The Neurobiology of Narcolepsy-Cataplexy

THE NEUROBIOLOGY OF NARCOLEPSY-CATAPLEXY

MICHAEL S. ALDRICH Department of Neurology, Sleep Disorders Center, University of Michigan Medical Center, Ann Arbor, MI, U.S.A.

(Received 17 July 1992)

CONTENTS 1. Introduction 533 2. Clinical aspects 533 2.1. Sleepiness and sleep attacks 533 2.2. Cataplexy and related symptoms 534 2.3. Clinical variants 534 2.3.1. Narcolepsy without cataplexy 534 2.3.2. Idiopathic hypersomnia 534 2.3.3. Symptomatic narcolepsy 534 2.4. Treatment 534 3. Pathophysiology 535 4. Neurobiological studies 535 4.1. The canine model of narcolepsy 535 4.2. Pharmacology of human cataplexy 537 4.3. Postmortem studies 537 5. Genetic and family studies 537 6. Summary and conclusions 539 References 539

1. INTRODUCTION 2. CLINICAL ASPECTS Narcolepsy is a specific neurological disorder Narcolepsy has a prevalence that varies worldwide characterized by excessive sleepiness that cannot be from as little as 0.0002% in Israel to 0.16% in Japan; fully relieved with any amount of sleep and by in North America and Europe the prevalence is about abnormalities of rapid eye movement (REM) 0.03-0.06% (Dement et al., 1972; Honda, 1979; Lavie sleep. About two-thirds of patients also have brief and Peled, 1987). The onset of narcoleptic symptoms, episodes of muscle weakness usually brought on by usually in the second or third decade of life, may emotion, referred to as cataplexy. The disorder gener- occur over a few days or weeks or it may be so ally begins in adolescence and continues throughout gradual that the loss of full alertness is unrecognized life. by family or friends. The onset may be attributed to Over the past three decades, there has been remark- a variety of antecedent events including psychological able progress in understanding the fundamental or physical stress, minor or major head trauma, mechanisms responsible for narcoleptic symptoms. infections, drug abuse, or pregnancy, but whether Narcolepsy is associated with the premature onset of these precipitants contribute to development of the REM sleep and studies of the neuroanatomical and syndrome or are just coincidental i~ uncertain. neurochemical substrate of REM sleep, combined with pharmacological, physiological, genetic and 2.1. SLEEPINESS AND SLEEP ATTACKS molecular studies of narcolepsy have led to striking advances. This article provides an overview of the Excessive sleepiness and cataplexy are not the only clinical syndrome followed by a presentation symptoms of narcolepsy but they are usually the most of current views of its pathophysiology and neuro- troublesome. Sleepiness is more characteristic than an biology. increase in actual sleep time; the amount of sleep obtained over 24 hr is increased only slightly because the daytime sleep episodes are offset by frequent nocturnal awakenings. Narcoleptic sleepiness is similar to the sleepiness Address correspondence to M. S. Aldrich, Sleep Dis- that occurs with sleep deprivation in normal people: orders Center, Department of Neurology Taubman Center it is most apparent in boring, sedentary situations and 1920/0316, University of Michigan Medical Center, 1500 is partially relieved by stimulation or movement. East Medical Center Drive, Ann Arbor, MI 48109-0316, Narcoleptic sleepiness fluctuates from minute to U.S.A. minute and varies between patients: it may be mild

