Summary

In this session, Daniel Ives, founder of Shift Bioscience, introduced their transcriptomic driver clock which enabled them to identify putative drug targets for safer cellular rejuvenation, which might avoid the challenges coupled to therapeutic use of Yamanaka factors. Daniel goes into detail on what the clock enables, how they are planning to validate these putative targets, his wishlist for tools that could help speed up and de-risk longevity and aging focused drug development, and their upcoming capital raise.

Presenters

Daniel Ives, Shift Bioscience

At Shift Bioscience Daniel is making drugs for cellular rejuvenation in humans through the application of machine-learning ‘driver’ clocks to cellular reprogramming. Daniel was inspired to pursue a Longevity mission in 2009 after reading Aubrey de Grey’s 2007 book ‘Ending Aging’. As a U. Cambridge PhD student, Daniel joined Ian Holts group at the Mitochondrial Biology Unit to pursue damage-removal strategies for mitochondrial DNA mutations, which cause…

Here’s what I call a friendly race for rejuvenation. I’ve just cherry picked eight of my favorites, whom I like for a couple of reasons. Firstly, they’re about age reversal. So it’s not just slowing it down, stopping it, it’s about bringing yourself back to a younger biological age, which is effectively age reversal. On one half is blood based rejuvenation, changing the blood milieu to make it rejuvenated. And then there’s this alternative paradigm, which is epigenetic reprogramming, basically cellular rejuvenation. The blood based rejuvenation has seen the most inroads in the clinic, probably because it’s so safe, because you’re just playing around with the blood composition. And a lot of this is gonna be familiar to the audience.
So Greg Fahy at Intervene Immune reversed GrimAge by a couple of years. That was persistent, which was really exciting. And I think they’re in Phase 2 with this cocktail of a couple of anti diabetics and a growth factor.
Tony Wyss-Coray at Alkahest is using an altered blood plasma fraction. He’s done things in Parkinson’s, the age related disease, which is a much higher bar than rejuvenation itself – you’ve got the aging, and then you’ve got the unraveling on top of aging. So really impressive to see things going on there.
Irina Conboy at IMYu is translating therapeutic plasma exchange (for rejuvenation), a clinically approved therapy, so let’s see what it does outside of its original indication.
Harold Katcher at Nugenics Research has shown you can reverse methylation age in rats, so far our best window into aging, by about 50%. It’s not like an incremental amount of reversal, it’s considerable and you basically restore all of the organ functions to young levels.
And then on the reprogramming side, which is, you could argue, maybe a more comprehensive type of rejuvenation, you’ve got David Sinclair and Life Biosciences. Taking pluripotency factors, playing with the composition of the factors, reducing them to OSK, reversing methylation biomarkers of age, restoring functions, which is a fantastic result.
Turn Bio is doing great things. They’ve got an mRNA based system and slightly more factors that are comprehensive (w/r rejuvenation)
And then you’ve got a couple of new efforts. Well, Calico is not so new, but the initiative is relatively new – trying to go deeper into reprogramming. Looking below the surface of what’s going on, trying to deconvolute the pathways. And there’s a recent paper in which they showed that just OS was sufficient to rejuvenate and it had about half the impact on the identity of the cells – which is the big worry, we want to rejuvenate but we don’t want to turn into a bag of stem cells.
And then something that I’m excited about – Joe Betts-Lacroix and Retro Biosciences. Everything that their staff is saying is very exciting.

