In the late 1970s, behind the towering wrought-iron gates that rise up just east of where 66th Street meets York Avenue in Manhattan, a birdman was closing in on a revelation that would signal a momentous “sea change,” as one top-drawer scientist would call it. For so long, older neurologists had taught younger neurologists that the brain of adult vertebrates produced no new neurons. And then would come an out-of-the-blue finding, catching most people, but not everyone, by surprise. The amazement would soon lift, the disbelief blow away, and all that would remain would be a hungry incentive to try and steer yet another stem cell—this one from an organ thought to be nearly impossible to heal—toward therapies for countless grim disorders.

More than fifteen years had passed since Ernest McCulloch and James Till had snared evidence of a blood-making stem cell in adult mice. Following in their footsteps, investigators were at different stages of closing in on stem cells that populate other organs of grown

animals. There were those in muscle that rebuild muscle fiber; those deep in the walls of the small and large intestine that replenish the intestinal lining; those in the basal layer of the skin that serve to replenish skin; and those in the testis that generate sperm cells. Sometimes an organ even contained more than one category of stem cell. For example, investigators were increasingly appreciating that alongside the stem cell in the bone marrow that made blood cells lived a cousin stem cell that spewed out new bone, cartilage, and fat cells.

Research on rats had indicated that the liver most likely had no resident stem cell. Of the two ways that organs acquire new cells—either from the division of stem cells or the division of mature cells—the liver’s regenerative zeal seemed entirely due to the mature type. Liver transplants were demonstrating great promise by the mid- to late ’70s, and when a person donated half of his or her liver to someone in dire need of this organ, both donor and recipient could rest assured that their portion would grow back to its original size, usually in about a month’s time, thanks to the lightning-fast division of hepatocytes, the liver’s highly specialized cell that processes nutrients from the blood for the entire body. If only every human organ were so regenerative! “That’s why no one had been ready to believe there were stem cells in the liver—because hepatocytes were known to be the cells that responded to liver cell loss,” relates Stewart Sell, a pathologist at the Ordway Research Institute in Albany, New York.

In 1975, however, Sell, who at the time was based at the University of California, San Diego, stumbled onto something. While investigating liver cancer in rats, he had run into a small round cell that was incredibly proliferative. It just so happened that during his med school days in Pittsburgh, Sell had had Barry Pierce as an instructor, and, “because of the seed that Barry Pierce had put in my head,” he relates, he concluded that the rats’ liver cancer probably was due to these small, fast-multiplying cells, which must be stem cells. “If it hadn’t been for Barry Pierce, I would have gone along with the dogma,” which had the cells originating in the bile ducts and proliferating in response to liver injury.

Later that year, Pierce, who had moved to the University of Colorado, happened to invite Sell to address a gathering of pathology students and faculty. Sell shared his interesting news for the very first time: He believed he had unearthed a stem cell in the liver. Given the liver’s normally slow turnover of cells, he suspected that as stem cells go, this one was essentially inactive—certainly nothing like the intestine’s busy stem cell. When part of the liver was surgically removed, mature hepatocytes, and not stem cells, reconstituted it. Nonetheless, he was quite certain that he had hold of a stem cell, one remarkably passed over all this time. Sell remembers that a hand shot up in the audience. “Why does the liver have stem cells if it doesn’t use them?” a faculty member asked. Responded Sell, “Just because hepatocytes can regenerate the liver doesn’t mean stem cells aren’t needed.” He and others would go on to identify situations in which they indeed were badly needed. Should the whole liver feel the severe effects of viral hepatitis or a toxic chemical, whereupon its hepatocytes dangerously stop dividing, then its stem cells would kick in, in an attempt to revitalize the liver with cell progeny.

By and large, adult mammals didn’t stand out as regenerative creatures. Yet as the evidence grew that numerous organs had stem cells that spun out new cells, adult mammals were beginning to look more regenerative than anyone had ever stopped to notice. Search as they might, however, biologists found no signs of stem cells in the adult heart, pancreas, or kidney. And as for the brain, there was no point looking there for stem cells, at least not in higher vertebrates, most agreed. Ever since the great Spanish neuroscientist Santiago Ramón y Cajal brought the nervous system into fuller view in the early century, the conventional wisdom was that a mammal’s neurons were made primarily before birth, during embryogenesis. Neurologists somberly advised, “Take care of the brain cells you are born with, because there’s no replacing them.” Or as Cajal had choicely put it: “Once the development was ended, the founts of growth and regeneration … dried up irrevocably.”

