by Chip Walter
The more scientists inspected the genomes of the animal world, the stranger their discoveries became. Another was an immense creature that the indigenous Inupiat people haul out of the Arctic Ocean onto the packed, hardtack beach of Barrow, Alaska, every fall and spring. Barrow (now known as Utqiaġvik) is the northernmost city in the United States: the last slip of the planet that North America has to offer before giving up nothing more than wind, snow, and the harsh gale-swept waters of the Beaufort Sea.
When John “Craig” George arrived in Barrow in 1977 to take up work as a young lab assistant and wildlife biologist for the Naval Arctic Research Laboratory (NARL), he told his boss he wasn’t sure he would last very long. It wasn’t that he minded shoveling animal dung, or keeping track of elk and bear and walrus out on the tundra. But where was the stately beauty of Alaska? Certainly not here!
Barrow, you see, was a little like the mole rat of Alaska. Even today, the place looks like some wind-whipped, gold rush settlement out of a 1950s B movie, an architect’s nightmare pocked with World War II-era Quonset huts and their drab descendants: ramshackle, prefab metal and pressed-wood buildings with an occasional geodesic dome here or there, each sitting on pylons hammered into the ice-hard permafrost to keep it from listing into the tundra when summer briefly makes its way to what the locals call “the top of the world.” Here, the streets don’t even require paving; they stay that cold.
Nevertheless, 40 years after George’s arrival, he had grown to love the place, and was still living there when he made a remarkable discovery about life span inside the body of the great creature in question: a bowhead whale.
Although most whaling has been outlawed worldwide, under a special relationship with the federal government and the Alaska Eskimo Whaling Commission (AEWC), the Inupiat people of Utqiaġvik are allowed to harvest bowheads for a few weeks a year. The deal was arranged because the whales are central to the native culture in the way buffalo once were for the tribes of the American Plains and West. Only so many whales can be harvested, and their meat, under the agreement, must be freely distributed to the natives of the village. Nothing can be sold either at a restaurant or store.
George was a great fan of bowheads, and knew just about everything about them. They were immense, among the largest animals on Earth, often as long as an 18-wheeler and weighing 45 to 65 tons. Largely because of their gargantuan dimensions, they had been revered by the Inupiats for more than 2,500 years. During the vast majority of those years, Inupiat whalers hunted the immense animals using long wooden harpoons, hafted with broad, flint blades. But beginning around 1890, those harpoons were replaced when Yankee whalers from New England started heading north. The Inupiat worked with the white whalers and took up the new technology that came with them: metal harpoons called Temple toggles, which were catapulted from a kind of cannon on deck.
One spring day in 1992, George arrived with one of his colleagues, Billy Adams, just as one of the whales was being hauled on shore. The whale was mature: 50, maybe 60 years old. Everyone figured he was getting pretty long in the tooth, because the consensus then was that all whales lived about the same length of time: no more than 75 years.
As the two biologists were looking the whale over, Adams noticed a divot in its back. It appeared strange enough that George asked if he could see that section of the whale as they were preparing to butcher and distribute it to the villagers. That was part of George’s job at NARL, keeping tabs on all animals in the region.
So George cuts his way into the 18 inches or so of blubber to see if he can get his hand into the divot, and he feels a harpoon. Normally, that wouldn’t be very unusual. Harpoons were sometimes found in the great animals. But when he pulled the harpoon out, the hair on George’s neck stood on end. In his hand, he held a large blade of slate that had been carefully hafted into the end of a long pike of bone or ivory: beautifully crafted and shaved to a razor’s edge in the shape of a triangle five inches long and four inches wide. Stone Age technology. Nothing like it had been created since the Yankee whalers had arrived with their Temple toggles 120 years earlier!
George knew enough about whale craft and history to understand what this meant. Some Inupiat hunter, coming in for the kill, riding in a whaling boat made of wood and sealskin no larger than the whale itself, had stood up in the frigid, open sea and thrust the harpoon into its back more than 120 years ago. That was when Ulysses S. Grant was president, and Jules Verne wrote Around the World in Eighty Days!
