101 | David Baltimore on the Mysteries of Viruses
What this covers
David Baltimore, Nobel laureate and architect of the classification scheme bearing his name, sits down with Sean Carroll to chart the territory of viruses—what they are, how they exploit cells, and why we should have seen the pandemic coming. Baltimore moves from the molecular mechanics of viral infection through the history of his own discoveries in reverse transcription, then pivots to argue that society has failed in its obligation to build defenses against entire classes of viruses rather than reacting after each outbreak emerges.
The conversation ranges widely across virology's landscape. Baltimore establishes that roughly half of human DNA originated from viruses, that megaviruses discovered in recent decades blur the boundary between living and non-living, and that the Speed-Accuracy Tradeoff in Replication explains why viral mutation rates dwarf those of cells—a difference of five to six orders of magnitude. He traces how Host-Virus Co-evolutionary Equilibrium works in endemic infections, and why viruses jumping species without that evolutionary history become voracious: HIV in monkeys poses no serious threat, but its arrival in humans proved catastrophic. He explains how vaccines and antiviral drugs operate on fundamentally different principles, and why designing either involves hazards—vaccines can sometimes worsen infection, and drugs face a Market Failure in Public Goods problem: the profit motive cannot predict which virus will matter, so companies won't develop broad-spectrum inhibitors against coronavirus classes without public funding and incentives. The discussion also touches on gene editing and global governance, cancer's viral origins, and why consciousness remains biology's most formidable unsolved puzzle—a jump in complexity that dwarfs even the leap from DNA to protein.
Baltimore argues that viruses are a vast, inventive, and genuinely living part of the biosphere whose diversity defies simple classification, and that the current pandemic was foreseeable—so society should proactively build a public health and pharmaceutical armamentarium against whole classes of viruses rather than reacting after each outbreak.
- Viruses are highly diverse, likely with multiple independent evolutionary origins, and routinely insert genetic material into host genomes (e.g., ~50% of human DNA originated as viruses).
- Non-equilibrium viruses that jump species cause the most damage because they haven't co-evolved with the new host.
- Drugs against entire classes like coronaviruses are technically feasible but neglected because the profit motive can't predict which virus will matter.
Society must proactively develop broad antiviral drug classes and catalog reservoir viruses before emergence.
- Society should proactively build an armamentarium of drugs that inhibit whole classes of viruses—since all coronaviruses are related and do the same basic things, the pharmaceutical industry could, given time and money, make inhibitory molecules effective against coronaviruses generally without knowing which specific one will emerge; we failed to do this because we never credited nature's ability to produce a human pandemic, and we should dedicate ourselves now to never being caught unprepared again.
“We ought to have in our armamentarium, drugs that will inhibit the growth of coronaviruses. All coronaviruses are related to each other... they wouldn't have to know which coronavirus it was in order to have inhibitory molecules. But we never spent any time trying to do that.”
- Epidemiologists feel this pandemic is not an outlier but should have been expected: SARS and MERS warned that coronaviruses could appear in new, dangerous forms; bats are an especially effective reservoir for viruses that can jump to humans, and the whole natural world harbors threats, so we should be systematically cataloguing viruses—especially in bats—and assigning companies the responsibility to make sure we can inhibit them.
“I think most people feel, epidemiologists feel that this is not an outlier, that we should've expected this. We were given warning with SARS and MERS that coronaviruses had the ability to appear in new forms”
- The recurring history of public health is that once a problem stops being imminent, focus and resources shift elsewhere; the lesson of this pandemic should be that we cannot do that—we must keep our eye on the entire world of potential viruses, and the obstacle is willpower, money, and political will rather than a fundamental scientific barrier.
“the history of public health is that when something stops being an eminent problem, that we lose our focus on it and we start using our resources for other things, and this ought to be the lesson of lessons, that we can't do that, we have to keep our eye on the ball”
- As long as antiviral development depends on the profit motive, society will not pre-emptively make drugs against future pandemics because the profit motive cannot know which virus is worth making a drug for; therefore non-profit research labs must be funded and incentivized through mechanisms like prizes to fill this gap.
“We have to find a way to fund them, to incentivize them, to make it... We gotta put up prizes, whatever it takes.”
- Baltimore does not expect a chemical inhibitor (drug) against COVID-19 in less than a couple of years; the main hope is to luckily find an existing safe drug that happens to inhibit the coronavirus, but there's no actual reason such a drug should exist because viral proteins look nothing like our own proteins—which makes them great immune targets but means we likely haven't already made drugs against them.
“I don't see us making a chemical inhibitor, a drug against COVID-19 in any less than a couple of years.”
Viruses are obligate intracellular parasites with minimal shared biology beyond cellular dependence.
