The Peter Attia Drive - #70 - David Sinclair, Ph.D.： How cellular reprogramming could slow our aging clock (and the latest research on NAD)

00:00:00.000 Hey everyone, welcome to the Peter Atiyah drive. I'm your host, Peter Atiyah. The drive

00:00:10.880 is a result of my hunger for optimizing performance, health, longevity, critical thinking, along

00:00:15.940 with a few other obsessions along the way. I've spent the last several years working

00:00:19.660 with some of the most successful top performing individuals in the world. And this podcast

00:00:23.620 is my attempt to synthesize what I've learned along the way to help you live a higher quality,

00:00:28.360 more fulfilling life. If you enjoy this podcast, you can find more information on today's episode

00:00:33.000 and other topics at peteratiyahmd.com. Hey everybody, welcome to this week's episode

00:00:43.360 of the drive. I'd like to take a couple of minutes to talk about why we don't run ads on this podcast

00:00:48.580 and why instead we've chosen to rely entirely on listener support. If you're listening to this,

00:00:53.820 you probably already know, but the two things I care most about professionally are how to live

00:00:59.060 longer and how to live better. I have a complete fascination and obsession with this topic. I

00:01:04.760 practice it professionally and I've seen firsthand how access to information is basically all people

00:01:09.920 need to make better decisions and improve the quality of their lives. Curating and sharing this

00:01:15.100 knowledge is not easy. And even before starting the podcast, that became clear to me. The sheer volume

00:01:20.380 of material published in this space is overwhelming. I'm fortunate to have a great team that helps me

00:01:25.800 continue learning and sharing this information with you. To take one example, our show notes are in a

00:01:31.600 league of their own. In fact, we now have a full-time person that is dedicated to producing those

00:01:36.260 and the feedback has mirrored this. So all of this raises a natural question. How will we continue

00:01:42.440 to fund the work necessary to support this? As you probably know, the tried and true way to do this

00:01:47.980 is to sell ads. But after a lot of contemplation, that model just doesn't feel right to me for a few

00:01:54.200 reasons. Now, the first and most important of these is trust. I'm not sure how you could trust me if I'm

00:02:00.400 telling you about something when you know I'm being paid by the company that makes it to tell you about

00:02:05.080 it. Another reason selling ads doesn't feel right to me is because I just know myself. I have a really

00:02:11.280 hard time advocating for something that I'm not absolutely nuts for. So if I don't feel that way

00:02:16.460 about something, I don't know how I can talk about it enthusiastically. So instead of selling ads,

00:02:21.580 I've chosen to do what a handful of others have proved can work over time. And that is to create

00:02:27.420 a subscriber support model for my audience. This keeps my relationship with you both simple and

00:02:33.480 honest. If you value what I'm doing, you can become a member and support us at whatever level

00:02:39.460 works for you. In exchange, you'll get the benefits above and beyond what's available for free.

00:02:44.300 It's that simple. It's my goal to ensure that no matter what level you choose to support us at,

00:02:50.040 you will get back more than you give. So for example, members will receive full access to the

00:02:57.520 exclusive show notes, including other things that we plan to build upon. These are useful beyond just

00:03:03.800 the podcast, especially given the technical nature of many of our shows. Members also get exclusive

00:03:09.960 access to listen to and participate in the regular ask me anything episodes. That means

00:03:16.340 asking questions directly into the AMA portal and also getting to hear these podcasts when they come

00:03:22.280 out. Lastly, and this is something I'm really excited about. I want my supporters to get the best

00:03:27.280 deals possible on the products that I love. And as I said, we're not taking ad dollars from anyone,

00:03:32.120 but instead what I'd like to do is work with companies who make the products that I already love and

00:03:36.960 would already talk about for free and have them pass savings on to you. Again, the podcast will

00:03:43.880 remain free to all, but my hope is that many of you will find enough value in one, the podcast itself

00:03:51.300 and two, the additional content exclusive for members to support us at a level that makes sense

00:03:57.360 for you. I want to thank you for taking a moment to listen to this. If you learn from and find value

00:04:02.440 in the content I produce, please consider supporting us directly by signing up for a

00:04:07.520 monthly subscription. I guess this week is professor David Sinclair. Now that name may sound familiar to

00:04:12.860 some of you because I've already interviewed David. In fact, his first interview appeared podcast back

00:04:18.300 in November, I believe in 2018. I wanted to bring David back on the podcast for several reasons. First

00:04:23.080 of all, he's always interesting to speak with, and we spend a lot of time speaking off podcasts and I want

00:04:28.720 to be able to share those discussions with people. Second, he has a new book that is coming out. In

00:04:34.020 fact, we have timed the release of this podcast to coincide with the release of that book, which is

00:04:37.680 tomorrow, September 10th. And third, we wanted to revisit some of the ideas around NAD, NR, nicotinamide

00:04:48.040 riboside that is, and longevity. These are still to this day among some of the questions I get asked

00:04:53.580 most about. And even though truthfully, I don't believe this is even at the top three level, you

00:04:59.880 know, most interesting questions in longevity. For whatever reason, people want to know all about

00:05:03.740 it. And therefore, I wanted to provide a little bit more insight into that. So as a bit of a

00:05:07.580 refresher, David's a professor in the Department of Genetics at Harvard Medical School. He's the co-director

00:05:12.720 for the Biological Mechanism of Aging program, or it's a center actually. He's best known for his work

00:05:19.740 in understanding why we age, how to slow its effects. We talk in the very first podcast about

00:05:24.500 his role in the discovery of sirtuins and the treatment thereof. Let's see. I think, you know,

00:05:31.200 you can go back and listen to his bio from that. Let's talk about what we talked about here. We

00:05:34.760 talk about this idea of the information theory of aging, and this is really a big part of what

00:05:40.120 David's book is all about. And I knew that David was going to be interviewed by a lot of people for

00:05:45.200 the book. So I felt that it was probably going to be most helpful to listeners if I interviewed him

00:05:52.020 on some of the more technical details of that book, which deals with, again, this information

00:05:56.700 theory of aging. Basically, what is the clock that determines our aging? What does it look like? And

00:06:01.760 perhaps most importantly, can it be manipulated? There are lots of other topics that are covered

00:06:05.800 in his book, such as the ethics of delaying death to significant degrees. You know, what would it

00:06:11.260 mean to humanity if we could live for, you know, hundreds of years? Those are very important

00:06:14.720 questions. I don't touch on any of them in this interview, and I suspect you'll hear a lot about

00:06:18.000 that in some of the other interviews. So if this topic is of interest to you, one, I recommend you

00:06:21.900 buy the book. I really enjoyed it, and I learned a lot. And that's saying something because I don't

00:06:26.860 really learn a lot reading books about aging, unfortunately, anymore. The books are written

00:06:31.000 at such a low level, but that's not the case here. This is really good. Secondly, I think the

00:06:35.980 discussion we have around NAD, NR features some really up-to-date stuff, including a couple of

00:06:42.080 papers that were published in the weeks leading up to our talk. And so again, I don't think this

00:06:47.000 is the final word on the subject, but I do think that you'll come away from this with an even more

00:06:51.900 nuanced appreciation for the potential benefits, if there are any, of supplemental agents that pose

00:06:59.000 a promise of increasing NAD. So without further delay, please enjoy my second conversation with David

00:07:05.840 Sinclair. Hey, David, thanks for swinging by. Thanks for having me back. Congratulations,

00:07:12.540 first of all, on the almost-released book. Well, thank you. That's an interesting time,

00:07:18.260 just waiting for the thing to drop September 10th. Yeah. Yeah. Well, by the time people hear this,

00:07:23.300 it will be dropping. So I kind of want to just start with the book because I think I said this to you

00:07:27.900 last time when we were together maybe a couple of months ago, but once you start recording a podcast,

00:07:33.260 you end up listening to podcasts a little bit differently because now you're sort of thinking

00:07:38.060 about it through the lens of the interviewer as well. And similarly, once I think you're trying

00:07:42.740 to write a book, all of a sudden you have much more interest in other people writing books,

00:07:47.060 not just for the content, but the actual process and things like that. So I remember we connected on

00:07:52.100 this about a year ago and I won't lie. I'm kind of secretly jealous and envious that you're done.

00:07:58.660 Yeah. Well, the scary thing is as soon as you're done,

00:08:00.920 your agent says, what's your next book? Oh, well, that's not even what I was thinking. I was

00:08:05.880 thinking the bigger issue is the moment you're done, you think, oh wait, there's just one more

00:08:09.900 thing I want to say. There is that. At some point the editor said, no more changes, David,

00:08:14.500 you're done. But it's a heck of a lot of work. Anybody who's written a book, real respect for

00:08:19.360 those people. And you're in the throes of yours right now, right? Yeah. So you had sent me your proposal

00:08:24.660 like a year ago, maybe longer actually, definitely longer than a year ago. And it's probably a great

00:08:30.480 example of how proposals and books often differ quite a bit. I'm generally the book I think ends

00:08:34.800 up being so much better, so much richer because there were things in the book that weren't necessarily

00:08:39.520 in the proposal that kind of caught me off guard, especially the first part of the book, which is

00:08:43.400 kind of what I want to talk about today. I think the second part of the book where you talk about

00:08:47.220 the societal implications for a longer lived population are interesting, but I'm going to let

00:08:54.080 somebody else talk about that with you. I want to talk about sort of the biology of this, but this

00:08:57.820 idea of, I mean, you introduce a very important mathematician who wrote a paper that I remember

00:09:03.860 seeing in my engineering studies, paper from the 1940s. So let's just start with that. Why is

00:09:11.240 Shannon relevant to the story of aging? Because I certainly didn't learn anything about him as an

00:09:16.540 anti-aging researcher 25 years ago. Well, Shannon's one of the most respected guys in math. What he did

00:09:22.740 in the 1940s was essentially figured out how to mathematically encode information and make sure

00:09:29.260 that information gets to the receiver. It's called the information theory of communication. What it led

00:09:33.840 to is what the world we have around us, the internet, TCPIP protocols. What that has to do with aging,

00:09:39.820 I wrote down in the book, it began really when I was just a postdoc in Lenny Guarante's lab at MIT.

00:09:45.520 Interesting, Claude Shannon was at MIT too. We discovered, to our surprise, that what was

00:09:52.240 controlling aging in yeast in part were these sirtuin genes. Now, sirtuins in science in aging are pretty

00:09:59.000 famous, but what a lot of people don't know is that the word sirtuin stands for a gene called sirtu,

00:10:05.680 which in yeast we found responds to dietary restriction, heat, starvation, and it allows

00:10:12.480 yeast to live about 30% longer if you upregulate it, or put just another copy of that gene in the

00:10:17.280 yeast cells. And Matt Caberline, who's now a famous professor, he was the graduate student who did that

00:10:22.980 experiment to put extra sirtu gene into the yeast, and they lived longer. So that was a massive

00:10:27.440 breakthrough, but what SIR stands for is really telling. It's actually been forgotten over the last

00:10:32.480 25 years. But SIR is an acronym for Silent Information Regulator. Silent Information Regulator.

00:10:40.900 What does that mean? This is a gene that controls other genes. It switches them on and off. Its main

00:10:45.480 job is to keep genes silent, and that allows cells to be dividing, be healthy. But what they do that is

00:10:53.400 not appreciated is when the cells are stressed, too much temperature, a broken chromosome,

00:10:57.700 the protein that this gene makes, this SIR gene, it leaves the silent regions, and it goes to repair

00:11:04.580 the problem or fix the problem. So in the case of DNA breaks, what we did was we broke the chromosome

00:11:08.980 of the yeast. The SIR enzyme protein leaves where it should be, goes to repair the DNA, and when it's

00:11:15.780 all fixed, it comes back again to shut those silent genes down. And you might say, that's a weird thing to

00:11:21.020 do as an organism. Well, this is a highly conserved process. It's found in yeast cells.

00:11:26.400 Just make sure the listener know what you're saying. You have a gene that's being silenced

00:11:30.480 that then gets broken. The thing that is silencing it leaves its silencing post to go and repair that

00:11:37.960 which is broken, only to come back and silence it. So from an expression standpoint, the world outside

00:11:43.400 hasn't changed. That's right. And the reason that I think it's set up that way is that these genes that

00:11:48.080 get turned on by the absence of the SIR protein help with the problem. They turn on DNA repair.

00:11:54.140 They hunker down. These are survival programs that's in every cell on the planet, I believe.

00:11:59.860 And the SIR enzyme that we're talking about is a master regulator of that survival circuit.

00:12:05.300 So why does that have anything to do with information theory? Well, what I think is going on is that

00:12:11.300 this program that's set down in our cells and when a yeast cell is young, which genes should stay on and

00:12:18.540 which genes should be off, gets messed up. It gets lost in the noise. And what's the noise?

00:12:24.240 It's this constant having to go and repair DNA or respond to too much heat or some other imminent

00:12:30.380 threat. Yeah. Doing it once doesn't matter. One cut to a chromosome, a broken chromosome is not going

00:12:35.120 to kill a yeast cell if it's repaired. But what it does over time is that cells lose their which genes

00:12:43.240 should be on and off. The genes that should be kept off in the yeast cell come on. One way to think

00:12:47.220 about this, a very simple analogy is Hurricane Katrina. Okay, so the SIR protein are the first

00:12:53.360 responders. They rush down to fix the problem. They do their job. But what happens is most of them come

00:13:00.240 back home, mow the lawn, pay the bills, and everything's good. But some of them get stuck down

00:13:04.740 there. They marry someone or maybe they get lost on the way home. They can't afford a ticket home on the

00:13:09.660 airplane. And you do that a hundred times. Or in our case, in our lives, every cell is getting a

00:13:15.740 broken chromosome every day. So you're doing this. How many days do we live? 30,000 days. This is a

00:13:22.860 problem, I think. Over time, you actually end up losing that program. And our cells lose their

00:13:29.480 identity. And the hallmark of aging in yeast is the loss of cellular identity. They become sterile. They

00:13:35.180 don't mate. Dysfunctional. So Claude Shannon figured out how to preserve information. You keep a

00:13:41.680 repository of the original data, a hard disk drive backup. And he said, if the receiver of that

00:13:49.080 information, which in his idea was the radio receiver of the signal in World War II and after,

00:13:55.920 in our world, it might be our email. In aging, it's ourselves in the future. And we lose that

00:14:02.180 program of keeping ourselves young. Cells and selves. And what Shannon said is that there may

00:14:09.380 be a repository of that information. If our cells are right, if they're doing a good job,

00:14:13.800 there should be a way to reset the system to get that complete email back again. And I think there

00:14:20.340 is one. We have some early evidence from mice that we can actually find that hard disk drive

00:14:26.580 and reinstall the software so that it's pristine again. And we find that we can actually improve the

00:14:32.160 health quite dramatically of parts of a mouse's body. Now, to really get into this the way you do in

00:14:37.880 the book, I think we have to take a few steps back and assume for a moment that a listener doesn't

00:14:44.320 know much about DNA beyond sort of the high level stuff, but maybe doesn't understand what an epigenetic

00:14:51.440 modification is, what things like methylation mean and how that occurs. So let's go back and go through a

00:14:57.780 little bit of that stuff because I think to get a firm understanding of this will enable the other

00:15:02.080 things you're going to talk about to make more sense. So we could use a mouse, we could use a

00:15:05.560 person. It doesn't really matter for the purpose of this discussion, but let's use a person because

00:15:08.680 I think it resonates with a listener more. So you and I have lots of DNA. We've got somewhere between

00:15:14.440 20,000 and 30,000 genes. These genes are made up of lots of this coding stuff and there's coding and

00:15:21.840 non-coding segments of these DNA, but basically it's a whole bunch of strung together nucleotides.

00:15:29.160 Now, some of them are not working at that point in time, right? They're sitting there, but they're

00:15:34.080 not actually getting turned into RNA to be turned into proteins, correct? Yeah, absolutely. We don't

00:15:40.700 want to turn all our genes on because then we wouldn't have any different cell types. We'd all just

00:15:44.980 be a giant blob of cells, a giant tumor probably. So when you take a piece of my skin and you take a

00:15:51.300 piece of my liver or you take a piece of my eye, nerve in the back of my eye, they all have the

00:15:57.760 same DNA, right? Yeah. Except for your sperm or your egg and some immune cells. I don't have any

00:16:03.640 eggs by the way, just for the record. Yeah. Okay. I can believe that. But your point is some of them

00:16:09.800 are getting turned on. So the DNA that are sitting in the epithelial cells of my skin are being instructed

00:16:15.120 to express proteins in a certain way. And that's what ends up producing skin versus the nerve in the back of

00:16:20.920 my eye versus my liver, et cetera. And what's interesting is the skin and the nerve at one

00:16:26.080 point were the same cell types when you were an embryo. But the pattern, we say pattern of gene

00:16:31.340 expression, the way the cells turn on different genes allowed nerves to become nerves and hopefully

00:16:36.120 stay nerves for most of your life. But I think towards the end of your life, that's the problem.

00:16:39.980 It's they're losing their identity and reverting back to something that was more primordial,

00:16:44.180 more like a skin cell. And the way the cells do that is really interesting. So that your DNA,

00:16:48.280 if you stretch it out of a cell, it'd be about six meters long. And there are little proteins that

00:16:52.720 wrap that length of DNA up inside a microscopic cell. And it's essentially like spooling a hose

00:17:00.160 in your garden. And then you spool those spools on top of each other. So it's a way of packaging

00:17:04.440 something very long, but very, very thin and tiny. But if it's packed up in a spool,

00:17:08.580 it won't be read by the cell. And that's what these serproteins do. The serproteins

00:17:12.160 actually maintain that structure of that spool. But if a cell needs to read the gene,

00:17:17.600 they remove the serproteins. And now that bundle can open up and the gene can be read.

