8 | Carl Zimmer on Heredity, DNA, and Editing Genes
What this covers
Science writer Carl Zimmer discusses how our understanding of heredity has grown far more intricate than earlier textbook models suggested. Rather than opening with a restatement of the thesis, this conversation with host Sean Carroll explores the historical arc of heredity science—from how practical concerns of livestock breeders in the 1700s and 1800s drove study of inheritance, to how Gregor Mendel's pea experiments became codified as law while overlooking messier reality. Zimmer traces how even Charles Darwin engaged with emerging ideas about mental illness and heredity, and how concepts like "race" and "blue blood" were constructed in fifteenth-century Spain to encode social boundaries as biological facts. The episode then pivots to what modern genetics reveals and conceals: the multiple channels through which traits pass between generations—not just nuclear DNA, but mitochondria, microbiome, and epigenetic marks—and how nature versus nurture operate inseparably, as when human height rose inches in a century through better diet, not genetic change.
The conversation turns to what people actually encounter when they receive genetic results. Zimmer argues that most consequential traits are polygenic, shaped by many genes and environment alike, making single-gene prediction myths dangerous when misapplied by consumer genetics firms. He examines CRISPR—its real mechanism, its genuine medical promise for diseases like sickle cell anemia, and its actual constraints—against science-fiction fantasies of designer babies. He dismisses the scenario that gene editing will split humanity into separate species, since sexual mixing will either dilute any introduced variant across the whole population or let it vanish. On the regulation question, he reframes it as explicitly political and social, not scientific, and takes a skeptical view of inequality arguments against editing that would require banning many other parental advantages. Throughout, he emphasizes why better genetics education matters: people now receive real genetic data in their inboxes, and Mendel's peas no longer suffice.
Zimmer argues that heredity is far more complex than the simple DNA-blueprint model, involving multiple inheritance channels (nuclear DNA, mitochondria, microbes, epigenetics) and gene-environment interactions, so emerging gene-editing technologies must be evaluated against this messy reality rather than science-fiction fantasies of designer babies.
- Most meaningful traits are influenced by many genes and by environment, defeating one-gene-one-trait intuitions
- Inheritance flows through multiple channels beyond nuclear DNA (mitochondria, microbiome, possibly epigenetics)
- CRISPR's real limits (safety, logistics, polygenic traits) make many feared scenarios biologically implausible
Most traits humans care about depend on many genes and environment, making single-gene explanations misleading or false.
- For most traits we care about, including seemingly simple ones like height, there is no single gene; they are influenced by many, many different genes, making the idea of 'the tall gene' meaningless, unlike simple cases such as blood type which depends on one gene.
“for most of the traits that we actually really care about or think about, and even seemingly simple ones like height, they are influenced by many, many, many, many, many different genes... you can't say that, 'Oh, I have... Do I have the tall gene?' It's just meaningless.”
- Average human height has risen several inches over the past century not because of inherited genetic changes but because of improved environment such as better diets and childhood medicine, demonstrating that traits depend on the combined influences of genes and environment (nature and nurture).
“all over the world, the average height of people is several inches higher than it was a century ago. And that's not because we are now inheriting a different set of genes, it's just that in that respect, the world got to be a better place.”
- Genes that strongly impact health, such as those causing Huntington's disease, early-onset Alzheimer's, or dramatically raising cancer risk, are rare because natural selection is not fond of them, so only a small fraction of people who sequence their genomes will find such high-impact variants; most people find only modest risk modifiers offset by protective variants.
“the genes that really have a strong impact on your health... genes that cause Huntington's disease, or genes that raise your risk dramatically of getting early onset Alzheimer's... these are rare. Natural selection is not fond of these genes... Therefore, they're rare.”
- Even a complete list of all three billion base pairs given to a computer with perfect knowledge could not fully predict an organism, because development emerges through cells making proteins, talking to each other, taking environmental cues, and migrating; DNA permits good predictions for some traits like eye color and crude ones for height, but not the whole organism.
“there are a lot of things that you can predict based on DNA... I bet if I looked at your DNA, I could get a pretty good idea what your eye color is. And I might be able to make some very crude predictions about... your height... But then there's this, the DNA makes... the cells are talking to each other, and they're taking in cues from the environment... And that is how we end up the way we are.”
- Many companies make dubious claims from DNA, offering special exercise programs, DNA diets, or even wine recommendations (e.g. Vinome) by cherry-picking a few variants from the literature and over-extrapolating; for example, a sequencing company told Zimmer his muscles were built for power based on one variant while ignoring the many other genes influencing muscles, illustrating why such firms should not be trusted to design babies.