People lived under the impression not only that their brain cells

were irreplaceable, but that virtually thousands were dying every second. The brain’s regression was thought to start in adulthood and accelerate in old age, the loss of neurons linked to the ravages of time and possibly made worse by such lifestyle habits as heavy drinking and drug taking, not to mention degenerative calamities like Alzheimer’s or Parkinson’s disease. The inescapable reality that disease and injury to the brain could be so deadly only reinforced the belief that this noble organ had no way of repairing itself. It seemed devoid of stem cells that might forge new neurons, and its mature neurons seemed incapable of division. How could they divide? scientists would exclaim to each other, given their often long and branching processes, the fibers (axons and dendrites) that connect them to any number of other neurons. What a tangle it would be if they did divide!

While it might lack stem cells, the adult brain was perceived to have a bit of regenerative oomph nevertheless. Neurons could sprout new connections and often did so when the brain was under duress. In addition, the nervous system had a second type of cell—glial cells—that divided throughout adulthood. Yet glia, which were looked upon as the “glue” that supports neurons, were ranked as fairly plebeian in comparison to the kingly neurons. Neurons were the conduits of everything learned and remembered, which strengthened the impression that they and their connections must be as fixed and irreplaceable as memories were fixed and lasting. “It was hard to imagine how brain cells and their processes could divide and give rise to other cells that would integrate into the existing circuit—especially in the hippocampus, which people thought was involved in memory. It would somehow screw up the acquisition of memories,” observes neuroscientist Fred Gage. Or as Pasko Rakic, another neuroscientist, once observed, you would have new cells that never went to elementary school.

Not that researchers had never glimpsed neurogenesis—the ability to generate new neurons in the brain and spinal cord—in adult animals. They had, and for several decades. But either the animals

reported on were “lower” on the totem pole, their biology deemed different from that of more complex animals, or the science behind these reports was viewed by other scientists as less than persuasive, which is what befell a string of reports in the 1960s by Joseph Altman.

Originally from Budapest and trained as a psychologist, Altman was working in MIT’s Psychophysiological Laboratory when, challenged by the paucity of insights into the brain and behavior, he began examining the brains of rats. “Like everyone else,” he recounts, “I believed that neurogenesis in mammals was a prenatal phenomenon”—something that largely happened before birth. Yet scanning the brain tissue of ten young adult rats, he glimpsed what he thought were new neurons. “The first paper I published was a question mark: Are there new neurons? In two to three years I was one-hundred-percent convinced there was neurogenesis after birth—even into adulthood in some regions.” In 1968 he relocated to Purdue University, where he continued collecting leading evidence.

Altman was in a good position to go brain-cell sleuthing, early user that he was of a novel way of tracing newly made cells. The method worked off an injectable form of thymidine, a chemical that when introduced into the bloodstream was taken up by a cell’s DNA as a cell prepared to divide into two cells. Rendering thymidine radioactive turned it into a tracer. Once inside about-to-divide cells, it ended up in the nuclei of the daughter cells, tagging these new cells and making them easily identifiable.

For more or less a decade, Altman ran one after another thymidine-labeling experiment on rats, cats, and a few guinea pigs, and when he checked his autoradiograms (film exposures that reveal the presence of radioactive thymidine in cells), time and again thymidine’s peppery traces jumped out at him. As he communicated to the rest of the world through a series of papers, he was seeing new cells in the mammal cerebellum, a hind region involved in muscle movement; in the olfactory bulb, which aids smell; and also in the hippocampus, a region dear to memory. The new cells he was spot-

ting weren’t just glia, the brain’s far-and-away most common type of cell. He also was seeing new neurons, the cells whose connections were regarded as the brain’s force and fire. Of this he had no doubt, he says today, for he had full confidence in his tools and the incredible vistas of cells that they were showing him.

To other neuroscientists, some of Altman’s sightings in the mammalian brain weren’t outlandish. Studies had shown that neurogenesis could keep sputtering along for a while, postnatally. For example, cells in a human infant’s cerebellum were known to keep dividing and making new cells for as long as two years after birth. Instead, it was Altman’s claims of new neurons in places where they had seldom, if ever, been described in the brains of older mammals that flummoxed his colleagues. New neurons in the hippocampus and the cortex? These regions were the seats of higher intelligence, places where the circuitry was deemed far too complicated to allow for the production or incorporation of new neurons. Altman admits that his evidence of neurons in the cortex was inconclusive; but of those in the hippocampus—“I was without doubt.”

