Skip to main content
NCBI home page
The Journal of Physiology logoLink to The Journal of Physiology
. 2018 Nov 22;597(1):283–301. doi: 10.1113/JP276467

Glycinergic neurotransmission in the rostral ventrolateral medulla controls the time course of baroreflex‐mediated sympathoinhibition

Hong Gao 1, Willian S Korim 2, Song T Yao 2, Cheryl M Heesch 3, Andrei V Derbenev 1,4,✉
PMCID: PMC6312423  PMID: 30312491

Abstract

Key points

  • To maintain appropriate blood flow to various tissues of the body under a variety of physiological states, autonomic nervous system reflexes regulate regional sympathetic nerve activity and arterial blood pressure.

  • Our data obtained in anaesthetized rats revealed that glycine released in the rostral ventrolateral medulla (RVLM) plays a critical role in maintaining arterial baroreflex sympathoinhibition.

  • Manipulation of brainstem nuclei with known inputs to the RVLM (nucleus tractus solitarius and caudal VLM) unmasked tonic glycinergic inhibition in the RVLM.

  • Whole‐cell, patch clamp recordings demonstrate that both GABA and glycine inhibit RVLM neurons.

  • Potentiation of neurotransmitter release from the active synaptic inputs in the RVLM produced saturation of GABAergic inhibition and emergence of glycinergic inhibition.

  • Our data suggest that GABA controls threshold excitability, wherreas glycine increases the strength of inhibition under conditions of increased synaptic activity within the RVLM.

Abstract

The arterial baroreflex is a rapid negative‐feedback system that compensates changes in blood pressure by adjusting the output of presympathetic neurons in the rostral ventrolateral medulla (RVLM). GABAergic projections from the caudal VLM (CVLM) provide a primary inhibitory input to presympathetic RVLM neurons. Although glycine‐dependent regulation of RVLM neurons has been proposed, its role in determining RVLM excitability is ill‐defined. The present study aimed to determine the physiological role of glycinergic neurotransmission in baroreflex function, identify the mechanisms for glycine release, and evaluate co‐inhibition of RVLM neurons by GABA and glycine. Microinjection of the glycine receptor antagonist strychnine (4 mm, 100 nL) into the RVLM decreased the duration of baroreflex‐mediated inhibition of renal sympathetic nerve activity (control = 12 ± 1 min; RVLM‐strychnine = 5.1 ± 1 min), suggesting that RVLM glycine plays a critical role in regulating the time course of sympathoinhibition. Blockade of output from the nucleus tractus solitarius and/or disinhibition of the CVLM unmasked tonic glycinergic inhibition of the RVLM. To evaluate cellular mechanisms, RVLM neurons were retrogradely labelled (prior injection of pseudorabies virus PRV‐152) and whole‐cell, patch clamp recordings were obtained in brainstem slices. Under steady‐state conditions GABAergic inhibition of RVLM neurons predominated and glycine contributed less than 25% of the overall inhibition. By contrast, stimulation of synaptic inputs in the RVLM decreased GABAergic inhibition to 53%; and increased glycinergic inhibition to 47%. Thus, under conditions of increased synaptic activity in the RVLM, glycinergic inhibition is recruited to strengthen sympathoinhibition.

Keywords: baroreflex, glycine, GABA, rostral ventrolateral medulla, patch‐clamp

Key points

  • To maintain appropriate blood flow to various tissues of the body under a variety of physiological states, autonomic nervous system reflexes regulate regional sympathetic nerve activity and arterial blood pressure.

  • Our data obtained in anaesthetized rats revealed that glycine released in the rostral ventrolateral medulla (RVLM) plays a critical role in maintaining arterial baroreflex sympathoinhibition.

  • Manipulation of brainstem nuclei with known inputs to the RVLM (nucleus tractus solitarius and caudal VLM) unmasked tonic glycinergic inhibition in the RVLM.

  • Whole‐cell, patch clamp recordings demonstrate that both GABA and glycine inhibit RVLM neurons.

  • Potentiation of neurotransmitter release from the active synaptic inputs in the RVLM produced saturation of GABAergic inhibition and emergence of glycinergic inhibition.

  • Our data suggest that GABA controls threshold excitability, wherreas glycine increases the strength of inhibition under conditions of increased synaptic activity within the RVLM.