- Viruses have very little in common with each other beyond not being cells: they cannot make their own proteins or their own energy, so they must be inside a cell to reproduce; this minimal requirement is met in enormously diverse ways, producing many distinct kinds of viruses.
“Viruses really only have in common that they're not cells. They can't make their own proteins, they can't make their own energy, so they have to be inside a cell in order to reproduce themselves.”
- Bacteria are cells that are completely self-sufficient—given nutrients they can absorb energy and multiply on their own—whereas a virus cannot do that and must get inside a cell, which is the fundamental difference between the two.
“bacteria are cells that are completely self-sufficient. So if you give them nutrients... they can absorb that energy and grow, multiply... and a virus can't do that. A virus has to get inside a cell.”
- Viruses' diverse use of genetic material (DNA or RNA, single or double strand) should be seen not as haphazard but as inventive—viruses have managed to use genetic information in ways that higher cells do not.
“I wouldn't call it haphazard. I would call it inventive. Viruses have managed to use genetic information in ways that higher cells don't.”
- There exists a class of megaviruses—some literally larger than bacteria, with hundreds or thousands of genes—discovered mostly in oceans in recent decades, and some of these megaviruses are themselves parasitized by smaller viruses; they retain some genes for protein synthesis and energy production, blurring the line between viruses and other organisms.
“there's a class of megaviruses which were discovered only in the last couple of decades, in oceans mostly, and some of these megaviruses actually have parasitic viruses that grow on them.”
Vaccines and antivirals work through fundamentally different mechanisms targeting immune priming versus viral inhibition.
- Vaccines and antiviral drugs are fundamentally different beasts: drugs are chemicals that physically interfere with viral growth, while vaccines present something resembling a virus that doesn't cause disease, prompting the immune system to make antibodies so that a real infection is met by an already-primed system that reacts fast enough that we usually never notice the infection.
“what we do is to make something that looks to the immune system like a virus but doesn't cause disease, and then we give that to people, and people react to it by making antibodies against the virus... And then if we get a real virus infection, the system is all set up and ready to go, and it reacts much faster”
- Baltimore is reasonably confident a vaccine against the novel coronavirus will be found within the next couple of years because it appears to be a pretty ordinary virus, and we have historically been able to make vaccines against most ordinary viruses such as measles, mumps, chickenpox, smallpox, and polio.
“it looks to me like this is a pretty ordinary virus, and against most ordinary viruses we have been able to make vaccines... I'm reasonably confident that they will find a way to make a vaccine within the next couple of years.”
- The immune system evolved as a general capability to recognize any foreign protein sequence and make antibodies that bind to it—whether that protein is from a virus or something else—so vaccination works by spurring this general system, not by teaching it a specific answer it couldn't reach on its own given enough time.
“the immune system reacts to any foreign protein by making antibodies that will bind to it... The immune system is evolved as a very general way of recognizing foreign protein sequence, and if it sees foreign protein sequence, it reacts to it. So it's a very general capability.”
- Designing a vaccine is uncertain because many things can go wrong: inactivating a virus can also destroy the protein's ability to stimulate an immune response, and more insidiously, a vaccine can sometimes make infection worse by actually helping the virus get into cells, so proving a vaccine is safe requires ruling out this enhancement effect.
“more insidiously, sometimes it makes the infection worse rather than better. It actually helps the virus get into cells.”
- Vaccines were made before antibodies were understood: Jenner used cowpox injected into humans so the immune system would be prepared to fight smallpox, exposing the body to something much less dangerous than smallpox, and this was very effective.
“we were making vaccines before we knew about antibodies. So Jenner made the first vaccines... that was cowpox injected into humans so that a human would think they'd seen smallpox, but in fact they'd seen something much less dangerous”
Cross-species viral jumps cause pandemic danger because viruses lack co-evolutionary equilibrium with new hosts.
- Viruses fully dedicated to one species co-evolve into equilibrium with their host—the virus becomes 'not too greedy' so it doesn't kill the host, and the host evolves to limit viral replication; but a virus that jumps from another species into humans hasn't co-evolved with them and so can become voracious and dangerous, which is why cross-species viruses cause the most trouble.
“if a virus is part of the ecosystem of an animal, take a human, the virus evolves to be not too greedy so that it doesn't kill off its host, and the animal or human evolves to fight against the virus to minimize the amount of virus that can be made. And they come into an equilibrium”
- HIV is native to monkeys, in which it does not cause serious disease, but when it jumped to higher apes like chimpanzees and to humans it caused havoc because it had not developed a modus vivendi with these new hosts.
“HIV is a virus that is native to monkeys. It doesn't cause serious disease in monkeys, but when it jumps to higher apes like chimpanzees or jumps to humans, then it causes havoc because it hasn't developed that modus vivendi.”
Viruses have inventively diverse genomic strategies and mutation rates orders of magnitude higher than cellular DNA.