00:17:23.160 And there are different layers of the epigenome. There's the superficial layer where the serproteins

00:17:27.460 are, and there are other proteins going around turning genes on and off. But there's a really

00:17:31.780 deep layer. The deepest layer is what's called DNA methylation. And cells can permanently,

00:17:37.000 for decades, mark a gene to be silent by putting these chemical groups called methyls

00:17:42.340 next to or in a gene. And that tells the cell, spool that up as tight as you can. And never,

00:17:49.760 unless I tell you to, reveal it again. And that's the reason our brain doesn't turn into a liver

00:17:54.760 one morning when we wake up.

00:17:56.320 Now, let's define methylation only because it's such a buzz term right now. I think everybody and

00:18:01.200 their brother is reading about, oh my God, am I a bad methylator? Am I a good methylator? Do I have an

00:18:06.680 MTHFR mutation? But let's sort of demystify all this stuff. A methyl group is a carbon group with

00:18:13.680 three hydrogens on it. So we're talking a very basic molecule in the broader architecture of

00:18:19.040 organic chemistry. And you're saying that when you put one of those methyl groups, when you attach

00:18:23.880 one of the carbons on that methyl to literally a carbon on one of the ends of DNA, you have the

00:18:30.020 ability to program it to not carry out the function of expressing itself. Is that a safe way to say

00:18:37.840 that?

00:18:38.160 It is. This is the underlying code that tells a cell what type it is. And there's a guy called

00:18:42.540 Conrad Waddington from the 1950s who didn't know that there were these methyl groups. So what he

00:18:47.460 imagined was the embryo or the fertilized cell is at the top of a hill and it rolls down the hill.

00:18:53.820 And if it lands in one valley, it's a nerve cell and it stays in the valley. But if it rolled into

00:18:59.620 another valley, it's a skin cell and stays a skin cell. And that's a great metaphor for how an embryo

00:19:05.400 and then eventually a baby is formed out of 27 billion cells that a baby is made of. Waddington

00:19:11.900 didn't think about as much, if at all, was what happens after that baby's born? What happens 80 years

00:19:18.020 later to his Waddington landscape? What happens to those valleys and those hills? And what I think is

00:19:23.160 happening is that we're having not just erosion of those hills so that the cells don't stay where

00:19:27.460 they should, but they're getting jostled by these DNA breaks and this reorganization of the

00:19:31.780 seroproteins and others. So the cells start to migrate up over the valley into other new valleys.

00:19:38.300 And now your neurons in your brain, your nerve cells are starting to behave a little bit like

00:19:42.040 skin cells or liver cells. And I think that's what's underlying many, if not all of the aspects

00:19:46.900 of aging that we eventually will succumb to.

00:19:49.520 And what role does entropy play in this? Because when you think about these hills and valleys,

00:19:54.280 there's a place where things want to settle out. There should be a place where things settle and

00:19:59.660 stay put. But the figure you have in one of your talks, which shows this projection right into the

00:20:06.680 future, it starts to get blurry at the end. In other words, it starts to look like there's more chaos in

00:20:12.640 the system. Is that just the natural drive of entropy within our systems? Is that the way it's

00:20:18.820 expressed, I guess, is what I'm asking?

00:20:20.540 Right. So mathematically, it appears very much like a loss of information and introduction of noise.

00:20:27.040 Because these methyl groups that are laid down in a pristine, precise fashion when we're young,

00:20:33.180 there are other methyls that accumulate over time in different places, what appears to be

00:20:38.460 randomly. So it's a loss of the original pattern. And that's why we talk about entropy. Now,

00:20:44.440 anyone who's familiar with the second law of thermodynamics says, okay, we're screwed. We're

00:20:48.580 never going to be young again, because you've lost information, similar to falling into a black hole.

00:20:53.280 You're never coming out.

00:20:54.460 Or even let's think of more obvious examples, like you can't unfry an egg. Once proteins become

00:21:00.800 denatured, once the clear part of the egg becomes white, you don't get to make it clear again.

00:21:05.940 Yeah. Or even worse, if your genome is a compact disc or DVD, you scratch that up,

00:21:10.840 and you can't read it again. Or even worse, you break a piece off, that information is lost. You

00:21:15.520 throw it in the trash. But what I think exists in cells, and you have some evidence, is that,

00:21:21.140 like Shannon suggested for the internet or information, is that if you have a backup copy,

00:21:26.920 and now going back to the genome, there seems to be something in cells that tells them these methyl

00:21:31.760 groups, the program that was laid down when you were a baby, is still there. And cells can access that

00:21:37.580 somehow, to say, all these other things that have happened since you were born, or since you were

00:21:41.660 a teenager, that's just noise. That's crap. Ignore that. In fact, when I tell you, and in my lab, we can

00:21:47.580 reprogram cells to go young again, to read the right pattern, there's a process that we're just beginning

00:21:53.440 to understand that says, this stuff is noise. Get rid of that. Ignore that. Get rid of it. But these other

00:21:59.420 methyl signals, these little flags on the genome that have been there since we were babies,

00:22:03.320 that's the good stuff. Keep that. And in that way, these spools of the hose, the DNA loops that

00:22:10.900 have become untangled and messed up as we've gotten older, reset back to being young again.

00:22:17.440 So let me make sure I understand that. If you could take the version of David Sinclair that was born,

00:22:22.840 and you could look at every single piece of methylation across every single strand of DNA on

00:22:29.260 every single gene in every single chromosome, and you had a picture of that, and you knew what

00:22:34.180 that always looked like, fast forward 50 years, some of those methyl groups are gone. You've lost

00:22:41.300 methylation in some places, and presumably in many more places, there are now methyl groups added that

00:22:46.260 were not present. So you've had both an addition and subtraction of methyl groups. And that now looks

00:22:51.560 like a different picture. Are you suggesting that at least to the first order, if you restored the

00:22:58.260 methylation status to what it looked like when you were born, you'd have a younger phenotype?

00:23:03.860 That's exactly what I'm saying.

00:23:05.100 That wasn't like a canned question. I wasn't even asking that rhetorically. Because on one level,

00:23:09.240 that seems really complicated. But on one level, it actually seems kind of simple.

00:23:12.820 You know what I mean? That's sort of what's weird about it.

00:23:14.460 What's amazing about it is that we don't even understand how it fully works yet. We just know that

00:23:18.640 there is this backup copy that we can access. At least we have early evidence of it.

00:23:22.900 So we have this manuscript that says that if we turn on a few key genes in the body of a mouse

00:23:30.400 or in a cell, it will quite literally not just act younger and turn on young patterns of genes.

00:23:38.200 When you measure its age by counting these methyl groups where they are, it is young again. And so

00:23:44.420 this backup copy exists. But what warps my mind, it blows my mind, is that there is something in the cell

00:23:51.620 that we've had all along in our lives that allows the cell to reset.

00:23:56.660 Can I ask a question? Tell me if this is the right time to answer this question.

00:23:59.360 A lot of what you're saying sounds like, hey, isn't that quote unquote a stem cell?

00:24:03.720 It's very much like that. Because what we've done in the field is we've taken the knowledge from

00:24:08.500 the reprogramming field, which is using what are called Yamanaka genes. And they're used in the

00:24:14.880 field right now by scientists, by companies to take an adult cell. I could take your skin cell,

00:24:20.080 for example, grow it in my lab, and I could make a stem cell out of that. How does that work?

00:24:24.760 It strips pretty much all of the methyl groups off the DNA. And that's a reset, not back to being

00:24:30.760 young, but by being primordial, by being pre-embryonic. In other words, methylation by definition

00:24:37.280 implies an aging in some level of a cell. So technically, your most primordial cell had no

00:24:43.660 methyl groups on its DNA? Right, right. Very few. And as we get older, and it's not... So what we've

00:24:48.500 discovered as a field is what, again, a few years ago, I wouldn't have believed it, but

00:24:52.260 it is true, is that this clock starts aging from conception. So even as we're growing in

00:24:58.700 the womb, our DNA is accumulating these chemical changes, these methyls. And that extends throughout

00:25:04.460 life forever. And if you gave me a sample of an embryo, or a baby, or a teenage girl, or

00:25:11.560 an 80-year-old, I and some other labs in the world can read that DNA. And by the pattern,

00:25:17.800 I could tell you exactly the age of that cell or that tissue.

00:25:21.100 It's almost like you're describing carbon dating of a cell, for lack of a better word.

00:25:25.640 No, that's perfect. That's the best way I've heard it described yet.

00:25:28.340 And I didn't realize that the fidelity was as great as you've just described it. I understood that you

00:25:33.000 could look at methylation and tell the difference between a baby and a 50-year-old. But if I heard you

00:25:37.740 correctly, you're saying that if you took a fetus at one month of development versus a newborn,

00:25:44.720 a full eight months later, you would actually be able to distinguish those

00:25:48.340 on the basis of methylation. I mean, obviously, phenotype and many things would give that away.

00:25:52.900 We would. And in fact, the pace of change is very rapid in an embryo.

00:25:56.540 What is the fidelity of this when you look at one, two, three, four, five-year-olds? Is it literally

00:26:02.020 that precise that it can measure the age of DNA to within a year?

00:26:07.820 I don't know what the latest statistics are. For a human blood sample, it was 95% accurate

00:26:13.720 for chronological age. But what we've realized is that the clock actually changes depending on how

00:26:18.740 you live your life. Well, I was just about to say, this is not that interesting if there's nothing

00:26:22.780 you can do about it. It's just one more reminder of your birth certificate. It's only relevant if

00:26:27.660 either things speed it up, which we'd like to know what those things are to avoid them,

00:26:32.020 or things can slow it down or even, as you said, reverse it, although that seems too good to be

00:26:36.800 true, right? Well, it's even more important than your birth certificate. Your birth certificate just

00:26:40.520 tells you when you're born. This clock tells you how fast you're aging. And a few labs, Steve Horvath

00:26:46.500 is one of the inventors of the clock. We actually call it the Horvath clock. He is able to estimate

00:26:51.960 not just how old you are, but predict when you're going to die with high accuracy. And that's really scary

00:26:56.960 that we are predestined based on our lifestyle up to that point, how long we're going to live.

00:27:01.860 And there are some things that slow it down. Exercise is one good thing. Calorie restriction.

00:27:06.700 Smoking does the opposite. So right now at this moment, having not eaten in a couple of days,

00:27:11.240 I am slowing my clock. You are. So there are probably mechanisms. Actually, we know of mechanisms

00:27:17.140 that are turning on sirtuins. So your NAD levels go up when you're fasting and sirtuins do a better job of

00:27:23.020 both repairing the DNA and keeping these- They have more substrate now to do their job.

00:27:27.560 Right. And just like in yeast, our cells need to do both at the same time, repair DNA and keep the 0.59

00:27:32.120 genes silent. That should be silent. But if we're lazy and we eat a lot of food and we don't exercise,

00:27:38.520 those programs that are designed to keep the clock from accelerating aren't as active. So that's what

00:27:45.020 we think is going on. And that's why if you have smoked a lot of your life or you don't exercise,

00:27:49.900 your clock will, on average, and for most people that are tested, will be older than your actual

00:27:54.560 age. And so the only thing that you can do is either slow that down. Right now, we don't know

00:27:59.360 how to reverse it in humans, but in mice, we're getting glimpses of how to actually literally make

00:28:04.260 a cell half its age that it once was. And how does this clock work in its prediction relative to

00:28:11.400 other things that get a lot of attention, such as the length of the telomeres of a cell? And maybe

00:28:17.000 define for the listener what telomeres are, since some people might not be familiar.

00:28:20.700 Our chromosomes, all 46 of them, are linear DNA molecules. So they have ends, two ends each

00:28:25.840 chromosome. And those ends need to be protected. And those ends are called telomeres. And a good

00:28:32.680 analogy of-

00:28:33.260 It's like the hard piece of a shoestring.

00:28:35.460 Exactly. That's what an aglet is. But as we get older, they get chewed back and eventually they 0.98

00:28:39.620 become so frayed that the cell recognizes the end of a chromosome as though it would be a broken

00:28:44.300 piece of DNA. And they shut the cell down trying to repair it and try to stick it together with

00:28:48.880 another piece. And you end up with a whole bunch of genetic chaos. One thing that cells do to prevent

00:28:54.420 from becoming a tumor is they shut themselves down and they become senescent. And that's a whole other

00:28:59.500 problem for the body once you've got a zombie cell that's not dividing and is in panic mode.

00:29:04.960 But telomeres do erode as we get older. So they've served as a pretty good clock. But what happens is

00:29:12.000 that cells can divide faster in different tissues. Some tissues don't even divide.

00:29:16.580 So it's not like a universal clock like the one that I just described that Horvath and others

00:29:22.060 discovered.

00:29:22.700 There was a paper that came out maybe six months ago in science after- remember that you had the

00:29:27.380 twin astronauts. One astronaut was at the space station for a year or the other. His twin brother

00:29:31.140 was on Earth for the same period of time, of course. And the paper was interesting in that- and there

00:29:35.820 were two, actually. I'm trying to remember. There was one of the two papers I read. So it's possible I'm

00:29:39.780 going to misrepresent this because I don't have all of the data. But what I remember struck me is

00:29:44.380 interesting enough that I was surprised nobody else was talking about it. Or if they were,

00:29:48.060 I was somehow missing it. Was they made a big deal about the fact that a lot of changes occurred

00:29:52.740 in space. And certainly some of these were very obvious and predictable. You could imagine bone

00:29:57.780 density going down, muscle mass, things like that. But they talked about how there was this dramatic

00:30:01.840 difference in the length of the telomeres of these two twins who presumably would have- I assume

00:30:07.980 they had measured pre and post. So it wasn't. But what was really interesting was within

00:30:12.320 something like two to three days of being back on Earth, the twin that was in space had a complete

00:30:18.380 reversion to what his brother was when he was back on Earth. And maybe I'm just skeptical, but that

00:30:25.500 made me a little less interested in telomeres as a particularly relevant metric of the age of an

00:30:32.000 individual. Would you interpret that differently?

00:30:34.100 No, I agree with that. That something that can change within days is less interesting as a clock

00:30:40.560 compared to something that seems to be immutable and ticking every day of your life. That's the

00:30:45.900 holy grail. And it looks like we may have found something like that in the field.

00:30:48.920 And Horowitz is a pretty young guy.

00:30:50.500 Yeah, he's younger than me. So he's in his forties, I think.

00:30:52.960 And he's a mathematician?

00:30:54.140 He is.

00:30:54.640 At UCLA.

00:30:55.320 He needed to be a mathematician because to find the clock, you can't just read all the methyl groups.

00:31:00.220 You actually need to train a computer that uses machine learning to find out which are the ones

00:31:04.100 that change with age and those others that just randomly change.

00:31:07.680 I know I've read a tiny bit about this, enough to be dangerous, right? So Horowitz used the data

00:31:12.620 from like Illumina or something like that. There was publicly available data. He had maybe 8,000

00:31:16.820 samples to study and presumably enough of it was longitudinal. In other words, he must have had some

00:31:21.680 samples of the same people. That would give you one piece of data. But I guess if he knew the age of

00:31:27.640 the persons, you could train a machine, as you're saying, I guess, to see how much methylation has

00:31:33.580 occurred. I guess I apologize if this question goes beyond your level of expertise on this subject.

00:31:37.560 It sounds like he's someone I should interview as well. Is he able to, or are you able to,

00:31:41.560 in your lab, look at methyl groups and know if they've been there for a long period of time,

00:31:47.940 if they're new, if they are additions, or if they basically, they were largely inherited? I mean,

00:31:53.480 do you have that capacity?

00:31:54.480 We do. That works because we know from building the clock that at age, five years of age,

00:32:00.660 this is the pattern that was likely there for an average human. And this is the pattern of an 80

00:32:05.680 year old and all in between.

00:32:07.200 But that would be true at a macro level. But I'm saying at the level of a given gene,

00:32:11.240 are you actually able to infer that?

00:32:13.700 Yes. So Horowitz's clock is built, there were a number of clocks. His latest one is built on a few

00:32:17.900 hundred sites on the genome that are very specific. And for reasons that he doesn't understand,

00:32:23.520 and I think we are beginning to understand, those are the ones that are sensitive to age

00:32:28.500 and that tick along. And some tick differently in different tissues. But he recently published a

00:32:34.520 universal clock. And another group at Harvard published a universal clock for mice, which means

00:32:38.820 it doesn't matter if I give you a blood sample or a skin sample or a brain sample,

00:32:43.600 by looking at that precise location on the genome next to gene X, and you look at 300 of them or so,

00:32:50.340 then you can actually say exactly what age that mouse is, independent of what part of that mouse

00:32:55.720 you're given.

00:32:57.100 That's kind of amazing when you consider the fact that if you didn't know that a priori,

00:33:02.160 you had to start with all of the genes and look at all of the potential sites of methylation.

00:33:08.240 I agree.

00:33:08.920 But just be an unbelievable problem.

