What It Will Take to Create 21st-Century Mammoths, Dodos, and Thylacines

Colossal Biosciences has generated a flurry of headlines in recent years, as the ‘de-extinction’ company announced plans to resurrect the woolly mammoth, the Tasmanian tiger, and, most recently, the dodo bird, developing a bioengineering toolkit along the way that has prompted investment from outfits like In-Q-Tel, a CIA-funded venture capital firm. Colossal has also acquired a stellar lineup of geneticists, including leading paleogeneticist Beth Shapiro, to help it in its quest to see these proxies of extinct species walk the Earth.

Last month, Shapiro—author of How to Clone a Mammoth: The Science of De-Extinction (2015) and Life As We Made It (2021)—leveled up her involvement with the company from an advisory capacity to its chief science officer.

While an exact version of an extinct animal cannot be created, scientists hope they can (to paraphrase the line from Moneyball) recreate the creatures in the aggregate. That means endowing Asian elephants with the long hair and cold resistance of a mammoth and making facsimile dodos spring forth from chicken eggs. Just last month, Colossal said it had engineered elephant stem cells that can be converted into an embryonic state, a big step toward its beyond-elephantine goal. In April, the company said it would give $7.5 million in 2024 to academic institutions undertaking ancient DNA research.

Shapiro recently spoke with Gizmodo about Colossal’s goals and her new role at the company. Below is our conversation, lightly edited for clarity.

Gizmodo: Things are moving so fast. When we last spoke, the dodo project had not even been announced. There was this open question of, well, how do you even go about de-extinction with birds? Colossal CEO Ben Lamm recently said that he thinks it’s more likely that we’ll have a dodo before a mammoth, just due to the artificial womb issue.

Shapiro: Artificial womb technology seems pretty hard. But that is so cool. Like the ability to try to figure out the placental interface and really understand some really foundational biology is exciting to me. I mean, that’s a field that I’ve never imagined that I would be in. And then, when I look at that team that’s working on the artificial womb, it’s engineers and developmental biologists and people who really care about trying to figure this out. It’s impressive. But yes, that’s probably a long time frame. The timing of a different species really varies. For every species that’s a candidate for de-extinction, there are different technical and ethical and ecological challenges associated with them. If we’re just focusing on the technology to get us to a gene-edited embryo, there are different hurdles with birds, as you say.

The strategy that Colossal—as well as some other academic research teams around the world—are using is this strategy to edit primordial germ cells. Primordial germ cells are cells that will eventually be either sperm or eggs, depending on the biological sex of the embryo. When a chicken egg is laid, that embryo is about 24 hours old. At that point, you couldn’t just edit it. There’s too many different cell types, there’s too many cells. You just couldn’t do that. But these primordial germ cells are migrating around the outside of the embryo, trying to establish themselves in the gonads that are developing at that point. And then, you can stick a needle into the egg and suck out some of those primordial germ cells without injuring the developing embryo, and then you can inject those into a dish in a lab. With the right culture conditions, those cells will survive. And you can edit them. Then, you can reinject them into an embryo at the same developmental stage, where they will migrate around the outside of the embryo, establish in the gonads. That chick’s DNA will be perfectly normal, unedited, but some of its gonads—and, if we’re using a lineage that doesn’t make any of its own gonads, which is the goal, then all of its gonads—will be edited. You can then fertilize a chick with edited eggs, with edited sperm, and you will have offspring that contain those DNA edits.

Once we figure that out—and that is a technological hurdle that we need to figure out, the right culture conditions, how to get the edits in, et cetera—once once we get that down, then it’s all a little bit easier, because you have eggs, and fast generation times, and multiple generations, and things like that. That is way easier than an elephant that has a 22-month gestation, right?

Isaac Schultz, Gizmodo: You’ve been working with Colossal for a while, and you’re leaving a couple other big gigs to go full-time there. Why the switch, and why now?

