In memoriam: Yves Frégnac, influential and visionary French neuroscientist

Adrien Peyrache first encountered Yves Frégnac some 20 years ago. Peyrache, now associate professor at McGill University, had signed up for a graduate neuroscience class about the visual system, taught at École Normale Supérieure in Paris on Saturday mornings. On the first day of class, Frégnac showed up late and disheveled. Peyrache was ready to dismiss both the class and the teacher.

“Then he started talking, and that was it,” Peyrache recalls. He found the material mesmerizing. “It changed my life forever.”

Frédéric Chavane, director of the Institut de Neurosciences de la Timone, was also enthralled by Frégnac’s teaching more than 30 years ago. But Chavane figured only 5 percent of the class could truly comprehend what Frégnac was saying. A few years later, when Chavane was training under Frégnac for his Ph.D., he asked about the difficult lectures and was told they were hard “on purpose.” Frégnac wanted to show students the full complexity of neuroscience—if the 5 percent who persevered should come and work with him later (as Chavane did), that was even better, Frégnac said.

Frégnac, who died on 18 October at the age of 73, built his career by meeting neuroscience’s complexity straight on. If he saw a deviation in the neuronal responses under his electrode, or if he doubted an established view in the field, he leaned into exploration. Over his 40-year career, Frégnac used experiments and computational modeling to study the early visual processing, helping to refine our understanding of it, his colleagues say.

He also changed the nature of neuroscience research in France. “He was really a pioneer of system neuroscience” in the country, says Brice Bathellier, director of the research team on auditory system dynamics and multisensory processing at the Institut Pasteur. Frégnac “defended the idea that we need to develop a more computational and a more systems-scale level of analysis in our science,” Bathellier says. “Today, of course, there is no question about it, but 25 years ago it was not obvious at all.”

But Frégnac is perhaps best known for combining experimental, computational and theoretical approaches to study the brain, and instilling that philosophy in a generation of researchers. That, says Gilles Laurent, director of the Max Planck Institute for Brain Research, is “one of his great legacies.”

F

régnac grew up in Lyon, but in 1970 he moved to Paris to study engineering at the prestigious Supélec school (École Supérieure d’Electricité). There, he ran the students’ sports office, wrote a cinema column for the school’s newspaper, and interned at the Flight Test Center in Istres, where he rode in a Fouga Magister fighter plane.

But after two years of the program, Frégnac asked for a break to study biology. This “double curriculum” was his idea, says his daughter Katherine Frégnac, and though his plan was “innovative at the time,” Frégnac managed to convince his school to allow it.

By graduation, he had earned a master’s degree in biophysics, and also diplomas in modeling and neuroscience. Yet it was the recommendation from one of his professors that Frégnac visit Michel Imbert at the Collège de France that set his career trajectory in motion.

Frégnac later recalled the visit in an essay in the annual magazine of l’Académie François Bourdon. “I entered a dark room,” he wrote. “Bright stimuli were moving in all directions on a screen. Shrill sounds like machine-gun bursts were mixed with the voices of the researchers.”

He had found a group of “explorers of the brain,” he wrote, who were recording the electrical activity of neurons in the visual cortex of a mammal, using tungsten electrodes only a few micrometers thick. The staccato noises were biological signals converted to sound, an audio version of the brain’s electricity.

The visit was a revelation, he recalled. Frégnac stayed and worked with Imbert, listening to the brain’s electrical chatter by day and often having late-night discussions with Imbert and Nobel laureate physicists also interested in the brain. Frégnac knew he had found his life’s work, and he later called those years “pure joy.”

U

nder Imbert, Frégnac obtained a Ph.D. in human biology in 1978 and in neuroscience in 1982. He published papers on how visual neurons develop their selectivity to visual features, and after finishing his program he published a massive review of the development of feature selectivity in the visual system, which became (and remains) an essential reference in the field, Chavane says. “And this, imagine, is done by a guy just getting out of a PhD.”

Frégnac next launched his own team at the Institut de Neurobiologie Alfred Fessard in Gif-sur-Yvette, just south of Paris, and began investigating existing hypotheses in vision neuroscience, probing the work of pioneers such as David Hubel and Torsten Wiesel. Working with collaborators, Frégnac published landmark work on cortical plasticity and Hebbian learning mechanisms in the primary visual cortex. In a series of experiments, he and his team demonstrated that by repeatedly modifying the response of neurons to visual stimuli, they could change the neurons’ functional profile, such as their orientation preference or the nature of their receptive fields.

