Table of Contents

Inhibitory Subpopulations in preBötzinger Complex Play Distinct Roles in Modulating Inspiratory Rhythm and Pattern

  1. Kaiwen Kam
  1. Stanson Toshok Center for Brain Function and Repair, Chicago Medical School, Rosalind Franklin University of Medicine and Science, North Chicago, Illinois 60064

  1. Author contributions: K.K. designed research; Z.C., J.S., and K.K. performed research; Z.C., J.S., and K.K. analyzed data; K.K. wrote the paper.

 
Next Section

Abstract

Inhibitory neurons embedded within mammalian neural circuits shape breathing, walking, and other rhythmic motor behaviors. At the core of the neural circuit controlling breathing is the preBötzinger Complex (preBötC), where GABAergic (GAD1/2+) and glycinergic (GlyT2+) neurons are functionally and anatomically intercalated among glutamatergic Dbx1-derived (Dbx1+) neurons that generate rhythmic inspiratory drive. The roles of these preBötC inhibitory neurons in breathing remain unclear. We first characterized the spatial distribution of molecularly defined preBötC inhibitory subpopulations in male and female neonatal double reporter mice expressing either tdTomato or EGFP in GlyT2+, GAD1+, or GAD2+ neurons. We found that the majority of preBötC inhibitory neurons expressed both GlyT2 and GAD2 while a much smaller subpopulation also expressed GAD1. To determine the functional role of these subpopulations, we used holographic photostimulation, a patterned illumination technique, in rhythmically active medullary slices from neonatal Dbx1tdTomato;GlyT2EGFP and Dbx1tdTomato;GAD1EGFP double reporter mice of either sex. Stimulation of 4 or 8 preBötC GlyT2+ neurons during endogenous rhythm prolonged the interburst interval in a phase-dependent manner and increased the latency to burst initiation when bursts were evoked by stimulation of Dbx1+ neurons. In contrast, stimulation of 4 or 8 preBötC GAD1+ neurons did not affect interburst interval or latency to burst initiation. Instead, photoactivation of GAD1+ neurons during the inspiratory burst prolonged endogenous and evoked burst duration and decreased evoked burst amplitude. We conclude that GlyT2+/GAD2+ neurons modulate breathing rhythm by delaying burst initiation while a smaller GAD1+ subpopulation shapes inspiratory patterning by altering burst duration and amplitude.

Previous SectionNext Section

Significance Statement

Inhibition is critical for control of rhythmic motor behaviors, such as breathing, walking, and chewing. We reveal functional differences among embedded preBötC inhibitory microcircuits that contribute to excitation–inhibition balance governing distinct processes in the breathing motor program. We develop approaches for high-throughput analysis of spatial distributions of neurons in tissue and for dynamic functional manipulations via an important extension of holographic photostimulation to sequences of patterns that more closely mimics physiological neural activity. Our study challenges the current understanding of the role of inhibition in breathing and provides novel insights into specific roles of inhibitory neurons in neural circuits controlling breathing that are applicable to other rhythmic motor behaviors in health and disease.

Previous SectionNext Section

Introduction

Breathing, walking, and other rhythmic motor behaviors are controlled by networks of excitatory and inhibitory neurons that coordinate and adjust muscle activity to meet physiological demand. At the core of the neural circuit controlling breathing is the preBötzinger Complex (preBötC), a compact nucleus in the ventrolateral medulla that generates breathing rhythm and organizes other rhythmic orofacial movements (Smith et al., 1991; Moore et al., 2014; Ashhad et al., 2022). In preBötC, a recurrently connected network of predominantly glutamatergic Dbx1-derived (Dbx1+) neurons generates rhythmic inspiratory drive to inspiratory premotoneurons and motoneurons and sends inspiratory-related signals throughout the brain (Koizumi et al., 2008; Bouvier et al., 2010; Gray et al., 2010; Ashhad et al., 2022). Also within preBötC are neurons that are glycinergic (expressing the glycine transporter GlyT2; GlyT2+; Winter et al., 2009; Morgado-Valle et al., 2010; Koizumi et al., 2013) and GABAergic (expressing either or both isoforms of glutamic acid decarboxylase, GAD1/GAD67 and GAD2/GAD65; GAD1/2+; Erlander et al., 1991; Kuwana et al., 2006; Koizumi et al., 2013). GlyT2+ and GAD1/2+ neurons are functionally and anatomically intercalated among Dbx1+ neurons, with many, but not all, showing inspiratory-modulated activity (Kuwana et al., 2006; Winter et al., 2009; Morgado-Valle et al., 2010; Koizumi et al., 2013; Oke et al., 2023).

Inhibition regulates many aspects of breathing (Shao and Feldman, 1997; Ezure et al., 2003; Ezure and Tanaka, 2004; Kubin et al., 2006; Smith et al., 2007; Dutschmann and Dick, 2012; Janczewski et al., 2013; Sherman et al., 2015; Baertsch et al., 2018) and plays a role in the etiology of neurodevelopmental disorders with respiratory phenotypes, such as Rett syndrome (Abdala et al., 2010; Chao et al., 2010). The proposed roles for preBötC inhibitory neurons include ensuring the alternating silence of antagonist inspiratory and expiratory muscles (Ezure et al., 2003; Janczewski et al., 2013), modulating motor nerve burst pattern (Shao and Feldman, 1997; Janczewski et al., 2013), inducing apneas for regulatory and protective reflexes and volitional behaviors (Ezure and Tanaka, 2004; Kubin et al., 2006; Janczewski et al., 2013), contributing to preBötC excitation–inhibition balance (Ramirez and Baertsch, 2018; Ashhad and Feldman, 2020), and generating rhythm (Smith et al., 2007).

There are, however, conflicting data and interpretations over the role of preBötC inhibitory neurons in breathing (Smith et al., 2007; Janczewski et al., 2013; Marchenko et al., 2016). Pharmacological blockade of inhibition (Shao and Feldman, 1997; Dutschmann and Paton, 2002; Janczewski et al., 2013; Marchenko et al., 2016) and optogenetic perturbations of inhibitory neurons (Sherman et al., 2015; Ausborn et al., 2018; Baertsch et al., 2018; Fortuna et al., 2019; Hulsmann et al., 2021; Vafadari et al., 2023) produce a variety of respiratory effects. These manipulations may have affected unspecified numbers and types of inhibitory neurons in the adjacent Bötzinger Complex (BötC) and potentially throughout the medulla, obscuring functionality specific to preBötC inhibitory neurons. Thus, a clear picture for the role of preBötC inhibitory neurons in breathing remains elusive (Abdala et al., 2015; Feldman and Kam, 2015; Dick et al., 2018; Ramirez and Baertsch, 2018; Ashhad et al., 2022).

To shed light on the molecular heterogeneity and function of inhibitory neurons and microcircuits within preBötC, we used double reporter mice in which red and green fluorescent proteins were expressed in Dbx1+, GlyT2+, GAD1+, or GAD2+ neurons. We found that the majority of preBötC inhibitory neurons express both GlyT2 and GAD2 postnatally while a much smaller subpopulation also expresses GAD1. To determine whether molecularly defined preBötC inhibitory neurons showed any functional differences, we used holographic photostimulation, which can dynamically and selectively excite small sets of neurons within a population, in rhythmic medullary slices from double reporter mice (Lutz et al., 2008; Zahid et al., 2010; Kam et al., 2013b). Holographic photostimulation of GlyT2+ neurons in the preinspiratory period delayed burst initiation, whereas photostimulation of GAD1+ neurons during inspiration increased burst duration and decreased burst amplitude, pointing to distinct roles for preBötC inhibitory subpopulations in inspiratory rhythm and pattern generation.

Previous SectionNext Section

Materials and Methods

Experimental procedures were carried out in accordance with the United States Public Health Service and Institute for Laboratory Animal Research Guide for the Care and Use of Laboratory Animals. All animals were handled according to approved institutional protocols at Rosalind Franklin University of Medicine and Science (#B14-16, #B18-10, #B21-03). All protocols were approved by the Rosalind Franklin University of Medicine and Science Institutional Animal Care and Use Committee (Animal Welfare Assurance #A3279-01). Animals were subject to regular veterinary inspection and only to the experimental procedures detailed here and maintained on a conventional 12 h light/dark cycle with ad libitum access to food and water. Every effort was made to minimize pain and discomfort.

Transgenic mice

Rosa26LoxP-Stop-LoxP-tdTomato (JAX Stock No. 7914; Madisen et al., 2010) mice were crossed with either GlyT2EGFP (Zeilhofer et al., 2005) or GAD1EGFP (JAX Stock No. 7677; Chattopadhyaya et al., 2004) BAC transgenic mice to generate Rosa26LoxP-Stop-LoxP-tdTomato;GlyT2EGFP and Rosa26LoxP-Stop-LoxP-tdTomato;GAD1EGFP mouse lines. These mice were then crossed with either Dbx1cre (Bielle et al., 2005), GlyT2cre (Foster et al., 2015), or GAD2cre (JAX Stock No. 19022; Taniguchi et al., 2011) mice to generate Dbx1tdTomato;GlyT2EGFP, Dbx1tdTomato;GAD1EGFP, GAD2tdTomato;GlyT2EGFP, GAD2tdTomato;GAD1EGFP, and GlyT2tdTomato;GAD1EGFP double reporter mice. Animal genotypes were verified via real-time polymerase chain reaction using primers specific for cre recombinase or by direct visualization of fluorescent reporter.

Histology and imaging

Male and female neonatal double reporter mice [postnatal days 0–4 (P0–P4)] were anesthetized by inhalation of 5% isoflurane and transcardially perfused with 4% formaldehyde solution (freshly depolymerized from paraformaldehyde powder in PBS). A block of tissue from the pons to the rostral cervical spinal cord was isolated, fixed overnight in 4% formaldehyde solution, and cryoprotected in 30% sucrose. The fixed brainstem–spinal cord block was sliced coronally on a Leica CM3050S cryostat in 60 μm intervals from the obex to the superior vestibular nucleus.

Sections were coverslipped in PVA-DABCO (Valnes and Brandtzaeg, 1985) and imaged on a confocal Leica SP8 microscope. Facial motor nucleus and nucleus ambiguus (NA) were identified by the absence of GlyT2+, GAD1+, and GAD2+ neurons in reporter mice and used as landmarks to identify preBötC. For each section, the preBötC was centered within a 450 × 450 μm field and a z-stack (1.5 μm optical section) obtained using a 20× objective.

Z-stacks were imported into CellProfiler 4.2.1 (Carpenter et al., 2006) one channel at a time and run through a custom-designed image processing pipeline to automatically extract the total cell count and three-dimensional coordinates of cell centroids. Briefly, image intensity was first rescaled to utilize the full dynamic range. Images were then shrunk to half their size and spatially averaged to reduce oversegmentation and increase the uniformity of signal within labeled cells. Resized images were subject to spatial averaging and morphological operations to enhance soma shapes. A sequential threshold was then applied, followed by further morphological operations to remove artifacts and fill holes. A watershed algorithm was then used to segment cells across three dimensions. Image processing parameters were determined empirically for each channel. Fluorescence intensity, shape, and position of detected cells were then measured and written to an output file. Image processing pipelines are available from the corresponding author's website (https://sites.google.com/site/kwkamlab) or the CellProfiler pipeline repository (https://cellprofiler.org/published-pipelines).

