Located in just one small region of the brain, hypocretin neurons have wirelike projections connecting to areas that promote alertness and regulate sleep/wake transitions. The receptors for hypocretin are spread around the brain, and engagement stimulates a network of arousal neurons. The network resembles an orchestra, with each section dominantly producing a different neurotransmitter: dopamine, norepinephrine, serotonin, or histamine. When someone is awake, each section is playing in concert with the others. Peptides such as hypocretin act more slowly and stick around longer, compared with classic small-molecule neurotransmitters—so we can view hypocretin as raising or lowering a curtain for the performance, while the neurotransmitters make up individual notes of the music.

  WIZARDS OF LIGHT

  Two separate labs identified hypocretin, also known as orexin, in the 1990s.13 Both groups were interested in the hypothalamus, even though they weren’t looking for something related to narcolepsy. The term hypocretin comes from the hypothalamus and from its similarity to another hormone called secretin, while orexin refers to how it can stimulate appetite when injected into a mouse’s brain.

  Before hypocretin was discovered, some information was available from studying dogs with inherited narcolepsy. Like humans, the dogs experience cataplexy when they feel excitement—when they are given food or toys. With the dogs, scientists were able to test various drugs and gather information about which regions of the brain were altered. But after the identification of hypocretin, researchers could more finely dissect the relevant brain circuitry. Electrical recordings in rats have shown that hypocretin neurons are most active when the animal is awake and exploring, less if awake and quiet, and inactive during sleep. They increase firing several seconds before animals wake up.

  However, hypocretin neurons are not only “wake up” neurons; they also respond to novelty or excitement. With the optogenetics technique, by engineering light-sensitive proteins from algae into an animal’s brain, scientists can use light from a fiber optics cable to activate or suppress preselected populations of neurons. Experiments from the lab of hypocretin’s co-discoverer, the Stanford neuroscientist Luis de Lecea, show that hypocretin neurons start firing when a mouse is presented with a new toy, another mouse, or a snack—or is grabbed by a human hand or smells a potential predator.14 De Lecea has quipped that in mice, hypocretin-producing neurons are active during the four Fs: feeding, fighting, fleeing, and mating.15 “Hypocretin neurons’ activity correlates with all of those activities,” he said. “The key aspect is how they’re responsible for integrating inputs of arousal and motivation.”

  It is now possible to create a mouse that experiences cataplexy by deleting the hypocretin gene; the animals are immobilized when presented with something enticing, such as chocolate. Since mice missing the hypocretin gene are thought to retain the neurons that produce the peptide, hypocretin is probably what’s important, even though the neurons produce other neuropeptides. And despite their central roles, if hypocretin or the neurons that make it are taken away, mice can still sleep, eat, and mate—even if cataplexy may get in the way.

  In addition to regulating sleep and wake, several research teams have shown how hypocretin neurons are connected to cravings for drugs of abuse.16 At UCLA, Jerry Siegel and his colleagues have shown that in the human hypothalamus, chronic exposure to opioids appears to increase the number of hypocretin-producing cells—the opposite of narcolepsy.17 Opiate exposure does not appear to be causing hypocretin neurons to multiply, but it is somehow activating hypocretin production in cells that had the potential for it.

  The Rutgers neuroscientist Gary Aston-Jones, whose lab observed similar effects in rats exposed to cocaine,18 said he viewed hypocretin as “a molecule for translating motivation into action.” Both Siegel and Aston-Jones have suggested that the findings may offer a way to pharmacologically blunt drug cravings. By temporarily inducing a partial narcolepsy-like state, drugs that diminish hypocretin signals, such as the sleep aid suvorexant, may be helpful in treating addiction.

  SOCIAL CONNECTIONS

  Nobody has tried optogenetic techniques on people, but neuroscientists can use the presence of hypocretin in cerebrospinal fluid to infer the kinds of roles it plays in humans. Once released from the brain cells where they are made, peptides such as hypocretin diffuse into the ventricles, the reservoirs for cerebrospinal fluid.