It wasn’t until 2015 that scientists got around to sequencing the genome of bowhead whales. The news revealed what everyone had come to expect: Bowheads were by far the longest lived mammals on Earth. It wasn’t unusual for them to survive 215 years! Right now, some of those out there beyond Utqiaġvik might have been swimming the Beaufort Sea when Napoleon was marching to Moscow.
Other animals, like the quahog clam (the one used in clam chowder), were known to live 400 years. And the well-known ancient tortoises of the Galápagos Islands poked along for upwards of 100 years. And in 2016, scientists had confirmed that some Greenland sharks might swim the cold waters of the North Atlantic for 500 years. But those were fish and amphibians and mollusks. Bowhead whales were mammals—large and complex—and they routinely lived almost three times longer than your average, healthy human. Once again, evolution had set a different clock, a different beta. But why? And how?
Over the years, Craig George developed some theories. For one thing, new studies revealed that bowheads didn’t even begin to mate until they were 25 years old. There was a direct correlation between the time a mammal grew sexually active and how long it lived. Another factor was that like humans, bowheads usually had only one calf at a time, and took 14 months to gestate. And like humans, they required a lot of care after birth. So, over time, evolution tended to select bowheads that would live longer: the better to ensure the species’ offspring survived and had time to continue breeding.
And then there was that old issue related to calorie restriction. Bowheads were the only whales in the world that lived in cold water year-round. While all of the world’s other whales swam to warm waters to bear their children, bowheads spent every last minute of their existence among the frigid seas of the Arctic. That made food scarce, which meant evolution would select for bowheads that could survive longer without food.
This, in turn, would slow the rate of reproduction. Research had shown bowheads could go 18 months without eating a single plankton or shrimp, and still go strong. (This was one reason their blubber was, by far, the thickest among all mammals.) The animals simply needed longer life spans and reproductive cycles to survive, and the evolutionary lottery had allowed them to manage it.
Rabbits and rats, on the other hand, are good examples of species that live in a world where they rapidly proliferate in climates where food is plentiful. They have no reason to survive especially long. They can procreate like, well, bunnies, and then pass on to make room for their offspring. In more ways than one, this paralleled the Daf-2 gene mutations in worms, fruit flies, and mice, which also slowed their aging, or the effect of the drug metformin as it changes the insulin pathways in humans. On a molecular/cellular level, all of these actions cause the creatures to react as though they are living in an environment where food is scarce.
George had no particular opinions on that insight, but he knew one thing: Bowheads were astonishingly tough—strong and healthy to the end. The powers of their DNA repair were stunning. Throughout their lives, they seemed to continually repair the huge numbers of cells needed to keep them alive and swimming millions of miles. Never once, after a good thousand of his investigations over the years, had George found evidence of any cancer or dementia in a bowhead.
Whatever was going on, all the evidence indicated that evolution could somehow change key genetic pathways that lengthened the life of a species. And that begged the bigger question: If evolution could find a way to do this, could science do it too?
20 | THE STARS WERE REMARK
ABLE
The stars were remarkable! It was one of those nights when the whole spine of the Milky Way rises up like diamond dust and you can feel the great wheel of the cosmos hold you in its galactic hand.
Nevertheless, Riccardo Sabatini’s girlfriend was not happy. Here they were in one of the most beautiful places in the world: Tomales Bay, overlooking the Pacific Ocean up past San Francisco. And in a hot tub, no less, alcohol within easy reach…and what’s he doing? Checking his phone.
But Sabatini couldn’t help himself; he was a nervous wreck because he was in charge of the “Face Project”—the one Craig Venter put him on at Human Longevity, Inc., almost the moment the company had hired its first computer science team. Now, Sabatini’s group was on the cusp of knowing whether the whole manifold undertaking was going to succeed or go down like a bad WWE match. Very soon, the project’s first human face would reveal itself, downloaded in a great burst of digits from the Cloud, a little like stardust. Or at least that was what Sabatini was hoping.