Why this discovery effort on cellular reprogramming? Why is more discovery necessary? Why don’t we just go ahead with what we know? Because it is a powerful rejuvenation paradigm, but with risks. And how we mitigate those risks is an area of active pursuit.
As a reminder in the top left graph it is shown on the Yamanaka dataset that when looking at the DNA methylation clock, you can take a cell from age 60 down to the age 0 – in just 17 days. That’s the most rapid rejuvenation demonstrated. And the exciting thing is that up until about 15 days, the cells still retain their original identity – not 100%, but they’ve still got a memory of it and can revert back to their identity. That’s a sort of a sign of how far you can rejuvenate but not turn into a stem cell. At least according to the clock, but that’s just a biomarker. If you express a subset of these Yamanaka factors in the optic nerves of aged mice, you can restore a remarkable ability that only newborn mice have. If you crush the optic nerve of a newborn mouse, it regenerates. And after just a few days of life a mouse loses the ability to regenerate this optic nerve. However, if you put the factors back into the mouse and express them, an aged mouse can now regenerate its optic nerve just like a newborn mouse. So you regain these functions. So it’s quite a dramatic functional result.
On the right side, you can see increased lifespan in mice, but you have to be in this Goldilocks zone, not too little, because nothing happens, but also not too much, because you get these pluripotent teratomas – scary cancers. But if you’re just in the middle, which is basically partial reprogramming, you’ve got a nice lifespan extension, at least in progeroid mice in this case. I’m sure there’ll be some interesting follow up studies to come.
So this is a very promising paradigm. But these are primary pluripotency factors, let’s not forget that. They’re designed to make things into stem cells. And that is in the direction of cancer. So we need to proceed with care. When we do something in one system, it wouldn’t necessarily work in the other, we don’t know. So it’d be great to increase the safety window, so we don’t need to worry about walking the tightrope.

Where do we fit into this? We’ve identified some putative drug targets for safer cellular rejuvenation. By applying machine learning ‘driver’ clocks to cellular reprogramming. And we’re ready to validate these targets. We’ve got this list of genes that we want to throw into a cell. And now it’s just a matter of just bringing in slightly more financing, then testing these genes and seeing if the bioinformatics are as exciting as we think they are.
So let’s go back to the beginning. We set out in 2017, with the backing of Jonathan Milner who’s an angel in Cambridge, UK, to tackle mitochondrial DNA mutations. This is a type of damage that accumulates with age. And we had some promising molecules that could combat these mitochondrial DNA mutations.
Something I wanted to look at early on was whether we can show you in an non-arguable way that this is the right approach to therapeutic intervention in aging. And it was actually my investor, Jonathan, that put me in the direction of the epigenetic aging clocks. The moment I found out about these clocks, it presented a fantastic opportunity to audit our hypothesis. So if mitochondrial DNA mutations were really the be-all and end-all of aging, then they should have a dramatic effect on this epigenetic aging clock, which was an agreed upon measure of aging. I know there’s lots of measures of aging, but at the time it (the clock) was relatively new as far as something everybody could agree on. And so we’ve done a mouse experiment with this mouse called the mutator mouse – it has way more mitochondrial DNA mutations, and it gets older faster, you get premature aging. So it’s very provocative, the link between this damage and the aging phenotype. And we had a drug that could slow down some of these aging phenotypes. But there were two questions.
Does this mouse, which shows a premature aging phenotype, show acceleration of the clock? Is there this really strong link between mutations (mitochondrial DNA mutations) and the clock? That’s the first question. Second question was, does our drug impact the clock?
There was no such thing as a mouse clock service that we could send samples off to for measurements. So we recruited our first full time employee, who learned the methods directly from the lab in Cambridge that created the first multi-tissue mouse epigenetic ageing clock, so that we could do what was needed to do. And we measured the clock in our mice, and surprisingly, the mutator mouse, although it showed a premature aging phenotype on the outside, the clock wasn’t accelerated at all. So the clock showed just the chronological age of the mouse, even though the mouse looked dramatically older. So that was a surprise to us. Even more surprising was that we slowed down the epigenetic aging clock in these mice. So there didn’t seem to be a connection between the clock and the mutations, but a drug designed to reduce the mutations did slow down the clock. And we later found out that only in this mouse, we slow down the clock, if we use these drugs in wild type mice, we don’t slow down the clock. So there’s some interesting biology there!
But the main thing about this experiment was that it was trying to find out whether targeting mitochondrial DNA mutations is the best lever for therapeutic intervention in aging. Its certainly an interesting approach for a niche aspect of aging, but for us it (the result) was a pivot point. The clocks told us that perhaps this isn’t the best approach right now.
So where do we go to look for better levers? Around the same time, we took on an intern called Brendan Swain, and we were both really excited about (1) ageing clocks (2) the Tabula Muris Senis, which is a Fantastic aging resource. For those that are not familiar, the Tabula Muris Senis is a database with single cell transcriptomes of the aging mouse at multiple time points, from every single cell, and every organ in the (mouse) body. It is basically a playground for aging scientists, to try and come up with things and look for phenomena. And what me and Brandon were really excited about was the logical extension of aging clocks. Could we create a clock that enabled a CRISPR screen for aging? Could we knock out every gene in the genome, have a clock to get a (ageing) readout, and get a relationship between every gene in the genome and aging, so we could systematically go through everything and let the data do the talking – so the genes will reveal themselves according to their relationship to the clock.
And the technical solution we needed to solve for this CRISPR screen for aging was a single cell aging clock based on a single cell transcriptome. And this was the enabling technology. So we played around with this Tabula Muris Senis, trying to develop methods for accurate single cell ageing clocks. And the short story is we did manage to crack this problem and generate accurate ageing clocks with single cell transcriptomes. And they have advantages. They also have disadvantages. But we did crack this problem. And just before we were going to go ahead and do the CRISPR screen, we realized there were a couple of problems. Firstly, the screen we were planning would just be single CRISPR knockouts. On a 10x Genomics machine, you could do a single CRISPR knockout in each cell and get a simultaneous transcriptome from each cell. But that’s only single gene perturbations. So you limit your question to ‘how can I rejuvenate with a single gene?’ but the more exciting question is ‘how can I rejuvenate, full stop?’ whether that’s a combination of four genes, or five genes, we don’t know. So with the 10x system that we were going to use, we didn’t have access to this combinatorial space. So that was the first problem. And then secondly, we realized that the clock we generated was built out of genes. And these genes were very interesting, some of them were known aging genes. And the more we looked at them, the more we realized we were looking at a lot more than just a single cell aging clock, we were perhaps looking at the aging biology itself. I mean, the original epigenetic aging clock, it’s a big question. Like, what are they? What’s the underlying biology? But with these gene base clocks, those answers were easier to spot. And so now we’re at the The Milner Therapeutics Institute, inside the Jeffrey Cheah Biomedical Centre, and we’re going to try and validate these genes. So, that’s the high level story.

