Medullary dorsal horn neurons activated during voluntary diving express neurotransmitter receptors for glutamate and substance P

Abstract

The diving response is a cardiorespiratory reflex characterized by apnea, bradycardia and increased peripheral vascular resistance when animals submerge underwater. Afferent information initiating this reflex response projects from the nasal mucosa via trigeminal nerves to activate second-order afferent neurons within the medullary dorsal horn (MDH). The present study investigated the synapse between these trigeminal nerves and MDH neurons, specifically what neurotransmitter receptors are expressed by MDH neurons activated during diving. Male Sprague-Dawley rats were trained to dive repetitively through an underwater maze, and Fos was used as a measure of neuronal activation. Immunohistochemistry combined with epifluorescent microscopy was used to determine which ionotropic glutamate receptor subunits, or if the Substance P receptor NK-1, were co-expressed by activated MDH neurons. Results indicate repetitive diving primarily activates small oval-shaped MDH neurons, and these Fos-positive MDH neurons co-express AMPA glutamate receptor subunits GluA1 and GluA2/3, KA glutamate receptor subunit GluK5, and NMDA glutamate receptor subunit GluN1. Since these subunits are constituent parts of their respective receptor structure, these findings reveal the formation of AMPA, KA, and NMDA glutamate receptors on the cellular membrane of these activated MDH neurons. In addition, MDH neurons activated by repetitive diving co-express the Substance P receptor NK-1. This research indicates that MDH neurons activated by diving express neurotransmitter receptors for both glutamate and SP. This supports the assertion that trigeminal nerves innervating the nasal passages release glutamate and SP within their central synapses in the brainstem and can function as part of the afferent limb of a cardiorespiratory reflex.

Graphical Abstract

1. Introduction

The diving response, an autonomic reflex characterized by apnea, bradycardia and increased peripheral vascular resistance (Butler and Jones, 1997, Fitz-Clarke, 2018, Panneton, 2013), is initiated when animals submerge underwater. Afferent information involved in initiating this response projects from the nasal mucosa via trigeminal nerves to the spinal trigeminal nucleus caudalis (also known as the Medullary Dorsal Horn (MDH) due to morphologic similarities with the spinal dorsal horn (Capra and Dessem, 1992, Gobel et al., 1988). During voluntary diving in rats, second-order neurons located within the ventral MDH are activated, as shown by their expression of Fos (Mcculloch, 2005, Mcculloch et al., 2016, Panneton et al., 2010, Panneton et al., 2012). It is thought that these activated trigeminal neurons are part of the afferent limb of the diving response brainstem neuronal circuit pathway (Mcculloch et al., 2018, Panneton et al., 2006). It is likely the terminal projections of trigeminal nasal nerves release glutamate as a neurotransmitter, as infusion of the MDH with kynurenate (a general excitatory amino acid antagonist) eliminates the cardiovascular response to nasal stimulation (Panneton and Yavari, 1995). Additionally, during nasal stimulation with ammonia vapors (which can produce cardiorespiratory responses similar to the diving response in muskrats (Mcculloch and Panneton, 1997, Panneton, 1990) and rats (Hollandsworth et al., 2009, Panneton et al., 2008, Rybka and Mcculloch, 2006), activated neurons within the MDH co-express ionotropic glutamate (AMPA and NMDA, but not kainate) receptors (McCulloch et al., 2013). Substance P may also be released by trigeminal nasal nerves, as activated neurons within the MDH also co-express NK-1 receptors after nasal stimulation (McCulloch et al., 2013).

Glutamate is the primary excitatory neurotransmitter in the mammalian CNS and is released by presynaptic neurons into chemical synapses (Alexander et al., 2019, Dingledine et al., 1999, Meldrum, 2000, Watkins and Jane, 2006). Postsynaptic glutamate receptors are therefore essential for mediating most excitatory neurotransmission (Ozawa et al., 1998, Parsons et al., 2005). Glutamate receptors are divided into two distinct groups: metabotropic and ionotropic receptors (Meldrum, 2000, Raz and Perouansky, 2019, Watkins and Jane, 2006). Metabotropic receptors are coupled to GTP-binding proteins (G proteins) and regulate the production of intracellular messengers (Ozawa et al., 1998, Raz and Perouansky, 2019). Ionotropic glutamate receptors are directly coupled to ion channels, and are classified as alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionate (AMPA), kainate (KA), or N-methyl-D-aspartate (NMDA) based on their preferred agonists (Alexander et al., 2019, Bettler et al., 2021, Collingridge et al., 2009, Dingledine et al., 1999, Parsons et al., 2005, Watkins and Jane, 2006).

Ionotropic glutamate receptors are integral membrane proteins composed of four large subunits that form a central ion channel pore (Alexander et al., 2019, Collingridge et al., 2009, Traynelis et al., 2010). All ionotropic glutamate receptors can form heteromeric subunit assemblies that have different physiological and pharmacological properties (Alexander et al., 2019, Parsons et al., 2005). Receptor heterogeneity within each receptor family arises from the assembly of distinct subunits into cation-selective tetramers (Bettler et al., 2021). A subunit will only assemble with other subunits within its own AMPA, KA or NMDA family (Dingledine et al., 1999, Meldrum, 2000). Subunit stoichiometry is not fixed, and multiple receptor subtypes exist within each receptor family based on subunit composition. AMPA receptors assemble as homomers, or heteromers, that may be drawn from GluA1, GluA2, GluA3 and GluA4 subunits; KA receptors can be expressed as combinations of GluK1, GluK2 or GluK3 (collectively GluK1/2/3), GluK4 and GluK5 subunits; and NMDA receptors consist of GluN1, GluN2A, GluN2B, GluN2C, GluN2D, GluN3A and GluN3B subunits (Alexander et al., 2019, Bettler et al., 2021, Collingridge et al., 2009). GluN1 constitutes the functional unit, whereas GluN2 and GluN3 constitute modulating subunits in NMDA channel activities (Alexander et al., 2019, Bettler et al., 2021, Dingledine et al., 1999, Traynelis et al., 2010).

