Bringing Extinct Species Back to Life | Dr. Beth Shapiro

Episode Overview

• Stanford Professor of Neurobiology and Ophthalmology Dr. Andrew Huberman (who discloses at 0:02:17 that this podcast is separate from his teaching and research roles at Stanford) interviews evolutionary biologist Dr. Beth Shapiro, Chief Scientific Officer at Colossal Biosciences (which employs about 120 scientists), former UC Santa Cruz professor and HHMI Investigator, former board member of Revive & Restore, National Geographic Explorer, and author of How to Clone a Mammoth and Life as We Made It.
• The conversation explores the science, ethics, and ecological utility of using ancient DNA, genomics, and synthetic biology to "de-extinct" vanished animals—including the dire wolf, woolly mammoth, dodo bird, thylacine, moa, and bluebuck—while demonstrating that these exact same technologies are used to rescue endangered living species.
• Beginning with foundational evolutionary biology—such as why "species" is a human construct, how modern humans carry 2% to 5% Neanderthal DNA, and the physical limits of ancient DNA preservation—the episode transitions into real-world genetic engineering breakthroughs, including Colossal's three living dire wolves and cloned red wolves.
• The narrative culminates in urgent applications for ecosystem restoration and human medicine, highlighting how single-gene edits can protect endangered marsupials from invasive toxins, how cloning 40-year-old tissue rescued the black-footed ferret, and how a bespoke CRISPR base editor cured an infant ("Baby KJ") of a lethal genetic disease.

Key Concepts

• The Fluidity of "Species" and Taxonomic Naming: Biology does not care what species an organism belongs to; "species" is a human classification box created for communication. While Ernst Mayr's "biological species concept" defines a species by the ability to interbreed and produce fertile offspring, biologists also rely on the "genetic species concept" (DNA similarity thresholds) and the "geographic species concept" (such as distinguishing a Florida panther from a Texas panther for conservation). Historical naming conventions by Carl Linnaeus also yielded repetitive Latin names like Bison bison (and Bison bison bison for the plains bison), Gorilla gorilla, and Lama glama / Lama lama, while Europeans called unrelated African, Asian, and American animals "buffalo" simply because their hides looked like they would make a good "boof" coat.
• Polar Bear and Brown Bear Hybridization: Although brown bears and polar bears diverged about 500,000 years ago, they produce fertile "grolar" or "pizzly" bear hybrids whenever their habitats overlap. Ancient DNA from a >100,000-year-old Alaskan polar bear bone revealed ancestry in brown bears worldwide, and during the last Ice Age (~20,000 years ago), polar bears trapped on Southeast Alaska's ABC Islands interbred with expanding mainland brown bears, leaving all modern North American brown bears with polar bear ancestry. However, interbreeding only occurs between an induced-ovulator female polar bear and a male brown bear, and hybrids only survive as brown bears—likely because hybrids with brown bear DNA lack pure white fur and cannot successfully hunt seals on sea ice.
• Selective Breeding and Canine Biology: Human intervention has dramatically reshaped domestic dogs. For instance, the English Bulldog was bred as a cross between the Bullmastiff and the Pug for bull-baiting (selecting for a short snout to clamp down, a floppy face lacking pain receptors, and a low center of gravity), resulting in heads so large and hips so narrow that females require Cesarean sections to give birth. On the topic of canine behavior and genetics, Broad Institute and UMass researcher Elinor Karlsson showed that breed behavioral stereotypes do not map cleanly to genes; dog behaviorist Cesar Millan noted that different decades in the US demonized German Shepherds (post-WWII), Dobermans, English Bulldogs, and Pit Bulls, while Dr. Huberman's veterinarian uncle notes that most bites statistically come from Chows. Regarding canine reproductive health, Dr. Huberman notes his own male dog is 7 months old and intact, citing a previous guest from the Karolinska Institute who stated that in Scandinavia it is illegal to neuter a male dog without a health reason (whereas in Australia the inverse is true), and that male dogs can be given a less invasive vasectomy so they retain hormones for brain development.