Introduction

The arterial baroreflex provides a rapid negative‐feedback loop to maintain blood pressure (BP) via reciprocal activation or inhibition of the parasympathetic and sympathetic branches of the autonomic nervous system. Modulation of GABAergic inhibition of presympathetic neurons in the rostral ventrolateral medulla (RVLM) is a primary mechanism for depressor and sympathoinhibitory responses (Llewellyn‐Smith & Verberne, 2011; Guyenet et al. 2013). The RVLM receives GABAergic inputs from the midbrain, medullary regions and other brain areas (Llewellyn‐Smith & Verberne, 2011; Bowman et al. 2013; Guyenet et al. 2013). GABAergic neurons in the caudal VLM (CVLM) are the main source of baroreflex‐mediated inhibition in the RVLM (Sun & Guyenet, 1985; Schreihofer & Guyenet, 1997; Chan & Sawchenko, 1998; Dampney et al. 2003a; Dampney et al. 2003b; Llewellyn‐Smith & Verberne, 2011; Guyenet et al. 2013). Interestingly, co‐localization of GABA and glycine has been found in the RVLM (Llewellyn‐Smith et al. 2001; Stornetta et al. 2004). Focal electrical stimulation triggered GABAergic and glycinergic synaptic currents in presympathetic RVLM neurons, suggesting that they receive GABAergic and/or glycinergic synaptic inputs (Dun & Mo, 1989; Deuchars et al. 1997). On the other hand, under control conditions, blockade of GABAergic but not glycinergic neurotransmission in the RVLM increased sympathetic nerve activity and BP (Blessing, 1988; Guyenet et al. 1990; Amano & Kubo, 1993; Cravo & Morrison, 1993; Heesch et al. 2006), suggesting that RVLM neurons are primarily under sustained tonic GABAergic inhibition; however, possible mechanisms of glycine release and a potential role for co‐inhibition of presympathetic RVLM neurons by GABA and glycine are not known.

In general, GABA and glycine generate fast synaptic currents (IPSCs) by acting at distinct populations of postsynaptic ionotropic receptors. They can be co‐released from a single synapse (Jonas et al. 1998; O'Brien & Berger, 1999; Nabekura et al. 2004; Dufour et al. 2010; Takazawa & MacDermott, 2010) consistent with synaptic co‐localization of GABA and glycine (Llewellyn‐Smith et al. 2001; Nabekura et al. 2004; Stornetta et al. 2004; Dufour et al. 2010). In addition, GABA and glycine share a common vesicular transporter (Wojcik et al. 2006) and GABAA receptors (GABAAR) and glycine receptors (GlyR) are usually clustered together on postsynaptic cells (Levi et al. 1999; Fischer et al. 2000; Kneussel & Betz, 2000). Despite the fact that GABAergic and glycinergic neurotransmission are closely related, the kinetics of the current (i.e. rise and decay time) are different. Because GABA induces a more sustained synaptic inhibition than glycine, this difference has implications for the strength and timing of inhibition (Russier et al. 2002; Rahman et al. 2013; McMenamin et al. 2016). Moreover, both GABAAR and GlyR undergo developmental maturation with changes in receptor subunit composition and faster kinetics (Liu & Wong‐Riley, 2013).

By contrast to the information available on GABAergic inhibition of RVLM neurons, the role of glycine in the control of RVLM excitability and baroreflex function is less well understood. Previously, we reported that GlyR in the RVLM mediated GABAAR independent inhibition of BP and renal sympathetic nerve activity (RSNA) (Heesch et al. 2006). The present study aimed to determine the potential physiological significance of glycinergic inhibition, identify the mechanisms of glycine release, and evaluate the co‐inhibition of presympathetic RVLM neurons by GABA and glycine.

Methods

Ethical approval

Male Sprague–Dawley rats were purchased from ENVIGO (Indianapolis, IN, USA) and the Animal Resources Centre (Murdoch, WA, Australia). The animals were housed in an institutionally approved on‐site animal care facility under a 12:12 h light/dark cycle at a constant 22°C and 40% relative humidity, with food and water available ad libitum. Housing of animals and experiments were performed in accordance with the guidelines of the National Institutes of Health Guide for the Care and Use of Laboratory Animals and the Australian Code of Practice for the Care and Use of Animals for Scientific Purposes. All experiments were approved by the Institutional Animal Care and Use Committees of University of Missouri, Tulane University and the Florey Animal Ethics Committee, Melbourne, Australia. The investigators understand the ethical principles under which the journal operates and the work described in the present study complies with these animal ethics principles (Grundy, 2015).