  Recall chapter 6, in which epilepsy patients came to the hospital for seizure diagnosis, and researchers took the opportunity to probe their brains for changes in adenosine. A similar set of experiments was performed to look at hypocretin levels’ responses to different activities.19 When Siegel’s group at UCLA did this, they saw that hypocretin levels increased markedly when epilepsy patients were visited by family members or interacted with physicians or hospital staff (figure 12.1).

  However, if we want to understand how emotions trigger cataplexy, the pattern of hypocretin production during social interactions is suggestive but not specific enough. In de Lecea’s studies on mice, hypocretin neurons became activated during many situations reflecting novelty or stress, but not all of these situations trigger cataplexy. We need to integrate what is known about neurotransmitters with additional information about the regions of the brain involved.

  FIGURE 12.1. In people undergoing intracranial epilepsy diagnosis, hypocretin production appeared to increase during interactions with family or medical staff.

  Source: Ashley M. Blouin et al., “Human Hypocretin and Melanin Concentrating Hormone Levels Are Linked to Emotion and Social Interaction,” Nature Communications 4 (2013). Reprinted with permission ©2013.

  IS CATAPLEXY UNBRIDLED REM SLEEP?

  For decades, the leading theory for explaining cataplexy was as an uncontrolled manifestation of the loss of muscle tone that accompanies REM sleep. In the late 1950s, Michel Jouvet discovered that cats with lesions in a dorsal region of the pons, part of the brainstem, appeared to act out their dreams. This area, called the sublaterodorsal nucleus, orchestrates immobility during REM sleep. Spurred by the sublaterodorsal nucleus, neurons in the medulla and spinal cord send inhibitory signals to motor neurons throughout the body. Cataplexy was thought to represent a short circuit, so that the process gets activated while someone is awake.

  Just as an aside, hypnogogic hallucinations are not as well studied, probably because they are less predictable, and sleep paralysis is known to occur occasionally in neurotypical people, outside the context of narcolepsy. Both are similarly thought to involve the aberrant activation of REM-like programs.

  Current research is coming close to substantiating the connection between cataplexy and REM sleep.20 However, the relationship becomes more complicated the closer one looks. Mehdi Tafti, a sleep researcher at the University of Lausanne, said that cataplexy and REM sleep may share the same neural pathways for loss of muscle tone, but beyond that, they don’t necessarily have the same mechanisms. Cataplexy is not a uniform state and has its own distinct EEG features, he said.21

  Some pharmacological evidence supports the cataplexy–REM sleep relationship. As Ann’s experience demonstrates, various antidepressants reduce the frequency and severity of cataplexy, and this is attributed to their ability to suppress REM sleep. The correlation seems to work in the other direction as well.

  Some medications, such as the blood pressure medication prazosin, can exacerbate cataplexy. In one severe case from Michigan, a woman in her forties began taking prazosin for hypertension about a year after she began to display signs of cataplexy.22 The medication worsened her cataplexy to the point that she had “virtually continuous brief episodes of weakness of the face, neck, trunk, or legs.” She could barely walk across a room without cataplexy occurring. When prazosin was stopped, her cataplexy ratcheted back to face and limb weakness at times of strong emotion. Prazosin works by blocking one type of receptors for the neurotransmitter norepinephrine, so researchers could infer that signals from norepinephrine help increase muscle tone and stave off cataplexy.

  Current theories propose that signals that make people go weak in the knees with laughter are activated all the time during emotional experiences or social interactions. While it may seem odd to imagine this happening constantly, neurologists have observed that H-reflexes, or muscles’ responses to electrical nerve stimulation, weaken during laughter—even in neurotypical people.23 Hypocretin neurons appear to be part of a loop connecting several regions of the brain, with an offshoot leading to the brainstem. Stabilization by hypocretin usually prevents incoming emotional signals from causing widespread muscle weakness, but if hypocretin is not present, the circuit is imbalanced.