In the mid-1990s, intracellular recording arose as a new but difficult technique to measure the subthreshold activity of neurons, and this caught Frégnac’s attention. These subtle “whispers,” or “rumors,” as Frégnac sometimes called them, revealed a dialogue between the neurons of the primary visual cortex, suggesting this layer has a higher processing ability than was imagined by the “feedforward” model of vision—the idea that visual information is processed serially as it advances through the layers of the visual cortex.

Frégnac and his colleagues devised methods to amplify inhibitory signals, becoming among the first to confirm the existence in the primary visual cortex of a “shunting inhibition”—a debated nonlinear inhibitory process—and show that it helps shape the response properties of neurons. This research was “particularly heroic, I think,” Laurent says. It was a clever and difficult study, but it also “surprised a number of people and ruffled a few feathers,” Laurent says, and so “it didn’t get the reception that it should have done initially, but eventually did.”

I

n 2000, Frégnac created a research unit at the Centre National de la Recherche Scientifique campus in Gif-sur-Yvette. He began to bring together an interdisciplinary group of biologists, physicists and engineers. The initial cast included a physicist from Canada, a neurophysiologist who had just developed “sleeping” brain slices in ferrets at Yale University, and Kirsty Grant, a specialist in electric fish who was, by then, his wife.

The group, initially called the Unit of Integrative and Computational Neuroscience but now called the Unit of Neuroscience Information and Complexity (UNIC), gained a reputation as a place for collaboration. Inside the unit’s walls, students and postdoctoral researchers from computational and experimental fields moved “seamlessly between teams,” as described in an 2014 report by an independent group of experts, and the science the group produced was often of “outstanding quality and originality, intrinsically interdisciplinary in a way rarely seen anywhere in the world.”

In the 2000s, Frégnac helped develop the neuromorphic computing field and spawned international collaborations such as SenseMaker, Facets and BrainScaleS. These initiatives were precursors to his involvement with The Human Brain Project (HBP), the ambitious but troubled effort to better understand the brain by modeling it in a computer.

Frégnac was initially an enthusiastic supporter of the HBP. The initiative seemed like a natural extension of the multidisciplinary research he’d already engaged in, and he wanted to ensure the French neuroscience community participated in the project. But the HBP was soon criticized for changes in its focus and for cutting out its experimental and cognitive neuroscience arm. Frégnac added his voice to the criticism, outlining his views in a 2014 Nature op-ed co-authored by Laurent. By the time the HBP ended in 2023, Frégnac had reached full disappointment. He saw its scientific output as “fragmented and mosaic-like,” as he said in a Nature article last year, and he felt the project did not provide new understanding of the brain.

Frégnac met France’s mandatory retirement age in 2017, and he stepped down from directing the lab. He continued to give talks at conferences and taught courses, including a class called “Brain and Cognition” at his engineering alma mater, now called CentraleSupélec. He kept in touch with his colleagues over dinner, often picking up the check. Frégnac had a decades-long interest in photography (he had his first photography exhibit in 1978), and in his semi-retirement, he spent more time taking photographs. “He would most probably have become a professional photographer if not a scientist,” Katherine Frégnac says.

Yet it was his ability to think critically that held together his scientific relationships. Terrence Sejnowski, professor and laboratory head of the Computational Neurobiology Laboratory at the Salk Institute for Biological Studies, knew Frégnac for decades and found him to be “a very thoughtful, very deep thinker,” he says. “I could always go and get really good opinions from him, no matter what question I had.”

This October, at the annual Society for Neuroscience meeting in Chicago, Frégnac met with Martin Baum, senior acquisitions editor at Oxford University Press. Frégnac had been working on a critical review of modern-day neuroscience, including initiatives that he was involved in, such as the Human Brain Project and the Blue Brain Project. Tentatively titled “Manifesto: In Praise and Defence of a Better Science of Brain and Mind,” the book was only 60 percent finished. The two men grabbed a couple of chairs in the exhibit hall and discussed the manuscript.

Once Baum had flown back to Oxford, and Frégnac to France, Baum got the book approved for publication. He sent Frégnac a contract on 20 October, but he never received a response. Later that day, Baum learned that Frégnac had died unexpectedly at his home in Gif-sur-Yvette. “I was really shocked by this news,” Baum says. “I’m sure it would have been an exciting book. It might have—who knows?—might have gotten some people up the wrong way, but he wasn’t too worried about that.”