00:33:11.240 When Horvath and Hannum, the other guy that did this, when those papers came out,

00:33:15.240 it was really hard to believe. Because for the last 50 years, we've been dreaming of a clock.

00:33:21.280 But the evolutionary biologist said, no, it's just organisms wasting away, like a car breaks down,

00:33:28.260 lack of selection for longevity. There's no way there's going to be a clock,

00:33:32.300 because aging is not considered a genetic program. And it isn't. There isn't a program that tells us

00:33:37.600 we must age. No one I know in their right mind believes that. But there are processes like I

00:33:42.520 was describing about the movement of these serproteins in yeast and movement in mammals

00:33:46.620 like us that leads to a predictable change on our genome that changes the way genes are switched

00:33:53.080 on and off as we age in very precise locations. But what truly blows my mind is that the cell somehow

00:33:59.920 knows which are the young ones and which are the old ones. And when we tell it to, it can reset the

00:34:05.980 cell back to what it was. But you don't want to go too far. Your point about stem cells is well taken.

00:34:10.780 If you push it too far, and some labs have done that, one color spell monte at the Salk Institute

00:34:15.200 showed, you turn on this reprogramming, you can actually cause tumors in a mouse, or actually turn

00:34:21.240 them on really fast, and they'll die within two days. You can push those balls up that landscape from

00:34:26.420 the valleys far too quickly. And the cells, they don't just regain their identity, they go all the way

00:34:31.440 back to being basically an embryo. And that's not going to help anybody if we kill people after two

00:34:36.420 days. But in my lab, we do what's called partial reprogramming. We push them a little bit so that

00:34:41.400 they regain their youth, but they don't lose their identity. So let me think about this for a second.

00:34:46.620 So when you go back to the analogy of Shannon, which is you've got receiver operator trying to

00:34:53.100 communicate through a signal, through an electrical signal, there's a loss of fidelity in the signal.

00:34:58.280 So the receiver operator pair have to be able to compare this transmitted signal to a master

00:35:05.560 signal. When you bring that analogy down into our DNA, is it the expectation that within every cell

00:35:13.680 resides a master copy, or not necessarily within every cell? It appears to be within every cell,

00:35:19.540 because when we reprogram the animal, and currently we're choosing to reprogram the retina and restore

00:35:25.060 eyesight in old mice. The cells that get the reprogramming signal from the three genes we put

00:35:30.100 in, those cells will survive, will regrow, will restore their function back to being young again.

00:35:37.260 And if a cell next door doesn't get the reprogramming signal, it doesn't regenerate. So it appears to be

00:35:42.440 intrinsic for each cell. We can also do that in the dish. We can grow human cells, grow nerve cells,

00:35:48.560 mini brain in the dish. We can reprogram those to be young again and survive a stress,

00:35:53.400 such as chemotherapy and regrow, as though they were young embryonic cells again.

00:35:57.900 So let's talk about what this reprogramming means.

00:36:01.180 What I mean by reprogramming is that we can use technology that we use now to generate stem cells,

00:36:08.600 but to partially reprogram them so that they turn on the youthful pattern of genes that we once had,

00:36:16.020 that we know we lose. There's no question, as we get older, our cells don't turn on the genes

00:36:20.280 that they once did when they were young, and genes that should be on when you're young get switched

00:36:24.400 off. So we lose that. That's that noise, epigenetic noise, I call it. Reprogramming somehow resets that

00:36:30.820 pattern. In terms of the hose and the spooling of the DNA, what's actually going on is that genes that

00:36:36.480 were once tightly bundled up by the seroproteins and by methylated DNA are coming undone as we get

00:36:42.140 older, this noise. Reprogramming somehow tells the cell that region of the genome, package that up again,

00:36:50.280 get the seroproteins to go back there, get the methyl back there, or remove them, and get that gene to switch

00:36:56.260 off again, because that gene has no business being on in the retina. You might need it somewhere else, and

00:37:01.440 somehow, somehow a liver cell knows that that gene shouldn't be silenced. Maybe it should come on.

00:37:07.720 But there's a repository. So in Shannon parlance, he calls it the observer. His backup disk is called the observer,

00:37:14.000 and the observer keeps a signal, keeps the original signal, until it's needed. And if the receiver of the

00:37:20.020 signal does the checksum, you know what a checksum is, I think all your listeners will know, that

00:37:25.480 every time a signal is sent, we know that it is complete because all the math adds up. If that

00:37:32.160 doesn't add up... That's a signal that says there's been a transmission error.

00:37:35.980 Right? And instead of going back to the original sender, it goes to the observer who keeps a copy

00:37:42.180 of the original signal and gets the rest of the data. So if you're sending, let's say, a photo to

00:37:47.840 Instagram, often you're in the subway or it doesn't make it. Half of the picture will make it. But

00:37:54.300 Instagram computers will say, hey, I only got half this picture. It didn't add up. The checksum didn't

00:37:58.900 make any sense. Please resend the rest of that signal. It may be stuck in Denmark. It may be stuck in

00:38:04.500 Iceland somewhere. But the system's set up that way. And that's what I think is going on in every

00:38:09.760 cell in our body is that information is still there. Our DNA is largely intact. We haven't lost

00:38:15.500 the genes due to mutations. That was the old idea from the 1950s. That information is there. We just

00:38:20.940 don't access it because the cells don't know whether to spool up the DNA and hide it or to expose it to

00:38:26.200 turn on the right genes. And I'm still just kind of blown away by this notion that you can... You shared

00:38:32.540 of me once a story about how the Horvath clock was accurate enough that you could even predict how

00:38:37.920 much a person had smoked. Yeah, this paper recently came out. They had the records of packs per day

00:38:43.520 that were accurate based on, I guess, medical records. And then they also asked these patients,

00:38:50.640 how much do you think you were smoking over your lifetime? And they made up some number based on what

00:38:55.280 they remembered. The clock matched what the medical record said and not what the person said.

00:39:01.740 So I'm sure they weren't lying. No, of course not. Because they would have had to have told the medical

00:39:05.340 record in the first place. But presumably, it's easier over time to tell. Like, if you ask a person

00:39:11.220 every year how much they're smoking, you'll get a more accurate response of the aggregate smoke versus

00:39:15.540 at the end of... That's exactly right. What it tells us is that this clock, there's no lying here.

00:39:22.000 Your DNA doesn't lie. Your clock records probably every good activity and every bad activity that

00:39:27.780 you've had in your life. Talk about the ultimate wearable. Could you imagine... We love our rings

00:39:32.240 and our CGMs and all of these things. But imagine you had a little Horvath clock you could stick into

00:39:37.220 your interstitial fluid. It's coming. If you gave me your DNA, I could tell you how old you are

00:39:42.640 biologically. I would love to do that. In fact, you know what would be fun is let's do a longitudinal

00:39:47.680 sample. We'll sort of stress the system a little bit. So we'll do here's some DNA and then we'll do

00:39:53.740 a fast. You know, I'll go a week without eating or something like that. Because that's a pretty

00:39:57.720 extreme stress. Like a week of fasting should do it. In fact, we should do this with rapamycin dosing.

00:40:02.940 Like I'm curious as to whether the pulsatile dosing of rapamycin that I take, does that have an impact?

00:40:08.980 In other words, you start to wonder, are there final common pathways that so many of these higher

00:40:14.980 level interventions impact? I think level is the right word. The spooling of the DNA, this epigenome.

00:40:22.880 There are various levels. The superficial level are these transient proteins. We call them

00:40:27.480 transcription factors that jump across the genes and tell them to be read. Another level down are

00:40:31.740 these sirtuins that we work on. They actually chemically modify the spooling proteins. We call

00:40:36.500 them histones. And then the deepest level, the third level down, is this methylation clock, which

00:40:41.660 is very hard to reverse. As far as I know, the only way to really do it is using a stem cell

00:40:47.040 technology, this reprogramming factor is Yamanaka. In other words, your belief is, I want to come back

00:40:51.380 to that because I actually like that level system you just put forward. But the reason you're saying

00:40:56.060 that is the best your evidence suggests in the lab today, outside of using a vector to actually insert

00:41:03.040 new DNA or something, is you can change the rate, but you can't change the direction. Is that sort of

00:41:08.860 what you're seeing? Yeah. That's the summary of now hundreds of papers on this topic. It's a scary

00:41:13.600 thought. All of these interventions, rapamycin, NAD boosters, metformin, all the data's not in,

00:41:19.220 of course. We need to do more. But the first studies have said that they have relatively little impact

00:41:24.540 on this very deep clock. And they slow it down. That's fine. They stop the marbles or the balls from

00:41:30.580 jumping too far across valleys. But they don't get the balls to go back into the valleys that they were

00:41:35.660 once when we were young, which makes sense. You can't take rapamycin or NAD boost or metformin and

00:41:40.980 restore vision in a mouse like we are with reprogramming factors. So let's talk about

00:41:45.460 that. That's obviously now we're entering the future. So talk about that experiment.

00:41:53.140 Well, so what we did was we took three of the Yamanaka genes that are used to make stem cells

00:41:57.360 and we packaged them into what's called an AAV, adeno-associated virus. This is a virus,

00:42:03.400 by the way, as you're already used by many companies. There are a number of products on

00:42:07.420 the market. Some are used actually to fix eye diseases, genetic diseases. So this isn't some

00:42:12.660 crazy super science fiction story. This is medical FDA-approved drug development. We take these

00:42:20.380 vectors, viruses, we package these three genes in, which is not easy because they don't hold much.

00:42:25.360 So we whittled this down. We gave it an on-off switch, which is important because you don't want

00:42:29.520 to become a stem cell. You don't want your eye to develop a tumor, if it might. We did another trick,

00:42:34.600 which is important, which is we left off the fourth Yamanaka factor called MYC. Now MYC is a 0.63

00:42:40.660 well-known oncogene, gene that causes cancer. That didn't take a genius to leave that off.

00:42:45.940 But what was surprisingly rewarding to see was that the MYC gene was superfluous. We didn't need it

00:42:53.220 to reprogram cells to be partially young again. How did you program the on-off switch?

00:42:59.260 Oh, that's really interesting too. This is in the field. We use a system where we can feed

00:43:03.540 a cell or a mouse doxycycline and it's just an antibiotic. So my daughter who had Lyme disease

00:43:09.700 will tell you it's not great for long-term use, but just for a week or a month, it's fine.

00:43:15.460 We've given it to mice for their whole life. They seem to be okay. We set it up so that what's

00:43:20.540 called a doxycycline responsive gene. And so now when we have the virus-

00:43:25.900 So you can silence the gene with doxy. Basically you use something inert that has a trigger.

00:43:30.360 Yeah. We could have used a bunch of different chemicals, but we use this one because it's-

00:43:33.560 You understand it well and it's pretty benign.

00:43:35.680 And the FDA would likely approve it because we know a lot about it. We didn't re-engineer it so

00:43:40.020 that it was extremely tight. So we don't want any leaky gene if we don't want it. And we also

00:43:44.560 made its levels very low because we don't want to blow the system out.

00:43:47.640 So we made this new version of the virus and delivered it.

00:43:52.160 And an adenovirus, for folks who might not know what that term is, these are very common viruses.

00:43:57.740 Many of us have been exposed to these already.

00:43:59.780 Presumably everybody has. I mean, it'd be hard to not be exposed to an adenovirus at some point,

00:44:03.680 but unless you live in a bubble. And there's so many different ones. Yeah. And remind me,

00:44:06.880 is an adenovirus a DNA virus or an RNA virus? I can't even remember.

00:44:10.380 Yeah. I can't remember either.

00:44:11.360 I used to know that. Okay. So it's a basically, and how much DNA can you pack into it?

00:44:16.200 I'm pretty sure it's a DNA virus. So we package DNA in there. You can pack a 5.4

00:44:20.760 thousand base pair of DNA.

00:44:23.860 5.4. Okay. All right. And you go back to these mice and the phenotype of the mouse is what at this

00:44:31.520 point? It's a old mouse, middle-aged mouse?

00:44:34.800 Well, we've done three crazy things to a mouse to test. The first thing we did was we crushed the

00:44:39.660 optic nerve.

00:44:40.240 So you took a normal mouse and you crushed its optic nerve. So now it can't see.

00:44:44.580 Yeah. And only very, very young mice will regrow nerves. You break your spine, you're not going to

00:44:49.840 walk again. And so we thought maybe if we turn the age of those cells back to being extremely young,

00:44:56.020 they'll grow back. And there's nothing that works to grow all the way back, a healthy optic nerve.

00:45:01.940 So we did that. I won't reveal the punchline yet. We also, these are on collaborations. I should

00:45:07.180 credit my collaborators. Jigang He at the Children's Hospital in Boston and Bruce Cassandra's lab did

00:45:13.520 this experiment. I'm going to tell you about glaucoma. So he puts pressure in the eye. Pressure is one of

00:45:19.160 the large drivers of glaucoma and disrupts vision. And then the third experiment was to take just

00:45:24.840 regular old mice that were two years of age and reprogram those. Actually, they were 12 months of

00:45:30.620 age. We didn't go too far, but by 12 months of age, a mouse has lost a lot of its vision.

00:45:35.140 Interesting. Cause that's only about 35 year old human, right? Yeah. 40. Okay. It's about 40 year

00:45:39.860 old, but those mice, we thought even that's an old house. So most people can't restore vision in

00:45:45.420 anything. We have now treated much older mice and we do see some partial effects, but. Okay. So let me

00:45:51.420 recap that. So you've got the first group are young mice, but they've had their optic nerve surgically

00:45:58.540 traumatized or traumatized through compression. You have a second group. I don't recall the age,

00:46:03.740 but you increased intraocular pressure to mimic glaucoma. And then the third group, you just took

00:46:09.200 relatively old mice that had a natural decline in their vision. Right. And there are two ways that

00:46:14.820 Bruce's lab measures vision in mice. One is you expose, well, you show it a TV screen with moving

00:46:20.280 lines and the lines can be thick or thin. And if you've got really good vision, you can see the thin

00:46:25.440 lines and you can watch the mouse turn its head. And if it's blind or it can't see, it obviously

00:46:30.040 won't move its head. Now that's partially subjective. It's not bad, but you want something

00:46:35.780 completely objective. And you can't measure the electrical signal in the occipital cortex. Okay.

00:46:40.900 So we stick an electrode in the back of the eye and measure the signals that come like an

00:46:45.080 electrocardiogram for the retina. And that doesn't lie. You want both to agree with each other,

00:46:49.720 but that one we can measure before and after treatment. And there we can see what the virus

00:46:56.640 is doing. Now, one thing that I haven't mentioned, which your listeners may be interested in is when

00:47:00.840 do we turn on the virus? And because we want to treat patients, we're not just doing this for fun.

00:47:05.160 We turned on these Yamanaka genes in the virus after the treatment, not before. Because if you've 0.99

00:47:10.980 got glaucoma or you've damaged your optic nerve or you're old, you need to be able to reverse that,

00:47:15.680 not just prevent the damage. And so we did that. So these results that we've put out there,

00:47:19.380 actually show that in all three experiments, you regenerate nerves. And in the case of glaucoma

00:47:25.740 and old mice, they get their vision back. So let's go through actually technically how you do this.

00:47:33.080 So the adenovirus vector now contains the modified stem cell, basically the genes of a modified stem

00:47:40.900 cell. Is that a safe way to describe it? Well, a better way would be that these are genes that

00:47:44.980 specify us during embryogenesis. So it tells the cell,

00:47:48.720 but you took it back to, yeah, I was trying to figure out where you, so you basically

00:47:52.460 have the methylated pattern of the gene that mouse would have had during development.

00:47:59.420 Well, we think so. We've measured the clock in nerve cells after reprogramming and they are younger.

00:48:04.480 Steve Horvath helped us with that. But what's also interesting is we can look at which genes are

00:48:08.920 on and off. And if we look at the young mouse and the old mouse and look at which genes are on and

00:48:13.800 off, we can see that we've got noise now. Genes that shouldn't be on are coming on, such as genes

00:48:19.940 that are involved in stress responses. And other genes, interestingly, are coming on that shouldn't

00:48:25.060 be in the eye in a young mouse. And those are very interesting. Some of those are, I think about 80 of

00:48:29.860 them are taste receptors, meaning proteins that stick out of cells and smell. Actual taste receptors

00:48:35.240 in a mouse or homologues to other animals? Well, they're predicted to be taste receptors or smell

00:48:42.620 receptors. They're called olfactory receptors. But what they're doing in the back of the eye and why

00:48:46.660 they change with aging is so far a complete mystery. I could speculate that they're important

00:48:51.740 for signaling chemicals that have nothing to do with smell. Has this been identified, for example,

00:48:55.560 in humans? Do we know that humans also have olfactory nerves in their eyes? I don't believe

00:49:01.000 so. I think we were the first to figure that out. But there are sporadic accounts of it because it's

00:49:05.860 so weird. Very few people have paid attention to it. But anyway, the end of the story is that when we

00:49:10.140 reprogram the retina, those genes that were coming on during aging get turned off and vice versa.

00:49:16.420 How many genes are we talking about, David, in this particular process?

00:49:19.640 Just hundreds, hundreds of genes. In terms of statistical significance, depends where you want to draw

00:49:23.740 the line. So I think it might be a few hundred that are really significant. But the thing that

00:49:29.160 blows my mind about this result is that the genes that went down with aging just a little bit when

00:49:35.240 reprogrammed come up a little bit. And those genes that went way down with aging when you reprogram

00:49:40.120 are way up. So it's as though the cell knows that this gene should be put back to the way it was.

