Somewhere on the campus of Stanford University, among the stately oaks and tile-roofed temples of scientific inquiry, live Carla Shatz’s super-mice. They learn complex physical tasks more quickly than their ordinary cousins. If one eye is deprived of sight, they rapidly rewire their brains to compensate, then beat normal one-eyed mice on tests of visual acuity. They recover more readily from some brain injuries, too. What sets these rodents apart is their superior ability to form new neural connections, or strengthen existing ones, in response to experience.
Shatz is proud of her prodigies. But when I ask for permission to visit them, the pioneering neurobiologist turns me down. “They’re in a germ-free facility,” she says apologetically, glancing away from a video in which a tiny champion powers through a water maze. Only their handlers are allowed into the lab where the animals are kept, Shatz explains, and even they must shower and don sterile garments before entering.
That’s because the mice have incomplete immune systems. They’ve been bred to lack proteins — members of a family called MHCI (which stands for major histocompatibility complex class I) — crucial to fighting pathogens. The same mutation that gives them their supremely adaptive brains has left them with extraordinarily vulnerable bodies.
In the body, MHCI proteins are watchdogs, tagging infected cells for immune attack. In the brain, these proteins assume an entirely different role, helping to regulate neuroplasticity — the ability of neural circuits to reshape themselves at every stage of life. Shatz has spent more than a decade probing the latter function. Yet until she proved otherwise, few scientists thought MHCI or other so-called “immune molecules” were even present in a normally functioning brain. In fact, their absence was a basic tenet of neuroscience.
As you may recall from biology class, the brain enjoys what’s known as immune privileged status. A dense layer of cells called the blood-brain barrier protects the organ from germs circulating in the body, and from the immune cells that combat them. That is why, when you are battling the flu, neither the virus nor the inflammation directed against it spreads into your delicate cerebral neurons.
Unless the barrier is breached by injury, autoimmune disease or catastrophic infection, rarely does a T cell, a B cell or any of the immune system’s other shock troops get through. It was long believed that the immune system’s molecular sentinels — molecules like MHCI — were missing from within the brain, too.
Then, 15 years ago, Shatz stumbled across genes coding for MHCI in fetal cat brains. She soon unearthed MHCI proteins and receptors (molecules that bind with MHCI) in the unimpaired noggins of mice. They eventually turned up in healthy monkey and human brains as well. And as Shatz devoted herself to investigating MHCI and its entourage in those surprising places, she sparked a neuroscientific revolution.
Today, a growing number of researchers are examining the complex ways in which immune molecules affect the brain and nervous system. Manipulating such molecules, these scientists believe, may be key to treating many devastating neurological ailments, from autism and schizophrenia to Alzheimer’s and ALS. Shatz even dreams of a “plasticity pill” to restore the neural suppleness of stroke victims — and her latest experiments offer hope that it could someday come to pass.
“People thought we were crazy when we made this discovery,” says Shatz, a slender 66-year-old with high cheekbones and a halo of dark curls. “But my parents gave me some advice when I was young. They said, ‘Don’t worry about what other people think of you.’ ” She gazes pensively at the computer screen. “I probably generalized it too much.”
There are no conventional family memorabilia in Shatz’s office, a few steps from her cavernous lab. The space is sleekly functional, with louvered glass walls, and the photos on the shelves are of her professional kin: mentors, colleagues, former students.
One framed artwork, however, connects Shatz to her childhood, and the earliest influences on her own brain circuitry. It is an etching by her late mother, an artist who loved science. The image, a mysterious, gray blob surrounding a spot of intense red, was based on an electron-microscope photo of a neuromuscular junction (the connection between a nerve cell and a muscle) clipped from a magazine sometime in the 1980s.
“She was a very serious painter,” Shatz says, “and she really understood biology. I think she would have made a great MD. But she stayed home and brought us up — it was that generation.” Shatz’s father was an aeronautical engineer who helped design the guidance system for the Apollo 13 lunar module.
Growing up in West Hartford, Conn., Shatz absorbed her mother’s passion for the visual arts. But she was also, as she puts it, a “science nerd” — a gender-radical stance for a girl who entered her teens just before Eisenhower left office. Her parents encouraged her to follow her interests, no matter how unorthodox or disparate.
Shatz followed them to Radcliffe College, where she majored in chemistry but also took classes in design. A course with biochemist George Wald, who had won a Nobel Prize for his work on how the eye perceives color, helped spark her interest in studying the brain mechanisms behind vision. “I realized that I could have my cake and eat it too,” recalls Shatz. After she told her adviser she wanted to write her honors thesis on how people see, he sent her across the river to Harvard Medical School, where a pair of ambitious scientists, David Hubel and Torsten Wiesel, were investigating the plasticity of the visual system.
She spent a year studying with the duo, whose work eventually won them their own Nobel. Hubel and Wiesel were already renowned for having charted the architecture of the primary visual cortex (the area of the cerebral cortex, at the back of the head, that receives input from the retinas).
As part of this work, they found the region has a pattern of zebralike stripes — dubbed “ocular dominance columns” — made up of neurons that process information from the right eye, the left eye or from both. In the aftermath of the discovery, their focus was finding whether that structured pattern was determined entirely by genes, or whether experience also played a role.
