Professor of Neurobiology Stanford University School of Medicine Member of Stanford Neurosciences Institute
In his first two years of graduate school, Ben Barres (then Barbara Barres) was a neuroscience graduate student by day and a neurologist on nights and weekends. He began his PhD at Harvard Medical School in 1983 after completing full medical training: an MD from Dartmouth Medical School followed by four years of a neurology-focused residency at Cornell University hospitals in New York City.

“When I started graduate school, all of my medical school loans came due. I had to start paying them back, but my graduate-student stipend was $6,000—barely enough to live on. So I started to moonlight as a neurologist. I was covering a neurology practice at a local hospital, seeing patients and covering emergencies Friday night to Monday morning. On Monday mornings I would drag myself to the lab. It was very hard to take courses, be in the lab, and to work as a neurologist. After about two years, I came in Monday morning and my advisor saw how bad I looked and asked why I was doing this to myself. I told him that I had loans to pay and couldn’t on my stipend alone. And he said, ‘Oh, is that all? That’s the problem? That’s why you’re working?’ And I said, ‘Yes.’” Barres’s advisor was David Corey, then a young investigator who had Howard Hughes Medical Institute funding. “He had slots to pay postdocs and I was still the only one in his lab, so he said to me ‘Since you have an MD, even though you are a graduate student, you can also be a postdoc and I can pay you a postdoc salary,’ which was $17,000 at the time. ‘Would that be enough for you to quit the neurologist job? Will that work for you?’ And I said ‘Yes!’ I immediately started sleeping at night and doing well in the lab and enjoying life because I was not exhausted all the time. I realized that I am not a good multitasker. That was an early insight: it was not going to work for me to simultaneously be a clinician and a researcher. But I was glad I did the neurology training. My work on glia came directly from my medical training, but I have never tried to practice medicine again. And that was David Corey for you, the world’s best graduate advisor. He is wonderful and I am very lucky to have ended up in his lab.”

Over the last 25 years, Barres’s research on glia has fundamentally changed the image of these cells—highly abundant non-neuronal cells in the brain that were previously relegated to a supportive and structural role (glia means ‘glue’ in Greek)—delineating the important roles they play in how the brain functions in sickness and in health.

Here, Barres describes how his career path in science began when he was a six-year-old, talks about what ignited his interest in neurobiology, and tells how he changed his life at age 42.


Budding scientist. Barres was born in 1954 and raised in West Orange, New Jersey. “When I was about six years old, I remember I decided to be a scientist and my fraternal twin sister decided to be a nurse, and I went on to become a scientist and she became a nurse. There were no nurses or scientists in our family, so who knows where we got this idea,” he says. “I was a young geek; I was interested in science, period.” Computers and computer programming were just entering the mainstream when Barres started high school, and he tried to get as much experience as possible using these early computers and writing code for them. During high school and college, he spent his summers working at Bell Labs in New Jersey and took part in a science honors program at Columbia University, in which the university’s professors taught weekend science and computer classes to high school students.

Sexism encounters. At age 13, Barres decided he would attend MIT. “Don’t ask me why. I think I met someone that I admired who went to MIT,” he says. As a high school senior, he applied for early decision and was accepted. He entered MIT in 1972 thinking he would major in computer science. “I had a take-home, five-question final exam due at midnight that took me all day. I solved the last question late at night; I just suddenly saw the answer. The next day, the professor passed back the exams and said that no one had solved the last question. I went to him after class and told him that I had solved the problem and showed him my paper. He looked at me with disdain and said, ‘Your boyfriend probably solved it for you.’ He just couldn’t imagine, in 1973, that a woman could solve a problem that hundreds of men couldn’t solve. I was kind of indignant that he accused me of cheating, but it really didn’t occur to me until years later that it was sexism. I didn’t really think about those things then. I saw myself as a guy and felt that I was a guy inside, even though I was a woman. So I was a bit oblivious to stuff like that.” Barres also had a hard time getting into a lab to do an undergraduate thesis project, even though his male counterparts had no trouble finding a professor mentor. He did eventually join the biochemistry laboratory of Maria Linder, one of the few female members of MIT’s science faculty at the time. “Still, I loved it at MIT. I had a great time. The most famous faculty teach undergrad courses there. I had Salvador Luria, who had a Nobel Prize, as my first biology professor. They all radiated such passion and talked about their latest research in class. I loved science when I came in and still loved it when I came out, and that’s all that mattered.”