And just a bit of detail on the technical breakthrough, what we did was we took a gene expression matrix, gene expression from single cells. And we did dimensionality reduction. So we went from genes to pathways, and this distills the information relevant to a time course – so there are aging time courses, rejuvenation time courses (for instance when you express Yamanaka factors, that’s a time course of rejuvenation). There’s also time courses for age related diseases, which have an aging dimension, and a disease specific dimension. So this method is applicable to all of those scenarios, but we focus on aging and rejuvenation. So we did a dimensionality reduction, and distilled the information that’s interesting in that time course. And then we used these pathways as input for a machine learning model of the time course – so aging clocks models using elastic net regression. And then we have this machine learning model that retains information about the contribution of individual genes to the model. So it’s like a ranked list of genes, where one gene contributes the most to aging and then this one, and another one after that. And that ranking is key, it allows us to prioritize some biology over other other biology.

As for some advantages of the gene clocks. Firstly, they allow you to enrich driver biology from a time course, just as evidenced. In the right-hand image you see the relative weightings of certain genes in a clock trained on human fibroblast aging. And you see mitochondrial genes show up, you see ribosomal genes show up, you see a gene that’s sufficient on its own to drive progeria, or accelerate aging on its own. These are known drivers of aging or genes linked with aging. And it’s really interesting to see, it doesn’t necessarily mean that we’re better than another clock, but it’s encouraging to see these things.
The second thing about gene based clocks is that you can rapidly test causality, with mature technologies like CRISPR, or overexpression. You can knock out a gene, activate a gene with CRISPR, or you can overexpress a gene and see whether it affects aging according to lots of different measures.
Genes are also very easy to target therapeutically. So you can have a therapeutic mRNA to target a gene, or you can develop a molecule targeting the gene products.
And the last thing is we’ve got the single cell resolution. And that was the reason we created the clock to try and get this single cell resolution. But it turns out, it’s more of an academic interest than anything else (though time will tell)

And then just to show you some of the some of the things we’ve done with this clock. So this is basically a single cell, RNAseq time course of aging fibroblasts, so we’ve only got five time points in this image on the left. But we’ve basically got predicted age versus actual age. So we got a couple of samples towards the lower end and a couple of samples towards the higher end and then a middle aged person in the middle around 50. But you can see there are individual dots in there, so we can predict the age of individual cells, with a Spearman’s correlation coefficient of 0.94.
Interestingly, in the two samples that are older, one person seems to have aged a bit faster than the other. I’m sure there’s an interesting story there. And then when we look at the genes that make up this clock for aging fibroblasts, we see these genes show up, which were linked to aging before, which is really encouraging.