Role of glycine in the RVLM on arterial baroreflex responses

Adult male Sprague–Dawley rats (weighing 250–350 g, n = 6) were anaesthetized (pentobarbitone 60 mg kg−1 i.v.), paralysed (pancuronium bromide, 1 mg kg−1 i.v.) and artificially ventilated, as described previously (Korim et al. 2012). Following the paralysing agent, maintenance of anaesthesia was verified by a <10 mmHg change in mean blood pressure in response to paw pinch. Temperature was kept at ∼37.5 ± 0.5°C using a thermostat controlled heating pad, connected to a rectal probe.

The left renal sympathetic nerve was prepared and recorded using bipolar silver hook electrodes as described previously (Yao et al. 2015). The activity was amplified (10,000×) and filtered (100–1000 Hz). Signals were sampled at 10 kHz, and digitized. Neurograms were quantified by rectifying and time‐averaging with a 1 s time constant, followed by subtraction of the noise level (0%) and normalization adopting the baseline as 100%. Baseline nerve activity was taken as the average of 5 min recordings in arbitrary units prior to baroreflex challenge.

The right femoral vein and artery were cannulated for drug administration and arterial BP recordings, respectively. Heart rate and mean arterial BP were derived from the BP channel. Following ∼5 min of stable RSNA recordings (i.e. baseline), bolus injections of phenylephrine (5 μg kg−1 i.v.) were administered and baroreflex‐mediated inhibition of RSNA was assessed before and 10 min after blockade of glycine receptors in the RVLM (strychnine, 4 mm, 100 nL). Stereotaxic co‐ordinates and procedures for microinjections into the RVLM were as described previously (Korim et al. 2014).

Strychnine was diluted in a solution containing 2% red fluorescent latex microbeads diluted in artificial cerebrospinal fluid (aCSF, in mm: 128 NaCl; 2.6 KCl, 1.3 NaH2PO4, 2 NaHCO3, 1.3 CaCl2 and 0.9 MgCl2). Fluorescent latex beads were injected to identify the centre of injection. Following assessment of baroreflex function, animals were killed with a pentobarbitone overdose (100 mg kg−1), transcardially perfused with PBS followed by 4% paraformaldehyde in 10 mm phosphate buffer (pH 7.4). Brains were removed and processed for histology, and microinjection sites were identified histologically as described previously (Korim et al. 2014).

Tonic glycinergic inhibition in the RVLM

In 28 adult male Sprague–Dawley rats (weighing 245–360 g), anaesthesia was induced with isoflurane and slowly transitioned to inactin (100 mg kg−1 i.v.). Rats were artificially ventilated (room air supplemented with O2) and paralysed with continuous i.v. infusion of gallamine triethiodide (25 mg kg−1 h−1 for first 15 min; then 12.5 mg kg−1 h−1 for the remainder of the experiment). Neuromuscular blockade in ventilated animals was used for all in vivo experiments to eliminate potential changes in respiration and pulmonary reflexes that could contribute to the observed responses. Depth of anaesthesia was assessed at 15 min intervals as described above and supplemental anaesthetic (inactin, 10 mg kg−1 i.v.) was administered as needed. Placement of vascular catheters, recording of arterial pressure and RSNA procedures were conducted as described above. During the experiments, brainstem nuclei were targeted based on stereotaxic co‐ordinates, and functionally defined by appropriate BP and RSNA responses to microinjection of glutamate (30 nL, 10 mm). All CNS microinjections were bilateral and administered serially with ≤1 min between the left and right side. At the end of experiments, rats were given a lethal dose of a solution that resulted in death (Somnasol; Henry Schein® Animal Health, Dublin, OH, USA; 390 mg pentobarbital sodium + 50 mg phenytoin sodium in 1 mL i.p.), transcardially perfused, and brains were harvested; proper pipette placement was subsequently verified by histological evaluation using methods similar to those described previously (Heesch et al. 2006).

