COURTESY OF STANFORD UNIVERSITYIn November 1986, Emmanuel Mignot arrived at Stanford University’s Center for Sleep Sciences and Medicine for a 16-month stint as a research associate. His goal was to find effective drugs to treat narcolepsy; his study subjects belonged to a colony of canines that suffered from the malady. “[When I got there], the dogs were being maintained, but not much was being done with them other than some chemistry studies on known neurotransmitters,” says Mignot, a professor of psychiatry and behavioral sciences at Stanford University and now director of the center. “As a pharmacologist, I wanted to study potential treatments for narcolepsy and understand the molecular biology to improve treatment in humans.”
The first narcoleptic dog, a French poodle named Monique, was brought to Stanford in 1974 by William Dement, the so-called “father of sleep medicine,” who had founded the center in 1970, the first in the world dedicated to the study of sleep. Dement and other researchers there established a full breeding colony in 1977 when dogs with a genetic form of the neurological disorder were discovered—initially, some puppies from a litter of Dobermans and, later, some Labradors. Narcoleptic dogs and humans both exhibit a combination of symptoms: perpetual sleepiness, cataplexy—muscle paralysis attacks triggered by emotions—and abnormal rapid eye movement (REM) sleep. While the condition in humans and dogs is treatable, there is no cure.
To study which narcolepsy drugs increased wakefulness and decreased cataplexy in the dogs, Mignot and psychiatry professor Seiji Nishino used a food-elicited cataplexy test: administration of the drug followed by release into a room with pieces of food on the floor and careful observation. “The dog would rush into the room and be so happy to eat the treats, and then would have an attack and collapse on the floor.” The researchers counted the number and duration of the attacks after treatment with a drug at various doses. In humans, cataplexy episodes are triggered by a positive emotion such as laughter at a joke or pleasant surprise. “For the dogs, it is food or the joy of playing. That is what is great about dogs as a model for this condition. When you give a treatment to a rat or mouse and they stop having cataplexy, you really don’t know if it is because they don’t feel good or if it is a genuine effect. But the dogs show you emotions like humans. I knew all of these dogs by name. They were my friends. I could see if they were worried or didn’t feel well.”
Mignot worked mostly with the Dobermans and Labs, but there were also dogs donated to the colony that seemed to have a sporadic form of narcolepsy, “There was Vern, a miniature poodle; Wally, a big poodle; Tucker, a mutt; and Beau, my beloved dachshund.” Using the cataplexy test in animals along with in vitro studies of the drugs’ chemical properties, Mignot and Nishino found that antidepressants suppress cataplexy by inhibiting adrenergic reuptake, and that amphetamine-like stimulants promote wakefulness in narcoleptics by increasing the availability of dopamine. “We improved the then-current treatments and started to understand the kinds of chemicals important to regulate narcolepsy symptoms.”
“I want to study the genetics of 40,000 people with sleep issues. . . . I think this will help us crack open the mysteries of sleep.”
But Mignot wanted to understand the molecular mechanism of narcolepsy, so he turned his focus to the genetic basis of the disorder. A lack of genetics training and no map of the dog genome to guide him did not deter Mignot. He has tirelessly pursued this previously little-studied and, so far, only known neurological disorder that fundamentally perturbs the nature of sleep states.
Here, Mignot talks about pursuing a master’s, PhD, and MD simultaneously, the paper retraction that has been the most difficult episode in his career so far, and his unexpected devotion to a Chihuahua.
Sir Mix-a-Lot. The youngest of six siblings, Mignot had a penchant for collecting fossils and for conducting chemistry experiments in the bathroom of his family’s home in Paris. “I bought chemicals sold by a Chinese shopkeeper on Rue Saint-Dominique to do all kinds of experiments, mixed them, and occasionally made mistakes. There were burn marks and projections on the walls of my bathroom.” In high school, the self-proclaimed “nerd with glasses” became interested in biology, and, after graduation in 1977, went to study for a medical degree at the René Descartes University Faculty of Medicine in Paris.
Collecting degrees. “In the second year of medical school, I got bored from all of the memorization.” He took the entrance exam for the prestigious École Normale Supérieure (ENS), which gives students freedom to pursue their academic interests at other institutions while providing a stipend, housing, and the support of professor mentors. He passed, and entered the ENS in 1979. Mignot worked towards a master’s in biochemistry, and then a PhD in molecular pharmacology while still continuing his medical studies. “Nothing was set up for MD-PhD programs at the time. It was all in parallel, which was crazy. I had an exam every few weeks,” says Mignot. In 1984, he received both his medical degree and, later, a PhD from Pierre and Marie Curie University.