This method can be applied to any time course, aging is just the first area of interest for us – that’s where we went to look. If you take Yamanaka’s public data, you can train a clock on rejuvenation. So actual rejuvenation means DNA methylation based rejuvenation, so we train our gene based predictor against that DNA methylation readout and get a very accurate clock with a 0.95 Spearman’s correlation coefficient.
And then we get a bunch of genes. And these genes are really exciting because they don’t look like pluripotency genes, a lot of them look very familiar from the aging side. And we’re particularly interested in the genes I’ve labeled because of their expression trajectory. Gene ‘1’ starts with a low level of expression. And very quickly, like in the first three days of the rejuvenation time course, it peaks and then it sustains. So this is very similar to the rejuvenation behavior – you don’t have any rejuvenation for the first three days, with Yamanaka factors, and then you get sustained rejuvenation. So we’re very interested in the type of genes that map to this rejuvenation pattern.
And then gene two is the opposite. So a gene with high expression that dips quickly, and then it stays down.

So now we’ve got this list of targets. And we just want to extract combinations. We’re going to extract lots of different combinations. There’ll be non-rejuvenating gene combinations, basically genes that are neutral. Genes that our clock arrives at, but they’re just flat.
And then we’re going to also find a pluripotent and rejuvenating combination. We want to move beyond this, because that’s what we have with the Yamanaka factors. But we expect to see combinations of pluripotent and rejuvenating genes.
But what we’re really setting out to do here is find combinations of genes that are non-pluripotent and rejuvenating in combination. And then also the interface between all the other genes, really trying to separate the rejuvenation part of cellular reprogramming from the pluripotency part. This is a gross oversimplification, but the idea is we have this list, and now we’re throwing these genes into our cell system where we can rapidly validate them.

This is our very high level roadmap. We’ve created this gene driver clock, originally to perform a CRISPR screen for aging, but there’s much more we can do with it. We’ve just protected this (with an EPO filing), which makes this talk much easier! And what we want to do next with this funding round, is that we want to distill the minimal set of genes sufficient for safely rejuvenating a human cell type. We don’t know what we’re going to arrive at upfront, whether it will be Yamanaka factors, or something slightly safer where we will actually be able to fully separate the pluripotency from the rejuvenation biology, and we’re confident we can do this. There are others like Calico that have shown that you can start to separate the biology but they are nowhere near exhausting the path ahead. Once we arrive at that minimal gene set, that’s a big IP milestone.
And then the real work begins, where we’ll start the therapeutic development. For a general purpose rejuvenation therapy, we’d like to develop that ourselves, but if we know one of the genes is linked to a particular disease indication, we don’t want to have to build a specialist pharma company again, so we’d probably partner with pharma for that one. But that’s the next stage. So one step at a time.

And just some of the characters that are working at Shift. I originally had training in mitochondrial biology, but then through the pursuit of rejuvenation, not just mitochondrial rejuvenation, I’ve ended up in a very different place.
Brendan was an intern with us in 2018. And we maintained a conversation and this is basically the direction we’re going now, it’s Brendan’s doing, so he deserves most of the credit.
Romina and I worked together when I was a mitochondrial biologist and we brought her in so we could do these first epigenetic ageing clock measurements.
Steve is not my brother, he is my dad. He’s not taking rejuvenation drugs, but he does look quite young. He’s a serial entrepreneur and saved us a lot of effort and dead ends and making stupid mistakes.
Once we’ve got more money, we can recruit all of the other positions mentioned on the slide. We are recruiting at least one lab scientist right now, because we need to see if these genes are gonna work or not (and there is no time to lose).

