WHEN CHIP QUINN left Benzer’s Fly Room, he took the memory project with him. In a Fly Room of his own at Princeton University, he went on poisoning flies and running their children through his Teaching Machine. In this way he found more slow learners to keep company with the original dunce, including amnesiac, smellblind, turnip, and rutabaga. But Quinn was groping. Memory was a black box, and he did not know how to get inside it. As one of Benzer’s student’s students observed afterward, dunce, turnip, and cabbage all showed “essentially the same overt phenotype: stupidity.” Quinn could not distinguish whatever was wrong with dunce from whatever was wrong with rutabaga.
One day a grad student of Quinn’s, Ronald Booker, fixed a fly to the end of a wooden stick with a dab of wax. He tied a slipknot in a fine wire to make a noose and looped the wire over one of the fly’s legs. He cinched it tight by pulling on it with a forceps and snipped the wire so that one end dangled loose. Then he suspended the little fly over a pool of salt water so that the wire just brushed the surface. The fly waved and kicked its leg, and the wire lifted out of the water and dipped back in. Whenever the wire touched the water, the fly got an electric shock. Normally, Quinn’s wild-type flies were uncertain learners, as Benzer’s nemesis Jerry Hirsch had pointed out. But hanging over the electrified pool, better than nine out of ten wild-type flies learned to keep that leg up.
Then Booker tried pinching off the fly’s head with a hot forceps or cutting it off with a razor blade while the fly hung over the water. Not only did the fly survive decapitation, it learned to keep the wire out of the water even better than a fly with a head. Flies have nerves outside the brain, just as human beings do, and apparently those nerves were learning the lesson. Quinn and Booker could only assume that the flies learned better without their heads because they were less distracted.
However, dunce, turnip, and rutabaga flunked this leg-lifting test with or without their heads, just as they flunked all of Quinn’s other learning tests. Something was different in both their genes and their nervous systems. But to find out exactly what, Quinn wrote in 1981, “we will need more sophisticated tools than hot forceps and a razor blade.”
That year, Quinn took on a new postdoc, Tim Tully, who was probably as passionate about the study of memory as Quinn had ever been. When Tully was a boy, a day of childhood memories had been knocked out of him in a Christmas accident. Hitting a tree with his sled had produced a kind of hollow of amnesia; he had lost all memory of the hours preceding the accident. (A quick dunk in ice water will do the same thing to a fruit fly.) When he was in college, studying with Jerry Hirsch, Tully decided that finding the secret of memory was the only thing he wanted to do in his working life. And after studying the literature he decided that Benzer’s method of genetic dissection was a more promising tool than Hirsch’s method of breeding and crossbreeding flies. He told Hirsch that Benzer was discovering molecules; Hirsch was not. Tully often says that Hirsch regarded him as a reincarnation of Judas for going over to Benzer. “And in a way perhaps I was,” Tully says. “Although I never kissed him on the cheek.”
Tully joined Quinn at Princeton and took stock of Quinn’s operation. He was just as blunt with Quinn as he had been with Hirsch. In his view, Quinn’s problem was the inefficiency of his screen. Quinn’s headless fly had proven that flies are capable of much better performances than they ever achieved in Quinn’s Teaching Machine.
Tully has a knack for gizmos and gadgets. He designed a new Teaching Machine. He piped in the odors silently and smoothly on soft breezes so that all the flies would get the same doses of his perfumes. He lined the shock tubes in such a way that every fly would feel every jolt. He designed each part of the machine to run quietly and efficiently so that the flies would not be distracted from their lessons and so that virtually every one would get virtually the same lesson every time. He used a little plastic elevator to lift the flies gently from one stage of the experiment to the next. Designing and building this machine took Tully four years.
Human eyes can see dim red light, but fly eyes cannot. So when Tully was ready to run his new Teaching Machine, he turned out the lights and watched the flies by the glow of a low-watt darkroom safelight. In the dark and quiet of the new machine, fruit flies acted much calmer than they ever had in Quinn’s, and Tully was delighted to find that better than nine out of ten of them now learned their lessons. Tully had gotten rid of the static—which is what Quinn and Booker had done much more crudely with the hot forceps and the razor blade.
Now, because the flies in the new Teaching Machine were not distracted from their lessons, Tully’s experiments were not blurred by the random noise of good learners that failed to learn. And now Tully could discern varieties of stupidity, a wide spectrum of troubles in Quinn’s mutants. He discovered that rutabaga, for instance, is able to learn; it just forgets fast. In other words, rutabaga is a memory mutant but not a learning mutant. Other mutants have trouble learning, but they hold on to what they do learn. They are learning mutants but not memory mutants.