So tantalizing were Altman’s findings that other investigators scrambled to replicate them. No one came forward with completely convincing data, however. In the late ’70s, Boston University’s Michael Kaplan also published accounts of new cells in rat brains, but his tests were viewed as not being extensive enough. As for Altman, some biologists would outright dismiss the Purdue biologist’s claims or view them as an exception to the ironclad rule that grown mammals experienced no new neurons. Conceivably, what Altman was seeing in rats was a phenomenon left over from leeches and other lower animals; or a phenomenon that lingered for a brief time after birth. The impression that emerged was, the phenomenon was nothing to dwell on, because it wasn’t generally found in mammals. Altman himself feels that, in a field rife with competitors, his work was just plain ignored. “People acted like our papers, which we published one after another, had never appeared,” he says today.

Altman and Shirley Bayer—Altman’s wife and a first-rate neurobiologist in her own right—would continue to study the process of neurogenesis in early development, making up their minds to do so with minimal contact with the professional community that had let them down. Data compiled by Bayer in the early ’80s, which Bayer feels were passed over by colleagues, provided the strongest proof yet of the continuous birth of new neurons in the adult mammalian brain. She and her Purdue team reported in Science that in rats of ages one month to one year, neurons in the dentate gyrus layer of the hippocampus increased by roughly thirty-five to forty-three percent. Moreover, she also gleaned unmistakable clues that neurogenesis kept happening straight into a rat’s old age.

Accounts of new neurons forming in the nervous systems of a bevy of nonmammalian creatures, including fish, amphibians, insects, and crustaceans, kept accumulating, as they had for some time. Scientists observed new neurons materializing in the respective brain centers of a cricket, a milkweed bug, a monarch butterfly, and a praying mantis; in the spinal cord of a stingray; in the retina of an old goldfish. Neurogenesis often appeared to be a lifelong occurrence in these creatures. Yet most scientists continued to assume that only less complex animals, mainly ones that were coldblooded and whose brains and eyes grew throughout adulthood, displayed neuronal growth.

The brains of adult mammals and other higher vertebrates make no new neurons. Fernando Nottebohm, the birdman behind Rockefeller University’s tall front gates, had encountered this basic tenet while he was studying zoology at Berkeley in the early ’60s. Now, in the late ’70s, as he probed the brains of songbirds to learn more about the mechanisms that enabled them to sing, he had to heed this as-

sumption. But the further his research progressed—both in a ground-floor lab in Gasser Hall and at the university’s Field Research Center in Millbrook, New York—the more he questioned what he had been taught. It was dawning on him that the very bird strains he’d been attracted to ever since his boyhood roamings in Argentina might be telling him something quite different.

“I had a passion for birds as a child. I couldn’t take my eyes off them,” shares Nottebohm, whose face—whether it’s his bushy eyebrows, direct gaze, or the shape of his mouth—contains something avian about it, although the rest of him—square-shouldered, with a distinguished deportment—is not at all bird-like. Nottebohm did most of his growing up in Buenos Aires, but his instincts as a naturalist were honed at his parents’ second residence, a ranch situated on the fertile pampa of central Argentina that stretches south of Cordoba. His hobby there, he recounts, “was to go walking or riding through the country noticing birds and animals.” Prolific numbers of South American songbirds, such as mockingbirds, Chingolo sparrows, and finches, beguiled the boy with their whistlings. Not to be outdone were guira cuckoos, burrowing owls, and scissor-tailed flycatchers that, while not songbirds, invariably caught young Nottebohm’s eye with a certain chirp or curious habit.

During his sophomore year at Berkeley, Nottebohm had taken a course taught by Peter Marler, an authority on vocal learning in songbirds, and had come away spellbound. Vocal learning refers to a young bird’s—or human’s—ability to imitate sounds made by its elders. Not only are birds and humans both vocal learners par excellence, outdoing other animal groups in this regard, but a similar biology possibly underlies the ability in both to vocalize. Marler himself had been a student of the field’s originator, William Thorpe, an animal behaviorist at the University of Cambridge. It was Thorpe, in particular, who had extended the use of the sound spectrograph, an instrument invented for monitoring sounds undersea during wartime, into a tool for rendering a visual display of birdsong on paper. The retooled instrument did away with the drudgery of transcribing

birdsong by ear, says Marler, “and made it possible to grapple with all the detail, variability, and complexity of birdsong. I like to think it was the beginning of the whole science of birdsong.”