And these are some of our advisors. Jonathan Milner is a fantastic Angel based in Cambridge, UK. And he’s founded many very interesting companies.
Ken Raj has been a coauthor on many of the landmark epigenetic aging clock studies, a fantastic guy. He’s given us some great advice over a period of time.
Wolf Reik let us into his lab very generously. And he’s doing some really interesting things on aging, particularly multi-omics. So at the moment the aging clocks are confined to one omic layer, but if you’ve got multi omics, you can train clocks across layers, which has some great benefits.
And Aubrey has been really supportive for a long time.
And then David Billington, he’s got some serious experience in big pharma companies.

And now just some supporters, I’m based at Milner Therapeutics Institute (please hyperlink), which is an accelerating environment. It’s not necessarily an accelerator, but it’s like working in a big Institute. We’re a small company embedded inside, so we can just do things faster.
I’m really pleased that Allison accepted us on to the Foresight Institute accelerator. We have been going for four years, but we’re very small. So we’re only three full time employees. And I appreciate all the help we can get.
We were part of Creative Destruction Labs, which is like a virtual Y Combinator, a Canadian based effort. And I was actually put in touch with somebody that’s been very valuable to us through this effort.
Accelerate@Babraham is an accelerator where the whole Cambridge UK life science ecosystem comes together to help new companies, and we were also given a nice cheque and lab space.

And since I am talking to this great group, it’s a good place to ask for technical solutions. So if anybody is working on these technical solutions, this would help accelerate things. I don’t know if any of these things are possible, but if you don’t ask, you don’t get.
The first thing is a skin and blood clock, but for biological age. So there’s something called the skin and blood clock at the moment, but it’s for chronological age. You can measure aging, in fibroblasts, and that also tracks aging in the blood. So there’s a connection between what’s going on in vitro, in your dish, and what’s going on in real humans. And that’s just for chronological age. If we want to develop therapeutics, they’re going to have to target diseases. And so what we would really like is something like a GrimAge that we can use in a dish, or a SystemsAge from Morgan Levine that we can use in a dish. So when we’re testing things in a dish, we can say – look, our time to heart disease is going up in response to this intervention – we’d like to know that as soon as possible. So that would be a fantastic way to de-risk the project substantially, by showing we’re actually pushing away age-related diseases with a prophylactic intervention.
And secondly, maybe a little bit simpler, a biological age clock for mice. So if you could track pathological age related changes, and then train a clock including those changes and chronological age, then we can test our interventions and see whether those mice move further away from age related pathologies. Again, it would help to de-risk things. Because with GrimAge and SystemsAge, you have to do human studies. But we can’t just teleport to human studies, we have to get there.
Thirdly, once we’re in human studies, it would be useful to have biological age clocks that are disease optimized. So if there was a GrimAge for heart disease, and it was optimized to give us the best possible resolution, that would be great. If we were targeting a certain disease, having a clock corresponding to that disease would be great. Obviously if we have that in a mouse, even better and if we can have that in a dish (in vitro) even better.
And number four is single cell multiomics. If we could just do multiomics for a time course, whether that’s aging or rejuvenation, and train a clock that picks out the best biology, whether it’s a CpG site, gene, or gene regulators (from chromatin accessibility data). We could then cherry pick the best features, not just from the molecular layer we’re confined to by our particular ‘omic platform, but also different ones, whether that’s a transcriptome or something else. Something else. So if anybody can come up with some good multiomics methods that would be great.
As an example, Wolf Reik has developed something called scNMT-seq and it measures three molecular levels. They’ve got plans for four and five, so that’s going to be fantastic.
And then lastly, basically just datasets for age related disease time courses. There’s a relative abundance of aging time courses and rejuvenation time courses. But age related disease time courses are a little bit harder to come by. But when you’ve got these clock methodologies, you can train them on aging or disease. So you can potentially pull out the driving biology for disease specific side, and also the aging side. And when you put those two together, you’ve really got something interesting.

If you want to get hold of me, my email address is [email protected].

Q&A

I have two kinds of questions here. The first one is, whether I am understanding it right that you’re working on data from mice? And on fibroblasts that you’ve taken from mice? And if so, why not humans? The reason I’m asking is that I have data that I haven’t published where we took humans and gave them Metformin and placebo or Acarbose and placebo, you know, ITP drugs, and it was a crossover study, and we did the same in animals. So we have transcripts from humans and from animals treated for the same time, equivalent dose of drugs. And to our surprise, the transcripts are not the same, although the upstream regulators are the same, the transcripts are not the same. So I think we have to be careful how we go from animal to human and vice versa. And of course, you know that the clocks are not the same, right. So, I think, from a drug development perspective, you should really base this only on humans, right? Otherwise you might be confused with the experiments.