CellProfiler output and confocal image stacks were loaded into Igor Pro for manual review and editing. Custom-written code was utilized to overlay the coordinate information extracted from CellProfiler onto the raw images from each channel. A three-dimensional distance threshold was then applied to determine whether neurons expressed both markers, i.e., were coexpressing, and marked accordingly. Results from automated cell detection were checked manually by reviewing markers indicating tdTomato-, EGFP-, and coexpressing neurons overlaid on the confocal z-stack, which allowed discrimination between coexpressing neurons and single-fluorescent neurons that overlapped rostrocaudally. Automated results were edited to add missed neurons, subtract false-positive signals, separate single-expressing neurons that were detected as coexpressing, and join coexpressing neurons that were detected as singly expressing. As neither Dbx1 nor any of the inhibitory markers delineate preBötC boundaries, we opted to use a hemicylindrical volume that hews roughly to a functionally defined rostrocaudal extent 360–480 µm caudal to caudal pole of facial motor nucleus and a transverse area corresponding roughly to the lateral part of the paragigantocellular reticular nucleus or the medial part of the lateral reticular nucleus (Ruangkittisakul et al., 2014). Thus, cells within a 150 µm semicircle centered at the ventral pole of NA were counted, and counts from left and right sides were summed for each rostrocaudal section.

Slice electrophysiology

Neonatal double reporter mice (P0–P4) of either sex were anesthetized with isoflurane and killed by thoracic transection. Brainstems were removed in cold artificial CSF (ACSF) containing the following (in mM): 124 NaCl, 3 KCl, 1.5 CaCl2, 1 MgSO4, 25 NaHCO3, 0.5 NaH2PO4, and 30 glucose, which we aerated with 95% O2 and 5% CO2. Brainstems were then pinned to an agar block with the rostral side up. We cut a single 550–600-μm-thick transverse medullary slice with the preBötC at the rostral surface. To obtain slices with the preBötC at the surface, the rostral cut was made above the first set of hypoglossal (XII) nerve rootlets at the level of the dorsomedial cell column and principal lateral loop of the inferior olive, and the caudal cut captured the obex (Ruangkittisakul et al., 2014). Slices were placed rostral side up in a chamber and perfused at 2–3 ml/min with 28–30°C recording ACSF. Baseline recording ACSF consisted of the cutting ACSF solution with K+ raised to 9 mM and Ca2+ maintained at 1.5 mM for a robust burst rhythm (Kam et al., 2013a). In all experiments, slices were allowed to equilibrate for 30 min to ensure that the frequency and magnitude of XII population activity reached steady-state. Respiratory activity reflecting suprathreshold action potential firing from populations of neurons was recorded from XII motor nucleus or nerve roots or as population activity directly from the preBötC using suction electrodes (tip size, ∼50 µm). For whole-cell patch-clamp recordings, preBötC GlyT2+ or GAD1+ neurons were selected visually using fluorescence microscopy and patched under IR-DIC video microscopy. Patch electrodes with a 2–3 MΩ tip resistance were fabricated using a three-stage custom program on a Flaming-Brown micropipette puller (P-97, Sutter Instruments) and filled with solution containing the following (in mM): 135 K gluconate, 1.1 EGTA, 5 NaCl, 0.1 CaCl2, 10 HEPES, 2 MgATP, and 0.3 Na3GTP. In whole-cell current-clamp recordings, membrane potentials were brought to −55 mV using a negative bias current. For whole-cell recordings, experiments in which variations in series and input resistance exceeded the initial value by 25% were discarded. Population and single-cell recordings were amplified using a MultiClamp 700B (Molecular Devices), filtered at 2–4 kHz, integrated, and digitized at 10 kHz. Integration was performed on custom-built Paynter filter with a 20–100 ms time constant. Digitized data were analyzed off-line using custom procedures written for Igor Pro (WaveMetrics). Semiautomated event detection of respiratory-related activity recorded in XII output or preBötC population recordings was performed using custom procedures written in Igor Pro. Multiple criteria, including slope and amplitude thresholds, were used to select events automatically, which were then confirmed visually.

Holography

Holographic photostimulation was performed on a Phasor SLM system (Intelligent Imaging Innovations) mounted around an epifluorescence upright microscope (Zeiss Axioscope). We used a 405 nm diode laser (CUBE 405–100; Coherent) to uncage MNI-glutamate (0.5 mM) and depolarize targeted neurons. An iterative Fourier transform algorithm, implemented in Slidebook 6 (Intelligent Imaging Innovations), computed the phase pattern on the SLM corresponding to the desired distribution of light intensity at the focal plane of the objective. A blocker was inserted in the intermediate Fourier plane to block the unmodulated light component (zero order spot) and replicate patterns (Golan et al., 2009). Neurons expressing a single-fluorescent protein were targeted by centering 10 μm circular spots over the somata of the selected neurons. The laser intensity per spot was set at 1–3 mW, which was previously determined to elicit bursts of action potentials in the targeted neurons without the spread of glutamate to neighboring neurons (Kam et al., 2013b). Each laser stimulation consisted of five to 10 0.8 ms pulses, delivered at 200 Hz.

Experimental design and statistics

Data are expressed as mean ± standard deviation or median [interquartile range (IQR)]. Box plots show median, IQR, and 10 and 90% thresholds. In figures displaying average data, data points from individual experiments are displayed in gray. To determine sample sizes, a priori power analysis was performed using G*power 3.1.9.2 (Faul et al., 2009). In this program, the appropriate statistical test along with a desired power of 95% at a 5% significance level was selected for each type of experiment with effects sizes based on published or preliminary data. Sample sizes represent the number of biological replicates for each group. Statistical analysis was performed using Igor Pro. Data were tested for normality using the Shapiro–Wilk test. One-way ANOVA was used for parametric multiple comparisons followed by a post hoc Tukey’s test to determine statistical significance. For data that were not normally distributed, Wilcoxon signed rank tests were used for paired comparisons of two groups. Multiple comparisons were evaluated using Friedman test for repeated measures designs. Post hoc Wilcoxon signed rank tests with Bonferroni–Holm corrections were then used to determine significance for within group comparisons of repeated measures designs. Statistical significance was set at p < 0.05. Details of statistical designs can be found in the Results.

Previous SectionNext Section

Results

GAD1 defines a distinct preBötC inhibitory subpopulation

We first examined the spatial distribution of neurons expressing the transcription factor Dbx1, which defines a predominantly excitatory preBötC subpopulation necessary for breathing (Bouvier et al., 2010; Gray et al., 2010) and/or the glycinergic inhibitory marker GlyT2 (Zeilhofer et al., 2005). We used P0–P4 double reporter mice of either sex expressing the red fluorescent protein tdTomato in Dbx1+ neurons and enhanced green fluorescent protein (EGFP) in GlyT2+ neurons. Double reporter mice were generated using a cre-lox strategy crossing Dbx1cre (Bielle et al., 2005) with Rosa26LoxP-Stop-LoxP-tdTomato (Madisen et al., 2010) in combination with GlyT2EGFP BAC transgenic reporter mice (Zeilhofer et al., 2005). We counted neurons expressing tdTomato, EGFP, or both fluorescent proteins within a 150 μm radius hemicylinder ventral to nucleus ambiguus that extended rostrocaudally from 360 to 480 μm caudal to the caudal pole of facial motor nucleus, a region corresponding anatomically to preBötC (Ruangkittisakul et al., 2014). Within this region, there were similar numbers of Dbx1+ and GlyT2+ neurons (Dbx1+: 662.8 ± 189.9; GlyT2+: 636.3 ± 132.8; n = 4; Fig. 1A–C). Coexpression of Dbx1 and GlyT2 in this region was observed (175.3 ± 62.5; n = 4; Fig. 1A–C), constituting 26.4% of Dbx1+ neurons and 27.5% of GlyT2+ neurons. No tdTomato labeling was detected in preBötC from cre-negative Dbx1+/+;Rosa26LoxP-Stop-LoxP-tdTomato animals (data not shown).

Figure 1.
View larger version:
Figure 1.

Spatial distribution of preBötC inhibitory subpopulations. Characterization of the number and spatial distribution of molecularly defined preBötC neurons from (A–C) Dbx1tdTomato;GlyT2EGFP, (D–F) GAD2tdTomato;GlyT2EGFP, (G–I) GAD2tdTomato;GAD1EGFP, and (J–L) GlyT2tdTomato;GAD1EGFP mice. A, D, G, J, Maximum z-projections of EGFP, tdTomato, and merged confocal z-stacks of a 60 µm section from the preBötC of the specified genotype. The dotted semicircle represents the area where neurons are counted. Arrows indicate image orientation (D, dorsal; M, medial). NA, nucleus ambiguus. Scale bar, 50 µm. B, E, H, K, Orthogonal y–z sections through the confocal z-stack at the vertical dotted line indicated in A, D, G, and J show neurons coexpressing both fluorescent proteins (arrows) as well as singly expressing neurons (arrowheads) that may overlap rostrocaudally and are distinguishable in the z-stack. C, F, I, L, Rostrocaudal distributions of EGFP (green), tdTomato (red), and coexpressing (yellow) neurons. Counts are the sum of neurons from left and right preBötC from 60 µm sections at the specified level, defined relative to the caudal pole of facial motor nucleus (VIIc). Venn diagrams representing the average total counts of preBötC EGFP (green), tdTomato (red), and coexpressing (yellow) neurons are shown as insets at the top. Total counts are the sum of neurons from left and right preBötC across three 60 µm sections that span the 360–480 µm rostrocaudal extent of preBötC.

To determine the molecular phenotype of preBötC inhibitory neurons, we characterized the pairwise spatial distribution and levels of coexpression of inhibitory markers in the preBötC of P0–P4 transgenic double reporter mice expressing either tdTomato or EGFP in GlyT2+, GAD1+, or GAD2+ neurons. Double reporter mice were generated using a cre-lox strategy, crossing GlyT2cre (Foster et al., 2015) or GAD2cre (Taniguchi et al., 2011) with Rosa26LoxP-Stop-LoxP-tdTomato in combination with GlyT2EGFP or GAD1EGFP BAC transgenic reporter mice (Chattopadhyaya et al., 2004; Zeilhofer et al., 2005).

