Idea Man

by Paul Allen

  The neuroscience community initially greeted our tool with some skepticism. The Allen Institute was the new kid on the block, funded by a technologist with no track record in the field. Our industrial-scale approach was unorthodox, and it took a while for people to believe that the data was really free, no strings attached.

  Less than two years after our launch, in September 2006, the full set of data for the mouse brain atlas was released on schedule, three years after we’d begun. With our first database complete, we made public these findings:

  The brain contains more genetic activity than we had thought. Scientists had previously estimated that two thirds of the genome was expressed somewhere in the brain. The Allen Brain Atlas shows that the proportion is closer to 80 percent, which helps explain why drug therapies designed for other organs often have side effects.

  Genes are expressed in distinct areas. Most of them are switched on in very specific subsets of cells or particular regions of the brain. This discovery has unveiled how biochemistry varies in different parts of the brain and how it relates to their specialized functions. By understanding these variations, scientists will be better equipped to modulate biochemical activity in diseased brain structures.

  Previous brain maps were sometimes inaccurate or incomplete, even on a gross anatomical level. As they defined gene expression patterns, our scientists came across previously unnoticed structural subdivisions. These findings have refined the understanding of how the brain is partitioned, a key to better diagnoses and therapies.

  After announcing the completion of the mouse brain atlas in Washington, D.C., I met Francis Collins, the former head of the Human Genome Project and now director of the National Institutes of Health. Early on, we’d felt a certain tension radiating from his agency; some at the NIH may have viewed our institute as a competitor. But Collins congratulated me wholeheartedly that day, and our institutional relationship has grown ever since. (In 2009, the Allen Institute won a GO Grant through the NIH as part of President Obama’s stimulus package.)

  Over time, any resistance to our work has dissolved. Susumu Tonegawa, a Nobel laureate and director of MIT’s Picower Center for Learning and Memory, called the mouse brain atlas “a breakthrough in neuroscience. It’s a new, extremely powerful approach to try to understand the brain. I would say it’s revolutionary.” Time said it was “a go-to source for researchers studying everything from multiple sclerosis to brain tumors.” In 2007, after Allan Jones coauthored a paper on the atlas in Nature, user hits on the Web site rose to record levels and are now up to a thousand visitors a day.

  WHEN I TRAVEL to scientific institutions, I’m delighted to hear stories of how our atlas has helped the field. “We use it every day,” a Stanford neurology professor told the Associated Press after a brief Web site glitch sent worried graduate students pouring into his office. “We can’t imagine life without this tool anymore.” Like the genome sequence, the brain atlas can save grad students and postdocs years of grind-it-out preliminaries. Researchers can track expression patterns for their gene of interest from our database, and that becomes their starting point. It’s like handing prospectors a map of a region’s diamond reserves. They can concentrate on the digging, knowing they’ve been directed toward something of value.

  One of the livelier topics at our charrette was breadth versus depth. Should we map the entire brain or focus on a single region? The wisdom of a whole-brain atlas is now clear. A Harvard researcher found a receptor gene expressed in the hypothalamus, in one of the few neurons in the brain linked to obesity; the atlas has accelerated his quest for a safe and effective drug therapy for appetite control. At the Seattle Swedish Neuroscience Institute at Swedish Medical Center, another researcher uses our data to zero in on genes with abnormal activity levels in glioblastomas, a lethal form of cerebral tumor. We’ve heard similar stories from researchers on Alzheimer’s, epilepsy, Down syndrome, and just about any other process or disease associated with the brain.

  I’m especially excited about the institute’s role in what may be a landmark study on the origins of autism, the spectrum of brain disorders that impairs a person’s ability to communicate, express emotion, and form social bonds. The project began in 2008, when Autism Speaks funded an effort led by Eric Courchesne of the Autism Center of Excellence at the University of California–San Diego. Courchesne had already established that autism was characterized by excessive brain growth in infants and toddlers, notably in the cerebellum and frontal lobes, but his studies had been limited by low-resolution imaging. Meanwhile, other studies had identified dozens of suspect genes but couldn’t tell where they were located or what they might be doing. Courchesne wanted to pinpoint both the genes and their locations to understand what might be setting off the disorder on a molecular level. Fortunately, he had rare postmortem brain tissue from both autistic and normal children. (Previous autism studies had mostly used adult brains, a major drawback for work on a developmental disorder.)