533 534 M.S. ALDR1CH

and easily overcome by stimulation or it may be 2.3. CLINICAL VARIANTS severe and overwhelming as soon as the patient sits down. The inability to obtain full alertness after any 2.3.1. Narcolepsy without cataple.xy amount of night-time or daytime sleep is the feature In the one-third of patients with narcolepsy who do that distinguishes narcoleptic sleepiness from sleepi- not have cataplexy, age of onset, premature occur- ness associated with sleep deprivation. rence of REM sleep and severity of sleepiness are Although spontaneous daytime sleep episodes may similar to patients with narcolepsy-cataplexy, but appear paroxysmal, similar to epileptic seizures, they sleep paralysis, hypnagogic hallucinations and severe are preceded by electrophysiological evidence of in- nocturnal sleep disruption are less common (Rosen- creasing drowsiness; thus sleep "attacks" are simply thal et al., 1990). As cataplexy may develop decades episodes of falling asleep. Because narcoleptics lack after the onset of sleepiness, the eventual occurrence personal experience of full alertness and become of cataplexy can never be excluded in a given patient. habituated to chronic drowsiness, they may not per- Genetic and family studies (see below) suggest that ceive sleepiness in the usual way; this lack of aware- narcolepsy without cataplexy is probably a less severe ness probably accounts for descriptions of "sleep manifestation of the same disorder. attacks". Chronic sleepiness leads to memory disturbances, visual problems and episodes of amnesia lasting for 2.3.2. Idiopathic hypersomnia seconds to 30 minutes or more, associated with semi-purposeful activity, referred to as automatic Patients with this syndrome have excessive sleepi- behavior. Memory problems are probably caused by ness without cataplexy or abnormalities of REM drowsiness and brief episodes of sleep with associated sleep. Although the syndrome is unrelated to nar- impaired attention and concentration rather than by colepsy in most cases, its occurrence in some family a specific disturbance of brain regions involved in members of narcoleptics and the fact that some memory. Failure of fusion induced by drowsiness in patients with this syndrome eventually develop cata- exophoric patients is the likely cause of blurred and plexy and REM sleep abnormalities suggest that double vision. Episodes of automatic behavior occasionally it may be a variant of narcolepsy. usually occur during monotonous or repetitive activities and may be associated with brief pauses in speech 2.3.3. Symptomatic narcolepsy or movement, irrelevant words or remarks and non- sensical activities. Acquired brain lesions as a cause of narcolepsy have been of interest since cases of narcolepsy were described following the epidemic of encephalitis lethargica more than 60 years ago. Although post-en- 2.2. CATAPLEXY AND RELATED SYMPTOMS cephalitic narcolepsy is now extremely rare, either Brief episodes of weakness brought on by excite- from death of affected patients or a change in disease ment or emotion, called cataplexy, occur in about patterns, narcolepsy can develop in association with two-thirds of patients and are almost unique to brain lesions adjacent to the third ventricle, an area narcolepsy. Precipitants of cataplexy include laugh- concerned with sleep-wake regulation (see Aldrich ter, anger, embarrassment, excitement and athletic and Naylor, 1989 for review). Although many such activities. With severe attacks, there is complete patients have the HLA marker associated with idio- atonic areflexic paralysis of striated muscles sparing pathic narcolepsy (see below), this is not enough to only respiratory muscles; the more common milder suggest that narcolepsy can occasionally arise from episodes can cause facial weakness, slurred speech, brain lesions alone in the absence of the usual genetic drooping eyelids, weakness of grip, head nodding, or substrate. slight buckling at the knees. Episodes last from a few seconds to several minutes; consciousness is not 2.4. TREATMENT impaired initially but longer episodes may be associated with auditory, visual, or tactile hallucinations Although narcoleptic symptoms respond to medi- and the onset of REM sleep. cation, they are usually not fully controlled. Thus, Two other symptoms, sleep paralysis and hypna- maintaining alertness and avoiding cataplexy during gogic hallucinations, are common in narcolepsy. Sleep specified times of the day, such as during work, paralysis refers to episodes occurring at sleep onset or school, or driving, are the main goals of therapy. For upon awakening of inability to move, sometimes sleepiness, amphetamines and related compounds associated with a sensation of struggling to move or such as methylphenidate and pemoline are the most awaken. They last up to several minutes and end effective medications. Many clinicians also rec- spontaneously or after mild stimulation. Hypnagogic ommend one or more short naps each day; such naps hallucinations also occur during the transitions be- often provide a period of relative alertness lasting tween sleep and wakefulness. They resemble dreams 1-2 hr. Tricyclic anti-depressants are the most effec- in that the apparent reality of visual, tactile, or tive medication for the treatment of cataplexy and auditory features is usually unquestioned, but they permit almost complete control of the symptoms in differ because some awareness of the environment is about 80% of patients. However, patients with Severe preserved. Unlike cataplexy, sleep paralysis and hyp- cataplexy may develop tolerance to their effects and nagogic hallucinations can occur in otherwise normal a temporary withdrawal of medication may be re- people after sleep deprivation or a change in sleep quired to restore efficacy. Abrupt discontinuation of schedule. anti-cataplectic medication can lead to nearly con- NEUROBIOLOGY OF NARCOLEPSY~ATAPLEXY 535 tinuous incapacitating cataplexy lasting for several ements for the generation of REM sleep are located hours or days. in the ports. The rostral ponfine tegmeutum and areas adjacent to and within the locus coeruleus contain neurons, referred to as "REM-on" cells, that are 3. PATHOPHYSIOLOGY selectively activated during REM sleep and that are essential for its expression (Sakai, 1988; Siegel, 1989). Although the essential clinical features of nar- Muscarinic agonists injected into the pontine tegmen- colepsy have been known for more than 100 years, turn facilitate the appearance of a REM sleep-like the discovery in the 1960s that patients with nar- state in a dose-dependent and site-dependent manner colepsy have a remarkable predisposition to enter (Baghdoyan et al., 1984; Vanni-Mercier et al., 1989); directly into rapid eye movement (REM) sleep at the these and other studies have demonstrated that many onset of sleep provided the first clue to its pathophys- of the REM-on neurons are almost certainly cholin- iology and neurobiology (Rechtschaffen et al., 1963; ergic and that cholinergic neurons play a central role Takahashi and Jimbo, 1963; Vogel, 1960). As REM in the generation of REM sleep. The muscarinic sleep usually begins about 80-100 minutes after the receptor M2 subtype appears to be largely respon- onset of REM sleep, the tendency to enter REM sleep sible for the REM sleep-inducing effects of muscar- within minutes of falling asleep provides a useful inic agonists, although MI and M3 receptor subtypes diagnostic tool for narcolepsy as well as an indication may play minor roles (Imeri et al., 1991; Velazquez- that impaired control and regulation of REM sleep Moctezuma et al., 1989, 1991). are key features of its pathogenesis. Located within the same regions as pontine REM- REM sleep is characterized by cortical activation on cells are REM-off cells, cells that are selectively with low amplitude desynchronized EEG activity, silent during REM sleep. Many of these REM-off skeletal muscle atonia and rapid eye movements. neurons, such as serotonergic neurons of the raphe Compared with NREM sleep, brain metabolic ac- nuclei and noradrenergic neurons of the locus tivity is increased and there is greater variability of coeruleus, are monoaminergic and appear to play a heart rate, blood pressure and respiration; it is the permissive role through modulation of activity of state in which most dreams occur. Evidence collected REM-on neurons (Jones, 1991; Siegel, 1989). over the past three decades has provided strong Two areas of the medulla are critically involved in indications that many of the symptoms of narcolepsy the expression of muscle atonia during REM sleep, are due to intrusion of elements of REM sleep into the nucleus magnoceilularis and the nucleus parame- the waking state. Thus, the weakness associated with dianus (Lai and Siegel, 1988). Certain cells in these cataplexy and the immobility that occurs with sleep regions have high rates of discharge in REM sleep paralysis are associated with muscle atonia that is with lower discharge rates during NREM sleep and physiologically almost indistinguishable from the wakefulness (Siegel, 1989). Atonia can be induced by atonia of REM sleep. Similarly, hypnagogic halluci- stimulation of these regions: glutamate induces ato- nations probably represent dream imagery occurring nia in the nucleus magnocellularis via non-N-methyl- at the interface between wakefulness and sleep. D-aspartate receptors whereas acetylcholine induces However, disordered REM sleep is not the sole atonia in the nucleus paramedianus via muscarinic pathophysiological feature of narcolepsy. Many day- receptors (Lai and Siegel, 1988; Lai and Siegel, 1991). time sleep episodes consist only of NREM sleep These regions appear to mediate REM sleep atonia suggesting that there is a broader problem of im- and the atonia induced by infusion of muscarinic paired sleep-wake regulation. Frequent nocturnal agonists into the pontine tegmentum. awakenings indicate an impaired ability to sustain sleep that complements the daytime inability to re- 4.1. ThE CANINE MODEL OF NARCOLEPSY main awake. Periods of sleep that have physiological features of REM and NREM sleep, along with The existence of an animal model of narcolepsy, evidence that an element of wakefulness is often still discovered in the early 1970s, has permitted extensive present during sleep onset REM periods (Hishikawa, investigations of this condition. Canine narcolepsy 1975), suggest that narcolepsy is associated with is characterized by juvenile onset of sleep-onset indistinct boundaries between wakefulness, REM REM periods, cataplexy, excessive sleepiness and sleep and NREM sleep (Broughton et al., 1986). sleep-wake abnormalities similar to those seen in Episodes of REM sleep without the usual associated human narcolepsy (Foutz et al., 1981; Kaitin et al., muscle atonia and the occurrence in some patients of 1986; Kushida et al., 1985; Lucas et al., 1979; Mitler the REM sleep behavior disorder (Schenk et al., and Dement, 1977). Cataplexy in these animals is 1989) support the notion that the sleep-wake state readily precipitated by food, a response that has been boundaries are impaired. standardized as the food elicited cataplexy test (FECT) (Delashaw et al., 1979). Studies using the FECT have demonstrated that 4. NEUROBIOLOGICAL STUDIES four classes of compounds exacerbate cataplexy (Table 1). Cataplexy is associated with increased The relation between impaired regulation of REM acetylcholine release in the medial pontine reticular sleep and clinical symptomatology suggests that ab- formation and infusions of muscarinic M2 receptor normal activity of neural systems involved in REM agonists, but not M I receptor agonists, into the sleep may contribute to the pathogenesis of nar- medial pontine reticular exacerbates cataplexy (Reid colepsy. As shown by studies using transections, et al., 1992). lesions and stimulation techniques, the essential el- Several classes of drugs inhibit cataplexy; the most 536 M, S. ALDRICH

