The Peter Attia Drive - #401 ‒ How curiosity transforms medicine： extraordinary discoveries that changed modern healthcare

00:00:00.000 Hey, everyone. Welcome to the Drive podcast. I'm your host, Peter Atiyah. This podcast,

00:00:16.540 my website, and my weekly newsletter all focus on the goal of translating the science of longevity

00:00:21.520 into something accessible for everyone. Our goal is to provide the best content in health and

00:00:26.720 wellness. And we've established a great team of analysts to make this happen. It is extremely

00:00:31.660 important to me to provide all of this content without relying on paid ads to do this. Our work

00:00:36.960 is made entirely possible by our members. And in return, we offer exclusive member only content

00:00:42.720 and benefits above and beyond what is available for free. If you want to take your knowledge of

00:00:47.940 this space to the next level, it's our goal to ensure members get back much more than the price

00:00:53.200 of the subscription. If you want to learn more about the benefits of our premium membership,

00:00:58.020 head over to peteratiamd.com forward slash subscribe. Welcome to a special episode of

00:01:05.840 The Drive. Today, we're going to take a look at some critical pieces of history in modern

00:01:10.980 medical science. Now, we normally focus on this podcast on the findings and applications of

00:01:18.060 medical research. But in this episode, we're going to instead look at that process with an emphasis

00:01:23.860 on how apparently and at times seemingly irrelevant basic research can be the stepping

00:01:30.400 stone upon which medical revolutions are built. Now, we could have presented this as a series of

00:01:36.540 findings, but I think the argument works best if you actually see it play out to see who these

00:01:42.960 scientists were, what they were genuinely trying to do, and what it looked like in the moment

00:01:48.740 before anyone understood what they stumbled into. So rather than argue the thesis, I want to show

00:01:54.720 it to you story by story. By the time we're done, you should hopefully understand where several of

00:02:01.020 the most consequential drug classes of the last 50 years actually came from. And hopefully more

00:02:07.340 than that, you'll have a different framework for thinking about where medical progress actually

00:02:13.340 comes from and how that informs how we as a society should value the basic science research

00:02:20.820 that fuels medical innovation. So without further delay, I hope you enjoy this special episode of

00:02:27.780 the drive. In the summer of 1961, a young Japanese biochemist named Osamu Shimamuru and his wife,

00:02:41.420 Akemi, and his mentor, Frank Johnson, loaded into a station wagon in Princeton, New Jersey,

00:02:46.760 and drove 3,000 miles to the northwest corner of Washington State. Their destination was a place

00:02:54.060 called Friday Harbor on San Juan Island. They went there for jellyfish, specifically Acoria

00:03:01.360 Victoria, a small, mostly transparent jellyfish that drifts in the cold waters of the Pacific

00:03:07.620 Northwest. Its umbrella is rimmed with tiny organs that emit a faint green light. Shimomura was

00:03:15.340 building a career studying the chemistry of bioluminescence. They scooped them up one at a

00:03:21.200 time with shallow dip nets, brought them ashore, and cut the luminous rings off the umbrellas with

00:03:27.220 scissors by hand, one jellyfish at a time. The goal that first summer was 50,000. They came back

00:03:36.780 the next summer. And the next. For 19 consecutive years, Shimomura and his family and a rotating

00:03:45.480 cast of students returned to Friday Harbor. By the time they finally stopped in 1988,

00:03:52.480 they had cut the bells off approximately 850,000 jellyfish, drawn from a population

00:03:58.660 of about 1 million pulled from the bay. 19 summers, a million jellyfish.

00:04:07.340 If you had walked up to Shamimura on a dock in 1965 and asked him what he was doing,

00:04:13.600 he would have told you, honestly, that he was trying to understand how a jellyfish produces

00:04:18.760 light. That was the whole project. He wasn't trying to cure a disease, he wasn't designing a

00:04:24.260 drug, and he wasn't setting out to revolutionize the very process of scientific discovery. He was

00:04:31.260 a man with scissors cutting rings off jellyfish because he wanted to know how the animal glowed.

00:04:37.780 Here's what he found. When he ground up the jellyfish rings and purified the proteins,

00:04:43.240 the first thing he isolated was a protein that emitted blue light, not green, and only emitted

00:04:50.660 that light in the presence of calcium ions. He called it aquarin. That, by itself, was a beautiful

00:04:57.740 piece of biochemistry. Nobody had ever seen a calcium-triggered light-producing protein before.

00:05:05.260 But while he was purifying aquarin, he kept noticing in the background a trace contaminant,

00:05:11.120 a second protein that didn't glow on its own. But when you shone the right wavelength of blue

00:05:17.860 light on it, it fluoresced bright green. He purified it, and he named it plainly green

00:05:25.620 fluorescent protein, or GFP. The jellyfish, it turned out, was running a two-protein optical

00:05:34.480 system. Aquarin generated blue light from a calcium signal. That blue light then excited

00:05:42.300 GFP sitting right next to it, which absorbed the blue light and re-emitted the energy as green.

00:05:50.520 And that's the glow you see in the water. An elegant piece of jellyfish biology.

00:05:57.800 Shimomura published it, and he and his family went back to catching more jellyfish,

00:06:02.220 mostly to study Aquarin. GFP, the green protein, sat there, in the literature,

00:06:10.200 basically ignored for almost 30 years. And here's where the story becomes extraordinary.

00:06:17.960 Because GFP, it eventually turned out, has three properties that nobody had any reason to expect

00:06:25.160 from a random jellyfish protein. First, it folds itself. You don't need any helper machinery to

00:06:33.920 make it work. Second, it builds its own fluorescent core all by itself from three of its own amino

00:06:41.580 acids and using nothing but oxygen. And third, and this is the part that changes biology,

00:06:49.640 GFP works in almost any cell, in almost any organism on Earth.

00:06:56.120 Because it's a protein, that means we can put the gene into an organism, and the organism will make GFP.

00:07:04.000 Because GFP folds itself and creates its own fluorescent core, it doesn't rely on any jellyfish-specific signal.

00:07:12.000 So we can put the gene for GFP into a bacterium, and it glows.

00:07:17.160 Put it in a worm, the worm glows.

00:07:19.740 Put it into a mouse, into a plant, into a human cell in a dish,

00:07:23.780 they all glow under the presence of blue light.

00:07:27.840 Think about what that means.

00:07:29.300 We can make specific parts of biology glow.

00:07:33.580 And if you can attach the gene for GFP to any other gene you care about,

00:07:38.060 the protein made from that gene will carry a little green flashlight with it

00:07:43.660 everywhere it goes, inside a living cell. For the first time in history, you could watch biology

00:07:49.740 happen in real time in living tissue. You could watch a single protein move from one part of a

00:07:57.380 cell to another. You could watch a cancer cell crawl through tissue and metastasize in a living

00:08:03.940 mouse under a microscope. You could watch an embryo develop, cell by cell, in a transparent

00:08:10.800 zebrafish. You could mark a stem cell and follow every one of its descendants for the lifetime of

00:08:17.840 an organism. None of this was possible before GFP. None of it. Now, it's so routine, you develop

00:08:26.500 GFP transgenic bacteria in a freshman lab class. Once biologists figure out what they had in the

00:08:35.380 1990s, the biology research exploded. A researcher at Columbia named Martin Shelfie expressed GFP

00:08:42.980 in bacteria and worms. They glowed. A chemist at UCSD named Roger Chen spent the next decade

00:08:49.660 engineering variants cyan, yellow, orange, eventually red. Another team developed GCAMP

00:08:57.260 fusing GFP to a calcium-binding protein so GCAMP would only grow green light in the presence of

00:09:04.720 calcium signals. We could now watch neurons fire and watch sperm fertilize an egg.