In GAD2tdTomato;GlyT2EGFP mice (GAD2cre × Rosa26LoxP-Stop-LoxP-tdTomato × GlyT2EGFP), GlyT2+ and GAD2+ neurons were found in approximately equal numbers with GlyT2+ neurons slightly outnumbering GAD2+ neurons (GlyT2+: 722 ± 106.7; GAD2+: 658.3 ± 107.1; n = 3; Fig. 1D–F). Coexpression of GlyT2 and GAD2 was high (GlyT2+/GAD2+: 431 ± 66; n = 3) with 59.7% of GlyT2+ neurons expressing GAD2 and 65.5% of GAD2+ neurons expressing GlyT2 (Fig. 1D–F). In contrast to the high density of GAD2+ and GlyT2+ preBötC neurons, significantly fewer GAD1+ neurons were observed in preBötC. In GAD2tdTomato;GAD1EGFP mice (GAD2cre × Rosa26LoxP-Stop-LoxP-tdTomato × GAD1EGFP), the number of GAD1+ neurons was much lower than GAD2+ neurons in preBötC (GAD2+: 792 ± 125.4; GAD1+: 41 ± 13.9; n = 3; Fig. 1G–I). Most GAD1+ neurons expressed GAD2 (GAD2+/GAD1+: 26.3 ± 10.4; n = 3) with 64.3% of GAD1+ neurons expressing GAD2 but only 3.3% of GAD2+ neurons expressing GAD1 (Fig. 1G–I). A similar distribution was observed in GlyT2tdTomato;GAD1EGFP mice (GlyT2cre × Rosa26LoxP-Stop-LoxP-tdTomato × GAD1EGFP). In these mice, the number of GAD1+ neurons was much lower than GlyT2+ neurons (GlyT2+: 662 ± 81.1; GAD1+: 60.3 ± 24; n = 3; Fig. 1J–L). Most GAD1+ neurons expressed GlyT2 (GlyT2+/GAD1+: 43.3 ± 18.5; n = 3) with 71.8% of GAD1+ neurons expressing GlyT2, but only 6.5% of GlyT2+ neurons expressing GAD1 (Fig. 1J–L). Across all genotypes, the number of preBötC GAD1+ neurons (50.7 ± 20.5; n = 6) was significantly lower than the number of preBötC GlyT2+ (669.7 ± 106.3; n = 10), GAD2+ (729.7 ± 136.5; n = 6), and Dbx1+ (662.8 ± 189.9; n = 4) neurons (one-way ANOVA; F(3,22) = 45.2; p = 1.4 × 10−9; post hoc Tukey’s test; GAD1+ vs GlyT2+: p = 5.1 × 10−9; GAD1+ vs GAD2+: p = 7.2 × 10−9; GAD1+ vs Dbx1+: p = 3.0 × 10−7). Thus, we found that inhibitory neurons were heterogeneous in their expression of these molecular markers and suggest that two major subpopulations of inhibitory neurons in the preBötC of neonatal mice are GAD2+/GlyT2+ coexpressing neurons and GAD1+ neurons that likely also express GAD2 and/or GlyT2.

Threshold preBötC GlyT2+ neuron stimulation delays burst initiation

We sought to examine whether these inhibitory preBötC subpopulations had distinct functional roles in inspiratory rhythm and pattern generation. A major issue with functional studies involving preBötC inhibitory neurons in vitro and in vivo is the difficulty in restricting perturbations to preBötC neurons due to the high density of inhibitory neurons in nearby BötC and numerous inhibitory afferents terminating in preBötC from other medullary sites. To selectively target preBötC GlyT2+ neurons and assess their functional role, we used holographic photostimulation in rhythmically active medullary slices obtained from neonatal Dbx1tdTomato;GlyT2EGFP (Dbx1cre × Rosa26LoxP-Stop-LoxP-tdTomato × GlyT2EGFP) double reporter mice (Lutz et al., 2008; Zahid et al., 2010; Kam et al., 2013b). In all cases, we only selected neurons expressing either tdTomato or EGFP. Glutamate uncaging of a 10 μm soma-centered spot over a single preBötC GlyT2+ neuron adjacent to Dbx1+ neurons resulted in a burst of action potentials (4.79 ± 2.07 action potentials; n = 5; Fig. 2A–C), similar to the responses of Dbx1+ neurons to photostimulation (Kam et al., 2013b; Sun et al., 2019). To determine whether adjacent neurons would be activated by diffusion of the uncaged glutamate liberated from the 10 μm wide holographic spot, we moved the spot 10 and 20 μm away from the soma and measured the whole-cell responses of neurons during uncaging. The number of action potentials elicited by holographic uncaging fell significantly when the spot was moved 10 μm away from the soma (1.61 ± 1.54 action potentials; n = 5) and decreased almost to zero when the spot was moved 20 μm away from the soma (0.11 ± 0.25 action potentials; repeated measures ANOVA; F(2,8) = 15.2; p = 0.002; post hoc paired t test with Bonferroni–Holm correction; 0 vs 10 μm: p = 0.03; 0 vs 20 μm: p = 0.007; 10 vs 20 μm: p = 0.08; n = 5; Fig. 2B,C).

Figure 2.
View larger version:
Figure 2.

Stimulation of preBötC GlyT2+ neurons during endogenous rhythm increases TE. A, DIC and fluorescence images of a Dbx1tdTomato;GlyT2EGFP mouse. Scale bar, 20 µm. B, Overlay of whole-cell current-clamp responses of a GlyT2+ neuron (white arrow in A) to photostimulation (green arrow) centered over the soma (left), 10 µm off the soma (middle), and 20 µm off the soma (right). Calibration, 20 mV, 200 ms. C, Group data showing the number of action potentials elicited by holographic photostimulation as a function of distance of the holographic spot from the soma. *p < 0.05, compared with soma-centered spot (0 µm), repeated measures ANOVA, post hoc paired t test with Bonferroni–Holm correction, n = 5. D, Fluorescent images of preBötC from Dbx1tdTomato;GlyT2EGFP mouse. Scale bar, 20 µm. E, Left, Fluorescent tdTomato image of preBötC. White arrows mark targeted neurons. Scale bar, 20 µm. Right, Integrated XII (∫XII) trace shows response to photostimulation of 8 Dbx1+ neurons. Calibration, 5 s. F, Left, Fluorescent image showing 4 preBötC GlyT2+ neurons targeted for photostimulation (white arrows). Scale bar, 20 µm. Right, Overlay of cycles of ∫XII activity under control conditions with no stimulation and when stimulation occurred during postinspiratory (post-I), expiratory (exp), and preinspiratory (pre-I) phases. Green arrows mark time of stimulation. Calibration, 2 s. Average control inspiratory burst shape (bottom left) and burst shape when stimulation occurred (bottom right) during the inspiratory burst (insp). Green arrow marks time of stimulation. Calibration, 0.2 s. G, Grouped amplitude (amp), burst duration (TI), and interburst interval (TE) for cycles with no stimulation (ctrl) and when stimulation of 4 GlyT2+ preBötC neurons occurred during post-I, exp, and pre-I. TI and amp are also shown when stimulation occurred during bursts (insp). *p < 0.05, Friedman test, post hoc Wilcoxon signed rank test with Bonferroni–Holm correction, n = 9. H, Left, Fluorescent image showing 8 preBötC GlyT2+ neurons targeted for photostimulation (white arrows). Scale bar, 20 µm. Right, Overlay of cycles of ∫XII activity under control conditions with no stimulation and when stimulation occurred during postinspiratory (post-I), expiratory (exp), and preinspiratory (pre-I) phases. Green arrows mark time of stimulation. Calibration, 2 s. Average control inspiratory burst shape (bottom left) and burst shape when stimulation occurred (bottom right) during the inspiratory burst (insp). Green arrow marks time of stimulation. Calibration, 0.2 s. I, Grouped amplitude (amp), burst duration (TI), and interburst interval (TE) for cycles with no stimulation (ctrl) and when stimulation of 8 GlyT2+ preBötC neurons occurred during post-I, exp, and pre-I. TI and amp are also shown when stimulation occurred during bursts (insp). *p < 0.05, Friedman test, post hoc Wilcoxon signed rank test with Bonferroni–Holm correction, n = 9.

To confirm that targeted neurons were in preBötC, we examined Dbx1/GlyT2 colocalization, which is lower in preBötC, and whether Dbx1+ neuron stimulation could evoke bursts and entrain rhythm, a property of preBötC Dbx1+ neurons (Bouvier et al., 2010; Gray et al., 2010; Kam et al., 2013b; Sun et al., 2019). Consistent with the Dbx1+ neurons being in preBötC, we observed little colocalization of Dbx1+ and GlyT2+ in our stimulation field, which was limited by widefield fluorescence to ∼50 μm below the rostral surface of the slice (Fig. 2D). We then photostimulated tdTomato-only–expressing Dbx1+ neurons and calculated the probability of stimulating a burst (Pstim) in XII motor output by dividing the number of successfully evoked bursts by the number of laser stimuli. Holographic photostimulation of 8 Dbx1+ neurons at intervals slightly shorter than the endogenous rhythm resulted in generation of inspiratory bursts in a median of 94% [IQR 76–95%] of stimuli (n = 9; Fig. 2E), consistent with these neurons playing a role in inspiratory rhythmogenesis.

In the same field, we selected 4 or 8 EGFP-only–expressing GlyT2+ neurons for low-frequency stimulation, which would allow sampling of all phases of endogenous rhythm. The interval between stimuli was greater than two times the average period to allow for unstimulated cycles in between stimulated cycles. The phase of stimulation was defined as the time of stimulation after a burst divided by the average interburst interval (TE) of unstimulated cycles (control TE). We categorized the calculated phases of stimulation into those occurring during the postinspiratory period (post-I; <20% of control TE immediately following a burst), during expiration (Exp; 20–90% of control TE after post-I), during the preinspiratory period (pre-I; 90–110% of control TE following Exp if stimulation did not occur during a burst), and during inspiration (stimulation occurring during a burst) and measured the effect on TE.

Photostimulation of 4 preBötC GlyT2+ neurons adjacent to the photostimulated Dbx1+ neurons during post-I and Exp did not affect TE, but pre-I stimulation resulted in a 9.2% increase in TE (5.67 s [IQR 4.71–6.42 s]) compared with control (5.19 s [IQR 4.23–5.57 s]) that was statistically significant (Friedman test; χ2(3) = 10.2; p = 0.02; post hoc Wilcoxon signed rank test with Bonferroni–Holm correction; ctrl vs post-I: p = 0.6; ctrl vs Exp: p = 1; ctrl vs pre-I: p = 0.004; n = 9; Fig. 2F,G). Amplitude (amp) and duration (TI) of the burst subsequent to photostimulation of 4 GlyT2+ neurons was not significantly different from control bursts (Friedman test; amp: χ2(3) = 3.3, p = 0.35; TI: χ2(3) = 0.3, p = 0.95; n = 9; Fig. 2F,G). While stimulation of 8 GlyT2+ neurons during post-I and Exp also did not affect TE, stimulation during pre-I produced a significant 43.8% increase in TE (6.44 s [IQR 4.40–9.15 s]) compared with control (4.48 s [IQR 4.04–5.71 s]; Friedman test; χ2(3) = 14.1; p = 0.003; post hoc Wilcoxon signed rank test with Bonferroni–Holm correction; ctrl vs post-I: p = 1; ctrl vs Exp: p = 0.7; ctrl vs pre-I: p = 0.004; n = 9; Fig. 2H,I). Amplitude and TI of the subsequent burst following stimulation of 8 GlyT2+ neurons at any phase were not significantly different from control (Friedman test; amp: χ2(3) = 1.3, p = 0.7; TI: χ2(3) = 0.3, p = 0.6; n = 9; Fig. 2H,I). When 4 or 8 GlyT2+ neurons were activated during an inspiratory burst, we observed no significant effects on inspiratory burst amplitude or duration (Wilcoxon signed rank test; 4 GlyT2+ amp: p = 0.1; 4 GlyT2+ TI: p = 0.2; 8 GlyT2+ amp: p = 0.2; 8 GlyT2+ TI: p = 0.07; n = 9; Fig. 2F–I).