TABLE 1, RECEPTOR PHARMACOLOGY OF CANINE CATAPLEXY Site of action Exacerbate cataplexy Inhibit cataplexy Noradrenergic alpha- 1 receptor Antagonists (~ 1b > c~ la) ~ Agonists2 Noradrenergic alpha-2 receptor Agonists (non-imidazole type) 3 Antagonists 3 Dopaminergic D-I receptor --- Antagonists 4 Dopaminergic D-2 receptor Agonists4 Antagonists4 Muscarinic receptor Agonists5 Antagonists ~ References: 1. Mignot et al., 1988; Mignot et aL, 1989; 2. Nishino et aL, 1992; 3, Nishino et al., 1990; 4, Nishino et al., 1991; 5. Delashaw et al., 1979; Foutz et aL, 1981.

potent effects are produced by drugs that increase glandin E2 and its methyl ester (Nishino et al., 1989). synaptic availability of norepinephrine by blocking Siegel et al. (1991), recording single unit activity reuptake or by enhancing release (Table 1). Although from the medial medulla during canine cataplectic fluoxetine and other serotonergic reuptake inhibitors attacks, identified cells that were activated during also inhibit cataplexy, their effects are probably medi- cataplexy (Fig. 1). The cataplexy-on cells, which were ated at least in part by their aminated metabolites, not cholinergic and which were in the same region as which inhibit norepinephrine reuptake (Nishino the glutamate sensitive cells mediating atonia de- et aL, 1992). Furthermore, the efficacy of tricyclic scribed above, were active during REM sleep; how- anti-depressants in the treatment of canine cataplexy ever, most REM-on cells were not activated during is directly proportional to their inhibition of norepi- cataplexy. nephrine reuptake (Foutz et al., 1981). Compounds that increase synaptic availability of dopamine These studies support the view that cataplexy is have little effect on cataplexy (Nishino et al., 1990, physiologically and pharmacologically closely related 1991; Renaud et al., 1991), but dopamine DI and to REM sleep. The premature onset of REM sleep in D2 receptor antagonists are inhibitory, perhaps by narcolepsy and the intrusion of muscle atonia into acting at presynaptic receptors on noradrenergic wakefulness that occurs with cataplexy could be a neurons. Cataplexy can also be inhibited by prosta- result of inadequate monoaminergic inhibition of

AcUve waking

......

Un#

Wlklng/caWplexy

EOO - ' ~ "

EMG

U~Jt i.I II lib l,ll i.h.l, tt 1611"11 liJJiJl w | ,,mu.o,ql b~--

NREM sleep so ,v

~ x ~ ~"~q' ~l~f~'~I'v~'~a'~,""v~c'('*~i ~'l,~'%v, ",'~, ,,','~ - .-, -- - ,,,,, -.~-.,.~,~....~r~,. ,~,.- ~ ~...... ~"~-~-~'q*'l EOG eMO t

UrJR lli~ II ~ I ; ~ t i ~1 T t lil]l 16 i I 1 IIJ III I ITI I II;ll~llll [| [~U III rail