Neuropeptides and kinins are important messengers in the nervous system (Hill and Oliver, 2007). The neurokinin Substance P (SP), an 11-amino acid neuropeptide, is a highly conserved member of the tachykinin peptide family that is widely expressed throughout the animal kingdom (Garcia-Recio and Gascón, 2015, Rodríguez and Coveñas, 2022, Zieglgänsberger, 2019), and was the first member of the tachykinin family to be identified (Hökfelt et al., 2001, Mantyh, 2002, Rodríguez and Coveñas, 2022). The observation that SP is expressed in the dorsal root ganglia (DRG) and dorsal horn but not in the ventral horn was taken as evidence supporting its involvement in primary afferent sensory neurotransmission (Zieglgänsberger, 2019). SP preferentially activates the neurokinin-1 receptor (NK-1; (Garcia-Recio and Gascón, 2015; Zieglgänsberger, 2019; Ilias, 2020; Rodríguez and Coveñas, 2022), and elicits excitatory effects as a neurotransmitter or neuromodulator in both the central and peripheral nervous system (Datar et al., 2004, Garcia-Recio and Gascón, 2015, Hill and Oliver, 2007). NK-1 receptors are upregulated in neurons of the spinal cord superficial laminae in rat models of pain and stress (Bradesi et al., 2009), but ligand-induced internalization of NK-1 receptors into spinal dorsal horn neurons can also be triggered by non-noxious somatosensory stimulation (Honore et al., 1999).

In a previous study we investigated MDH neurons and the neurotransmitter receptors they express after nasal stimulation with ammonia vapors in anesthetized animals (McCulloch et al., 2013). Although volitional diving in conscious animals and nasal stimulation with ammonia vapors in anesthetized animals can produce a similar efferent response (see (Mcculloch, 2012, Panneton, 2013) for reviews), in the present study we wanted to investigate trigeminal neuronal circuitry in a behaving animal without the possible effects of general anesthesia. Consequently we asked: what neurotransmitter receptors are expressed by MDH neurons activated in conscious Sprague-Dawley rats diving volitionally through an underwater maze? Our first hypothesis was that MDH neurons activated by diving would co-express ionotropic glutamate receptor subunits from different classes of glutamate receptors, specifically AMPA and NMDA receptors. The presence of these receptor subunits would therefore indicate the presence of assembled AMPA and NMDA receptors. This would strengthen the assertion that trigeminal nerves innervating the nasal mucosa release glutamate as a neurotransmitter to activate second-order neurons within the MDH during underwater diving. Our second hypothesis was that MDH neurons activated by diving would co-express NK-1 receptors. This would suggest that trigeminal nerves innervating the nasal mucosa also release SP as a neurotransmitter during underwater diving.

2. Methods

All studies were approved by the Midwestern University IACUC (MWU #1719). Male Sprague-Dawley rats were purchased from a commercial vendor (Harlan, Indianapolis IN) at 3 weeks of age (approximately 25 g each). Rats were housed in pairs within the Midwestern University Animal Facility in accordance with the NIH guidelines for the care and use of laboratory animals. Rats were provided with food and water ad libitum and were provided with cage enrichment. Cohorts of rats (8–12 rats per cohort) progressed together through swim and then dive training (see below) until they were adults (300–350 g). Several previous studies have indicated that non-diving rats produce little to no Fos-positive neurons within the ventral MDH (Mcculloch, 2005, Mcculloch et al., 2016, Panneton et al., 2010, Panneton et al., 2012). Accordingly, the decision was made to exclude swimming rats and non-diving cage control rats in the experimental design of the present study since we were interested in investigating the neurotransmitter receptors expressed in MDH neurons that are  activated by repetitive diving.

2.1. Animal training and experimental protocol

Each cohort of rats was trained to first swim and then to dive under water through a 5 m long water maze constructed from Plexiglass. For details of the training procedure see McCulloch (2014). Briefly, cohorts were trained 5–6 times per week for 10 weeks, with 3–5 trials per day and with 5 min separating each trial. The water in the maze was changed for each training session, and was maintained at 30 ± 2 °C. The rats were dried with a towel between each trial. Initially the rats were trained to swim to the finishing area of the Plexiglas maze. The length of the swimming distance was increased during each training session until the rats could swim the entire 5 m of the maze. Once this was completed, a vertical barrier that blocked the surface of the swimming channel was inserted into the maze, causing the rats to dip under the barrier to engage in their first dive. Gradually horizontal subsurface barriers were inserted into the maze channels which extended the distance the rats had to dive underwater before resurfacing and swimming the remaining distance to the finishing area. Once the dive training was completed, the rats could dive underwater through the entire 5 m maze in 10–15 s.

After all rats in each cohort had mastered the diving protocol, rats engaged in repetitive diving. In this protocol each rat dived though the maze every 5 min for 2 hrs, for a total of 24 dives (Mcculloch, 2005, Mcculloch et al., 2016). Then the rats were placed back in their cage to rest for 1 hr to allow neuronal production of Fos. The rats were then deeply anesthetized with a concentrated pentobarbital solution (Euthasol, Virbac) and then transcardially perfused with saline and then 4 % paraformaldehyde. The brain and rostral spinal cord were removed and placed overnight in a 20 % sucrose-paraformaldehyde post-fix solution. Next the brainstem was blocked and then cut from the rostral spinal cord to the inferior colliculus using a freezing microtome that produced 50 μm coronal slices. Brainstem tissue slices were collected in a 24-well plate and stored in a cryopreservative solution at −20°C for later immunohistochemistry.

