by Ryan Bradley
A critical appreciation
A car is better even than a garage, that ur-Californian invention space, birthplace of Google and Apple, Mattel and Maglite, Disney, HP, and the Predator drone. A garage is cliché, static, earthbound, and homely. But a car—it’s made for motion. And a car plying a winding two-lane highway, cutting north through rolling hills and oak valleys as it makes its way into the redwoods? And that the thing being invented wasn’t so much a thing as a concept, a process, a means for amplification, for copying millions of times over the blueprint of life? That such a thing was dreamed up upon the asphalt near the redwoods and the Pacific, while the buckeyes were in bloom—I will humbly submit that this particular story of the invention of the polymerase chain reaction, or PCR, by one strange, difficult, and maybe half-crazy future Nobel Prize winner piloting a little silver Honda one spring night may be the greatest, most Californian invention tale of them all. It’s an untidy story that echoes on into the future still, reshaping entire industries and creating new ones, and—this being a California story—film franchises, too. So let’s zoom in on that moment, that spring night in 1983 when Kary Mullis was driving fast and thinking expansively.
The drive north through Mendocino County, toward his cabin in the Anderson Valley, was quiet and still. The car hugged the turns and the air hung heavy, perfumed with the buckeye blooms. It was evening, and his girlfriend and coworker, Jennifer Barnett, was asleep in the passenger seat. Mullis’s thoughts were back at the lab in Emeryville, where he worked with oligotide for Cetus, a biotech company.
Oligotide are DNA and RNA strands snipped short and turned into bits of nucleic acid, standard size and useful for experiments. Mullis, a chemist, was part of the DNA synthesis group at Cetus. It was his responsibility to manufacture custom oligotide for company labs. Mullis was a bit of a loose cannon. A jerk by many accounts, including his own. Tom White, a department head at Cetus, recalled how Mullis often proposed wild ideas during the company’s scientific retreats, “some of which were flatly wrong.” Mullis, White recalled, “wasn’t really familiar with some of the most basic aspects of molecular biology.” And he was abrasive, combative, suspicious, violent, and held a grudge. He once threatened to show up at Cetus with a gun because he sensed that one coworker was making moves on another with whom he was romantically involved.
But Mullis could also be playful. That’s how he described his work: as “playing in the lab” or “playing with oligonucleotides.” He’d heat up and cool down the strands, watching as they denatured, like an egg in a frying pan, and renatured, a fried egg turned back into a translucent yolk surrounded by clear egg white. It was here, in the early 1980s, that Mullis began to ponder the possibility of automated DNA replication, the steps that would eventually lead him to PCR.
The essential problem with oligonucleotides and DNA in general was that there were limits to how many standardized DNA snips one could synthesize. Replication, or amplification, got around this problem by simply copying the single strand you wanted to work with. Of course, the fact that DNA is also the building block of life, a genetic fingerprint and blueprint, makes replication all the more powerful. As Mullis put it, “DNA molecules in our cells are our history.…They are the stuff of which our future will be crafted. All of the organs of all the plants and animals of Earth and organs that have never been in light of the moon or sun, will be ours to explore—to use and adapt to our needs. Our will be done on Earth as we sail off to the stars in heaven.”
Grandiose stuff, which he recounts in his book, Dancing Naked in the Mind Field, a collection of essays released in 1997 that covers wide-ranging subjects, from PCR to astrology, global warming to the O. J. Simpson trial. There on that winding road to Mendocino, Mullis’s moment of insight came after following his usual torrent of thoughts into an unexpected eddy.
He was working on a different problem, trying to identify a single nucleotide at a given position on a molecule of DNA—specifically, how to identify the base-pair mutation that causes sickle-cell anemia. To get to a single molecule of DNA you must first tease out a single strand. DNA is, after all, a double helix. You can separate the strands—a natural process, one that occurs during cell division—in a lab using heat. Scientists had been doing this since the 1950s, soon after they discovered the enzymes that repair and replicate DNA. Polymerase is one such enzyme. It copies DNA, but requires an extra strand of nucleic acid to do so. As in a printer (indeed, PCR is often referred to as “DNA photocopying”), the nucleic acid in this process is called a primer. In a lab, these primers must be acquired, harvested, and made ready.
That’s where Mullis and his oligotide came in. The oligotide was the primer. If he could more easily identify and snip and copy a strand of that nucleotide, he figured he could speed up his work and make his job easier. (It may be an underreported phenomenon: the root of genius being simply trying to reduce the drudgery of one’s work.)
The primers are made up of the four nucleotide base pairs you probably learned about in high school and may have since forgotten. But they’ll sound familiar: adenine, thymine, guanine, cytosine. ATGC. These base pairs provide the map to replication. They’re the blueprint that polymerase follows. They tell the polymerase when to start copying, and when to stop. They’re also useful for other things, such as identifying genetic mutations that can lead to disease, in this case, sickle-cell anemia.