lo • I ! FIO. 1. Single-unit activity from the medial medulla of a narcoleptic dog recorded during wakefulness, non-REM sleep, REM sleep and cataplexy elicited by play or food. The unit is selectively activated during REM sleep and during cataplexy, Only a fraction of REM-on c~lls are activated during cataplexy. Reproduced with permission from Siegel et al. (1991). NEUROBIOLOGY OF NARCOLEPSY--CATAPLEXY 537 cholinergic REM sleep mechanisms or of excessive increased concentrations of dopamine and 3,4-dihy- excitation of cholinergic or glutamatergic brainstem droxyphenylacetic acid (DOPAC) in the amygdala systems. (Miller et al., 1990); and (5) increased concentrations of norepinephrine in portions of the brainstem (Miller et al., 1990). Preliminary studies suggest that 4.2. PHARMACOLOGY OF HUMAN CATAPLEXY there are no major changes in adenosine or benzo- The clinical pharmacology of human cataplexy is diazepine receptors (Bowersox et al., 1986; Hawkins consistent with the canine model. Amphetamines and et al., 1991). While these studies are not conclusive, related stimulants, commonly used to treat narcolep- they are similar to the findings in human narcoleptic tic sleepiness, increase synaptic availability of norepi- brain and suggest an alteration in cholinergic and nephrine. Prazosin, a noradrenergic alpha-1 receptor monoaminergic metabolism. antagonist, causes an exacerbation of human cataplexy (Aldrich and Rogers, 1989; Guilleminault et al., 1988) whereas tricyclic anti-depressants that block 5. GENETIC AND FAMILY STUDIES norepinephrine reuptake inhibit cataplexy. Although cataplexy is also inhibited by fluoxetine (Langdon Although the occurrence of narcoleptic symptoms et al., 1986), a relatively specific blocker of 5-HT among family members of persons with narcolepsy reuptake, the inhibition may be due to the effects of has been recognized fol~ many years (Baraitser and its primary metabolite, norfluoxetine, which is active Parkes, 1978; Daly and Yoss, 1959; Kessler et al., at noradrenergic synapses. Drugs that increase synap- 1974), the prevalence of confirmed narcolepsy among tic availability of dopamine have relatively little effect family members, as opposed to subjective complaints on cataplexy and other narcoleptic symptoms despite of sleepiness, is only about 1% (Guilleminault et al., the finding that cerebrospinai fluid levels of dopamine 1989). Only 3% of narcoleptics have a first degree and its metabolite homovanillic acid are reduced in relative with narcolepsy and only one-third of those, sufferers of human narcolepsy (Montplaisir and God- or 1% of all narcoleptics, have more than one bout, 1986; Parkes et aL, 1974). These findings are affected first degree relative (Guilleminault et al., consistent with the view that defective noradrenergic 1989). The risk of developing narcolepsy is about 18 modulation of REM sleep mechanisms contributes to times greater in first degree relatives of narcoleptics human and canine narcolepsy. As noradrenergic pro- than in the normal population (Guilleminault et al., jections systems are also intimately involved in main- 1989). This mode of inheritance is consistent with an tenance of wakefulness, a defect in noradrenergic autosomal dominant pattern with low penetration. metabolism may also contribute to a broader prob- The first clues to the genetic basis of narcolepsy lem of sleep/wake regulation. were provided by studies of Class II human leukocyte antigens (HLA). Expression of these antigens is gov- 4.3. POSTMORTEM STUDIES erned by the major histocompatibility complex (MHC) on chromosome 6 (Fig. 2). Studies performed Postmortem studies of human narcoleptic brain in the 1980s, initially in Japan and subsequently in tissue have not shown consistent structural lesions at Europe, North America and Africa, demonstrated the light microscopic level. Neurochemical studies, that the HLA-DR2 antigen, found in 20-35% of the however, suggest that monoaminergic function is normal population, has an incidence that exceeds abnormal. Binding of ligands specific for dopamine 95% in European and Caucasian patients with nar- D1 and D2 receptors is increased in areas of the colepsy and is close to 100% in Japanese patients striatum (Aldrich et al., 1992; Kish et al., 1992). (Billiard and Seignalet, 1985; Juji et al., 1984, 1988; Preliminary studies suggest that there may also be an Kramer et al., 1987; Langdon et al., 1984). There is alteration in the balance of adrenergic alpha-2 and also an extremely high incidence of HLA-DQwl, alpha-I receptors in selected brain regions (Aldrich equaling or exceeding that of HLA-DR2 in all narco- et al., 1991). Studies of the norepinephrine metabolite leptic groups assessed to date. In African-Americans 3-methoxy-4-hydroxyphenylglycol (MHPG) and of with narcolepsy HLA-DR2 is found in only about the serotonin metabolite 5-hydroxyindoleacetic acid 65% but HLA-DQwl is present in more than 90% (5-HIAA) have shown increased concentrations and (Matsuki et al., 1992; Neely et al., 1987). These a trend towards increased ratios of monoamine studies strongly suggest that a gene in close proximity metabolites to neurotransmitters in a number of to the HLA-DR and DQ regions has a major role in brain regions (Kish et al., 1992). While medication the pathogenesis of narcolepsy. The HLA-DP region taken before death may have contributed to these is not involved in the narcoleptic haplotype (Olerup, findings, there is the suggestion of augmented mono- 1990). amine turnover and increased monoaminergic nerve In patients with narcolepsy who do not have terminal activity in at least some brain areas (Kish cataplexy, the DR2 marker occurs almost as fre- et al., 1992). quently as it does in narcolepsy-cataplexy syndrome Studies of canine narcoleptic brain have also (see Aldrich, 1992), suggesting phenotypic variability shown neurochemical abnormalities such as: (1) in- for the HLA associated susceptibility gene or genes. creased numbers of muscarinic receptors, probably of The occurrence of idiopathic hypersomnia in some the M2 subtype, in the brainstem nucleus reticularis DR2 positive relatives of narcoleptics suggests that gigantocellularis (Kilduff et al., 1986); (2) increased some cases of this entity may represent a third dopamine D2 receptors in the amygdala (Bowersox phenotype (Harris et al., 1987). et al., 1987); (3) increased alpha-2 adrenoceptors in Recent analyses have permitted subtyping of the the locus coeruleus (Fruhstorfer et al., 1989); (4) HLA-DR2 antigen into DRwl5 and DRwl6 sub- 538 M.S. ALDRICH

