For Joe Betts-LaCroix, the answer to every problem seems to be a startup — and an unconventional one at that. An autodidact raised in Oregon and shaped by his eccentric parents, he took a circuitous path to being a founder.
Growing up, he mostly hated school. After graduating, he spent years reading books and tinkering with electronics in the basement of a home he shared with “musicians, artists, and weirdos” before enrolling at a local college in Massachusetts. Betts-LaCroix then transferred to Harvard, where he fell in love with college and structured learning. He earned his undergraduate degree in environmental geoscience and later conducted research in biophysics and robotics systems at other institutions.
In 2000, Betts-LaCroix founded his first startup, OQO Inc. The company made tiny personal computers and received plenty of accolades. But he wanted to do something bigger, more important. “I wanted to do something,” he said, “that I wouldn’t end up feeling cynical about afterward.”
“We want the hard task. We want to help healthy people stay healthy longer.”
Joe Betts-LaCroix
And so, in the early 2010s, he began diving into the scientific research on aging and longevity. He was shocked to find the conversation in Silicon Valley was steeped in marketing, not actual science. “Evidence in tech is pretty straightforward,” he said. “Biology is this kind of black box.”
After studying the literature and starting a nonprofit that held salons with leading longevity scientists and researchers, he decided to found a startup — Retro Biosciences — with an ambitious goal: to find a way to add 10 years of healthy, vital life to humanity.
For the first few years, the company operated in secret. Backed by $180 million in funding from OpenAI co-founder Sam Altman, it eschewed a more traditional strategy of focusing on one treatment or drug. Instead, it pursued five — a riskier, more expensive approach. The result: Today, Retro Biosciences is at the vanguard of longevity research and ready to go to clinical trials with a promising treatment for Alzheimer’s and other neurodegenerative diseases. Freethink sat down with Betts-LaCroix to talk about artificial intelligence, cell rejuvenation, and young blood.
You have a very ambitious goal of adding 10 years to people’s lives. How did you arrive at it?
We had a bunch of meetings and discussions early on about what does this really mean? And for whom? But it’s kind of cheating to say, “Oh, we’re gonna take extremely unhealthy people and increase their lifespan by taking away the things that make them extremely unhealthy.” We decided it wasn’t our job to extend the healthy lifespan specifically of smokers, because if that were the case, we should probably make some kind of behavior modification app. We want the hard task. We want to help healthy people stay healthy longer.
I guess some other questions that pop up around our goal are, “Okay, why not nine years or 11 years?” Or the kooks on one side will say, “Why aren’t you helping humanity live to 1,000 years?” And I’m like, “‘Cause I don’t know how.” And on the other side, they’re saying, “It’s hubris to say that you’re going to make such a change.” I think 10 years is actually a really nice sweet spot. I’d classify us as critical optimists. Most things, especially in biology, don’t work. So it’s a lot of work to keep plowing through and find the few that do.
How do you understand aging from an evolutionary perspective?
I don’t think there’s a single magic bullet for a healthy person to just add 10 years to your healthspan. Aging is essentially a default process. It’s everything else that goes wrong after you subtract the needed health to get to reproduction. We’re selected to get to the point that we can successfully raise a generation or two of kids and support them through their growing-up process. And in order to get there, evolution has granted us a bunch of really incredible powers. We can repair breaks in DNA, and we can heal our skin when it gets sliced. And we have these regulatory mechanisms that keep our blood sugar within a range of not too high, not too low. It’s amazing what has evolved, but only as much as needed. It’s sort of like a Toyota that’s designed to go for 200,000 kilometers or so, then a bunch of stuff is wearing out.
“A 95-year-old person can start having zero-year-old blood.”
Joe Betts-LaCroix
There have been evolutionary incentives for other species to live longer. Bowhead whales live on the order of, say, 230 years. There’s just a different predator-food ratio environment in the sea, and it makes sense for them to have longer lifespans. And so they do better at the cellular level, including repairing DNA mutations. There’s some very real possibilities, for instance, in borrowing some of the mutations that they’ve developed throughout their evolution and bringing those processes or amplifying those processes back in humans.
Our biology isn’t that different from theirs. It wasn’t that long ago that we diverged. And it may be a lot easier to help people live, on average, 10 more years because we’re able to pick up the low-hanging fruit and fix a few things. Technology is moving really, really fast. It’s pretty exciting to work on.
What are some of the most promising rejuvenation treatments you’re working on?
There are a lot of really promising avenues that we plan to release as therapeutics. The first thing that we’ll bring to humans is based on autophagy, which is a cellular process of waste recycling. It converts old, broken, accumulated waste proteins, which can be toxic, and breaks them back down into the amino acids that make up proteins. And so you can make new proteins out of them.