Sure. So the data I presented was actually data from human fibroblasts. There was an aging time course and that was human fibroblasts, so human subjects. And the rejuvenation time course was human fibroblasts. So I think Yamanaka basically took them at age 60 and turned them back. But we have done this on mice as well, and it shows similar results. So we did all of the experiments to create our method on the Tabula Muris Senis mouse data, but then we applied it to human cells, and we see really exciting stuff in the human cells.

The second question for you is whether you are putting some positive controls in. Did you also treat some of the fibroblasts with Metformin or Rapamycin or NAD or something that actually has the effect on epigenetics, to get the length field of what you’re doing in relation to general therapeutics?

(Daniel Ives) Yeah, I think that’s an obvious thing to do that we haven’t done so far. Looking at the gold standard interventions like rapamycin and checking what they do to our clock definitely makes sense. And if it does the expected things, the clocks are meaningful or more meaningful.
(Karl Pfleger) So the nice thing about this platform with regard to humans versus mice is that they should be able to look at both sets of clocks and actually limit to things that work in both. And of course, we care about them working in humans, but if they go down a clinical path, then working in mice is a good thing, if it’s the same thing that can work in both. But the platform should allow filtering – if there are enough leads – down to things that work in both.

Single cell sequencing is a little biased towards less expressed genes such as transcription factor, which is really important for aging and pathway generation. So how did you solve this problem?

(Daniel Ives) With respect to the rejuvenation time course, it wasn’t actually single cell data. There were cells that were sorted for a cell surface marker of pluripotency. So it’s actually bulk measurements. And we can use the same method on bulk measurements. So we don’t have this coverage problem, like you said, we would have in single cell data.
(Brendan Swain, inventor of clock method) I don’t want to reveal too much about the method just because of IP concerns, but the method isn’t just based on the absolute variance of genes, which should be the major contributor to that problem of picking out genes that are the most variable across the dataset, which is kind of how a lot of this kind of analysis is done. We’ve made great efforts to allow genes to contribute equally to the clock, and make sure that we’re not favoring the loudest shouters.

And my second question is whether you are doing a whole methylome analysis for identifying cell biological age, which is coupled with CRISPR? Or are you doing a specific methylation signature?

(Daniel Ives, mishearing ‘methylome’ as ‘metabolome’). We’ve started on the transcriptome. We were at the mercy of the public datasets to begin with, it’s heavily based on Tabula Muris Senis, and aging time courses in human cells. And we’re just starting to generate our own bespoke datasets to the systems that we’re going to be using to try and validate our genes. And so we will be looking to do beyond transcriptomics when we start doing things ourselves. But in the early days, we were at the mercy of the public datasets, so I don’t think there was anything there at the time.

Is all the data limited to fibroblasts?

No. So in the Tabula Muris Senis, which is mice, you’ve got a lot of different cell types, certain cell types have richer time courses. So as far as using our clocks is concerned, we’re better off with a richer time course. On the human side, and on the human rejuvenation side specifically, there are reprogramming paradigms from five different cell types. With reprogramming data from endothelial cells and four other cell types, you can train multi-tissue rejuvenation clocks across all of those and see what the rejuvenation biology is. Because obviously it works across lots of tissues, the potential is higher from a translation perspective. So not just fibroblast (rejuvenation), we are shooting a bit higher.

What would actually really help you move this project forward now? If anyone is on the call or even afterwards for those that will watch it on YouTube? What are specific ways in which they can reach out to you, in as actionable a way as possible?

Beyonds the wish list, which is maybe a little bit farther in the future, immediately we just want to test these genes (putative safe rejuvenation genes) as fast as possible. And we’re doing a lot of these things for the first time. So we’re looking for people, preferably in Cambridge UK, that can help us. We need a lab scientist at the very minimum. And we’re raising a certain amount of capital. It’s not a ridiculous amount, but it’s a lot more than we’ve raised before. So if there’s anybody out there that’s very passionate about clocks, and trying to go beyond Yamanaka factors (for safe cellular rejuvenation), explore the unknown, see what’s there, let us know. It’s just exciting to have an approach where you can reach some of that. If any of that sounds exciting to you, and you want to help move this forward, then please get in touch. The support now would be so much more appreciated than say, in six months time. We really need to get going now. Because they’re (the genes) are just sitting on the list. It would be a tragedy if the lead genes were sitting on that list for any longer than necessary. So we just want to start testing them.

