Normal plasma levels of orexin A (hypocretin-1) in narcoleptic patients

Methods.

Thirty-one subjects were included in the current study. In every subject, a lumbar puncture was performed for diagnostic purposes after informed consent. Eleven patients with narcolepsy (eight men, three women, mean age 42.3 ± 18.9, mean body mass index [BMI, kg/m²] 26.5 ± 5.0) suffering from clear-cut cataplexy, excessive daytime sleepiness, and showing multiple sleep-onset REM periods during diagnostic polysomnography were included. All but one patient carried the HLA-DR2 antigen. Twenty patients suffering from noninflammatory neuropsychiatric disorders (schizophrenia [n = 5], somatoform disorder [n = 3], affective disorders [n = 4], adjustment disorder [n = 2], personality disorder [n = 1], obsessive–compulsive disorder [n = 1], migraine [n = 1], PD [n = 1], dementia [n = 2]) were included as control subjects (14 men, 6 women, mean age 44.0 ± 18.7, mean BMI 25.2 ± 4.4). The groups did not differ with respect to age (t[1,29] = 0.25, NS), sex distribution (Fisher’s exact test, NS), and BMI (t[1,29] = 0.79, NS).

Lumbar punctures were performed between 8 and 9 am with the patient in a sitting position. Immediately thereafter, 10 mL of blood were withdrawn, stabilized with ethylenedinitrieo tetraacid acid (1 mg/mL blood) and aprotinin (300 KIU/mL blood) and centrifuged for 10 minutes at 4000 × g at 4 °C. CSF and plasma were frozen at −80 °C and −20 °C, respectively.

Prior to the determination of orexin A levels, plasma was extracted with a SPEC-3ML-C18AR column (Ansys Diagnostics, Lake Forrest, CA). Orexin A levels were determined by a highly sensitive commercially available radioimmunoassay (Phoenix Pharmaceuticals, Mountain View, CA). The detection limit was 40.0 pg/mL. Each sample was tested in duplicate, and the mean intra-assay coefficient of variation was 9.25%. The assay did not cross-react with orexin B (human), agouti-related protein (83–132)-amide (human), neuropeptide (human, rat), α-melanocyte-stimulating hormone, and leptin (human).

Fisher’s exact test was used to compare the number of detectable versus nondetectable orexin A values in CSF between both groups. Age, BMI, and plasma orexin levels were compared between groups using the Student’s t-test. Correlation coefficients were calculated according to Pearson’s method. In the figure, data are depicted as the mean ± SEM. The level of significance was set to p < 0.05.

• Download figure
• Open in new tab
• Download powerpoint

Figure. CSF orexin A levels (A) and plasma orexin A levels (B) in pg/mL, mean ± SEM, in 11 narcoleptic patients and 20 control patients with various neuropsychiatric disorders. For statistics see text. Horizontal bars indicate the arithmetic mean.

Results.

As shown in the figure, panel A, 9 of 11 narcoleptic patients showed orexin A levels below the detection limit in CSF whereas in every control subject orexin A was detectable (χ², p < 0.05). In contrast, orexin A was detectable in plasma of all patients (range: 175 to 847 pg/mL). Plasma orexin A levels did hardly differ between the narcoleptic and control groups (see the figure, panel B; t[1,29] = 0.07, NS). In control subjects, orexin A levels clearly were lower in plasma compared with CSF (487.2 ± 154.2 vs 739.0 ± 211.0 pg/mL, t[1,19] = 4.23, p < 0.05). The plasma concentrations of orexin A did not correlate with CSF concentrations (r = −0.04, NS; this correlation was computed in controls only because of the virtual absence of orexin A in the CSF of patients with narcolepsy). Furthermore, plasma orexin A levels did not correlate with age (r = 0.16, NS) and BMI (r = −0.01, NS). Also, sex had no influence on plasma orexin A concentrations (431.0 ± 216.2 pg/mL [female subjects] versus 508.0 ± 150.4 pg/mL [male subjects], t[1,11.3] = 0.98, NS).

Discussion.

In the current study, we found very low CSF levels of orexin A in patients with narcolepsy, confirming one earlier report.1 In our sample, CSF orexin A levels in all patients with narcolepsy were below the range of values observed in controls. In the study of Nishino et al.,1 one patient showed a CSF orexin A concentration in the range of the control population, and one patient’s value was even above this range. Hence, orexin A deficiency in the CNS seems to be present in most, but not in all, patients with narcolepsy. In our sample, the control subjects showed CSF orexin A levels higher than those reported by Nishino and colleagues, and the values showed higher variability. The reasons for this discrepancy are unclear but could be related to differences in methodology and characteristics of the samples.

We showed here for the first time that orexin A can be detected in human plasma. In control subjects with various neuropsychiatric disorders, plasma orexin A levels were approximately 30% below those observed in CSF. However, plasma and CSF levels did not correlate. This finding suggests that systemic orexin might be derived from CNS-independent sources, such as neuronal cells within the gut.6 The results of the current study indicate that orexin production by these sources is preserved in human narcolepsy despite the virtual absence of orexin synthesis in the central compartment.2-3⇓

It is known that orexin deficiency in human narcolepsy is not genetically determined in the majority of patients,3 indicating that yet unknown environmental factors are of pivotal importance. Because of the tight association of human narcolepsy to the HLA-DR2/DQB1*0602, it is tempting to speculate about a possible involvement of autoimmune phenomena.7-8⇓ Because we found orexin A in normal quantities in the systemic circulation, it is unlikely that any kind of a presumed autoimmune attack on the orexin system is directed against the orexin A molecule. Instead, it is likely that such processes specifically target hypothalamic orexin-producing cells.

The orexin system might play a central role for future developments in the treatment of narcolepsy and other disorders of excessive daytime sleepiness. Animal experiments showed that orexin A readily crosses the blood–brain barrier.9 However, this is unlikely to be the case in humans, because CSF orexin A levels in patients with narcolepsy are extremely low despite normal plasma orexin A levels. Hence, it may be necessary to develop small, nonpeptide molecules targeting orexin receptors.