We’re working on a particular stage of the autophagy process that tends to be broken in age-related diseases — when protein waste is building up inside neurons and the lysosomes. We have discovered a compound that restores the acidity in lysosomes, and this is a known problem in, for instance, Alzheimer’s disease. For that program, we have this compound called RTR242, and we have been through tons of preclinical testing and proving that it works in various animal models. We’re now gearing up to do a clinical trial in Australia around the end of the year.
We have incredibly positive data from the earlier stages. We’ve shown in Alzheimer’s disease models in mice that the lysosomal acidity is restored by our compound. We’ve also shown that in healthy mice it’s not raised to unhealthy levels. So it has the ability to do this selectively, which creates a good ratio of effectiveness to toxicity. We’ve also shown that it operates more generally, or at the mechanistic level, so that it can apply to other diseases where the late-stage autophagy process has failed. So that’s really encouraging.
The immune system is obviously critical to healthy aging. How is Retro Biosciences trying to address its breakdown with biotechnology?
There aren’t many cell types that can be replaced in the body. Fortunately, blood stem cells are amenable. Eighty percent of all the human cells in your body are from your blood. If you have 95-year-old blood stem cells, the resulting system you get is not as good as the one you would get if you were 26.
Blood stem cells are a pathway that’s been plowed for decades by clinicians working to steadily improve the process known as a bone marrow transplant, motivated by patients with leukemia. Typically, they replace the blood stem cells — and all the progenitor cells as well — with an extremely harsh process that’s worth it if you’re otherwise going to die. And then you get a transplant of blood stem cells from some other person that’s as closely matched as possible from an immune perspective. But it’s never perfect, except in the case of identical twins.
“I’d classify us as critical optimists.”
Joe Betts-LaCroix
We can now take cells from your own body and reprogram them back into pluripotent stem cells — like the kind that made you in the first place. Every one of you started off as a single cell, and that cell then divided and divided and differentiated. That’s the process of differentiation. In 2006, this brilliant lab in Japan run by this dude Shinya Yamanaka figured out how to reverse that process. They could take one of these highly specialized cells that is already developed in us as an adult and turn back the differentiation and make it back into a pluripotent stem cell.
Our process for replacing the old immune system with a young immune system is based on first taking one of your adult cells, reprogramming it back to a pluripotent stem cell, then a very complex, delicate process of differentiating it back into a blood stem cell and then reintroducing it into your body. One thing that happens when you reprogram an adult cell back into a pluripotent stem cell: the adult age of the starting cell is wiped away, and you have these pluripotent stem cells that are basically age zero. That means a 95-year-old person can start having zero-year-old blood. It’s one of these holy grail projects that the stem-cell community has been working on for decades, but only just as of late last year, we finally cracked it. It’s incredible. And that means that we can go through this whole process of making your own blood stem cells that are zero age that you can replace your old ones with.
How does artificial intelligence factor into the work that you’re doing?
After the genome was sequenced, one of the big changes that hit biology was that it started becoming an information technology. It went from being this very analog kind of thing, with test tubes and increases and decreases in degree of various chemical concentrations, to suddenly being a string of numbers. Around the mid-teens, gene sequencing started getting cheap enough that you could sequence every cell in a whole test tube of cells, which means that we started getting incredibly large quantities of biological information.
I love hanging out with brilliant professors and researchers in the biological fields. But I just kept having this nagging feeling that each one of them kind of only understands maybe a few dozen genes. Since we have 20,000 genes — and they’re all interacting with each other, and there’s actually millions of proteins generated from them — I don’t care how brilliant of a Harvard professor you are, we’re gonna need help here. The question is how to model biology in a way that’s useful. We needed higher-level models. And the higher-level model that emerged in 2022 was ChatGPT.
AI is really good at looking at linear strings of stuff ’cause that’s what text is. We decided to partner with OpenAI to see if we could build a bionative model that could start to understand some things about how proteins work. And so we trained a small version of GPT-4 with this giant mountain of bionative data and then started looking at the outputs. Of course, it’s super young and untamed, and it’s gonna be years before we can figure out how to steer it to do exactly the kinds of things we want it to do and so on. But it’s a little surge of optimism. Maybe we could use that to start trying to figure out how we can fix some things that are broken in biology?
Where do you think your company will be 10 years from now?
It’s impossible to predict the future, but I would say we have up to three drugs on the market that can significantly improve human health and especially increase the likelihood that people can still be healthy and feeling good in later years in their lives.