So scintillating were Marler’s lectures about how various animals had evolved specific traits in tune with their surroundings that he nearly hypnotized his students, Nottebohm relates. The Englishman’s descriptions of why birds communicated and the sounds they used for this purpose were so “irresistibly interesting,” says Nottebohm, that he made up his mind to devote his own career to songbirds. By looking into how vocal learning evolved in birds, he hoped to pocket new insights into the evolution of human vocalization, and maybe even gain fresh clues as to whether the acquisition of language had paved the way for human consciousness. (Despite this last hope, Nottebohm admits that after nearly forty years of studying vocal learning, he is none the wiser about consciousness. “We all believe it’s there, but if you can’t quantify a phenomenon, you have nothing to grasp,” he remarks. “And with consciousness, there’s nothing to grasp, because you can’t weigh it or measure it.”)

When Nottebohm was seven or eight, he somehow captured a white canary that had flown into his family’s large garden in Buenos Aires. Placing it in a cage in the courtyard, he listened to his friend’s song for years to come—and even ended up an imitator. When relaxed and wandering around Rockefeller’s Millbrook field center, which he directs, Nottebohm easily breaks into a melodic whistling.

Now, in the late ’70s, he often thought of his canary and how it had been a harbinger of things to come. More than the cheerful chaffinch, the slightly squeally zebra finch, or any other bird that he had so far studied, it was the effervescent canary that was providing Nottebohm with clues about how a bird sang. A few years earlier, in 1976, he had mapped out the canary’s song system, which, located in the bird’s forebrain, is comprised of two small yet distinct areas of neurons. Impulses conveyed by the large neurons in these regions travel down neural fibers and activate the muscles of the syrinx, the bird’s vocal organ. Instead of involving cells throughout the brain,

the biology of birdsong appeared to be quite confined, which was terrific news for Nottebohm and other like-minded researchers, since it meant that figuring out the biology behind birdsong might be that much easier.

A question that occupied Nottebohm, one that had been around for some time, was, why is it that male birds so often are the singers, whereas females sing much less or not at all? Of the world’s approximately 4,000 species of songbirds, those that migrate are most likely to exhibit this male-female discrepancy. The behavior may have evolved because migrating males often fly ahead of the females and use song to mark their territory and attract a mate. Nottebohm wondered if some anatomical difference between singing males and nonsinging females possibly explained their vocal differences, and later in 1976 he and his first postdoctoral student, Arthur Arnold, confirmed as much. They observed that a male canary’s song center was as much as three to four times larger than a female’s. If a larger song center accounted for the male’s singing ability, this size difference made perfect sense. Yet it also came as a surprise. Such a blatant structural difference between a female and male brain had never before been encountered in a vertebrate species.

Nottebohm wasn’t entirely sure what to make of this finding, except that there had to be more of something in a male’s song center—more blood vessels, or more cell connections, for instance—that made it physically bigger and supported the outflow of song. Could it be that there were more neurons?

From decades past came a handful of reports that he read with great absorption. The gist of these findings was that female canaries could sing as vibrantly as males if injected intramuscularly with testosterone. Bird importers in the ’40s and ’50s apparently had been wise to this trick and used it to hoodwink pet-store owners. Accounts tell of storekeepers complaining that they would sell a melodious canary to a customer, only to have the customer reappear and demand his or her money back, all because the canary had suddenly stopped singing. The storekeepers suspected the importers of boost-

ing the female birds’ testosterone levels, which made the females sing, but only for so long.

Since testosterone had this effect on female canaries, their brain’s song center must be extremely hormone-sensitive, Nottebohm realized. But what exactly did the ruse of implanting testosterone do to the song center? Enlarge it, as the singing male’s larger song system implied? But how? Did the hormone cause blood capillaries to expand? Did it make neurons physically bigger? Did it cause a significant increase in the overall numbers of glial cells or neurons? Nottebohm would quickly shake off this last idea. “People would have found it laughable,” he says today, given the assumption that the adult brain made no new neurons.

To find out what was going on in the female’s brain, in 1979 Nottebohm set about the logical experiment. He implanted capsules containing testosterone under the breast skin of several female Belgian Waserschlager canaries and waited for hormonal changes to ensue. Sure enough, just as demonstrated by other scientists, in about a week’s time the lady canaries were singing as robustly as choir boys. Some weeks later he proceeded to euthanize the birds, remove their brains, and weigh and measure them. He couldn’t quite believe his results; but then again he could. The testosterone that outwardly had turned the females into singers inwardly had more than doubled the size of the female’s two forebrain song-center regions. It was the first time that a hormone had been observed to create such a gross anatomical change in the brain. Of far greater importance, it was eye-opening testimony that the vertebrate brain was far more flexible and mutable than people realized. Here again, the brain was making more of something. Could it be cells? If new cells were being born, that presented the incredible picture of a regenerative brain wherein dwelled proliferating cells that had stemness—neural stem cells. Nottebohm was beginning to warm up to the notion.