• Archaic Human Admixture and the 5% That Makes Us Human: Dr. Shapiro's husband, Ed Green, worked in Svante Pääbo's group when they assembled the Neanderthal genome and identified the Denisovan lineage from a pinky bone in Russia's Denisova Cave. Humans shared a common ancestor with chimpanzees and bonobos 3 to 5 million years ago (accumulating about 100 genetic differences per generation from copying errors), but diverged from Neanderthals 300,000 to 500,000 years ago. When Homo sapiens dispersed out of Africa into Europe, they interbred with Neanderthals (Dr. Shapiro notes she does not know what is known about hip or birth canal size restricting archaic human interbreeding). While an individual modern human carries 2% to 5% Neanderthal DNA (averaging ~3%), stitching together all fragments across all living humans reconstructs over 90% to 95% of the Neanderthal genome. The remaining ~5% where no living human carries Neanderthal DNA pinpoints the exact genomic regions where a fetus required the modern human version to survive, shortening the evolutionary window for studying uniquely human traits by an order of magnitude.
• Ancient DNA and Modern Human Traits/Disease: Some Neanderthal alleles rose to high frequency in modern humans via positive selection: variants in Latin American populations predispose carriers to Type 2 diabetes; a variant studied in a Colombian trial alters pain sensation (Dr. Shapiro hedges she cannot remember whether it caused more or less pain); and during the COVID-19 pandemic, a Neanderthal-derived allele at ~50% frequency in Asian populations caused severe outcomes by making lungs more susceptible to viral entry, while another Neanderthal allele was protective. Both Neanderthals and woolly mammoths also carried variants in the MC1R gene associated with reddish hair (though Dr. Shapiro notes she does not know if mammoth mummies bleached red or were genetically red, adding there is genomic evidence for reddish hair, while Dr. Huberman notes redheads on average require more anesthesia). Additionally, scientists isolate ancient plague DNA from European graveyards to study immune evolution, and ancient human DNA shows that tall stature genes were introduced to Europe ~4,700 years ago by migrating Yamnaya Steppe people and driven to high frequency in Northern Europe (such as the Netherlands, where height has plateaued) via mate selection— just as Dr. Shapiro suspects blue eyes evolved via sexual selection (noting she is 5 feet tall).
• Physical Limits of Ancient DNA Preservation: Dinosaurs went extinct over 66 million years ago and their bones have mineralized into rock, making dinosaur de-extinction impossible. After death, DNA is degraded by UV light (as proofreading enzymes lose energy), freeze-thaw cycles (water expansion physically shatters strands), and microbial decay from bacteria and fungi. The oldest DNA ever recovered comes from a mammoth bone dating to roughly 1 to 2 million years ago (Dr. Shapiro hedges that exact dating in that time window is difficult). While hot, wet environments destroy DNA—causing Dr. Shapiro to fail at extracting DNA from hundreds of dodo bones from swampy Mauritius—Colossal sequenced a high-quality dodo genome from a specimen in the Danish Museum of Natural History that had been brought to Europe alive on a ship. Interestingly, the very first ancient DNA paper was published in 1984 by Allan Wilson's group at UC Berkeley, who sequenced a tiny DNA fragment from the dried skin of an extinct quagga.
• Avian Genetics, Sex Determination, and Parthenogenesis: Birds are among the most endangered animals on Earth, yet they cannot be cloned via somatic cell nuclear transfer (SCNT, used for Dolly the sheep) because their egg cells cannot be accessed at the right developmental stage. Avian genomes also feature tiny microchromosomes and, in some species, a "germline-restricted chromosome" present only in sperm and egg cells. Across vertebrates, sex determination evolved multiple times: mammals use XY (where males carry the Y-linked SRY gene, and fertility clinics can sort lighter Y-bearing sperm from heavier X-bearing sperm via centrifuge before ICSI), birds use WZ (females are heterogametic), and alligators and crocodiles use temperature-dependent incubation. In a famous case in the California condor, a female produced a viable chick without a male (parthenogenesis) because meiosis failed to fully separate her chromosomes, leaving two maternal copies in the egg. In humans, inheriting two copies of a chromosome from one parent (uniparental disomy/imprinting) causes disorders like Prader-Willi or Angelman syndrome (while Trisomy 21 causes Down syndrome), and a mammal could never parthenogenetically produce a male without a Y chromosome.