“we already have all sorts of companies out there that are offering really dubious claims based on looking at your DNA... There's even a company that's called Vinome... They will recommend wine to you, based on your DNA.”
Gene editing could treat disease but designer-baby scenarios are implausible and genetic prediction scores explain only modest outcome variation.
- CRISPR could treat hereditary diseases by, for example, removing a sickle-cell patient's bone marrow stem cells, tweaking their DNA so they produce the proper hemoglobin instead of the misshapen kind that causes sickle cell anemia, and returning the cells to make healthy blood; clinical trials on this could start very soon.
“if someone has sickle cell anemia, you take some of their bone marrow cells out... You tweak their DNA so that they can now make hemoglobin that they need... and you put the cells back into people, and they make healthy blood cells, that's the hope. And there are clinical trials that could start very soon on that.”
- The argument against CRISPR based on inequality is weak because it implies we currently live in a paradise of equality; genes are not the only thing influencing children's later success, and many other advantages of wealthy parents (like SAT prep classes) already shape outcomes, so singling out gene editing for a ban raises the question of whether all such advantages should be illegal.
“I have a problem with these arguments against CRISPR based on inequality, because they all make it sound like we are living in a paradise of equality today.”
- Studies link genes to how long people stay in school, and one can compute an education score from a million DNA variants, but using it to track a kindergartener would be ridiculous because the score predicts only a small amount of the variation in school outcomes; many people with high genetic scores drop out early and many with low scores reach grad school, so it is just one variable among several and is not a scenario worth treating as deterministic.
“this score it only predicts a small amount of the variation in how people do in school. So chances are that your score would be very wrong. So lots of people with a high genetic score who drop out early from school. There are a lot of people with a low genetic score who go into grad school. It's just one variable among several.”
- The science-fiction scenario that CRISPR will split humanity into two species (rich edited people versus poor) is implausible because people have lots of sex without respecting arbitrary boundaries, so any introduced genes will either disappear entirely from the gene pool, as most variants do, or diffuse around the whole world, making these supposed ethical dilemmas not worth treating as real.
“people just don't work that way, animals in general don't work that way. People have sex, lots of sex... Whatever genes might get introduced into some rich person will either disappear entirely from the human gene pool eventually... Or will just kinda diffuse around all over the world”
- CRISPR by itself cannot deliver a science-fiction future of designer babies because it would require everyone to undergo in vitro fertilization, currently a difficult drawn-out process with possibly inherent limits, plus many other advances in reproductive technology and stem cell research before widespread embryo editing could even be possible.
“CRISPR alone would not deliver you into that science fiction future, you'd have to have all sorts of other advances in reproductive technology and stem cell research, and all the rest of it before this could even be possible.”
Epigenetic molecules and mechanisms switch genes on and off, controlling how cells divide and remain heritable across cell generations.
- Genes are attended to by molecules that switch them on or off: proteins clamp onto genes to shut them down or activate them, and DNA coiled around spools cannot be used to make protein; these epigenetic controls are long-lasting and inherited by daughter cells when a cell divides, which is why dividing skin cells make skin cells rather than brain cells.
“there are proteins for example, that will clamp on to genes and they can essentially shut them down... You can coil DNA around spools and then basically anything that gets coiled up, any gene just can't be used to make a protein... when a cell divides, those same controls will in effect be inherited by the two new cells. So that's why your skin cells, when they divide, they make skin cells”
- Women have two X chromosomes but must keep one shut off to avoid poisoning themselves with too many X-chromosome proteins; specialized RNA molecules wrap around one X chromosome and silence it, demonstrating that some genes produce functional RNA rather than proteins.
“in women, women have two X chromosomes, they need to keep one of them shut off or they're gonna basically poison themselves with too many proteins from the X chromosome... there are these RNA molecules that basically wrap around the X chromosome, one of the X chromosomes in women, and silence it.”
- The idea that an experience can alter how our genes work and pass that alteration to future generations has good evidence in plants and tiny worms and tantalizing mouse experiments (such as odor-shock conditioning seemingly affecting pups via sperm), but the molecular mechanism for transgenerational epigenetic inheritance in mammals is unknown and many scientists doubt it makes sense.
“we could have an experience that alters how our genes work and that alteration can get passed down to future generations. That's the crux of epigenetics and heredity... there's definitely good evidence for it happening in plants... When you get to mice, there's a very tantalizing experiments.”
Brains impose categorical thinking on heredity despite it operating through stable balances and ongoing competition, not absolute laws.
- Humans have brains that like categories and tend to turn observations into absolute laws, such as treating Mendel's observations as Mendel's Law or treating male and female as exception-free categories, but heredity does not work like that; many apparent patterns are actually stable balances produced by ongoing competition, a kind of detente.