On the heels of that test, the Rockefeller scientist completed another experiment that would spectacularly sew together the gathering edges of his research. He was assisted by his wife, Marta, who

often lent a hand with his experiments. In the field, for instance, she had sat confined for hours in a blind in Trinidad recording the bugling calls of orange-winged Amazon parrots. At Rockefeller’s field center in Millbrook, which lay a few miles from the Nottebohms’ residence and consisted of a large Tudor-style fieldstone gatehouse that stood between field and forest, chief among her talents was analyzing by eye the birdsong that a sound spectrograph imprints on endless yards of film, a meticulous job that is now done by computer. In addition, she edited her husband’s writing into “understandable English,” by his own claim. Overall, Marta was Nottebohm’s tireless sounding board, “a refreshing, uncorrupted intelligence,” as he describes her, whom he relied on more than he did the scientific community, which he viewed as being filled with theories that could lead one astray.

Ornithologists had long known that male songbirds sang most brightly in the spring, but as summer set in sang less and less, until their song all but disappeared in mid- to late summer, which was molting time. In the fall, a bird started getting up to song speed again and by January had acquired a whole new repertoire. A detailed comparison made by the Nottebohms of bird brains in spring and in fall brought the fascinating discovery that the canary’s song center changed size according to the season: in singing season, namely spring, it was bigger; by early fall, when the singing season was over, it was smaller—much smaller. Over the course of those few months it could shrink by as much as half. The Nottebohms weren’t terribly surprised to further discover that the birds’ levels of testosterone rose and fell directly in parallel with the song center’s swelling and shrinking.

When testosterone levels rose, what biological change swelled the song center, making a bird sing more? Testosterone itself wouldn’t account for this swelling. Rather testosterone was causing something to grow and make more of itself. In various papers, including the above experiment’s write-up in Science in 1981, appropriately titled “A Brain for All Seasons,” Nottebohm aired his hunches. Maybe the

song center’s expansion was due to an increase in either the connections between neurons or the numbers of neurons or glia. His lab soon got evidence. In spring, at the peak of singing season, the brains of male canaries indeed had a greater complexity of fibrous connections than in late summer. Yet he suspected a proliferation of new brain cells was also happening. If the brain harbored new cells—or stem cells that gave rise to new cells—one might never know they were there unless the right markers were used.

Nottebohm’s was “a lone voice” in those days, describes Steven Goldman, a former graduate student of the Argentinian. Few others were raising the possibility that the vertebrate brain might be fashioning new neurons. Moreover, Goldman adds, Nottebohm’s familiarity with a bird’s song system made him uniquely qualified to utter such an iconoclastic theory. Casting about for a laboratory to work in, Goldman, then a first-year student in Rockefeller and Cornell’s joint M.D.-Ph.D. program, had been drawn to Nottebohm’s lab both because Nottebohm’s focus on bird brains was “so far out there,” recalls Goldman, but also because of his own burgeoning interest in brains and the then-science-fictional prospect of repairing them. In his freshman year at the University of Pennsylvania, where he majored in psychology and biology, he had written a report on a Russian psychologist’s attempts to transplant the heads of dogs. “I’ll never forget the photographs of those poor two-headed dogs,” says Goldman. “It really made me interested in brain regeneration.” Nottebohm’s bird research also took Goldman “back to West Philly,” he says, where as a boy he would escape his neighborhood’s roughneck street life by slipping away to the nearby confluence of parkland and golf courses to watch birds. “So on some subliminal level, Fernando’s was an obvious attraction. It was an opportunity to look at brain plasticity and at the same time have an excuse to study birds,” relates Goldman, who today heads the division of gene and cell therapy at the University of Rochester Medical Center.

Nottebohm’s initial aim had been to get to the bottom of how birds learned to sing, but around the time Goldman joined his lab,

the birdman was unexpectedly veering off into much bigger territory. When he surveyed the field center’s many dozens of bright-yellow canaries tinged with green, which were housed in an outdoor aviary at the time, he was aware that he just might be looking at an animal whose brain—a brain that could easily be confused with a chickpea if you didn’t have your glasses on, Nottebohm would tell students—had surprising things to say about all vertebrate brains.