00:49:47.440 And I have some ideas, but I have no evidence how that actually happens.

00:49:53.700 Sorry to interrupt, but does that correlate to how much methylation is on the gene?

00:49:58.800 In other words, you talked about this gene was down a little bit and it corrected to going up a

00:50:04.080 little bit. This one was down a lot and it corrected to go up a lot. Is the a little versus a lot

00:50:10.100 correspond closely to the extent of methylation?

00:50:12.940 We don't know that yet. You're right at the cutting edge of the work in my lab. One prediction

00:50:17.380 would be that those chemical changes map to those genes that we see change, but we haven't

00:50:23.800 overlapped those data sets yet. But it's the next exciting possibility. One thing that we do see is

00:50:29.140 that if we stress a cell out, let's say we break its chromosome for a day, the cell actually has these

00:50:35.020 hypersensitive regions that will open up and stay open. And that's evidence for epigenetic noise.

00:50:42.240 And so what I think is going on is that cells in their daily response to broken DNA and other

00:50:48.420 things like UV damage when we go to the beach and don't wear sunglasses, that back and forth of,

00:50:53.480 oh, we've got to go put out that fire. We've got to go repair that chromosome has led to these

00:50:57.200 regions of the chromosome that are sensitive to opening up. And once they're open up and the

00:51:02.380 methyl groups have removed or been added in the wrong place, they stay that way. Unless we reprogram

00:51:07.880 them, somehow the cell knows, oh, that one screwed up. Let's go back to being young again.

00:51:13.080 So let's play devil's advocate for a moment. In clinical medicine, we're actually 20 years,

00:51:17.800 almost exactly, to the first time someone used one of these adenovirus vectors to treat a patient.

00:51:24.720 Wasn't it 20 years ago at Penn, Jesse? I don't even remember his last name now. It begins with a G,

00:51:30.000 I believe, blanking on this. But basically a patient died, right? And that sort of really,

00:51:34.960 really changed the trajectory of genetic engineering. And I always thought that as

00:51:40.700 tragic as that death was, it was a little bit of a distraction from what I always took to be was a

00:51:46.380 much bigger issue, which was how penetrant can you get these vectors? How ubiquitous can you get these

00:51:53.840 vectors? Let's use an example. If somebody had the belief that APOE4 gene was predisposing them to

00:52:01.980 Alzheimer's disease, which it is, right? It's a predisposition. It's not a fait accompli.

00:52:06.400 The likelihood that you could engineer a vector to swap every copy of APOE4 with a new copy of APOE2,

00:52:13.540 for example, seems very improbable. If you're doing that once they're an adult, I mean, if you're going

00:52:18.340 to have a shot at doing that, as there's reports of this in China, I haven't really followed this,

00:52:23.060 but there's all this talk about these CRISPR babies where they're engineering out LP little

00:52:27.520 a and things like that in embryos. But how technically difficult is this to do in a complex

00:52:35.160 organism like a mouse or ultimately a human where you need to change enough of the gene to actually

00:52:43.120 put the right program in? And I'm blanking on what the analogy is going back to Shannon's

00:52:49.900 observer. Because that looks sort of binary. It's sort of like receiver, operator, observer.

00:52:55.600 Okay. But here it's like, don't you have to get enough of the corrected version of the gene in

00:53:01.320 to actually get the optic nerve back? I don't know if my question makes sense, by the way.

00:53:05.200 It is binary in the sense that if a cell doesn't get the virus, it's not going to be reprogrammed.

00:53:10.040 So the current limit to technology right now is getting it into the cell.

00:53:13.640 It's how many of the cells can actually get the virus is really what I mean. Yeah.

00:53:16.820 At the cellular level, it's binary, but it's sort of analog across all the cells, right?

00:53:21.160 Yeah, it is. And that's the problem we face. That's why it's going to take a while to reprogram

00:53:25.580 an entire human being. These viruses like the liver, for example, and you don't want to

00:53:30.840 give yourself liver cancer just trying to reprogram your nose or whatever.

00:53:35.520 That wins the Darwin Award, right? If you're the person who does that.

00:53:38.700 I cured my wrinkles, but I died of teratoma. I would predict that at some point,

00:53:42.440 people are going to try and use this for cosmetic effects as well. We don't know

00:53:46.560 what we can rebuild. We know we can restore retina. We can regrow optic nerves. We don't

00:53:51.320 yet know how much we can reprogram an entire animal, let alone an entire human. But one thing

00:53:57.220 that has held us back in my lab is that we can't deliver these AAVs, these viruses, to every cell

00:54:03.880 in a mouse. Now, the good news is we may not have to. Maybe we only need to reprogram a quarter of

00:54:10.580 them and we get to be a lot healthier and younger. We don't know that yet. But in the eye, the reason we

00:54:14.940 chose the eye as well, besides the fact that we really like a challenge, is that the eye has

00:54:19.220 drugs approved for viral delivery. It's relatively insensitive to the rest of the immune system. So

00:54:26.220 even if you've got antibodies against, it's a protected organ. And then we've got the fact

00:54:31.480 that we can deliver a lot of virus and in fact, about half of the optic nerves in the back of the

00:54:36.480 eye. Can I get a vote on the next tissue to work on here? Sure. Prostate. You have a lot of the

00:54:41.340 similar features. You have a very immune privileged location. Every single man will get prostate

00:54:46.680 cancer in his life. What we can't figure out is which ones are going to die of it. Imagine you

00:54:51.540 could take the prostate gland of a man who's 50. So a 50-year-old has, I don't know, something like

00:54:58.600 a 40% chance of already having prostate cancer. Again, fortunately, most of those will not go on to

00:55:03.720 kill. But the prostate causes a lot. There's really no need for your prostate gland once you get beyond a

00:55:09.740 certain point. But rather than eradicate it, imagine just restoring it back to a young prostate and

00:55:15.380 getting rid of the cancer and getting rid of some of the hypertrophy and the other things that come

00:55:18.980 along with it. I mean, it just strikes me as a very discrete immune protected organ that you could

00:55:25.680 go after. I'm sure there's 10 problems I haven't thought of that a urologist is listening to this

00:55:30.720 sort of. Yeah. I think it makes sense, except from a business standpoint, where diseases that have

00:55:36.200 much more need than long-term effects. So you're saying that loss of vision is acute enough that

00:55:42.260 there's a reason to do something about it immediately as soon as you're experiencing

00:55:45.560 visual loss. Exactly. Or you're having heart failure. You've got no other choice but to try

00:55:49.540 this or you're going to die. Those are the early low-hanging fruit to help people. We need to figure

00:55:56.100 out, is this safe? There's some risk, right? There's no risk-free drug except for fish oil, which

00:56:00.920 we don't know anymore if that's highly effective. So there are risks. And with this technology that

00:56:06.660 we don't yet know what those risks are, we're going for the eye because it's not likely to

00:56:12.840 cause any problem. We've done long-term studies in the mice many months, no effect on anything

00:56:19.940 negative. If it was perfectly safe, I think the prostate would make a lot of sense. If we had a

00:56:25.100 drug on the market, a doctor could try that in a clinical trial off-label. But I think the future

00:56:29.300 looks bright if we can get one tissue or one organ be reprogrammed into a safe way.

00:56:35.200 So cardiac myocytes, how technically challenging will that be? And do you have a mouse model for

00:56:39.540 that yet? A heart failure mouse model? We have hypertrophy. That's the best model we have in

00:56:45.360 my lab. There may be better ones. The problem with hypertrophy is they retain contractility,

00:56:50.560 right? So it's sort of, you almost want a mouse model where they've lost some of the contractility.

00:56:55.620 That strikes me as a, I don't know, I could be wrong, but that strikes me as the easiest place

00:56:59.340 to try to figure this out. Right. Yeah. I need to research this. I wonder if

00:57:02.960 digoxin would be one way to go. Interesting. So what steps exist between this proof of concept in mice

00:57:12.580 to an actual human clinical trial, even in phase one?

00:57:17.480 Well, the good news is about an eye study is that you go straight to phase two.

00:57:21.360 Because you have enough other adenoviral vectors out there that they give you your safety

00:57:25.760 in phase one? Well, my understanding is that in the eye, it's a special condition where healthy

00:57:31.580 volunteers don't want their eye injected with virus. Ah, of course. So you're not going to do a

00:57:36.120 dose escalation in, although sometimes they make you do dose escalation in a diseased population.

00:57:41.000 For example, with cancer drugs, a lot of times the phase one is still done in cancer patients.

00:57:44.700 Right. Well, the advice that I've been given is that we could go to a phase two immediately.

00:57:49.700 And so the obstacle is surprisingly not getting into patient, doing the trial. It's actually

00:57:56.160 making enough virus. There's such a gold rush and interest in gene therapy that making these

00:58:03.600 adenoviruses can take a year. And it's about twice as much as it cost a couple of years ago.

00:58:08.360 Gene therapy has gone from... Wow. So that's kind of the opposite of what you think of with Moore's

00:58:12.340 Law on the sort of transistor side where more interest, more technology should get cheaper.

00:58:18.680 It will, but we're in that uptick of supply versus demand. And the demand is huge. And it's

00:58:24.260 partly because it's the hottest thing now. Being able to edit the genome, correct genetic diseases,

00:58:29.320 and now reprogram the body. This is a massively interesting area that holds a huge amount of

00:58:35.120 promise. And presumably profit if there's promise.

00:58:37.580 Well, otherwise you're not going to get people to put their hard-earned money behind it

00:58:40.740 unless they're very philanthropic. The point is that, I think you raised this earlier, I know you

00:58:45.120 raised it earlier. That the gene therapy used to be the pariah of medical treatments because it was

00:58:50.500 thought to be dangerous, risky, probably won't work. That's gone at 180. And now if you have a gene

00:58:56.380 therapy company, there have been a few that have been sold in the billions recently. These are

00:58:59.600 extremely hard. Everything that's new, breaking new ground, investors are all over it.

00:59:04.700 How do the vectors today differ from the vectors 20 years ago? And I don't even remember the story

00:59:09.340 particularly well of when this boy died. Presumably he died of sepsis or something sort of related to,

00:59:16.060 but not the direct proximate result of the gene therapy.

00:59:19.720 My understanding is these were different types of viruses that could integrate into the genome and

00:59:23.660 cause a mutation. And that led to problems, not least with standing up the good blood tumors.

00:59:29.340 The viruses today don't cause cancer. They don't integrate and they don't have any negative

00:59:33.900 side effects other than an immune reaction. How is that possible that this is not putting,

00:59:39.200 sorry, you're saying the problem used to be that it was putting its DNA in as well as a packaged DNA

00:59:44.900 that it was carrying with it? Yeah. That's my recollection that they had a propensity to

00:59:48.480 integrate into the genome. So where is gene therapy today with things like sickle cell and some of the

00:59:53.380 really obvious, like if you're thinking about what's the poster child for gene therapy,

00:59:58.120 it's sickle cell, it's thalassemia, it's cystic fibrosis. Like you can rattle off 20 diseases that

01:00:05.600 are tailor-made for this because they're single known gene mutations. Yeah. Well, people who are

01:00:11.300 interested in this should Google it. Clinicaltrials.gov. Clinicaltrials.gov has a bunch of these.

01:00:15.280 Oh yeah. Many, many. This is a great example of how asleep at the wheel I am. It's like my day job

01:00:20.400 to pay attention to medicine, but you get siloed into one thing that I'm interested in. And I'm,

01:00:24.520 my knowledge of gene therapy is 18 years old. It's kind of embarrassing. I'm embarrassing myself as I,

01:00:30.400 the more I'm talking, the more I'm embarrassing myself. Don't be so hard on yourself. Even I can't

01:00:33.540 keep up. My head is spinning and this is my day job. I absorb a lot of this because it through

01:00:37.540 osmosis, I'm at meetings and whatever, but every day I have to read or at least skim 50 papers just

01:00:43.380 to keep up. And in my area, let alone someone else's. So there's no way you or certainly anyone who's

01:00:51.240 not following this for a living can keep up. So you're saying that right now there are active

01:00:55.700 clinical trials ongoing for people with these really obvious candidate genes.

01:00:59.760 Sickle cell is a company that looks really promising, has clinical trials I believe in

01:01:03.540 progress. So yeah, I think it won't be long, maybe just a few years before these diseases

01:01:08.620 are correctable. Now that cost a lot, and this is probably a topic we shouldn't jump into because

01:01:13.380 it's a complete diversion, but because they're one shots, then to recoup the cost that it cost,

01:01:20.120 these are extremely expensive. The spark therapeutics drug that treats a type of retinal

01:01:24.760 degeneration, it's one or two injections. I think it's two injections, but it's in the

01:01:31.480 high hundreds of thousands per treatment. That's a lot of money and some politicians are up in arms

01:01:36.340 about that. So companies are getting more innovative in how this might play out. And I'm aware of one

01:01:43.440 biotech that said, if in five years you're not cured, you don't have to pay for the treatment.

01:01:48.580 It's interesting. I've also heard of models, this sounds even more controversial, where the cost

01:01:54.940 of the treatment is basically taken out as future earnings of an individual whose health

01:02:01.400 is restored. So you're sort of tying it back to a gain in productivity. So if someone who can't see

01:02:09.160 has their vision restored, well, they're going to presumably be able to make more money or do X,

01:02:16.480 Y, and Z in a more productive fashion over the remainder of their life. And a portion of that

01:02:20.520 gets paid back to remunerate the cost. I mean, I got to say, I am really glad that that's not the

01:02:26.340 problem I have to solve because I can really empathize with both sides of these debates.

01:02:31.340 I don't, it's very difficult for a drug company to sort of find the motivation to do these things

01:02:35.500 without some clarity around how these things can be priced. And at the same time, it seems criminal

01:02:42.600 to say it's going to cost a couple hundred thousand dollars for someone who's born with

01:02:46.720 sickle cell anemia to be free of sickle cell anemia. I mean, I'm glad that smarter people

01:02:51.220 than me get to figure that one out. I want to go back to this thing because I'm still sort of

01:02:55.120 wrapping my mind around this idea of the Horvath clock. So two totally unrelated thoughts. The first

01:03:01.660 is in the short term, can we use this as a way to measure our progress with the interventions we

01:03:09.320 have at our disposal? So remember a few months ago, you and me and Nir were sort of hanging out

01:03:14.060 in Boston. I don't even remember what we were doing. We were just sort of talking about some

01:03:17.600 stuff, but I made this argument that we already have some pretty amazing quote unquote drugs out

01:03:25.380 there. There's the really obvious ones like rapamycin that are actual drugs, but then there are other

01:03:31.320 quote unquote drugs like exercise. Exercise is a very potent drug. Fasting is a very important drug.

01:03:37.680 If you want to use that terminology loosely, but we still don't really know how to dose these drugs

01:03:42.420 that well. And again, with the case of rapamycin, I talked a little bit about my use of it. My reading

01:03:47.440 of the literature says pulsatile use of rapamycin, probably the right way to do it. Constitutive use,

01:03:53.120 probably not. But truthfully, should it be pulsed every three days, every five days, every seven days,

01:03:58.420 every 10 days? At what dose? Not really sure. Fasting, you and I have talked about this all day long,

01:04:04.560 could come up with an infinite number of permutations and combinations for how one should

01:04:08.780 fast. And my fear is we won't get really good answers in these because one, the biomarkers we

01:04:16.000 have today are far too crude to really tell you what's going on. Even slapping a continuous glucose

01:04:23.440 monitor on 24 seven doesn't come close to giving you this insight. Looking at IGF, in my opinion,

01:04:29.120 certainly doesn't give you this insight. We need something deeper to your point, right? We need to go

01:04:33.180 deeper, deeper, deeper. And because these interventions are not essentially profitable,

01:04:38.420 there's really no great incentive for the biomedical community to be studying them. But yet those are

01:04:45.340 some of the best interventions we have. And for many of us, we'll never have the opportunity to have

01:04:49.700 an adenovirus shoved down our retinal cavity to fix our eyes. It's going to be, hey, how do we eat?

01:04:58.080 How do we sleep? How do we exercise? And how do we take drugs that are currently available? So do you

01:05:04.740 think there's an opportunity to do this, to use these clocks, to look at the extent of methylation

01:05:10.660 and epigenetic change within our DNA to, as you said, even though none of these things are likely to

01:05:16.460 reverse it the way the intervention that you described is, if we study rates of change,

01:05:22.240 that could be a great first order proxy.

01:05:23.960 Right. But it's still not going to tell you within a month if what you've just done is working.

01:05:29.240 If every year a person had a look at their clock and you could say, hey, David, I'm sort of making

01:05:36.340 this up, but David, since I saw you last year, your genetic clock sped up nine months relative to the

01:05:43.520 12 months of chronologic aging that you've undergone. Keep up the good work. I can't tie it to what you've

01:05:49.200 done, but would I be able to at least directionally say whatever you've done in the last year

01:05:52.740 has been directionally correct versus if you showed up after the following year and it said,

01:05:58.860 oh gosh, David, you've aged two years epigenetically. Something's not good. Now you might say, yeah,

01:06:05.160 I got a divorce and got fired. Well, okay, no guff. It's not good, but that doesn't offer much help.

01:06:10.460 But anyway, I'm just sort of thinking about this through purely selfish reasons. I'm going to be

01:06:13.480 completely transparent. I just want to know what to do.

