Nickolina Lauc - The Path To Reliable Ageing Clocks

1:09AM Aug 5, 2024

Speakers:

Leslie Kenny

Keywords:

glycans

aging

clock

disease

years

women

change

predict

looked

drug

epigenetic

maximum lifespan

biomarker

lichens

predictive

individual

published

prediction

estrogen

biomarkers

Thank you very much. So I will talk about clocks. We've heard a little bit about them today. And I'll mention a number of them. But keep in mind, I'm an expert on one. So watching the glycan side, and I'll name all the experts on the others. So first, why should we measure biological age? Why would we need this as a marker? Well, we know that chronological age is not a good enough measure of health, all of us and I'll give you two examples. We recently tested a number of centenarians in Costa Rica, and there were 100 100, free under five. And none of them actually had any diagnosed condition at that point, none that we can diagnose with traditional health care. And they're already achieving extreme longevity. While on the other hand, we have friends, we have a colleague, who started working with a company when she was 19. By 16, she got a diagnosis of rheumatoid arthritis. Then at 20, she got a diagnosis of MS. And she has a life expectancy which is greatly reduced to where she's living. So biological age should give us a better indication of health than chronological age, and also this risk of disease and mortality, especially in people who are not aware of it or younger individuals who are have this accelerated risk, but it's going undetected. And aging clocks have been the mission for a very long time. Actually, the first assumed biomarker of systemic aging were telomeres. But then we learned that telomeres are really good marker of aging of a single cell. But they're not as good of a marker of aging on the system level. Because we have cells, trillions of cells with many different ages. So we collect the sample and try to calculate an average age, it doesn't equal a very good systemic age. And for systemic age, now we have two very good candidates. So the first, molecular systemic age clocks were the glycan clock and the safe harbor DNA methylation clock. And by complete coincidence, they were published on the exact same date in 2014. So last year, we celebrated their 10th anniversary, and the first likened clock was developed in creation. And Steve Horvat, who's the inventor of the epigenetic clock, his surname means creation. So we decided to hold a conference in Croatia to celebrate our anniversary. And if you Google, actually, I have a link on the other page, like we shared with you afterwards, you can see all the recordings if you look for the longevity symposium and see if so, the epigenetic clock and mentor, he was actively looking for aging. Looking at the epigenome, we were not looking for it, we stumbled on it by complete random coincidence, by looking at a molecule that's mainly neglected in research. And we didn't have a lot of tools to look at it at scale in the past. In fact, our lab pioneered high throughput glycomics. Back in 2007, we analyze the first 1000 Human glycans together with a lab in Dublin led by Polina. And to date, our small, small lab and are a research institute in Croatia, lifted 85% of the total human glycan, we generate 85% of high throughput, like studies globally. And this is a review paper by chemical reviews. So these are just published, like comes out 191 published like comes our lab generated 160,000 of those. And we this in collaboration with all the top universities from Oxford Park, Harvard, pretty much everyone. And we found that when you look at these large cohorts, you see aging, especially on IgG glycans, we would see that certain structures would accumulate with age, and other structures would decline with age. And if you see on the graphs here, if I can love figure out a ticker, women are red. And they have a very different aging curve to men. Men have a very liberal path. And then women we see this acceleration that's linked to menopause. And I'm going to get into that a little bit later. So then we're curious. Can we see this agent in every population and together with the human glycan project, we looked at 27 different populations around the world 100 individuals in each and we saw that this prediction that the glycan was associated with like expected lifespan, but actually the strongest association was development index of a country. And we had a replication cohort in In the UK, of British minorities, the separate cohort, where we got samples from the first and second generation of British minorities, Indian, British, Caribbean, and Eastern European, what. And we saw that in the first generation there glycans would look like their original country. But in the second generation, their glycans will start to be more similar to white British, despite the ethnicity. So there is a very strong impact of environment. And that goes beyond ethnicity. And a number of these clocks are good predictors of mortality and morbidity. So the key thing they have to do, if you have an accelerated clock, Does it predict disease? And Does it predict mortality? And this was a nice comparison done by Ricardo Mariani. On the generation Scotland biobank, where he looked at if you have an accelerated like an age, or a grim age or other epigenetic ages, it, how does it predict diabetes? And how does it predict all cause mortality. And interestingly, although our clock is not modeled on reality, like the Grim age, were equally predictive of all cause mortality as women age, and fino each, which encompasses Clinical Biomarkers was less predictive than the grim age things like an age and predicting diabetes, so they are more predictive than some of the traditional markers we have. And maybe the most interesting fun finally, we've had in the recent years is that these clocks don't well, the clocks correlate because all of them correlate with logical age. But if you see if you look at acceleration in these clocks, epigenetic clocks, and glycan clocks are on an opposite spectrum, we do not correlate at all, which is very interesting, because both clocks are predictors of mortality and morbidity, which means that aging is not one thing. It's multifaceted. There's a number of different mechanisms driving it. And maybe the most, something that Steve Horvat did recently, he developed a clock to predict maximum lifespan in humans. And then he tried to correlate that clock with his clock to predict mortality and morbidity. And these clocks didn't coordinate, which means whatever governs our maximum lifespan is very different to what kills us before we get out there. And that's one of the key things we're focusing on. But it also means that there's different theories of aging, like one is a program. If we reprogram it, can we extend maximum lifespan, maybe, but also, we need to fix the things that kill us in between. And that's the part that we're focused on. Now, then, there was an idea that if we connect all of these different omics, we will develop the ultimate agent clock, and that was done in Edinburgh with the ricardas biobank, they tried to they looked at all these different multi omics aging clocks, and connected in one case here in the middle, created a multi omic or mega mega clock, they put them all together. And the pluses are prediction of hospitalization. And you can see in this ultimate clock, there was zero prediction of health outcomes, it basically became the perfect measurement of your chronological age. So putting these two, this, these things together doesn't work, we have to look at them independently. And these are potentially very early mechanisms of disease that start with aging itself. But they're not one thing. And different clocks predicted different things. But there are talks which are stronger at predicting mortality and morbidity. And if I tell you that glycans are a great biomarker for this, I would be very biased. But when somebody else says it for us, then we can, we can just agree. This was 10 years ago, this was a theory. But now we have more and more independent data coming out of how important like how important lichens are. And this is a paper in preprint. That's just about to be published in Nature Communications, done in way Cornell and Doha, where they looked at over 6300 molecular trades, and which ones of these were associated most strongly with age. And now the 20 associations nine of these were glycans, and the rest were predominantly sex hormones.

