia, helping or not. But generally, when we are having issues with elevated blood sugar and inflammation in our bodies, that's tending to polarize these microglia to being that M1 phenotype that cannot phagocytize, as your work so elegantly demonstrated in the APOE study. So I think, you know, we look at many straws on the camel's back, and I think vitamin D is a big straw, But it can be, you know, worsened or offset based upon these other inroads, based upon, I believe, the importance of our daily choices. So that's what I was thinking about. But we'll deal with that next time when you've done all that vitamin D work. I'm so excited about it. So thank you again. And I sure hope I get to meet you at a conference in person someday soon.

That'd be great. Thank you so much, David.

Bye for now. I think it's clear that Dr. Marzi's incredible work powerfully reinforces a central theme we've been discussing many times on this podcast, and certainly in the books I'm writing, that our genes are not our destiny. That what matters is how our genes are regulated, how they are expressed. And the factors that control that include our environments, our immune signaling, our metabolism, and powerful influence of our lifestyle choices. These signals shape microglial behavior over time. Dr. Marzi's research gives us a molecular framework for understanding why factors like pesticide exposure, inflammation, and metabolic stress increase risk for neurodegenerative conditions, and why interventions that support immune balance and cellular resilience may have such incredible impact. So I'm really grateful to have had this opportunity to speak with Dr. Sara Marzi. I've been following her work for a long time. I'm thankful that she joined us today and very thankful for all of her pioneering work, which is really at the intersection of epigenetics, meaning controlling our gene expression, the microglia, the brain's immune system, and ultimately brain health. And this is exactly the kind of science that empowers us to think differently and therefore very proactively about protecting the brain, where technology enhances insight without replacing wisdom. So thank you for joining me here on The Empowering Neurologist and for helping us share in what appears to be a brighter future. We'll be back soon. I'm Dr. David Perlmutter. Bye for now. Bye.