00:09:12.320 We have an entire pallet of fluorescent reporters used in modern biology. All of them are descendants

00:09:19.520 in one way or another from the protein Shimamura pulled out of the jellyfish on a dock in

00:09:25.600 Washington state. I cannot overstate how much of modern biology runs on this. If you walk into

00:09:32.480 Essentially any major biology or biomedical research lab in the world today,

00:09:37.780 GFP or one of its descendants is in use somewhere on the bench.

00:09:42.840 Drug development pipelines depend on it.

00:09:45.160 Cancer research depends on it.

00:09:46.920 Neuroscience depends on it.

00:09:48.960 Molecular cellular systems biology.

00:09:52.380 GFP is in virtually every lab.

00:09:54.980 There is essentially no significant area of modern biology that isn't touched

00:09:59.580 somewhere in its toolkit by a protein that came out of a jellyfish. Because Shimamura had a

00:10:05.820 seemingly medically irrelevant, simple, basic curiosity, he wanted to know what makes jellyfish

00:10:13.400 glow. In 2008, the Nobel Prize in Chemistry was awarded to Oshamu, Shimamura, Martin,

00:10:21.720 Chalfi, and Roger Chen for the discovery and development of green fluorescent protein.

00:10:26.980 And every single one of those advances—the gene therapies, the cancer screens, the neuroscience—all

00:10:34.040 of it—exists because a man and his family spent 19 summers in a small bay in Washington

00:10:40.660 state cutting the bells off a million jellyfish, figuring out how jellyfish glow.

00:10:47.520 I want to start there because I think the jellyfish story makes a point about a persistently

00:10:52.500 underappreciated truth in modern medicine. A staggering fraction of what we now consider

00:10:58.920 modern medicine, the drugs in your medicine cabinet, the diagnostic tests you had run on

00:11:03.860 your blood, the gene therapies that are now starting to cure diseases we used to consider

00:11:08.000 death sentences, or even the foundational tools used every day by the laboratories that fuel these

00:11:13.540 discoveries, so much of it traces back to research that, at the moment it began, looked like it had

00:11:19.960 nothing to do with treating human disease. Somebody was studying a jellyfish, studying

00:11:26.480 this snake or that lizard's venom, studying fungus growing on rice, a microbe in a hot spring,

00:11:33.740 a repeating pattern in some single-celled organism's DNA. And later, sometimes decades later,

00:11:42.140 that research turned out to be the foundation of work that has improved or even saved millions of

00:11:47.860 lives. I think this is worth pausing on. The reflexive way we typically talk about medical

00:11:54.020 progress is in terms of intent, in terms of medical goals. We say things like, we're working

00:12:00.520 on a cure for Alzheimer's, or they're developing a drug for obesity. And of course, there is

00:12:06.600 enormous, important, intentional problem-solving science. In fact, some of the most important

00:12:12.300 advances in modern medicine have come from highly directed efforts. RT therapy, checkpoint inhibitors,

00:12:19.380 monoclonal antibodies, kinase inhibitors, antiviral therapies, mRNA vaccines, these required

00:12:26.020 extraordinary intentional engineering optimization, clinical development, and persistence.

00:12:31.940 The point isn't that directed science doesn't work. It obviously does. The point is that many

00:12:37.900 of those successes ultimately rest on a foundation of biological discoveries that came first.

00:12:44.720 But what I want us to think about today is that an enormous amount of the most consequential

00:12:50.040 medical progress does not come from people sitting in a room asking how to cure a disease.

00:12:54.920 It comes from people sitting in a room or going out in the field and asking,

00:12:59.440 how does this thing in nature actually work? The reason this can be so productive is something I

00:13:07.260 find to be genuinely humbling. Nature hasn't solved the problems we care about in medicine.

00:13:14.480 Evolution is not optimizing for longevity, and it certainly isn't optimizing for preventing

00:13:20.160 Alzheimer's disease, atherosclerosis, osteoporosis, or any of the chronic diseases that emerge long

00:13:26.340 after reproduction. But evolution has generated something else, an almost unimaginably vast

00:13:34.580 toolkit of molecules, signaling pathways, defense systems, metabolic adaptations,

00:13:40.520 and biological strategies. Nature has been running experiments across millions of species

00:13:46.480 for roughly 4 billion years. And while it hasn't solved our problems for us,

00:13:53.040 it has often already created some of the components of the solutions we're looking for.

00:13:59.140 The challenge is recognizing them when we see them.

00:14:02.820 While modern medicine is perhaps a few hundred years old,

00:14:06.860 nature has been running experiments across millions of species in parallel for billions of years.

00:14:14.280 We often find that the mechanisms we need already exist.

00:14:17.700 The molecule we need may already exist.

00:14:21.100 Sometimes we don't even know what the problem is and nature already has a trick up its sleeve.

00:14:25.760 we just have to be paying attention long enough and close enough to notice it. And once you

00:14:30.740 internalize that, a lot of medical history starts to look different. The methods look different.

00:14:36.720 The methods should look different, because if nature has already touched so many of these

00:14:41.860 problems, then sometimes the most productive thing a scientist can do is not invent something new,

00:14:48.240 it's to find the organism that already invented it. That principle is where we're going to spend

00:14:54.620 our time today. The natural world has solved a lot of problems. Once we find those solutions

00:15:01.340 and really recognize them for what they are, we can build on them and translate them into research

00:15:07.400 and clinical tools. And I think the next story is perhaps one of the cleanest illustrations of that

00:15:13.300 idea, finding a tool developed by nature and directly putting it into medical practice.

00:15:19.720 In the late 1960s, a Japanese biochemist named Akiro Endo was working at the Sankyo Drug

00:15:28.520 Company in Tokyo, and he was thinking about cholesterol. By that point, the basic biochemistry

00:15:34.780 of cholesterol synthesis was understood. Conrad Bloch and Fyodor Leinen had won the Nobel Prize

00:15:41.160 for working it out in 1964. Every biochemist in the field knew that the rate-limiting step

00:15:48.120 in cholesterol synthesis was a single enzyme called HMG-CoA reductase. And everyone understood

00:15:55.280 in principle that if you could just inhibit that enzyme, you should be able to lower the body's

00:16:01.200 cholesterol levels. The problem was nobody had any idea how to find such an inhibitor.

00:16:08.060 The medicinal chemistry of the era was not capable of designing one from scratch.

00:16:13.920 So Endo asked a different question.

00:16:16.500 He didn't ask, how do I design a molecule that inhibits HMG-CoA reductase?