  A team of neurologists at the University of Bologna has been gathering information on regions of the brain activated during cataplexy.24 To set the scene, they asked children and teenagers with narcolepsy, newly diagnosed and unmedicated, to choose short movie clips—Roadrunner and Coyote cartoons or YouTube cat videos, for example. For each young study participant, a video that worked well in evoking cataplexy was shown while they were wearing EEG electrodes and inside a magnetic resonance imaging scanner.

  During cataplexy events, imaging detected increases in blood flow in several brain regions, including the amygdala and part of the prefrontal cortex associated with pleasurable stimuli. The amygdala is known for its role in processing fear and fear-associated memories, but neuroscientists believe it is just as important as an “intensity detector” for positive emotions.25

  If a study participant simply laughed at a cartoon, the same regions were not activated. The Italian researchers concluded that during laughter-induced cataplexy, the activated regions of the brain are the same regions that physiologically process positive emotion and amusement or reward.

  Some additional evidence supports the idea that the amygdala’s function is impaired in people with narcolepsy. Usually humans reflexively blink more quickly in response to unpleasant or disgusting stimuli, a process controlled by the amygdala, but people with narcolepsy do not display this response.26 A model of how cataplexy occurs is that the amygdala sends signals that activate a REM-like paralysis program within the spinal cord, and those signals are successful because they are unopposed by hypocretin neurons.

  Recent research suggests an important role for the peptide oxytocin, known for its roles in both childbirth and in promoting social bonding.27 Oxytocin is supposed to drive us to seek out social situations and draw our attention to social cues. Preliminary findings indicate that in mice that model narcolepsy, oxytocin-sensitive neurons in the amygdala are active just before the onset of cataplexy. The involvement of oxytocin may begin to explain why in humans, laughter with friends or loved ones is so often a trigger for cataplexy. Oxytocin could be the spark of the emotional short circuit.

  TURNING A KEY IN A LOCK

  Neuroscientists have made considerable progress in understanding what hypocretin does and how its absence leads to cataplexy and disturbance of sleep and wake. However, all that research hasn’t provided people with narcolepsy with new ways to manage their symptoms—yet.

  A proposal for a nasal spray containing hypocretin generated excitement—but not a viable product. It took time for pharmaceutical companies to begin developing hypocretin receptor agonists, which mimic hypocretin’s effects on its receptors. Because the potential market for insomnia was much larger, drugs that antagonize the same receptors such as suvorexant (FDA approved in 2014) came first. But another hurdle was pharmacological: finding candidate compounds that cross the blood-brain barrier and activate the receptors. De Lecea said it has been a hundred times more difficult to find an agonist, something that activates the receptor mechanism like turning a key in a lock, than an antagonist, which obstructs the same lock.

  In 2017, Takeda Pharmaceutical Company began testing hypocretin receptor agonists in clinical trials. Other firms are developing similar drugs. These compounds are supposed to imitate the signals from hypocretin the brain is missing in narcolepsy type 1. Eventually, these could compete with the array of stimulants, antidepressants, and other medications that many people with narcolepsy now use. Whether they will be better than previous medications remains to be seen. Like conventional stimulants, hypocretin receptor agonists may also carry the risk of abuse or unpleasant side effects.28

  CHAPTER 13

  FRUSTRATING AND MOSTLY FRUITLESS

  This is something like the finding that everyone who had cataplexy had red hair. While it might not cause narcolepsy, it might indicate who could develop it.

  —American Narcolepsy Association newsletter, 1984

  This chapter takes its name from a 2006 commentary by the Harvard sleep neurologist Thomas Scammell. At the time, researchers studying narcolepsy had clues to how hypocretin was being eliminated from the brain but were not able to discern the specific mechanism.1 A few years after publication, the 2009–2010 H1N1 flu pandemic pushed the field forward. In the aftermath, researchers were able to obtain evidence that narcolepsy type 1 arises through the immune system making a disastrous mistake, coming from a combination of genetic susceptibility plus an environmental trigger. It is possible that a similar mechanism can explain other sleep disorders, such as narcolepsy type 2 or IH, but the full picture is not yet available. Plus, because the definitive laboratory test for narcolepsy type 1 has not been widely available until recently, many people currently diagnosed with narcolepsy do not know which category they actually fall into.