We hypothesized that the specific effect of pre-I stimulation of 4 or 8 GlyT2+ neurons on TE was due to preinspiration being a particularly sensitive period of increasing synchronization during rhythmogenesis (Kam et al., 2013b; Ashhad and Feldman, 2020; Ashhad et al., 2022). A similar process may be initiated by photostimulation of 4–9 preBötC neurons (Kam et al., 2013b). The latency observed when triggering synchronous activity in a small group of Dbx1+ neurons may represent the same assembly mechanism occurring during pre-I (Kam et al., 2013b; Sun et al., 2019). If true, stimulation of GlyT2+ neurons following stimulation of 8 Dbx1+ neurons should produce a prolonged latency to burst generation, analogous to the increased period observed when 8 GlyT2+ neurons were photostimulated in pre-I during endogenous rhythm. We therefore took advantage of the ability of the holographic system to display multiple patterns to stimulate Dbx1+ neurons and GlyT2+ neurons sequentially and examine how inhibition affects this assembly process. We first established a baseline by holographically photostimulating 8 Dbx1+ neurons at intervals just shorter than the endogenous interval. This protocol resulted in entrainment of the rhythm (Pstim: 94.1% [IQR 89–94%]; n = 9; Fig. 3A,E). The latency between laser stimulation of 8 Dbx1+ neurons and burst initiation was 0.15 s [IQR 0.13–0.16 s] (n = 9; Fig. 3A,E).

Figure 3.
View larger version:
Figure 3.

Stimulation of preBötC GlyT2+ neurons increases the latency to burst initiation for evoked bursts. A–D, Left, Fluorescent images of preBötC from a Dbx1tdTomato;GlyT2EGFP mouse. Right, Overlaid traces from the ∫XII recording of evoked bursts triggered by (A) 8 Dbx1+ neurons, (B) 8 Dbx1+ neurons followed by 4 or 8 GlyT2+ neurons after a 50 ms delay, (C) 8 Dbx1+ neurons followed by 4 or 8 GlyT2+ neurons after a 100 ms delay, and (D) 8 Dbx1+ neurons followed by 4 or 8 GlyT2+ neurons after a >100 ms delay (during the evoked burst). White arrows indicate targeted neurons, red arrows indicate time of 8 Dbx1+ stimulation, and green arrows indicate time of 4 or 8 GlyT2+ stimulation 50, 100, and >100 ms after Dbx1+ stimulation. Dotted lines indicate average beginning and end of burst evoked by 8 Dbx1+ neurons. Calibration, 0.2 s. E, Grouped data showing amplitude (amp), burst duration (TI), probability of stimulation (Pstim), and latency to burst initiation following stimulation of Dbx1+ and GlyT2+ neurons. *p < 0.05, Friedman test, post hoc Wilcoxon signed rank test with Bonferroni–Holm correction, n = 9.

We then examined the effects of stimulating 4 or 8 GlyT2+ neurons 50 or 100 ms following stimulation of 8 Dbx1+ neurons (Fig. 3B,C). While sequential stimulation of 4 or 8 GlyT2+ neurons 50 ms after stimulation of 8 Dbx1+ neurons did not significantly alter the latency preceding the burst (Fig. 3B), increasing the interval between 8 Dbx1+ and 4 or 8 GlyT2+ neuron stimulation to 100 ms significantly increased the latency by 45.8% for 4 GlyT2+ neurons (0.21 s [IQR 0.19–0.22 s]) and by 48.6% for 8 GlyT2+ neurons (0.22 s [IQR 0.20–0.25 s]) compared with 8 Dbx1+ neuron stimulation alone (Friedman test; χ2(4) = 26.8; p = 2 × 10−5; post hoc Wilcoxon signed rank test with Bonferroni–Holm correction, 8 Dbx1+/50 ms/4 GlyT2+ vs 8 Dbx1+: p = 0.9; 8 Dbx1+/50 ms/8 GlyT2+ vs 8 Dbx1+: p = 0.3; 8 Dbx1+/100 ms/4 GlyT2+ vs 8 Dbx1+: p = 0.004; 8 Dbx1+/100 ms/8 GlyT2+ vs 8 Dbx1+: p = 0.004; n = 9; Fig. 3A–C,E). Pstim and amplitude and duration of evoked bursts were not significantly affected by 4 or 8 GlyT2+ neuron stimulation at 50 or 100 ms intervals (Friedman test; Pstim: χ2(4) =0.6, p = 1; amp: χ2(4) = 1.51, p = 0.8; TI: χ2(4) = 5.8, p = 0.2; n = 9; Fig. 3B,C,E). We increased the delay between 8 Dbx1+ neuron stimulation and either 4 or 8 GlyT2+ neuron stimulation to activate inhibitory neurons during the evoked burst (Fig. 3D). GlyT2+ neuron stimulation during an evoked burst did not significantly alter burst amplitude or duration (Wilcoxon signed rank test; 4 GlyT2+ amp: p = 0.4; 4 GlyT2+ TI: p = 0.3; 8 GlyT2+ amp: p = 0.1; 8 GlyT2+ TI: p = 0.9; n = 9; Fig. 3D,E). Thus, stimulating 4 or 8 preBötC GlyT2+ neurons specifically delays burst initiation.

Threshold preBötC GAD1+ neuron stimulation prolongs burst duration

The ability to target specific numbers of neurons allowed us to directly compare whether the smaller subpopulation of preBötC GAD1+ neurons differed functionally from preBötC GlyT2+ neurons, in which GAD1 expression was largely absent. To assess the functional role of the GAD1+ preBötC subpopulation and specifically target these neurons, we used holographic stimulation in rhythmic medullary slices from Dbx1tdTomato;GAD1EGFP (Dbx1cre × Rosa26LoxP-Stop-LoxP-tdTomato × GAD1EGFP) double reporter mice. Again, we selected only neurons that expressed either tdTomato or EGFP for photostimulation. Glutamate uncaging of a 10 μm soma-centered spot over a single preBötC GAD1+ neuron adjacent to Dbx1+ neurons resulted in a burst of action potentials (5.03 ± 0.65 action potentials; n = 3; Fig. 4A–C), similar to the responses of Dbx1+ and GlyT2+ neurons to photostimulation (Fig. 2A–C; Kam et al., 2013b; Sun et al., 2019). As with GlyT2+ neurons, the number of action potentials elicited by holographic uncaging fell significantly when the spot was moved 10 μm away from the soma of GAD1+ neurons (2.08 ± 1.18 action potentials; n = 3) and decreased almost to zero when the spot was moved 20 μm away from the soma (0.13 ± 0.23 action potentials; repeated measures ANOVA; F(2,4) = 72.13; p = 0.0007; post hoc paired t test with Bonferroni–Holm correction; 0 vs 10 μm: p = 0.01; 0 vs 20 μm: p = 0.003; 10 vs 20 μm: p = 0.08; n = 3; Fig. 4B,C).

Figure 4.
View larger version:
Figure 4.

Stimulation of preBötC GAD1+ neurons during endogenous bursts increases TI. A, DIC and fluorescence images of a Dbx1tdTomato;GAD1EGFP mouse. Scale bar, 20 µm. B, Overlay of whole-cell current-clamp responses of a GAD1+ neuron (white arrow in A) to photostimulation (green arrow) centered over the soma (left), 10 µm off the soma (middle), and 20 µm off the soma (right). Calibration, 20 mV, 200 ms. C, Group data showing the number of action potentials elicited by holographic photostimulation as a function of distance of the holographic spot from the soma. *p < 0.05, compared with soma-centered spot (0 µm), repeated measures ANOVA, post hoc paired t test with Bonferroni–Holm correction, n = 3. D, Fluorescent images of preBötC from Dbx1tdTomato;GAD1EGFP mouse. Scale bar, 20 µm. E, Left, Fluorescent tdTomato image of preBötC. White arrows mark targeted neurons. Scale bar, 20 µm. Right, Integrated XII (∫XII) trace shows response to photostimulation of 8 Dbx1+ neurons. Calibration, 5 s. F, Left, Fluorescent image showing 4 preBötC GAD1+ neurons targeted for photostimulation (white arrows). Scale bar, 20 µm. Right, Overlay of cycles of ∫XII activity under control conditions with no stimulation and when stimulation occurred during postinspiratory (post-I), expiratory (exp), and preinspiratory (pre-I) phases. Green arrows mark time of stimulation. Calibration, 2 s. Average control inspiratory burst shape (bottom left) and burst shape when stimulation occurred (bottom right) during the inspiratory burst (insp). Green arrow marks time of stimulation. Calibration, 0.2 s. G, Grouped amplitude (amp), burst duration (TI), and interburst interval (TE) for cycles with no stimulation (ctrl) and when stimulation of 4 GAD1+ preBötC neurons occurred during post-I, exp, and pre-I. TI and amp are also shown when stimulation occurred during bursts (insp). *p < 0.05, Friedman test, post hoc Wilcoxon signed rank test with Bonferroni–Holm correction, n = 7 for post-I, exp, and pre-I, n = 8 for insp. H, Left, Fluorescent image showing 8 preBötC GlyT2+ neurons targeted for photostimulation (white arrows). Scale bar, 20 µm. Right, Overlay of cycles of ∫XII activity under control conditions with no stimulation and when stimulation occurred during postinspiratory (post-I), expiratory (exp), and preinspiratory (pre-I) phases. Green arrows mark time of stimulation. Calibration, 2 s. Average control inspiratory burst shape (bottom left) and burst shape when stimulation occurred (bottom right) during the inspiratory burst (insp). Green arrow marks time of stimulation. Calibration, 0.2 s. I, Grouped amplitude (amp), burst duration (TI), and interburst interval (TE) for cycles with no stimulation (ctrl) and when stimulation of 8 GlyT2+ preBötC neurons occurred during post-I, exp, and pre-I. TI and amp are also shown when stimulation occurred during bursts (insp). *p < 0.05, Friedman test, post hoc Wilcoxon signed rank test with Bonferroni–Holm correction, n = 7 for post-I, exp, and pre-I, n = 8 for insp.

We again confirmed the location of the field in preBötC by examining colocalization of Dbx1+ and GAD1+ and determining whether photostimulating 8 tdTomato-only–expressing Dbx1+ neurons entrained inspiratory rhythm measured in XII motor output. We did not observe coexpression of Dbx1 and GAD1 in our limited target fields (Fig. 4D). Photostimulation of 8 Dbx1+ neurons produced robust entrainment of the rhythm (Pstim: 89.5% [IQR 80–97.2%]; n = 7; Fig. 4E), consistent with the field being in preBötC.

Following the protocol for preBötC GlyT2+ functional studies, we photoactivated 4 or 8 EGFP-only–expressing GAD1+ neurons during endogenous rhythm. In contrast to GlyT2+ stimulation, there were no statistically significant effects of the number of GAD1+ neurons stimulated or the phase on TE (Friedman test; 4 GAD1+: χ2(3) = 6.4, p = 0.09; 8 GAD1+: χ2(3) = 2.3, p = 0.5; n = 7; Fig. 4F–I). The amplitude and duration of the burst following stimulation of 4 or 8 GAD1+ neurons at any phase was also not significantly different (Friedman test; 4 GAD1+ amp: χ2(3) = 0.8, p = 0.9; 4 GAD1+ TI: χ2(3) = 1.5, p = 0.7; 8 GAD1+ amp: χ2(3) = 3.5, p = 0.3; 8 GAD1+ TI: χ2(3) = 3.5, p = 0.3; n = 7; Fig. 4F–I). In contrast to GlyT2+ stimulation, photostimulation of 4 or 8 GAD1+ neurons during inspiratory bursts significantly increased burst duration by 21.2% (ctrl: 0.29 s [IQR 0.20–0.35 s]; 4 GAD1+: 0.35 s [IQR 0.26–0.60 s]) and 63.5% (ctrl: 0.24 s [IQR 0.21–0.33 s]; 8 GAD1+: 0.39 s [IQR 0.29–0.46 s]), respectively, while amplitude was not significantly affected (Wilcoxon signed rank test; 4 GAD1+ vs ctrl amp: p = 0.2; 4 GAD1+ vs ctrl TI: p = 0.04; 8 GAD1+ vs ctrl amp: p = 0.7; 8 GAD1+ vs ctrl TI: p = 0.02; n = 8; Fig. 4F–I).