  That was where we came in. Our high-resolution techniques enabled us to look more deeply into cellular structure, and to focus in particular on genes that are normally expressed in cells in specific cortical layers. We could then tell if those cells were in the right places in autistic children or not. In essence, we could trace autism’s fingerprint.

  We began by sectioning brain tissue from both the autistic and control cases. In each, we explored a part of the frontal lobe tied to attention, working memory, and “theory of mind”—the ability to understand that other people have their own perspectives on the surrounding environment. Our goal was to find out whether this area of the brain was organized differently in autistic children.

  Using our catalog from the mouse brain atlas and the narrower Allen Human Cortex Study, alongside data from two normal control brains, we identified about twenty genes that were strongly and consistently expressed in the target area. Of those twenty, five had already been implicated in autism. We then studied more than two thousand slides to compare those genes’ expression in the autistic and normal brains. We expected to find abnormalities throughout the subregion, and we did. But we were surprised by the form they took: small, self-contained areas that were unusually dense with neurons yet showed a sharp decrease in the expression of most of the target genes. These pathological patches, as our scientists called them, existed in all of the autistic subjects. The brain tissue surrounding them appeared completely normal.

  Here was powerful evidence that autism might be a focal disorder in self-contained local areas in the brain. Most of the pathological patches were measured in millimeters and were easy to overlook in the tight mesh of neurons unless you examined one layer of cortex at a time. Decades of experiments with lower-resolution MRIs had failed to detect them.

  The UCSD–Allen Institute study represents a new and powerful way of doing large-scale neuropathological research at the molecular level. Beyond confirming that vaccines cannot possibly cause autism, it reveals clues that may clarify the disorder’s developmental origins and help explain why children vary across its spectrum. By understanding the cellular basis for autism, scientists may be able to devise new interventions, from early diagnostic testing to drug and other therapies. It now even seems possible that we might establish the root causes of autism, along with schizophrenia and other diseases, within our lifetimes.

  OUR AMBITIONS HAVE continued to grow. The Allen Human Brain Atlas, a four-year project scheduled for completion in 2012, represents a leap in scale and complexity. (The human brain is two thousand times as large as its mouse counterpart; flatten out the human cortex, and it’s the size of a seventeen-inch pizza.) The challenges begin with finding suitable tissue. We need brains from “normal” adults between twenty and sixty-eight years old, with no local injuries, drug addictions, or history of neurological or psychiatric disease. (Suicides are ruled out by definition.) And because brain tissue deteriorates within twenty-four hours of death, it can be challenging to get it in time. Thanks to the institute’s rel
ationships with NIH-funded brain tissue repositories on both coasts, we have received three brains so far and hope to get ten to complete the initial atlas. That might seem like a small sample for building a reference brain map, but only a small percentage of human genes vary in their pattern of expression across individuals.

  Given the amount of staining and scanning involved, it would be impractical to submit the entire human brain to analysis. Following the recommendation of our advisory council, we compromised. As a first step, we’re building a comprehensive 3-D atlas that will cover all expressed genes in all areas and offer something for every specialty. Because this first 3-D cut can’t get down to the cellular level, we’ll also provide a finer-resolution database for up to five hundred genes of especially high value to researchers in each of the major brain structures. Together, these two approaches will furnish unparalleled information about the normal human brain.

  Our second game-changing project is the Allen Mouse Brain Connectivity Atlas. At our charrette, Richard Axel pointed out that human behavior is primarily controlled not just by the expression of individual genes, but even more so by the physical and biochemical pathways that excite or inhibit billions of interdependent neurons. Most current research in this area is limited to efforts to define cell-to-cell or region-to-region connections. Our goal is to tackle the brain as a whole and to illustrate in detail how neurons are wired throughout.