  OLD-SCHOOL DEFINITIONS

  Beginning in the nineteenth century, clinicians recognized that narcolepsy could run in families. One of the original descriptions of narcolepsy, from the German psychiatrist Karl Westphal, was of a mother and son who both experienced cataplexy and sleep attacks. Familial links helped strengthen neurologists’ case for narcolepsy being a biological disorder.

  In the 1950s, Yoss and Daly at the Mayo Clinic studied a Minnesota family in which seven out of sixteen siblings—twelve out of twenty-four people in the family they interviewed—had what they labeled as narcolepsy.2 Their descriptions sound familiar. Taken to the clinic by her insistent daughter, one forty-three-year-old woman said that she had been “fighting sleep all of my life.” That woman’s brother denied abnormal sleepiness but said he had three times driven into a ditch on the way home from work, and his family said that he sometimes fell asleep at the dinner table.

  It was just before the discovery of sleep-onset REM in daytime naps; all Yoss and Daly could really do was ask their patients questions. For most patients they studied, the family links were not as obvious. Yoss and Daly’s survey of four hundred people with narcolepsy found that about a third had a family member with the disorder.3 Other investigators found that this figure varied anywhere from 10 to 50 percent. The inheritance pattern for narcolepsy was more sporadic than for other diseases.

  When advances in genetics in the 1980s made it possible to search more systematically, Japanese researchers included narcolepsy in a program investigating several complex disorders. A driving force behind this effort was the psychiatrist Yutaka Honda at the University of Tokyo. In the 1960s, Honda had helped organize Japan’s first narcolepsy support group, and he was a pioneer in the use of antidepressants to control cataplexy.

  Honda, working with Takeo Juji, a specialist in transfusion medicine, observed an association between narcolepsy and one form of an HLA gene called HLA-DR2.4 The discovery was the first clue that narcolepsy had a connection with the immune system. HLA (human leukocyte antigen) genes were known to shape whether a transplant is perceived by the host’s immune system as foreign and subject to attack. When someone needs a bone marrow transplant, HLA type is the main factor determining whether a donor is a match. The HLA locus is the most polymorphic and diverse in the human genome, probably because a variety of HLA genes provides an evolutionary advantage in the arms race against pathogens.5 By that time, some HLA gene variants had been linked to other diseases such as rheumatoid arthritis and multiple sclerosis. The DR2 variant doesn’t sensitize people to autoimmune disorders in general—it is protective against type 1 diabetes.

  The link between narcolepsy and HLA-DR2 was very tight—all the Japanese narcolepsy patients were DR2 positive. But around a quarter of the general population is also DR2 positive. HLA genes may set the stage, but genetics do not determine whether someone is going to develop narcolepsy. They only indicate whether someone is susceptible.

  Later studies of narcolepsy genetics including African Americans and European Americans revealed that another HLA genetic variant called DQB1*0602 was a better marker than DR2 for population groups outside Japan.6 Again, most people who have this risk factor—still the strongest identified—do not develop narcolepsy. With identical twins, usually just one will have narcolepsy. Some other factor has to propel the disease to appear.

  Throughout the 1980s, some tension existed between Honda and sleep researchers at Stanford over the proper definition of narcolepsy.7 Among sleep specialists in the United States, the MSLT was widely adopted. In contrast, for Honda and his colleagues, displaying sleep-onset REM during an MSLT was not specific enough, and only patients with a history of cataplexy should be diagnosed with narcolepsy.8 To emphasize the distinction, the Japanese group used the term “essential hypersomnia” instead of “narcolepsy without cataplexy.” Some specialists outside Japan, such as the United Kingdom’s David Parkes, reportedly preferred Honda’s strict, old-school definitions. Parkes was known to tell trainees, “A good sleep centre has far more need of a psychiatrist than an EEG machine.”9