00:16:21.560 He asked, who in nature is already working on one?

00:16:25.920 His reasoning went something like this.

00:16:28.060 Fungi and molds spend their entire existence at war with bacteria.

00:16:33.280 They live in the same wet, decaying, nutrient-rich environments,

00:16:38.700 and they're in constant chemical competition for resources.

00:16:42.820 Over hundreds of millions of years of that competition,

00:16:46.600 fungi have evolved an enormous arsenal of weapons,

00:16:50.380 small molecules designed to poison, disable, or starve their bacterial competitors.

00:16:56.700 We already had one famous example.

00:16:59.960 Penicillin, discovered by Alexander Fleming, is exactly that kind of weapon,

00:17:04.280 a chemical that a fungus uses to kill bacteria.

00:17:09.080 Endo's insight was this.

00:17:10.380 many bacteria, in order to grow, need sterols, or sterile-like molecules, to build their cell

00:17:17.540 membranes. And if you were a fungus trying to starve out your bacterial neighbors, one extremely

00:17:23.960 effective weapon would be a molecule that blocks bacterial sterile synthesis. Which means somewhere

00:17:31.060 out there, in the near endless catalog of compounds that fungi have spent hundreds of millions of

00:17:36.960 years inventing, there ought to be a molecule that inhibits the cholesterol synthesis pathway.

00:17:43.440 Specifically, there ought to be a molecule that inhibits HMG-CoA reductase. Some fungus,

00:17:50.780 somewhere, had already designed the drug to solve this problem. Endo sought out to find it.

00:17:57.540 So he and his colleague, Masao Kuroda, started screening. Over the next two years,

00:18:03.280 they tested more than 6,000 microbial strains, most of them fungal, for the ability to inhibit

00:18:11.260 HMG-CoA reductase. For two years, nothing. Finally, in March of 1972, they got a hit.

00:18:20.940 A strain of penicillium citronum growing on a rice sample from a grain shop in Kyoto

00:18:27.040 produced an extract with potent inhibitory activity against the enzyme. Endo shared his

00:18:33.760 results with the pharmaceutical company Merck, who went searching for fungi on their own for a

00:18:38.760 related compound they could in turn bring to the clinic. Shortly thereafter, they isolated a

00:18:45.140 closely related compound from Aspergillus tereus, a different fungus, called it lovastatin, and got

00:18:51.900 approved in the United States in 1987. From lovastatin, you eventually get simvastatin,

00:18:58.920 pravastatin, atorvastatin, rezuvastatin, the entire class of drugs that by any reasonable

00:19:05.420 estimate has prevented millions of cardiovascular events and millions of premature deaths.

00:19:12.340 I want you to notice the structure of what just happened because it's the cleanest possible

00:19:16.980 version of the argument I made a few moments ago. Akyro Endo did not invent the starting point for

00:19:24.340 statins. Penicillium Citrinum invented it. Aspergillus Tereus invented it. What Endo and

00:19:32.720 the generations of medical chemists who followed did was transform that starting point into a safe,

00:19:39.680 scalable, clinically useful class of medications. These fungi needed a way to inhibit sterol

00:19:48.000 synthesis in order to survive, for a different reason than ASCVD prevention to be sure,

00:19:54.440 but a need to inhibit sterol synthesis all the same. So Endo asked the right question of Nature,

00:20:01.440 who has already solved this problem? And Nature, sitting in a grain shop in Kyoto,

00:20:07.260 handed him the answer. A mold growing on a piece of rice that had been working on the

00:20:13.800 cholesterol problem for a few hundred million years before Aikiro Endo started looking for

00:20:20.040 the answer. He just had to be the one who thought to ask. Now, I say this as the cleanest case

00:20:27.800 because Endo went looking on purpose. He had a hypothesis about where nature might have already

00:20:36.120 solved the problem. And he turned out to be right. But most of the stories I want to tell you next

00:20:41.180 are not like that. And most of them, people who found a paradigm-shifting answer were not even

00:20:47.440 looking for it at all. They were looking at something else entirely. A snake, a hot spring,

00:20:54.280 a salt pond, a desert lizard. And what they found there turned out, in retrospect, to change

00:21:00.680 everything. Which takes us to South America. In Brazil, in the first half of the 20th century,

00:21:08.980 Bothrop's Gerarica, the Brazilian lancehead pit viper, was a serious public health problem. It

00:21:16.400 still is. The Gerarica lancehead is a beautiful, irritable, zigzag-patterned pit viper about three

00:21:25.980 to five feet long, and its venom causes a sudden catastrophic drop in blood pressure. In 1901,

00:21:34.560 the Butantan Institute in Brazil was founded, in no small part to understand venomous snake bites.

00:21:41.300 They began collecting, studying, and stockpiling pit viper venom. Eventually, venom began to be

00:21:49.740 leveraged as a research material, lyophilized, abundant, locally available, and it stopped being

00:21:57.000 just an object of study itself. Pharmacologists could use it as a reagent to probe mammalian

00:22:03.660 physiology. In the late 1940s, the Brazilian pharmacologist Mauricio Rocha e Silva was working

00:22:11.420 on circulatory shock. At the time, circulatory shock was thought to be mediated by histamine.

00:22:17.720 He wanted to know whether proteolytic enzymes, common ones like trypsin or proteases in gerarica

00:22:27.680 venom, released histamine from plasma. So he exposed blood to gerarica venom. The venom did

00:22:36.080 not release histamine from blood samples. It released something else. This was the discovery

00:22:41.720 of a new peptide, a potent vasodilator, which also produced a slow contraction in the small

00:22:48.980 intestines of guinea pigs. He named this peptide bradykinin, from the Greek for slow movement.

00:22:57.340 This was a massive discovery. In the decades since, bradykinin has turned out to be a fundamental

00:23:03.440 player in mammalian cardiovascular biology, an endogenous regulator of blood pressure,

00:23:09.620 vascular permeability, and inflammation. But our story doesn't stop here. A decade and a half

00:23:15.580 later, Rocha Isilva had a graduate student named Sergio Ferreira, and he handed Ferreira a puzzle.

00:23:24.120 Bradykinin generated in plasma by venom was more active than a synthetic bradykinin produced in a

00:23:32.160 test tube. Why? Ferreira's answer, published in 1965, was that the venom also contained a separate

00:23:39.940 family of peptides that potentiated bradykidin's effects. Bradykinin potentiating factor, or BPF,

00:23:49.060 by blocking the breakdown of bradykinin. The venom proteases caused bradykinin to be released

00:23:56.680 in the plasma, and BPFs in the venom protected that bradykinin from degeneration. The BPF work

00:24:05.520 landed Ferreira a postdoctoral position in London, in the vein lab at the Royal College of Surgeons.

00:24:13.100 In the lab, another postdoc named Kevin Ink was studying angiotensins in pulmonary circulation,

00:24:19.840 having found that angiotensin-1 is converted to angiotensin-2, a vasoconstrictor, in the lungs.

00:24:27.960 Ferrara brought powdered BPF with him to London from Brazil,

00:24:32.280 and his work with Kevin Ng converged on an unexpected finding.