• Engineering the Woolly Mammoth and Dire Wolf: Whole genomes from 1.5-million-year-old mammoths show that the woolly mammoth's closest living relative is the Asian elephant—they are closer to each other than Asian elephants are to African elephants. Their genomes are ~99% similar (comparable to the ~99% similarity between humans and chimpanzees, depending on how it is calculated; Dr. Shapiro also mentions the popular claim that humans share 85% of their DNA with a banana, immediately hedging that she does not know if that is true). Because individual mammoths differ by millions of DNA letters, Colossal focuses on editing the fixed differences that make a mammoth a mammoth into Asian elephant cells, while also developing artificial wombs (currently in mice) so they do not have to rely on 100 Asian elephants—which have a 22-month gestation and reach sexual maturity at age 14—as surrogates. Similarly, Colossal created three living dire wolves—twin males Romulus and Remus (almost 2 years old, weighing at least 120 pounds at one point, though Dr. Shapiro hedges she does not have their latest measurements) and an ~18-month-old female named Khaleesi—by making 20 edits to a gray wolf genome. Because the exact ancient dire wolf genes for a light coat risk causing oculocutaneous albinism (blindness and deafness) in gray wolves, Colossal engineered the white coat using safe genetic variants from living domestic dogs and gray wolves. The dire wolves are hand-reared, kept on hormones to prevent inbreeding, and will never be released into the wild to avoid competing with endangered gray wolves.
• Trophic Cascades and Megafauna Ecosystem Restoration: Reintroducing keystone species transforms physical landscapes. At Pleistocene Park in northeastern Siberia, Sergey and Nikita Zimov introduced Canadian bison, wild horses, deer, and muskox; in winter, these herbivores scrape away insulating snow to graze, exposing the ground to deep freeze, slowing summer permafrost melt, and fostering deep-rooted plant diversity. Similarly, reintroducing gray wolves to Yellowstone curbed overpopulated herbivores, allowing riverside shrubs to recover and altering river flow. Conversely, the extinction of Tasmania's top predator, the thylacine (Tasmanian tiger), left low-diversity Tasmanian devils unchecked to spread lethal, transmissible Tasmanian devil facial tumor disease via biting (a topic Dr. Huberman previously discussed with HHMI biologist Jared Rutter), as no predator was present to cull sick animals early.
• Genetic Rescue of Endangered Species: The same synthetic biology toolkit used for de-extinction rescues living species from genetic bottlenecks and invasive threats:
• American Chestnut Tree: The USDA recently deregulated the first gene-edited organism for conservation—an American chestnut engineered to resist an imported Chinese fungal blight that killed a billion trees in a decade during the early 20th century.
• Northern Quoll: Australian partners of Colossal identified a single-letter DNA change (one amino acid in one gene) from toad-eating mammals on the other side of the world and engineered it into northern quoll cells in a dish, granting the endangered marsupial resistance to invasive, toxic cane toads.
• Black-Footed Ferret: Listed in the first endangered species class of 1966 after farmers poisoned prairie dogs, the ferret was thought extinct until a dog named Shep killed one outside Meeteetse, Wyoming about a decade later and a taxidermist recognized it. All ~500 ferrets released annually descend from a tiny founder group (including a prolific male named Scarface). To fix inbreeding, US Fish & Wildlife, the San Diego Frozen Zoo, and Revive & Restore cloned 40-year-old frozen tissue from an unrelated line, producing Elizabeth Ann in 2020 (who could not reproduce due to what Dr. Shapiro recalls as an ovary wiring/egg-release issue) and a second fertile clone, while also editing in plague immunity from domestic ferrets.