2.2. Immunohistochemistry

A 1 in 3 series of the stored brain tissue was then used for immunohistochemistry to identify neurons that expressed Fos, and thus were activated during repetitive diving, and specific neurotransmitter receptors or neurotransmitter receptor subunits. Serial antibody dilutions, including negative controls, were conducted for all antibody combinations to ensure optimal labelling and to minimize background labelling. PBS (or PBS with 2 % Triton) was used for all immunohistochemical washes between incubations. Tissue was first blocked in 10 % normal donkey serum for 1 hr. The tissue was then incubated overnight with a cocktail of an anti-Fos primary antibody and a second primary antibody directed against a specific neurotransmitter receptor or neurotransmitter receptor subunit (see Table 1). The next day the tissue was blocked for 1 hr in 10 % bovine serum albumin (BSA). The tissue was then incubated for 2 hrs with a fluorescent-tagged secondary antibody directed against the Fos primary antibody (green, AF 488). The tissue was again blocked for 1 hr in 10 % BSA, and then incubated for 2 hrs with a fluorescent-tagged secondary antibody directed against the primary neurotransmitter receptor antibody (red, AF 594). The tissue free-floating in PBS was rearranged into serial order, mounted on slides, and then coverslipped using hard set mounting media (Vector H-1700).

Table 1.

Primary and secondary antibody combinations used to visualize Fos-positive neurons and neurotransmitter receptor or receptor subunits. For catalog product numbers: C = Chemicon; I = Invitrogen; M = Millipore; NB = Novus Biologicals; P = Proteintech; SC = Santa Cruz; SYSY = Synaptic Systems. For fluorescent tagged secondary antibodies: AF488 = AlexaFluor (green) and AF594 = AlexaFluor (red). N = number of animals used to visualize that antibody combination.

AMPA receptor (Glutamate)
GluA1 and Fos (N = 6)
	Fos	GluA1 receptor subunit
Primary Ab:	goat (SC, sc-52G); 1:1000	rabbit (M, AB1504); 1:10,000
Secondary Ab:	donkey AF488 (I, A11055); 1:500	donkey AF594 (I, A21207); 1:500
GluA2/3 and Fos (N = 3)
	Fos	GluA2/3 receptor subunits
Primary Ab:	rat (SYSY, 226017); 1:500	rabbit (C, AB1506); 1:500
Secondary Ab:	donkey AF488 (I, A21208); 1:5000	donkey AF594 (I, A21207); 1:500
GluA4 and Fos (N = 3)
	Fos	GluA4 receptor subunit
Primary Ab:	goat (SC, sc-52G); 1:1000	rabbit (M, AB1508); 1:2000
Secondary Ab:	donkey AF488 (I, A11055); 1:500	donkey AF594 (I, A21207); 1:500
Kainate receptor (Glutamate)
GluK1/2/3 and Fos (N = 4)
	Fos	GluK1/2/3 receptor subunits
Primary Ab:	goat (SC, sc-52); 1:1000	mouse (C, MAB379); 1:5000
Secondary Ab:	donkey AF488 (I, A11055); 1:500	donkey AF594 (I, A21203); 1:1000
GluK4 and Fos (N = 3)
	Fos	GluK4 receptor subunit
Primary Ab:	rat (SYSY, 226017; 1:1000	rabbit (NB, NBP176852); 1:400
Secondary Ab:	donkey AF488 (I, A21208); 1:1000	donkey AF594 (I, A21207); 1:1000
GluK5 and Fos (N = 3)
	Fos	GluK5 receptor subunit
Primary Ab:	rat (SYSY, 226017); 1:1000	rabbit (NB, BPB176853); 1:2000
Secondary Ab:	donkey AF488 (I, A21208); 1:1000	donkey AF594 (I, A21207); 1:2000
NMDA receptor (Glutamate)
GluN1 and Fos (N = 5)
	Fos	GluN1 receptor subunit
Primary Ab:	goat (SC, sc-52); 1:1000	rabbit (I, PA534599); 1:2000
Secondary Ab:	donkey AF488 (I, A11055); 1:1000	donkey AF594 (I, A21207); 1:2000
GluN2A and Fos (N = 3)
	Fos	GluN2A receptor subunit
Primary Ab:	rat (SYSY, 226017); 1:1000	rabbit (P, 19953-I-AP); 1:100
Secondary Ab:	donkey AF488 (I, A21208); 1:500	donkey AF594 (I, A21207); 1:1000
GluN2B and Fos (N = 3)
	Fos	GluN2B receptor subunit
Primary Ab:	rat (SYSY 226017); 1:1000	rabbit (P, 21920-I-AP); 1:200
Secondary Ab:	donkey AF488 (I, A21208); 1:1000	donkey AF594 (I, A21207); 1:500
Neurokinin-1 receptor (Substance P)
NK-1 and Fos (N = 3)
	Fos	NK1 receptor
Primary Ab:	rat (SYSY, 226017); 1:1000	rabbit (NB, NB300–101); 1:2000
Secondary Ab:	donkey AF488 (I, A21208); 1:1000	donkey AF594 (I, A21207); 1:1000

2.3. Microscopy

The tissue was visualized using an epifluorescent microscope (Nikon E600 or Nikon Ti2-E). The area of interest was the MDH, which extends approximately 1400 µm from the pyramidal decussation caudally to the obex rostrally (Fig. 1A). In each rat, a 1:3 immunohistochemical series yielded approximately 9 sections, from caudal to rostral, that contained the MDH. Additionally, the afferent signal from the nasal passages projects bilaterally to both the left and right MDH, thus creating two MDH hemisections per section. Using both left and right MDH, an average of 18.5 ± 1.4 MDH hemisections were inspected per rat. Tissue from between 3 and 6 rats was used to visualize each of the Fos and receptor or receptor subunit pairings (see Table 1). Tissue from some rats were used to visualize more than one Fos and receptor pairing.

Fig. 1.

