Silencing of Hypoglossal Motoneurons Leads to Sleep Disordered Breathing in Lean Mice

Abstract

Obstructive Sleep Apnea (OSA) is a prevalent condition and a major cause of morbidity and mortality in Western Society. The loss of motor input to the tongue and specifically to the genioglossus muscle during sleep is associated with pharyngeal collapsibility and the development of OSA. We applied a novel chemogenetic method to develop a mouse model of sleep disordered breathing Our goal was to reversibly silence neuromotor input to the genioglossal muscle using an adeno-associated viral vector carrying inhibitory designer receptors exclusively activated by designer drugs AAV5-hM4Di-mCherry (DREADD), which was delivered bilaterally to the hypoglossal nucleus in fifteen C57BL/6J mice. In the in vivo experiment, 4 weeks after the viral administration mice were injected with a DREADD ligand clozapine-N-oxide (CNO, i.p., 1mg/kg) or saline followed by a sleep study; a week later treatments were alternated and a second sleep study was performed. Inspiratory flow limitation was recognized by the presence of a plateau in mid-respiratory flow; oxyhemoglobin desaturations were defined as desaturations >4% from baseline. In the in vitro electrophysiology experiment, four males and three females of 5 days of age were used. Sixteen-nineteen days after DREADD injection brain slices of medulla were prepared and individual hypoglossal motoneurons were recorded before and after CNO application. Positive mCherry staining was detected in the hypoglossal nucleus in all mice confirming successful targeting. In sleep studies, CNO markedly increased the frequency of flow limitation n NREM sleep (from 1.9 ± 1.3% after vehicle injection to 14.2 ± 3.4% after CNO, p < 0.05) and REM sleep (from 22.3% ± 4.1% to 30.9 ± 4.6%, respectively, p < 0.05) compared to saline treatment, but there was no significant oxyhemoglobin desaturation or sleep fragmentation. Electrophysiology recording in brain slices showed that CNO inhibited firing frequency of DREADD-containing hypoglossal motoneurons. We conclude that chemogenetic approach allows to silence hypoglossal motoneurons in mice, which leads to sleep disordered breathing manifested by inspiratory flow limitation during NREM and REM sleep without oxyhemoglobin desaturation or sleep fragmentation. Other co-morbid factors, such as compromised upper airway anatomy, may be needed to achieve recurrent pharyngeal obstruction observed in OSA.

Keywords: chemogenetic; neuromuscular activity; obstructive sleeep apnea; sleep; upper airway.

Figure 1

Localization of AAV5-hSyn-hM4 (Gi)-mCherry DREADD in the hypoglossal nucleus. Fluorescent microscopy images (× 10) show mCherry expression spanning the hypoglossal nucleus. 12N denotes the hypoglossal nucleus; 4V denotes the fourth ventricle.

Figure 2

Localization of AAV5-hSyn-hM4 (Gi)-mCherry and fluorescently labeled cholera toxin B (CTB-AF488) in the hypoglossal nucleus. Fluorescent microscopy images (× 20) shows (A) CTB labeled motoneurons within the rostral hypoglossal nucleus following genioglossal injection; (B) mCherry expression in the same section; (C) merged mCherry and AF488 images. 4V denotes 4th ventricle.

Figure 3

Localization of AAV5-hSyn-hM4 (Gi)-mCherry and fluorescently labeled cholera toxin B (CTB-AF488) in the hypoglossal nucleus. Fluorescent microscopy images (× 20) shows (A) CTB labeled motoneurons more caudally, at the obex level of the hypoglossal nucleus, following genioglossal injection; (B) mCherry expression in the same section; (C) merged mCherry and AF488 images. ^* denotes central canal.

Figure 4

In-vitro activation of hypoglossal neurons that contain inhibitory DREADDs. (A) Representative example demonstrates a depression of action potential firing of DREADDs-containing hypoglossal neuron recorded in current-clamp configuration after 1-min CNO (10 μM) application. (B) The summary data from 7 neurons illustrate significant (P < 0.001, 1-way ANOVA) inhibition of firing activity of hypoglossal neurons that started as early as 2 min post CNO application and lasted for at least 20 min post CNO application. ^*** P < 0.001.

Figure 5

(A) A representative trace of NREM sleep after saline and after CNO delivery in lean mice. Panels shows compressed recording of EEG, nuchal EMG, respiratory flow, effort and pulse oximetry (SpO2); The asterisks indicate flow limited breaths; (B) The shaded area on the right panel is decompressed.

Figure 6

Percentage of flow limited breaths in both NREM sleep (1.9% ± 1.3% baseline vs. 14.2% ± 3.4% CNO, p < 0.05) and REM sleep (22.3% ± 4.1% baseline vs. 30.9% ± 4.6% CNO, p < 0.05). ^*p < 0.05.