“Once I get in the lab, I can’t leave, I just get so excited. It’s an addiction. I will always choose doing experiments over sleeping.”

Neuroscience spark. As a sophomore, Barres took a course called Psychology and the Brain, taught by neuropsychologist Hans-Lukas Teuber. It was 1973 and the term “neurobiology” had not yet been coined. “He talked about figuring out what parts of the brain did what and just hooked me. That was when I got the idea to become both a neurologist and a neuroscientist,” he says. Barres switched his major from computer science to premed. His goal was to get his medical degree first and follow it with a doctorate in neurology. After graduating in 1976, Barres entered Dartmouth Medical School. “I started with medical school first, which turned out to be fortuitous for me. When I started the PhD program at Harvard, I was more mature about the way I studied and more focused. I had decided to do a PhD rather than a specialized postdoc after residency because I didn’t feel I had broad enough training in neurobiology, which was just starting to explode,” he says. But before Harvard, Barres completed four years of a residency program in neurology and became a certified neurologist. “Although my plan was to simultaneously practice and do research, I was much more aware after my residency that there was very little that neurologists could do to help their patients. There were few actual treatments for the conditions I was seeing, so it was becoming less appealing to me to actually practice neurology.”

First glimpses. In Corey’s lab at Harvard, Barres learned how to patch clamp and made recordings from glial cells, which happened because Corey was just setting up his own lab there and Barres had just learned how to culture rat-derived glial cells in his prior rotation. “The patch clamp was the first time you could do good recordings, because glia are such tiny cells.” At the time, glial cells were thought to be passive neuron-supporting cells. But Barres began to notice that different types of glial cells had different types of ion channels. He had become captivated with these highly abundant brain cells back in medical school because their role in the normal brain was a mystery. While in Corey’s lab, Barres cranked out six publications, including five in Neuron. In 1988, he and Corey were among the first to show that glial cells do indeed have ion channels. Barres also developed an antibody-based technique, called panning, to purify glial cells, which demonstrated a range of glial cell types in the rat brain. “I had these beautiful glial cells to record from, but we couldn’t keep them alive in culture for more than a few hours because the necessary growth factors had not been identified yet.”


Moving on. Barres first encountered his postdoc advisor while poring over the literature on glial cells. “It was all dreck, really descriptive and not clear, except for Martin Raff’s work. He was defining glial-cell markers.” Barres met Raff during a visit Raff made to neighboring MIT, and after the British researcher served as Barres’s unofficial second graduate advisor, Barres joined Raff’s University College London lab in 1990. There, he adapted his panning technique to purify each of the major classes of glial cells, including astrocytes and oligodendrocytes. His motivation was to combine glia and neurons in culture to begin to tease apart their interactions. Barres went on to show that about half of oligodendrocytes in the rat optic nerve die during development and to identify the growth factors necessary for their survival in culture. He also demonstrated that the point of glial apoptosis was to form a one-to-one match with a myelinated axon.

More than glue. Barres moved to Stanford to set up his own lab in 1993. As a graduate student and postdoc, he had devised tools to purify and culture glia. In his own lab, he began to study how glial cells communicate with neurons, and to what end. Barres and then postdoc Frank Pfrieger showed that, in vitro, glial cells are necessary for the formation of functional synapses between neurons. Then in 2001, Barres’s lab found that neurons need glial cells to make and stabilize mature synapses in vivo as well. “That was a complete surprise. The dogma was that neurons intrinsically have all of the machinery to form synapses,” he says.