- Most RNA viruses enter the cytoplasm or nucleus and make more of themselves, including messenger RNA that hops onto existing ribosomes; a common strategy is for the virus to interfere with the cell's own messenger RNA so it can't get on ribosomes, freeing them up for full viral use.
“often what the virus does is to interfere with the cell's messenger RNA so that the cell's messenger RNA can't get on ribosomes. That frees up the ribosomes so that the virus can take full advantage of it. This was something that I worked out actually in my thesis 60 years ago.”
- RNA-genome viruses overcome the greater chemical fragility of RNA simply through sheer numbers: an infected person makes literally billions of new viruses, so the fact that some copies are imperfect or don't survive is not a significant problem.
“they're making literally billions of new viruses, so the fact that the... If they have an RNA genome, that it is more chemically fragile than it is, is overcome by numbers”
- Viral and cellular DNA are chemically identical, but cells devote enormous attention to the precision of genome duplication, making errors very rare, whereas viruses treasure speed over accuracy and are willing to accept—even want—genetic change so they can adapt; the difference in mutation rate is on the order of five to six orders of magnitude.
“Viruses, first of all, treasure speed rather than accuracy... They're willing to accept some genetic change. In fact, they may want some genetic change because they want to mutate and be able to adapt to new circumstances through mutation.”
Retroviruses use reverse transcription to permanently insert RNA-derived DNA into host genomes.
- Information cannot flow from protein back to RNA because the genetic code is read in one direction—an exquisitely complex machine would be needed to reverse the protein-synthetic process—and Crick correctly anticipated this irreversibility, even predicting the existence of an adapter (later confirmed) to translate code into protein structure.
“there's an enormous structural barrier, because it's literally a code in RNA that gets transferred into the structure of protein, so to go back into the code would take a very complicated machine. You can't just go backwards in the protein synthetic machinery.”
- The central dogma as stated by Francis Crick was that information flows DNA to RNA to protein and—critically—cannot flow from protein back to RNA; Crick never claimed information couldn't flow backward from RNA to DNA, so Temin and Baltimore's 1970 discovery of reverse transcription did not actually violate the central dogma even though many claimed it did.
“What I said very strongly was, "It couldn't flow from protein back to RNA." So what Howard Temin and I showed in 1970 was that information could flow backwards from RNA to DNA. And many people said we violated the central dogma... But if you believe Francis Crick, and I believe Francis Crick, he had already taken that into account.”
- A class of RNA viruses (retroviruses) carries an enzyme in the virus particle that copies its RNA into DNA as the first thing it does on entering a cell; that DNA then enters the nucleus and inserts itself end-to-end at a roughly random location into the cell's DNA, permanently and insidiously adding new genetic information to the cell for the rest of its life.
“there's a class of RNA viruses that carry with them in the virus particle an enzyme that can copy RNA into DNA... it copies the RNA into DNA as the first thing it does when it gets inside a cell. And then that DNA goes into the nucleus... and it literally inserts itself end to end into the cell's DNA”
- Drugs to fight AIDS were available within about five years of HIV's discovery because pharmaceutical companies already had on the shelf compounds that selectively attacked reverse transcriptase, a viral enzyme target; this jump-started a successful chemical campaign of synthesizing inhibitors of reverse transcriptase and other viral proteins to control HIV and AIDS.
“we had on the shelf drugs, pharmaceutical companies had on the shelf, drugs that could selectively attack reverse transcriptase. And that was the reason that we had drugs to fight AIDS within about, what, five years of the discovery because they were on the shelf”
Viral DNA integrated into host genomes drives cancer through insertion of mutated cellular genes.
- Cancer is induced by genes: because viruses are so small, the only way they could cause cancer is via genes, and the Varmus-Bishop discovery showed that viral cancer-inducing genes were actually mutated cellular genes the virus had picked up—so cancer arises from genes that may be carried by viruses or be endogenous mutations, and the same kinds of genes are involved across species.
“people who were sensitive enough to numbers realized that viruses were so small that the only thing they could be causing cancer with is genes.”
- The bulk of human cancer is not caused by viruses—common cancers like lung, breast, and prostate show no evidence of viral involvement—though some human cancers (head and neck, cervical) are virus-induced, and in animals there are many more cases of virus-induced cancer, which is where most viral cancer is studied.
“we think that the bulk of human cancer is not caused by viruses”
- Cancer-inducing viruses work by inserting their own DNA into the cell's DNA, so that every time the cell divides it carries the viral sequences to the next generation; these sequences cause synthesis of specific proteins that take over the cell, and in an animal this produces a tumor of cells all carrying the viral sequences.
“There are some times when viruses actually insert their own DNA into the DNA of the cell, and actually, that's how cancer-inducing viruses work because they become part of the cell's DNA, and now every time the cell divides, it carries the virus along to the next generation.”