New to narcolepsy. Mignot became interested in the effects of drugs on the brains of psychiatric patients, studying how different compounds affected the metabolism of neurotransmitters in the brains of rats, and pursued a residency in psychiatry to complement his laboratory research. In 1986, he was offered a professorship in pharmacology at the Paris V University School of Medicine. But first, Mignot needed to complete the mandatory military service that he had deferred. “Instead of going to a former French colony to practice medicine, I convinced the French government to send me to Stanford to study modafinil, a wakefulness-promoting drug created by a French pharmaceutical company called Lafon Laboratories for the treatment of narcolepsy. I had never heard about [narcolepsy] during medical school—it must have been a single line in my textbooks. I discovered that Stanford was doing work on sleep and that Dement had started a colony of narcoleptic dogs there. I thought I could study these animals and figure out how modafinil worked.”
So Mignot came to Stanford for 16 months as part of his military service with financial support from Lafon Laboratories. “The company had claimed modafinil worked by a novel mechanism, unrelated to how stimulants work,” says Mignot. But Mignot found that modafinil bound the dopamine transporter, inhibiting the reuptake of the neurotransmitter, boosting wakefulness. “This is a similar mode of action as Ritalin, but the company was claiming otherwise. It took 10 years for my results to be validated, finally, by Nora D. Volkow, now director of the National Institute on Drug Abuse, who showed . . . that indeed the drug displaces the dopamine transporter at doses that increase wakefulness in humans.”
Going to the dogs. At Stanford, Mignot immersed himself in his work with the dog colony. “I worked all the time and came home just to sleep. I was definitely not very successful with girls then, because I smelled like dog all the time. I spent all day with the dogs, going to the facility, hugging, playing, and working with them. When we bred them, sometimes the mothers rejected their puppies so we had to come in every few hours, even in the middle of the night, to bottle-feed the puppies. Even after I took a shower, you could still smell the dogs. It was a strange part of my life.”
From pharmacology to genetics. Mignot kept extending his stay at Stanford. “After a few years I realized our pharmacology studies were never going to lead to narcolepsy’s cause. We needed to find the genetic cause in the dog.” In 1988, he resigned a faculty position in Paris—which was being held for him even as he continued to extend his time at Stanford—deciding to search for the mutated gene responsible for narcolepsy in dogs. In 1993, Mignot became the head of the Center for Narcolepsy at Stanford. A connection between an immune gene, the human leukocyte antigen (HLA) allele HLA-DR2, and narcolepsy in humans had already been identified by Yutaka Honda at the University of Tokyo, so Mignot’s lab tried to ascertain whether the same connection was true in the dogs or if the immune gene was simply a genetic linkage marker. These were the days before the dog or human genome had been sequenced, so the work took Mignot’s lab 10 years, and almost 200 narcoleptic Dobermans and Labradors: years of painstaking chromosome walking experiments, DNA fingerprinting, and the construction of a bacterial artificial chromosome library of dog genomic pieces. “What helped us a lot was that we knew the Dobermans and Labs had the same genetic defect because we interbred and got narcoleptic puppies—what’s called a complementation test.” In 1999, Mignot’s team identified the mutated gene as hypocretin receptor 2, whose protein binds hypocretin (also called orexin), a neuropeptide that regulates arousal and wakefulness. Several weeks later, after seeing these findings, Masashi Yanagisawa’s lab independently published a confirmation, showing that hypocretin knockout mice also have narcolepsy.
In parallel narcolepsy studies across ethnic groups, Mignot’s lab found that it was not the initial HLA-DR2 allele that predisposed humans to narcolepsy, but another, nearby HLA gene, DQB1*0602.
Humans are not like dogs. “After we found the gene, the research went fast. We decided to look at hypocretin itself and see if it’s abnormal in humans.” Mignot’s lab sequenced the genes for the hypocretin receptor and its ligand in narcoleptic patients, expecting mutations in either to be rare because of the known HLA-narcolepsy linkage and the fact that most cases in humans, unlike in dogs, are not familial. Only one documented case, a child who had narcolepsy onset at six months of age, has been found to harbor a hypocretin gene mutation. “I think you need to knock out both receptor 1 and 2 in humans to get the full narcoleptic phenotype,” says Mignot. “Those with just one mutation may be more prone to tiredness but not full narcolepsy.”