00:24:37.160 By adding BPF to Kevin's preparations,

00:24:41.340 it turned out that the same enzyme that was degrading bradykinin

00:24:45.740 was also converting angiotensin-1 to angiotensin-2.

00:24:49.840 Both jobs were inhibited by BPFs. This enzyme apparently had two synergistic jobs,

00:24:58.680 destroy the vasodilator, gradykinin, and produce the vasoconstricting angiotensin 2.

00:25:07.020 They found that this enzyme was angiotensin-converting enzyme, or ACE. Now, ACE had

00:25:14.280 previously been discovered, but the identification of a potent ACE inhibitor caught the attention

00:25:20.980 of chemists at Squibb. The venom peptides themselves were not viable drugs. They're not

00:25:27.580 orally active and they have a short half-life. Squibb's chemists took the smallest active BPF

00:25:34.720 peptide from the Gerarica venom and used it as their structural lead and designed a small

00:25:42.000 orally active molecule that inhibited ACE in the same way that venom peptide did.

00:25:48.980 This is conventionally remembered as one of the earliest examples of rational,

00:25:53.360 mechanism-based drug design. They called the molecule Captopril, and it was approved by the

00:25:58.940 FDA in 1981 as the first oral ACE inhibitor in clinical use. And Captopril, Begat, Enalapril,

00:26:07.640 lisinopril, ramipril, and the rest of the ACE inhibitor class, and conceptually began the

00:26:14.200 angiotensin receptor blockers, losartan, valsartan, and so on. The renin-angiotensin-aldosterone

00:26:22.120 system is now one of the most heavily drugged pathways in all of medicine. These drugs are

00:26:28.500 foundational in the treatment of hypertension, heart failure, chronic kidney disease, post-MI

00:26:34.680 remodeling, and diabetic nephropathy. They are plausibly among the most consequential drug

00:26:41.260 classes ever developed. So let's sit with the causal chain for a moment. Millions of lives

00:26:47.300 saved or extended because the Butantan Institute was stockpiling local snake venom. A clinical

00:26:54.580 question, how does this venom kill people, produced a research reagent. The reagent used

00:27:00.780 to probe a wrong hypothesis about shock produced the accidental discovery of bradykinin, a fundamental

00:27:08.140 signaling molecule in cardiovascular and inflammation biology. A discrepancy in a

00:27:14.120 bradykinin bioassay produced BPFs. BPFs, carried to a London lab studying lung vasculature,

00:27:23.000 produced the recognition that one enzyme sat at the center of two opposing blood pressure systems,

00:27:29.060 And that recognition, handed to two medicinal chemists, produced a world-changing drug class.

00:27:36.340 For the first 70 years of the 20th century, nobody was trying to develop a hypertension

00:27:41.980 drug.

00:27:42.840 The molecule fell out of 70 years of fieldwork and basic biology, each step asking a question

00:27:48.540 the previous step hadn't anticipated.

00:27:51.420 The question of how a Brazilian pit viper venom kills its prey could have saved thousands

00:27:57.340 of lives in Brazil.

00:27:58.460 Instead, through the curiosity of basic medicinal scientists, tens of millions of lives around the

00:28:06.860 world have been saved or extended. For a more technical audience, you already know the importance

00:28:13.880 of the next story, and it's therefore somewhat impossible to overstate. Anyone who has spent

00:28:19.480 any time in a biology lab, including simply taking a freshman year biology lab class at university,

00:28:25.480 knows how fundamentally this work has changed biology and medicine forever. It is the technique

00:28:33.060 that opened the door to the last few decades, the golden age of biology research, and it begins

00:28:40.340 at a hot spring. In the mid-1960s, a microbiologist at Indiana University named Thomas Brock started

00:28:46.780 taking summer trips to Yellowstone National Park. He wasn't a typical hiker. He was a microbial

00:28:53.040 ecologist who became interested in extremophiles, organisms that live in extreme environments.

00:29:00.160 Yellowstone, with its hundreds of geothermal features, has a higher concentration of accessible

00:29:05.760 high-temperature aquatic habitats than just about anywhere else on Earth. At the time,

00:29:12.060 the prevailing belief in microbiology was that bacteria could grow in conditions up to about

00:29:18.700 55 degrees Celsius, after which they might survive, but the heat prevented their molecular

00:29:25.580 machinery from successful reproduction. As temperatures keep rising, the belief

00:29:30.540 was that proteins may be denaturing, membranes dissolving, and the whole thing a mess of chaos

00:29:37.260 and thermodynamics. Nothing could survive. Hot springs near boiling temperatures were therefore

00:29:43.680 assumed to be necessarily sterile. Brock didn't believe it. He started looking through Yellowstone's

00:29:50.960 hot springs, and he was right. At a place called Octopus Spring, he observed pink filamentous

00:29:57.980 bacteria growing in water at 88 degrees Celsius, well above the supposed ceiling for life.

00:30:06.020 Over the next several years, he and an undergraduate student named Hudson Fries

00:30:11.060 worked out the technique to isolate and culture these organisms, no small task given their

00:30:17.220 preferred resting temperature. And in 1969, they published a paper successfully characterizing

00:30:24.300 a species of thermophilic bacteria they'd isolated from mushroom spring in Yellowstone's

00:30:31.840 lower geyser basin. They fittingly named it Thermus aquaticus, Latin for hot and of the water.

00:30:40.000 And that was their project.

00:30:41.460 Two scientists asking whether anything was alive in nearly boiling water.

00:30:47.760 The answer was yes.

00:30:49.060 They published the paper.

00:30:50.380 They moved on.

00:30:51.780 Now we jump forward 15 years.

00:30:54.440 In 1983, a biochemist named Cary Mullis was working at a biotech company in California

00:31:01.740 called Cetus.

00:31:03.400 And he was thinking about how to copy DNA in a test tube.

00:31:07.260 The conceptual idea, which he famously had on a drive up to Mendocito, was beautiful.

00:31:13.800 You separate the two strands of DNA double helix by heating it near boiling, effectively

00:31:19.780 melting the hydrogen bonds that keep the sister strands together.

00:31:24.580 You then cool the solution, which allows short, specific DNA sequences, called primers, to

00:31:31.480 bind to the regions you want to copy as the temperature returns to normal.

00:31:36.480 Then, you have a polymerase enzyme extend the DNA, which will start at those primers

00:31:43.280 bound to their target sequence, thus creating copies of the parent strand in your region

00:31:49.960 of interest.

00:31:50.640 You now have two copies of double-stranded DNA sequence of your choosing, having started

00:31:57.760 with a single double-stranded parent, and then you repeat the process.

00:32:02.140 each cycle doubles your DNA. After 30 cycles, you have turned one molecule into a billion copies.

00:32:12.420 This technique is known as polymerase chain reaction, or PCR. But there was just one problem.

