Professor of Genetics at Harvard Medical School and one of the most prominent geroscientists, George Church works on gene therapies that can potentially reverse age-related diseases. We had the opportunity to interview this prolific researcher and entrepreneur, who is involved in dozens of startups, on topics ranging from the current state of gene therapy to his recent attempt to auction off his genome, one of the first sequenced human genomes in the world, as an NFT.
What have been the successes and the failures of gene therapy in recent years? What do you expect to happen in the next few years?
So, most of the big failures of gene therapy happened at the very beginning, around the year 2000, almost two decades ago, when a couple of people died from an LMO2 oncogene, and one person died from an immune reaction to an adenovirus vector. So, that was 20 years ago. Fast forward to now, and gene therapies are mostly succeeding, hundreds of them are in clinical trials, you have dozens that have been approved by the FDA.
That’s gene therapy broadly, and there are even a few specifically CRISPR-based, which means editing. Gene therapy is, classically, adding a gene, while CRISPR is typically subtracting a gene, and those also are beginning to work for sickle cell disease and retinal disease, and so on. Both categories are looking pretty good recently, in the last few years.
There’s also a great deal of work on delivery, and you can even include, as honorary members of the gene therapy category, the Pfizer and Moderna mRNA gene delivery COVID vaccines. That’s obviously a huge hit. So, I would say that the gene therapy field is looking quite good.
Do you think that the success of the vaccines can change the momentum, maybe in terms of policy and regulation?
Absolutely. Also, the speed. This is radically new technology, and yet, messenger RNA for gene delivery was approved in ten months. And so, we can hope that this is generalizable to other, similar therapies.
Do we have any successes with gene therapy specifically in the rejuvenation field?
That’s a little further out. I mean, most are in pre-clinical animal trials or in very tiny, not yet approved human trials. But yes, pre-clinical animal trials are looking good. We published two papers, one on three genes that spread from the site to which they were delivered systemically, and another one about three genes that are localized; these are the so-called Yamanaka factors that cause rejuvenation. Those are two different studies, one headed by Noah Davidsohn and the other by David Sinclair.
Therapies for animals is something that one of your many start-ups, Rejuvenate Bio, has been doing. Recently, it has secured another 10 million in funding, and it has been alive for a few years, so what is the situation there?
Rejuvenate Bio, and Noah Davidson in particular, was involved in both of those papers, and those were both in mice, and they since have taken one of those two combination gene therapies, three different genes, things like fibroblast growth factor (FGF21), the soluble form of the TGFß receptor, and aKlotho. Anyway, that three-gene combination has then been moved into dog trials.
So, this is not an animal model, this is an actual veterinary product, because people do care quite a bit about their pets, and some of them would even clone their pets, but cloning is not really what you want. You don’t want another member of that breed, you want to have the adult live a healthier life. And, hopefully, in a couple of years, that will be available publicly for dogs, and about the same time, we’ll have the same drugs in clinical trials for humans.
You seem to be very interested – and invested – in new technologies like blockchain. How do you think information technology, and AI in particular, can help us fight aging?
Most of our work with homomorphic encryption and blockchain, NFTs and so forth, has been in Nebula, and it’s been focused on security of data, but the other aspect is machine learning and AI for designing therapeutics, for proteins in particular, and we published four papers recently, and those four papers launched three different companies: Nebula, Manifold Bio, and Dyno. And they range from making new enzymes to making new protein therapies to making new viral capsid delivery systems.
In each case, we don’t just use machine learning but large synthetic libraries, and that combination is much more powerful, and even each one of them separately, very synergistic. You can think of large libraries as a kind of an analog external device that allows you to test in the real world things that would be very difficult to simulate. And so, we can make libraries of millions of different viral capsids, for example, to find those that have the right tissue homing pattern.
So, things are moving towards in silico?
This is a combination of in silico and “in naturo”. It sometimes is called natural computing, when you use certain physical systems as a supplement. In silico normally means literally silicon-based, Von Neumann architecture, digital binary computers. But you can think of this as a whole new kind of computing where you can actually synthesize a lot of things that are very difficult to simulate.
In on article, you were called “a serial biotech entrepreneur”. Do you think an abundance of private initiative is the way to go when it comes to fighting aging, or maybe governments should play a greater role? Do you see any shifts in policy and regulation recently?
I think, and some of my colleagues agree, that we need to re-educate the FDA to be more interested in preventative medicine and in aging as a disease, and I think that’s a fine goal, but that could take time, and that’s uncertain. I think it’s easier and probably better to just accept that something that works on the core, fundamental components of aging will also reverse several different types of diseases simultaneously. We are working on eight different diseases.
In a way, diseases of aging are even better than biomarkers, because they are really what we care about. That doesn’t require the FDA to think revolutionary new thoughts or wrap their heads around something strange. It also has advantage over preventative medicine. I love preventative medicine, we work on it, but to convince a cautious federal agency to give a dose of something powerful to someone who is already healthy… This might make them less healthy. Toxicity will stop the clinical trial.
So, most preventative medicine has been very benign, a very “do no harm” sort of thing. But in this case, if you work on eight different diseases of aging, and some of them have a fairly early onset, and you can actually show reversal, you will get approval. Then you will have, as a side benefit, preventing all the other diseases. You will not only cure the early-onset disease, you will prevent all the others. That is, if you have the right thing.
That sounds to me a bit like the rationale behind the TAME trial.
Except they are trying to prove that aging is a disease and that it can be prevented. I think both of those are radical enough to require a change in policy. But it is a clinical trial, so it is working. That policy shift is working.
My understanding is that they are trying to prove that metformin delays the onset of all major aging-related diseases.