Chromosome 6 Long arm Short arm <_

jJ Major Histocompatibility Complex

I ClaSSDregionlI t Class III [ B ClaSScI A t

DP subregion DQ subregion DR subregion I illiil I' il Ilil IIIII i I ...... I 131 ct2 !31 t~l et2 132 txl 131 131 132 [33 ctl FtG. 2. Major histocompatibility complex (MHC) genes, located on the short ann of chromosome 6, control the expression of antigen-binding cell surface glycoproteins. Class I antigens, expressed on nucleated cells and platelets, are designated HLA-A, -B and -C. Class III genes code for a variety of proteins including those involved in complement activation. Class II antigens, expressed on lymphocytes, monocytes, macrophages and dendritic cells and designated HLA-DP, -DQ and -DR, are made up of four subunits of alpha and beta chains. types. In addition, the DQw6 subtype of DQwl has and DQ(beta) are identical in narcoleptics and con- been identified. The DRwl5 and DQw6 subtypes trois with DRwl5/DQw6/Dw2 haplotypes (Hollo- are strongly associated with narcolepsy as is the man et al., 1987; Honda et al., 1986; Kuwata et al., Dw2 antigen, which is detected with mixed 1991; Lock et al., 1988; Olerup et al., 1990; Uryu lymphocyte cultures rather than by serologic testing et al., 1989; Wilner et al., 1988). These findings and (Honda et aL, 1986; Juji et al., 1988). The evidence that the HLA marker came from different DRwI5/DQw6/Dw2 haplotype is carried by more ancestral sources in some familial cases (Guillem- than 90% of narcoleptics and is specifically associ- inault et al., 1989) suggests that narcoleptic suscepti- ated with susceptibility to narcolepsy, not to the bility is not caused by a mutation of the genes symptom of somnolence, as its incidence is not responsible for the DRwI5/DQw6/Dw2 haplotype. increased in sleep apnea or in other disorders causing The occurrence of the DRwI5/DQw6/Dw2 haplo- excessive sleepiness (Rubin et al., 1988). type in asymptomatic relatives of narcoleptics and in The DRwI5/DQw6/Dw2 haplotype encompasses control subjects and the presence of DQBI-0602 in several polymorphic genes, named DRBI-1501, almost 40% of asymptomatic African-Americans DQA1-0102 and DQB1-0602, that encode the provides strong evidence that the HLA-related sus- DRwI5 and DQw6 antigens. Two of these, DRBI- ceptibility gene is not sufficient to induce narcolepsy. 1501 and DQB1-0602, occur only with the Concordance for narcolepsy is rare in monozygotic DRwI5/DQw6/Dw2 haplotype in Caucasian and twins (Guilleminault et al., 1989; Montplaisir and Japanese populations and are therefore potential Poirier, 1987), indicating that unknown environmen- candidates for the HLA-related narcolepsy suscepti- tal factors play a major role in the development of bility gene. However, in African and African- narcolepsy. American subjects, some haplotypes show a dis- Despite the strong association of HLA markers sociation between DRBI-1502 and DQBI-0602. with narcolepsy, increasing evidence suggests that the Matsuki et al. found that the DQB1-0602 gene was syndrome is genetically heterogeneous. There are well present in 100% of African-American narcoleptics documented cases of DR2/DQwl negative nar- but in only 38% of control African-Americans (Mat- colepsy and in some narcoleptic families with two or suki et al., 1992). The DQBI-0602 was present even more affected individuals, not all narcoleptics carry in subjects who did not carry DRwl 5. These findings the DRw 15/DQw6/Dw2 haplotype. In other families, suggest the possibility that the HLA-related nar- all affected members are negative for HLA-DR2 colepsy susceptibility gene is identical to the DQBI- (Dahlitz et al., 1992; Guilleminault et al., 1989; Singh 0602 gene. However, other genes closely linked to et al., 1990) and studies of DNA band patterns from DQBI-0602 could also be responsible and remain to the HLA region generated by a number of restriction be investigated. enzymes in two such families suggested that there had Attempts to identify a unique DNA sequence been no recombination in the HLA region and that conferring susceptibility to human narcolepsy have so the occurrence of narcolepsy was unrelated to HLA far been unsuccessful. Restriction fragment length haplotype (Ditta et aL, 1992). Thus, a gene unrelated polymorphisms (RFLPs) for DR(beta), DQ(alpha) to the HLA region appeared to be responsible, NEUROBIOLOGY OF NARCOLEPSY-CATAPLEXY 539 suggesting that narcolepsy is genetically hetero- However, the existence of some families in which geneous. narcolepsy does not cosegregate with HLA haplo- The inheritance of canine narcolepsy in Doberman types suggests genetic heterogeneity with the exist- pinschers and Labrador retrievers is consistent with ence of a second narcolepsy susceptibility gene the presence of a single fully penetrant autosomal unrelated to the HLA region. This second gene is recessive gene, designated canarc-l. This gene is not probably not the fully penetrant autosomal recessive linked to the canine equivalent of the MHC region gene, designated canarc-l, that is responsible for the (Dean et al., 1989; Matoyama et al., 1989); instead it narcolepsy that occurs in some large canine breeds. is tightly linked to a polymorphic DNA band that is Discordance for narcolepsy in monozygotic twins strongly homologous with the human switch region indicates that environmental factors are required for of the mu immunoglobulin heavy-chain gene (Mignot full expression of the syndrome. et al., 1991b), which is involved in determining the Despite the striking advances reviewed in this immunoglobulin class of certain activated B-lympho- manuscript, many aspects of narcolepsy remain un- cytes. clear including the precise identity and function of the Although full expression of canine narcolepsy and HLA-associated susceptibility gene and the ad- spontaneous cataplexy requires homozygosity for ditional gene that is implicated in families in which canarc-l, cataplexy may develop in heterozygotes narcolepsy does not cosegregate with HLA haplo- following administration of a combination of types. The additional environmental, developmental physostigmine, a muscarinic agonist and prazosin, an and genetic factors that lead to disease expression and adrenergic alpha-I antagonist (Mignot et al., 1991a). contribute to phenotypic variability are unknown. This biological activity of a single copy of canarc-1 The function of the recessive gene responsible for suggests that this gene, if present in humans, could canine narcolepsy is unknown as is the role, if any, contribute to narcolepsy susceptibility. However, in of the homologous human gene. Although neuro- preliminary studies of five families in which narcolep- chemical and neuropathological studies implicate a tic susceptibility was not associated with HLA haplo- derangement of noradrenergic metabolism, it is un- type, narcolepsy was not associated with unique certain whether these changes are primary or sec- RFLP patterns generated by the mu-switch H-24 ondary. Research that sheds light on these genomic probe gene, suggesting that the genes in the uncertainties may help to identify more effective immunoglobulin heavy chain locus do not contribute treatments for this chronic disorder. to susceptibility to human narcolepsy (George et al., 1992). Although the associations of human narcolepsy with the HLA-D region and canine narcolepsy with REFERENCES the immunoglobulin mu-switch region suggests that ALDRICH, M. S. (1992) Narcolepsy. Neurology" 42(SnppL 6), 34-43. narcolepsy may have an immunologic pathogenesis, ALDRICH, M. S. and NAYLOR, M. W. (1989) Narcolepsy associated there is little evidence of immunologic abnormalities with lesions of the diencephalon. Neurology 39, 1505--1508. in human narcolepsy (Frederikson et al., 1990; ALDRICH, M. S. and ROGERS, A. E. (1989) Exacerbation of human Honda et al., 1986; Rubin et al., 1988; Strohmaier cataplexy by prazosin. Sleep 12, 254-256. et al., 1988). The onset of canine cataplexy is associ- ALDRICH,M. S., HOLLINGSWORTH,Z. and PENNEY,J. B. (1991 ) Adrcn- ated with a mild CSF pleocytosis (Gaiser et al., 1989) crgic receptor autoradiography of human narcoleptic brain. Sleep Res. 20A, 280. and it remains possible that a transient immune ALDRICH, M. S., HOLLXNGSWORTH,Z. and PENNEr J. B. (1992) Dopa- mediated event that is not readily detectable is in- mine receptor autoradiography of human narcoleptic brain. Neu- volved in the initiation of human narcolepsy. rology 42, 410-415. BAGHDOYAN,H. A., RODRIGO-ANGULO,M. L,, MCCARLEY, R. W. and HOnSON,J. A. (1984) Site-specific enhancement and suppression of desynchronized sleep signs following cholinergic stimulation of 6. SUMMARY AND CONCLUSIONS three brainstem regions. Brain Res. 306, 39-52. BARAITSER, M. and PARKES, J. D. (1978) Genetic study of the nar- Human narcolepsy is a chronic neurological dis- coleptic syndrome. J. Med. Genet. 15, 254-259. order characterized by excessive daytime sleepiness, BILLIARD, M. and SEIGNALET,J. (1985) Extraordinary association episodic muscle weakness (cataplexy), premature on- between HLA-DR2 and narcolepsy. Lancet 2, 226-227. set of rapid eye movement (REM) sleep and impaired BOWERSOX, S. S., KILDUFF, T. S, KAIT1N, K. I., DEMENT, W. C. and CIARENELLO, R. D. (1986) Brain benzodiazepine receptor charac- sleep--wake regulation. A canine model for nar- teristics in canine narcolepsy. Sleep 9, I 11 I 15. colepsy exists that shows many physiological, phar- BOWERSOX,S. S., KILDUFF,T. S., FAULL, K. F., ZELLER-DEAMIClS,L., macological and behavioral similarities to the human DEMENT, W. C. and CIARENELLO, R. D. (1987) Brain dopamine disease. Neurobioiogical studies of human and canine receptor levels elevated in canine narcolepsy. Brain Res. 402, narcolepsy suggest that abnormal regulation of REM 444--448. sleep, probably related to disturbed monoaminergic BROUGHTON,R., VALLEY,V., AGUIRRE,M., ROBERT,J., SUWALSKI,W. and DUNHAM, W. (1986) Excessive daytime sleepiness and the and cholinergic brainstem activity, accounts for much pathophysiology of narcolepsy-cataplexy: a laboratory perspec- of narcoleptic symptomatology. tive. Sleep 9, 205-215. Genetic and family studies have demonstrated a DAHLITZ, M., VAUGHAN,R., KAYED, K. and PARKESJ. D. (1992) The strong association of human narcolepsy with the narcolepsy susceptibility gene. Lancet 339, 1297. HLA haplotype DRwl5/DQw6/Dw2 and with the DALY, D. D. and Yoss, R. E. (1959) A family with narcolepsy. Proc. Mayo Clin. 34, 313-319. DQBI-0602 gene in the HLA region. The two main DEAN, R. R., KtLDUFF, T. S., DEMENT,W. C. and GRUMET, F. C. (1989) phenotypes of narcolepsy--nareolepsy-cataplexy Narcolepsy without unique MHC Class II antigen association: syndrome and narcolepsy without cataplexy--are studies in the canine model. Hum. Immun. 25, 27-35. both strongly associated with the HLA markers. DELASHAW,J. B., FOUTZ, A. S., GUILLEMINAULT,C. and DEMENT, W. 540 M.S. ALDRICH