Hubel and Wiesel wanted to learn, among other things, why a child with a cataract — unlike an adult — may lose sight permanently in the affected eye if the obstruction isn’t removed promptly. To find out, they deprived cats and kittens of sight in one eye for prolonged periods.
In kittens with one eye sutured shut, they discovered, the ocular dominance columns changed radically. The stripes devoted to the sighted eye expanded, while those devoted to the obstructed eye shriveled. These changes occurred only during a “critical period,” the scientists observed, when the young animal’s neural circuitry was still developing. In adult cats, as in humans older than 6, such plasticity was greatly reduced.
Shatz witnessed some of these landmark experiments and they intrigued her. When she graduated in 1969, she knew she wanted to explore the eye-brain connection further. But how? Her two doctor uncles urged her to go on to medical school, but Shatz wasn’t sure.
A few years earlier, her paternal grandmother — a brilliant woman, the first in her family to go to college — had been crippled and made mute by a stroke. “Here were my uncles, both of them neurologists, and they couldn’t do a damn thing for her,” Shatz recalls thinking. “There wasn’t enough research.” She decided to become a neurobiologist.
In 1976, after returning to Hubel and Wiesel’s lab for doctoral studies, she became the first woman to receive a Ph.D. in neurobiology from Harvard. In 1978, she was hired by Stanford, soon becoming the first woman to become a tenured professor there in basic science. Commandeering a lab of her own, she set out to expand on her mentors’ work, and she wound up wandering into utterly new terrain.
Fire and Wire
Hubel and Wiesel had uncovered the principles of how neuroplasticity works in the brain: Basic neural architecture is hardwired. For instance, the eye is genetically programmed to connect with the visual, not the auditory, part of the brain. Fine-tuning that circuitry — connecting the eye to a specific part of the visual cortex — is shaped by experience.
To understand how this works, it helps to remember a bit more basic biology. When neurons are stimulated, they spike with enough electric current to send neurotransmitters across a tiny gap called a synapse, where other neurons, in turn, are similarly stimulated and provoked. As the signal passes down the line, entire circuits are put in play.
Neurons that handle lots of robust, well-synchronized signals sprout more neurotransmitter-generated terminals, and their connections with other neurons along the signaling pathway grow stronger. (Neuroscientists have a slogan for this: “Cells that fire together wire together.”) Synapses that transmit few, weak or out-of-sync signals are pruned away. In this way, brain circuits are remodeled with use.
By the late 1970s, researchers had learned much about how the process plays out in the mammalian visual system. Genetically programmed molecules guide embryonic nerve fibers from the light-sensitive retina toward the visual cortex, where images are perceived. At first, the connections are highly approximate; the stripes of the ocular dominance columns are partially formed in humans and other mammals at birth. Then, input from the eyes helps refine the neural pathways until, over a period of months to years depending on the species, the columns mature and stabilize.
But major questions remained. Where, exactly, did the influence of genes leave off and experience begin? What biochemical processes allowed sensory input to change the brain’s wiring? Why was neuroplasticity greater in juveniles than in adults, not only in the visual system but other areas as well?
Searching for answers, Shatz decided to focus on the lateral geniculate nucleus (LGN), a clump of tissue shaped like a piece of elbow macaroni, set behind each eyeball, that serves as a relay station shunting visual signals to the ocular dominance columns.
Layers of the LGN start forming around the 47th day of gestation in fetal cats. The layers are almost completely formed by the time the animal is born, around day 60. It had long been assumed that this process was hardwired until the animal opened its eyes, at which point experience finished the job. Shatz, however, suspected that the layering might be guided by spontaneous nerve impulses from the retina when the animal was still in utero.
In 1988, she tested that hypothesis on cat fetuses using a poison called tetrodotoxin, which prevents nerve impulses from firing. On the 42nd day of gestation, she surgically removed the fetuses, fitted them with tiny tanks that pumped the neurotoxin into their brains, then returned them to the uterus. (“Carla has good hands in the lab,” recalls Stanford neurobiologist Sue McConnell, then one of Shatz’s postdocs.)
Outwardly, the fetuses continued to develop normally. But when their brains were examined two weeks later, Shatz’s hunch proved correct: The LGN’s layers had failed to develop.
Shatz now had indirect evidence that the layers were formed in response to retinal signaling, coded for not by experience but by genes. But it wasn’t until three years later that she was able to observe the signaling in action. Working with her colleague Denis Baylor, she harvested retinas from fetal cats and ferrets, putting each piece in a dish of nutrient medium to keep it alive. Then she placed each piece of retina over a grid of electrodes wired to sense collective neural activity.
The electrodes detected waves of synchronized nerve impulses sweeping through the retinal tissue, in patches of up to 100 cells at a time. “It was like neighborhoods of nerve cells in the eye were placing phone calls to the brain to check connections,” Shatz recalls.
In a developing fetus, she surmised, those phone calls would reach groups of nerve cells in the LGN; as their synaptic connections with the retinal neurons strengthened, the LGN neurons would begin forming “area codes” of their own.
Meanwhile, synapses with poor connections would be trimmed away. After birth, visual input would further sort out the LGN’s wiring, until layers associated with right and left eyes were complete. It would take much more research, however, to determine exactly how the process worked.