00:32:20.640 The DNA polymerases people used at the time were derived from E. coli, and they fell apart when

00:32:27.640 you heated them. So when every cycle of PCR required denaturing the DNA at near boiling

00:32:34.720 temperatures, the polymerases could not survive the process, which meant you had to add fresh

00:32:41.200 enzyme manually at every cycle. It worked, but it was tedious, expensive, error-prone,

00:32:49.020 and basically not practical at any meaningful scale. The scientists at CETIS famously developed

00:32:56.520 a robot to add fresh polymerase after each round of PCR in its thermocycler, dubbed Mr. Cycle.

00:33:05.400 But this was still nowhere near a high-throughput, high-fidelity, cost-effective means for amplifying

00:33:12.240 DNA. The team at Cetus needed a polymerase that could survive being heated to 95 degrees Celsius

00:33:20.320 over and over without falling apart. They needed an enzyme that could survive in near

00:33:26.760 boiling water. They turned to Thermus aquaticus. Polymerase is fundamentally necessary for life

00:33:34.640 to continue. Every organism necessarily requires a means of copying their own DNA

00:33:40.500 in order to pass that DNA onto new cells or new generations. Thermus aquaticus could apparently

00:33:47.340 do this at temperatures of at least 88 degrees Celsius. In the mid-1970s, the Thermus aquaticus

00:33:54.660 polymerase, commonly referred to as TAQ or TAC polymerase, had been isolated and partially

00:34:02.880 characterized, in part because the species had been deposited in public culture collections

00:34:09.100 after Brock's work. The CETIS team built on this, purified TAC polymerase in their lab,

00:34:14.900 and plugged it into their PCR protocols. With this, the entire technique and all of biological

00:34:22.420 sciences transformed. Now you could load the reaction, set the thermocycler, walk away,

00:34:29.240 and come back to a billion copies of your sequence. We had a near endless supply of any

00:34:36.920 DNA sequence we desired from any species we desired ready to go after a few hours of incubation

00:34:45.240 on a cycling heating block. Mullis won the Nobel Prize in Chemistry in 1993 for inventing PCR,

00:34:54.180 and deservedly so, but it was only possible because Brock and Freeze had asked decades earlier

00:35:00.960 whether anything lived in extreme temperatures. They found life in an 88 degree Celsius hot

00:35:09.100 spring. They found a way to collect and culture those species, and they shared that species with

00:35:15.140 scientists everywhere to see, study, and utilize on their own. And I want us to think about what

00:35:22.160 that unleashed. PCR is not a drug. PCR is something more fundamental than a drug. PCR is the enabling

00:35:30.060 technology for essentially all of modern molecular biology. I'm sorry, this list is long. It's long

00:35:38.160 enough to make the point, but it's not long enough to be nearly exhaustive. DNA is the code upon which

00:35:45.360 all of life is programmed, and PCR is the tool we use to interact with that code. PCR is everywhere.

00:35:52.880 It's the basis of genetic testing, forensic DNA analysis, cancer mutation panels, prenatal screening, vaccine development, ancestry testing, transgenic lab animals, fluorescent reporters like GFP, the cloning of any recombinant therapeutic protein, the entire field of genomics, the development of every gene therapy, and viral infection testing.

00:36:19.120 All of it depends on PCR. So all of it depends on TAC polymerase. And all of it traces back

00:36:26.420 to a microbiologist wondering whether anything could live in hot water. Without that 1969 paper,

00:36:33.760 modern biology as it exists today simply could not exist. There's no plausible alternative history

00:36:40.040 in which we get to where we are without somebody somewhere going to look in the boiling water for

00:36:46.560 life. We just got lucky that that somebody was Tom Brock. And perhaps we got especially lucky

00:36:52.320 that he got to do it in advance without anybody asking what disease he was trying to cure.

00:36:59.180 And yet the discovery is so fundamental, so irrevocably intertwined with modern biology

00:37:05.320 and medicine, that it's hard to imagine how any disease at all could have been cured in the last

00:37:11.520 30 years without his work, his curiosity leading him to search for signs of microbial life

00:37:18.100 inside of Yellowstone hot springs. This next story is, to me, a particularly exciting illustration

00:37:25.760 of the principle because its ramifications were understood so recently that we are really in the

00:37:32.180 midst of watching the timeline play out. The story begins in the late 1980s in the small

00:37:38.180 Spanish coastal town of Santa Pola near Alicante on the Mediterranean. Santa Pola is famous for

00:37:45.460 its salt flats, vast, shallow, evaporative ponds where seawater concentrates, creating dramatically

00:37:52.920 saline pools of water. These salt ponds support a strange community of organisms, salt-tolerant

00:37:59.920 or even salt-requiring microbes called halophiles, many of them archaea, that have adapted to thrive

00:38:07.500 at salinity levels that would kill virtually anything else. A young Spanish microbiologist

00:38:14.520 named Francesco Mojica was doing his PhD on one of these archaeas, Halophorex mediterinae,

00:38:23.000 at the University of Alicante. His original project, in fact, was almost a different project

00:38:29.040 entirely. He was studying how this organism responds to changes in salinity, but someone

00:38:34.680 beat him to the punch and published it first. So along the way, he kept stumbling over a weird

00:38:39.820 feature of its genome. There were these clusters of short, regularly spaced palindromic repeats,

00:38:47.840 sequences of about 30 base pairs that repeated over and over, separated by unique, non-repeating

00:38:53.900 spacer sequences of regular length. Now, just parenthetically, for those of you who might not

00:38:59.800 be familiar with a palindrome. A palindrome is something that when read left to right and right

00:39:05.460 to left is the same. So the word race car is a palindrome because regardless of which way you

00:39:10.920 read it, it's the same way. And of course, my favorite palindrome, which I just learned of

00:39:15.500 recently from one of our research analysts, is the phrase, go hang a salami, I'm a lasagna hog.

00:39:23.420 I encourage you to prove to yourself that that is indeed a palindrome. Anyway, these were everywhere 0.99

00:39:28.640 in the genome, but he didn't know what they were. In fact, nobody knew what they were.

00:39:33.040 He wasn't the first to see them, exactly. A Japanese group led by Yoshizumi Ishino had noted

00:39:38.720 similar repeating patterns in E. coli in 1987, left as effectively a footnote in the discussion

00:39:46.640 section of their paper. Mohika was the first to identify them explicitly as a feature to

00:39:53.380 investigate. The function was a complete mystery, and most molecular biologists who came across

00:40:00.120 these sequences shrugged and moved on. Mojica didn't move on. He spent the better part of a

00:40:06.880 decade, the entire 1990s and into the early 2000s, trying to figure out what these repeats were

00:40:13.980 doing. The breakthrough came in 2003. By that point, the genomics revolution was in full swing,

00:40:21.380 and there were enough sequenced genomes in the public databases that Mojica could do something

00:40:27.120 he couldn't do before. He could take the unique spacer sequences, the bits between the repeats,

00:40:33.900 and ask the sequencing database, have you seen this sequence before? He sent it through the NIH

00:40:40.560 database known as BLAST, and the answer came back, yes. Some of those spacer sequences exactly

00:40:48.220 matched DNA from bacteriophages, viruses that infect bacteria and archaea. That was the key.