He knew it was silly to be gawking at a smartphone, rather than beckoning the stars with the love of his life. But the Face Project was a big deal—a remarkable challenge. Venter had put the team—specifically Sabatini and Franz Ochs, who had created the first version of Google Translate, plus a whole passel of other brilliant Silicon Valley computer scientists—to a test: Predict what someone’s face would look like based on their DNA, and only their DNA. No pictures, no video. Nothing but the information in their genes.14
When Venter first laid out the idea, HLI’s software wizards told him it was insane. Couldn’t be done. At least not in any reasonable length of time—maybe 10 years. Venter looked at them all like a saddened rector who had caught the altar boys sipping the church wine. “Of course you can do this,” he said. “If I were doing it, I could probably get it done in a couple of weeks.”
Venter knew that wasn’t true, of course, not if he were given a strapping dose of Kurzweilian augmented intelligence. But that was one of Venter’s ways of motivating the teams that worked with him, and they knew it. The whole crew was a bunch of type A, utterly geeked-out programmers; otherwise, Venter wouldn’t have hired them. So all he had to do was toss a challenge their way, tell them they couldn’t do it, and wait for them to prove him wrong.
Now, in October 2015, they had run the numbers, tweaked the great batches of machine-learning algorithms, thrashed the Cloud with data, and were furiously crunching the artificially intelligent numbers, waiting for the first models to emerge. The immensity of the data was shocking. The team had spent months sequencing thousands of genomes at 300 gigabytes each, plus all of the medical, biological, and historical information on each patient: age, weight, skin color, eyes, personal medical histories. Next, they applied boatloads of computer algorithms to shuffle the trillions of possibilities into facial models that made sense. Word at one point was that the project had so burdened Amazon’s cloud computing systems that they crashed. “We broke the Cloud!” Venter told me. But the really big question was: Would the idea work?
And that was why Sabatini was eyeing his phone. Then, at last, the answer came. There, with the Big Dipper twinkling brightly above him, Sabatini saw…success! The team had created something most people in the genetic field had believed utterly impossible: a full-on human face that looked almost precisely like images of the originals, each one made possible by extracting nothing more than specific bits of a particular human’s deoxyribonucleic acid.
It was a beautiful thing.
Of course, if you simply looked at the science, it all made perfect sense; after all, the genetic artifacts of a face had to be buried in the genome somewhere. Where else would they be? It was the genome that made one’s face possible, did it not? But it was one thing to know there was a needle in the haystack, and quite another to actually find it. Sabatini and his wizards had done that. Did someone say this would take 10 years? The team did it in eight months! They had extracted precisely the right nuggets of data from the double helix needed to form the picture of a real and accurate face. And they had done it by turning molecules into digits.
That was really the purpose of Venter’s Face Project: to provide a test case that revealed how artificial intelligence—or, as some put it, machine learning—could extract the revelations hidden within every human genome. If one could accurately predict what a person looked like based on his or her genes, it demonstrated that in time all of that other information could be extracted, too: every molecule that constituted a “self,” including how you might die.
21 | HERE BE DRAGONS
By the time Human Longevity, Inc.’s Face Project was under way, the company’s business plan had been well set. The idea wasn’t to simply aggregate genetic information; these days, anyone could do that. No, HLI would gather genetic information with an accuracy and depth that would reveal nothing less than how the great gears of evolution constructed both the human species and the individual known as “you.”
The Face Project was only one of many undertakings that Venter had simmering throughout the corridors at HLI. Also included were a series of research collaborations with Genentech, the J. Craig Venter Institute (JCVI), and King’s College London. The King’s College connection allowed HLI to get its hands on the genomes and microbiomes of 2,000 twins. Deals were made with cancer institutes and insurance companies from the United Kingdom to South Africa. Within two years, the company had grown to 200 employees in San Diego and Silicon Valley. Venter hired an assortment of top-rung executives, including Brad Perkins, former chief strategy and innovations officer at the Centers for Disease Control and Prevention as chief medical officer; Bill Biggs to handle genomic sequencing; and, eventually, Ken Bloom as president.
