In 1999, it was discovered that a genetically induced deficiency of hypocretin neurotransmission in dogs and mice leads to a complex of symptoms remarkably similar to that seen in human narcolepsy.6'7 These findings focused attention on the role of hypocretin in the pathophysiology of human narcolepsy.
A mutation screening of the hypocretin genes was performed in 74 patients, including patients with and without a family history of narcolepsy and patients with and without HLA DQB1*0602.8 Of these, only one patient carried a disease-causing mutation: a G-to-T transversion in the hypocretin signal peptide. The disease phenotype of this patient differed significantly from the classical pattern: cataplectic attacks with a frequency of five to 20 per day had started at the age of 6 months. This low mutation yield was not surprising from a clinical point of view, as typical narcolepsy does not run in families. More unexpectedly no mutations in the hypocretin genes were identified in the uncommon familial cases. Up till now there have been no reports about any other human case or family with an identified defect (including polymorphisms) in the pre-prohypocretin or the hypocretin receptor genes.9,10 However, there are known associations with other genes, which may be considered as susceptibility genes. The strongest association known is with the HLA DQB1*0602 allele, present in over 90% of narcoleptic patients.11 As about 20% of the population carries this allele and only 0.05% of the population develops narcolepsy, this association is not very specific. Interestingly, the uncommon familial cases show a much weaker association with this HLA type.12 Associations with polymorphisms in the genes for monoamine-oxidase-A, tumor necrosis factor alpha, and catechol-O methyltransferase have also been reported.13,14
Despite the lack of an identified genetic cause, hypocretin deficiency as cause of human narcolepsy remained an attractive hypothesis. To further study this an assay for the measurement of hypocretin was developed and cerebrospinal fluid (CSF) hypocretin-1 measurements were performed blindly in a small series of patients and controls.15 The astonishing finding was that 7 out of 9 patients had an undetectable low hypocretin-1 level, in contrast to the control group having levels about 7 times the detection limit, and all in a narrow range. Subsequently, a large extension study was performed to confirm these findings and establish the sensitivity and specificity.16,17 Several hundreds of subjects were included: healthy controls, patients with typical narcolepsy, atypical narcolepsy, idiopathic hypersomnia, sleep apnea, restless legs, periodic limb movements disorder, insomnia patients as well as patients suffering from a wide variety of neurological disorders. From these data the sensitivity and specificity of hypocretin deficiency for narcolepsy turned out to be very high, 87% and 99% respectively. Typical cataplexy, characterized by symmetrical weakness particularly triggered by laughter, was the best predictor of hypocretin deficiency in these studies. Narcoleptic patients with atypical or no cataplectic attacks, HLA DQB1*0602 negative patients and those suffering from familial forms of narcolepsy often show normal hypocretin-1 levels. These findings suggest other possible causes, such as "downstream" defects in the hypocretin system, involvement of substances that modulate hypocretin cells, non-genetic acquired receptor defects, and even leave the possibility for causes other than hypocretin deficiency. The identification of these potential causes is hampered by the limited knowledge concerning the hypocretin system.
The only other relatively frequently occuring disorder in which undetectable hypocretin-1 levels were found is the Guillain-Barre syndrome.18,19 The explanation is unclear, and the deficiency is probably transitory. Whether this deficiency is accompanied by clinical symptoms has yet to be established. The Guillain-Barre studies were retrospective in nature and concerned patients with severe paralysis requiring artificial ventilation, preventing a reliable assessment of narcoleptic symptoms.
The identification of hypocretin-1 deficiency as a specific marker for human narcolepsy without any overt genetic explanation raised the question about the most plausible cause. Two controlled post-mortem neuropathological studies shed some light on this crucial question. One study used in situ hybridization against hypocretin mRNA in frozen brain tissue.8 In the two brains a complete disappearance of hypocretin mRNA was demonstrated. Melanin-concentrating hormone mRNA, a component of adjacent non-hypocretin cells, was preserved in these patients, just as in control subjects. To extend these findings, hypocretin-1 and 2 peptide levels were measured in brain tissue, cortex and pons, of eight control subjects and six patients.8 Both hypocretin-1 and hypocretin-2 were absent in the brains of the narcoleptic patients. Furthermore, there were no signs of inflammation in two brains studied: tumor necrosis factor-alpha levels and HLA DR2 expression were normal. There were also no obvious signs of gliosis or global cellular loss in the region of interest. In the other study, immunohistochemistry after antigen retrieval in fixed tissue was used (four brains - one from a patient without cataplexy).20 A dramatically reduced number of neurons expressing hypocretin 1 and 2, when compared with control brains, was found. Five to 15% of hypocretin neurons were preserved. Interestingly, the highest number of remaining neurons was found in the one patient without cataplexy. In this study an increased number of GFAP stained astrocytes in the narcoleptic brains was found, supporting a degeneration of neurons.
These findings suggest that normally functioning hypocretin cells are present in patients early in life, and that narcoleptic symptoms appear only after degeneration of the majority of cells. However, it must be kept in mind that these assumptions have not yet been proven. In particular it has not been proven whether the hypocretin cells degenerate or only have an impaired hypocretin synthesis or release. The main reason for this lack of knowledge is the lack of a hypocretin-independent marker of hypocretin cells. Structural studies using advanced MRI techniques have not settled this question; voxel based morphometry yielded contradictory results.21-23 Additional factors impeding the identification of the possible cell loss are the low number of cells and their anatomical distribution: they are not clustered in a nucleus but are intermingled with other cells in quite a large area of the perifornical hypothalamus.
Although CSF hypocretin-1 deficiency in humans correlates with the presence of excessive daytime sleepiness (EDS), it is most specifically correlated with the presence of cataplexy,17 similar to the findings in animal models. Unfortunately, we are not informed whether undetectable levels reflect a total absence of (biological active) hypocretin, although this seems likely.24 The exact role of hypocretin deficiency, respectively hypocretin loss, in the pathophysiology of human narcolepsy has therefore not been established. The extent of the hypocretin deficiency may possibly determine whether patients only suffer from EDS or do also develop cataplexy. This is suggested by one of the post mortem brain studies, showing more residual hypocretin cells in the only patient without cataplexy.20 However, firm conclusions cannot be drawn from one patient without information regarding CSF hypocretin-1 levels. Alternatively, the difference between narcolepsy with and without cataplexy may be explained by a differential disturbance in the function of hypocretin 1 and 2 or their respective receptors, as the type of lesion in the various animal models leads to differential behavioral phenotypes (see below).25,26 However, a completely different etiology, independent of a disturbed hypocretin transmission may also be possible.
To return to the historical debate about narcolepsy and the question whether the presence of cataplexy is a prerequisite for the diagnosis, and whether this is a disease sui generis, we still have no definite answer. The few cases with isolated long lasting EDS, typical REM sleep abnormalities on polysomnography testing and undetectable levels of hypocretin-1 challenge this view.17 However, the very high sensitivity and specificity of undetectable levels of hypocretin for the presence of cataplexy underscores that cataplexy is a cardinal symptom.
Was this article helpful?