Gene editing poses dual-use risks requiring international governance, though therapeutic disease correction is unambiguously beneficial.
- The danger of gene editing is that the same power used to fix disease genes could be used to alter cosmetic or enhancement traits (intelligence, height), and controlling this is a huge challenge because there is no international law or treaty ensuring common global behavior—so what cannot be done in one country may be done elsewhere—which is why Baltimore organized international meetings to build awareness and concern even though they produced no definitive answers.
“if we do that, do we open a Pandora's box and provide the opportunity to not change genes that are bad, but to change genes that are cosmetically less than exactly what we would want?”
- Roughly 50% of human DNA originated as viruses; we carry around many pieces of DNA whose origins lie entirely outside ourselves, called parasitic DNA because it exploits DNA's properties to propagate, and some of it entered our genome in ancestral organisms millions of years ago because it takes a very long time to remove anything once it is in the genome.
“something like 50% of our DNA originated as viruses. We're carrying around in ourselves all sorts of interesting little pieces of DNA that have origins totally outside of ourselves.”
- Among bacteria, DNA from one bacterium can transfer into another (horizontal gene transfer), but this does not happen in humans or other multicellular organisms—we never acquire pieces of rabbit or mouse DNA; the only mechanism that inserts foreign DNA into higher organisms is viruses, and large-scale genome sequencing would reveal any such transfers if they occurred.
“what we generally mean when we say horizontal gene transfer, is that among bacteria, DNA from one bacterium can go to another bacterium and be inserted. And that's something that really doesn't happen as far as we know in humans or in higher organisms”
- Using gene editing to modify human heredity is 'all good and no bad' in the specific case of inherited genetic diseases—getting into affected families' genes to stop them passing Huntington's disease, sickle cell anemia, and thousands of other heritable diseases would be a good thing for the human race.
“the place where it seems to me it's all good and no bad is with certain kinds of diseases that are in our genome and that are inherited in families, and that if we could get into those families, into their genes and modify them so that they don't pass Huntington's disease, they don't pass sickle cell anaemia... that that would be a good thing for the human race.”
Viruses are living matter that reproduce and evolve despite requiring host cells, with uncertain evolutionary origins.
- RNA likely came first in the origin of life—not because RNA is easier to make than DNA (it isn't) but because RNA is not a rigid rod like the DNA double helix; being a single, 'wigglier' strand, RNA can do more, can also store information, and so a world with only RNA is easier to imagine than one with only DNA.
“It's because RNA isn't a rigid rod. DNA is this double helix that winds around itself and forms a long rod. RNA is generally just one strand, and so it's wigglier, and it can do more.”
- Viruses should be considered living matter because they reproduce, evolve, and are part of the DNA/RNA world; there is still a fundamental difference between living organisms and non-living matter, and viruses fall on the living side despite needing a host cell to reproduce.
“I've always been comfortable calling viruses life because they evolve, they do everything that living systems do. The only thing is they have to be inside a cell to reproduce.”
- Although all viruses use DNA or RNA and the same genetic code as the rest of life, they probably did not all evolve from a single universal precursor; DNA viruses and RNA viruses are very different, and marine bacterial viruses, plant viruses, and animal viruses likely represent disparate evolutionary stories without a common base.
“I don't think all viruses evolved from a universal precursor. I think that there have been multiple evolutions of viruses that don't have a common ancestor.”
Common cold's genetic diversity across hundreds of viral species defeats vaccination in a way single viruses do not.
- The common cold cannot be vaccinated against because it is not one virus but many—hundreds in humans, including rhinoviruses, coronaviruses, adenoviruses, both DNA and RNA viruses—all producing similar symptoms; its sheer variety, not any single intractable target, is what defeats a vaccine.
“the common cold is not something that we can make a vaccine against, because it's so... Has so much variety.”
Viruses provide ecosystem benefits by nutrient recycling despite harming individual organisms.
- Viruses can benefit an ecosystem while harming individual organisms—for example, in the ocean viruses constantly kill bacteria and break them open, enriching the water as a nutrient soup, which is good for the ocean ecosystem even though it is bad for the individual bacteria.
“the ocean is a rich soup because viruses are constantly killing off cells and breaking them open. Now, is it good for the... It's good for the oceans, but is it really good for the bacteria? I don't think so.”
Consciousness and neural function represent biology's most profound outstanding questions.
- The most interesting outstanding biological question is the origin of consciousness; neuroscience is the future of biology, and despite generations of work we still don't understand how the brain works or how it produces consciousness—the jump from neurons to consciousness is orders of magnitude more challenging to think about than even the jump from DNA to protein.
“the most enigmatic piece of it is consciousness. How do we... I've got consciousness of the room in front of me... but how do I do it? How do I make the images out of biology, out of neurons?”