At the heart of DNA sequencing, even in these early, primitive days, lies a beautiful truth: base pairs are complementary. That is, if you know one side of a base, you’ll know the other. For this reason, the standard process required only one oligonucleotide primer. But, Mullis thought, for this specific goal, what about two?
Driving fast on a dark, winding road, Mullis slipstreamed into the eddy. If he had two oligonucleotides going, what now? If they remained in the solution, the polymerase would copy them, too, following the roadmap set out by the target DNA. So, instead of having a single copy of DNA, you’d have two. Run it again and you’d have four, then eight, then sixteen, and before very long there’d be millions of copies from the original DNA sample.
Mullis had a background in computer programing and often found himself drawn to automating processes in the lab. More tinkering, but of a different kind. If it worked—and it was at this point nothing more than a fever dream down a dark and winding road—he could harness, amplify, and replicate DNA, which he described, in its natural state, as “a tractless coil, like an unwound and tangled audio tape on the floor of the car in the dark.”
Once Mullis arrived at his cabin in the redwoods he did not sleep. Instead, he drew “little diagrams on every horizontal surface that would take pen, pencil, or crayon.” He had with him a good bottle of a local Cabernet, and in the fuzzy morning light he “settled into a perplexed semi-consciousness.” The only thing that stirred his thoughts and troubled his mind was whether or not these chain reactions were already in use. Surely, he thought, he would have heard about it, “and so would everybody else, including Jennifer, who was presently sunning herself beside the pond, taking no interest in the explosions that were rocking my brain.”
At this point his own recollection flashes forward, sprinting past several years of development at Cetus, skipping right to the glory of 1993, and his Nobel Prize in chemistry. The story does not end there, however, for Mullis or PCR. PCR, in fact, was just getting started.
Just what is a polymerase chain reaction? And is it different from PCR? The latter stands for the former, but it also stands for something else, something more. The reaction itself is, well, a reaction: the constant building of genetic material from base pairs spurred on by polymerase, dreamed up by Mullis on that drive. But PCR, the way it’s incorporated into labs throughout the world today, is different.
The problem with defining PCR from its origins is that the process itself has changed considerably, and even calling it “a process” is slippery. Paul Rabinow, an anthropology professor at UC Berkeley and the author of Making PCR, doesn’t like describing PCR as process. Same goes for “technique,” Rabinow writes, because doing so fixes it to a point, which serves to “eliminate the history of PCR’s invention.” Maybe, he suggests, it’s a concept. But that, too, incorrectly places too much of the credit with Mullis. There were a dozen others, at Cetus and elsewhere, without whom PCR would not have existed, or would have remained only in Mullis’s mind and in the scribbled notes in his cabin. Or, even more likely, someone else would have dreamed it up. Maybe it was in the air, like so much sweet buckeye scent.
Rabinow proposes that we think of PCR as an “experimental system.” And it is useful to think about PCR holistically, not as a single thing, but many things. It’s not one technique but many, just as it’s several concepts, with many inventors, arriving out of an entire milieu unique to California, and the biotechnology industry, at that particular moment in time, where liberal thought met serious capital and eddies of thought could turn into a series of innovations that could revolutionize all avenues of science, from biology to genetics to ecology to conservation work, but also medical science and criminal justice. Viewed in this light, PCR isn’t something created, but something inevitable that emerged in this place, at this time, much like the chain reaction itself.
Soon after Mullis’s initial insight, only months into the testing that would, over the years, develop into PCR, Mullis was at a dinner party at his best friend’s house. The chemist Albert Hofmann was there. Hofmann had invented LSD in Basel, Switzerland, in 1943, only he didn’t quite realize what he had done. It was only after decades, as LSD took on a life of its own in laboratories and beyond, that it dawned on Hofmann just what had happened—how it had happened to him. “Kind of like PCR,” Mullis recalls in Dancing Naked.
Mullis believes in fate and star signs. After listing the exact time, date and place of his birth (“I was born at 17:58 Greenwich Mean time on December 28, 1944, in Lenoir, North Carolina.”), he concludes, “You can find out more about me from that than you can from reading this book.” Later he writes that the reason he became a biochemist had to do with Mercury and Mars being in conjunction in Sagittarius: “I was not going to specialize in something well-defined and manageable.” Mullis is more than a bit nuts. He does not believe it’s possible for humans to cause the planet to overheat, or create a hole in Earth’s ozone layer. He believes the notion that our emissions are causing the temperature of the planet to rise is “about as ridiculous as blaming the Magdalenian paintings for the last ice age.” He also does not think AIDS is related to HIV.
What Tom White said about him held true. He was unafraid to very publicly, loudly, rudely hold forth with definitive conclusions based on no evidence. Not a trait associated with good science, or with scientists.