• Red Wolf, Panthers, and Island Foxes: Alongside the dire wolf, Colossal cloned endangered red wolves using >75% red wolf ancestry discovered in Gulf Coast coyote-like animals by Princeton's Bridgett vonHoldt. While inbreeding gave isolated Florida panthers crooked tails, cowlicks, and undescended testicles until Texas panthers were translocated in the mid-1990s, California's Channel Island foxes survived tiny isolated bottlenecks without predators for so long that they purged their deleterious recessive alleles and thrive today with nearly zero genetic diversity. Colossal is also working on the Mauritian pink pigeon, the bluebuck, and—led by the Indigenous Ngāi Tahu Research Centre on New Zealand's South Island—the extinct moa.
• Human Gene Editing Ethics and Bespoke CRISPR Cures: While human germline editing remains controversial—exemplified by a researcher in China (formerly a Stanford postdoc) who was punished and had his lab shut down after using CRISPR on human embryos to disrupt the CCR5 HIV receptor, with parallel claims of hippocampal/brain enhancement (a story broken by Antonio Regalado), alongside at least four IVF companies now deep-sequencing embryos to predict traits like height and IQ—somatic gene editing is already saving lives. "Baby KJ," born at Children's Hospital of Philadelphia (CHOP) with a lethal urea cycle deficiency causing blood ammonia buildup, became the first child cured of a genetic disease via synthetic biology after an academic, industry, and NIH team designed and safety-tested a bespoke liver-targeted CRISPR base editor in 6 months, curing him with three doses at 6 months old.
• Episode Sponsors Disclosed by Dr. Huberman: Joovv medical-grade red light therapy devices (up to $400 off select products at joovv.com/huberman); Eight Sleep smart mattress covers Pod 6 (up to $200 off at eightsleep.com/huberman); AG1 vitamin, mineral, probiotic, and adaptogen supplement, including the newly launched AG1 Next Gen/Pro with 5g creatine monohydrate, calcium HMB, and zinc carnosine (free bottle of omega-3 coenzyme Q10 with first subscription at drinkag1.com/huberman); and Function lab testing and biomarker analysis platform offering over 160 tests for $365/year or $1/day ($50 credit with code HUBERMAN at functionhealth.com/huberman).

Quotes

• At 0:00:00 - "So our dire wolves, they have 20 edits that we picked. And we sequenced genomes from fossil dire wolves, we learned from those genomes what genetic changes made those animals bigger, more robust, light-colored in coat, and then we engineered those changes into a gray wolf genome to recreate the dire wolf." - Dr. Shapiro explaining how Colossal Biosciences engineered living dire wolves from fossil DNA data.
• At 0:00:21 - "You know, I often get the question of, 'Why are you thinking about bringing extinct species back to life? Why aren't you thinking about helping living species not become extinct?' And the answer is, we are doing both. It is the same tools. It's the same technology, it's the same needs." - Dr. Shapiro clarifying that de-extinction and modern endangered species conservation rely on the exact same biotechnology toolkit.
• At 0:03:51 - "Biology doesn't care what species you are. Species is a human concept. We have this incredible proclivity to want to put things into boxes so that we can talk about them." - Dr. Shapiro deconstructing the idea that taxonomic boundaries are rigid biological laws.
• At 0:09:02 - "Biology doesn't care about your species concept. Your species concept is something that you have adopted to have a conversation with another person about a particular topic." - Dr. Shapiro explaining why biologists use different definitions of species (biological, genetic, geographic) depending on the conservation or research goal.
• At 0:10:31 - "And today, most people have somewhere between 2 and 5% of their DNA that is from this admixture event, this hybridization event between anatomically modern humans dispersing out of Africa and the Neanderthals that were already in Europe." - Dr. Shapiro detailing the genetic legacy of human-Neanderthal interbreeding.
• At 0:18:57 - "What's most interesting, and I think is relevant to your question, is that those hybrid bears only ever live and survive as brown bears. We see no evidence of brown bear DNA getting into polar bears." - Dr. Shapiro illustrating how ecological selection pressures prevent hybrid brown-polar bears from surviving in Arctic hunting environments.