01:06:15.380 Yeah. I think the clock gives us the ability to do that. We did it with telomeres, but the field did it

01:06:19.900 with telomeres, but it wasn't as accurate as this new clock. So you could do that. You could every

01:06:25.260 year, even every six months, do your DNA methylome, the Horvath clock, and have a look at your rate of

01:06:32.300 change. If you've got stored blood samples, you could go back and see what that change was back

01:06:37.980 in time. It'd be interesting. Everybody could save a blood sample and go back in time too.

01:06:43.040 Are there any companies that are doing this?

01:06:44.560 There are, they just have sprung up. I can't remember their names off the top of my head,

01:06:49.260 but you can find them on the internet. They'll tell you your DNA methylation age. Yeah.

01:06:54.980 Interesting. I was actually just looking at something totally unrelated to this,

01:06:58.820 but equally outside of my wheelhouse the other day, which was, it was looking at something around

01:07:03.600 endogenous versus exogenous AGE formation. And I was like, how has no one come up with a company

01:07:09.320 to measure this? Because that's the crux of everything, right? It's not about,

01:07:12.520 people get so phosphorylated about how many AGEs they're eating and they don't realize it's the

01:07:17.140 endogenous production of AGEs that are far bigger issue. This is like screaming commercial application.

01:07:23.280 Well, yeah. So Steve Horvath and I were approached by someone who wanted to start a company

01:07:26.840 that would measure your age and then tell you what supplements to take. And he and I didn't believe

01:07:33.140 that the science was rigorous enough yet to say what would correct the clock. But those experiments

01:07:38.260 that you're talking about could actually do that.

01:07:40.040 So the second thing I want to come back to on this is something I think we may have touched on

01:07:44.220 very briefly in our first podcast is around senescent cells. And this is one of those things where you

01:07:50.580 see the picture on the front cover of science or nature and it's mind boggling. You've got this

01:07:54.920 old mouse, this old decrepit mouse sitting right next to what looks like a spry, plump, young,

01:08:01.980 beautiful mouse. And the punchline is, guess what? They're the same age,

01:08:05.960 but in the beautiful young one, we took this subset of cells called senescent cells and we

01:08:13.000 killed them. So maybe explain what that is at a high level and how it overlaps with or differs from

01:08:21.160 everything you've just described with respect to the methylation clock.

01:08:24.380 Well, if we just for a second go back to the jostling of the epigenome and the noise that's

01:08:29.420 introduced, what we've shown in my lab is that creating this noise and the way we do it in my

01:08:35.500 lab, we like to use a broken DNA to distract the proteins. What we see is that the early stages of

01:08:42.260 aging due to this epigenomic noise leads to the loss of cellular identity. The very end stage of

01:08:47.540 that process is that the cells check out of the cell cycle so they don't divide anymore,

01:08:53.060 but they don't die. They just sit there and they're stuck in this emergency state. There are other

01:09:01.500 things that can cause senescence loss of the ends of the chromosomes. The telomeres will also cause a

01:09:05.460 cell to say, whoa, got a real problem here. Got a broken piece of DNA at the end of my chromosome.

01:09:10.780 Let's shut this down before we become a tumor. And it's thought that these cells, these senescent

01:09:16.440 cells are really important to prevent cancer from taking off because they shut themselves down and

01:09:21.200 they stay like zombies in our tissues for decades and they don't die. The problem is that they don't

01:09:28.580 just sit there. They're actually, they're sort of poisoning the well. They're in a state of stress

01:09:32.760 and they're saying to these cells around them, oh my God, I'm panicked. You guys should be panicked too.

01:09:37.680 So they, they send out chemicals and proteins that stress the other cells. The other cells are now

01:09:42.880 in a panic state and their epigenome, I believe is getting disrupted and accelerated as well.

01:09:48.620 So if I'm hearing you correctly, you're saying the senescent cells can be part of the cause of

01:09:54.660 the methylation and epigenetic interference with the non-senescent, presumably dividing cell or active

01:10:03.680 cell. Yeah. And so there's some evidence for that. Like I said, we can disrupt the epigenome,

01:10:07.680 they get senescent. Those senescent cells you can now put next to normal cells and they will

01:10:11.740 induce senescence or cancer in those cells or make them tumorigenic, we call it.

01:10:16.700 The other experiment that was beautiful was done by Jim Kirkland at the Mayo Clinic. He took some

01:10:22.160 senescent cells and he implanted them a little bit in a little dab into the peritoneal cavity,

01:10:28.140 the lining of the gut under the skin. And those mice ended up having signs of premature aging,

01:10:34.940 higher blood sugar and other things. So a little bit of senescent cells goes a long way. And that's

01:10:39.440 what's scary because if you take fat from a young and an old mouse or a young and an old human,

01:10:45.340 we can actually stain them. We can color them whether they're senescent or not. And when they

01:10:51.600 turn senescent, we can stain them blue. There's an enzyme they make called beta glycosides. If you

01:10:56.540 stain young fat, it actually looks white. But if you look at middle-aged, it's pale blue and an older

01:11:03.740 mouse or an older human, 50 years old, my age now, it's dark blue. It's packed with these senescent

01:11:10.100 cells in the fat. And no wonder the rest of the body is in a panic state if these fat cells are

01:11:15.560 now sending out the emergency screaming signal. What is the phenotypic identification of a senescent

01:11:20.580 cell? Well, that's still debated at conferences, but we all agree that they turn blue, that beta

01:11:25.940 glycosides. There are other genes that come on that signal DNA damage, one called P16, one called P21.

01:11:33.520 These are genes that cause cells to check out of the cell cycle and stop cancer. If you lose these

01:11:38.100 genes actually, it predisposes you to cancer. It makes sense. There are other issues, which is

01:11:42.960 that some senescent cells don't have that particular signature. And there are other cells that are

01:11:46.920 non-senescent that do have those signatures, presumably, right? Not so many. Well, there are

01:11:50.940 some. There's a mouse that was made by Ned Sharpless, who's now the head of the FDA. So he made a mouse

01:11:56.680 that would fluoresce with a firefly luciferase, a glowing mouse that was under the control of the P16 gene.

01:12:05.560 So that if P16 came on, that cell or that tissue would glow green. So we had that mouse. We found

01:12:11.400 that if a cell or a tissue got stressed, let's say a mouse had an infection or it got damaged or was

01:12:18.060 nibbled on or for some reason was just stressed out, the P16 gene, that fluorescent signal came on.

01:12:24.700 So that you're right, there are other things that can turn this on. So it's not definitive.

01:12:27.720 There is actually no definitive way to tell senescence in a tissue versus none besides

01:12:34.780 this blue stain, which seems to be pretty good. Wow. Going back to the third group of mice that

01:12:40.920 you worked on, this was the one that was just older. Presumably it had senescent cells.

01:12:47.020 It had senescent optic neuron cells. They weren't senescent yet. I see. So you'd stain those. You knew

01:12:53.180 that you didn't have any senescent cells that were contributing to the visual deficits that

01:12:57.980 were slowly accumulating. We looked because I was curious as to whether we saw that. I don't

01:13:02.900 believe we saw senescence. The reason that it's also highly unlikely is that we don't know how

01:13:07.300 to reverse senescence, even with reprogramming. Well, so that's exactly the question I was going

01:13:10.800 to ask, which is if you saw them, did they change in the presence of the reprogramming or did they

01:13:16.080 still stay there? That's the next experiment. Actually, in my lab, we need to know if that's true.

01:13:19.780 There's a postdoc listening to this who's cursing right now, because I think I rattled off a couple

01:13:23.800 of next experiments for your lab, right? Well, there's 30 people in my lab, so we can split the

01:13:28.620 work. That's good. But yeah, it's an important question. But what I think is going to turn out

01:13:32.540 is that if you're pre-senescent and you've just lost your identity, then that's reversible. We see

01:13:40.240 that in the lab, in the retina. But if you've clicked over into this zombie state, then you're in this

01:13:46.860 state that may be possibly never reversible. I would never say never, but that's a lot more

01:13:53.060 difficult. So what does that mean for the future? How does that fit into your CD analogy? So if you've

01:13:58.820 got a scratched CD and the scratches are being caused by a senescent actor that's out of the

01:14:04.660 replicating pool, but is poisoning the well, you're telling me you can buff out the scratches in the CD,

01:14:11.060 but you can't get rid of the scratcher. And you're accumulating more and more scratchers over time.

01:14:15.380 You are. But theoretically, it should be possible to even get a senescent cell to grow again

01:14:20.440 because it should have all of its DNA still there. Now, there are issues. If it's lost its telomeres,

01:14:25.320 it won't like being brought out of senescence. Or it's senesced because it's full of mutations.

01:14:29.300 That can also cause senescence. So there are reasons why you wouldn't want to get a senescent

01:14:33.000 cell to start growing again. You might cause cancer. But there are other reasons cells check out

01:14:38.140 that isn't due to loss of information. So we've got a pretty badly scratched CD that we can polish

01:14:44.220 with reprogramming. But if you've gouged it so deep that even reprogramming can't work,

01:14:50.240 then we need something else probably. But I wouldn't say for sure.

01:14:53.440 So this is now, let's get into sort of a little bit more of the sci-fi speculation.

01:14:57.560 Where do you think would be the ideal application for this clinically in terms of in humans,

01:15:04.120 explicitly for the purpose of longevity and not disease treatment? So we've already talked about

01:15:10.840 lots of applications on the gene therapy side. Is your belief that that is the only application,

01:15:16.820 that using reprogramming as an anti-aging tool is a precise tool that goes after the specific

01:15:24.820 anti-aging phenotypes? Oh, look, your skin is more wrinkly and more saggy. Boom, here's a virus.

01:15:31.860 Ah, your vision is deteriorating. Boom, here's a virus. Ah, your heart muscle doesn't pump as hard.

01:15:36.860 Boom, here's a virus. Versus a global approach that says, no, we're going to go right to the master

01:15:44.520 CD and somehow we can restore your methylation pattern to that of you at your birth. Which of

01:15:52.980 those two paradigms do you think is the sort of quasi sci-fi, but quasi, if you had to guess,

01:15:59.520 predicted approach? Well, I think within our lifetimes, we'll see the first one.

01:16:02.700 The glaucoma patients are waiting for this. We're working on starting a clinical trial,

01:16:08.100 hopefully within the next 18 months. So this isn't as far away as you might think. How long

01:16:12.380 before we can treat other diseases? I think that it's going to depend on the severity of the disease,

01:16:17.000 the FDA. But I could imagine within this decade, there are multiple diseases, heart, maybe not skin,

01:16:22.620 I don't know about that, but other severe disease will be tackled either one by one. But we're not

01:16:27.940 just waiting for that. In my lab, we've already dosed mice with the virus intravenously to see what

01:16:32.580 would happen. The good news is that they're all still alive. They're all still happy.

01:16:36.020 No evidence of cancer. What's the expression pattern?

01:16:38.320 Ah, well, this is the issue. Majority is in the liver, some in the gut.

01:16:42.940 That's an adeno feature, isn't it? Adeno really loves the liver.

01:16:45.980 Yeah. If somebody had an adeno that got into every cell evenly, we'd be set. That science fiction

01:16:50.720 future could be within our lifetimes where we get a dose when we're young of adenovirus. We get to age 40

01:16:57.800 and we're starting to experience the signs of aging. Our eyesight isn't as good as at night. We have

01:17:02.720 to hold the menu a bit far away. I'm there, by the way. Yeah, me too. I just, I hit it, man. I hit it

01:17:07.940 at 45 and a half. I can't believe it. It happened overnight and it's exactly the scenario you

01:17:15.120 described. It's the restaurant, tiny print, dark. Well, you're losing your vision. If you had been

01:17:20.860 infected with the adenovirus, your doctor or your doctor, you could have someone prescribe a course of

01:17:25.260 doxycycline for a month. And if we're right, you'll get your vision back and you'll get who

01:17:31.200 knows what back. That's amazing. It's funny with all of these incredible, as you describe it, this

01:17:36.860 huge re-uptick in the interest around gene therapy, I have to believe some of these companies are

01:17:41.980 looking for better vectors as well. I mean, if the fact that I remember adeno has a predisposition

01:17:46.240 for the liver, that tells you it's 200 years ago's news. Are there other viruses that appear more

01:17:51.000 promising? Well, there are dozens in academia now. They come with greater risk, of course, right?

01:17:55.080 No, these are slight variations on what's currently used in that clinic. And different companies are

01:18:00.100 using different vectors. So AV9 is good for muscle. So companies are going for muscular dystrophy with

01:18:05.460 AV9. AV11 and 2 are good for the eye. There's a menu now of dozens. But they're basically all still

01:18:11.640 adenoviruses. They are. There are slight tweaks on the proteins that tell the virus where to go in.

01:18:16.460 But what I hear is that some companies, I forget if it's Roche, but that's my recollection,

01:18:21.720 has made millions of different varieties. And if that's true, we may be able to choose any tissue

01:18:28.600 we want. In other words, we're not getting away from the tissue-specific paradigm.

01:18:32.760 Well, eventually we will. Why not? I mean, if it's on the market, what's going to stop somebody

01:18:37.220 from trying this anyway? Because I think it gets to a theoretical question, which comes down to,

01:18:42.200 you sort of alluded to it earlier. Why do I need a new car eventually? Can I just keep replacing each

01:18:47.900 individual failing part? Or at some level, do I need a new car? Is there some final thing that it

01:18:53.920 becomes impossible to replace? Like the chassis just, especially you're in Boston, like us Californians,

01:19:00.780 we don't, our chassis never rust. But at some point you could replace the engine and it's not enough.

01:19:07.860 What are you going to replace? The axle. Replace this, you're going to replace that. So if you take this

01:19:12.100 organ-specific approach, the skin, the eyes, the heart, the lungs, the brain, is there something

01:19:19.320 else that ultimately is going to lead to our demise? Or is that effectively just the accumulation of

01:19:26.200 enough senescent cells that the gouges in the CD become so deep that even an organ-specific approach

01:19:33.120 ultimately fails? Where I'm really going with this is, is there even theoretically an argument for

01:19:38.840 cellular immortality? Theoretically. This is as close as we've come to finding a way to actually

01:19:44.540 live for thousands of years. I don't know about immortality. I think that the problem with

01:19:49.120 what I'm calling the information theory of aging, which is what I wrote about in my book,

01:19:53.680 is that we do lose information. Every cell does experience mutations. It's not perfect.

01:20:00.360 Though I know I can take some of your cells and clone you and make a young version of Peter,

01:20:05.240 but it doesn't work for all cells. And so ultimately, if you're a thousand years old,

01:20:10.420 you may have lost a lot of the genetic information. But epigenetic information,

01:20:15.440 because there's this backup drive, this observer that we have found exists in cells somehow that

01:20:21.760 you can tap into, as long as the genome, the DNA strands are still largely intact,

01:20:26.920 we can reverse aging. But it's an information loss issue. So the gouging that we get may scratch

01:20:34.840 some of the foil in the CD or the DVD and you lose a little bit of the song, which you'll never

01:20:39.880 get back. Now, that said, if I was George Church, who's in my department, he would say,

01:20:44.820 no big deal. By the time a thousand years goes by, we can replace anything. And that's probably true.

01:20:50.740 Whole system, regrow bones. Why not grow a whole new heart, put it in? That's going to be doable.

01:20:56.220 So what I think the future holds is the following. A lifestyle where you're monitoring yourself

01:21:00.520 with devices, devices tell you what to eat, when to eat. If you want to pay attention,

01:21:05.060 you don't have to listen. But when there's a problem, you're notified. You've got a tumor

01:21:08.960 somewhere in your body. We've detected it. Go have that killed before it grows. That's going to be 20

01:21:13.380 years ahead of what we can do now for patients. That'll keep you young and healthy for a lot longer.

01:21:19.200 In the meantime, we'll learn, thanks to guys like you and clinical trials, what to do to live longer,

01:21:25.860 whether it's the perfect exercise, the perfect combination of diet.

01:21:30.080 But is that necessary? I mean, if you really stop to think about it, couldn't you just make

01:21:34.000 the case that all of this nonsense that people like me do and all of this ridiculous effort that

01:21:39.640 goes into fasting and rapamycin and this exercise and sleeping and all of this fun killing activity of

01:21:47.760 my existence, if I could reprogram, why would I do any of this?

01:21:51.860 Well, yeah. I mean, who could argue with that?

01:21:53.320 I mean, like I could literally go and get a biggest, most beautiful pizza burger imaginable

01:21:59.940 right now and not worry about any of this stuff.

01:22:02.960 Right. That's if it comes true. So, so far, so good. You know, if you can restore vision,

01:22:07.160 I'm sure you can restore a lot of parts of the body. Let's say reprogramming doesn't work that

01:22:12.000 well. And you say you get senescent cells, your organs eventually lose their information in their

01:22:16.980 genome. Then what? Well, you can delete senescent cells. You can take someone else's organs or

01:22:22.880 grow your own in the dish or have a pig grow your own. With all that put down, if that all worked,

01:22:28.580 I would challenge anybody to say that that wouldn't allow people to live a lot longer.

01:22:33.280 There's still people out there who say we're never going to make it past on average 80, 85,

01:22:38.840 let alone to 100, let alone to 120.

01:22:41.500 Well, I don't know. I think it's much easier to imagine an upward movement to 100, for example.