We next determined whether GAD1+ stimulation could affect inspiratory rhythm and pattern generation using sequential stimulation. Photostimulation of 8 Dbx1+ neurons in slices from neonatal Dbx1tdTomato;GAD1EGFP mice recapitulated the effects on Pstim and latency observed with Dbx1tdTomato;GlyT2EGFP reporter mice (Fig. 5A,E). When Dbx1+ stimulation was followed by stimulation of 4 or 8 GAD1+ neurons with a 50 or 100 ms delay, no significant differences in Pstim, latency, burst amplitude, or burst duration were observed (Friedman test; Pstim: χ2(4) = 4.3, p = 0.4; latency: χ2(4) = 6.9, p = 0.1; amp: χ2(4) = 12.7, p = 0.01; TI: χ2(4) = 2, p = 0.7; post hoc Wilcoxon signed rank test with Bonferroni–Holm correction; amp: p > 0.05 all comparisons; n = 6; Fig. 5A–C,E). To determine the effect of GAD1+ stimulation during the evoked burst, we increased the delay following 8 Dbx1+ stimulation. Stimulating 4 or 8 GAD1+ neurons during bursts evoked by stimulating 8 Dbx1+ neurons prolonged burst duration by 59 and 93%, respectively (ctrl: 0.24 s [IQR 0.21–0.32 s], n = 6; 4 GAD1+: 0.39 s [IQR 0.30–0.50 s], n = 5; 8 GAD1+: 0.47 s [IQR 0.26–0.59 s], n = 6; Wilcoxon signed rank test; ctrl vs 4 GlyT2+ TI: p = 0.03; ctrl vs 8 GlyT2+ TI: p = 0.03; Fig. 5D,E). While stimulating 4 GAD1+ neurons during evoked bursts did not affect burst amplitude, stimulating 8 GAD1+ neurons during evoked bursts caused a statistically significant 38.5% decrease in amplitude (ctrl: 6.9 a.u. [IQR 3.6–9.4 a.u.], n = 6; 8 GAD1+: 4.2 a.u. [IQR 1.7–7.9 a.u.], n = 6; Wilcoxon signed rank test; ctrl vs 4 GAD1+ amp: p = 0.4; ctrl vs 8 GAD1+ amp: p = 0.03; Fig. 5D,E). We therefore demonstrate that stimulation of 4 or 8 preBötC GAD1+ neurons during endogenous and evoked bursts increases burst duration and stimulation of 8 preBötC GAD1+ neurons during evoked bursts decreases burst amplitude, revealing distinct functional roles for preBötC GlyT2+ and GAD1+ neurons.

Figure 5.
View larger version:
Figure 5.

Stimulation of preBötC GAD1+ neurons during evoked bursts alters burst pattern. A–D, Left, Fluorescent images of preBötC from a Dbx1tdTomato;GAD1EGFP mouse. Right, Overlaid traces from the ∫XII recording of evoked bursts triggered by (A) 8 Dbx1+ neurons, (B) 8 Dbx1+ neurons followed by 4 or 8 GAD1+ neurons after a 50 ms delay, (C) 8 Dbx1+ neurons followed by 4 or 8 GAD1+ neurons after a 100 ms delay, and (D) 8 Dbx1+ neurons followed by 4 or 8 GAD1+ neurons after a >100 ms delay (during the evoked burst). White arrows indicate targeted neurons, red arrows indicate time of 8 Dbx1+ stimulation, and green arrows indicate time of 4 or 8 GAD1+ stimulation 50, 100, and >100 ms after Dbx1+ stimulation. Dotted lines indicate average beginning and end of burst evoked by 8 Dbx1+ neurons. Calibration, 0.2 s. E, Grouped data showing amplitude (amp), burst duration (TI), probability of stimulation (Pstim), and latency to burst initiation following stimulation of Dbx1+ and GAD1+ neurons. *p < 0.05, Wilcoxon signed rank test, n = 5 8 Dbx1+/>100 ms/4 GAD1+, n = 6 all other groups.

Previous SectionNext Section

Discussion

Inhibition plays several roles in modulation of breathing, but how these functions are governed by inhibitory neurons in the preBötC has not been clearly determined (Abdala et al., 2015; Feldman and Kam, 2015; Dick et al., 2018; Ramirez and Baertsch, 2018; Ashhad et al., 2022). We found that, while most preBötC inhibitory neurons in neonatal mice express GlyT2 and GAD2, GAD1 delineates a distinct inhibitory subpopulation. Holographic photostimulation of a small number of GlyT2+ neurons during the preinspiratory phase or following threshold Dbx1+ stimulation resulted in an increase in latency to burst initiation. Photostimulation of the same number of GAD1+ neurons did not delay endogenous or evoked inspiratory burst generation but did prolong burst duration when stimulation occurred during endogenous or evoked inspiratory bursts. Activation of more GAD1+ neurons during inspiration decreased evoked burst amplitude. These results point to distinct functional roles for preBötC GAD1+ and GAD1− inhibitory neurons in the breathing central pattern generator.

Expression patterns of Dbx1, GlyT2, GAD2, and GAD1 in double reporter mice confirm the molecular heterogeneity of preBötC and point to its broad organization into larger excitatory Dbx1+ and inhibitory GlyT2+/GAD2+ subpopulations with a smaller inhibitory GAD1+ subpopulation. preBötC Dbx1+ neurons are predominantly glutamatergic with a small minority expressing GlyT2 or GAD1 (Winter et al., 2009; Bouvier et al., 2010; Gray et al., 2010; Morgado-Valle et al., 2010). While we did not see significant overlap of Dbx1 and GlyT2 in acute slices, perhaps due to the limited sensitivity and optical resolution of widefield fluorescence imaging, we found more overlap (26.4%) between Dbx1+ and GlyT2+ neurons in our histological analysis than previously reported (Bouvier et al., 2010; Gray et al., 2010). In one study using a combined in situ hybridization (ISH) and immunohistochemistry (IHC) protocol that may have limited sensitivity, only ∼9% of Dbx1+ neurons in a 500 × 720 μm cylindrical column representing the ventrolateral medulla expressed GlyT2, with ∼82% expressing the glutamatergic marker VGlut2 (Gray et al., 2010). Another study using a similar combined ISH and IHC approach found that 81.7% of Dbx1+ neurons in a 300 μm3 box representing preBötC expressed VGlut2 (Bouvier et al., 2010). preBötC, like many functionally defined structures in the reticular formation, does not have clear anatomical borders, despite the position of these structures being highly stereotyped (Feldman and Kam, 2015). We opted to use a hemicylindrical volume that hews roughly to a functionally defined rostrocaudal extent 360–480 µm caudal to caudal pole of facial motor nucleus and a transverse area that follows cytoarchitectural boundaries (Ruangkittisakul et al., 2014). In our study, averaging cell counts at specific rostrocaudal levels across double cre-lox and BAC transgenic reporter mice that amplify or even overrepresent gene expression (due to cre leakage) in a smaller, hemicylindrical field may have inflated the fraction of Dbx1+/GlyT2+ coexpressing neurons. On the other hand, GlyT2, GAD1, and GAD2 mRNA in Dbx1+ preBötC neurons are found at levels that were not significantly different from non-Dbx1+ preBötC neurons (Hayes et al., 2017), suggesting that the number of preBötC Dbx1+ neurons that express these inhibitory markers, while small, may not be inconsequential.

While previous studies have shown GlyT2 and GAD1 coexpression as well as GlyT2 and GAD2 coexpression in preBötC neurons (Koizumi et al., 2013; Hirrlinger et al., 2019; Oke et al., 2023), ours is the first to suggest that GAD1 expression may define a small subpopulation of preBötC inhibitory neurons distinct from the majority of GAD2+/GlyT2+ preBötC neurons. Coexpression across these inhibitory markers was not complete, so singly expressing neurons or neurons expressing other combinations, such as GAD1+/GAD2+ or GAD1+/GlyT2+ double expressing neurons, could constitute other functionally distinct subpopulations that differ in their phasic activity or connectivity with other breathing-related neurons. Indeed, a recent study suggested that GlyT2+ only neurons were predominantly inspiratory whereas GlyT2+/GAD2+ neurons fired tonically (Oke et al., 2023). Coexpression of GlyT2 and GAD2 is developmentally regulated (Hirrlinger et al., 2019; Oke et al., 2023), so the distribution of molecularly defined inhibitory subpopulations, and perhaps their function, may change during maturation.

To determine whether preBötC GAD1+ and GAD1− inhibitory populations play functionally distinct roles in inspiratory rhythm and pattern generation in vitro, we used holographic photostimulation to restrict activation to a small number of neurons and to establish that the targeted inhibitory neurons were in preBötC by showing that activating adjacent Dbx1+ neurons entrained inspiratory rhythm. We found that photostimulating 4 or 8 GlyT2+ preBötC neurons during the preinspiratory period produced a phase delay. With evoked bursts, GlyT2+ photostimulation only produced an effect when the interval between Dbx1+ and GlyT2+ stimulation was >50 ms, suggesting a window during which threshold GlyT2+ activity can alter the emergent burst initiating process. These data are consistent with brief optogenetic activation of GlyT2+ neurons causing phase-dependent delays in inspiratory activity in vivo with phase delays being strongest when stimulation occurred in the preinspiratory period (Sherman et al., 2015). Prolonged optogenetic stimulation extends the expiratory period and can produce apnea (Sherman et al., 2015; Fortuna et al., 2019). Repeated optogenetic stimulation of vesicular GABA transporter (VGAT+) inhibitory neurons that include GlyT2+ neurons terminates inspiration or increases breathing frequency by eliciting rebound excitation (Baertsch et al., 2018; Fortuna et al., 2019; Hulsmann et al., 2021). Effects elicited by optogenetic stimulation may be due to activation of a greater number of preBötC GlyT2+ neurons; stimulation of GlyT2+ neurons outside of preBötC, including those in BötC and inhibitory afferents from more distant brain regions; and/or activation of a broader inhibitory population due to the use of the VGAT promoter.

We also found that activation of 4 or 8 preBötC GAD1+ neurons unexpectedly increased burst duration when stimulation occurred during the inspiratory burst and that stimulation of 8 GAD1+ neurons also decreased evoked burst amplitude. Previous studies pointed to a role for inhibitory neurons in decreasing burst amplitude and duration. Pharmacological blockade of GABAA receptors increased amplitude and modestly increased duration (Shao and Feldman, 1997; Janczewski et al., 2013; Baertsch et al., 2018). Optogenetic GlyT2+ and VGAT+ neuron stimulation in vivo decreased inspiratory duration or terminated inspiration (Sherman et al., 2015; Baertsch et al., 2018; Hulsmann et al., 2021). However, these studies did not specifically activate GAD1+ preBötC neurons. Our data are thus the first to show how targeted activation of GAD1+ preBötC neurons affects inspiration.