• At 0:23:33 - "If we were to go around the world today and pick out all of the pieces of Neanderthal DNA that exist in humans today, we would put together more than 90%, possibly more than 95% of the Neanderthal genome just from people who are alive today." - Dr. Shapiro revealing how nearly the entire Neanderthal genome survives in fragments scattered across the modern human population.
• At 0:25:33 - "We've narrowed that list down even further to that other 5%. What's going on in that 5% of the genome where no living person has Neanderthal DNA? That is where the stuff that the baby had to have the human version in order for that baby to survive. And that's where we look to see what it is that made us human." - Dr. Shapiro explaining how evolutionary "deserts" devoid of Neanderthal DNA pinpoint uniquely Homo sapiens traits.
• At 0:27:30 - "During the pandemic, one of the first alleles that was discovered to be associated with negative risk or bad outcomes of COVID was a gene that came from Neanderthals. So it's a Neanderthal-derived gene that was at something like 50% frequency—so way above the 3% average frequency—in Asian populations." - Dr. Shapiro connecting ancient hominin hybridization to modern infectious disease susceptibility.
• At 0:32:40 - "We can't bring a dinosaur back to life because we don't have dinosaur DNA. The oldest DNA that we have ever recovered from a bone is from a mammoth bone. A mammoth bone probably dates to around 1 to 2 million years ago." - Dr. Shapiro establishing the hard chronological boundary of DNA survival vs. rock fossilization.
• At 0:37:48 - "With the animals, you get a mosaic landscape where there's some parts that are moist and wet, there are other parts that have been exposed and they're dry and colder, and you get broader diversity of plants that are coming back where these animals are. So they are essentially recreating their ecosystem just by being there." - Dr. Shapiro describing how megafauna at Pleistocene Park insulate permafrost and restore tundra biodiversity.
• At 0:43:06 - "My scientific explanation is that meiosis didn't fully separate and she ended up having a fertilized egg..." - Dr. Shapiro explaining how a critically endangered California condor achieved a "virgin birth" (parthenogenesis) without a male.
• At 0:49:38 - "And we know that the closest living relative of a mammoth is an Asian elephant. And in fact, mammoths and Asian elephants are more closely related to each other than Asian elephants are to African elephants." - Dr. Shapiro clarifying the evolutionary tree of elephants and why Asian elephants serve as the genomic template for mammoths.
• At 0:54:07 - "Instead, what we're doing is focusing on where all of those mammoths are the same as each other, but different from elephants, and thinking that those places—like we were talking about with humans—those are the places that are important to make a mammoth a mammoth." - Dr. Shapiro outlining the comparative genomics strategy for identifying essential mammoth traits across variable individuals.
• At 0:59:24 - "And so instead, we brought back the dire wolf light-colored coat using different edits that we know are safe in a gray wolf background because there are light-colored domestic dogs and gray wolves. So we used the edits that we know are safe because they exist in living dogs." - Dr. Shapiro demonstrating how animal welfare takes precedence over strict genetic literalism when engineering de-extinct traits.
• At 1:04:41 - "The very first gene-edited organism created specifically for the purposes of conservation, ecosystem restoration, was deregulated by the USDA this week. It's the American chestnut tree." - Dr. Shapiro highlighting a milestone regulatory approval for synthetic biology in forest restoration.
• At 1:14:47 - "There are mammals that eat toxic toads that live on the other side of the planet, even that eat cane toads, right? And they have a single letter change that changes one amino acid in one gene that allows them to eat that toxic cane toad and not die. Our Australia, Colossal Australia partners have made those changes in a quoll." - Dr. Shapiro showing how a single base-pair edit can save an endangered native predator from an invasive toxic species.
• At 1:25:25 - "If we say that those technologies are too risky, we are accepting the outcome of doing nothing, which is also a decision. And I think that's where a lot of times the conversation gets lost. Doing nothing is not 'not deciding.' Doing nothing is saying we accept the fact that we are going to have a future that is less biodiverse than the present." - Dr. Shapiro framing the central ethical imperative for using genetic engineering in conservation.