Beth Shapiro: Ben has been trying to get me to come on board full time since the beginning, as I’ve been working with them as an advisor in my role as the lead paleogeneticist. It’s always been attractive. I’m really excited about the potential for developing tools that have direct application to biodiversity conservation. It would be great if a single person in an academic job could contribute to this, but the landscape is such that that just really isn’t possible. Being able to take the helm of science at Colossal… it’s way outside of the scope of anything that I would have been able to accomplish as an individual academic. I’ve seen the group develop and evolve, and I’ve just been consistently impressed. I wrote the book that basically said this was too hard and wasn’t going to happen, and I go and I see all the things that they’re doing and I think, “Wow. They’re actually going to get there.” And as these new tools and technologies develop, Ben has promised to make them available to conservation at no cost. Which is fantastic. Any progress that we can make with birds, for example—these are among the most endangered species that are out there, and yet we can’t really do some of the basic things that we need to do to make DNA edits to bird genomes. So if we can make some foundational discoveries, they have tremendous impact across biodiversity conservation.

I’ve been thinking about it for a long time, but it’s really hard to leave an academic role. You have a big lab and a lot of people who count on you. And I wanted to make sure that everybody who was in my group at UCSC has everything that they need to be able to finish their PhD or their current postdoc or whatever the training is. The timing for me wasn’t so much about when exactly I wanted to jump into Colossal, but making sure that I was taking care of everybody in my lab at UCSC.

Gizmodo: I want to hear about the time horizons of de-extinction, which is obviously what everyone is obsessed with. When is all of this happening? 

Shapiro: That’s also something that I’m really not in a position to comment on. I’m trying to figure out where everything is. There’s only one date that has actually been officially announced from Colossal, and that is that Eriona (Hysolli) and George (Church) and Ben confidently believe that they can have a mammoth by 2028. There’s a lot of scientific discovery that has to happen between now and then, and ideally we could predict exactly when we’re going to make discoveries and then we can build on those. But that’s just not the way biology works. Biology is dirty and complicated. It’s not like software.

Gizmodo: You mentioned an elephant’s 22-month gestation. In an artificial womb, would it still take 22 months, or can you can you accelerate the gestation process?

Shapiro: I don’t know. We don’t fully understand how to create an artificial placenta at the moment. We don’t really understand the intricacies of the developmental process. This is all information that we will learn along the path. I would assume for now that we need the 22 months, because there’s probably a lot of interesting biology that happens in those 22 months, and it’s a very large embryo that’s born. There has to be a lot of time for just resources to be turned into an animal. But this is something that we will learn.

Gizmodo: You wrote the book on how de-extinction was not possible, once upon a time. You’ve also written about how humans have changed the surface of this planet. How much has the de-extinction landscape changed since you wrote those books, since 2015 and even 2021?

Shapiro: There’s been a lot of progress in gene editing and the precision of on-target gene edits and being able to move large bits of DNA into a genome at the same time. All of that is stuff that we would need. There’s been a lot of work done in ancient DNA: we now have many more mammoth genomes, which makes it easier for us to compare all of the mammoth genomes that we have and all the elephant genomes that we have, and ask where all the mammoths are the same as each other but different from the elephants, which is helping us to narrow down what edits we would need to make. But we still don’t have an artificial womb. We’re still in the process of learning about if we need to use elephants or, if we need to use animals in any of these processes, exactly how we would do that. All of science has moved forward, and I think we have now the core of each of these technologies, but mostly developed for model organisms, or agricultural animals, or people. The core of the technologies are all there, but now it’s how do we get them to be applied to these species that we often don’t think about when it comes to developing tools for genome editing, embryo transfer, and things like that.

Gizmodo: It does feel kind of like—I don’t want to say chicken or the egg, just given that we’ve already kind of covered chicken and eggs—but there’s an interesting conversation happening between the technology that exists and the de-extinction projects. Because the de-extinction projects, as I understand it, are supposed to yield these new technologies to feed back onto the other side of that conversation.

Shapiro: De-extinction is a moonshot, right? Personally, I would love to see these technologies developed and applied to the conservation of species that are still alive. So how do we get there? We have to get there with a moonshot, with a crazy goal that we can direct all of our energy towards actually solving these problems.

I was at a meeting last week at (Geneva Science and Diplomacy Anticipation). We were talking about all of the ecosystems that are in peril around the world, and we can talk in circles forever about how we need these technologies, but we don’t really know where to start. If we just had a moonshot, like one ecosystem that we thought we could come together as an international community to save, then we start saying, “Okay, here is the list of technologies that we need to be to be working through, and this is a path to get there.” And that is what de-extinction is for genetic rescue. There is a moonshot that says, “We want to create a mammoth.” Well, what do we need to make a mammoth? We need advances in ancient DNA, we need advances in selecting loci for editing the genotype to phenotype relationship. We need advances in actually making edits to DNA and large-scale edits to DNA. We need advances in cell culture for elephants. We need advances in learning about how to actually have elephants be happy and healthy in captive breeding environments, if we’re going to go that way. We need to develop an artificial womb. All of these are technologies that have application across genetic rescue and also even human health landscapes. By giving us this moonshot—by saying we’re going to get to a mammoth—we have created a path. We have created a moonshot that forces us through these technologies in a way that I think otherwise we might not get there.