01:22:48.360 I mean, look, that's sort of my point of view, right? I think genetically, I'm probably engineered

01:22:52.200 to stick around until I'm in my early to mid 80s. Again, this is just looking at my parents,

01:22:57.260 which once you get over 80, your genes become a far bigger predictor of your longevity than

01:23:02.480 in your 60s through virtually uncoupled. I feel like I've got a great roadmap on what it means to

01:23:07.440 get to be 100, which is still stochastic. There's no guarantees, but it's like, how would you stack the

01:23:12.400 odds in your favor kind of thing? But it's an entirely another animal to imagine a world where

01:23:17.000 you can take individuals and even get them to be 200. That's a really big leap.

01:23:23.080 And I would have said three years ago, it's impossible.

01:23:27.000 What did you think three years ago would have been the limits of our technology? And that's again,

01:23:33.160 three years ago, you were thinking, I could give you more NAD. I could give you more things to

01:23:39.760 activate your sirtuins. I could tweak your mitochondria this way versus that way. Based

01:23:45.000 on that level of manipulation, what were the limits you thought we were...

01:23:48.200 So my thinking was, having come from the calorie restriction world, that animals that are

01:23:53.240 calorie restricted live at best 30% longer. And they're healthy, which is great.

01:23:58.640 And the CR mimetics like rapamycin and metformin, they can, depending on how sick the animal is,

01:24:03.840 but let's say even rapa in a non-sick mouse can give you 30% more life.

01:24:07.860 Right. So as long as we're like a mouse, we could live 30% longer. So, you know, 30% of 80.

01:24:14.860 It's a big deal. That gets you to a hundred right there.

01:24:17.120 Yeah, that's a hundred. That's why I thought someone who does all these things has a better

01:24:22.460 chance of reaching a hundred than ever before. But what I didn't take into account in those

01:24:27.200 numbers, most people don't when they think about this is if we make it to a hundred. Okay. So that

01:24:32.000 means I'm still alive in the year 2069. What technologies do they have in 2069? Is reprogramming

01:24:41.220 a common thing? Probably it will be. Yeah. So in other words, this is the optionality play.

01:24:45.760 It's if you add 20 years of life, extending someone from 80 to a hundred, you have to take

01:24:52.660 into account the probability that things come online during that period of time that can also

01:24:57.800 impact the very variable that you're trying to manipulate. Yeah. And already every month that

01:25:02.640 we stay alive, we get an extra week of life. That's how technology is going currently.

01:25:05.820 Wait, say that again. Extra month of existence. You'll be able to live an extra week.

01:25:10.680 Come on. That seems too good to be true. That's a 25% plus up. That can't be right.

01:25:15.620 I'll check on it. Did I hear you right? I'll tweet that out if it's right. But that's what I

01:25:19.260 recall hearing. I'm only questioning it just based on it seems too good to be true.

01:25:23.440 So every additional month of life is offering a week of additional life extension just based on

01:25:29.280 the technologies associated with it. I mean, because there's no evidence of that to date,

01:25:33.660 is there? Because we really have seen a compression of life expectancy over the last two years. Haven't

01:25:38.880 we actually seen? So up until I think, again, I don't want to, I'm probably misquoting this,

01:25:44.500 but directionally, I think this is right. Up until about 2015, life expectancy was increasing at

01:25:50.680 about 0.4% per year. That's now crested. I don't believe that in this environment that we live in,

01:25:58.600 in other words, we have figured out a way to eat, stress, and not exercise our way out of all of the

01:26:06.300 technological benefits that have come our way. And that's, I think, why you're seeing this sort

01:26:10.760 of cresting, right? Where we've solved all of the infectious problems that gave us most of our

01:26:15.360 longevity gains way back in the day. We figured out that you shouldn't drink out of your sewer and

01:26:21.140 we've learned to wash our hands and we've got great antibiotics. But these chronic diseases that

01:26:25.860 are killing us now, the force that's driving them, which I think is basically food, sleep,

01:26:32.980 lack of exercise, stress, et cetera, et cetera. I feel like those things are weighing down on us more

01:26:37.360 than modern medicine is giving us tools to fight back. And at the very least, they're at a standoff.

01:26:42.280 Certainly in the US, that's true. It's not true for all countries.

01:26:45.520 That one week to a month thing might be a non-US stat then.

01:26:48.800 Oh yeah. It's a global increase in the maximum lifespan.

01:26:52.160 So that might take into account, again, I don't follow this research closely enough to say it,

01:26:56.300 but it might be that, well, you're getting more vaccines in the hands of people who are

01:26:59.780 otherwise unvaccinated and getting fresh water and food into people who otherwise don't have it.

01:27:03.880 Is that, do you think?

01:27:04.360 It's not. No, I have to correct what I said because it's very important. It's a graph of the

01:27:09.420 average lifespan of the longest lived country at the time. So Japan has been leading that 0.76

01:27:14.700 for the last decade or so. So that's important. It means its average lifespan is increasing

01:27:18.500 at the top end. So if a country uses all the new technology and has good healthcare and people

01:27:24.120 don't eat themselves to death or take opioids, that continues to march up. What doesn't seem to

01:27:30.020 change, in fact, if anything has plateaued or reversed, is the maximum lifespan of humans,

01:27:35.560 which if you believe Jean-Calman, she was 122. People debate that, but some people have made it

01:27:41.900 117, 180, no question about it. But that seems to be our current limit. But did those people take

01:27:47.040 care of themselves? Absolutely not.

01:27:48.340 No, not at all.

01:27:48.960 They're the opposite.

01:27:49.780 Yeah. Nir has done great work on that topic.

01:27:51.840 They smoke and drink themselves to an untimely death at 117. But what if they had access to the

01:27:58.240 knowledge that we have now, which is lifestyle, some of these medicines that we think can help?

01:28:02.640 Well, maybe they would help those people to get beyond that. But to really go beyond that,

01:28:08.080 I think you need something really new. And that's why I'm a lot more optimistic than I was having seen

01:28:14.740 what reprogramming has in terms of potential to be able to not just slow the clock down, which is

01:28:20.160 inexorable, seemingly inexorable, but now actually get cells to go back in time.

01:28:25.000 Well, I mean, it's interesting. I could talk about this for a lot longer. I think at this point,

01:28:29.680 I mean, I think the readers will, I think, enjoy your book greatly. I've deliberately avoided asking

01:28:34.260 you, I think, some of the questions that are also on my mind about what are the implications of this?

01:28:38.960 I think I get asked these questions a lot and I just defer. I just punt. I just say, look,

01:28:44.320 I'm not even trying to solve a societal issue. I'm interested in the longevity of the individual.

01:28:50.520 And that's a hard enough problem. So that's kind of the one that I want to think about and put all of

01:28:54.720 my energy into. But in your book, you really do actually try to ask the broader question, which is

01:29:00.360 what is the implication of a society where people can live to 200? It would change a lot of things. So

01:29:05.600 I'll let the readers either hear you on other podcasts to talk about that or do what's even

01:29:10.600 better and just actually read the book themselves. But I want to kind of come back to some things that

01:29:15.280 we touched on really briefly in our first discussion that I've had so many follow-up questions on.

01:29:20.620 And so if you still got a little, you don't have to go back to Boston tonight, do you?

01:29:24.300 No.

01:29:24.820 Okay. So let's talk a little bit about your own personal habits around stuff. So

01:29:29.220 let's start with metformin. We've talked a little bit about metformin. There are a couple of papers

01:29:35.120 that have come out kind of recently that have suggested, hey, maybe metformin in the metabolically

01:29:40.400 healthy person and or the person who's exercising is either not effective or potentially blunting

01:29:45.460 the effects of that. How are you reading those papers?

01:29:47.940 Well, I thought what you wrote online was excellent. I think about the same way, which is

01:29:51.880 we've known that one of metformin's main effects is to quote unquote poison mitochondria. It inhibits

01:29:57.640 a group of proteins that generates energy in mitochondria. And the response of the body when it has a bit

01:30:04.200 of inhibition of this is to say, wow, I'm low on energy. Let's build up those factories. So

01:30:10.240 mitochondria are often called the power packs or the battery packs of the cell. They generate chemical

01:30:15.320 energy. If you're young or you exercise or you calorie restrict, you'll have more greater area

01:30:20.420 or activity of these mitochondria. So more is better in general for humans. More mitochondria is

01:30:25.020 good. And as we get older, we lose that ability. Metformin slightly inhibits that activity and the

01:30:30.780 response is to make more of them. Problem is that if you're constantly inhibiting those mitochondria,

01:30:37.180 that's not going to be seemingly helpful with this new study for building up mitochondria

01:30:42.220 after exercise. Now you could argue that maybe you don't need to have more mitochondria after

01:30:47.500 exercise, but I think you probably would benefit from more mitochondria. So what you suggested that

01:30:52.620 I think makes a lot of sense, though we need to prove this or at least test it, is that if you're

01:30:58.440 exercising, don't take metformin on the days where you do intense exercise, maybe the next day after

01:31:02.940 to let your body recover and build up mitochondria. This leads to something that I think is a very clear

01:31:09.140 theme in all the work that I've done, the work that others are doing and what you've been talking

01:31:13.960 about, which is pulse your biological stress. Put your body in a state of anxiety or fear, adversity,

01:31:21.000 but you don't want to do it all the time. Your body needs a chance to recover. If you want to take a

01:31:26.120 supplement, maybe don't take it every day. Same with rapamycin. You don't want that on all the time.

01:31:30.780 In fact, I wouldn't take rapamycin, certainly if I was exercising, because it's going to tell the cell

01:31:35.060 to hunker down and not grow. And you may not even heal after exercise as well with rapamycin.

01:31:41.260 So I think that the view that the combination of hunker down fast, but then exercise on alternative

01:31:48.160 days or take the supplement and exercise on alternative days, they make a lot of sense.

01:31:52.700 Do you think anybody's going to be able to probe this? I mean,

01:31:55.500 NEAR is working on obviously getting TAME funded, and that's obviously asking a slightly different

01:32:01.640 question that's really going after sort of a non-diabetic population. But is it going to be

01:32:07.280 able to look at this? Do you think it will be able to tease out this issue that we're talking about? Or

01:32:11.680 is the study not going to be structured to be able to answer this question?

01:32:15.060 I've not heard of anybody who's testing that directly. Usually there's just one variable.

01:32:19.680 And has this changed the way you take metformin?

01:32:21.960 It has, but I need to put a caveat is that I don't take metformin regularly anyway. I need to find the

01:32:28.060 right time to take it. And before I was taking it when my stomach felt in good shape. And you can

01:32:32.220 tell when your stomach feels out of whack. Either you've eaten a big meal the night before, or you're

01:32:36.740 just not feeling right, a little bit of heartburn. Under those cases, I don't take metformin because

01:32:40.860 it does a number on my stomach, which is great if you don't want to eat, but also I'd prefer not to

01:32:45.100 always have a sore stomach. So I was already timing it. So now I just take metformin when I know I'm

01:32:52.000 going on a long trip and I'm not going to exercise. You know, I'm on planes and trains. That's a good time to

01:32:57.040 rebuild your body. And then if I'm at home and I'm exercising a couple of times a week, I'll lay

01:33:02.140 off the metformin. And then what about rapamycin? Have you ever revisited that? I did take it just

01:33:07.820 as an experiment, but haven't been taking it regularly. One of the things I do when I'm fasting

01:33:12.440 is I'm not taking those things, obviously. So any period of fasting longer than a day, those things

01:33:19.240 get stopped. Again, that's sort of a, an idea that's not sort of supported necessarily by

01:33:25.320 evidence one way or the other. When we last spoke, you were taking resveratrol. You noted that you

01:33:31.260 were taking it with sort of a fattier meal. Is that still something you're doing?

01:33:36.000 The evidence of resveratrol just continues to be good. It certainly does no harm. I do take it with

01:33:41.620 my tiny bit of yogurt in the morning, which I make myself. I mean, I grow yogurt at home. I miss that.

01:33:46.660 That does no harm. I don't have any negative effects. My cardiovascular system seems great.

01:33:52.120 Do you think that resveratrol, like where do you think it ranks in sirtuin activators? There's

01:33:57.680 others on the market out there. There's other supplements, even like terastilbene that are

01:34:02.220 sold that you can buy online. You very eloquently described the story. And if people haven't listened

01:34:08.100 to our first podcast, obviously this is a great opportunity to hit pause, go back and listen to it

01:34:12.820 because you talk about sort of the novelty of resveratrol and how it was sort of the first sort

01:34:17.320 of built for purpose, custom. This is what it should look like. Oh, let's go get it.

01:34:22.520 That's been over 10 years, hasn't it? Wasn't that like 2006, 2007?

01:34:25.540 We first showed that it activated the enzyme sirtuin in yeast and extended lifespan. That was 2003.

01:34:31.600 Jeez. Yeah. So do you still think resveratrol is the best if I, because this is something I

01:34:35.580 haven't done yet. I just haven't been able to convince myself that it's just one more thing I

01:34:40.700 need to add to my already complicated regimen. And if I wanted to start taking sirtuin activators,

01:34:45.500 would you recommend resveratrol? If so, at what dose? Or would you recommend I take something

01:34:48.740 different? And again, you can speak to me and you don't have to give advice to anybody else.

01:34:52.140 Well, yeah, I never give recommendations, but I continue to take resveratrol because it's cheap.

01:34:57.300 It's harmless as far as we know. But the evidence keeps stacking up that long-term is beneficial.

01:35:03.240 I mean, it's not going to cure diabetes. It's not as powerful in human studies or in mice as

01:35:08.540 rapamycin. No question. But does it extend the lifespan of a mouse that's eating a Western diet?

01:35:14.700 Absolutely. It does. That's been done many times. It almost seems like it falls potentially into the

01:35:18.840 metformin trap, which is the metformin data in metabolically unhealthy people. It's pretty hard

01:35:24.400 to argue that metformin is beneficial. Paradoxically, the people who are most obsessed with this stuff

01:35:29.940 are already doing so many of the other quote unquote good lifestyle things that you wonder,

01:35:35.460 is it possible that you're already doing such a good job of all the other things you manage with

01:35:40.040 respect to your health, that the resveratrol is neutral?

01:35:42.640 Well, maybe if you're optimized like you are, I think as good as it can get. But if you're elderly 0.98

01:35:47.520 and you're not exercising, you know, wheelchair, what else are you going to do? There is some data

01:35:53.000 in my lab that I'll share with you that we haven't published yet, but I think it's interesting to

01:35:56.380 mention. And I presented it at a meeting in Rotterdam last week for the first time to a big audience.

01:36:02.340 So let me just tell the audience quickly what resveratrol is. It's a plant molecule,

01:36:05.720 we'll get it from red wine in very small quantities, but the amount that we're giving

01:36:10.100 the mice and human studies, it's a lot. It's hundreds of times more than that.

01:36:13.640 So you can't drink your way into enough resveratrol?

01:36:16.420 No, but you know, a glass of wine.

01:36:18.140 But that hasn't stopped people from trying.

01:36:19.920 Right. The molecule, I take a gram of resveratrol in the morning. It's a high dose. It seems to be

01:36:23.840 fine.

01:36:24.040 What did the ITP study way back in the day as a human equivalent, or is it too difficult to make

01:36:29.240 that normalization?

01:36:30.400 So ITP showed that if you have a healthy mouse on a lean diet, it doesn't extend their lifespan.

01:36:35.100 Sorry. Yeah. And then in your lab, you took obese or unhealthy mouse on crappy diet,

01:36:40.840 extended lifespan.

01:36:42.060 Well, it depends where you start the diet. It was extensive. It was 20, 30% depending on how

01:36:46.500 you count it.

01:36:46.980 And the dose roughly was what?

01:36:49.300 We did two different doses. They both worked. One was 24 mg per kg. Another one was 240 mg per kg.

01:36:55.560 That's a big difference. That's a 10X difference.

01:36:57.660 Yeah. Right. And the lower dose was just as effective.

01:37:01.340 What did the log higher dose do that the lower dose didn't?

01:37:04.580 It kept the mice from gaining weight.

01:37:06.440 All right. So you're closer to the 24 mg per kg. No, I guess if you're taking a gram,

01:37:11.020 yeah. You're sort of in between those, right?

01:37:12.900 Yeah. It's on the high end. But the result is the following. We had a science paper published

01:37:17.780 in 2013 where we went to the effort of making or finding research for a mutation in the SIRT1

01:37:25.160 gene that we had published is likely the way resveratrol works. And that mutation blocked

01:37:31.900 resveratrol's ability to activate the SIRT1 enzyme. And that's been heavily debated and highly

01:37:39.660 controversial. It's one of the big controversies in my career. So we were forced, if not encouraged,

01:37:45.320 to do better. So we went back. We found this mutation that blocked the activation of this enzyme.

01:37:49.540 And if we're right, then resveratrol won't work if you've got this mutation in a cell. And we found

01:37:55.480 that was true. The drugs that were in development, the super potent ones, also blocked by this one

01:38:00.520 mutation. What the mutation did was it made an enzyme that couldn't be moved. It had a stiff elbow.

01:38:06.400 And without the bending of the protein at the elbow, resveratrol couldn't activate it anymore.

01:38:10.280 And we know this very clearly. It's been well published and cited. But here's the big experiment.

01:38:16.440 And it's been 10 years in the making. You take the mutation, you put it into a mouse, not just a cell,

01:38:22.340 a mouse. That takes a couple of years. It took us a couple of years. And now we have a mouse that

01:38:26.640 isn't normal. It's missing one amino acid in an enzyme that renders it-

01:38:32.220 Susceptible to resveratrol.