Anatomical orientation of the trigeminal medullary dorsal horn (MDH). A) After repetitive diving, Fos-positive cells were counted bilaterally in the ventral portion of the MDH, between the pyramidal decussation caudally and obex rostrally. Superficial is MDH laminae I&II (blue); Deep is MDH laminae III-V (green); Paratrigeminal neurons are located within spinal trigeminal tract (red). Cross-sections from Paxinos and Watson (Paxinos and Watson, 1998). B) Epifluorescent low magnification (4x) image of a labeled hemisection. Scale bar = 500 µm. C) Enlarged epifluorescent image (10x) of boxed area in B were used to delineate MDH regions. Scale bar = 100 µm. D) Higher magnification (40x) epifluorescent images were used for neuronal identification within thick (50 µm) tissue sections. Photomicrograph shows a representative example of immunohistochemical labeling within the MDH. Thin arrows show Fos-positive only neurons; medium arrows show receptor subunit positive neurons that are not Fos-positive; and large arrowhead shows a colocalized neuron that is both Fos-positive and receptor subunit positive. Note the visible dark nucleolus within the Fos-positive nucleus (upper Fos-positive only neuron). Note the halo of receptor subunit labeling either with (arrowhead) or without (medium arrows) a labeled Fos-positive nucleus. Note also that the Fos-positive and receptor-positive neuron is small and oval-shaped containing little cytoplasm. Locations of MDH neurons were then plotted on low magnification (10x) photomicrographs (in panel C). Superimposed in panel C are locations of single labelled Fos-positive only neurons (crosses) and colocalized Fos-positive and neurotransmitter receptor positive neurons (filled circles) in each of the 3 MDH regions. Scale bar = 10 µm. Abbreviations: MDH, medullary dorsal horn; Py, pyramidal tract; Pyx, pyramidal decussation; sp5, spinal trigeminal tract, Sp5I, spinal trigeminal nucleus interpolaris.

Fos immunoreactivity is located in cell nuclei whereas immunoreactivity for receptor subunits is expressed predominantly in the cell membrane of neuronal somata and its proximal dendrites (Tse et al., 2008). Accordingly, tissue was first inspected for the presence of Fos in the nucleus of the neuron (indicating activation of that neuron) and then for the neurotransmitter receptor or receptor subunit in the cellular membrane (indicting expression of that neurotransmitter receptor in that Fos-positive neuron). Receptor-only neurons were not counted. The number of single labelled (Fos-positive only) and colocalized (Fos-positive and neurotransmitter receptor positive) neurons were counted bilaterally within the ventral superficial MDH (laminae I and II) and deep MDH (laminae III-V; Fig. 1A). Neurons within the paratrigeminal nucleus, located adjacent to the MDH within the spinal trigeminal tract, were also counted. Fos-positive neurons were not counted in other brainstem regions.

To aid in the counting and to reduce fluorescent fading of the chromophores, photomicrographs were taken through red and green filters with a digital camera (Nikon DS-Qi2) and associated imaging software (Nikon Elements). Low magnification (4x and 10x; Fig. 1B,C) images were used to delineate the anatomical boundaries of the MDH. Subsequent microscopic tissue analysis of the MDH at higher magnifications (20x or 40x) enabled identification of Fos-positive neurons at different focal planes within the thick (50 µm) tissue sections (Fig. 1D). Cross-inspection of the higher magnification red and green microscopic images was used to verify the presence of single labelled or colocalized neurons. The location of these MDH neurons was then identified and plotted on the low magnification (10x) photomicrographs (Fig. 1C). Neuronal counts within each of the superficial MDH, deep MDH and paratrigeminal regions were quantified using these low magnification plots.

Morphologic measurements from 55 randomly selected Fos-positive MDH neurons colocalized with neurotransmitter receptor subunits were collected using magnified 40X) images and Nikon Elements. Collected morphological information included cell shape (oval, pear-shaped, or fusiform), short and long axes, and cross-sectional area. Neurons were classified as small, medium or large according to the criteria of Ide and Killackey (1985) based on length of the long diameter (small neurons, less than 15 µm; medium neurons, 15–22 µm; long neurons, greater than 22 µm).

2.4. Data presentation

Photomicrographs were adjusted with Nikon Elements, and figures were composed and labeled using CorelDraw (Corel). Neuronal data are presented as the number of Fos-positive neurons ± SE per hemisection, and as the percentage of Fos-positive neurons that co-expressed the specific neurotransmitter receptor or receptor subunit. Statistical differences were tested with oneway ANOVAs (SigmaStat 14.5, SPSS) with p < 0.05 set as the level of significance.

3. Results

Rats engaged in repetitive diving took an average of 13.4 ± 0.4 s to dive through the underwater maze.

3.1. Fos-positive neurons

The majority of MDH neurons that were both Fos-positive and receptor subunit positive after repetitive diving were small and oval-shaped (see Figs. 1D and 2). Neurons were oval (81.1 %), pear (16.4 %), or fusiform shaped (1.8 %); pyramidal neurons were not encountered. Of the Fos-positive neurons, 60.0 % were small (long axis 12.25 ± 0.33 μm; short axis 8.50 ± 0.23 μm; cross-sectional area 82.79 ± 3.53 μm²), 36.4 % were medium (long axis 17.11 ± 0.31 μm; short axis 9.25 ± 0.37 μm; cross-sectional area 120.14 ± 5.40 μm²), and 3.6 % were large (long axis 25.74 ± 3.66 μm; short axis 11.81 ± 1.03 μm; cross-sectional area 226.52 ± 15.49 μm²). Small neurons were oval (100 %); medium neurons were oval (55 %), pear-shaped (40 %), or fusiform (5 %); large neurons were either oval (50 %) or pear-shaped (50 %). The morphologic characteristics of the Fos-positive neurons were similar regardless of the type of receptor subunit they co-expressed.

Fig. 2.