Well, delaying the time of onset is closely related to longevity, and I think that both delaying onset and longevity tend to require longer clinical trials. They tend to take longer, because the variation of age of onset is quite high, and to show that you’ve actually delayed onset, therefore, requires possibly decades to show. On the other hand, reversal of the early stages of a serious diseases could be shorter. You know, I’m hoping that all these approaches are going to work, it’s just that I think that the most conservative one is aiming at something that will affect multiple diseases.
Now, the problem is that metformin and, probably, the gene therapy we’re working on, as well may only affect a certain fraction of the aging process. Our colleague Olshanski has said that if you completely eliminate cancer, which is an age-related disease, you might extend life by three years. So, you may have to do many things right all at once in order to bring it down. For example, if there are ten different predators that eat you, you may have to extinct all ten of them to actually be able to survive. It’s not sufficient to take out one or two.
It’s the theory of competing risks.
So, you are saying that the gene therapy you are working on is about eliminating each one of those predators?
All at once. The nice thing about gene therapy is that there is a direct connection between most of the phenomena that has been noted for, say, the nine different pathways of aging. Those are associated with genes, and so there is a direct connection between the gene associated with the disease and the therapy, while if you go through a small molecule, you’re depending on serendipity or a very exhaustive screening of small molecules to find the ones that do the right thing. So, metformin was serendipitous, rapamycin was serendipitous, this wasn’t something from first principles, the way gene therapy is.
Circling back to your numerous startups: do you fear what many people call “the valley of death”: the loss of interest that might happen in a few years if there are no major successes?
I think that’s true for any category of therapy, but there are relatively few cases when the entire category fails. Usually, it’s individual drugs. I think it’s important that we have some kind of success. We don’t say: “If we don’t get everybody to be 150-years-old, we failed”. We instead say: “Well, if we can help dogs avoid microvalve disease, then we’ve got a billion-dollar drug. If we help humans avoid several diseases of aging and extend their lives by two or five years, that’s enough of a success to drive the field forward”.
So, “the valley of death” is only if you get nothing, and the thing is, we’re in a more productive place than we’ve ever been. By “we” I mean the whole field. We have exponential technologies for reading and writing DNA, we have amazing breakthroughs in cell biology and new categories of therapies and delivery vehicles.
In your last interview with us, you said that you don’t believe in a maximal lifespan for humans: “there is no law of physics or other reason for an upper barrier”. Yet, many people in our field shy away from such bold conclusions. Some say that even talking about that alienates people and hurts our cause. What is your take on this?
That’s just a statement on physics. Balancing that, I clearly state that my goal is not longevity, not even modest longevity. It’s just reversal of diseases of aging, which really is classic medicine. I was just answering the very specific question posed to me – do you know what the upper limit is? And the answer is that I don’t know of an upper limit other than maybe the heat death of the universe. But that doesn’t mean that I’m aiming for that. I’m aiming for something that could get FDA approval in a few years, which is the reversal of multiple diseases of aging.
Which takes me to the next question: do we even know how to aim at life extension?
I don’t think we do. I think if we get serious aging reversal, it’s something that we can continue to improve on, just like we improved on transportation from the first wheel to rocket ships, or the way we moved from being able to sequence a few base pairs of DNA to being able to sequence the entire biosphere. The thing is to get the foot in the door where we’re actually working on the core processes of aging, on the clock that decides that a mouse is going to die in two years and bowhead whales die in two hundred.
If we get to that core thing, then we can keep improving it, and, if you keep reversing, there is a chance that you can do that for a very long time. But I don’t think this should be the goal. It’s hard to do clinical trials on that. The trial where you show that you’ve extended life by even fifty years would be a very expensive, very long clinical trial. So, let’s just focus on things we can do in weeks.
Reversal of aging and age-related diseases – do you think any of this will happen in the nearest future, do you have a forecast to make?
It’s hard to predict the outcome of a clinical trial, that’s why we do clinical trials. You could predict when they could start, and then you can predict that they will last the normal period of time. And we have a couple of small molecule drugs already in clinical trials, and we have these gene therapies in pre-clinical animal trials. We should be getting results from those in the next four or five years – results, not final approval for general use.
That’s probably a decade away before we have that kind of feedback. But that’s just about feedback, we can’t predict how many diseases it will affect or how many years it will reverse. That’s going to be empirical more than theoretical. We have a solid theoretical foundation based on a lot of experience, but it’s now in the hands of clinical trials.
At least, like you said, we’re in a much better place as a field now than 20 years ago.
Yes, we’re not only in a better shape in terms of knowledge of age-related diseases, we’re also in a much better shape in terms of exponential improvement in the technologies that we need for testing, for synthesis of many different. We have so many shots on goal, possibly thousands of drugs in development simultaneously.
I just have to ask: you made waves by auctioning off your genome as an NFT. The auction was supposed to happen on June 10th. Was it delayed or cancelled?
That was premised on a new technology and people’s enthusiasm for it. I think the NFT market had a correction around the time that we were thinking about doing it. But it did achieve the goal of getting people interested again in some of the things that you could do with these security measures. Our main focus is on allowing people to share parts of their genome without giving away their genome. So, it allows the physician to look at the parts they need.
It allows, on occasion, researchers partial access without sacrificing the overall security of your genome. You should own your genome, and you should only allow it to be used by people you want it to be used by. So that was the main message. Because we have this strange situation where we have this incredibly affordable genome now, it has come down from three billion dollars for a poor-quality genome to 300 dollars for a clinical-grade diploid genome, and still not that many people know that or take advantage of it.
I thought that might be the reason behind you doing that. So, NFT for you is a technology that can be put to actual use – for instance, by allowing people to get paid for their genomic information?
Well, to be compensated for their genomic information. Some people will be doing it for purely altruistic reasons, while others should be properly compensated, and that should be done in a secure manner.