That this is a study which is being published in a couple of months time but we have some data from it, where we looked at acceleration of aging on an individual level because everything before was covered level. So if you look at individual level, these are six other twins follow the refer two years in Karolinska Institute in Sweden, and on an individual level, we see this very high prediction of mortality and also incidents dimension. So aging clocks on an individual level, particularly the glycan one are very important for your future health. And the Hallmark behind our clock is inflammation. Inflammation is a very old aging theory, it was published on all the way back in 2006, by Claudia Franceschi. But only recently, we added it to the home hallmarks of aging, although it is connected to a number of them. And we know that this mechanism leading to many different diseases has a common cause. And inflammation is part of our three out of five that deaths globally are conditions connected to it. So it's we know it's important for health outcome. But can we measure it? And here the problem is, same as we have with all disease, we walk into it blindly. Because our clinical markers are designed to detect disease when you're symptomatic, when the problem is very apparent. And it's severe. Well, actually, we know that these chronic conditions start decades before we can currently detect it. And the markers we use for inflammation are doing the same thing. CRP interleukins. All of this is when there's already a severe problem. Well, this we call it sterile, systemic inflammation is far harder to detect. And we have never had, we don't really have many biomarkers to do. So what, we have a biomarker that does it, but it's not yet popularized. Let's put it that way. This is where we go to other conferences, we try and preach about like puns, and hopefully somebody reads the research. But glycans are this very neglected. Oh, and they'll just realize these are my old slides. But that's my fault. We're changing them last minute. But the role of glycans in biology is to enable multi celled life. They've evolved since we become multicellular, their way their way that we change our G or change our function in biology, without needing to change our genes. And we were never going to turn genes and on and off for epigenetics, we can real time add function to our molecules for glycans, who would incorporate all those genetic and epigenetic influences, and connect them to what's happening in our environment real time. So do you're not familiar with them by name, you are familiar with them by process. For example, our blood groups are defined by lectins. All of our immune communication happens to glycans, cancer, would use glycans to disguises itself to evade our immune system, and even conception happens to lichens, sperm would coat itself in these healthy looking well, these anti inflammatory glycans to evade the woman's immune system. So they're present in every process we have since we're multicell. And often they get confused with something called glycation. And glycation is this random damage to a protein by sugars or the food we're eating. And it's a, it's not a regulated or enzymatic process. glycosylation, or what we measure where we see this more immune aging is enzymatic, it's regulated, it happens within the cell when the proteins are being developed. And majority of proteins and lipids are glycoproteins and glycolipids. We also see that likeness change in a large number of diseases in specific ways. So if you look at the curve of this circle, each one of these little like the G one G two s represents a different light cone, and the abundance of it and you'll see the different diseases have different glycosylation signatures, so that we are looking at systemic inflammation, it is quite specific when we look at the details of what glycans are doing. And we see that these changes in disease resemble aging, where you can have a young person with a diagnosed condition who has the same profile of a healthy old person who has no condition. So it's present both in disease and aging. And if we go down to detail, particular conditions will be particular signatures will be predictive of certain diseases. And they have a quite high AUC meaning accuracy. So for example, in hypertension, just by looking at glycans 98% of the time, we can predict somebody's developing hypertension six years ahead of time, in rheumatoid arthritis, just by looking at certain glycans, you can predict a diagnosis 10 years ahead of time, with 92% accuracy. So they will in the future become Clinical Biomarkers that we apply in medicine. To tell you you will have you're heading towards a problem. But they are modifiable. So they're not fixed enough fixed like our genes. For example, here, the signature we see predicting cardiovascular disease response to physical activity. And in women, the signature is very heritable. It's 70% heritable. And we still see significant response to physical activity. So this is a modifiable risk for diabeetus, the signature receipt predicting insulin resistance and diabetes responds to caloric restriction and weight loss. So we can tell somebody, Hey, you're heading towards this problem in 510 years time, it's time to do something about it now, before you actually get to a diagnosis, and you end up on a drug for the rest of your life. And then, how do we apply them to the individual. So if we do something exciting, and science doesn't always apply to the individual, so does this apply to the individual. And we have this evidence now, we've been tracking people longitudinally for decades. And this is one example of us. Mike Snyder from Stanford, who's tested himself 200 times over 10 years. And you can see how his structures are slowly changing with aging, some are going up and down. And then you have two events of age reversal. One is linked to a lifestyle change, physical activity running. The other one is linked to a diabetic drug, not Metformin. So you can see how stable it is, and then how well it responds with intervention. And in the case of Mike, we use the scientific method, meaning the analysis is run once, so you see a little bit of noise, but it's one to 3%. In the commercial method, we do triplicates, meaning that our error margin is less than 1%. So change in one year is a meaningful change. And you can see this as a woman's example. Women have different challenges to men, for example, pregnancy, we see accelerates aging by 10 years, but you can fully recover. And in this case, she was heading towards Hashimotos, but didn't have significant symptoms. So she managed it with diet. And when she reduces her fasting, she actually sees beneficial change, and her lamps go to normal. And what activates the changes for her is stress when she's stressed, she doesn't eat throughout the day.