With Nottebohm’s guidance and encouragement, in early 1981 Goldman launched an experiment to solve the puzzle: Did a canary’s song center get bigger because of added new brain cells? Or was this change entirely due to increased quantities of neuronal fibers? The plan was to implant testosterone capsules in female canaries; inject the birds with radioactive thymidine; wait a few weeks to give the testosterone time to do whatever it did to the bird’s brain; and then inspect the birds’ song centers. If doses of testosterone did cause an upsurge of new glial cells or new neurons—bingo!—the thymidine should make the new cells spottable. Since neurons supposedly didn’t reproduce, Goldman’s own hunch was that multiplying glial cells were the unknown element that was making the song center get bigger. Nottebohm took a more wait-and-see stance. He felt it was entirely possible that new neurons were behind the swelling, but so might be some entirely unforeseen phenomenon. Like a latter-day Abraham Trembley, Nottebohm didn’t like pinning himself to any one theory.

The study was small—only four birds—and unfolded largely in Nottebohm’s Manhattan lab. Three weeks out, Goldman euthanized the canaries, but it wasn’t until several weeks after that—after flushing them with formaldehyde to preserve their organs, removing their brains, embedding the small orbs in paraffin, slicing them thinly, then mounting each slice onto a glass slide—that he was ready to begin the really interesting part.

Perched on a stool in Nottebohm’s office one evening, Goldman adjusted the magnification on an Olympus light microscope from low to high, and with that small gesture visually entered a cross sec-

tion of brain. His gaze immediately settled on a crowd of cells whose nuclei contained black speckles of radioactive thymidine. The cells included new glia, and—there was no doubt in his mind—new neurons. “I was sure they were neurons because they were large, large like the pyramidal neurons that populate the song center,” says Goldman. Amazement kept him glued to the microscope until well past midnight, that winter night in 1981.

When Nottebohm saw the results, despite being ready for whatever outcome, he too was astounded, yet also vigilant. “I thought the profiles of the cells were very distinct and different from glia. There seemed to be no question we were looking at neurons,” he recalls. “But I think I always left the escape hatch open, in case they were cells that were only neuron-like” and not the real McCoy.

This small study was a teaser. A larger study was sorely needed that could verify this remarkable sighting of new neurons and further address a slew of questions. If the new cells really were neurons, where were they coming from—the song center or some other region? The scientists had noted a sprinkling of new cells in the wall of the lateral ventricle, one of the brain’s passages for cerebrospinal fluid. Were cells born there, and did they then migrate the short distance to the song center?

On June 10, 1981, Nottebohm and Goldman began essentially the same experiment over again, this time using eighteen birds. The study would become Goldman’s Ph.D. thesis—as celebrated a thesis, as it turned out, as a young neurologist-in-training could wish for. By euthanizing testosterone-treated birds at varying intervals, some of the researchers’ questions began to find answers. In birds euthanized early on, lots of new labeled cells were seen in the ventricle wall and no place else, whereas in birds euthanized a few weeks later, hardly any labeled cells were present in the wall, yet considerable numbers appeared in the song center. Quite possibly, new cells were born in the ventricle wall and then traveled to the song center.

This brain display was something! On any given day, the number of new neurons that a canary’s brain was churning out was ap-

proximately one percent of the song center’s neurons, which wasn’t inconsequential. And while it had been difficult to imagine how neurons in more advanced brains could possibly divide to give rise to new cells without causing a tangle of connections, here in the bird was an answer. Immature cells—neural stem cells—that hadn’t yet sprouted axons and dendrites were dividing, and their offspring were becoming neurons.

Another piece of this discovery was that the new cells seemed to originate in the ventricular wall. This same scenario occurs in the mammalian developing fetus when new neurons emerge from this wall and migrate in droves to their final work stations in the brain. But cell migration in the adult brain wasn’t necessarily anticipated, because the brain wasn’t thought to make new neurons. Then again, Joseph Altman had pointed to this very spectacle in the adult rat brain with some insistence, and in his papers mentioned others before him who had also observed cell proliferation in the adult brain, including man’s. “These facts would suggest a high rate of cell proliferation in adult rats in [the subependymal layer of the lateral ventricle] and the migration of labeled cells to as yet undetermined brain regions,” wrote Altman in 1965. In a few years, he determined that one place that cells migrated to in young-adult rat brains was the olfactory bulb, the brain’s forwardmost point, and named this path the “rostral migratory stream”—rostral, as in forward.