Where do you see this if it is successful? I mean, you already talked a little bit about immediate next steps. Where do you see this going in potentially like 5 to 10 years? Are there any very high hanging fruits that you hope to be able to eventually get to?

Drug development is a very long and painstaking process, so you’ve got to set expectations, but at the same time, the whole COVID response, the goalposts were moved for that. I’ve got a feeling that if there was something really exciting and safe moving forwards, there would be a certain amount of energy pushing that forwards faster than expected. In 5 to 10 years time, I’d expect something (a safe rejuvenation drug) to be in the late clinical stages. There’s so many steps in between, and I think getting things right is more important than the timeline. But obviously, the faster the better, because if you get these drugs to work, we’d rather have these drugs sooner. I think even before that, it would be really great if we could find a list of safe rejuvenation genes and engineer these into a mouse embryonic stem cell, so that you can just induce these genes in a whole mouse and not worry about getting drugs to every part of the body. J ust induce the mouse and then watch the ageing clocks. Do the clocks go down? And can you keep resetting biological age to infinity, having the first mouse where you can do that. That would be really eye opening, to see what would happen there. So I think generating that mouse embryonic stem cell will be a really exciting thing that isn’t so far in the future. But you’d have to give that a bit of time. There’ll be people that aren’t specialists, or Longevity enthusiasts, and they would just want to see the mouse live longer.

You already talked a little bit about the race at the beginning of the talk, do you have any ideas and potential collaborations going forward even with other folks in that race? Because I think what’s so cool about you and Nathan who’s on this call is that you really try to pull the whole ecosystem together, you’ve been doing quite a lot of work on this. I think it’s really interesting how you position yourself really in this ecosystem like a rising tide?

Yeah, I think Steve Horvath’s Clock Foundation, just developing these clocks, to basically speed the drug development. That’s great. I think Morgan Levine is doing some really cool stuff enhancing the clocks. That’s an obvious one. And I think it’s very hard for me to analyze what’s out there and try and connect that to what I’m doing. So if you think you can help what we’re doing, obviously I’m going to take the time to listen to you and see if that is a good fit. And so I’m just keeping an open door, and it served me well.

Do you think that there’s any potential risks or downsides that you see, any potential reasons why this might not work that you could falsify in the next few years or so? Like is there anything that keeps you up at night? You’re already testing this right, is there anything else where you see a potential bottleneck that we need to overcome that maybe people in this call could even advise on?

Yeah, the thing that keeps me up at night is translatability. We could get very excited about what we’re doing in our specific system and then we don’t have a connection to all the other systems. So, Steve Horvath’s mammalian clock is a really great way to try and dispel some of these fears, because it’s connected to so many different contexts of aging. When you move that clock, you’ve really moved something. So that’s one way to do it. But yeah, it’s just making sure we’re not system specific. So if we do find a minimal set of safe rejuvenation genes, we just want to broaden from our first cell type to multiple cell types and make sure that it works in mice. We have to go through mice to get to humans, whether we like it or not. Maybe organoids can also help.. Basically we should test all of the systems. There’s no reason to go narrow and try and save pennies (or cents) when we could just increase the breadth and translatability. So I think translatability is the biggest worry. And then the safety side, can we fully deconvolute from pluripotency. That’s not a foregone conclusion. So we just need to complete the study. There are stil discovery elements to what we’re attempting but there’s a lot of promise from what we’ve done so far. The opportunity is there and we just have to make the most of it.

Seminar summary by Bolek Kerous.

Gaming the Future: The Book!

Drug Target Discovery for Cellular Rejuvenation by ‘Driver’ Clocks | Daniel Ives, Shift Bioscience

Summary

Presenters

Daniel Ives, Shift Bioscience

Presentation: Daniel Ives

Q&A

Search Foresight Institute