In 2000, Mignot’s and Nishino’s groups reported that hypocretin was not present in narcoleptic patients’ cerebrospinal fluid—a test still used diagnostically today. The same year, independent studies from Mignot’s laboratory and that of Jerome Siegel at the University of California, Los Angeles, found that the lack of hypocretin was not due to gene mutations but to the fact hypocretin cells were missing in the brains of narcoleptic patients. HLA genes were well known to be associated with many autoimmune diseases, and Mignot hypothesized that hypocretin was missing due to an autoimmune attack against hypocretin-secreting neurons. What the abnormality is in those narcolepsy patients with normal hypocretin levels remains a mystery.
Mignot Moves Forward
Still a missing link. “I have been working on this [autoimmunity] hypothesis for 10 years, and we see that this hypothesis is more and more likely, but we cannot find any direct proof. It’s frustrating, but that kind of struggle is the story of my life.” All known autoimmune diseases result in the generation of antibodies in patients, but antibodies against hypocretin or the hypocretin cells have never been detected. So Mignot’s lab tested whether T-cells were the immune component attacking hypocretin. In 2013, his lab published a study identifying the T-cell culprits. But the study was retracted by Mignot himself one year later, when Mignot’s group couldn’t reproduce the results after the scientist who did most of the experiments had left the lab. “It was really painful and the worst time in my career.”
A new lead. “In 2010, a lot of people suddenly started to develop narcolepsy after receiving the Pandemrix vaccine against swine flu. It’s very odd. We still don’t understand why this particular vaccine increased the risk of narcolepsy.” Mignot thinks that a component of the vaccine or the virus itself triggers the immune system to attack hypocretin-producing neurons. “So now I am doing a lot of studies comparing the different vaccines and the wild-type virus to try to understand what could be common to produce this response. I think the vaccine will give us a final clue to isolate the immune T-cells involved in narcolepsy.”
Genetics of sleep. Mignot’s lab is working on a genome-wide association study, which shows that the genetic variants linked to narcolepsy are mostly immune-related, similar to Type 1 diabetes, celiac disease and other autoimmune diseases, further supporting the autoimmune hypothesis. Mignot is also getting a large human study off the ground. “I want to study the genetics of 40,000 people with sleep issues to see if there are genetic traits that cause people to sleep well or not sleep well, to need more sleep or less sleep. This hasn’t been done yet. I think this will help us crack open the mysteries of sleep.”
A new companion. “The dog colony was officially dismantled in 2000 after we found the canine narcolepsy gene. The dogs were adopted and we got Bear, a narcoleptic Schipperke. He passed away over a year ago. I loved that dog and miss him a lot. He was an unusually kind soul. Three months later, a breeder from Vermont called and said he had a narcoleptic Chihuahua. I flew to Vermont and adopted Watson and he’s been with us ever since. I never would have thought to adopt a Chihuahua, but now I can’t think of life without Watson. He is faithful and cuddly. I really think you can bond with any dog.”
The journey continues. “This story of narcolepsy, it’s a difficult story. Finding the gene was very difficult, and finding the autoimmune connection should have been trivial, but it has been an ordeal because there is absolutely no collateral damage. As [Stanford neurologist] Larry Steinman said to me, it’s like a ‘hit and run’—it looks like it was cleaned up and the players disappear. It’s hard, but by learning about this disease, we may discover other diseases where a similar autoimmune destruction happens in the brain but we have never realized it. I wouldn’t be surprised if some forms of depression and schizophrenia have an autoimmune basis in the brain. By experience, the more difficult it is, the most interesting the answer will be.”
- Identified the gene for hypocretin receptor 2, which, when mutated, causes an inherited form of narcolepsy in Dobermans and Labradors
- Identified how antidepressant and stimulant drugs work as treatments for narcolepsy
- Identified DQB1*0602 as the main human gene associated with narcolepsy
- By genome-wide association, found immune polymorphisms, such as one in the T-cell receptor alpha, that also predispose people to the disease, further suggesting the disease is autoimmune
- Found that human narcolepsy, unlike canine narcolepsy, is not caused by mutations in the hypocretin receptor 2 gene but is due to an immune-mediated destruction of hypocretin-producing neurons in the brain