High magnification epifluorescent photomicrographs showing immunohistochemical labeling examples of Fos-positive single and colocalized neurons. Figure shows labelling of: (a) AMPA glutamate receptor subunit GluA1; (b) AMPA glutamate receptor subunit GluA2/3; (c) KA glutamate receptor subunit GluK5, (d) NMDA glutamate receptor subunit GluN1, and (e) NK-1 SP receptor. Left side of figure shows single labelled (Fos-positive only) neurons: left columns show Fos-positive neuron (green); middle columns show neurotransmitter receptor subunit or receptor (red); and right columns show composite labelling. Note that these Fos-positive neurons do not have distinct receptor subunit labelling. Right side of figure shows colocalized Fos-positive and neurotransmitter receptor subunit positive neurons: left columns show Fos-positive neuron (green); middle columns show neurotransmitter subunit or receptor (red), and right column shows composite labelling. Note that these neurons have distinct halos of receptor subunit labelling surrounding Fos-positive nuclei. Scale bars = 10 µm.

The grand mean of total Fos-positive neurons (single labelled Fos-positive neurons plus colocalized Fos-positive neurons) per hemisection was 2.98 ± 0.51 total Fos-positive neurons in the paratrigeminal nucleus, 10.76 ± 1.23 in the superficial MDH, and 15.16 ± 2.80 in the deep MDH (Table 2). The grand mean of total Fos-positive neurons per hemisection was significantly different between each of these three MDH locations (p < 0.001). By MDH region, 43.8 ± 3.4 % of the activated Fos-positive neurons were located in the deep MDH, 42.7 ± 2.7 % were located in the superficial MDH, and 13.5 ± 1.6 % were located in the paratrigeminal nucleus. The percent of Fos-positive neurons in the paratrigeminal nucleus was significantly less than in the other two locations (p < 0.001). The total number of Fos-positive neurons per hemisection (i.e. Fos only plus colocalized Fos neurons) were similar within each of the three MDH regions (i.e. paratrigeminal nucleus, superficial MDH, and deep MDH) across the 10 separate receptor subunits that were investigated (Table 2).

Table 2.

Counts of Fos-positive neurons per hemisection from rats trained to dive underwater. Counts per hemisection are given for neurons that that were only Fos-positive (Fos only) or were both Fos-positive and neurotransmitter receptor subunit positive (Colocalized). The total number of Fos-positive neurons (sum of Fos only and Colocalized; Total Fos) is also given. Counts are given for the three MDH regions: paratrigeminal nucleus (Para), superficial MDH (laminae I & II; Super), and deep MDH (laminae III-V; Deep). Fos-positive neuronal counts per hemisection are given for immunohistochemical pairings with 10 different receptor subunits (See Table 1). The grand mean (± Standard Error; SE) for Fos only, Colocalized, and Total Fos are given. Within each of the three MDH regions, the total Fos-positive neuronal counts per hemisection were not dissimilar between the 10 different receptor subunits that were investigated. a = significantly greater than GluN2A within Para (p < 0.05); b = significantly greater than GluA1 and GluK1/2/3 (p < 0.05) within Deep. 1 = significantly different from the grand mean of the other two MDH regions (p < 0.001).

	Fos only			Colocalized			Total Fos
	Para	Super	Deep	Para	Super	Deep	Para	Super	Deep
AMPA (GluA1)	2.25	2.17	2.46	0.79	4.77	0.66	3.04	6.94	3.12
AMPA (GluA2/3)	0.92	6.11	7.77	2.43	7.55	13.67	3.36	13.66	21.44
AMPA (GluA4)	1.24	5.61	7.94	0.47	1.85	2.77	1.71	7.46	10.71
Kainate (GluK1/2/3)	1.14	9.88	4.55	0.16	0.49	0.06	1.30	10.38	4.61
Kainate (GluK4)	3.89	9.92	9.77	0.00	0.00	0.00	3.89	9.92	9.77
Kainate (GluK5)	0.90	2.99	11.69	3.61	9.69	4.94	4.50	12.68	16.63
NMDA (GluN1)	1.50	4.19	3.62	4.27	15.61	18.47	5.77a	19.79	22.10b
NMDA (GluN2A)	0.34	6.91	16.64	0.04	1.74	8.33	0.38	8.65	24.97b
NMDA (GluN2B)	1.45	7.88	18.97	0.74	2.98	9.71	2.19	7.28	9.61
NK-1	1.80	2.85	4.18	1.89	4.43	5.43	3.69	7.28	9.61
Grand Mean	1.54	5.85	8.76	1.44	4.91	6.40	2.981	10.761	15.161
± SE	0.31	0.89	1.77	0.49	1.53	1.95	0.51	1.23	2.80

3.2. Colocalization of receptor subunits in Fos-positive neurons

The number of colocalized Fos-positive neurons (and therefore the percent of colocalized neurons) depended upon which neurotransmitter receptor subunit was immunolabeled (Table 2; Fig. 2, Fig. 3). Within the paratrigeminal nucleus, Fos-positive neurons predominantly co-expressed AMPA glutamate receptor subunit GluA2/3 (74.6 ± 4.3 %), KA glutamate receptor subunit GluK5 (82.5 ± 5.1 %), NMDA glutamate receptor subunit GluN1 (70.9 ± 5.0 %), and SP NK-1 receptors (51.4 ± 8.4 %). Other AMPA, KA, and NMDA receptors subunits were expressed to lesser degrees by paratrigeminal Fos-positive neurons. Within the superficial MDH, Fos-positive neurons predominantly co-expressed AMPA glutamate receptor subunits GluA1 (70.4 ± 6.7 %) and GluA2/3 (54.2 ± 6.8 %), KA glutamate receptor subunit GluK5 (77.3 ± 2.1 %), NMDA glutamate receptor subunit GluN1 (75.8 ± 5.6 %), and SP NK-1 receptors (57.5 ± 3.7 %). Other AMPA, KA, and NMDA receptors subunits were expressed to lesser degrees by superficial MDH Fos-positive neurons. Within the deep MDH, Fos-positive neurons predominantly co-expressed AMPA glutamate receptor subunit GluA2/3 (63.9 ± 8.1 %), NMDA glutamate receptor subunit GluN1 (79.7 ± 3.9 %), and SP NK-1 receptors (60.2 ± 5.9 %). Other AMPA, KA, and NMDA receptors subunits were expressed to lesser degrees by deep MDH Fos-positive neurons. GluK4 receptors subunits were not co-expressed by any Fos-positive neurons in any of the trigeminal regions inspected, although it was noted that there were neurons immunolabeled with the GluK4 receptor subunit in other areas of the brainstem. GluK1/2/3 receptors subunits were only minimally expressed by Fos-positive neurons in the three trigeminal regions inspected.