And we've tested many different interventions and how you can change your licensing, like in the age, I can talk, we won't have time, only a few minutes, but I'll highlight a few. With a few warnings. There's no drug that has affected everyone. In fact, I'll switch to this 20 years ago, Dr. Alan roses said that majority of drugs don't work or 90% only work in 30 to 50% of people. And we will apply this principle in health care, but we haven't really applied it in preventative care. And the word of caution is Metformin, we have seen that in individuals, Metformin has a significant effect in this example of Tim Spector who's been tracked for 10 years. And you'll see that he has a drug with metformin, which continues to stay there while he's on the drug. But then, when we did a placebo controlled trial, with healthy overweight individuals, we only saw affecting one individual, or 10%, which wasn't significant. And then when we test the same drug in diabetics, we see 50% of people respond. But But diabetes, only 10% of the population with developed diabetes, there is another condition or natural event that will happen to 50% of the population. And that's menopause. And this age acceleration that we see with menopause is equal to the changes we see with disease and aging. And we do see that this can be rescued with estrogen. This is a placebo controlled trial with estrogen and you see significant effect. If you replace it, and you see accelerated aging if you don't, of course, when you look at it in a non controlled environment. So this is when in the clinic, majority of women are improved proving but then you see two people or two women going the opposite direction. And that's the case with every drug never works for everyone, even if it works as a general rule. Although I think estrogen is the only drug where we have hard evidence to say that it is a life standard. And this comes from the UK Biobank they looked at 406 commonly prescribed drugs and they identified 14 that prolonged lifespan, and now do these 14 for worse. And this is actually a response to a joke. So Steve had a conference last year where he said at the end of the conference, he'll tell us a joke. And at the end the joke was the best thing you can do for your longevity too. De is be a woman. And then I a few months later I came to see him as like Steve, I think we can improve your job. This is a, this is a cohort from Western University. It's a HIV biobank followed over 1015 years. And they look at glycans. And they've seen, they've had a small cohort of individuals born at male who transition into women, and go on straight estrogen. And interestingly, they had a follow up, where they see a change in the health outcome, which was less coronary plaque. So the hardest form of evidence we need with a new biomarker is that if you alter the biomarker with an intervention, it changes the outcome. And that's what we have here. You alter the biomarker, it changes in anti inflammatory way, and you change the health outcome. And this is the only evidence we have for an agent clock to predict an outcome in this way. But so the joke is, the best thing you can do for longevity today is maybe transition into it. But I don't doubt that many people will. So there is an alternative. And we see that actually the beneficial effect of testosterone is mediated rest reject. So if you blocked testosterone entirely men, and you give it back, it's only positive if you allow aromatase, if you block aromatase, then it is pro inflammatory. So this beneficial effect is mediated by estrogen. end to end, if you agree that the future of healthcare should be personalized, and it should be preventative. We believe that like this can unlock this future for us. They incorporate all of our unique influences from genetics epigenetics to environmental influences, they help us evaluate on an individual level if something does or doesn't work. And they also helped us preempt disease up to 10 years ahead of time. And that's it. Thank you.