  A comprehensive brain circuit map demands new techniques for tracing connections, and the complete data set could run as large as several quadrillions of bytes. But if we succeed, this kind of diagram could dramatically expand our knowledge of how nerve cell communications are altered by disease and how new therapies might most effectively intervene.

  DURING MY FIRST flush of excitement over the mouse brain atlas, I met Eric Kandel, the Columbia University neuroscientist who won the Nobel Prize in Physiology or Medicine for his work on memory storage in neurons. I told him, “We’re going to know so much more about the brain in the next ten years.”

  Dr. Kandel gently applied the brakes. “I’ve been working in this field for fifty years,” he said, “and not in my lifetime—and probably not in yours—will we understand the brain.”

  I was reminded of a question I’d put to the assembled luminaries at our charrette: “How many Nobel Prizes will need to be won in neuroscience before we really know how the brain works?” Their responses ranged between twenty-five and fifty.

  That’s a long, long way from here. In the meantime, I’m confident that our atlases will help those future laureates and speed them on their way.

  CHAPTER 22

  ADVENTURE

  I wasn’t raised as an adventurer. As a child, I traveled through books, the way my mother did. The piles of National Geographic in our basement depicted the larger world out there, but I didn’t envision myself as a globetrotter. Then, as a young man at Microsoft, I simply lacked the time to explore. All that changed when I became ill at twenty-nine. I started scuba diving in Hawaii; I came to love France and its culture and cuisine, so different from what I’d known. Still, the last thing I thought I’d ever own was a yacht. Here was my image of boating: a society of snobs who drank Scotch and smoked cigars and wore double-breasted blazers and captain’s hats. I wanted nothing to do with it.

  But friends kept telling me about the great trips they’d had to Alaska, and that the only way to do it was to charter a boat. In 1992, I rented an eighty-five-footer and took my family up the Inside Passage. We saw a whale swim underneath us, and many others spouting. We dined on fresh spot prawns bought from passing fishermen. In Anan Bay, we watched bears gorge on salmon swimming upstream, so numerous that (as the Indian saying went) you could have walked across the water on their backs.

  I had the time of my life. A boat seemed like the best possible way to share my budding passion for exploration, not to mention a terrific platform for my newfound passion for scuba diving. The following year, I was able to buy Charade, 150 feet long and five hundred tons, with a crew of ten. It had five staterooms and a Jacuzzi up top. I thought, My gosh, I’m buying more boat than I’ll need.

  Fast forward to a few years later, when my captain said, “What’s your ultimate boat, Paul?” I told him that I’d been absorbed by the undersea world ever since my parents took me to see Jacques Cousteau’s The Silent World, one of the first documentaries with underwater color cinematography, including shots from a twoman submarine. I said that I’d love to have my own sub to take my explorations literally down to the next level. While I wasn’t after size for its own sake, a bigger boat could accommodate more of my friends on our far-flung journeys. I also wanted to upgrade my onboard recording studio with a full digital console. Dave Stewart had an idea for a shipboard concert stage, with audience seating on the aft deck.

  That’s how Octopus was born, in the spirit of Cousteau’s underwater adventures. I went to Espen Øino, the naval architects based in Monaco, and they created a two-foot model that looked reasonable. Then the work started. It took a full year for more than a hundred draftspeople to design Octopus, and three years more for two companies to construct it. Midway through the process, the prime contractor invited me to their shipyard in Kiel, Germany, to show me how they built submarines for the German and Turkish navies. One of them had a torpedo, which piqued my interest. At the end of the tour, I asked them, deadpan, “Could I add a torpedo tube to my yacht?”

  The two engineers looked at each other, and you could see the deutsche mark signs going off in their heads. One of them said, “Ja, ja, we can add the tube, it’s possible to add the tube, ja.”