“We like to take biology and put it into categories and try to come up with absolutes... I think we just have brains that really like categories, but heredity just does not work like that.”
- Mendel was forgotten for about 50 years partly because, when his mentor urged him to replicate his pea findings in another plant, he chose hawkweed, which reproduces asexually by kicking out male DNA and cloning, so the clean 3-to-1 ratios he saw in peas did not appear, undermining his traction.
“he went and studied another common garden plant and it turned out that this other one, hawkweed... it doesn't reproduce in the neat sexual way that peas do... Boom, those lovely 3 to 1 patterns that Mendel saw with peas, they're just not there at all when he looks at another species”
CRISPR is a repurposed bacterial immune system enabling precise DNA cutting and editing but carries serious uncontrolled safety risks.
- CRISPR is a natural bacterial immune system in which bacteria make molecules that store information about viruses and then create molecules that zero in on particular stretches of virus DNA and cut it; scientists repurposed this to target and cut any chosen DNA stretch and substitute new DNA, creating a powerful molecular tool now used routinely to discover, for example, which genes a cancer cell needs to survive.
“it's basically an immune system for bacteria. They make molecules that can essentially store information about viruses... some scientists said 'Well, we could use that, and we could maybe cut whatever DNA we want.'... they could zero in on particular stretches of DNA, and make cuts, and then substitute in new DNA.”
- CRISPR is revolutionary but problematic because it works by chopping DNA, a radical act cells defend against since DNA breaks can make cell descendants cancerous; moreover, a recent study showed that cutting out a target segment can remove long stretches of needed DNA, and repair may shuffle DNA in harmful ways, raising serious safety concerns about using CRISPR-edited cells in the body.
“it's editing that involves chopping DNA, and that is a pretty radical thing to do to DNA, and cells don't like it... it can lead to all sorts of damage that can ultimately cause a cell's descendants to become cancerous... there's a recent study that showed that sometimes when scientists try to cut one particular segment of DNA out, they cut out a long stretch that includes that particular target.”
Human gene editing via mitochondrial replacement has begun with divergent national legal and regulatory responses already in place.
- If people are going to attempt human gene editing, the question of whether to pass laws preventing it or to put regulations allowing certain uses shifts the issue from a scientific question to a social and political one, as already illustrated by divergent national responses to mitochondrial replacement therapy.
“Should we pass laws to prevent that? Or, do we have... do we put regulations in place to allow certain uses of it for certain things? And then really you're shifting from a scientific question to a social or a political one.”
- Human genetic engineering has already begun via mitochondrial replacement, in which an egg's nuclear DNA is transferred into a donor egg with healthy mitochondria to prevent devastating mitochondrial disease, producing so-called three-parent babies; the US has banned it, Mexico has no laws so a doctor performed it there semi-secretly, while Britain debated it openly in Parliament and licensed a regulated clinic.
“genetically engineering humans has already begun because we were talking about mitochondria before... People call that three-parent babies... in the United States, that has been banned... a doctor in New York... actually went to Mexico... But meanwhile in Britain, they talked about this very quite explicitly and had discussions in Parliament... they're going to allow this to go forward under a lot of regulation.”
Mitochondria inherited maternally have their own DNA and descended from ancient symbiotic bacteria integrated into eukaryotic cells.
- One theory for why mitochondria are inherited only from the mother is that two batches of mitochondria from different parents would not play well together, operating differently and potentially causing problems, so excluding the father's mitochondria avoids this conflict.
“One theory is that if you have two batches of mitochondria inside a person coming from different people, they're going to not play well together, that they're gonna operate differently and that could actually cause problems.”
- Mitochondria are the powerhouses of the cell that generate fuel using oxygen and nutrients; they have their own DNA separate from the nucleus, divide and copy their DNA on their own, and are descended from bacteria that entered our single-celled ancestors about 1.8 billion years ago in a symbiotic relationship and became permanent residents unable to live outside our cells.
“About 1.8 billion years ago, when we were single-celled, the ancestors of mitochondria somehow ended up inside of our cells in maybe some symbiotic relationship... and then they became basically permanent residents, so that they couldn't live outside of our cells anymore.”
- Although nuclear chromosomes are a 50-50 split between parents, mitochondria are inherited entirely from the mother because the sperm uses and then destroys its own mitochondria and does not deliver them to the egg; because mitochondrial DNA is not shuffled by meiosis, it remains nearly identical across maternal generations, making it powerful for tracing genealogy.
“when you look at our chromosomes, we are a 50-50 split between our parents. But you look at our mitochondria, it's all mom, just all mom.”