The surprises didn’t stop there for the Rockefeller scientists. New neurons even showed up in the song centers of females that hadn’t received testosterone. This finding amazed the Rockefeller detectives the most, since it hinted that the vertebrate brain might be giving birth to cells all the time. Female canaries weren’t altogether mute; sometimes they sang a little at the end of breeding season, when their testosterone levels were known to rise a bit; and they also issued a range of calls. Possibly the new neurons contributed to this vocal behavior, the scientists guessed.

If they were to challenge the dogma that held that adult vertebrate brains were incapable of regenerating neurons, Nottebohm and

Goldman knew they had better be dead certain that the new cells they were glimpsing in a canary’s brain were really neurons. Searching back through the literature, they would unearth Joseph Altman’s papers posting evidence of neurogenesis in rat, cat, and guinea pig. Initially they were disappointed; they had imagined they were the first ones to spy new neurons in adult advanced vertebrates. And yet Altman’s research trek became helpful to them, and not only because his findings supported their own. Learning about the “phenomenal resistance” that Altman had encountered made them twice as cautious about how to proceed, remembers Nottebohm. “Neither Steve nor I wanted to make fools of ourselves. We didn’t want to encounter what Altman had encountered.”

As part of Goldman’s thesis defense, two distinguished neuroscientists evaluated the song-center cells, and they, too, judged them to be neurons. Their opinions and Goldman’s lengthy scrutiny with an electron microscope left the Rockefeller researchers confident that their finding was real. Writing up the particulars for publication, they were exuberant to think that their discovery might set the old dog of dogma on its ear. Even Nottebohm, a man of decorum, admits coming close to jumping up and down. “When we had what we considered strong data that new brain cells continue to be produced in adulthood, and in the widespread manner we saw in birds,” he recounts, “I had no doubt that this was the nicest thing I’d ever do.”

They announced their exceptional news at the Society for Neuroscience’s annual meeting in the fall of ’82. Their paper followed in April, by which time the skeptics were out in full force. As far as many authorities in the neurology world were concerned, Nottebohm and Goldman’s news was too exceptional to be true.

But Nottebohm, in his genteel way, held firm. And when he spoke at conferences, he would pose the even more heretical notion that just maybe neurogenesis was taking place clear across the animal kingdom—even in the human brain. “I had little doubt that this would prove to be a general event,” he notes today. “I didn’t believe in a species inventing a trick like that and keeping it to itself.” Yet the

skeptics contended that the phenomenon that the Rockefeller team was seeing in birds must be restricted to lower species. Primates? Out of the question. They raised other concerns. How could new cells in the wall of the ventricle possibly know enough to crawl in the direction of the song center? How could the Rockefeller investigators be absolutely sure that thymidine, when injected, wasn’t labeling glial cells as well as neurons, making it seem as though there were more neurons than there really were? Most important, did their supposed neurons walk the walk and talk the talk of neurons? Did they snugly integrate into the existing neural network?

A particularly strong gust of mistrust broadsided Nottebohm and Goldman in ’85 in the form of a widely discussed paper in the prestigious journal Science. Its author, a highly respected and trusted Yale neurobiologist, was one of the two outside examiners of Goldman’s thesis who had agreed that the canary brain appeared to be making new neurons. However, upon conducting a similar investigation of his own with rhesus monkey brains, the scientist reported that he had failed to find so much as a smidgen of evidence of new neurons in these primates. A fixed population of neurons was too crucial to the workings of learning and memory to imagine that neurogenesis occurred in adult advanced vertebrates, he concluded.

As time went by, Nottebohm and Goldman—Goldman having left Nottebohm’s city lab in late 1983 to begin his residency at the Cornell-New York Hospital across the street—had reason to become even more certain of their evidence, at least in birds. In 1984, Nottebohm and his postdoc John Paton nailed proof that thymidine-tagged cells were honest-to-goodness neurons. The scientists literally impaled new cells in the brains of living canaries with ultrathin electrodes and showed that they were as excitable as neurons ought to be and capable of relaying impulses. Nottebohm today feels that this was the experiment that “brushed aside all ambiguities. Suddenly people believed what we were saying!”

In case anyone didn’t, in 1988 another of Nottebohm’s postdocs, Arturo Alvarez-Buylla, made further progress into the adult bird

brain. Astonishingly, neurogenesis wasn’t restricted to the canary’s song center. New neurons turned up throughout its forebrain, an area that included the memory-involved hippocampus and a region comparable to the mammalian cortex. Cells that likely were neural stem cells were dividing in the ventricular wall, and some of the resulting new cells were self-renewing, while others migrated to the forebrain and became neurons. This revelation wasn’t unlike Leeuwenhoek’s when he squinted through his hand lens and spied minuscule animals moving around in a pot of rainwater. “He gaped at their enormous littleness,” writes Paul de Kruif. Alvarez-Buylla admits to gaping, too, the day that he and Nottebohm were staring into a double-viewing microscope and saw a parade of small, young cells migrating toward the forebrain. This cell dispersement showed up in the brain of every songbird and nonsongbird the scientists inspected. “There were no impediments in the adult brain stopping cells from migrating long distance,” Alvarez-Buylla notes today. Was the same activity occurring in mammals? If it was, the implications for treating the human brain were tremendous.