Fig. 3.

Percent colocalization of Fos-positive and neurotransmitter receptor subunit neurons. Percentages for each neurotransmitter were determined by dividing the number of colocalized Fos-positive neurons by the total number of Fos-positive neurons present within each location (See Table 2). (A) Paratrigeminal nucleus; (B) Superficial MDH; (C) Deep MDH. AMPA glutamate receptor subunits include GluR1, GluR2/3, and GluR4. Kainate glutamate receptor subunits include GluK1/2/3, GluK4, and GluK5. There were no colocalized neurons for GluK4. NMDA glutamate receptor subunits include GluN1, GluN2A, and GluN2B (N = 3). NK-1 receptor binds Substance P. Mean data points from individual rats are also included for each plot.

4. Discussion

In mammals, underwater submersion initiates the cardiorespiratory reflex known as the diving response. Part of the diving response central neuronal circuitry involves activation of neurons with the spinal trigeminal nucleus, specifically the MDH. The activation of neurons within the trigeminal nucleus would constitute the afferent limb of the circuit that receives and processes the initial sensory signal from the nose and nasal passages. The present study provides novel evidence to show that many MDH neurons activated by repetitive diving co-express ionotropic glutamate receptors. This strengthens the assumption that trigeminal nerves innervating the nasal mucosa release glutamate as a neurotransmitter to activate second-order neurons within the MDH during underwater submergence. Additionally, a significant portion of MDH neurons activated by repetitive diving co-express the receptor NK-1, suggesting that SP is also released within the MDH by trigeminal nerves innervating the nasal mucosa. These current results are also very similar to our previous study that used ammonia vapors to stimulate the nasal passages of anesthetized rats (McCulloch et al., 2013). This suggests that similar trigeminal circuitry within the MDH is used both during repetitive diving in conscious animals and nasal stimulation in anesthetized animals.

4.1. MDH neurons express Fos after repetitive diving

The c-fos gene and the Fos protein are markers of cellular activity and are easily expressed after a wide range of stimuli (Coggeshall, 2005, Lara Aparicio et al., 2022, Morgan and Curran, 1991). While there are limitations to the technique, Fos has been studied in several experimental animal models since the 1990s, and its use as a marker of cellular activation for many physiological processes, such as sensory information processing, is useful and reliable (Kovacs, 2008, Lara Aparicio et al., 2022). However, since Fos can be produced in numerous signaling pathways, its expression does not provide specific information about a given pathway, including inclusion of inhibitory processes (Coggeshall, 2005, Dragunow and Faull, 1989, Kovacs, 2008, Lara Aparicio et al., 2022). Immunohistochemistry is generally used to detect the presence of the Fos protein, and it is widely used to enable in vivo identification of activated cells (Coggeshall, 2005, Kovacs, 2008, Lara Aparicio et al., 2022, Morgan and Curran, 1991).

In the present study, Fos-positive neurons were found in the ventral MDH between the pyramidal decussation caudally and the obex rostrally in rats repetitively diving through an underwater maze. As an internal control, and as would be expected, the total number of Fos-positive neurons within each of the three MDH regions were similar and varied little between the 10 separate receptor subunits being investigated (Table 2). Previous studies have shown increases in Fos labelling within the ventral MDH in diving rats but not in swimming or non-diving control rats (Mcculloch, 2005, Mcculloch et al., 2016, Panneton et al., 2012). Trigeminal nerves innervating the nose and nasal passages project to this same ventral MDH location (McCulloch et al., 2018). Thus neurons within the ventral MDH activated during diving would likely receive and process the afferent signals from the nose and nasal passages involved in initiating the diving response (Mcculloch et al., 2018, Panneton et al., 2006), including reflexively produced apnea (Pérez et al., 2023, Perez et al., 2020).

Morphologically, Fos-positive neurons within the MDH activated during diving were generally small and oval shaped. This is not unexpected, as histological subnuclear regions of the spinal trigeminal caudalis have been previously described as containing densely packed small oval or fusiform neurons with a thin rim of cytoplasm in the rat (Ide and Killackey, 1985, Ivanov et al., 2024, Li et al., 1999, Li et al., 2000, Sedlacek et al., 2007, Voisin et al., 2002), cat (Kerr, 1966, Kerr, 1970), and human (Gobel et al., 1988, Olszewski, 1950). Neurons within the MDH superficial laminae (laminae I and II) receive primary afferent projections from craniofacial regions (Darian-Smith, 1973), receiving sensory information relayed primarily via glutamate released from trigeminal nerves (Largent-Milnes et al., 2014, Todd, 2010). Within both the spinal dorsal horn (Todd, 2010, Wu et al., 2010) and medullary dorsal horn (Sedlacek et al., 2007) local circuits are formed from excitatory (glutamatergic) and inhibitory (GABAergic) neurons. Projection neurons located in lamina I send information from these local circuits toward the thalamus and parabrachial nuclei (Kobayashi et al., 2024). While the MDH is most concerned with the relay of nociceptive and thermal input (Hu, 1990, Sedlacek et al., 2007), this nucleus does mediate tactile input, and 32 % of neurons within the deep laminae (laminae III-V) respond maximally to innocuous mechanical stimulation (Hayashi, 1985, Mchaffie et al., 1994).