(1979) Cholinergic mechanisms and cataplexy in dogs. Exp. KUSHIDA, C. A, BAKER, T. L. and DEMENT, W. C. (] 985) Electroen- NeuroL 66, 745-757. cephalographic correlates of cataplectic attacks in narcoleptic DEMENT, W., ZARCONE, V., VARNERV. et al. (1972) The prevalence of canines. Electroencephalogr. clin. Neurophysiol. 61, 61 70. narcolepsy. In: Sleep Research, Vol. 1, p. 148. Eds. M. H. CHASE, KUWArA, S., TOKUNAGA, K., JIN, F.. JuJl, X., SASAKI.T. and HONDA, W. C. STERNand P. L. WALTER. Brain Information Service/Brain Y. (1991) Narcolepsy (correspondence). New Eng/. J Med. 324, Research Institute, UCLA: Los Angeles. 271-272. DITTA, S. D., GEORGE, C. F. P. and StNGH, S. M. (1992) HLA-D-re- LAI, Y. Y. and SIEGEL, J. M. (1988) Medullary regions mediating gion genomic DNA restriction fragments in DRw 15(DR2) familial atonia. J. Neurosci. 8, 4790--4796. narcolepsy. Sleep 15, 48-57. LAI, Y. Y. and SIEGEL, J. M. (1991) Pontomedullary glutamate FOUTZ, A. S., DELASHAW,J. 1~., GUILLEMINAULT,C. and DEMENT, W. receptors mediating locomotion and muscle tone suppression. J. C. (1981) Monoaminergic mechanisms and experimental cata- Neurosci. II, 2931-2937. plexy. Ann. Neurol. 10, 369-376. LANGOON, N., WELSH, K. I., VANDAM, M., VAUGHAN, R. W. and FREDERtKSON, S., CARLANOER, B., BiLLiARD,M. and LtNg, H. (1990) PARKES, D. (1984) Genetic markers in narcolepsy. Lancet 2, CSF immune variables in patients with narcolepsy. Acta neurol. 1178-1180. stand. 81,253-254. LANGDON, N., SHINDLER, J., PARKES, J. D and BANDAK, S. (1986) FRUHSTORFER, B., MIGNOT, E., BOWERSOX, S., NISHINO, S. and Fluoxetine in the treatment of cataplexy. Sleep 9, 371 373. DEMENT, W. C. (1989) Canine narcolepsy is associated with an LAVIE, P., SO, A. K. L. and PELED, R. (1987) Narcolepsy is a rare elevated number of alpha 2-receptors in the locus coeruleus. Brain disease in Israel. Sleep 10, 608~09. Res. 500, 209-214 LOCK, C. B., SO, A. K. L., WELSH, K. I., PARKES, J. D. and TROWS- GAISER, C., BURTCH, M., RADDE, L., DEMENT, W. and STEINMAN, L. DALE, J. (1988) MHC Class II sequences of an HLA-DR2 nar- (1989) Evidence for an autoimmune etiology in canine cataplexy. coleptic. Immunogenetics 27, 449-455. Sleep Res. 18, 230. LUCAS, E. A., FOUTZ, A. S., DEMENT, W. C. and MITLER, M. M. GEORGE, C. F. P., OTT, R. N., S1NGH, S. M., GELB, M,, GRU~'T, F. C. (1979) Sleep cycle organization in narcoleptic and normal dogs. and MIGNOT, E. (1992) IgM (mu switch) region restriction frag- Physiol. Behav. 23, 737-743. ment length polymorphism in human familial narcolepsy. Sleep MATSUKI, K., GRUMET, F. C., LIN, X., GELB, M., GUILLEMINAULT, Res. 21,201. C., DEMENT, W. C. and MIGNOT, E. (1992) DQ (rather than DR) GUILLEMINAULT,C., M1GNOT, E., ALDRICH, M., QUERA-SALVA,M. A., gene marks susceptibility to narcolepsy. Lancet 339, 1052. TIBERGE, M. and PARTINEN, M. (1988) Prazosin contraindicated in MIGNOT, E., GUILLEMINAULT,C., BOWERSOX, S., RAPPAPORT, A. and patients with narcolepsy. Lancet 2, 511. DEMENT, W. C. (1988) Role of central alpha-I adrenoceptors in GU1LLEMINAULT, C., MmNOT, E. and GRUMET, F. C. (1989) Familial canine narcolepsy. J. clin. Invest. 82, 885-894. patterns of narcolepsy. Lancet 2, 1376-1379. MIGNOT, E., GUILLEMINAULT,C., BOWERSOX, S. et al. (1989) Central HARRIS, L., DOUGLASS, A. B., PAZDERKA, F. and DOSSETOR, J. B. alpha I adrenoceptor subtypes in narcolepsy-cataplexy: a dis- (1987) Patterns of HLA-DR2 inheritance in families with nar- order of REM sleep. Brain Res. 490, 186-191. colepsy and hypersomnolence. Sleep Res. 16, 347. MIGNOT, E., HUNT SHARP, L., ARRIGONI, J., NISHINO, S., GUILLEM- HAWKINS, M., O'CONNOR, S., RADULOVACKI, M., BOWERSOX, S., INAULT, C., DEMENT, W. C. and CIARENELLO. R. (1991a) Gene MIGNOT, E. and DEMENT, W, (1991) Radioligand binding to dosage effects in canine narcolepsy: induction of cataplexy-like adenosine receptors and adenosine uptake sites in different brain behaviors in heterozygous dogs after drug challenge. Sleep Res. regions of normal and narcoleptic dogs. Pharmac. Biochem. Behae. 20, 294. 