Expression of a poly-glutamine-ataxin-3 transgene in orexin neurons induces narcolepsy-cataplexy in the rat

Abstract

The sleep disorder narcolepsy has been linked to loss of hypothalamic neurons producing the orexin (hypocretin) neuropeptides. Here, we report the generation of transgenic rats expressing a human ataxin-3 fragment with an elongated polyglutamyl stretch under control of the human prepro-orexin promoter (orexin/ataxin-3 rats). At 17 weeks of age, the transgenic rats exhibited postnatal loss of orexin-positive neurons in the lateral hypothalamus, and orexin-containing projections were essentially undetectable. The loss of orexin production resulted in the expression of a phenotype with fragmented vigilance states, a decreased latency to rapid eye movement (REM) sleep and increased REM sleep time during the dark active phase. Wakefulness time was also reduced during the dark phase, and this effect was concentrated at the photoperiod boundaries. Direct transitions from wakefulness to REM sleep, a defining characteristic of narcolepsy, occurred frequently. Brief episodes of muscle atonia and postural collapse resembling cataplexy were also noted while rats maintained the electroencephalographic characteristics of wakefulness. These findings indicate that the orexin/ataxin-3 transgenic rat could provide a useful model of human narcolepsy.

Figure 1.

A–H, Colocalization of orexin neuropeptide and the orexin/ataxin-3 transgene product (A, B) and disappearance of orexin-positive neurons (C–F) and projections (G, H) in orexin/ataxin-3 hemizygous transgenic rats. The LH region of wild-type rats (A) and orexin/ataxin-3 hemizygous transgenic littermates (B) was stained using antibodies against orexin-A (brown) and against the Myc tag of the transgene (black). All transgenic animals showed nuclear staining of the transgene product only in orexin neurons. Wild-type animals showed a normal distribution of orexin neurons in the LH at 4 weeks (C) and 17 weeks (E) of age. In contrast, orexin/ataxin-3 transgenic littermates showed a notable reduction of orexin-immunoreactivity by 4 weeks of age (D) and a virtually complete loss of immunoreactivity at 17 weeks of age (F). No difference in MCH neuronal population in the perifornical area could be found at 17 weeks between wild-type rats (E, inset) and their orexin/ataxin-3 hemizygous transgenic littermates (F, inset). Dense orexin-containing projections in the thalamic paraventricular nucleus of wild-type animals (G) were undetectable in transgenic animals at 17 weeks of age (H). f, Fornix; 3V, third ventricle. Scale bars: A, B, 40 μm; C–H, insets, 200 μm. A–F, Bright-field microscopy; G, H, dark-field microscopy.

Figure 2.

A, B, Representative dark-phase hypnograms of a wild-type rat (A) and an orexin/ataxin-3 hemizygous transgenic littermate (B). The hypnogram of the transgenic animal shows more rapid cycling between vigilance states. It also exhibits direct transitions from wakefulness to REM sleep (indicated by arrowheads). W, Wakefulness; NR, NREM sleep; R, REM sleep.

Figure 3.

Representative EEG–EMG recordings from an orexin/ataxin-3 hemizygous transgenic rat. A, Normal transition from wakefulness to NREM sleep. The EEG signal increases in amplitude and slows in frequency, whereas neck muscle tone diminishes in the EMG signal. B, Abnormal direct transition from wakefulness to REM sleep. The EEG signal starts with the typical mixed-frequency, low-amplitude, wakefulness pattern, which directly gives way to the regular low-amplitude pattern of REM sleep, dominated by θ activity in the 6–10 Hz range, with concomitant complete neck muscle atonia. C, Cataplexy-like event. Although the EEG characteristic of wakefulness remains unchanged, a sudden and transient complete neck muscle atonia occurs, after which normal muscle activity suddenly resumes. Note that θ activity in the EEG signal is less than that recorded during REM sleep, and that no visual indications of NREM sleep are apparent during this interval.

Figure 4.

Power spectra of EEG–EMG recordings from wild-type rats and their orexin/ataxin-3 hemizygous transgenic littermates. A, Representative EEG power spectra of wakefulness, NREM sleep, and REM sleep in a wild-type rat. B, Representative EEG power spectra of an orexin/ataxin-3 transgenic littermate. SOREM (i.e., sleep-onset REM) episodes designate those REM sleep episodes that follow an abnormal transition from wakefulness. Note that the mean EEG frequency distribution of SOREM episodes is essentially identical to that recorded during normally occurring REM sleep in the same animal. The signal amplitudes in A and B have been normalized in each animal to allow comparison across animals. C, D, Representative absolute EEG power spectra of REM sleep and SOREM episodes (C) and of wakefulness and cataplexy-like events in an orexin/ataxin-3 hemizygous transgenic rat (D). E, Average EMG signal power recorded during the designated vigilance states from a representative orexin/ataxin-3 hemizygous transgenic rat. Data are expressed as mean ± SEM.

Figure 5.

A, B, Time spent each hour (in minutes; mean ± SEM) in REM sleep (A) and wakefulness (B) for wild-type rats and their orexin/ataxin-3 hemizygous transgenic littermates. Significant differences between the genotypes (t test; p < 0.05) are marked by asterisks. The dark period is denoted by the horizontal bar.