Microbes constitute a non-random heritable channel passed through specific hosts with biological filters selecting which species persist.
- As a baby moves through the birth canal it gets slathered in bacteria, some of which the mother encourages to grow there, and breast milk contains both bacteria and indigestible sugars that feed bacteria; there is debate over whether specific species are seeded into babies early, and Caesarean sections may deprive infants of this seeding, which matters because an improper early microbiome can cause lifelong problems such as impaired immune function.
“as a baby moves through the birth canal during delivery, it gets slathered in bacteria... the mother's milk contains bacteria as well, as well as bacteria food... if your microbiome isn't quite right when you're young, that can lead to problems throughout your life, your immune system may not work properly”
- The human microbiome is not a random collection of critters; the same species and strains recur across people, human microbiomes resemble each other far more than they resemble a chimpanzee's, suggesting the body applies filters that determine which microbes get a place.
“if you look at human microbiomes compared to a chimpanzee's, they're gonna be a lot more similar to each other than chimpanzees are, and it seems like we have filters... not everybody gets a seat in the bus.”
- Some species pass down certain microbes as faithfully as their own genes, constituting another channel of heredity; cockroaches depend on a bacterium that makes protein compounds they cannot, build special organs to house it, and physically transport these microbe-carrying cells into their eggs so offspring are born with the bacteria ready, much like humans are born with mitochondria.
“Cockroaches actually depend on one species of bacteria to help them to eat food... when it comes time for the female to produce her eggs... these cells that carry these microbes, they just start crawling, and they make their way through the cockroach's body to the cockroach's eggs... so, after these eggs get fertilized... then the cockroach is born with these bacteria ready to go, just like we are with our mitochondria.”
Consumer DNA testing reveals much less than marketed; people misinterpret results by panicking over associations or dismissing findings.
- Consumer companies like 23andMe and Ancestry perform genotyping, surveying about a million marker spots in the genome and inferring intervening DNA because people share stretches of identical variants, whereas whole-genome sequencing tries to determine every single letter, revealing deeper information.
“What 23andMe or Ancestry does is they do something called genotyping. Basically, they look at maybe a million markers... When you get your whole genome sequenced, that means that you're trying to figure out as best as you can, every single letter in your genome”
- The widely repeated claim that microbes outnumber human cells 10 to 1 is probably false; the ratio is more like 1 to 1, with about 37 trillion human cells and a similar order of bacterial cells, though the full microbiome including viruses and fungi may be higher.
“we always say your microbes outnumber your own cells by 10:1 which is always fun to say... It turns out probably not to be true. It's probably more like 1:1.”
- People interpreting genetic test results make two mistakes: being disappointed at finding nothing dramatic (when good news is good), and panicking over a disease-associated gene without digging deeper into what a study actually found, how big a risk it conferred, and how large the study was, since tiny preliminary studies are often wrong and many supposed disease mutations have been overturned.
“you have to dig down that extra level and say, 'What exactly did this study find?'... how big was this study? A lot of these studies when they're preliminary, is just like a hundred people, they're tiny. And tiny studies are often wrong. So there are plenty of mutations that were thought originally to cause diseases that we now know do not.”
Mendel discovered genes are discrete invisible factors combining mathematically rather than blending parental traits.
- Mendel demonstrated that heredity involves discrete invisible factors (later called genes) that combine mathematically, not a blending of parents, as shown by the wrinkled-versus-smooth pea trait where a recessive factor can hide and reappear in about a quarter of offspring.
“if you cross a smooth pea and a smooth pea together, you might be surprised to suddenly have a quarter of the peas being wrinkled... that wrinkled factor can hide because it's what we would now call... recessive.”
- Some pieces of DNA break Mendel's law of equal segregation by overriding the 50/50 chance of a chromosome ending up in an egg or sperm, as shown by fly strains that almost always produce daughters because a gene on the female chromosome ensures the flies have no sons, exemplifying the ultimate selfish gene.
“there are genes, or pieces of DNA that basically just totally break Mendel's law completely, and just override that 50/50 split... there was a gene that was basically hijack, it was sitting on the female chromosome in flies, and was basically, ensuring that these flies did not have any sons”
DNA functions as a four-letter text whose sequence changes propagate when copied and passed to descendants.
- Genes function like texts made up of units from a four-letter alphabet, with over three billion letters in human DNA; changing the letters or cutting and pasting chunks of text produces changes in the organism, and these changes are passed down when DNA is faithfully copied.
“it's almost like genes are like texts. They're made up of these units, where they're like letters made up from a four-letter alphabet and we have over three billion letters in our DNA, and change the letters, or cut and paste chunks of text, and you get changes to us.”