During those long early years of R and D, Tully had the same public relations problems that had always plagued Benzer’s school. Most biologists found it hard to take a fly in a Teaching Machine seriously, and Tully had trouble getting any funding for his project. When James Watson first got interested in Tully’s work, Watson’s advisers urged him to stay away. But Watson installed Tully in the laboratory next to his office at Cold Spring Harbor, and slowly Tully began dissecting the act of memory in his flies.
For years, Tully had been reading the papers of the neurobiologist Eric Kandel, who is now at Columbia University. Kandel was working on learning and memory in a giant snail, Aplysia. When Kandel tickled the siphon of the snail, it flinched and contracted its gill. But when Kandel tickled the snail ten or fifteen times, it flinched less and less. The snail was learning to ignore him. If he smacked the poor snail on the head, it suddenly became much more responsive again.
Kandel and his colleagues managed to trace the circuit of neurons that governs these simple reflexes. The team stuck a microelectrode into a single neuron and recorded the characteristic pattern of bursts of electricity, like a series of dots and dashes in Morse code, that made the gill flinch. In one experiment, Kandel sent precisely that same series of spikes into a motor neuron. The gill flinched. Kandel was learning the snail nerve’s language, the way King Solomon in the legend learned the languages of the beasts, fowl, and fishes.
From there Kandel played with some of the molecules that compose the message; this was the work that fascinated Tully. One nerve sends another nerve a message across the synapse between them. The message takes the form of a packet of molecules—or a burst of packets. When the snail gets knocked on the head, a message shoots down a nerve from its head and activates an enzyme called adenylate cyclase in the gill nerve. This enzyme helps the gill nerve make a second compound called cyclic adenosine monophosphate, or cAMP. The cAMP then cranks up the nerve cell’s sensitivity so that the next time the siphon is tickled, that cell will release more packets across the synapse. In this way, the memory of the blow to the head has left a trace in the snail: the snail will flinch harder the next time it is tickled in the siphon because it has just been struck. It is a simple case of learning and memory, reduced to a few molecules.
At Caltech, Benzer and his student Duncan Byers were following Kandel’s work too. Since cAMP matters so much in the learning and memory of the snail, they decided to take a second look at their fly mutant dunce. They were thrilled to discover that dunce makes a crippled form of the enzyme cAMP phosphodiesterase. Without that enzyme, dunce has problems metabolizing cAMP. So at this level, at least, the language of the fly and the snail seemed to be the same. Benzer and his students wondered if the enzyme cAMP might be part of a universal mechanism of learning and memory in the living world.
Meanwhile, to study the formation of long-term memory in his new and improved Teaching Machine, Tully gave a set of wild-type flies ten lessons one after the other, without a break between them, as if the flies were cramming for a final exam. He gave another set of wild-type flies ten lessons with rest periods in between, as if they were preparing for their exam by studying every other day. The class of crammers forgot everything fast; they made only short-term memories. But the class that took breaks still remembered the lessons after a week.
During the first half of the 1990s, Tully and a growing team of colleagues, postdocs, and graduate students defined step after step in the making of these lasting memories at the molecular level. They did the work by using the old memory mutants in ingenious new combinations in Tully’s Teaching Machine, along with some brand-new mutants that he and his lab discovered along the way, including latheo, named after Lethe, the river in Hades that caused all the souls in the underworld to forget their past, and linotte, named after the French expression tête de linotte, for which the English equivalent is “birdbrain.”
Kandel and his laboratory had discovered that cAMP turns on certain key genes involved in making permanent memories—but which genes, they did not know. Their work suggested that a certain protein—cAMP-responsive element-binding protein—might be involved in the cascade. The gene that makes this protein is known as creb. In 1994, one of Tully’s chief partners in this work, Jerry Yin, cloned one of the fly’s creb genes. The creb gene can make a protein in one of two alternate forms, and these two forms have opposite functions in the brain. One form activates certain genes; the other deactivates those same genes. In other words, one form is an on switch and the other form is an off switch.
When they found that switch, Yin, Tully, Quinn, and a few other collaborators began one of the most remarkable genetic engineering projects in the study of genes and behavior. They mixed a DNA transformation cocktail containing a creb gene: the on switch. They injected this cocktail into the eggs of wild-type flies. When they had made and bred these transgenic flies, they loaded a class of them into the Teaching Machine.