00:40:58.020 Mojica realized that the repeats weren't random junk. They were a record. Each spacer was a

00:41:06.060 snippet of DNA that the organism had captured from a past bacteriophage infection. The clusters

00:41:14.520 of repeats were the index. Between the repeats were the codes for memories of past infections.

00:41:23.500 This looked like an immune system, specifically an adaptive immune system. The bacterium,

00:41:30.720 or archaean, was in effect vaccinating itself against viruses by holding onto fragments of

00:41:38.700 the phage's genomes and using that information to recognize and destroy them on future encounters.

00:41:46.020 This was a stunning insight. It implied that bacteria and archaea, the simplest organisms

00:41:52.720 in biology, had a sophisticated adoptive immune system with memory. We now call that system

00:41:59.600 CRISPR, clustered regularly interspaced short palindromic repeats, and the nearby proteins

00:42:08.540 were deemed CAS, short for CRISPR-associated. Mojica wrote up the paper. He submitted it to

00:42:16.120 Nature in 2003. It was rejected without external review. He sent that to the Proceedings of the

00:42:23.700 National Academy of Science, rejected, lacking, quote, novelty and importance. He sent it to

00:42:30.880 molecular microbiology, rejected. Nucleic acids research, rejected. Increasingly desperate and

00:42:38.980 now afraid he was going to be scooped, he eventually submitted it to a smaller journal

00:42:45.000 called the Journal of Molecular Evolution, which, after 12 months of review and revision,

00:42:51.340 finally published it in February 2005. This is another place where we see that revolutionary

00:42:57.380 science isn't simply underfunded. Once the revolutionary insight is seen and documented,

00:43:02.780 even then, the field can still miss what it's looking at. The discovery that bacteria have an

00:43:08.920 adaptive immune system, arguably one of the most important findings in molecular biology over the

00:43:13.400 past 50 years, was rejected by four of the top journals in the field. The reviewers and editors

00:43:20.020 did not see the significance of what they were looking at. Mojica was studying salt pond archaea.

00:43:25.760 He was a Spanish microbiologist with a small lab, and his findings sounded like esoteric

00:43:31.040 microbial biology. It took a lesser journal to publish what is, in retrospect, an absolute

00:43:37.120 landmark paper. From there, the timeline is short. Labs around the world began to catch on,

00:43:44.780 or discover other features on their own. Danisco, yeah, the yogurt company, provided direct

00:43:51.320 experimental evidence that CRISPR really does work as an adaptive immune system in bacteria.

00:43:57.340 Bacteria are necessary for dairy fermentation. Bacteria are susceptible to outbreaks

00:44:02.400 of phage infections. This is an enormous practical problem for industrial dairy.

00:44:08.240 Danisco's 2007 paper in the journal Science validated Mojica's hypothesis and identified

00:44:15.100 Cas9 as the protein that recognizes CRISPR-encoded sequences to cut and destroy bacteriophage

00:44:22.360 DNA.

00:44:23.380 In 2012, Jennifer Doudna and Emmanuel Charpentier, and around the same time a group at the Broad

00:44:30.160 Institute and MIT led by Feng Zhang, who we've previously had on the podcast, showed

00:44:35.700 that you could take this bacterial immune system, program it with a custom guide RNA,

00:44:41.400 and use it to cut DNA at any specific location in any genome you wanted.

00:44:48.280 In bacteria, the goal of the cut was to destroy the phage's ability to reproduce.

00:44:54.060 Doudna and Charpentier showed that you could instead provide a new template of DNA

00:44:59.760 alongside the CRISPR system, and instead of destroying the genome,

00:45:03.660 you could incorporate new DNA into the cut site.

00:45:07.000 You could modify genes within the genome of a living organism. Doudna and Charpentier won the

00:45:14.020 2020 Nobel Prize in Chemistry for this. The implications here are enormous. Most gene

00:45:20.060 therapy trials had been discontinued. Their typical goal was to add a new gene to people

00:45:26.380 with genetic disorders. These extra genes don't always stick around for long, so durability could

00:45:32.200 be a concern. And if the gene itself was toxic, not just losing function like in Huntington's

00:45:38.120 disease, the normal variant's efficacy could be limited by the competition with the host's own

00:45:44.240 pathological gene. We had no way to actually modify the gene variant that led to a disease,

00:45:50.300 but now we do. And in December of 2023, the FDA approved cascovy, the first CRISPR-based therapy

00:45:57.640 ever brought to clinic. It's the treatment for sickle cell disease. Sickle cell disease until

00:46:02.980 very recently was a disease with limited options and shortened lifespans. CRISPR therapy isn't a

00:46:08.160 cure in the simple sense. It's complex, expensive, and requires bone marrow conditioning. But for the

00:46:13.620 patients who can access it, the results are extraordinary. The vaso-occlusive crisis that

00:46:18.720 defined their lives essentially stopped. From a strange DNA pattern in a Spanish salt pond 0.95

00:46:25.920 archaea to a therapy that durably treats sickle cell disease. And even recently, a phase 1b trial

00:46:33.780 posted incredibly promising results for a CRISPR-derived treatment of familial hypercholesterolemia.

00:46:40.040 30 years since the salt ponds, CRISPR is curing diseases. And again, look at the structure of

00:46:46.140 what happened. Francisco Mojica did not set out to cure sickle cell disease. He set out to understand

00:46:52.560 why the genome of an obscure, salt-loving Archeon had repeated sequences in it. The system he

00:46:59.120 uncovered turned out to be a programmable molecular tool of staggering power. We, humans, the species

00:47:07.000 that builds the technologies, didn't invent CRISPR. Bacteria invented CRISPR somewhere around

00:47:13.180 3 billion years ago as a defense against viruses that have been trying to exploit them since

00:47:20.060 shortly after life began. Mohica just noticed it. We didn't go looking for a tool to edit the

00:47:26.500 genome. We found it when wandering through salt flats. The last story I want to tell is perhaps

00:47:33.660 the most relevant story in terms of disease burden. It's a little less basic exploration

00:47:41.240 than some of our previous examples, but it's also the most tangibly relevant. Every listener to this

00:47:47.620 podcast, hell, probably every person in the English-speaking world at this point knows these

00:47:52.420 drugs, but very few may know where they came from. In the late 1980s, a clinical endocrinologist named

00:47:58.040 John Eng was working at the Bronx VA Medical Center in New York. He was a hands-on primary

00:48:04.280 care endocrinologist who spent most of his time taking care of veterans with diabetes.

00:48:09.100 But he had a side project. He worked on the bench in the laboratory of Rosalind Yalow,

00:48:14.020 the Nobel Prize-winning physicist and biochemist who, decades earlier, had invented the radio

00:48:19.320 immunoassay technique. Yalow's lab had developed unusually sensitive methods for detecting and

00:48:26.300 characterizing peptide hormones. And as an endocrinologist working with diabetic patients,

00:48:32.340 Aang wanted to use the tools to hunt for novel peptides that affected the pancreas.