Communication enablers. Barres was on his way to studying the function of glia cells in vivo, but first, he needed to get a handle on some of the molecules glial cells used to communicate with neurons. “If we didn’t know the glia-secreted molecules, we couldn’t do a knockout mouse and ask what glial cells are doing in vivo. The purified cell experiments were tools to generate hypotheses that we could test in vivo,” says Barres. In 2005, the lab showed that two thrombospondins, glycoproteins secreted by immature astrocytes in the developing brain, promote synapse formation both in culture and in vivo, and that thrombospondin was sufficient to induce in vitro formation of neuronal excitatory synapses—structurally complete and presynaptically active, but postsynaptically silent—even in the absence of astrocytes. Then, in 2012, Nicola Allen, a postdoc in the lab, found that two other astrocyte-secreted molecules, glypican 4 and 6, are necessary for neurons to form fully functional glutamate receptor–dependent synapses. The lab and Barres’s former students are still working to understand the other molecules that work together to sculpt neuronal synapses.

Smart glia. Another function of astrocytes, discovered by then postdoc Won-Suk Chung, was that astrocytes eat synapses by phagocytosis not only during developmental pruning, but also in the adult brain. “Astrocytes are sensing neuronal activity and making decisions about which synapses to eat or not eat, and we think this implies a critical role for astrocytes in synaptic plasticity that underlies experience. It’s another demonstration of how smart glia are and the remodeling and restructuring that goes on in the brain,” says Barres.

Glia gone bad. Most recently, Barres’s lab showed that astrocytes can go rogue. Aberrant astrocytes in the mouse brain, rather than promoting neuronal connections, induce death of other types of glial cells and of neurons. The team detected this type of astrocyte activity in brain samples from patients with multiple sclerosis, amyotrophic lateral sclerosis, Alzheimer’s and Parkinson’s diseases, and in individuals with brain injury. Barres is now trying to address how these rogue cells arise, what neurotoxins they secrete, and how they may be involved in neurodegenerative diseases. “We now have some evidence that it is neuronal sickness or injury that induces these astrocytes, and the implication is that this glial reaction may be partly causing the degeneration in the brain. We haven’t proven that but that’s the next exciting paper!”


Being different. “I was at Stanford with my own lab for two years before I changed sex. But I had been confused about my gender from when I was a little kid, maybe even 3 years old. I knew there was something different about me and I was confused and ashamed about what it was. I never discussed it with anyone until I decided to change sex. But my parents must have been aware of it, because every Halloween I was dressing as a football player or an army man, and whenever I was allowed to choose my dress, I was dressing as a guy. And I am sure my parents thought ‘What is going on with this kid?’ But we never discussed it.”

An eye on human disease. Stemming from his medical and graduate school experiences, Barres created and is the director of the Masters of Science in Medicine program at Stanford, with the goal of exposing PhD students in basic science to clinical medicine. “Twenty, twenty-five years ago, basic scientists were not expected to work on disease. Studying disease was considered a second-class scientific activity. The focus in graduate programs is on the model systems, and we never gave them the tools to study human disease. We want to enable researchers to study human diseases. [The program] teaches students the language, anatomy, and pathology, and the major questions in the field we have, which leads to curiosity,” says Barres.

Lab addict. “Once I get in the lab, I can’t leave, I just get so excited. It’s an addiction. I will always choose doing experiments over sleeping. I would still probably be in Corey’s lab if he hadn’t stopped paying me.” 

Greatest Hits

  • Devised a purification technique called panning to isolate rat ganglion cells and then different types of rodent glial cells
  • ·Demonstrated that glial cells are necessary for neuronal formation of functional synapses and to maintain synaptic function in culture
  • ·Identified the important synapse-inducing molecules secreted by astrocytes, including thrombospondins
  • Found that microglia and astrocytes phagocytose synapses both during development and in the mature brain, providing evidence that glia are involved in synaptic plasticity throughout life
  • Found that a type of highly neurotoxic “reactive” astrocyte is generated after acute brain injury and in neurodegenerative diseases