Each fly in that classroom carried an extra on switch. What is more, Tully and his team had mixed the DNA transformation cocktail in such a way that the extra on switch would flip on only when the fly was warm. This is one of the routine tricks of genetic engineering that are now possible in Fly Rooms; the gene involved is called a heat shock promoter, because heat will shock the gene into promoting the action of the other genes in the DNA transformation cocktail.
In a cool room these flies learned normally, as expected. Now came the test. Tully took a bottle of the flies and dunked it in warm water. Inside the brains of those flies, the extra switch flicked on. Three hours later, Tully gave the flies a single lesson. The flies remembered that lesson for a week. And as one drosophilist wrote in a commentary on Tully’s work, “For flies in the wild, one week is long enough to arrive at midlife crisis.” It was as if a boy of twelve were told a phone number once and he still remembered the number at the age of forty. With the extra on switch, the flies had the equivalent of a photographic memory.
Tully and his lab had proven that they had found at least a piece of the machinery of memory. In the same stroke, they had changed memory in the direction that every thinking generation has dreamed of.
FOR AN ARROWSMITH, this discovery, building as it did on a century of cumulative scientific effort, would have brought blurred nights of almost religious awe. For Watson, who strains every nerve every day to support the Cold Spring Harbor Laboratory, the significance of the moment was more mundane. According to laboratory legend, when Tully first told Watson, Watson screamed out, “We’re going to be rich!”
Watson had come to agree with Benzer that the path through the gene would lead to a greater understanding of the mind and brain. But unlike Benzer, Watson thought immediately of the control of illness, the enhancement of the mind, and the commercial and political possibilities of anything that pointed in those directions; which is why he had decided to gamble on Tully. Soon after Tully’s discovery he was talking it up to Wall Street investors, with one of his self-mocking smiles: “Insect brains.”
“Well, you see, genetics in a sense restricts our freedom,” he says, sitting in the vast suite of offices that he inhabits as director of the Cold Spring Harbor Laboratory. “I mean, we’re born with a lot of things, which means that we can’t do everything. And some people have more freedom than others. It’s not an—” A pause. “It’s, um, slightly uncomfortable,” he adds in a confiding tone. “It’s really uncomfortable if you don’t have them, and if you do have them, wondering whether you deserve them.…”
Watson has an acute sense of himself and what he represents. His Nobel Prize certificate hangs on the wall behind his desk. The bell tower that tolls the hours outside one of his office windows has a staircase shaped like a double helix. To Watson it seems an omnipresent fact of personal experience that genes give some of us more freedom than others. “Well, the freedom not to be sick,” he says, “or the freedom to have your brain work right. As distinct from most people,” he adds, “whose brains don’t work right.”
BECAUSE OF THE family likeness of genes in the tree of life, no one at Cold Spring Harbor was amazed to discover that mammals have a creb gene very much like the fly’s. In another laboratory at Cold Spring Harbor, a molecular biologist named Alcino Silva tried the same sorts of experiments with mice that Tully had practiced upon his flies. By now mice had reached the point where flies were in the 1960s: they had become powerful tools of genetics. Anyone with a computer in the mid-1990s had access to maps with 7,377 markers on twenty pairs of mouse chromosomes. If Silva found an interesting quirk in the memories of his mice, he could quickly track it down to a gene on a chromosome. Then he could clone that gene, insert it into a mouse egg, and see if it transformed the behavior of the mouse, just the way Tully and his crew had done with flies. The mouse genome is about the same size as the human genome, and for virtually every gene in us there is a corresponding gene in the mouse. It is true that the genes do not always do the same things, but gene for gene we are clearly quite closely related. Even our chromosomes are the same: if you chop mouse chromosomes into about two hundred pieces and put them together rearranged, you can make human chromosomes.
Not only can mouse workers take a single human gene and put it into a mouse and see what it does, they can even take a big fragment of a human chromosome from a human being—a fragment bearing hundreds or even thousands of genes—insert it into a mouse egg, and see what it does. And even before the first cloning of mice in 1998, hundreds of inbred mouse strains were so nearly identical that they provided geneticists with virtual armies of identical twins.
The creb gene in a mouse is an on switch and an off switch virtually identical to the fly’s. So Silva tested a strain of mice with a mutant creb gene. In one experiment, he tried putting a mutant mouse in a round, steep-sided pool called a Morris Water Maze. When the mouse found itself in a Morris Water Maze, it tried to save itself. The only escape was a small island that was submerged just under the surface of the dark, opaque, deep-dyed liquid somewhere in the pool. After a few swims a normal mouse could learn to find the island. But the mouse with the mutant creb switch searched for the island at random all through the pool again and again. Every time, it made land purely at random. Without creb it could never remember where it had found the island the time before. It may even have forgotten there was an island to be found.