38, 1-6. MIGNOT, E., WANG, C., RATTAZZI, C., GAISER, C., LOVEIT, M., HISHIgAWA, Y. (1975) Sleep paralysis. In: Narcolepsy, pp. 97-124. GU1LLEMINAULT, C., DEMENT, W. C. and GRUMET, F. C. (1991b) Eds, C. GUILLEMINAULT, W. C. DEMENT and P. PASSOUANT. Genetic linkage of autosomal recessive canine narcolepsy with a Spectrum: New York. mu immunoglobulin heavy-chain switch-like segment. Proc. natn. HOLLOMAN, J. D., BELL, J. I., KILDUFF, T. S., DEMENT, W. C., Aead. Sei. U.S.A. ~, 3475-3478. GUILLEMINAULT,C. and MCDEVlTT, H. O. (1987) HLA-DR restric- MILLER, J. D., FAULL, K. F., BOW~RSOX, S. S. and DEMENT, W. C. tion-fragment-length polymorphisms in narcolepsy. J. Neurosei. (1990) CNS monoamines and their metabolites in canine nar- Res. 18, 239-244. colepsy: a replication study. Brain Res. 509, 169-171 HONDA, Y. (1979) Census of narcolepsy, cataplexy and sleep life MITLER, M. M. and DEMENT. W. C. (1977) Sleep studies on canine among teenagers in Fujisawa city. Sleep Res. 8, 191. narcolepsy: pattern and cycle comparisons between affected and HONDA, Y., JuJl, T., MATSUKI, K., NAOHARA,T., SATAKE, M., INOKO, normal dogs. Electroencephalogr. elin. Neurophysiol. 43, 691--699. H., SOMEYA, T., HARADA, S. and DOl, Y. (1986) HLA-DR2 and MONTPLAISIR, J. and GODBOUT, R. (1986) Nocturnal sleep of nar- Dw2 in narcolepsy and in other disorders of excessive somnolence coleptic patients revisited. Sleep 9, 159 161. without cataplexy. Sleep 9, 133-142. MONTPLAtSIR, J. and POIRIER, G. (1987) Narcolepsy in monozygotic IMERI, L., BIANCHI,S., ANGELI,P. and MANCIA, M. (1991) Differential twins. Neurology 37, 1089. effects of M2 and M3 muscarinic antagonists on the sleep-wake MONTPLAISIR, J., DECHAMPLA1N, J., YOUNG, S. N. et aL (1982) cycle. Neuro. Rep. 2, 383-385. Narcolepsy and idiopathic hypersomnia: biogenic amines and JONES, B. E. (1991) Paradoxical sleep and its chemical/structural related compounds in CSF. Neurology 32, 1299-1302 substrates in the brain. Neuroscience 40, 637-656. MOTOYAMA, M., KILDUFF, T. S., LEE, B. S. M., DEMENT, W. C. and JuJl, T., SATAKE, M., HONDA Y. and Dot, Y. (1984) HLA antigens in MCDEVlTT, H. O. (1989) Restriction tYagment length polymor- Japanese patients with narcolepsy--all the patients were DR2 phism in canine narcolepsy, lmmunogenetics 29, 124-126. positive. Tissue Antigens 24, 316-319 NEELY, S., ROSENBERG. R., SPIRE, J.-P., ANTEL, J. and ARNASON, B. JUJI, T., MATSUKI, K., TOKUNAGA, K., NAOHARA, T. and HONDA, Y, G. W. (1987) HLA antigens in narcolepsy. Neurology 37, (1988) Narcolepsy and HLA in the Japanese. Ann. N. Y. Aead. Sei. 1858-1860. 540, 106-114. NISHINO, S., MIGNOT, E., FRUHSTORFER, n., DEMENT, W. C. and KAIT1N, K. I., KILDUFF, T. S. and DEMENT, W. C. (1986) Evidence for HAYAtSHI, O. (1989) Prostaglandin E2 and its methyl ester reduce excessive sleepiness in canine narcoleptics. Electroencephalogr. cataplexy in canine narcolepsy. Pro('. HatH. Acad. Sei. U.S.A. 86, c/in. NeurophysioL 64, 447-454, 2483-2487. KESSLER, S., GUILLEMINAULT, C. and DEMENT, W. (1974) A family NISH1NO, S., HAAK, L., SHEPERD, H., GUILLEMINAULT, C., SAKAI, T., study of 50 REM narcoleptics. Aeta neurol, stand..~0, 503 512. DEMENT, W. C. and M1GNOT, E. (1990) Effects of central alpha- KILDUEF. T. S., BOWERSOX, S. S., KAITIN, K. I., BAKER, T. L., 2 adrenergic compounds on canine narcolepsy, a disorder of CfARENELLO, R. D. and DEMENT W. C. (1986) Muscarinic cholin- rapid eye movement sleep. J. Pharmac. exp. Tl~er. 253, ergic receptors and the canine model of narcolepsy. Sleep 9, 1145 1152. 102-106. NISHINO, S., ARRIGONI, J., VALTIER, D., MILLER, J. D., GUILLEM- KISH, S. J., MAMELAK, M., SLIMOVITCH,C. el a[. 0992) Brain neuro- INAULT, C., DEMENT, W. C. and MIGNOT, E. (1991) Dopamine transmitter changes in human narcolepsy. Neurology 42, 229-234. D2 mechanisms in canine narcolepsy. J. Neurosei. II, KRAMER, R. E., DINNER, D. S., BRAUN, W. E., ZACHARY, A. and 2666-2771. TERESl, TERESt, G. A. (1987) HLA-DR2 and narcolepsy. Arch. NISHINO, S., ARRIGONI,J., SHELTON, J., DEMENT, W. C. and MIGNOT, Neurot. 44, 853-855. E. (1992) Further analysis of the effects of serotonergic and NEUROBIOLOGY OF NARCOLEPSY-CATAPLEXY 541