• At 1:36:55 - "And then give him a CRISPR medicine, a bespoke CRISPR medicine as a six-month-old kid, three times, and he is cured of this disease and will live a normal life." - Dr. Shapiro recounting the historic cure of "Baby KJ" at CHOP using a customized CRISPR base editor.
• At 2:00:34 - "When we announced the dire wolves, we announced that we had cloned red wolves, which is the most endangered species of wolf in North America, an endemic wolf that again is living in a very small population in captive breeding in the Carolinas." - Dr. Shapiro demonstrating how Colossal applied its canine cloning pipeline to immediately bolster an endangered native wolf population.

Takeaways

• Reframe "doing nothing" as an active ecological choice: Recognize that rejecting genetic engineering in conservation is not a neutral stance; because human activity alters habitats faster than natural evolution can adapt, inaction is an active decision to accept a less biodiverse future.
• Prioritize animal welfare over genetic literalism in synthetic biology: When engineering traits into living genomes, substitute potentially harmful ancient variants (such as fossil dire wolf coat genes linked to albinism, blindness, and deafness) with safe, validated genetic variants from living relatives.
• Leverage public excitement around de-extinction to fund endangered species rescue: Use high-profile "moonshot" projects (like the mammoth, dodo, and dire wolf) to build and fund the exact reproductive and genomic tools needed to save critically endangered animals like the red wolf, northern quoll, and Mauritian pink pigeon.
• Preserve cryobanked tissue in "Frozen Zoos" before genetic bottlenecks occur: Maintain biobanks of living cells from threatened species so that decades-old genetic diversity can be cloned back into inbred populations, just as 40-year-old tissue restored the black-footed ferret gene pool.
• Search admixing wild populations for "ghost" ancestry of endangered species: Use genomic sequencing to screen hybridizing wild populations (such as Gulf Coast coyotes carrying >75% red wolf DNA) and clone those individuals to recover lost genetic diversity for captive breeding programs.
• Apply targeted single-gene edits to help native wildlife survive invasive species: Rather than relying solely on slow habitat cleanup, identify natural toxin- or disease-resistance alleles in other species and engineer them into threatened native wildlife—such as giving northern quolls single-amino-acid cane toad resistance or giving black-footed ferrets domestic ferret plague immunity.
• Use time-limited genetic switches when deploying gene drives: Instead of eradicating nuisance vectors entirely, engineer mosquitoes that cannot transmit malaria or dengue—building upon foundational vector biology like HHMI Director Leslie Vosshall's work on mosquito olfaction—or design temporary, generation-limited gene drives to suppress invasive, flammable weeds like Mediterranean cheatgrass so native deep-rooted plants can recover.
• Never plant unsolicited foreign seeds outdoors: Follow the biosecurity warning Dr. Huberman received from UC Santa Cruz plant biologists when sent unsolicited "free seeds for gardening" from China in 2016/2017—never plant unknown imported seeds, as foreign pathogens or weeds (like the blight that killed a billion American chestnut trees) can devastate native ecosystems.
• Partner with Indigenous communities to guide ecological restoration: Model conservation initiatives on collaborations like Colossal's moa project, which is directed by the Māori Ngāi Tahu Research Centre on New Zealand's South Island to align biotechnology with local cultural and ecological stewardship.
• Evaluate canine sterilization options that preserve developmental hormones: When managing male dogs that should not breed, consult a veterinarian about a vasectomy rather than traditional castration so the animal remains sterile while retaining endogenous hormones for brain and physical development.
• Develop artificial wombs (ectogenesis) to protect endangered surrogates: Invest in ex-utero gestation technologies to scale the recovery of slow-reproducing megafauna and endangered species without burdening scarce living females (like Asian elephants with 22-month pregnancies).
• Build rapid-response regulatory and clinical pipelines for bespoke CRISPR therapies: Use the 6-month collaboration between CHOP, industry, and the NIH that cured "Baby KJ" of a lethal urea cycle deficiency as a blueprint for delivering customized somatic base-editing cures to infants with rare genetic diseases.