00:48:38.380 He'd read something interesting in the literature. Researchers at NIH had reported

00:48:43.160 that the bite of certain venomous reptiles, including notably the Gila monster, a stocky,

00:48:50.220 slow-moving, beaded-skinned lizard native to the deserts of the American Southwest,

00:48:56.460 caused dramatic inflammation and enlargement of the pancreas in bitten animals. Something in the

00:49:03.600 venom was hitting the pancreas hard. Why would a desert lizard have a venom that targets the

00:49:10.180 pancreas. Heng didn't know, but he thought it was worth investigating. The Gila monster also had

00:49:16.880 another biologically interesting property. It eats very rarely, sometimes only three or four

00:49:24.760 meals a year. And yet between meals, it appears to maintain remarkably stable blood glucose levels

00:49:32.960 through long stretches of fasting. Whatever the lizard's pancreas was doing, it was doing it 0.58

00:49:39.680 differently from ours. So Aang ordered some dried Gila monster venom and started running it through 0.98

00:49:46.200 the assays he's inherited from Yalo's lab. Working with his colleague Jean-Pierre Roffman, 0.96

00:49:52.840 who would shuttle samples back and forth from Brooklyn to the Bronx in his Toyota Camry,

00:49:58.620 Aang systematically separated the peptide components of the venom. He found two peptides,

00:50:05.260 one had been described before. The second was new. Published in 1993, he named it Exendin-4.

00:50:14.180 When he sequenced Exendin-4, the result was remarkable. The peptide showed about 53%

00:50:21.640 amino acid sequence similarity to a human hormone called GLP-1, glucagon-like peptide 1,

00:50:30.020 a gut hormone produced in the intestine that stimulates insulin secretion in response to a

00:50:37.380 meal. Oddly, the lizard had a venom peptide that activated, in a mammal that it had bitten,

00:50:43.640 the same receptor as our own meal-induced insulin-stimulating hormone. That explained

00:50:50.420 the pancreatic inflammation. The bite was hitting GLP-1 receptors like a hammer.

00:50:57.640 Xsendin-4 had a key property that GLP-1 does not.

00:51:03.100 Native GLP-1 has a half-life in the bloodstream of about two minutes.

00:51:08.200 An enzyme called DPP-4 chews it up almost as fast as the gut makes it.

00:51:13.920 That's actually fine for what GLP-1 normally does,

00:51:17.020 but it makes the native hormone useless as a drug.

00:51:20.260 You'd have to be infusing it in constantly.

00:51:23.480 Exendin-4, by contrast, is resistant to DPP-4.

00:51:27.980 Its half-life is on the order of hours, not minutes.

00:51:31.500 It acts on the GLP-1 receptor with a durability that GLP-1 itself lacks.

00:51:38.560 Eng had found a long-acting GLP-1 receptor agonist

00:51:42.560 sitting in the saliva of a desert lizard, ready to use.

00:51:47.860 Eng tried to get the VA to patent the discovery.

00:51:50.500 The VA declined.

00:51:52.240 Eng eventually patented it himself, licensed it to a small biotech, and the Wrights eventually

00:51:58.320 made their way to Eli Lilly and Amelin. In 2005, the FDA approved synthetic Xenon-4 under the

00:52:06.720 trade name Bieta, generic name Xenotide, for the treatment of type 2 diabetes. It was the first

00:52:13.700 GLP-1 receptor agonist ever brought to market. And then the floodgates opened.

00:52:20.220 Exendin-4 was a foreign Gila monster peptide, so many patients developed immune responses to the

00:52:26.720 therapy. The therapy required twice daily injections, provided carefully around meal

00:52:32.280 windows, and included side effects like kidney strain and nausea. That said, in patients who

00:52:38.240 tolerated it, it worked. Novo Nordisk, Lilly, and others realized that this was an enormous

00:52:44.800 biological lever, not just a way to improve glycemic control in diabetes, but, as it turned

00:52:50.840 out, an extraordinary tool for weight loss. As we've talked about at length, GLP-1 signaling

00:52:56.500 also suppresses appetite, slows gastric emptying, and changes how the brain weighs reward signals

00:53:02.560 around food. Subsequent drugs were not direct extendin-4 analogs, but modifications of the

00:53:10.000 human GLP-1 sequence engineered to reduce immunogenicity, increase half-life, and in some

00:53:17.080 cases add alternative targets in addition to GLP-1. We now have liraglutide and semaglutide

00:53:23.740 along with terzepatide, which adds GIP agonism in addition to GLP-1. We've talked about these

00:53:31.560 drugs at length and we'll continue to talk about them as the science develops. Suffice it to say,

00:53:35.980 these drugs are reshaping medicine in real time. They're reshaping the treatment of type 2 diabetes

00:53:41.380 and they're reshaping the treatment of obesity. The cardiovascular outcome data is changing how

00:53:46.560 we think about metabolic disease as a driver of atherosclerotic risk. And there are ongoing

00:53:51.720 trials in heart failure, kidney disease, sleep apnea, alcohol use disorder, addiction more

00:53:56.740 broadly, and possibly Alzheimer's disease. We don't yet know how broad the indication space

00:54:02.420 will turn out to be, we do know that this is one of the most consequential drug classes of the last

00:54:07.780 50 years. And the proximate origin was a clinician at a Bronx VA hospital wondering what he might

00:54:14.760 learn about endocrinology by studying how a desert lizard's bite caused pancreatic inflammation.

00:54:22.380 Now, these stories are, I think, incredibly interesting in their own right, and maybe I'm

00:54:27.000 just a sucker for research. They are each, in my opinion, fascinating illustrations of the

00:54:32.600 scientific process. But I also want to bring this back to the larger argument because I think that

00:54:37.780 point matters. Let's look at what we just walked through. Modern cell biology runs on a protein

00:54:43.440 from a jellyfish. Statins come from a fungus growing on rice. The ACE inhibitor class came

00:54:50.120 from a snake whose venom kills by dropping blood pressure. PCR, and by extension, essentially all

00:54:56.280 modern molecular biology came from a microbe in a Yellowstone hot spring. CRISPR, the foundation

00:55:02.120 of modern gene editing, came from the genome of a salt pond organism. And the GLP-1 class of drugs

00:55:08.740 came from a desert lizard that's hardly known outside of trivia. With the exception of statins,

00:55:14.800 the original scientist was not trying to cure a disease. Shimomura wasn't trying to revolutionize

00:55:21.380 cell biology. Ferreira was studying proteins in snake venom. Brock was studying life in hot

00:55:28.440 springs. Mojica was studying salt ponds. Eng was curious about pancreatic function.

00:55:34.840 If you had taken any one of these projects in the year it started and put it in front of a

00:55:39.580 translational impact review panel, a panel that asks what disease will this cure and on what

00:55:44.580 timeline, many of them would have struggled to get funded. Some almost certainly would have failed

00:55:50.320 outright. I want to study how a jellyfish glows is not a fundable proposal in that framework.

00:55:56.720 In fact, Douglas Prasher, who cloned the GFP gene, could not get NIH to fund him to continue.