noradrenergic uptake inhibitors on canine eataplexy. Sleep Res. SIEGEL, J. M. (1989) Brainstem mechanisms generating REM sleep. 21, 66. In: Principles and Practice of Sleep Medicine, pp. 104-120. Eds. NISHINO, S., ARRIGONI,J., SHELTON,J., FRUHSTORFER,B., GU1LLEM- M. H. KRVG~, T. Ro~4 and W. C. DEMENT. WB Saunders: INAULT,C., DEMENT,W. C. and MIGNOT,E. (1992) Further charac- Philadelphia. terization of the alpha-I receptor subtype involved in the SIEGEL,J. M., NmNHUlS,R., FAHRINGER,H. M. et aL (1991) Neuronal regulation of cataplexy. Sleep Res. 21, 245. activity in narcolepsy--Identification of cataplexy-related cells in OL~UP, O., SCHAF~R, M., HILLERT, J. and SACHS, C. (1990) The the medial medulla. Science 252, 1315-1318. narcolepsy-associated DRwl5, DQw6, Dw2 haplotype has no StNGH, S. M., GEORGE, C. F. P., KRYGER, M. H. and JUNG, J. H. unique HLA-DQA or -DQB restriction fragments and does not (1990) Genetic heterogeneity in narcolepsy. Lancet 335, extend to the HLA-DP subregion, lmmunogenetics 32, 41--44. 726-727. PARKES, J. D., FENTON, G., STRUTHERS, G. et al. (1974) Narcolepsy STROHMAIER, P., MUELLER-ECKHARDT, G. and MEIER-EWERT, K. and cataplexy. Clinical features, treatment and cerebrospinal fluid (1988) Cellular approach for detecting narcolepsy-specific alter- findings. Quart. J. Med. 43, 525-526. ations in DR2 haplotypes. Hum. Immun. 22, 221-225. RECHTSCHAFFEN, A., WOLPERT, E. A., DEMENT, W. C., MITCHELL, S. TAKAHASm, Y. and JIMBO, M. (1963) Polygraphic study of A. and FtSm~R, C. (1963) Nocturnal sleep of narcoleptics. Elec- narcoleptic syndrome, with special reference to hypnagogic hallu- troencephalogr, clin. Neurophysiol. 15, 599~509. cinations and cataplexy. Folia Psychiatr. Neurol. Jpn. (Suppl.) 7, REID, M. S., GEARY,J. N., SIEGEL,J. S., DEMENT,W. C. and MIGNOT, 343-347. E. (1992) Acetylcholine release in the medial pontine reticular URYU, N., MAEDA, M., NAGATA,Y., MATSUKI, K., JuJI, T., HONDA, formation is enhanced during cataplexy in the narcoleptic canine. Y., K^WAI, J., AN~O, A., Tsua, K. and INOKO, H. (1989) No Sleep Res. 21, 253. difference in the nucleotide sequence of the DQ(beta)l domain RENAUD, A., NISHINO,S., ARRIGONI,J., GUILLEMINAULT,C., SAKAI, between narcoleptic and healthy individuals with DR2, Dw2. T., DEMENT, W. C. and MIGNOT, E. (1991) Effects of selective Hum. Immun. 24, 175-181. monoamine uptake inhibitors and release enhancers on canine VANNI-MERcmR, G., SAKAt, K., LIN, J. S. and JOUVET, M. (1989) cataplexy. Sleep Bes. 20, 76. Mapping of cholinoceptive brainstem structures responsible for ROSENaMAL, L., ZORICK, F., MERLOTTI, L., WITTtG, R., ROEHRS, T. the generation of paradoxical sleep in the cat. Arch. hal. Biol. 127, and ROTH, T. (1990) Signs and symptoms associated with cata- 133-164 plexy in narcolepsy patients. Biol. Psych. 27, 1057-1060. VELAZQUEZ-MOCTEZUMA,J,, GILLIN,J. C. and SmROMANI,P. J. (1989) RUBIN, R. L., HAJDUKOVlCH,R. M. and MITLER, M. M. (1988) HLA- Effect of specific MI, M2 muscarinic receptor agonists on REM DR2 association with excessive somnolence in narcolepsy does not sleep generation. Brain Res. 503, 128-131. generalize to sleep apnea and is not accompanied by systemic VELAZQUEz-MoCTEZUMA,J., SHALAUTA,M., GILLIN,J. C. and SNmO- autoimmune abnormalities. Clin. lmmun. Immunopath. 49, MANI, P. J. (1991) Cholinergic antagonists and REM sleep gener- 149-158. ation. Brain Res. 543, 175-179. SAKAI, K. (1988) Executive mechanisms of paradoxical sleep. Arch. VO~EL, G. (1960) Studies in psychophysiology of dreams. III. The Ital. Biol. 26, 239-257. dream of narcolepsy. Arch. gen. Psych. 3, 421428. SCHENCK, C. H., BUNDLIE,S. and MAHOWALD, M. H. (1989) Nar- WIL~R, A., STEINMAN,L., LAVIE, P., PELED, R., FRIEDMANN,A. and colepsy, loss of REM atonia and the REM sleep behavior BRAUTBAR, C. (1988) Narcolepsy-cataplexy in Israeli Jews is disorder: a polysomnographic and clinical report on 10 patients. associated exclusively with the HLA DR2 haplotype. Hum. Im- Sleep Bes. 18, 300. mun. 21, 15-22.