Mullis also dropped a lot of acid. During another dinner, a friend named Brad gave him “what was called a double-domed 1000 microgram” of LSD. When it kicked in Mullis started laughing, got up from the table, and realized on his way to the couch “that everything I knew was based on a false premise. I fell down through the couch into another world.” Brad put “Mysterious Mountain” by Alan Hovhaness on the stereo and kept playing it, over and over. After some minutes or seconds or hours Mullis noticed that time did not extend smoothly, but that it was punctuated by moments. He fell into a crack between two moments and was gone. He lay on the couch for four hours. “My mind could see itself,” he recalled. Then he and Brad went for a car ride. He sat in the back and, going down Marin Avenue in Berkeley, felt very dizzy.
Another morning he woke from a bad trip huddled under his desk, unable to remember who he was, what he liked, what he did. He was terrified and sad. He looked out the window and saw children in the yard. One of the children was his own, but he didn’t know which. His wife at the time woke up. He looked at her in a way that spurred her to remind him that she was his wife. He looked harder, but still couldn’t remember. “I thought I loved books and music, but I couldn’t remember which books or what kind of music.” He had, he recalled, “annihilated” his personality. Slowly, as if from a distance, his life returned to him, “whole and undamaged.” He felt he now knew what it was be meaningless. Time marched un-smoothly on; the system had a life of its own, outside of him, that cabin, that car, the cracks, his life, the ride, the drive.
By the end of the 1980s, PCR had become a standard system and approach in most biotech labs. “PCR is doing for genetic material what the invention of the printing press did for written material—making copying easy, inexpensive, and accessible,” read a 1990 article in Breakthroughs in Bioscience, a publication published by the Federation of American Societies for Experimental Biology. Many labs were investigating ways to use PCR for detecting infectious diseases or, like Mullis’s initial question, to amplify and find potentially harmful mutations in genes. At Genentech, a South San Francisco–based biotech company, Stephen Bustin was on a team studying the effectiveness of HIV vaccines. The team’s leader, Jack Nunberg, had come from Cetus, joining Genentech through what would, decades later, become its parent company, Roche Molecular Diagnostics, which bought PCR from Cetus. Bustin detailed their work in his 2009 book The PCR Revolution.
The group’s goal was twofold. First, they needed to come up with an alternate method of amplifying and duplicating DNA. As PCR was becoming increasingly common, companies’ patent portfolios around the system were swelling. The second, related goal was to see if real-time PCR was possible—that is, a method of PCR that was novel, faster, and could be monitored in real time. The team used a probe for the oligonucleotide to enter the polymerase, and they used a florescent dye to watch the process. Under ultraviolet light, the fluorescing oligonucleotides transferred energy, which sped up the process. Then they added a closed-tube system and concentrated the formula. Before long, real-time PCR was at hand.
The Genentech team and others began using PCR for diagnostics, turning it toward ever smaller and more sensitive genetic material. A real-time PCR process was used, and FDA approved, for in vitro diagnostic assays. More than two decades later, in vitro PCR analysis would find another avenue for FDA approval when 23andMe, the Mountain View–based genetics-for-everyone-and-as-social-network company, was granted clearance to market an at-home test kit for Bloom syndrome, a recessive disorder. Real-time PCR was also, at Genentech and elsewhere, applied to early AIDS detection—for rooting out the virus’s DNA, rather than in the genetic material found in antibodies made against AIDS, which often doesn’t crop up until weeks or even months into an infection. Soon, PCR systems began to be used for early cancer detection. The amplification of cell DNA meant that scientists and doctors were able to see exactly when and where the cell replication cycle was occurring unchecked.
But the most Hollywood-ready uses of PCR had nothing to do with saving lives. No, catching crooks and cloning dinosaurs were what made PCR famous. The leap from DNA typing to genetic fingerprinting was a small one. Remember, the first step in the process was to amplify an already-small amount of genetic material. Finding that small amount, then identifying what made it unique and matching that to a suspect’s DNA, was a natural evolution from the biotech lab to the crime lab. Only, in the early days, throughout the 1980s and early 1990s, systems weren’t in place, the technology was young, and mistakes were rife. By the time CBS’s hit show CSI arrived, genetic fingerprinting had gone from a major plot point to a routine part of investigations.
Today, the most important use of genetic fingerprinting isn’t necessarily to catch criminals but to exonerate those wrongfully convicted, which it has done at least a hundred times in the past thirty years. In the last year alone, six inmates on death row were exonerated in part due to DNA evidence that didn’t match up with DNA found at the crime scene. Among these six was Henry Lee McCollum, whose name and story the Supreme Court Justice Stephen Breyer (a San Francisco native) recalled in his dissent in Glossip v. Gross, a dissent that questions the very constitutionality of the death penalty. A key line, and one that will echo into future legal battles, would not have been possible without PCR: “If McCollum had been executed earlier, he would not have lived to see the day when DNA evidence exonerated him and implicated another man.”