00:56:03.980 The system failed in exactly the way I'm describing. And yet, collectively, this body of

00:56:08.980 work, this catalog of seemingly irrelevant, curiosity-driven, what-is-going-on-in-nature

00:56:15.180 type of science produces foundational tools and therapies that are now improving or saving the

00:56:21.660 lives of conservatively hundreds of millions of people. There's almost nothing you can do in

00:56:28.080 modern medicine that doesn't, at some step in its supply chain, depend on at least one of the

00:56:34.320 discoveries I just described. I think we are systematically bad at predicting which curiosity

00:56:40.620 driven work is going to matter. But I'm not convinced there is actually a way to become

00:56:45.500 systematically good at it. I think this is just one of the most important and most underappreciated

00:56:50.640 facts about how scientific progress actually happens. And I think it has real consequences

00:56:56.320 for how we should think about science funding, science policy, and the cultural framing of

00:57:02.920 research as something that needs to justify itself by proximity to a cure. Now, I want to be careful

00:57:09.840 here. I am not arguing that all basic science is good or that we should fund every grant proposal

00:57:15.220 that comes through the door regardless of merit. There is an obvious element of survivorship bias

00:57:20.460 here. The basic science that we are talking about today is the basic science that is worth talking

00:57:25.640 about today. Almost none of today's curiosity-driven projects will become GFP or CRISPR.

00:57:33.520 But that's actually the point. The lesson is not that every exploratory project will become

00:57:38.580 transformative. The lesson is that we are extraordinarily bad at predicting ahead of time

00:57:43.940 which ones will. There is bad basic science, there is uninteresting basic science, and there is

00:57:49.900 rigorous, accountable peer review that we should absolutely preserve. What I'm arguing is something

00:57:55.960 more specific, that the criterion of near-term translational obvious impact, the criterion

00:58:03.140 that asks of every project, what disease will this cure, is not always the right filter to apply.

00:58:10.200 It seems obvious once you take a step back. Nature has been doing experiments for longer than we have.

00:58:17.500 Four billion years, give or take. Every problem we're trying to solve in medicine, cholesterol

00:58:22.460 regulation, blood pressure regulation, glucose metabolism, immune memory, repairing damaged DNA,

00:58:28.140 surviving extreme stress, every single one of these problems has been encountered in some form

00:58:33.520 by some organism at some point in evolutionary history, either to exploit these pathways in the

00:58:40.160 case of things like snake venoms, or to survive them in the case of extreme environments.

00:58:45.680 And in many cases, through millions or billions of years, evolution has shaped a reasonable

00:58:51.620 solution. Nature has, at a minimum, developed a remarkable toolkit. The history of

00:58:58.020 medicine is full of examples where we didn't invent the underlying mechanism. We discovered

00:59:03.600 it, understood it, and then engineered it into something useful. Many of the molecules are just

00:59:10.240 sitting out there, the mechanisms already laid out. We just have to be willing to go look,

00:59:16.680 including, and perhaps especially, in places that don't at first look medically relevant.

00:59:23.080 In retrospect, the work that turns out to matter

00:59:25.880 almost always begins as a question that is not yet obvious about a disease.

00:59:31.080 It begins with somebody who wants to understand

00:59:33.460 how some piece of nature actually works.

00:59:36.520 It's driven by curiosity.

00:59:39.060 And I want to make sure this is not interpreted as discounting

00:59:42.380 the enormous amount of work that follows.

00:59:45.360 Curiosity may drive an initial discovery,

00:59:47.840 But intentional translation is obviously a pivotal part of the process too.

00:59:54.360 An immune system in salt pond archaea cannot cure sickle cell disease.

00:59:59.320 We need scientists directly characterizing the human disease, manipulating the CRISPR system,

01:00:05.260 and going through years of dedicated focused experimentation before the basic discovery

01:00:11.340 becomes a human therapeutic.

01:00:12.880 So I do think it's worth seeing biomedical science as existing on a continuum, with necessary time, attention, and resources devoted to each step in the chain.

01:00:24.040 It is easy to look at recent medical advances and see something human-driven, human-designed, and intentional.

01:00:31.280 Humans-made CAR T cells and checkpoint inhibitors for cancer therapy, monoclonal antibodies, kinase inhibitors, mRNA vaccines.

01:00:39.360 What can be more difficult to see is that these products of human ingenuity would be impossible

01:00:44.900 without scientists dissecting the natural world first. Chimeric T-cells cannot exist without

01:00:51.680 understanding fundamental T-cell biology first. Checkpoint and kinase inhibitors can't be

01:00:57.860 developed if we don't know the checkpoints and kinases worth inhibiting. Monoclonal antibodies

01:01:03.580 make no sense without first understanding natural polyclonal antibody expression.

01:01:09.360 and you get the point. While it's true that many of these basic questions begin in obviously

01:01:15.720 disease-relevant systems, it still is in service of this thesis. The basic science,

01:01:22.280 the foundational base of knowledge, is an irrevocably necessary part of the process

01:01:28.540 of curing human disease. Someone, hundreds of thousands of someones really, had to be curious

01:01:35.400 about how the natural world worked at its fundamental level.

01:01:40.040 Only then, with the knowledge they gained,

01:01:42.360 can we effectively supplement our labs and our bodies

01:01:46.320 with designer molecules of our own creation.

01:01:49.720 What I hope you'll leave with today is an appreciation for this idea.

01:01:53.540 The drugs in your medicine cabinet did not just come from a clean idea in a lab.

01:01:59.100 They started with a million small acts of curiosity about how the world works.

01:02:04.500 And the lesson in humility here for me, the thing I genuinely take from all of this is that we should be much more generous as a culture and as a funding apparatus with the people who are out there asking those small, strange, sometimes apparently irrelevant questions because we are demonstrably terrible at predicting which one of them is going to change everything.

01:02:27.840 We need people who translate curiosity into human outcomes, but I can't help but be struck by the

01:02:35.520 fact that the apparently impractical question is sometimes the one that matters most.

01:02:42.860 Thank you for listening to this week's episode of The Drive. Head over to peteratiamd.com

01:02:49.240 forward slash show notes if you want to dig deeper into this episode. You can also find me

01:02:55.260 on YouTube, Instagram, and Twitter, all with the handle Peter AtiyahMD. You can also leave us

01:03:00.860 review on Apple podcasts or whatever podcast player you use. This podcast is for general

01:03:06.880 informational purposes only and does not constitute the practice of medicine, nursing,

01:03:10.900 or other professional healthcare services, including the giving of medical advice.

01:03:15.580 No doctor patient relationship is formed. The use of this information and the materials linked to

01:03:21.240 this podcast is at the user's own risk. The content on this podcast is not intended to be

01:03:26.680 a substitute for professional medical advice, diagnosis, or treatment. Users should not disregard

01:03:31.920 or delay in obtaining medical advice from any medical condition they have, and they should

01:03:36.180 seek the assistance of their healthcare professionals for any such conditions. Finally, I take all

01:03:42.120 conflicts of interest very seriously. For all of my disclosures and the companies I invest in

01:03:47.120 or advise, please visit peteratiamd.com forward slash about where I keep an up-to-date and active

01:03:54.200 list of all disclosures.

The Peter Attia Drive - July 27, 2026

#401 ‒ How curiosity transforms medicine： extraordinary discoveries that changed modern healthcare

Episode Stats

Transcript