Thank you, Nicola. That was fantastic. I think Nick Molina is also doing a fundraising round. And some of you may know that women only get 2% of all venture capital funding. So if you're interested, please find her. Now. In the meantime, be great to have some questions for Nicola. And I'm sure some of you do you have questions? Any questions for her? Professor Dennis noble? Nikki, could you

shout? The demonstration that like, like, I'm sorry, a better predictor of disease over a 10 year period is extremely important. And obviously needs to be looked at extremely carefully. And the reason I suggest that is that the study done on Poly genic scores as a prediction of disease being exceedingly disappointing. They were published by Hingorani, and his colleagues in the British Medical Journal just last October. And for the two big killers, cancer, and cardiovascular disease, the predictive ability of the polygenic scores is close to zero. Shocking. So people are going to be examining what you're claiming very carefully. Because what you're saying is true. It will change the way we look at predictive ability, fundamentally

100%. And so I didn't have time to go through a lot of data here. But of course, these are all research candidates. And the research has been going on for 15 years. And we probably need another 15 to get to where we want to go with any of these biomarkers. But we have shown in the mouse model, that if you altered lichens in the hypertension prediction, we started with correlation then with retrospective and prospective trials. And we've gone to an animal model where if you change these lichens, you actually change the disease. So we had a model of obese mice, where if we give them a certain glycan precursor, they they stay obese because of the diet, but they don't develop hypertension versus that. The one that the control that does so at least in animal models, we can show that this is actually a mechanism of disease. Now we need to do the same in humans. And I really hope that this field expands because glycobiology is so small, and we can only do so much as one lab. So we really need a lot more of the world looking into it so that we can get to its full potential.

Any other questions for Nikolina? yeah okay all right well Nicolina thank you so much