Biggs had been setting up sequencing labs for 20 years, basically since they existed. His name befitted him. He was a big man with graying blond hair that flopped over his high forehead, partial to loose slacks and Hawaiian shirts. Bloom had originally come into HLI to head its immunotherapy division, but took on the role of president in early 2016. He was an affable, articulate man who had spent years in the academic world, then created a health care company later bought by GE Healthcare, where he had spent several years as chief medical officer.
Biggs’s work would be especially important, because Venter had immediately bought two Illumina HiSeq X Ten Sequencing Systems (the grandchildren of Hunkapiller’s DNA sequencers from the HGP days), with three more in the pipeline. Venter’s goal was to initially sequence 2,000 human genomes a month, but accelerate to 40,000 by the end of the first year. By 2016, Biggs had installed 26 Illumina sequencers with the company knocking out over 700 human genomes a week: 60 terabytes of raw information and 240 terabytes after the information was analyzed. That was the equivalent content contained in 3,120,000 full-length movies a year.
Venter eventually crammed the place so full of sequencers that Biggs had to start giving them names like Obi-Wan, Leia, and R2-D2. It turned out that the staff found it easier to track mythical names for the contraptions than use random numbers. Generating these many genomes a week was what Brad Perkins liked to call “getting to scale.”
Bloom agreed. By his reckoning, HLI would need to analyze a good one million integrated genomes before it could hope to gain a really solid idea of what any human double helix could reveal. All of this was entirely consistent with Venter’s goal for HLI, but different from Calico and the SENS Research Foundation. He wasn’t shooting to attain immortality or radically extend human life—certainly not in the ways Kurzweil and de Grey had been talking about. Venter’s focus was on “extending health span” (or, as some of the gerontology wonks liked to put it, “compressing morbidity”). The goal was to get the maximum number of years of life that evolution had worked out for Homo sapiens, and to make them good ones rather than the painful, long goodbyes that marked the ends of so many lives.
Not that he was opposed to radical life extension. If Art Levinson and the researchers over at Calico could manage to
create a pill that jumped human life span to a healthy few hundred years, fine by him: He’d be the first to gulp it down. On the other hand, there might be consequences if everyone did that. Venter was halfway serious when he said men might have to be castrated to ensure long-lived humans didn’t cram the planet full of people.
Either way, there was plenty of work to be done. More deals and hardware followed at HLI: high-end imaging equipment, deeper machine learning expertise, more computing power, the sequencing of the microbiome, cancer cells, and tumors—whatever Venter could get his hands on to begin peeling away the obstreperous riddles the human genome held close. Then, he would apply all of that data to tackle whatever life-killing diseases he could before they took an early hold.
At its simplest, Venter saw the enterprise as a database company faced with solving the most complex translational challenge humans had ever looked in the eye. It really was like trying to decipher an incredibly complex foreign language. If such a mystery could be unmasked, the practice of medicine would change profoundly. Doctors would stop treating symptoms and start using patients’ personal genetic information to head off disease. Every human being might not live forever, but because their ailments would be treated before symptoms emerged, they would live far better, and, therefore, far longer.
And think what an improvement that would be! After all, nearly every cancer is curable at Stage I, often at Stage II. Early detection might be the simplest cancer treatment: far easier and less painful than radiation or chemotherapy or surgery. Or what if you knew in your 20s that you were genetically predisposed to heart disease? Maybe you would change your lifestyle. Or take some preventative drug. Or, very possibly, thanks to all of those new insights about the genome, use Crispr to rearrange a few of your genes to eliminate the problem entirely, before it became a problem. With technology like this, science could have saved the lives of people like John Venter, Fredric Kurzweil, and Sol Levinson.