01:38:33.600 Well, worse, recalcitrant. It's immune to the effects of resveratrol in the test tube.

01:38:37.740 And now we could repeat our 2006 study of the high-fat diet with and without resveratrol

01:38:44.540 and with and without this mutation. And I didn't know this was going to work. In fact,

01:38:48.660 in the history of pharmacology, I don't know if anyone's ever found one amino acid change that

01:38:54.000 blocks a plant molecule in the diet. That's very difficult because in a diet, these molecules

01:38:58.500 and plants are hitting probably hundreds of proteins. But we made this one change. And now we

01:39:02.940 could ask the question, definitively, if you give a mouse resveratrol, which benefits

01:39:07.620 still occur.

01:39:08.660 Yeah. What do you attenuate?

01:39:09.720 But also which are off target, which are working through something else. That's just as interesting.

01:39:14.500 And I really don't give a damn anymore about what the answer is. I just want to know.

01:39:19.160 And so we did the experiment with my student pretty much definitively. I would like to hear

01:39:23.260 anybody who can disagree with this statement, that resveratrol extends lifespan by activating

01:39:28.500 SIRT1.

01:39:28.960 And it begs the question, can we apply that to ourselves, which is during those periods of time

01:39:36.860 when we are not fully dialed in, would we benefit from resveratrol?

01:39:41.340 Well, that's the reason I'm taking resveratrol is I don't exercise it.

01:39:44.060 Do you pulse it as well?

01:39:45.000 No. Unfortunately, you caught me out. I like to take it every morning. I found that it's been good

01:39:49.600 to me. Health is great. I'm doing fine. It's one of the longest experiments I've ever done,

01:39:54.300 probably the longest, but it's ongoing. And because I'm changing other things all the time

01:40:00.560 to see what works, what doesn't, I've kept that constant.

01:40:04.040 So I would love to chat about, we've got a little bit more time here, if you're willing,

01:40:07.700 is I still probably get more questions about nicotinamide riboside and NAD specifically

01:40:15.280 than almost any other molecule that's sort of out there. We talked about, I'm writing a book now,

01:40:20.880 and part of that book, there's an appendix in it. And in the appendix,

01:40:23.920 what I'm doing is writing a short section on sort of the drugs, supplements, and hormones that I

01:40:31.140 think are most interesting. And so I'm including, of course, something on NAD and NR. I think I've

01:40:37.620 identified 17 or 19 drugs, supplements, hormones that I want to address in this appendix. I would

01:40:43.360 say that I get more questions about NR and NAD than all other 18 put together, maybe with the

01:40:49.740 exception of metformin. So this is a topic that just continues to interest people. I would say

01:40:55.760 that my understanding of it is sort of at the six out of 10 level, which is enough to be dangerous

01:41:00.440 and enough to be frustrated at the fact that it's not nine out of 10 level. And we talked about this

01:41:05.420 again the first time we spoke, but let's go back for a moment and explain why do people even care

01:41:09.860 about NAD or why should one care about their NAD levels?

01:41:13.020 Well, I've talked a lot about sirtuins today. These are the protectors of the genome and the

01:41:17.460 epigenome. They lose their activity over time. They have two things they require for activity,

01:41:22.920 for maximum activity. We mentioned resveratrol, which is an activator you can eat or take in

01:41:29.900 as a supplement. That's the accelerator pedal on this enzyme family. The fuel that they also need,

01:41:36.720 100% without it, they don't work, is NAD. NAD is a molecule that's in our bodies. We acquire it

01:41:42.500 every second of every day to exist. Our bodies use it for chemical reactions. And without it,

01:41:48.080 everything shuts down. And we're always making more and we're recycling it all the time. We have

01:41:52.300 many grams of it in our body. It's probably one of the top two molecules that's important for life

01:41:56.660 and one of the earliest that have ever evolved on the planet. The other one's ATP, which is chemical

01:42:00.340 energy. NAD is also used to be the most boring molecule in biology. You just had to learn by rote how

01:42:08.240 it was used by the body and recycled. And it was just a bunch of chemical reactions.

01:42:12.500 And it was forgotten about during the 1960s, 70s, and 80s. In the 1990s, especially in the 2000s,

01:42:19.000 was discovered that it also acts as the body's signaling molecule. And we think it tells the

01:42:23.900 body when you're exercised, when you're hungry, and is largely how calorie restriction works.

01:42:29.960 So we think that in organisms like worms, flies, yeast, more NAD is better. When you give them

01:42:35.060 more NAD, they live longer. And now the question is, is that true for humans as well? And the idea is

01:42:40.400 that by either replacing lost NAD or boosting it to levels that you would only get if you run

01:42:45.980 marathons constantly, you can turn on these sirtuin defenses and other aspects like DNA repair

01:42:51.680 proteins that need NAD.

01:42:53.500 So again, we could almost be back in the paradigm that we potentially are with metformin

01:42:58.340 and with resveratrol, which is, it might be that the less healthy you are, the more you could benefit

01:43:06.260 from supplementation or restoration, correct?

01:43:08.700 I believe that because in our animal studies and other studies that people have done,

01:43:12.840 the benefits of NAD and of resveratrol are seen predominantly in mice and humans that are

01:43:19.740 obese or have a disease. And so they replenish what's lost. That said, if you boost the levels in a

01:43:27.060 mouse of NAD, we published a few years ago, actually, no, it was about a year ago, that raising

01:43:33.020 them above normal levels in an old animal, get you back to having a young cardiovascular system

01:43:38.180 and they can run just as far as a young mouse. But when we gave NAD boosters to the young mice,

01:43:44.020 they didn't run further, but they did if we exercise them and gave them the NAD booster at the same time.

01:43:49.060 So it was the fuel, but not the trigger. I mean, you still needed to actually, it wasn't enough to

01:43:56.720 get the expression basically of the behavior. In a three month old, very young mouse, but in a 50

01:44:02.100 year old, I would say that at least speaking for myself, I already have some deficits. I'm not as

01:44:07.820 perfect or as healthy as I used to be. And so that may actually help more than it ever has before.

01:44:13.140 All right. So let's talk about boosting it. So the first question is, David, can I just go out and buy

01:44:17.960 NAD in a pill and take it? I think people sell NAD as a pill.

01:44:22.460 Let me reframe that. Is there a biological rationale for taking NAD orally?

01:44:29.320 Very few people have studied taking NAD orally. What we've studied in humans and in mice extensively,

01:44:35.060 maybe not as extensively as many would like, is giving precursors to NADs.

01:44:39.900 Because most people take NAD intravenously. That's sort of the typical way it's administered

01:44:43.840 in this country or elsewhere. Right. Right. But there's this adage and

01:44:48.780 there's some evidence that NAD doesn't directly get into cells. It's a large molecule. There's

01:44:55.220 some evidence that nerve cells take it up. But in general, it has to be broken down first

01:44:59.760 before it's taken up into cells and reconstituted inside the cell. That may work fine. I've heard

01:45:05.920 anecdotes that IV NAD is interesting, interesting results.

01:45:09.560 Although you could argue that the placebo effect coupled with the actual physiologic responses

01:45:15.920 one might have to nicotinamide could explain the quote-unquote reactions and the feelings that

01:45:21.820 people have to intravenous NAD. But is it safe to say that at this point in time, our scientific

01:45:27.340 understanding is that intravenous NAD is not sufficiently making it into cells and more

01:45:32.920 importantly, mitochondria? Is that a safe assumption?

01:45:35.420 Well, it gets into mitochondria because there's, at least if you believe this literature,

01:45:40.100 there's an NAD transporter that pulls it into mitochondria.

01:45:43.500 But not from the plasma.

01:45:44.700 Right.

01:45:45.080 Yeah, you'd have to make it into the cell first.

01:45:47.140 Yeah. I haven't seen convincing evidence yet. Now, I haven't read every paper on the planet,

01:45:51.580 but I'm unaware of...

01:45:53.000 Wait, you haven't read every paper on the planet?

01:45:55.160 Shame on you.

01:45:56.200 I know.

01:45:57.180 You don't come on this show without reading every paper on the planet.

01:45:59.620 Well, I'll sell my kids. The IV NAD needs a lot more clinical research. I agree with you.

01:46:06.180 Yeah, I'm a little skeptical on that. Okay. So then you said, okay, well, look,

01:46:09.480 we've got this idea where we can orally take something like nicotinamide riboside. And I can

01:46:14.120 go buy this on Amazon today.

01:46:15.520 Yeah, you can.

01:46:16.060 So NR for short, we're going to talk a lot about abbreviations. So NR becomes NAD how?

01:46:22.140 Right. So NR is nicotinamide riboside. It looks actually chemically similar to

01:46:26.920 how DNA is made, interestingly. It's what the riboside means. Nicotinamide is vitamin B3.

01:46:33.040 So it's partly a vitamin B3, partly a piece of DNA. So that is a molecule that cells suck up

01:46:40.040 through a transporter. It's well understood. They stick on a phosphate. It becomes NMN,

01:46:47.580 nicotinamide mononucleotide. And then the cells turn that into NAD. So it's two steps.

01:46:53.420 NR into cells, to NMN, to NAD. And then once it's into NAD, it's then recycled. It's turned into

01:47:02.020 nicotinamide when a sirtuin reacts with it.

01:47:04.740 Nicotinamide is abbreviated NAM.

01:47:06.680 Yeah, typically. And that's a version of niacin or vitamin B3. But many people ask me,

01:47:13.160 can I just take a high dose of vitamin B3? And there's some interesting things. You can raise

01:47:17.500 NAD by just taking vitamin B3, but you're missing out on the other components that the cell now has to

01:47:21.460 make, which is the riboside, the sugar, DNA part, and the phosphate. So it's not surprising that

01:47:27.060 other labs have shown that if your goal is to raise NAD in the body, at least in a mouse,

01:47:32.140 it's been studied that niacin isn't as effective as taking NR or NMN. And there's actually reasons to

01:47:40.400 avoid taking high doses of nicotinamide unless you're a cancer patient where it may help.

01:47:43.780 But nicotinamide, we showed back in 2002, is a really effective inhibitor of the sirtuins,

01:47:50.800 which are enzymes that you want to keep on. It's the whole point of raising NAD.

01:47:55.940 And so we try to avoid nicotinamide while raising NAD. And actually, I hadn't thought of this,

01:48:01.380 but it would be very useful if the field had a definition, which is the ratio of NAD to nicotinamide,

01:48:06.440 because that would give us an indication of the boosting the gas to the engine versus the brake.

01:48:12.020 Yeah. So right after you and I spoke last year, there was a paper that came out from Princeton,

01:48:18.760 Josh Rabinowitz's lab, that looked at oral nicotinamide riboside. It was a tracer study

01:48:24.940 that looked at mice where they gave them oral NR. And basically the question was, what is the fate

01:48:31.080 of this? Where is it going? And what that paper showed was the liver took, because this was oral,

01:48:36.620 of course, so that stuff gets, the NR gets absorbed out of the gut, presumably, and very quickly,

01:48:42.020 everything in the gut makes its way to the liver first. Hence, it's called this first pass effect.

01:48:47.180 And it was in the liver that most of that NR got turned into NAD. But the study didn't find that

01:48:55.120 much NR made it out of the liver. In fact, what the study, if I recall, and now it's been so long

01:49:00.600 since I've looked at it, but I think that they saw NAM, nicotinamide was up in the blood, but not

01:49:05.980 nicotinamide riboside, which you presumably will still wanted some of that leaving the liver to go,

01:49:11.000 to get into other cells. Because I'm assuming that you don't just need more NAD in the liver,

01:49:17.300 correct? Wouldn't you want it also in the muscles or other cells?

01:49:20.700 Well, yeah, you would. But there was a new study that came out that showed that if you give NR to

01:49:25.740 people in a clinical trial, they could get NAD levels raised in muscle as well.

01:49:30.960 Which study, this one was the-

01:49:31.940 Just was posted on Biowarka.

01:49:33.500 Yeah, yeah. This is the one that hasn't been peer reviewed yet, correct? And it also showed the

01:49:36.540 very high nicotinamide levels in the blood. Yeah. Yeah. Right. And so I think where the field is

01:49:41.380 now is it's trying to get the NAD levels high without-

01:49:44.600 Okay, yeah. That's the study that was using a very high dose of NR, correct? This is a thousand

01:49:48.400 milligrams? Right.

01:49:49.520 Yep. Okay. So that's taking-

01:49:51.340 On somewhat-

01:49:52.160 Four times the posted dose that's given when you buy the supplements online.

01:49:56.280 Yeah. It was a good study. Placebo controlled. They had average BMI was slightly higher. I think it

01:50:01.120 was in the high 20s. Average age was, I think, up in the mid-50s. It was a higher age group where

01:50:07.680 you'd expect some effects. So they proved at least what we had seen in mice, that you can

01:50:11.460 get NAD to rise beyond the liver.

01:50:13.320 How do you reconcile that? If that study demonstrated that there was an increase in NAD

01:50:18.840 in the muscle, how did it get there? It couldn't have got there from the liver. The liver can't,

01:50:24.740 to my knowledge, can't export NAD to the plasma, to the muscle. Does that imply

01:50:30.020 that the dose potentially in the Rabinowitz study was not high enough for enough excess

01:50:36.540 NR to leave the liver to make its way to the cells? Does it suggest, as at least one author

01:50:42.640 has suggested, potentially there was a methodologic error in the Rabinowitz study where through freezing

01:50:50.600 the samples, some of the NR was not detectable on thawing. Something that, by the way, I've asked

01:50:56.820 people on both sides of this. And I'm getting conflicting inputs on this, by the way. It's

01:51:01.900 very difficult to sort of understand this. Again, I don't think people are bad actors here. I think

01:51:06.360 it's complicated stuff and the assays don't lend themselves to necessarily working out every time.

01:51:12.840 But what is your best explanation for how the thousand milligrams of oral NR in humans made its

01:51:19.140 way into increasing muscle NAD?

01:51:21.740 Well, it's getting past the liver.

01:51:22.760 The NR is getting past the liver.

01:51:24.620 Right. Well, that's the simplest explanation. You've got to start hand-waving and saying,

01:51:27.900 oh, well, the liver then sends out an enzyme or a signal. I think that's possible.

01:51:33.820 They didn't look in the mitochondria. We don't know if the NAD made it into the mitochondria,

01:51:37.220 correct?

01:51:37.700 They didn't. They didn't. But there are a couple of recent studies that show that it's

01:51:41.380 very important for the NAD to go up in mitochondria particularly.

01:51:44.300 Yes. And I don't think that's been demonstrated, has it? At least not in healthy. We're going to come

01:51:48.560 back to the other study in a moment. If I recall that study you're talking about

01:51:52.380 showed a few improvements in certain inflammatory markers. Is that correct?

01:51:57.020 Right. There wasn't much change with the NR. It was some inflammation went down in the muscle.

01:52:02.260 And if anything, mitochondrial markers of activity were lower.

01:52:06.240 Were lower. That was something that didn't make a lot of sense. Although you could argue

01:52:10.600 if the mitochondria became more efficient, perhaps you needed less activity, but you start to wonder

01:52:15.600 if that becomes hand-waving as well. What's your interpretation of that particular finding?

01:52:20.240 I don't have a good explanation other than that's what happens and that's what we'll see

01:52:24.080 with other studies. I think we just need to check if other precursors do that because we

01:52:28.540 don't know if it's an NR-specific effect or if the whole class of molecules will do that

01:52:34.100 as well. That'll be interesting to see. But what I can say is that it's a surprise because

01:52:38.960 Johan Ulrichs, who's over in Switzerland, and myself and Matt Kaeberlein even, who's shown in

01:52:44.640 mitochondrial disease, that NMN, and in some cases NR as well, does boost mitochondrial activity.

