Vagal Sensory Neuron Subtypes that Differentially Control Breathing

Abstract

Breathing is essential for survival and under precise neural control. The vagus nerve is a major conduit between lung and brain required for normal respiration. Here, we identify two populations of mouse vagus nerve afferents (P2ry1, Npy2r), each a few hundred neurons, that exert powerful and opposing effects on breathing. Genetically guided anatomical mapping revealed that these neurons densely innervate the lung and send long-range projections to different brainstem targets. Npy2r neurons are largely slow-conducting C fibers, while P2ry1 neurons are largely fast-conducting A fibers that contact pulmonary endocrine cells (neuroepithelial bodies). Optogenetic stimulation of P2ry1 neurons acutely silences respiration, trapping animals in exhalation, while stimulating Npy2r neurons causes rapid, shallow breathing. Activating P2ry1 neurons did not impact heart rate or gastric pressure, other autonomic functions under vagal control. Thus, the vagus nerve contains intermingled sensory neurons constituting genetically definable labeled lines with different anatomical connections and physiological roles.

Figure 1. Genetic Control of Sensory Neuron Types in the Vagus Nerve

(A) RNA in situ hybridization experiments in the nodose/jugular complex revealed that P2ry1, Npy2r, and Gpr65 are expressed in subsets of vagal sensory neurons. (B) Two color in situ hybridization experiments for indicated genes revealed largely non-overlapping neuron populations. The numbers of cells expressing one receptor (red or green) or both receptors (yellow) were counted. (C) The indicated Cre lines were crossed with lox-L10-GFP mice, and in offspring, fixed cryosections of the nodose/jugular complex were imaged by fluorescence microscopy. Native GFP fluorescence (green) and a fluorescent Nissl stain (gray) were visualized. Scale bars, 100 mm. See also Figures S1 and S2.

Figure 2. Visualizing Vagal Afferents in the Lung

(A) Cartoon depiction of the neural tracing strategy, which involved infection of a Cre-dependent AAV (AAV-flex-tdTomato) and/or a Cre-independent AAV (AAV-eGFP) in the nodose/jugular complex of ires-Cre knockin mice. (B) Whole-mount analysis of native tdTomato fluorescence in a flattened lung lobe from a Vglut2-ires-Cre mouse infected with AAV-flex-tdTomato. Scale bar, 1 mM. (C) Whole-mount analysis (maximum projection of stacked confocal images) of native tdTomato (tdT) and GFP fluorescence in the nodose/jugular complex of a Vglut2-ires-Cre mouse infected with AAV-flex-tdTomato and AAV-eGFP. Scale bar, 100 µm. (D) Different ires-Cre lines were infected with AAV-flex-tdTomato and AAV-eGFP, and fibers were visualized in fixed lung cryosections by immunohistochemistry for tdTomato (red) and GFP (green). Scale bars, 1 mm. (E) Quantitative analysis of lung innervation in 17.5-mm² lung regions expressed as an area ratio of T/G-derived immunofluorescence (mean ± SEM, see Results and Extended Experimental Procedures for additional detail on T/G calculation). (F) High resolution image of a representative vagal afferent beneath the epithelial layer, (E), of a major airway (airway lumen, L). Scale bar, 20 µm. (G) Representative P2ry1 candelabra terminal (tdT fluorescence) at a neuroepithelial body (CGRP immunostaining, green). Scale bar, 20 µm. (H) The number of neuroepithelial bodies (NEBs) innervated by each neuron type after visualization with AAV-flex-tdTomato and normalization with AAV-eGFP (n = 3–5, mean ± SEM, **p < 0.01, and ***p < 0.001). See also Figure S3.