Gizmodo: You said that if you have one environment, you can develop a pathway to get there. Colossal’s working on several extinct species. So why those species, and how does the introduction of these species to habitats kind of rehab them?

Shapiro: The species are selected because they are really across the tree of life. We have a bird, we have a marsupial, and a placental mammal. These are three different species that have significantly different technical hurdles to get to the point where we’re going to have a gene-edited embryo. And so I think this allows us to try to develop technologies that are going to have application to taxa across the tree of life.

As far as the application to the landscapes, again, every species will have different ecological challenges associated with this. With the thylacine, for example, we have an environment where we have a recently extinct apex predator, and we know that that apex predator is still missing from that landscape, which is still sort of struggling to rebalance after the extinction of this individual. There are lots of opportunities to work with scientists to better understand what might happen when we reintroduce an apex predator into that landscape, to develop the tools that we would need to monitor what’s happening to that landscape, to work on developing relationships with local community members and indigenous groups, to see what they want in this landscape, and to collaborate with them on developing approaches to eventually release individuals into a landscape.

Gizmodo: When it comes to ecosystem monitoring, it seems like it would be in the company’s interest develop a digital twin, or something where you could see on fine scales exactly how the environment changes, depending on the number of species in the habitat, things like that.

Shapiro: There are modeling approaches that people have used before, not necessarily making digital twins. Ecosystems are complicated places, and your model can only be so complex and sophisticated that you can actually understand what it means when it goes wrong. There’s a chance that you make a model that’s so complicated that, when it breaks, you both haven’t learned anything about the ecosystem and you also haven’t learned about your model. So that’s not really useful scientifically. But it definitely is important.

Colossal has been working on a paper to try to estimate the carrying capacity of mammoths—the carrying capacity of Arctic ecosystems for mammoths—thinking about things like, how much food would there be? How much space would you need? How many other species are there? What would the feedbacks be as far as the climate goes? And so, there definitely is interest in trying to predict ecosystem impacts way before the potential of actually having any ecosystem impacts. Because clearly thinking about what would happen when we have animals that really are released on the landscape is critical to being able to make these projects move forward.

Gizmodo: A part of the proxy mammoth project is generating this ecosystem that hasn’t existed for a while. Colossal sells it as a form of climate change mitigation. Is that the idea for every species?

Shapiro: Different scientists have different opinions about the potential for impact on climate. I am in a camp that is different from the camp that George Church and Sergey Zimov are in, who really see mammoths as a potential for helping to the permafrost to cool down. I think we don’t really have enough data to know that that would be true. We don’t really understand the number of mammoths that we would need on the ecosystem. I think it’s important for us as a company to present all of the potential ideas that are out there, and then do the research that we need to figure out what the what the truth is.

I think for each species, though, there will be different ecological impacts. So with the dodo, for example: I can’t imagine that having dodos in Mauritius is going to have an enormous impact on global climate change. And so I think that answers your question, is every animal intended to deal with global climate change? No. The purpose of thylacine is to reintroduce an apex predator into an ecosystem and to help create a more resilient and robust ecosystem in the face of climate change. The same is probably true for a dodo and a mammoth. In my mind, that is what is most critical about genetic rescue technologies—and that includes de-extinction or creating proxy species for animals that are extinct—but also saving species that are in danger of becoming extinct. Because we know that ecosystems that are biodiverse, where there is redundancy in the trophic levels, where you have redundancy of the energy flow and the food web, are more resilient ecosystems. And so by creating approaches, by creating tools that can be applied to ensure that we have a future that is both biodiverse and filled with people, this is what I think these goals are.