Is threshold stimulation of GlyT2+ and GAD1+ preBötC neurons physiologically relevant? Manipulations of small numbers of neurons have revealed fundamental properties of neural circuits in mammals (Doron and Brecht, 2015) and provide unique constraints for computational models of preBötC (Feldman and Kam, 2015; Guerrier et al., 2015; Song et al., 2015; Phillips and Rubin, 2022; Ashhad et al., 2023). Changing activity of a few, or even single, neurons can impact neural circuit dynamics (Isaacson and Scanziani, 2011; Karnani et al., 2016a,b). Indeed, the cell number-dependent effects on evoked burst amplitude with GAD1+ neuron stimulation suggests that regulation of distinct aspects of inspiratory rhythm and pattern generation may be achieved by tuning the number of active preBötC inhibitory neurons.

This functionality may be mediated by regulation of excitation–inhibition (E–I) balance, where even minor alterations in the number or strength of inhibitory activity can alter neural circuit dynamics (Isaacson and Scanziani, 2011; Ramirez and Baertsch, 2018; Ashhad and Feldman, 2020). The overlapping connectivity and phasic activity of inhibitory and excitatory preBötC neurons supports such a role. Excitatory somatostatin-expressing and GlyT2+ preBötC neurons overlap extensively in their efferent targets and receive common inputs (Yang and Feldman, 2018; Yang et al., 2020). GAD1+ and GlyT2+ inspiratory preBötC neurons also overlap with glutamatergic preBötC neurons in their projections to XII premotoneurons and motoneurons (Koizumi et al., 2013). GlyT2+ and VGAT+ neurons have been found to be tonically active, expiratory-modulated, and inspiratory-modulated, some with intrinsic bursting properties (Ezure et al., 2003; Winter et al., 2009; Morgado-Valle et al., 2010; Baertsch et al., 2018; Oke et al., 2018; Burgraff et al., 2022) while most GAD1+ neurons display inspiratory-modulated activity (Koizumi et al., 2013). Thus, GlyT2+ and GAD1+ preBötC neurons coterminate and are coactive with excitatory Dbx1+ preBötC neurons during preinspiration and inspiration.

These data support a model whereby GlyT2+ and GAD1+ preBötC neurons contribute to E–I balance governing distinct respiratory-related processes (Fig. 6): GlyT2+ to burst initiation and inspiratory rhythmogenesis and GAD1+ to burst duration and amplitude and inspiratory patterning. In an endogenous cycle, E–I balance within the network is heavily tipped toward inhibition immediately following an inspiratory burst (post-I), gradually shifts toward excitation during the interburst interval (Exp), veers rapidly toward excitation during preinspiration, and reaches a zenith of excitability during the inspiratory burst (Fig. 6B). In this model, the key parameters governing the effects of preBötC neuron photostimulation are the phase-dependent level and rate of change of E–I balance and the proximity of these levels to thresholds for either burst initiation (Insp-On) or burst termination (Insp-Off). Dbx1+ neuron photostimulation triggers bursts by shifting E–I balance above the burst initiation threshold (Fig. 6C). Photostimulation of preBötC GlyT2+ neurons has little effect during post-I or Exp since excitability is already low and well under the burst initiation threshold and E–I balance is only gradually changing (Fig. 6D). However, during pre-I, when E–I balance changes rapidly and approaches the Insp-On threshold, photostimulation of only a few preBötC GlyT2+ neurons, perhaps due to strong projections onto rhythmogenic Dbx1+ neurons, deflects the balance away from threshold and delays burst initiation (Fig. 6A,D). During the burst, inhibition from GlyT2+ photostimulation, perhaps coupled with weak connectivity with pattern generating output neurons, cannot offset the strong network excitability triggered by crossing the burst initiation threshold.

Figure 6.
View larger version:
Figure 6.

Model of roles of preBötC inhibitory subpopulations in inspiratory rhythm and pattern generation. A, Proposed circuit diagram showing preferential connectivity of GlyT2+ neurons with rhythmogenic Dbx1+ neurons and GAD1+ neurons with pattern-generating inspiratory premotoneurons (pMN) and motoneurons (MN). Dashed lines indicate hypothesized projections. B, The balance (white) of Excitation (Exc, orange), pushing up, and Inhibition (Inh, green), pushing down, changes in level and rate during the postinspiratory (post-I), expiratory (Exp), preinspiratory (pre-I), and inspiratory (Insp) phases of the respiratory cycle. The crossing of the inspiratory-On (Insp-On, small dash line) and inspiratory-Off (Insp-Off, long dash line) thresholds triggers burst initiation and burst termination, respectively. C, Photostimulation of Dbx1+ neurons (purple arrow) shifts the balance toward increased excitation and may exceed the Insp-On threshold, resulting in burst initiation. D, GlyT2+ photostimulation (purple arrows) has little effect during post-I and Exp since E–I levels are low, only gradually changing, and well below Insp-On threshold. During pre-I, E–I balance changes rapidly, so the shift produced by GlyT2+ photostimulation away from Insp-On threshold prolongs pre-I and delays Insp. E, GAD1+ photostimulation (purple arrow) shifts E–I balance away from Insp-Off threshold, prolonging burst duration as activity-dependent signals, such as Na+ and Ca2+, take longer to accumulate and activate burst-terminating outward conductances.

preBötC GAD1+ neurons may project preferentially to pattern-generating preBötC output neurons, inspiratory premotoneurons, or XII motoneurons to clamp activity below maximum when active, even at low levels (Fig. 6A,E). We note that during sequential stimulation, GAD1+ stimulation produced a plateau in the burst shape, suggesting that this stimulation acted to prolong burst duration, not by increasing excitation, but by holding activity and pausing normal burst-generating processes that were then released when GAD1+ stimulation ended. This pause may prolong burst duration by slowing the accumulation of Na+ or Ca2+ and delaying the crossing of the Insp-Off threshold representing activation of burst-terminating Na+- or Ca2+-dependent outward conductances (Krey et al., 2010).

Activation of more preBötC GlyT2+ or GAD1+ neurons, as observed in optogenetic studies, may shift E–I balance toward inhibition and result in stronger inhibitory effects on both rhythm and pattern even when stimulation occurs outside of the preinspiratory or inspiratory period. Thus, while inhibition is not necessary for inspiratory rhythmogenesis (Shao and Feldman, 1997; Janczewski et al., 2013; Sherman et al., 2015), changes in GlyT2+ preBötC neuronal activity during preinspiration may allow precise, rapid regulation of burst timing by advancing or delaying bursts (Sherman et al., 2015; Baertsch et al., 2018), whereas GAD1+ preBötC neuron activity during inspiration modulates E–I balance onto pattern-generating neurons to increase or decrease burst duration and amplitude.

In summary, we discover functionally distinct inhibitory preBötC subpopulations defined by the presence or absence of GAD1. We find that threshold stimulation can precisely and rapidly regulate the initiation of inspiration through activation of GlyT2+ neurons and the duration and magnitude of inspiration through activation of GAD1+ neurons. We suggest that these modest changes in GlyT2+ and GAD1+ preBötC activity alter E–I balance in rhythm and pattern generating microcircuits, allowing for precise, dynamic control of breathing. Pathological disruption of E–I balance may then be responsible for the respiratory dysfunction observed in Rett syndrome and other neurodevelopmental conditions (Dani et al., 2005; Abdala et al., 2010; Chao et al., 2010; Nelson and Valakh, 2015). We conclude that the activity of distinct subpopulations of inhibitory preBötC neurons constitute functional microcircuits that can rapidly shape breathing, expand dynamic range, and confer lability that allows breathing to respond to changes in metabolic demand and environmental conditions and to coordinate with other respiratory-related behaviors, such as swallowing and vocalization.

Previous SectionNext Section

Footnotes

    • Received October 5, 2023.
    • Revision received April 19, 2024.
    • Accepted May 6, 2024.

  • GlyT2cre and GlyT2EGFP mice were generously provided by Dr. H.U. Zeilhofer. This work was supported by National Institutes of Health grant (NS097492).

  • ↵*Z.C. and J.S. contributed equally to this work.

  • The authors declare no competing financial interests.

  • Correspondence should be addressed to Kaiwen Kam at kaiwen.kam{at}rosalindfranklin.edu.
Previous Section
 