4.2. Glutamate receptor co-expression by Fos-positive neurons

Results from the present study indicate that MDH neurons activated by repetitive diving co-express ionotropic glutamate receptor subunits. Neurons within the paratrigeminal nucleus highly express AMPA (especially GluA2/3), KA (especially GluK5), and NMDA (especially GluN1) receptor subunits. Neurons within the superficial MDH highly express AMPA (especially GluA1 and GluA2/3), KA (especially GluK5), and NMDA (especially GluN1) receptor subunits. In addition, neurons within the deep MDH highly express AMPA (especially GluA2/3) and NMDA (especially GluN1) receptor subunits, but not KA receptor subunits.

The formation of functional glutamate receptors is based on the assembly and stoichiometry of individual receptor subunits (Bettler et al., 2021, Traynelis et al., 2010). However, the expression of subunits from different classes of ionotropic glutamate receptors in each of the MDH regions confirms the formation of the three subtypes of glutamate receptors (AMPA, KA, and NMDA) within the synapses located on these activated neurons (Alexander et al., 2019, Parsons et al., 2005). The presence of glutamate receptors would support the assertion that the neurotransmitter glutamate is released into synapses located on these neurons (Parsons et al., 2005, Raz and Perouansky, 2019), which then activates the MDH neurons and induces them to produce Fos during diving. Therefore, we accept our first hypothesis that MDH neurons activated by repetitive diving co-express ionotropic glutamate receptor subunits, specifically AMPA and NMDA receptor subunits. The involvement of ionotropic glutamate receptors (AMPA, KA, and NMDA) in the processing of somatosensory information within the trigeminal nucleus has been found previously (Han et al., 2008, Hegarty et al., 2007, Mcculloch et al., 2013, Tallaksen-Greene et al., 1992, Tang et al., 2001), and supports the present findings.

In addition to the AMPA subunit GluA2/3 found in all three MDH regions, we found a high expression of GluA1 subunits in the superficial MDH, but not in the paratrigeminal or deep MDH. The significance of this stoichiometric difference in subunit assembly is unknown but may relate to the specific functionality of the glutamatergic AMPA receptors within the superficial MDH that is not present within the other two MDH regions. In addition, we found that Fos-positive MDH neurons within the superficial MDH and paratrigeminal nucleus, but not the deep MDH, also co-expressed KA receptor subunits, although no MDH region expressed GluK4 subunits. At present the significance of KA receptors in the superficial MDH and paratrigeminal nucleus, but not the deep MDH, is unknown. However, it is interesting that in rats stimulated by ammonia vapors applied to the nasal passages to induce a cardiorespiratory response similar to the diving response, rather than by repetitive diving as in the present study, the presence of KA receptors on activated neurons was not found in any MDH region (McCulloch et al., 2013). The reason for this differential KA receptor expression currently remains unclear, but could be due to the use of anesthesia in McCulloch et al. (2013) and conscious non-anesthetized rats in the present study.

The kinetic properties of glutamate receptor families seem well designed to serve their respective roles in synaptic transmission. AMPA receptors mediate fast neurotransmission as their rapid kinetics are suitable for this purpose (Dingledine et al., 1999, Ozawa et al., 1998, Traynelis et al., 2010). AMPA and KA receptors are rapidly activated by high concentrations of glutamate with a high probability of opening but possess low-affinity binding sites for glutamate and thus are thought to deactivate quickly (Dingledine et al., 1999). In contrast, neurotransmission mediated by NMDA receptors occurs more slowly and lasts for a much longer period, since their kinetics are much slower than those of AMPA receptors (Ozawa et al., 1998, Raz and Perouansky, 2019). In addition, AMPA receptor channels are functionally dominated by GluA2 subunits that show low Ca²⁺ permeability, whereas NMDA receptors have high Ca²⁺ permeability characteristics (Bettler et al., 2021, Chen et al., 2000, Ozawa et al., 1998). Furthermore, since AMPA and NMDA are often colocalized at central synapses (Chen et al., 2006, Chen et al., 2000, Dingledine et al., 1999), the rapid opening and closing kinetics of AMPA receptors in response to neurotransmitter release can induce partial relief of voltage-dependent Mg²⁺ blockade of the more slowly activating NMDA receptors (Dingledine et al., 1999). It appears that functional AMPA receptors are important for the enhancement of synaptic efficiency and maintenance of neuronal dendritic spines where most central excitatory synapses reside, while NMDA receptors play more roles in long-term potentiation (LTP) and synaptic plasticity (Chen et al., 2000, Ozawa et al., 1998, Raz and Perouansky, 2019). Despite their widespread distribution throughout the CNS, the physiological significance of KA receptors remains largely unknown, although they may have more of a neuromodulatory role (Bowie, 2008, Lerma, 2006, Ozawa et al., 1998, Traynelis et al., 2010).

Based on the present evidence, activation of MDH neurons during repetitive diving appears to involve both fast and slow glutamatergic neurotransmission. The presence of AMPA receptors in the superficial MDH, deep MDH and paratrigeminal nucleus may mediate fast neurotransmission involved in the local MDH circuitry important for generating the cardiorespiratory reflex responses seen after both repetitive diving and nasal stimulation (McCulloch et al., 2013). In addition, the presence of NMDA receptors in these same MDH regions may involve long-term modulation of these synapses after repetitive stimulation. Colocalization of AMPA and NMDA receptors within each of the MDH regions may also be important for the functionality of neurotransmission through MDH circuitry, as it is within the vestibular nucleus (Chen et al., 2000, Tse et al., 2008), inferior olive (Chen et al., 2006), and forebrain regions (Wang et al., 2004).