  I let the image hang in the air for a few moments before telling them I was joking. And the guy nodded his head and said, “Ja, we could have added the tube, but getting the license for the torpedoes, that would have been difficult.”

  I’ve owned a couple of other yachts, Meduse and Tatoosh, but I was stunned by the sheer size of Octopus when it was delivered in 2003. At 414 feet, it was a third longer than a football field, more than twenty yards wide, seven stories high. At the time, it was the fourth largest yacht in the world, with the top three built for heads of state. (As the yacht industry continues to extend the realm of the possible, Octopus has dropped in the rankings and is now ninth largest overall.) It had a full-time crew of more than fifty and the most advanced nautical technology. When I first stood on the bridge, I felt as though I was on a spaceship.

  It took me six months to get used to owning something of that scale. But over the years since, Octopus has realized every mission I had in mind for her. All my passions come together in one moveable feast: a basketball court, a movie theater, a swimming pool. The recording studio has ocean views in all directions and is painstakingly soundproofed from engine noise and vibration; it’s about as good musically as any in the world. Dave Stewart has recorded there, and so has Mick Jagger. U2 once previewed their latest album on board and played it so loud that they burned out the speakers. We’ve had too many phenomenal jams in that space to count. Each year we host a shipboard party during the Cannes Film Festival, and the studio becomes a bandstand.

  But while Octopus is ideal for get-togethers, musical and otherwise, my very favorite spot on the boat might be the most intimate one, a little aerie that seats a few people in total quiet at the very top. I’ve looked out over the Venice rooftops from there, and the factories and naval yards along the Huangpu River in Shanghai. With a top speed of twenty knots, Octopus has the range for long-haul explorations in the tradition of Cousteau’s Calypso, the minesweeper that carried the oceanographer’s crew of scientists and adventurers. It’s less a Bentley than a Range Rover.

  There’s a glass-bottomed room where you can watch the stingrays and jellyfish swim by when you’re at anchor, and a remote-controlled robot vehicle with a high-definition camera that can descend to three thousand meters. But no video can capture the immediacy of deep-sea exploring in a submarine twelve hundred feet below the surface. The sub holds eight pe
ople and launches from an internal lagoon, like a yellow underwater bus. For some reason, Pink Floyd sets an ideal mood as the surface recedes and the dark envelops us. For the next half hour, we’re going down.

  Our most memorable dive was onto the hangar deck of the USS Saratoga, the aircraft carrier that was sunk in the Bikini Atoll nuclear bomb tests in 1946. Another time we explored an ancient Roman wreck in the Tyrrhenian Sea. The wooden hull had long since rotted away, exposing its cargo of hundreds of graceful, long-necked amphorae, ceramic vases two thousand years old. They were just outside my porthole, close enough to touch.

  Octopus has a reinforced steel nose to push aside small pieces of ice, and in February 2007 we traveled to the Antarctic. We had a rough crossing; the big boat doesn’t roll, thanks to stabilizing wings, but it can pitch in a head sea when the wind is on your bow. As we went south from Ushuaia, Argentina, toward the Antarctic Circle, the iceberg traffic got heavier. Our captain ceded control to a specialist in polar navigation, the ice master. He sat on the bridge with binoculars, in a seeming Zen state, and calmly intoned, “Bring the ship to 273, please. … Now bring the ship to 142.” He knew how to estimate the size and shape of the ice masses beneath the water, which were eight or nine times larger than what we could see above it.

  It was near the end of summer, with the days still tolerably warm (often in the forties) and an endless twilight fading to a night about four hours long. Antarctica is a monstrously beautiful landscape, dead quiet whenever the wind stops blowing. It’s a vast white canvas on which nothing has been written, except for chunks of ice a vivid blue where glacial pressure has squeezed out the air bubbles. You can helicopter to a mountaintop, seven thousand feet above sea level, and see fifty miles in every direction—with no sign of a living soul.