Now, for every species there is a different ecological outcome and a different ecological example. In Mauritius, for example, the Mauritian Wildlife Foundation has partnered with Colossal for thinking about dodo rewilding, which includes identifying habitats that dodos might be able to survive. Dodos became extinct because they’re a flightless bird that lays a single egg in a nest on the ground. And when people introduced rats and cats and pigs, they just ate the eggs that were on the ground. We know that if dodos are going to be able to be rewilded, we’re going to have to have a habitat that doesn’t have those particular invasive species in it. Now, the Mauritian Wildlife Foundation has a fantastic track record of being able to do on-the-ground conservation work in Mauritius. There’s an island right now called Île aux Aigrettes, where they’ve removed invasive species and even reintroduced giant tortoises, which is another species that became extinct in Mauritius. And they’ve seen that, after reintroducing giant tortoises, that a lot of the native plants started to rebound. They discovered that this was because the tortoises were eating the ebony seeds, and by passing through the digestive system of the tortoise, the ebony was in a better place to be able to germinate. And so I think there will be surprising interactions that are restored with something like a dodo that we can’t predict, but also just by creating these habitats that are ready for a dodo has immediate benefits to other species that are endemic to Mauritius that are endangered at the moment, because now there are reinvigorated landscapes. Just the idea of having the dodo has caused an increased investment by the Mauritian Wildlife Foundation in some of the work that’s going on in Mauritius.

Gizmodo: Before we move away from the dodo, I have to ask about your dodo tattoo.

Shapiro: It’s here, on my arm.

Gizmodo: Oh, neat. It looks like a vintage illustration.

Shapiro: It is, my tattoo artist drew it from a book. It’s a scientific illustration.

Gizmodo: Back to the technical challenges. Each of the species has its own sizable challenges. What is the technology that’s going to take the longest to develop, for a de-extinction in one of these three species to occur? 

Shapiro: The foundation for all of these technologies exists. It’s just tweaking them so that they’re applicable to these species that people haven’t worked with before. I would say that the biggest challenge is probably going to be if we want to make, like, enormous changes. If we want to get as close as possible to the species that used to be alive, then we’re going to need to make a lot of changes to the genome, not just a few tweaks that bring back these core phenotypes, which honestly, I think is sufficient. If we can bring back the core phenotypes with a few tweaks, then great. I say that, ‘it’s done: we’ve done de-extinction.’ But some people are more purist. We work with Andrew Pask, who really wants to make all of the changes that you would need to get 100% of the way to a thylacine. And to do that will require new tools to insert large pieces of DNA into genomes or to synthesize artificial genomes entirely from scratch. I think those tools are probably the things that are the longest timeline. But those aren’t actually that critical to Colossal meeting the goal of reviving these core phenotypes from the species that we’re targeting.

Gizmodo: How do you work with the diversity of opinions within Colossal? How do you have these conversations as you’re developing the technologies?

Shapiro: We all have the same final goal. We want to develop these technologies so that we can get to de-extinction, or whatever many of us are willing to accept is de-extinction, of these three species. But we also want to develop these technologies because we care about the future of the planet, and we want to be able to apply them to conserve biodiversity. And I think it’s healthy to be able to have conversations about exactly how we’re going to get there and exactly what the consequences of this are going to be, because it keeps us on our toes. It makes us keep reading. It makes us keep engaging people that have different opinions to ours. It makes us keep having these conversations so that we’re in a place to be able to learn more.

One thing I’m most excited about in this particular role is exactly that. I’ve been doing kind of the same thing for the last 25 years in my academic job, and suddenly I’m doing so many things that I have never done before. I’ve thought about these things, but I’ve never been in a position where I really have to understand everything about them so that I can be informed enough to make decisions and advise people as they’re moving forward. And I love it. I am so excited about the possibility of jumping in and learning so many new things. I feel revitalized as a scientist. And feel like the diversity of opinions that you encounter helps you keep that drive, keep that sort of excitement in your career—because prove that I’m wrong and I will change my mind. That is that is how science should work. And if everybody agrees, that’s boring. It’s also not going to get us anywhere.

Gizmodo: Can the public expect any more species to be added to Colossal’s de-extinction agenda soon?

Shapiro: I think we’ve got our hands pretty full at the moment. But you never know. Not soon.

Gizmodo: Anything else you’d like to highlight?

Shapiro: Colossal has a fantastic portfolio of conservation programs. They’re working across the world with partners like Rewild, for example. And this work is really critical. It’s really important work that I am very excited to be engaged with and to be pushing forward. And even though it’s not getting us to a mammoth, it will help us on that path and it will help us when we get there.