01:52:50.880 Now, these are mice and it may be unfortunate that humans are just not mice and that's where

01:52:54.920 it ends. I don't expect that. I hope it's not true, but the data won't lie. We'll do the clinical

01:53:00.500 trials. I'll all be blinded. We've got many trials to go, but there are differences between

01:53:05.600 NMN and NR. So curious to see if NMN has the same effect in humans as well. Those studies are

01:53:11.340 ongoing. We don't have any good data just yet. And do you think that NMN would be best

01:53:16.200 administered through a regular oral route or would you want to do it through an SL route,

01:53:22.820 somehow bypass the liver? Do you think that there are opportunities there with either NR or NMN to get

01:53:28.900 even higher plasma concentrations, but without this compensatory rise in nicotinamide that potentially

01:53:35.480 is harmful? Well, the SL route, I'm asked about a lot. Sublingual, put under your tongue,

01:53:39.640 try and get it taken up by that. And let me explain why this is the case because it might not

01:53:43.280 be obvious if you're listening because you might be saying, why would putting it under your tongue

01:53:46.460 be okay, but swallowing it not? And the reason is when a person swallows a medication, it goes through

01:53:53.520 from the stomach into the jejunum and the ilium, usually in the jejunum, which is the first part of

01:53:58.540 the gut after the stomach, it gets absorbed. And that blood supply goes straight to the liver

01:54:03.940 through this thing called the portal circulation. And so most drugs actually have to be designed with

01:54:10.240 that in mind, either immune to the liver's metabolism or designed such that they're pro drugs

01:54:17.060 and the liver actually turns them into the right drug. When you're talking about putting something

01:54:20.960 under your tongue, just like someone who, for example, carries around nitroglycerin, if they run

01:54:26.200 the risk of getting chest pain, that drug gets directly absorbed into circulation and doesn't go

01:54:32.400 through the liver. So I just explained that for the listener to make sure they understand why that

01:54:36.840 would be a potential advantage. Yeah. And also there's the complicating factor that our microbiome

01:54:41.800 will love to chew up that NR probably. And there's increasing studies showing that microbiome does eat

01:54:48.740 up some of the molecules that we're ingesting. So that niacin part of the molecule, nicotinamide,

01:54:54.220 it comes off pretty quickly. Even if a molecule in your fridge gets wet, you will start to lose that

01:55:00.260 nicotin might bond and it'll break off. And in the gut, some evidence that people have published

01:55:06.240 and some haven't points to the gut playing a major role in how much this actually gets into the body

01:55:12.600 and how. Typically the public, they're not doing this for a living. They don't see the brutal struggle

01:55:18.200 for academic survival going on. But now in the days of podcasts like yours, Peter, the public can

01:55:24.580 actually see this play out. Now that's good because the public can see what is the cutting edge of

01:55:29.520 science and make their own decisions and hear experts' opinion. But it's bad because it makes

01:55:34.100 it look like science is one giant food fight. But that's normal. Any new field will have these

01:55:39.800 disagreements about, is your assay working? Is there a transporter taking it up? Is the microbiome

01:55:44.620 destroying it? Do we need pro-drugs for liver or can we just put NMN under the tongue? And we don't

01:55:50.300 have any good answers, really good answers. I'm afraid to say that right now. But I can tell you what I

01:55:56.780 see emerging. I'm happy to give my opinion. These are not facts. These are opinions. And I think we're

01:56:02.060 all entitled to our own opinions, certainly not facts. My opinion is that the microbiome removes a

01:56:09.020 lot of the nicotinamide from NMN and of NR before it's taken up by the gut. There are some studies that

01:56:14.920 I've seen that aren't yet available that trace the movement of these molecules through an animal.

01:56:20.160 We don't do those in humans typically because they're very expensive. You need to have isotopically

01:56:23.740 labeled molecules, labeling different parts of the molecule. But then you can say, okay, where did

01:56:28.600 the nicotinamide go? Where did the sugar go? And so I've seen some of that data. Now it's not all

01:56:33.220 in agreement, but if I was to summarize it, I think there's a little truth in everybody's results. I

01:56:39.980 think there's truth that it makes sense not to put it all through the gut. It makes sense that if you

01:56:45.580 put a lot in the gut, that's also going to work. Some of it will get through. There's some truth in

01:56:51.180 that NMN gets broken down in the gut and then taken up by the gut and remade in the body into

01:56:57.960 NAD because you're basically just pulling apart a three-piece Lego set, putting it through the

01:57:04.180 screener, and then reassembling it on the other side. That seems to happen too. But also I've seen

01:57:09.240 data that looks convincing that some NMN and some NR gets straight into the body, goes to the liver,

01:57:16.360 some goes beyond the liver into the muscle. And so it's messy. And there's probably never going to

01:57:21.340 be one single answer to what's going on in the body with something this complicated. But here's

01:57:26.200 the way I view it, is that certainly for the members of the public, I don't think they care

01:57:30.960 if there's a transporter or not. They don't care what we want to disagree about a mass spectrometry

01:57:35.820 assay for NMN. We'll figure that out. That'll come out in the wash. What's important is,

01:57:41.440 does it work in a human? That's really all that matters. We know that these molecules do amazing

01:57:46.080 things in mice to health and in some cases to longevity. Although potentially the most important

01:57:50.900 study of this is not yet out yet, which is this ITP for NR in mice, correct? That should be the most

01:57:57.720 robust analysis of NR, should it not, in mice? It is. So they use mixed strains. They use a variety of

01:58:03.640 labs across the country. And so that's, it's considered a standard, but it's not definitive

01:58:09.580 because there are plenty of ways to dose, plenty of ways to deliver it, plenty of molecules in our

01:58:15.280 kit. But if it doesn't work, it's another data point. And so with ITP, we'll see. Maybe it doesn't

01:58:21.640 work. Johan Ulrichs over in Switzerland says, if you give NR to old mice, it does work. I took

01:58:27.000 standard lifespan a little bit. We don't know about NMN. We're running that experiment in my lab.

01:58:30.880 And that's no secret. So we'll see if that works or not. We'll see. I think ITP is a good start.

01:58:37.280 What I find somewhat frustrating is that they've never asked me for advice on how to dose or what

01:58:42.000 to give or anything. So yeah, I was surprised as well that you weren't involved in that. Now,

01:58:46.680 what you're basically saying is, look, in the end, does this stuff clinically work is all that

01:58:50.360 matters? Because there's really smart people out there saying, show me the evidence that increasing

01:58:55.400 intramuscular NAD matters. What if it's indifferent? What if this is true, true, and unrelated?

01:59:00.420 Very recently, a paper came out looking at megadose of oral nicotinamide riboside with

01:59:07.720 terastilbene in patients with ALS. And it was a minuscule study that had as many dropouts as it

01:59:15.460 had completers, if not more. But the gist of it was that on some what appear to be subjective

01:59:21.540 measurements of quality of life, there was an improvement in patients with ALS taking this very

01:59:28.240 high dose of nicotinamide riboside with terastilbene versus those taking a placebo.

01:59:33.580 And they measured some cardiac function as well.

01:59:35.940 I think they measured one pulmonary function called forced vital capacity. Yeah. So which is

01:59:40.700 how much air could you blow, which would be a pretty important pulmonary function, which is one of the

01:59:44.440 more important things that gets degraded in somebody with ALS. So the point is, for a very small

01:59:49.460 study that obviously didn't have any hard endpoints, it looked like a success. But I can't help but think

01:59:55.360 of what you talked about earlier. What if this is another example of something where to see the

02:00:02.500 effect, you have to be testing it in the most distressed organism. You don't like to talk about

02:00:07.900 people in that terms, but a person with ALS is under far greater distress than you are. And it's

02:00:13.780 certainly possible that in somebody who is that close to the physiologic limits of survival can actually

02:00:20.980 see a small benefit, which I think is what that study, assuming that study is replicated, which of

02:00:25.920 course is to your point, that's the nature of science. I mean, each experiment is nothing more

02:00:30.840 than a way to alter a probability of something likely to be true, but this would now make that case.

02:00:37.760 But that was my reading of that study, which was interesting, but I want to see that in someone

02:00:43.340 healthy. I want to see that in someone for the same reason I want to know what metformin is doing

02:00:48.560 in somebody who doesn't have diabetes. So someone really smart, I forget who it was on this topic,

02:00:54.500 once speaking very specifically about this, there is so much smoke out there that you have to believe

02:00:59.380 there's a fire, but I just don't know where it is. I think that's sort of how I feel.

02:01:03.100 Well, with the resveratrol experience that I've had in my career and with NAD, it wouldn't surprise me

02:01:10.740 if what you said is true, which is if you're in peak condition and you're young, you're not going to

02:01:15.360 see a big effect. If you've got ALS or some other disease that gives you low NAD levels,

02:01:21.120 so two ones are not working the way they should, then you'll see the benefits. That seems to be a

02:01:25.660 theme that's emerging. If that's true, that's still good because we're not always going to be super

02:01:30.580 healthy or able to run every day. There will be a come a time. And for the ALS patients, I'm sure

02:01:36.260 they're rejoicing that this could be true. That study was the first real believable hint. I'm using my words

02:01:43.140 very carefully, but that one looked like there may be some fire there in the ALS patients. Now,

02:01:49.000 it was a p-value of 0.01 and there was some subjectivity, but if you look at the placebo

02:01:53.000 versus the control, the placebos got worse and the drug experimental, most of them went up in

02:01:59.380 improvement in terms of life measurements. You didn't have to squint to see that result, which was

02:02:04.840 a nice thing. Now, we'll see. I mean, again, you've got to remember that they don't have pure

02:02:10.360 NR in this drug. It's a mixture of... It's NR with terastilbene, yeah.

02:02:14.100 And terastilbene is a very similar molecule to resveratrol.

02:02:16.880 Yeah, I was going to ask you, the study was, I believe they were using 1,200 milligrams combined.

02:02:24.460 So it was 1,000 of NR and 200 of terastilbene.

02:02:28.220 So that's six of those capsules that they sell.

02:02:31.060 That's right. Now, the question is, is terastilbene milligram for milligram as potent as resveratrol?

02:02:36.380 In other words, were those patients only taking one-fifth of the dose of a sirtuin

02:02:40.880 activator that you're taking?

02:02:42.160 Yeah, they would have been.

02:02:43.160 So in other words, 200 milligrams of terastilbene is about the same as 200 milligrams of resveratrol?

02:02:47.920 No one knows that, but resveratrol is just, again, these methyls. Resveratrol has three little

02:02:53.840 arms sticking out of two rings and two of those are methyls in terastilbene. So it's very

02:02:59.460 similar molecule to resveratrol. Whether or not it's superior, I don't believe it is. I mean,

02:03:05.620 there's some marketing that says that it's better. I haven't seen any data on that.

02:03:09.320 If this study were done without the terastilbene, it might be more interesting because we could then,

02:03:14.380 it's almost like you'd almost have a third arm that had either NR only or PT only. PT for the

02:03:20.040 listener being terastilbene.

02:03:21.240 Yeah. Well, that's the best way to do an experiment, but it's probably an extra few million dollars to do

02:03:25.480 that. Yeah. Interesting. So what is the current field looking like on the sirtuin activators?

02:03:32.060 Are there others coming down the pipeline? Is your lab working on next gen sirtuin activators?

02:03:37.600 What we're pushing hard on are pro drugs of NAD boosters. Hopefully get around all the

02:03:43.560 liabilities that we've discussed about these NAD boosters, such as better absorption, not digested by

02:03:49.780 the gut bacteria, is released in the gut, doesn't fall apart in your fridge. These are all good

02:03:56.380 things and hopefully is more potent than the natural molecule. And those, I've been working

02:04:01.420 with a team of chemists for the last five, six years with hundreds of different molecules that

02:04:06.060 have been put at least through animals and hopefully one day will be put into humans.

02:04:10.600 Our most advanced molecule in that class is in human studies right now at the Brigham Women's

02:04:15.500 Hospital. In what type of patient?

02:04:17.240 Well, these are healthy volunteers. So this is like phase one.

02:04:20.540 It's a phase one, yeah. So phase two would be if all goes well next year.

02:04:24.180 And what type of patient do you think would be most applicable?

02:04:27.140 I can't divulge what the company's thinking because they're paying for it, but I can say

02:04:31.040 that they're looking at diseases that are not common. And so very similar to what you're

02:04:35.800 sort of like this example of the, so people who are closer to the metabolic or to the sort

02:04:40.140 of cliff edge.

02:04:41.100 Well, there are a number of reasons for doing that. One is that there's good animal models for

02:04:45.200 some of these diseases where these molecules and relatives of them have worked. But also

02:04:49.900 there's that business reason, which is that trying to make a drug for obesity or longevity

02:04:54.260 is currently, no one would give you any money to do that, be very difficult. And so what's

02:05:00.420 hot, what people want to see is a fast track for a disease that is, has an unmet need where

02:05:07.440 patients are demanding something from the FDA and that there are actually bonuses, incentives

02:05:13.780 that the government has put in place to encourage people to make drugs for those diseases as

02:05:18.340 well.

02:05:18.540 Well, David, the last thing I want to talk about is how in the hell did you make those

02:05:22.820 beautiful, beautiful drawings in your book?

02:05:25.760 Oh, well.

02:05:26.540 I was surprised to learn that you had actually done those. You're an actual artist. You're

02:05:30.520 an artist who masquerades as a scientist.

02:05:32.140 Well, my enemies would say I'm a BS artist, but I do drawing as a hobby. And it was actually

02:05:37.820 a real pleasure that I was forced to do drawing. Hadn't done it since the 1980s, but I used to

02:05:44.200 do a lot of drawing. I was going to be an artist or at least an architect, a paid artist. I

02:05:48.980 very nearly became a computer graphic design guy before it was such a thing as Pixar. I loved

02:05:56.580 biology, though I ended up falling in love with lab work and thought aging was the biggest

02:06:00.580 thing that needed to be sold rather than drawing pictures. But the way I ended up having to

02:06:04.980 draw all these pictures, which are, you haven't seen the book, you can go on the website, lifespanbook.com

02:06:09.680 and I've got a lot of the drawings up there. Even now they're posted.

02:06:13.100 Well, by the time people hear this, the book will be out.

02:06:15.220 Well, fantastic. So unfortunately they're reduced to the size of a postage stamp. I drew them

02:06:18.780 the size of a foot by half a foot on a sketch pad.

02:06:22.100 That's what makes them look so impressive, though. That's what they look like. They're drawn

02:06:25.100 by a computer. They're so good. Because I'm trying to imagine you actually drawing them

02:06:29.840 that small. Not that it's any less impressive that you drew them at larger scale, but the

02:06:35.980 detail is unbelievable. They actually look like photographs that were then rendered into

02:06:39.640 sketches. That's how impressed I was.

02:06:41.320 That's how I originally did it. And the lawyers at Simon & Schuster said, you can't even use

02:06:45.620 that. You'd have to go and get permission from everybody who took the photo that I was

02:06:50.460 rendering. So I had to go back and basically make original art that I own. I had 28 days to

02:06:57.720 draw 28 human faces. And so it was fun. I really enjoyed it. I would get home sometimes

02:07:04.180 at 10 o'clock at night and have someone's face to draw. And what was actually helpful

02:07:08.460 was that I couldn't obsess over it because I have that kind of personality of perfection.

02:07:12.480 And I was forced. I've only got an hour or two to draw this and I would just sketch it

02:07:16.240 out. It actually wasn't that hard because these days you can hit control delete and or alt

02:07:21.080 delete and get rid of what you've just done if you don't like it. In the old days with

02:07:24.700 India ink, you make a sketch and a mistake, you forget it. It's over. Start again. Wasn't

02:07:29.160 that hard, but really enjoyed using that other part of my brain that I usually don't use.

02:07:33.620 Well, David, thank you very much for stopping by today. Congratulations on your book. I know

02:07:37.740 how much work goes into that and I think people are really going to enjoy it. I think it's

02:07:41.160 so funny. Like it's actually a pretty different look at a pretty common topic. I haven't done

02:07:46.700 the Google search on how many longevity books are in Amazon, but you probably need scientific

02:07:51.000 notation to count it. And most of them probably aren't worth reading to be honest, but yours

02:07:57.260 absolutely is. I think people will really enjoy it. Well, it's a fresh look. I think from

02:08:01.240 today it's been clear that there's a lot of new stuff in there. It's a different way of

02:08:04.500 looking at aging, this whole information theory idea. It's new. Nobody talks like I do about

02:08:09.440 aging. Very few of us, I should say. And what's exciting about it actually is I was writing

02:08:14.420 stuff down as it was happening in the lab. So readers won't just learn about what I do every

02:08:19.160 day and what my family does and what I think the future looks like, but also what it's like to be

02:08:24.240 part of these discoveries and how it feels and for the students and the impact on potential impact

02:08:30.440 on the world. So yeah, I'm very appreciative that you've had me on, allowed me to talk a bit about

02:08:35.660 what's in the book. I'm looking forward to reading yours when it comes out too.

02:08:39.880 Well, that'll be a couple of years from now, but that'll be great. We'll turn the tables.

02:08:43.160 Anyway, thanks so much, David. I really appreciate it.

02:08:44.960 Thanks, Peter.

02:08:45.260 You can find all of this information and more at peteratiamd.com forward slash podcast.

02:08:52.800 There you'll find the show notes, readings, and links related to this episode. You can also find

02:08:57.880 my blog at peteratiamd.com. Maybe the simplest thing to do is to sign up for my subjectively

02:09:03.140 non-lame once a week email, where I'll update you on what I've been up to, the most interesting papers

02:09:08.020 I've read, and all things related to longevity, science, performance, sleep, et cetera.

02:09:12.500 On social, you can find me on Twitter, Instagram, and Facebook, all with the ID peteratiamd. But

02:09:18.900 usually Twitter is the best way to reach me to share your questions and comments.

02:09:22.480 Now for the obligatory disclaimer, this podcast is for general informational purposes only and

02:09:26.860 does not constitute the practice of medicine, nursing, or other professional healthcare services,

02:09:31.860 including the giving of medical advice. And note, no doctor-patient relationship is formed.

02:09:36.640 The use of this information and the materials linked to the podcast is at the user's own risk.

02:09:42.340 The content of this podcast is not intended to be a substitute for professional medical advice,

02:09:46.860 diagnoses, or treatment. Users should not disregard or delay in obtaining medical advice for any medical

02:09:52.280 condition they have and should seek the assistance of their healthcare professionals for any such

02:09:56.840 conditions. Lastly, and perhaps most importantly, I take conflicts of interest very seriously. For all of

02:10:02.840 my disclosures, the companies I invest in and or advise, please visit peteratiamd.com forward slash about.

The Peter Attia Drive - September 09, 2019

#70 - David Sinclair, Ph.D.： How cellular reprogramming could slow our aging clock (and the latest research on NAD)

Episode Stats

Summary

Transcript