BP and RSNA responses to bilateral injections into the RVLM of strychnine (3 mm, 100 nL), followed by bicuculline (4 mm, 100 nL) at the peak response to strychnine, were evaluated in four groups of rats. Control rats (n = 5) received no other treatment. In a second group of rats, to first block output from the nucleus tractus solitarius (NTS), muscimol (60 nL, 2 mm) was microinjected bilaterally into the NTS (NTS‐X, n = 5). Responses to microinjection of strychnine plus bicuculline into the RVLM were tested 30–60 min after NTS blockade. Neuronal circuits within the CVLM were activated (disinhibited) by bilateral microinjection of bicuculline (100 nL, 4 mm) into the CVLM in a third group of rats (CVLM‐B, n = 6). Responses to strychnine plus bicuculline in the RVLM were then tested 7–10 min after disinhibition of the CVLM. In a fourth group of rats, the NTS was first blocked by muscimol, followed by disinhibition of the CVLM, and responses to strychnine plus bicuculline in the RVLM were then tested (CVLM‐B[NTS‐X], n = 6). RSNA was expressed as a percentage of the initial baseline values before any CNS microinjections were performed.

In a separate group (n = 6), rats were prepared as described above for bilateral microinjection of muscimol into the NTS. Arterial pressure and RSNA were recorded and the time course of NTS blockade was evaluated. Arterial baroreflex inhibition of RSNA in response to i.v. phenylephrine (5 μg kg−1) was tested before and 5, 30, 60, 90 and 120 min after bilateral NTS muscimol injections.

Slice electrophysiology

In 41 male Sprague–Dawley rats (4–6 weeks old), we studied 52 neurons. Pseudorabies virus 152 [PRV‐152; a retrogradely transported viral vector strain isogenic with PRV‐Bartha that reports enhanced green fluorescent protein (GFP)] was supplied by NCRR CNNV Virus Centre (Pittsburgh, PA, USA). PRV‐152 was used to identify neurons in the RVLM (Gao & Derbenev, 2013). Briefly, under isoflurane anaesthesia (2–3%), a small dorsolateral incision was made to expose the left kidney. Two injections (2 μL each) of PRV‐152 (1 × 108 plaque‐forming units per millilitre) were made into the cortex of the left kidney with a glass pipette with a tip diameter of 50 μm. To minimize discomfort during and after PRV‐152 infection, the animals were treated with a s.c. injection of buprenorphine (0.05 mg kg−1). After surgery, animals were maintained in a biosafety level 2 facility.

At 90–96 h after inoculation with PRV‐152, rats were anaesthetized with 2–3% of isoflurane and then decapitated. Transverse brainstem slices (∼300 μm thick, ∼2 per brain) containing PRV‐labelled RVLM neurons were made with a vibrating microtome (Series 1000; The Vibratome Company, St Louis, MO, USA) and stored at 34–37 °C. PRV‐152 labelled RVLM neurons were visualized and targeted for recording based on their fluorescence and via post hoc identification using the avidin‐Texas Red reaction (Derbenev et al. 2004). Recording pipettes were directed to these neurons using infrared deferential interference contrast optics (FN1; Nikon, Tokyo, Japan) with a 40× water‐immersion objective. Patch electrodes (2–6 MΩ) were filled with a solution containing (in mm): 130 K‐gluconate or Cs‐gluconate, 1 NaCl, 5 EGTA, 10 Hepes, 1 MgCl2, 1 CaCl2, 2–4 ATP and 0.2% biocytin (pH 7.2–7.4) (adjusted with KOH or CsOH). To demonstrate the distribution of PRV‐labelled neurons in the RVLM, a scanning laser confocal microscope (DMi8; Leica, Wetzlar, Germany) was used.