Jurassic Park wouldn’t exist without PCR, either. Michael Crichton, an MD, heard about how the system had been adapted and was being used by paleontologists to amplify trace amounts of ancient DNA, or aDNA. Amplify the DNA, then clone it—it’s just that simple, in fiction. The movie version swept PCR under the rug a bit. In the animated video explaining how, within this movie universe, it is possible to bring back dinosaurs, PCR is lumped in with a bunch of other “sophisticated techniques” the Jurassic Park scientists applied to the amber-preserved, mosquito-ingested dino-blood.
The brilliance of Crichton’s fiction is that it hews just close enough to fact that there’s more than a whiff of plausibility. Indeed, paleontologists then were analyzing the DNA of very old species, just not that old. At the time, they were examining the 140-year-old skin of the last quagga, an African horse, to prove it the zebra’s closest relative. A similar test proved the extinct and flightless New Zealand moa unrelated to the also-flightless kiwi. And paleontologists were beginning to use such aDNA analysis on a thirteen-thousand-year-old giant sloth fossil. Suddenly, amid a flurry of interest in Crichton’s book, the aDNA tests turned to even older samples: a twenty-million-year-old Miocene plant leaf, fossilized. To get the DNA, what was left of it, the rocks were ground up, the fossil destroyed. But a fragment was found—a 770-base pair from an extinct tree. Another team outlined how they had extracted DNA from a termite that was twenty-five to thirty million years old, fossilized in Oligo-Miocene amber. A chrysomelid beetle in Dominican amber was also found, cracked open, sequenced, and destroyed. And a priceless, 125-million-year-old weevil specimen encased in Lebanese amber was taken apart, too, because of PCR.
Tomas Lindahl, a Swedish scientist and cancer researcher, published a paper in Nature critiquing the technique and raising serious doubts about how DNA could withstand so much pressure, exposure to water, and oxidation over such a long period of time. Under even ideal conditions, DNA rarely survives longer than one hundred thousand years. Lindahl was right, but ignored. Two months after his critique, Nature published a paper on the weevil extraction and DNA testing, which was later found to be contaminated. The paper had to be withdrawn, a fact that attracted much less attention than its initial publication—on the same day Jurassic Park first hit the big screen.
Still, for zoologists and ecologists, PCR has transformed how we might monitor and understand our world. Detailed maps of seed dispersals and migratory patterns that reach back through time would not be possible without the ability to test small samples scooped up in the field. Apply the same logic that one would at a crime scene to a rarely seen animal and it’s far easier to know what it is, and where it’s going, picking up a bit of hair or urine or feces without disturbing the creature.
Of course, what PCR has really helped us understand is ourselves: where we’ve gone, how we spread across the planet, where we came from. The greatest, most expensive, and largest single scientific undertaking, which would have been impossible without PCR, is without a doubt the Human Genome Project—spanning decades, costing trillions, employing thousands. It was the moonshot of our biotech age.
But what have we learned since sequencing a full human genome? Now that sequencing one’s DNA costs as little as $1,000 and takes only a few hours, what now? What next?
There is far more work to be done, of course. More research needed, as they say. The Human Genome Project created more questions than answers. For starters, our DNA has far fewer genes than expected: not quite twenty thousand. And many seem to turn on, then off, or not function at all. A lot of what was in our DNA appeared, to a cadre of scientists, to be nothing more than junk. The geneticist Steve Jones is fond of saying of his field, “The more we learn, the less we understand.” That car ride through the country that led to PCR caused a ripple that grew into to a wave that hasn’t yet crested. Or as Mullis, the surfer, put it: “We are just here for the ride. And the ride is not smooth. It never has been smooth.”
All images courtesy of Dare DNA www.etsy/shop/DareDNA.
Rabinow, Paul, Making PCR: A Story of Biotechnology (Chicago: University of Chicago Press, 1997).
Mullis, Kary, Dancing Naked in the Mind Field (Darby, Pennsylvania: Diane Polishing Company, 1998).
Mullis, Kary, “The Unusual Origin of the Polymerase Chain Reaction,” Scientific American (April 1990): 56.
Bustin, Stephen A., editor, The PCR Revolution: Basic Technologies and Applications (Cambridge: Cambridge University Press, 2014).
Glossip v. Gross, United States Supreme Court (Stephen Breyer, dissenting, 2015), 4–5.
Crichton, Michael, Jurassic Park (New York: Random House)
Jurassic Park, directed by Spielberg, Steven (1990, Universal Picture Studios).
Lindahl, Tomas. “Instability and decay of the primary structure of DNA,” Nature, (April 1993): 709–715.