Figure 3. Characterization of Vagal P2ry1 and Npy2r Neurons

(A) Cartoon depiction of optogenetic strategy. The vagus nerve is surgically exposed in anesthetized mice and illuminated to activate ChR2-expressing sensory neurons. (B) Whole nerve electrophysiological recordings in Vglut2-ChR2 mice revealed light-induced action potentials. (C) Compound action potentials following brief optogenetic stimulation (arrow) in Vglut2-ChR2, P2ry1-ChR2, and Npy2r-ChR2 mice. A and C fibers were classified based on conduction velocity (Figure S4) x = 5 ms, y = 110 µV (Vglut2), 62 µV (P2ry1), 160 µV (Npy2r), and dashed inset, x = 1.45 ms. (D) The ratio of A to C fibers was calculated by integrating corresponding peak area in the compound action potential; dashed line: A/C ration of1 (n = 5–8, mean ± SEM, and *p < 0.05). (E) Calcium imaging of single neuron responses to capsaicin (2 µM) and KCl (50 mM) in acute ganglia cultures from P2ry1-ires-Cre; lox-L10-GFP and Npy2r-ires-Cre; lox-L10-GFP mice. (Left panels, color scale = 340/380 nm Fura-2 excitation ratio and right panels, neurons expressing GFP [green, native fluorescence] and responding to capsaicin [red] are superimposed and counted). Scale bar, 100 µm. (F) Representative traces for single neurons imaged in (E). See also Figure S4.

Figure 4. Remote Control of Breathing

(A) Respiratory effects following focal vagus nerve illumination (yellow shading) in lox-ChR2, Vglut2-ChR2, P2ry1-ChR2, Npy2r-ChR2, and Gpr65-ChR2 mice. Respiratory rhythms (representative traces) were measured using a pressure transducer via trachea cannula. Changes in respiration rate and minute volume were calculated over time, with each data point reflecting a 5 s bin. (B) Light-induced changes in respiration rate, tidal volume, and minute volume were calculated over the 10 s trial (n = 4–8 as indicated, mean ± SEM, and ***p < 0.001). See also Figures S5 and S6.

Figure 5. P2ry1 Neurons Trap Respiration in a State of Exhalation

(A) Theoretical models of lung volume changes during light-induced inhalation and exhalation trapping. (B) Representative data showing changes in lung volume following optogenetic activation of vagal afferents in P2ry1-ChR2 mice. (C) Percentage change in total lung volume evoked by light in P2ry1-ChR2 mice (n = 5) and control mice (lox-ChR2) (n = 8). Total lung volume was calculated by integrating lung volume across 10 s periods before and during light stimulation. (D) The percentage of time P2ry1-ChR2 mice and control mice were in a high lung volume state before, during, and after light stimulation. High volume state was defined as greater than mean volume during tidal breathing (mean ± SEM, *p < 0.05, and ***p < 0.001).

Figure 6. P2ry1 Neurons Acutely Control Breathing, but Not Gastric Pressure or Heart Rate

(A) Measurements (representative traces) of respiratory rhythm, heart rhythm, and gastric pressure following focal illumination (yellow shading) of the vagus nerve in anesthetized P2ry1-ChR2, Vglut2-ChR2, and control (lox-ChR2) mice. Heart rate was measured by ECG, with boxed insets showing rhythms before (1, left), during (2, middle), and after (3, right) light exposure. Intraluminal gastric pressure was measured using a pressure transducer inserted through the pyloric sphincter. (B) Changes in heart rate (normalized from a 30 s pre-stimulus period) were calculated over time, with each data point reflecting a 5 s bin. (C) Changes in heart rate and gastric pressure were calculated over the first 10 s or 3 min of light stimulation respectively (mean ± SEM and ***p < 0.001). See also Figure S5.

Figure 7. Non-Overlapping Central Projections of Vagal P2ry1 and Npy2r Neurons

(A) The nodose/jugular complex of P2ry1-ires-Cre and Npy2r-ires-Cre mice was infected with AAV-flex-tdTomato (AAV-flex-tdT) and AAV-eGFP. At 4 weeks after infection, fixed brainstem cryosections were analyzed by two color immunohistochemistry for tdTomato (red) and eGFP (green). Representative images of anterior and posterior brainstem containing the vagal projection field are shown (full rostral-caudal series, Figure S7). Solitary tract, sol; fourth ventricle, 4V; central canal, CC; area postrema, AP; L-NTS includes ventral, lateral, ventrolateral, interstitial, and intermediate NTS subnuclei; and M-NTS includes dorsolateral, dorsomedial, medial, and commisural NTS subnuclei. Scale bar, 100 µm. (B) Quantitative analysis of innervation by P2ry1 and Npy2r fibers in L-NTS, M-NTS, and AP, expressed as an area ratio of T/G fluorescence. Fluorescence was summed in every eighth section (25 µm) from Bregma −6.4 mm to −7.8 mm. (n = 4, mean ± SEM, *p < 0.05, **p < 0.01, and ***p < 0.001). See also Figure S7.