Nottebohm would come to suspect that as new cells were added to the adult avian forebrain, older cells must be dying. Otherwise the brain would keep getting bigger and bigger. Scientists may have missed neurogenesis in adult vertebrates for this very reason; if the brain maintains a stable number of neurons, you’d have to be sitting inside a living brain with a movie camera to know that neurons were perpetually added and subtracted. “When I realized that we were dealing not only with adult neurogenesis, but with a system of constant neuronal replacement, that’s the part that I found mind-blowing,” relates Nottebohm. “We were used to having skin cells slough off and be replaced; we knew the liver could replace itself, and that the lining of the gut did the same thing. But here the brain was doing it!” Nottebohm theorized that while most of the songbird forebrain maintains a constant population of neurons, the song center grows larger or smaller due to seasonal variations that encourage or discourage its recruitment of new cells.

The Nottebohm lab would hunt high and low for proof that a bird’s ability both to sing and relearn a complete repertoire each spring was a direct result of seasonally appearing new neurons. “Common sense almost demands it,” notes Steve Goldman. Yet the only evidence so far that neurons can support a particular function is a correlation between increased numbers of neurons in a chickadee’s hippocampus and the bird’s seed-storing behavior.

Nottebohm would give a lot of thought to the basic tenets that needed reexamination because of his lab’s revelation of new neurons in the bird. How memory works remains far from understood, although the theory that gets the most votes casts the synapses in the role of the chief repository for memories—a synapse being the juncture at which an impulse from one neuron jumps across a sliver of space into the extended arm of another neuron. Nottebohm was inclined to think that if neurons were being replaced, something tied to learning and memory conceivably warranted their replacement. Here he parted ways with Trembley and came up with a theory: Maybe neurons, and not synapses, are the centerpiece of long-term learning; perhaps neurons store memory in the form of changes in their genes, with new cells encoding new memories. “Our brain, I believe, has a limited capacity for acquiring new memories because of its size, and unless you can replace cells, you might come to the end of your learning potential,” shares Nottebohm. “Neuronal replacement could represent the brain’s attempt to remain young.”

Very gradually Nottebohm’s bird work would be accepted; Altman’s rodent studies as well. “It took a long time for opinion to turn around, as long a time as it takes a ship to turn around,” says Alvarez-Buylla. Of no surprise to Nottebohm, others would uncover evidence of new brain cells being born in many vertebrate brains—mammals, even. And the sightings of brain-cell regeneration and migration prompted researchers to realize there was now tremendous reason to try to check, or even reverse, a neurodegenerative disease. Fetal tissue transplantation, in which cells from the fetal brain are grafted into an adult brain that is under siege from, say,

Parkinson’s disease, had attracteed research attention for years. But now the adult brain’s regenerative nature offered a promising different approach. “There was a tremendous rush of excitement to think that you might be able to induce neurons to be produced in the adult brain, and that you might be able to direct them to diseased areas,” describes Nottebohm. Could researchers ever achieve the trick of getting stem cells inside the brain to differentiate into the actual type of neuron that was therapeutically needed? It was as though the brain was a newly discovered planet open to exploration. Anything seemed possible.

It was the spring of ’95. Fernando and Marta Nottebohm found themselves in Athens with a few free hours between planes and decided to pay a visit to the Parthenon. They were en route to Jerusalem, where Nottebohm had been invited to give a lecture at the Hebrew University covering all the special things that canaries had shown him about the vertebrate brain. By then, researchers were reporting evidence of new neurons in the brains of more than a few adult mammals—from tree shrews to cats to marmosets—although none yet had turned up in humans. Nottebohm would tell colleagues it was much more fun in the old days, when no one believed him.

Once atop the Acropolis, the couple sat down on a stone rampart to admire the city spread out below them. No sooner had they done so when a white canary—one exactly resembling Nottebohm’s feathered friend from his boyhood—flew out of nowhere, landed beside the birdman, and began to sing. Remarks Nottebohm, who to this day remains awed by the encounter, “That was really scary!”

This page intentionally left blank.