References

  1. ↵
    1. Abdala AP,
    2. Dutschmann M,
    3. Bissonnette JM,
    4. Paton JF
    (2010) Correction of respiratory disorders in a mouse model of Rett syndrome. Proc Natl Acad Sci U S A 107:18208–18213. https://doi.org/10.1073/pnas.1012104107 pmid:20921395
    Abstract/FREE Full Text
  2. ↵
    1. Abdala AP,
    2. Paton JF,
    3. Smith JC
    (2015) Defining inhibitory neurone function in respiratory circuits: opportunities with optogenetics? J Physiol 593:3033–3046. https://doi.org/10.1113/jphysiol.2014.280610 pmid:25384785
    CrossRefMedline
  3. ↵
    1. Ashhad S,
    2. Feldman JL
    (2020) Emergent elements of inspiratory rhythmogenesis: network synchronization and synchrony propagation. Neuron 106:482–497.e4 https://doi.org/10.1016/j.neuron.2020.02.005
  4. ↵
    1. Ashhad S,
    2. Kam K,
    3. Negro CAD,
    4. Feldman JL
    (2022) Breathing rhythm and pattern and their influence on emotion. Annu Rev Neurosci 45:223–247. https://doi.org/10.1146/annurev-neuro-090121-014424 pmid:35259917
    CrossRefMedline
  5. ↵
    1. Ashhad S,
    2. Slepukhin VM,
    3. Feldman JL,
    4. Levine AJ
    (2023) Microcircuit synchronization and heavy-tailed synaptic weight distribution augment preBotzinger complex bursting dynamics. J Neurosci 43:240–260. https://doi.org/10.1523/JNEUROSCI.1195-22.2022 pmid:36400528
    Abstract/FREE Full Text
  6. ↵
    1. Ausborn J,
    2. Snyder AC,
    3. Shevtsova NA,
    4. Rybak IA,
    5. Rubin JE
    (2018) State-dependent rhythmogenesis and frequency control in a half-center locomotor CPG. J Neurophysiol 119:96–117. https://doi.org/10.1152/jn.00550.2017 pmid:28978767
    CrossRefMedline
  7. ↵
    1. Baertsch NA,
    2. Baertsch HC,
    3. Ramirez JM
    (2018) The interdependence of excitation and inhibition for the control of dynamic breathing rhythms. Nat Commun 9:843. https://doi.org/10.1038/s41467-018-03223-x pmid:29483589
    CrossRefMedline
  8. ↵
    1. Bielle F,
    2. Griveau A,
    3. Narboux-Neme N,
    4. Vigneau S,
    5. Sigrist M,
    6. Arber S,
    7. Wassef M,
    8. Pierani A
    (2005) Multiple origins of Cajal-Retzius cells at the borders of the developing pallium. Nat Neurosci 8:1002–1012. https://doi.org/10.1038/nn1511
    CrossRefMedline
  9. ↵
    1. Bouvier J,
    2. Thoby-Brisson M,
    3. Renier N,
    4. Dubreuil V,
    5. Ericson J,
    6. Champagnat J,
    7. Pierani A,
    8. Chedotal A,
    9. Fortin G
    (2010) Hindbrain interneurons and axon guidance signaling critical for breathing. Nat Neurosci 13:1066–1074. https://doi.org/10.1038/nn.2622
    CrossRefMedline
  10. ↵
    1. Burgraff NJ,
    2. Phillips RS,
    3. Severs LJ,
    4. Bush NE,
    5. Baertsch NA,
    6. Ramirez JM
    (2022) Inspiratory rhythm generation is stabilized by I(h). J Neurophysiol 128:181–196. https://doi.org/10.1152/jn.00150.2022 pmid:35675444
    Medline
  11. ↵
    1. Carpenter AE, et al.
    (2006) CellProfiler: image analysis software for identifying and quantifying cell phenotypes. Genome Biol 7:R100. https://doi.org/10.1186/gb-2006-7-10-r100 pmid:17076895
    CrossRefMedline
  12. ↵
    1. Chao HT, et al.
    (2010) Dysfunction in GABA signalling mediates autism-like stereotypies and Rett syndrome phenotypes. Nature 468:263–269. https://doi.org/10.1038/nature09582 pmid:21068835
    CrossRefMedline
  13. ↵
    1. Chattopadhyaya B,
    2. Cristo GD,
    3. Higashiyama H,
    4. Knott GW,
    5. Kuhlman SJ,
    6. Welker E,
    7. Huang ZJ
    (2004) Experience and activity-dependent maturation of perisomatic GABAergic innervation in primary visual cortex during a postnatal critical period. J Neurosci 24:9598–9611. https://doi.org/10.1523/JNEUROSCI.1851-04.2004 pmid:15509747
    Abstract/FREE Full Text
  14. ↵
    1. Dani VS,
    2. Chang Q,
    3. Maffei A,
    4. Turrigiano GG,
    5. Jaenisch R,
    6. Nelson SB
    (2005) Reduced cortical activity due to a shift in the balance between excitation and inhibition in a mouse model of Rett syndrome. Proc Natl Acad Sci U S A 102:12560–12565. https://doi.org/10.1073/pnas.0506071102 pmid:16116096
    Abstract/FREE Full Text
  15. ↵
    1. Dick TE,
    2. Dutschmann M,
    3. Feldman JL,
    4. Fong AY,
    5. Hulsmann S,
    6. Morris KM,
    7. Ramirez JM,
    8. Smith JC
    , Respiratory Neurobiology Consortium (2018) Facts and challenges in respiratory neurobiology. Respir Physiol Neurobiol 258:104–107. https://doi.org/10.1016/j.resp.2015.01.014 pmid:25644207
    CrossRefMedline
  16. ↵
    1. Doron G,
    2. Brecht M
    (2015) What single-cell stimulation has told us about neural coding. Philos Trans R Soc Lond B Biol Sci 370:20140204. https://doi.org/10.1098/rstb.2014.0204 pmid:26240419
    CrossRefMedline
  17. ↵
    1. Dutschmann M,
    2. Dick TE
    (2012) Pontine mechanisms of respiratory control. Compr Physiol 2:2443–2469. https://doi.org/10.1002/cphy.c100015 pmid:23720253
    CrossRefMedline
  18. ↵
    1. Dutschmann M,
    2. Paton JF
    (2002) Glycinergic inhibition is essential for co-ordinating cranial and spinal respiratory motor outputs in the neonatal rat. J Physiol 543:643–653. https://doi.org/10.1113/jphysiol.2001.013466 pmid:12205196
    CrossRefMedline
  19. ↵
    1. Erlander MG,
    2. Tillakaratne NJ,
    3. Feldblum S,
    4. Patel N,
    5. Tobin AJ
    (1991) Two genes encode distinct glutamate decarboxylases. Neuron 7:91–100. https://doi.org/10.1016/0896-6273(91)90077-D
    CrossRefMedline
  20. ↵
    1. Ezure K,
    2. Tanaka I
    (2004) GABA, in some cases together with glycine, is used as the inhibitory transmitter by pump cells in the Hering-Breuer reflex pathway of the rat. Neuroscience 127:409–417. https://doi.org/10.1016/j.neuroscience.2004.05.032
    CrossRefMedline
  21. ↵
    1. Ezure K,
    2. Tanaka I,
    3. Kondo M
    (2003) Glycine is used as a transmitter by decrementing expiratory neurons of the ventrolateral medulla in the rat. J Neurosci 23:8941–8948. https://doi.org/10.1523/JNEUROSCI.23-26-08941.2003 pmid:14523096
    Abstract/FREE Full Text
  22. ↵
    1. Faul F,
    2. Erdfelder E,
    3. Buchner A,
    4. Lang AG
    (2009) Statistical power analyses using G*Power 3.1: tests for correlation and regression analyses. Behav Res Methods 41:1149–1160. https://doi.org/10.3758/BRM.41.4.1149
    CrossRefMedline
  23. ↵
    1. Feldman JL,
    2. Kam K
    (2015) Facing the challenge of mammalian neural microcircuits: taking a few breaths may help. J Physiol 593:3–23. https://doi.org/10.1113/jphysiol.2014.277632 pmid:25556783
    CrossRefMedline
  24. ↵
    1. Fortuna MG,
    2. Kugler S,
    3. Hulsmann S
    (2019) Probing the function of glycinergic neurons in the mouse respiratory network using optogenetics. Respir Physiol Neurobiol 265:141–152. https://doi.org/10.1016/j.resp.2018.10.008
  25. ↵
    1. Foster E, et al.
    (2015) Targeted ablation, silencing, and activation establish glycinergic dorsal horn neurons as key components of a spinal gate for pain and itch. Neuron 85:1289–1304. https://doi.org/10.1016/j.neuron.2015.02.028 pmid:25789756
    CrossRefMedline
  26. ↵
    1. Golan L,
    2. Reutsky I,
    3. Farah N,
    4. Shoham S
    (2009) Design and characteristics of holographic neural photo-stimulation systems. J Neural Eng 6:066004. https://doi.org/10.1088/1741-2560/6/6/066004
    Medline
  27. ↵
    1. Gray PA, et al.
    (2010) Developmental origin of preBotzinger complex respiratory neurons. J Neurosci 30:14883–14895. https://doi.org/10.1523/JNEUROSCI.4031-10.2010 pmid:21048147
    Abstract/FREE Full Text
  28. ↵
    1. Guerrier C,
    2. Hayes JA,
    3. Fortin G,
    4. Holcman D
    (2015) Robust network oscillations during mammalian respiratory rhythm generation driven by synaptic dynamics. Proc Natl Acad Sci U S A 112:9728–9733. https://doi.org/10.1073/pnas.1421997112 pmid:26195782
    Abstract/FREE Full Text
  29. ↵
    1. Hayes JA,
    2. Kottick A,
    3. Picardo MCD,
    4. Halleran AD,
    5. Smith RD,
    6. Smith GD,
    7. Saha MS,
    8. Del Negro CA
    (2017) Transcriptome of neonatal preBotzinger complex neurones in Dbx1 reporter mice. Sci Rep 7:8669. https://doi.org/10.1038/s41598-017-09418-4 pmid:28819234
    CrossRefMedline
  30. ↵
    1. Hirrlinger J,
    2. Marx G,
    3. Besser S,
    4. Sicker M,
    5. Kohler S,
    6. Hirrlinger PG,
    7. Wojcik SM,
    8. Eulenburg V,
    9. Winkler U,
    10. Hulsmann S
    (2019) GABA-glycine cotransmitting neurons in the ventrolateral medulla: development and functional relevance for breathing. Front Cell Neurosci 13:517. https://doi.org/10.3389/fncel.2019.00517 pmid:31803026
    CrossRefMedline
  31. ↵
    1. Hulsmann S,
    2. Hagos L,
    3. Eulenburg V,
    4. Hirrlinger J
    (2021) Inspiratory off-switch mediated by optogenetic activation of inhibitory neurons in the preBotzinger complex in vivo. Int J Mol Sci 22. https://doi.org/10.3390/ijms22042019 pmid:33670653
    Medline
  32. ↵
    1. Isaacson JS,
    2. Scanziani M
    (2011) How inhibition shapes cortical activity. Neuron 72:231–243. https://doi.org/10.1016/j.neuron.2011.09.027 pmid:22017986
    CrossRefMedline
  33. ↵
    1. Janczewski WA,
    2. Tashima A,
    3. Hsu P,
    4. Cui Y,
    5. Feldman JL
    (2013) Role of inhibition in respiratory pattern generation. J Neurosci 33:5454–5465. https://doi.org/10.1523/JNEUROSCI.1595-12.2013 pmid:23536061
    Abstract/FREE Full Text
  34. ↵
    1. Kam K,
    2. Worrell JW,
    3. Janczewski WA,
    4. Cui Y,
    5. Feldman JL
    (2013a) Distinct inspiratory rhythm and pattern generating mechanisms in the preBotzinger complex. J Neurosci 33:9235–9245. https://doi.org/10.1523/JNEUROSCI.4143-12.2013 pmid:23719793
    Abstract/FREE Full Text
  35. ↵
    1. Kam K,
    2. Worrell JW,
    3. Ventalon C,
    4. Emiliani V,
    5. Feldman JL
    (2013b) Emergence of population bursts from simultaneous activation of small subsets of preBötzinger complex inspiratory neurons. J Neurosci 33:3332–3338. https://doi.org/10.1523/JNEUROSCI.4574-12.2013 pmid:23426661
    Abstract/FREE Full Text
  36. ↵
    1. Karnani MM,
    2. Jackson J,
    3. Ayzenshtat I,
    4. Sichani AH,
    5. Manoocheri K,
    6. Kim S,
    7. Yuste R