- Although DNA was known since the 1800s and many mid-1900s scientists believed genes were made of proteins, elegant experiments showed that transferring DNA between microbes transfers traits while proteins do not matter, establishing DNA as the molecular basis of heredity once its structure was determined.
“it took some elegant experiments to demonstrate, no, actually, if you transfer DNA from one microbe to another, you transfer that trait, the proteins don't matter.”
Humans emotionally and conceptually cling to heredity as central to identity and kinship across generations.
- People cling to heredity profoundly because we use it to define who we are and our connection to both past and future, making it one of the most intimate concepts humans grapple with.
“we use heredity to define who we are, and also what is our connection both to the past and to the future. You can't ask for anything more intimate than that.”
Modern genetics education must move beyond Mendel because people now receive real genetic results requiring deeper understanding.
- Grade-school and high-school genetics education must improve and move beyond Mendel because genetic findings are no longer abstractions but real results people receive in their email inboxes from consumer DNA companies, which they cannot understand using only pea-plant experiments.
“I feel that our high school, grade school genetics has just gotta step up its game, because these things are not just abstractions that you learn about in high school and never think about again. People are getting these results in their email inbox.”
Genomic conflict occurs when parental gene copies pursue different evolutionary strategies within developing offspring.
- Genes from the two parents can be in evolutionary conflict inside the child: a father's gene copy may drive faster growth for his long-term evolutionary benefit while the mother's copy is turned off because excessive fetal growth drains her resources and reduces her lifetime reproductive output, producing a genomic tug of war that finds an optimal balance.
“dad's genes are maybe driving kids to grow faster, because that's good for the father's long-term evolutionary benefit. The mother meanwhile... too much growth is actually... Can really drain her resources... you will actually find that the man's copy of a gene is turned on, a woman's copy is turned off inside the child.”
The human genome contains only twenty thousand protein-coding genes plus regulatory RNA genes and switches amid extensive junk DNA.
- Humans have about 20,000 protein-coding genes that make up only 1-2% of the genome; roughly 10% more has its own functions including thousands of genes that encode functional RNA molecules and stretches that act as genetic switches, while much of the rest is junk including dead mutated genes and viral DNA.
“we have, as you say, 20,000 of these protein-coding genes. And they only make up about 1% or so, 1 or 2% of the human genome.”
Early psychiatry and breeding practice drove scientific study of heredity more than abstract curiosity.
- Early 1800s psychiatrists in France and elsewhere, trying to understand madness, noticed that institutionalized patients often had relatives with similar conditions, leading them to ask whether mental illness was hereditary; Darwin read extensively in psychiatry while developing his ideas about heredity.
“In the early 1800s, particularly in France... psychiatrists were trying to understand madness. Then they were struck by the fact that, when they would do questionnaires for their patients, their patients often had people in their family or more distant relatives, who also were institutionalized”
- The scientific study of heredity was driven by practical concerns rather than abstract curiosity: livestock and plant breeders like Bakewell, who created a new sheep breed in 1700s England, struggled with the variation in offspring, and by the 1700s-1800s nations viewed successful breeding as a source of national wealth.
“In the 1700s, someone named Bakewell in England became legendary because he created a new breed of sheep, and he did it by carefully breeding different kinds of sheep together and coming up with these rules of his own for how to breed them.”
Racial and hereditary concepts originated in 1400s-1500s Spain to distinguish groups and relied on pseudoscientific visible-vein theories.
- The concept of 'blue blood' and the modern use of the word 'race' originated in 1400s-1500s Spain, where people sought to distinguish themselves from Jews and Muslims; the idea was that a racially pure person's veins could be seen through translucent skin.
“It was in Spain in the 1500s when people in Spain were trying to distinguish themselves from Jews and Muslims. So this is where... And this is where the whole issue of race comes from. Literally, the word race starts to be used in Spain in this way.”
Cells are messier than imagined, transcribing much DNA into discarded RNA and managing chaos with recycling proteins.
- Cells are far sloppier than commonly imagined: much DNA is transcribed into RNA that the cell immediately shreds because it serves no function, and the cell manages this chaos with proteins that check whether RNA belongs and recycle it; this messiness makes intelligent design laughable.
“the cell is a lot sloppier than we think of it as being. A lot of DNA actually, is used by our cells to produce RNA molecules, and then the cell just immediately shreds all that RNA, because those were just accidental.”
Blogs shifted from future communication medium to obsolete platform.
- Blogs were once considered the future of communication, and the speakers note the irony that they have now become a thing of the past.
“back when blogs were the future.”