Reviving Extinct Species: Beth Shapiro on the Science and Ethics Behind De-Extinction

When scientists announced the creation of a dire wolf — extinct for thousands of years — social media erupted. But for Beth Shapiro, Chief Science Officer at Colossal Biosciences, this is far more than a "resurrection game" involving mammoths and dodo birds. It's a toolkit with the potential to save endangered species and transform human medicine.

In an episode of Andrew Huberman's podcast, the evolutionary biologist and former Howard Hughes Medical Institute investigator laid out the technical logic and ethical considerations behind de-extinction. One line of hers captured the entire project's essence: "People always ask why we're bringing back extinct animals instead of helping living ones avoid extinction. The answer is: we're doing both. With the same tools, the same technology, the same needs."

How Was the Dire Wolf "Made"?

Shapiro described the technical roadmap for the dire wolf project in precise detail. The team extracted and sequenced a genome from dire wolf fossils, identified the genetic variants responsible for the animal's larger size, greater strength, and lighter coat color, then engineered those changes into the gray wolf genome to reconstruct the dire wolf.

"We made 20 carefully selected edits," she emphasized. This was not a matter of randomly grabbing a few dire wolf genes and dropping them into a gray wolf — it was an extremely deliberate process.

One particularly illuminating detail involves coat color. Sequencing revealed that the dire wolf carried two genetic variants that lightened its fur, but in living gray wolves and dogs, similar mutations in the corresponding genes can cause oculocutaneous albinism — meaning blindness or deafness. The team therefore abandoned direct replication and instead used editing approaches already known to be safe within a gray wolf genetic background (since light-colored domestic dogs and gray wolves already exist), achieving the same pale coat without the risk.

"It had to be safe within the gray wolf genetic background. If it didn't pass the animal welfare threshold, we wouldn't do it."

The Debate Over "Is This Really a Dire Wolf?"

After the dire wolf announcement, some scientists sharply criticized Shapiro, arguing she had "no right to call it a dire wolf." Her response was almost philosophical: species are a concept invented by humans, and "biology simply doesn't care what species you belong to."

She walked through several coexisting species concepts — the "biological species concept," based on the ability to produce fertile offspring; the "genetic species concept," based on a threshold of genomic similarity; and the "geographic species concept," based on spatial distribution. De-extinction work doesn't follow the genetic species concept. "We're not trying to make something genetically identical to any individual that ever lived."

Shapiro invoked Neanderthals to illustrate how blurry species boundaries really are. Modern humans and Neanderthals interbred, and most people today carry between 2% and 5% Neanderthal DNA. More intriguingly, the specific 2–5% varies from person to person — if you assembled all the Neanderthal DNA fragments carried by every living human, you could reconstruct over 90% of the Neanderthal genome. This suggests the vast majority of those genes are harmless to humans.

Her position was unambiguous: "If people don't want to call it a dire wolf, fine — it doesn't matter. As long as it looks like a dire wolf and can fill the dire wolf's ecological niche, I'm happy to call it one."

The Real Goal of De-Extinction: Restoring Ecosystems

Why mammoths and dodo birds rather than, say, directly restoring the Florida panther? Shapiro's answer spans technology, ethics, ecology, and society.

On the technical side, dinosaurs cannot be revived. The oldest DNA ever extracted comes from a mammoth bone roughly one to two million years old, while dinosaurs went extinct 66 million years ago — their bones long since mineralized into rock, leaving no DNA behind. After an organism dies, DNA is continuously fragmented by UV radiation, freeze-thaw cycles, and microbial activity; only the cold conditions of Arctic permafrost allow it to survive over geological timescales.

Ecologically, Shapiro cited Siberia's Pleistocene Park as an example: when large herbivores return to the tundra, their winter foraging — digging through snow to reach vegetation — keeps the permafrost cold and slows its thaw, ultimately supporting richer and more diverse plant communities. The famous case of reintroducing gray wolves to Yellowstone similarly demonstrated the cascading effects apex predators have on entire food webs.

The goal of reviving the thylacine (Tasmanian tiger) is also ecologically motivated. Without a top predator, Tasmanian devils have been decimated by a contagious facial tumor disease; had an apex predator remained, sick individuals would likely have been culled naturally.

Pleistocene Park is a long-term experiment launched in 1996 by Russian ecologist Sergey Zimov in Yakutia, Siberia. Its central hypothesis is that before the end of the last Ice Age, large herbivores — mammoths, wild horses, bison — maintained vast "mammoth steppe" ecosystems through trampling, snow compaction, and grazing. These ecosystems had high albedo and kept permafrost temperatures low, serving as a critical carbon sink in the Northern Hemisphere. After these animals went extinct, mosses and shrubs took over, increasing insulation and accelerating permafrost thaw, releasing large quantities of methane. Experimental data show that areas with animal activity have permafrost temperatures roughly 2°C lower in winter than control areas. One of the key arguments behind Colossal's mammoth de-extinction project is the hope of eventually introducing engineered Asian elephant–mammoth hybrids into the park to slow Arctic permafrost thaw in the context of climate change — making de-extinction not just a conservation issue but directly relevant to climate change mitigation.

The Same Toolkit, Saving Living Species

Shapiro repeatedly stressed that the technology stack developed for de-extinction — multiplex genome engineering, cellular reprogramming, and wildlife iPS cell technology — is fully applicable to protecting species alive today.