4.3. NK-1 receptor co-expression by Fos-positive neurons

Results from the present study show that NK-1, the receptor for SP, is also highly expressed by MDH neurons activated by repetitive diving. This was observed in the superficial MDH, deep MDH, and paratrigeminal nucleus. Therefore, based on these data, we accept our second hypothesis that MDH neurons activated by repetitive diving co-express NK-1 receptors. These findings suggest that nerves innervating the nose and nasal passages release SP as a neurotransmitter, or perhaps as a neuromodulator, during repetitive diving. The NK-1 receptor is involved in cellular responses such as pain transmission, endocrine and paracrine secretion, vasodilation, and modulation of cell proliferation (Garcia-Recio and Gascón, 2015). Within the dorsal horn, glutamate and SP have been shown to coexist in primary afferent C-fibers (Zieglgänsberger, 2019) and are co-released by primary afferent neurons involved in mediating nociception (Battaglia and Rustioni, 1988, Randic et al., 1995). However, animals would not voluntarily dive underwater if this were a nociceptive event. Instead, the release of fast-acting neurotransmitters, such as glutamate and neuropeptides like SP, may contribute significantly to synergistic responses to external stimuli (Aicher et al., 1997, Zieglgänsberger, 2019) such as those occurring during underwater submersion. Tachykinins, such as SP, appear to be functionally involved in slow excitatory synaptic transmission (i.e. enhancement of NMDA-induced current) within the dorsal horn, and the role of neuropeptides in long-term modulation of synaptic transmission is well established (Aicher et al., 1997, Randic et al., 1995).

4.4. Diving response brainstem circuitry

The diving response is autonomic reflex that produces apnea, bradycardia and increased total peripheral resistance (TPR) when animals submerge underwater. The diving response, like all autonomic reflexes, consists of an afferent input, central CNS integration and an efferent response. In the present study, Fos-positive neurons were found in the ventral MDH between the pyramidal decussation caudally and the obex rostrally in rats engaged in repetitive diving. These activated MDH neurons likely receive and process the afferent input from the nose and nasal passages involved in initiating the diving response. As such, the present research contributes new information about the afferent limb of the diving response, namely that the nerves innervating the nasal passages are likely to release glutamate and SP within their central synapses, thus contributing to the activation of these MDH neurons.

In addition to the MDH, activated brainstem areas during diving include the caudal pressor area, nucleus tractus solitarius, rostral ventrolateral medulla, Raphe nuclei, A5, Locus Coeruleus, Kölliker-Fuse area, and subnuclei of the parabrachial nucleus (Mcculloch et al., 2016, Panneton et al., 2012). Since these specific brainstem regions are activated in diving rats but not in non-diving rats, this suggests these other brainstem regions constitute further integrative and/or efferent aspects of the diving response. A schematic brainstem circuit of the diving response, presented in the graphical abstract, indicates there must be three neuronal pathways projecting from the MDH that accounts for the three efferent aspects of the diving response, specifically apnea, bradycardia and increased TPR. For discussions on brainstem pathways and the specific neuronal locations that mediate the physiological aspects of the diving response, see Panneton et al. (2000), Panneton (2013), and Panneton and Gan (2020).

4.5. Study limitations and future investigations

A limitation of the present study is the lack of experiment control animals. Previously it was found that the only way to activate neurons within the MDH during diving is to have the animals dive underwater (Mcculloch, 2005, Panneton et al., 2012). Significantly fewer MDH neurons are activated during swimming or when dive-trained animals are left in their cages (Mcculloch, 2005, Panneton et al., 2012). Thus, swimming or cage control cohorts would not add to study results, because in these non-diving animals we would not be able to identify the MDH neurons that are part of the diving response. We reasoned that we would only use diving animals in this study because only in these animals would Fos identify the neurons of interest. If swimming or cage control animals were included in this study, we would encounter fewer Fos-positive neurons, but more importantly we could never be sure if the Fos-positive neurons in the MDH of swimming or cage control animals were part of the diving response circuitry. We recognize that in this study we were narrowly focused on a subset of MDH neurons (i.e. those neurons involved in the diving response) rather than the research being an inclusive study on the MDH itself.

Other limitations of the present investigation include is the lack of nuclear staining that specifically identifies neurons. In future studies co-staining with NeuN would ensure that identified Fos-positive cells are neurons rather than astrocytes (Gusel'nikova and Korzhevskiy, 2015). NeuN (or DAPI) would also identify the presence of a cell nucleus that is surrounded by the halo of receptor subunit only labeling. Additionally, staining tissue for Fos, AMPA receptors and NMDA receptors (as well as NeuN) to visualize colocalization within the same activated neuron would better determine the role of fast and slow neurotransmission within the MDH circuitry of the diving response. The addition of morphological analysis to the enhanced immunohistochemical staining would also help elucidate functional relevance in this regard. Finally, applying this more robust analysis to the different MDH regions (superficial MDH, deep MDH, and paratrigeminal nucleus) at different MDH levels (caudal, middle and rostral) would further increase the understanding of trigeminal nuclear functioning. Additionally, the staining for inhibitory (i.e. GABAergic) neurons could add to the understanding of functional local trigeminal reflex circuits. Clearly additional research is warranted.

5. Summary

In conclusion, conscious rats that repetitively dive underwater show the efferent responses to diving, including an intense bradycardia and increase in total peripheral resistance (Chotiyanonta et al., 2013, Panneton et al., 2010). As part of the afferent aspect of the diving response circuitry, the present research indicates neurons within the ventral MDH express Fos during repetitive diving. The present research also indicates that these MDH neurons activated by diving express AMPA, KA, and NMDA subtypes of glutamate receptors and NK-1 receptors for SP. Colocalization of AMPA and NMDA receptors within the ventral MDH may be important for the functionality of neurotransmission through MDH local circuitry. These results are very similar to those found previously by McCulloch et al. (2013) in which ammonia vapors were used to stimulate the nasal passages of anesthetized rats instead of repetitive diving (McCulloch et al., 2013). Together these studies parsimoniously suggest the nerves innervating the nasal passages release glutamate and SP within their central synapses in the brainstem and can function as part of the afferent limb of a cardiorespiratory reflex.

CRediT authorship contribution statement

K.M. DiNovo: Visualization, Supervision, Project administration, Methodology, Formal analysis. P.F. McCulloch: Writing – review & editing, Writing – original draft, Methodology, Investigation, Formal analysis, Conceptualization.

Conflicts of Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.