At Columbia, Kandel and his group were testing creb mice too. They used a white disk with holes around the rim, one of which was an escape tunnel that led back to the mouse’s cage. Under the pressure of bright hot lights and a loud alarm-clock buzz, each mutant mouse dashed around the disk looking for the one hole that it had found before, the hole that would lead home. A mouse with a ruined creb gene would race from hole to hole, leaving signs of anxiety everywhere it went. It would sniff around between the holes, run back to the middle of the disk, scratch its head. When test time was up, a creb mutant would still be scurrying around looking for that one hole that led home.
So creb seems to be a universal sort of toggle switch that an animal throws when it wants to remember something for a long time. Apparently these switches have been preserved for many hundreds of millions of years. No one yet knows how many other switches there are that allow a human being to remember events in uncanny detail for uncanny amounts of time. Jeff Hall has this gift, though he is not always happy to have it. With each year he looks more burdened by memories, lost causes, and unrighted wrongs. His father was the same way: a star reporter for the Associated Press who never took a note. Hall is a prodigy on the Gettysburg battlefield, which he haunts in his off-hours in his stiff blue Union cap. Whenever anyone asks him a question, he answers volubly, voluminously, and with something so much like gratitude in his voice that he sometimes draws crowds.
Hall once wrote an angry letter to Benzer citing Benzer’s lack of generalship, his defection from the field of behavior, and many other grievances. “I remain offended,” Hall wrote, and added, “unfortunately, I never forget a thing.”
THE CREATION of a fly with a photographic memory got so much press attention that one day in April 1995 Watson and Tully actually began to wonder if the work might be putting their lives in danger. Local, national, and international television and newspaper reporters were trooping through the Fly Room to see the flies perform in Tully’s Teaching Machine. And in the middle of the media storm, the Unabomber, who had just mailed his sixteenth bomb, mailed a letter to the New York Times, a letter written in the royal or revolutionary “we.” An excerpt of the letter ran in large type across the top of the front page. The letter explained the bomber’s hatred of genetic engineers, among other specialists:
We have nothing against universities or scholars as such. All the university people whom we have attacked have been specialists in technical fields. (We consider certain areas of applied psychology, such as behavior modification, to be technical fields.) We would not want anyone to think that we have any desire to hurt professors who study archaeology, history, literature or harmless stuff like that.
Watson dropped his copy of the Times on Tully’s desk with the headline facing Tully. “So that mad bomber particularly doesn’t like geneticists,” Watson said, looking very pale and tired. He and Tully stared at each other. In their adjoining offices they felt like sitting ducks. “I think we’ve got to ask that all our packages go through an X-ray machine,” said Watson.
“Oh, man,” Tully said, “I’m history. It’s just a matter of time before I lose my fingers.”
But that spring Tully exulted in the success of his experiment. “I’ve always known exactly what I wanted to do,” he said. “I didn’t know how long it would take and how resistant people could be when they have you pigeonholed.” Pigeonholed as hopeless—Watson has files full of letters advising him that it would be impossible to learn anything important about the human mind by poisoning the brain of a fly. “That has been frustrating. The whole miracle was Watson. It was an absolute angel that swooped down and said, you come to Cold Spring Harbor.”
And that spring Watson exulted too. “If I were a sort of graduate student now, I’d probably gravitate toward something like Tully’s lab,” he said. Most young molecular biologists are working on the genes that control the growth of embryos. “I don’t regard that as—” Watson snickered sarcastically. “You know, I don’t want to work on the twenty-third homeo-box.” Let someone else nail down the details of the gene cascades that make an egg grow into a fly, a zebra fish, a mouse, or a human fetus. “Exactly how it grows—we’re a long ways from totally describing that,” Watson said. But the process is no longer essentially mysterious. “You could say that sequential gene action is the basis of it all. You’ve got these cascades that produce the eye …” And he rattled off some of the fantastic gene-to-gene cascades that shape the eyes and the limbs and differentiate the head from the tail. This is the subject that Brenner and Benzer helped to open up when they turned away from behavior, and it is the field that fruitless is now helping to illuminate.
“I mean, Jesus, embryology’s solved!” cried Watson with dismissive hyperbole. “But I don’t know how memory works in the slightest. You know, no one can give any model of the human brain. And that’s why it just seems to me”—he snickered—“a bit of a challenge.”