The black-footed ferret is a prime example. This once-critically endangered species was rebuilt from a tiny founding population, leaving it severely lacking in genetic diversity. Scientists used tissue samples preserved for decades in a frozen zoo to produce a clone (Elizabeth Ann, born in 2020) carrying genetic variants absent from the living population. Plans are also underway to use gene editing to give ferrets resistance to plague — domestic ferrets aren't susceptible, which means the genetic basis for resistance can be found and replicated.

Another clever example involves the northern quoll in Australia. The invasive cane toad is toxic, and quolls that eat one die. Yet on the other side of the planet, certain mammals can eat toxic toads without dying — thanks to a single nucleotide change in one gene that alters a single amino acid. Colossal's Australian partners have already made that change in quolls, potentially allowing them to both avoid extinction and exploit cane toads as a food source.

"If we say these technologies are too risky, we're accepting 'do nothing' as the outcome — and that is also a decision." Shapiro pointed out that the only self-rescue mechanism available to endangered species is natural selection, but ecosystems are changing far faster than natural selection can keep up.

Wildlife iPS cell (induced pluripotent stem cell) technology is one of the core elements of this toolkit. iPS technology was pioneered by Japanese scientist Shinya Yamanaka in 2006. The principle is to "reprogram" differentiated somatic cells (such as skin cells) back to an embryonic stem cell-like pluripotent state by introducing specific transcription factors, after which they can be re-differentiated into eggs, sperm, or any tissue type. Applying this to wildlife means that even after an animal has died, as long as a small amount of skin or blood was cryopreserved, it may be possible to reconstruct gametes and even offspring in the future. This gives the "Frozen Zoo" concept an entirely new significance — the San Diego Zoo has been cryopreserving cell lines from hundreds of species since the 1970s, originally as an archive, but now a functional "genetic rescue library." The black-footed ferret case used exactly this kind of tissue collected and frozen decades ago; the cloned Elizabeth Ann carries alleles no longer present in the existing population, effectively "transfusing" new genetic diversity into a severely inbred species.

When the Technology Points at Humanity

The conversation reached its deepest point when it turned to human gene editing. Shapiro mentioned the case of "KJ," an infant with a urea cycle disorder who, through a collaboration among academia, industry, the NIH, and Children's Hospital of Philadelphia, received a customized CRISPR base-editing therapy at six months of age — and was cured. "Nobody accused those doctors of playing God."

She argued that humans have always been performing genetic selection — through mate choice. She cited height as an example: the tall-stature genes prevalent in Northern European populations were introduced roughly 4,700 years ago by steppe peoples (the Yamnaya culture). This is visible evidence of human genetic selection in action, "not because anyone was picking embryos in a petri dish."

On the question of social acceptance, both Huberman and Shapiro noted an interesting pattern: the real barrier is usually the delivery mechanism, not the technology itself. GLP-class drugs in pen-form injectors eliminated needle phobia and rapidly destigmatized the entire peptide wellness space. By the same logic, once brain-computer interfaces become non-invasive, the fear of "controlling the brain" will fade.

Base editing is a second-generation precision tool built on CRISPR, developed by the David Liu lab in 2016. Traditional CRISPR-Cas9 cuts both strands of DNA and relies on the cell's own repair machinery, which carries the risk of insertions or deletions. Base editing requires no double-strand break; instead, it chemically converts a specific base directly into another (e.g., C→T or A→G) — like using an eraser and pencil to correct a single letter without tearing the page — dramatically improving precision and safety. The urea cycle disorder KJ suffered from (OTC deficiency) is an X-linked genetic disease; patients cannot metabolize ammonia produced from protein breakdown, and without treatment can rapidly suffer brain damage or death. The milestone significance of this case lies in the fact that it was the first in vivo base-editing therapy designed from scratch for a single patient — researchers completed therapy design, lipid nanoparticle delivery system construction, and safety testing within months of diagnosis, demonstrating the rapid translational potential of personalized gene therapy from lab to clinic and providing a methodological template for treating a broad range of single-gene disorders.

Artificial Wombs and AI-Simulated Ecosystems

To enable gestating hundreds of mammoths simultaneously without occupying an Asian elephant's 22-month pregnancy, Colossal is developing artificial womb technology. Shapiro acknowledged this alarms many people, but pointed out the profound implications for human medicine — for instance, allowing a woman diagnosed with breast cancer during pregnancy to begin treatment early, or allowing a fetus requiring surgery to develop safely outside the womb. Modern NICUs (neonatal intensive care units) are, in a sense, already a primitive form of artificial womb.

Facing the uncertainty inherent in ecosystem interventions, Shapiro was open to using AI to build digital twins of ecosystems: "That's entirely within the realm of feasibility." Colossal already has a dedicated team producing "care reports" that deeply analyze the short-, medium-, and long-term effects of releasing de-extincted animals into ecosystems, in full consultation with local stakeholders and conservation biologists.

It's also worth noting that Colossal simultaneously announced the cloning of red wolves — the most endangered wolf in North America. By cloning coyote-like animals carrying more than 75% red wolf ancestry, the team is introducing new genetic diversity into the existing red wolf population. "To everyone who says 'you should be doing this instead of that' — take note: we're doing all of it."