    (2016a) Opening holes in the blanket of inhibition: localized lateral disinhibition by VIP interneurons. J Neurosci 36:3471–3480. https://doi.org/10.1523/JNEUROSCI.3646-15.2016 pmid:27013676
    Abstract/FREE Full Text
  37. ↵
    1. Karnani MM,
    2. Jackson J,
    3. Ayzenshtat I,
    4. Tucciarone J,
    5. Manoocheri K,
    6. Snider WG,
    7. Yuste R
    (2016b) Cooperative subnetworks of molecularly similar interneurons in mouse neocortex. Neuron 90:86–100. https://doi.org/10.1016/j.neuron.2016.02.037 pmid:27021171
    CrossRefMedline
  38. ↵
    1. Koizumi H,
    2. Koshiya N,
    3. Chia JX,
    4. Cao F,
    5. Nugent J,
    6. Zhang R,
    7. Smith JC
    (2013) Structural-functional properties of identified excitatory and inhibitory interneurons within pre-Botzinger complex respiratory microcircuits. J Neurosci 33:2994–3009. https://doi.org/10.1523/JNEUROSCI.4427-12.2013 pmid:23407957
    Abstract/FREE Full Text
  39. ↵
    1. Koizumi H,
    2. Wilson CG,
    3. Wong S,
    4. Yamanishi T,
    5. Koshiya N,
    6. Smith JC
    (2008) Functional imaging, spatial reconstruction, and biophysical analysis of a respiratory motor circuit isolated in vitro. J Neurosci 28:2353–2365. https://doi.org/10.1523/JNEUROSCI.3553-07.2008 pmid:18322082
    Abstract/FREE Full Text
  40. ↵
    1. Krey RA,
    2. Goodreau AM,
    3. Arnold TB,
    4. Del Negro CA
    (2010) Outward currents contributing to inspiratory burst termination in preBotzinger complex neurons of neonatal mice studied in vitro. Front Neural Circuits 4:124. https://doi.org/10.3389/fncir.2010.00124 pmid:21151816
    CrossRefMedlineGoogle Scholar
  41. ↵
    1. Kubin L,
    2. Alheid GF,
    3. Zuperku EJ,
    4. McCrimmon DR
    (2006) Central pathways of pulmonary and lower airway vagal afferents. J Appl Physiol 101:618–627. https://doi.org/10.1152/japplphysiol.00252.2006 pmid:16645192
    CrossRefMedline
  42. ↵
    1. Kuwana S,
    2. Tsunekawa N,
    3. Yanagawa Y,
    4. Okada Y,
    5. Kuribayashi J,
    6. Obata K
    (2006) Electrophysiological and morphological characteristics of GABAergic respiratory neurons in the mouse pre-Botzinger complex. Eur J Neurosci 23:667–674. https://doi.org/10.1111/j.1460-9568.2006.04591.x
    CrossRefMedline
  43. ↵
    1. Lutz C,
    2. Otis TS,
    3. DeSars V,
    4. Charpak S,
    5. DiGregorio DA,
    6. Emiliani V
    (2008) Holographic photolysis of caged neurotransmitters. Nat Methods 5:821–827. https://doi.org/10.1038/nmeth.1241 pmid:19160517
    CrossRefMedline
  44. ↵
    1. Madisen L, et al.
    (2010) A robust and high-throughput Cre reporting and characterization system for the whole mouse brain. Nat Neurosci 13:133–140. https://doi.org/10.1038/nn.2467 pmid:20023653
    CrossRefMedline
  45. ↵
    1. Marchenko V,
    2. Koizumi H,
    3. Mosher B,
    4. Koshiya N,
    5. Tariq MF,
    6. Bezdudnaya TG,
    7. Zhang R,
    8. Molkov YI,
    9. Rybak IA,
    10. Smith JC
    (2016) Perturbations of respiratory rhythm and pattern by disrupting synaptic inhibition within pre-Botzinger and Botzinger complexes. eNeuro 3:ENEURO.0011-16.2016. pmid:27200412 https://doi.org/10.1523/ENEURO.0011-16.2016
    Abstract/FREE Full Text
  46. ↵
    1. Moore JD,
    2. Kleinfeld D,
    3. Wang F
    (2014) How the brainstem controls orofacial behaviors comprised of rhythmic actions. Trends Neurosci 37:370–380. https://doi.org/10.1016/j.tins.2014.05.001 pmid:24890196
    CrossRefMedline
  47. ↵
    1. Morgado-Valle C,
    2. Baca SM,
    3. Feldman JL
    (2010) Glycinergic pacemaker neurons in preBotzinger complex of neonatal mouse. J Neurosci 30:3634–3639. https://doi.org/10.1523/JNEUROSCI.3040-09.2010 pmid:20219997
    Abstract/FREE Full Text
  48. ↵
    1. Nelson SB,
    2. Valakh V
    (2015) Excitatory/inhibitory balance and circuit homeostasis in autism spectrum disorders. Neuron 87:684–698. https://doi.org/10.1016/j.neuron.2015.07.033 pmid:26291155
    CrossRefMedline
  49. ↵
    1. Oke Y,
    2. Miwakeich F,
    3. Oku Y,
    4. Hirrlinger J,
    5. Hulsmann S
    (2018) Cell type-dependent activation sequence during rhythmic bursting in the preBotzinger complex in respiratory rhythmic slices from mice. Front Physiol 9:1219. https://doi.org/10.3389/fphys.2018.01219 pmid:30233397
    Medline
  50. ↵
    1. Oke Y,
    2. Miwakeichi F,
    3. Oku Y,
    4. Hirrlinger J,
    5. Hulsmann S
    (2023) Cell types and synchronous-activity patterns of inspiratory neurons in the preBotzinger complex of mouse medullary slices during early postnatal development. Sci Rep 13:586. https://doi.org/10.1038/s41598-023-27893-w pmid:36631589
    Medline
  51. ↵
    1. Phillips RS,
    2. Rubin JE
    (2022) Putting the theory into ‘burstlet theory’ with a biophysical model of burstlets and bursts in the respiratory preBotzinger complex. Elife 11:e75713. https://doi.org/10.7554/eLife.75713 pmid:35380537
    CrossRefMedlineGoogle Scholar
  52. ↵
    1. Ramirez JM,
    2. Baertsch NA
    (2018) The dynamic basis of respiratory rhythm generation: one breath at a time. Annu Rev Neurosci 41:475–499. https://doi.org/10.1146/annurev-neuro-080317-061756 pmid:29709210
    CrossRefMedline
  53. ↵
    1. Ruangkittisakul A,
    2. Kottick A,
    3. Picardo MC,
    4. Ballanyi K,
    5. Del Negro CA
    (2014) Identification of the pre-Botzinger complex inspiratory center in calibrated “sandwich” slices from newborn mice with fluorescent Dbx1 interneurons. Physiol Rep 2:370–380. https://doi.org/10.14814/phy2.12111 pmid:25138790
    Medline
  54. ↵
    1. Shao XM,
    2. Feldman JL
    (1997) Respiratory rhythm generation and synaptic inhibition of expiratory neurons in pre-Botzinger complex: differential roles of glycinergic and GABAergic neural transmission. J Neurophysiol 77:1853–1860. https://doi.org/10.1152/jn.1997.77.4.1853
    CrossRefMedline
  55. ↵
    1. Sherman D,
    2. Worrell JW,
    3. Cui Y,
    4. Feldman JL
    (2015) Optogenetic perturbation of preBotzinger complex inhibitory neurons modulates respiratory pattern. Nat Neurosci 18:408–414. https://doi.org/10.1038/nn.3938 pmid:25643296
    CrossRefMedline
  56. ↵
    1. Smith JC,
    2. Abdala AP,
    3. Koizumi H,
    4. Rybak IA,
    5. Paton JF
    (2007) Spatial and functional architecture of the mammalian brain stem respiratory network: a hierarchy of three oscillatory mechanisms. J Neurophysiol 98:3370–3387. https://doi.org/10.1152/jn.00985.2007 pmid:17913982
    CrossRefMedline
  57. ↵
    1. Smith JC,
    2. Ellenberger HH,
    3. Ballanyi K,
    4. Richter DW,
    5. Feldman JL
    (1991) Pre-Botzinger complex: a brainstem region that may generate respiratory rhythm in mammals. Science 254:726–729. https://doi.org/10.1126/science.1683005 pmid:1683005
    Abstract/FREE Full Text
  58. ↵
    1. Song H,
    2. Hayes JA,
    3. Vann NC,
    4. Drew LaMar M,
    5. Del Negro CA
    (2015) Mechanisms leading to rhythm cessation in the respiratory preBotzinger complex due to piecewise cumulative neuronal deletions. eNeuro 2:ENEURO.0031-15.2015. pmid:26465010 https://doi.org/10.1523/ENEURO.0031-15.2015
    Abstract/FREE Full Text
  59. ↵
    1. Sun X,
    2. Thörn Pérez C,
    3. Halemani DN,
    4. Shao XM,
    5. Greenwood M,
    6. Heath S,
    7. Feldman JL,
    8. Kam K
    (2019) Opioids modulate an emergent rhythmogenic process to depress breathing. Elife 8:e50613. https://doi.org/10.7554/eLife.50613 pmid:31841107
    CrossRefMedlineGoogle Scholar
  60. ↵
    1. Taniguchi H, et al.
    (2011) A resource of Cre driver lines for genetic targeting of GABAergic neurons in cerebral cortex. Neuron 71:995–1013. https://doi.org/10.1016/j.neuron.2011.07.026 pmid:21943598
    CrossRefMedline
  61. ↵
    1. Vafadari B,
    2. Tacke C,
    3. Harb A,
    4. Grutzner AA,
    5. Hulsmann S
    (2023) Increase of breathing rate mediated by unilateral optogenetic inactivation of inhibitory neurons in the preBotzinger complex in vivo. Respir Physiol Neurobiol 311:104032. https://doi.org/10.1016/j.resp.2023.104032
  62. ↵
    1. Valnes K,
    2. Brandtzaeg P
    (1985) Retardation of immunofluorescence fading during microscopy. J Histochem Cytochem 33:755–761. https://doi.org/10.1177/33.8.3926864
    CrossRefMedline
  63. ↵
    1. Winter SM,
    2. Fresemann J,
    3. Schnell C,
    4. Oku Y,
    5. Hirrlinger J,
    6. Hulsmann S
    (2009) Glycinergic interneurons are functionally integrated into the inspiratory network of mouse medullary slices. Pflugers Arch 458:459–469. https://doi.org/10.1007/s00424-009-0647-1 pmid:19238427
    CrossRefMedline
  64. ↵
    1. Yang CF,
    2. Feldman JL
    (2018) Efferent projections of excitatory and inhibitory preBotzinger complex neurons. J Comp Neurol 526:1389–1402. https://doi.org/10.1002/cne.24415 pmid:29473167
    CrossRefMedline
  65. ↵
    1. Yang CF,
    2. Kim EJ,
    3. Callaway EM,
    4. Feldman JL
    (2020) Monosynaptic projections to excitatory and inhibitory preBotzinger complex neurons. Front Neuroanat 14:58. https://doi.org/10.3389/fnana.2020.00058 pmid:33013329
    CrossRefMedline
  66. ↵
    1. Zahid M,
    2. Velez-Fort M,
    3. Papagiakoumou E,
    4. Ventalon C,
    5. Angulo MC,
    6. Emiliani V
    (2010) Holographic photolysis for multiple cell stimulation in mouse hippocampal slices. PLoS One 5:e9431. https://doi.org/10.1371/journal.pone.0009431 pmid:20195547
    CrossRefMedline
  67. ↵
    1. Zeilhofer HU,
    2. Studler B,
    3. Arabadzisz D,
    4. Schweizer C,
    5. Ahmadi S,
    6. Layh B,
    7. Bosl MR,
    8. Fritschy JM
    (2005) Glycinergic neurons expressing enhanced green fluorescent protein in bacterial artificial chromosome transgenic mice. J Comp Neurol 482:123–141. https://doi.org/10.1002/cne.20349
    CrossRefMedline

Navigate This Article

  1. Published online before print 10 May 2024, doi: 10.1523/JNEUROSCI.1928-23.2024 The Journal of Neuroscience, 44(25): e1928232024
  1. AbstractFree
  2. » Full Text
  3. Full Text (PDF)

ORCID

  1. Profile for Kam, K.https://orcid.org/0000-0002-8479-0542
Responses to This Article
Be the first to respond to this article
Submit a Response
View Similar Articles
Similar articles in this journal»
Advertisement