A finding like Tully’s explains why so many of the first molecular biologists turned from the gene to the brain in the first place, Watson added. Molecular biology had been getting routine. “Whereas the brain is slightly different. It just seems much—more fun.”
WITH THE CLONING and rewriting of the fly’s creb gene, it was clear that Tully was inside his problem. He was exploring from the inside the kinds of behavior that Pavlov, for instance, had studied from the outside when he conditioned dogs to drool or flinch at the sound of a bell. Just as Pavlov’s work became a cornerstone of behaviorism, which is the study of behavior from the outside, Tully’s work will be a cornerstone of the new study of behavior from the inside. It is now quite clear from studies in both these fields that all animals, even flies and snails, can learn to associate two events that come at them one after the other: a bell and a morsel of food; a smell and a shock. With flies, dogs, or human beings, when an experimenter slowly separates the two events, learning slowly falls away. In each species there seems to be some maximal time span after which there is no further associative change. Tully believes that this limit reflects a cellular property of neurons that can now be studied at the molecular level and will eventually be found to function essentially the same way across all life. What Pavlov explored in his dogs and Tully and his colleagues are now exploring in flies and mice, the power to learn and remember, is very old. Since we have the on-off switch, and so do flies and worms, it must be ancient, like sex and the clock. This, too, is a molecular invention that is older than the Cambrian explosion; and it is hard to imagine a more interesting molecular invention for us to begin to understand.
Tully once made a pilgrimage to Pavlov’s old institute in St. Petersburg and asked to see a list of the names of Pavlov’s dogs. The only dog’s name that he had ever been able to find in the literature was Birka, which means “Whitey.” No one at Pavlov’s former institute knew the dogs’ names either. Then Tully went to Pavlov’s apartment, which has been preserved just as he left it the day he died. A curator gave Tully a tour of the place. At the end of the tour, they sat down for stale coffee and biscuits, and the curator pulled out an old photo album filled with the names and mug shots of Pavlov’s dogs.
Now Tully plans to honor those names. In his view, behavior sits at the top of a biological pyramid. At the bottom of the pyramid there are genes in a cell. Then there are cells interacting to make a living system: say, a walking, talking, learned, arguing organism. Learning and remembering are among the most exquisitely complicated feats that a walking, talking organism can accomplish: the absolute top of the pyramid of organization. So there are bound to be hundreds of genes involved in learning and remembering, and Tully plans to find them all, one by one, and name them after Pavlov’s dogs.
When he and his lab engineered the fly with the photographic memory, Tully felt an amazing rush of vindication and exhilaration. They had made a beginning. He knew that much of his crew’s ability to do such interesting work derived from their position in time, high in Occam’s Castle. He thanked his stars that he had arrived on the scene just as it had become possible to grasp some of the first solid links between genes and behavior and to fulfill the mission he had been pursuing since he had been a student in college. “This is the point in time when we get to talk about these things,” Tully says. “This is the defining moment. Things finally have come to a head.” In one corner of his office at Cold Spring Harbor, he keeps a photograph of himself in Pavlov’s old apartment, wearing Pavlov’s hat. A few weeks after his eureka he crossed his office and peered at the photograph. “It really is a pilgrimage,” he said. “You can see it on my face.”
“It really is a pilgrimage.” Tim Tully in Pavlov’s old apartment in St. Petersburg, wearing Pavlov’s hat. Benzer in Darwin’s old house, south of London, writing at Darwin’s desk. (Illustrations credit 16.1)
In the same spirit, Benzer once visited Darwin’s house in the village of Down, not far outside London. He explored the place with the impudent curiosity he brings everywhere. His daughter Barbie was with him. She is used to her father’s ways; she remembers following him as he strolled past “No Trespassing” signs in O’Hare International Airport because he wanted to see the control tower (“Where everyone else gets ushered out, he gets a tour.”). At the Institut Pasteur, Benzer once outraged the institute’s French administrators by asking if he and his family could have Pasteur’s old apartment. (Absolutely not!)
Darwin’s study was roped off. Benzer admired Darwin’s chair, equipped with wheels so that Darwin could roll around among his files without getting up, and a blanket to keep him warm. “Barbie so impressed the guard that I got to sit in the chair,” Benzer says.
“No one ever gets to sit in that chair,” Barbie says.
The guard lifted the velvet rope, and Benzer climbed up into the chair. It was so high that his feet hardly touched the floor. “That was a great thrill for me,” says Benzer. “They didn’t have his slippers around, unfortunately, so I couldn’t put on Darwin’s shoes.”