Spontaneous IPSCs (sIPSCs) were examined at a holding potential of −10 mV. This experimental condition permitted us to exclude excitatory glutamatergic transmission without using glutamate receptor antagonists that could alter the presynaptic release of inhibitory neurotransmitters (Xu & Smith, 2015; Boychuk & Smith, 2016). GABA (100 μm) and glycine (500 μm) were bath applied. To segregate GABAergic from glycinergic IPSCs, we used bicuculline methiodide (30 μm). 4‐aminopyridine (4‐AP), a non‐selective, voltage‐dependent potassium channel blocker, was used to prolong the depolarization of neurons and increase the release of neurotransmitters from active synapses (Thompson, 1977). 4‐AP was dissolved in aCSF and bath applied for 3–5 min at final concentration of 2 mm. TTX, a voltage‐activated sodium channel blocker, was used to diminish action potential‐dependent neurotransmitter release. The frequency and amplitude of sIPSCs and miniature IPSCs (mIPSCs) were analysed within individual neurons as described previously (Gao & Derbenev, 2013). The mean phasic current (integrated current of IPSCs) was calculated using: I phasic = f × Q, where f is the synaptic frequency and Q is the charge transfer measured as the area under the IPSC (Park et al. 2006; Gao & Smith, 2010a). All of the chemicals for in vivo and in vitro studies were obtained from Sigma‐Aldrich (St Louis, MO, USA) and Tocris Bioscience (St Louis, MO, USA).

Experimental design and statistical analysis

The effect of RVLM strychnine on the time course of baroreflex inhibition was determined by the period between the response onset and return to baseline level. A D'Agostino & Pearson omnibus test was performed to verify a normal distribution of the data. Parametric statistical analysis was performed using a two‐tailed paired Student's t test. All statistical tests for baroreflex experiments were performed using Prism, version 5.0 (GraphPad Software Inc., San Diego, CA, USA).

The effects of NTS inhibition by muscimol and disinhibition of the CVLM, as well as the combination of the two, on tonic responses to strychnine and bicuculline in the RVLM were evaluated. Two‐way repeated measures (RM) ANOVA was used to compare mean arterial pressure (MAP) or RSNA values at baseline (prior to RVLM injections, 100%), peak response to RVLM strychnine and peak response to RVLM bicuculline, among the four groups. If a significant interaction between group (control, NTS‐X, CVLM‐B, CVLM‐B[NTS‐X]) and treatment (RM = base, strychnine, bicuculline) was present, analyses for treatment and group effects were performed followed by a post hoc Student–Newman–Keuls test (Sigma Plot, version 12.5; Systat Software Inc., Chicago, IL, USA). To evaluate the effects of pre‐treatments prior to RVLM injections, MAP and RSNA before and after NTS muscimol were compared by a two‐tailed paired t test and the biphasic effects of bicuculline in the CVLM (with and without prior NTS muscimol) were compared by two‐way RM ANOVA. To meet the criteria of normality and equal variance, in some instances, square root or log transformations were performed prior to statistical comparison. Data evaluating the time course of NTS blockade with muscimol were evaluated using one‐way RM ANOVA.

Effects of bicuculline, strychnine, TTX and 4‐AP on the frequency and amplitude of IPSCs in brain slice experiments were analysed within a recording using the Kolmogorov–Smirnov test (a non‐parametric, distribution‐free goodness‐of‐fit test for probability distributions), with at least 1 min of continuous activity being measured for each condition. The frequency and amplitudes of IPSCs, input resistance and the resting membrane potential within one neuron were analysed using a paired, two‐tailed Student's t test for Gaussian distributed values or a two‐tailed Wilcoxon matched‐pair signed rank test for non‐Gaussian distributed values. Multiple comparisons within the population for Figs. 8 and 9 were performed with one‐way ANOVA for Gaussian distributed values or Friedman test followed by Dunn's multiple comparisons test (Prism, version 7.0) for non‐Gaussian distributed values. Data are presented as the mean ± SEM. P < 0.05 was considered statistically significant.

Figure 8. 4‐AP increased GlyR‐mediated synaptic currents.

Figure 8

Open in a new tab

A–D, showing that application of 4‐AP (2 mm) significantly increased the frequency, amplitude, area of events and I phasic in presympathetic RVLM neurons. Additional application of bicuculline (Bic) (30 μm) decreased the amplitude, frequency and I phasic. E, continuous recording of sIPSCs from a presympathetic RVLM neuron voltage clamped at −10 mV in the presence of 4‐AP and bicuculline before and after strychnine (Str) (1 μm). Results were similar in three additional cells tested. The data confirmed that co‐application of bicuculline and strychnine abolished all the remaining sIPSCs. Numbers of replications are shown in parentheses. Bar graphs represent the mean ± SEM; open circles represent individual data points; * P < 0.05, ** P < 0.01, *** P < 0.001.

Figure 9. The component of GlyR‐mediated synaptic currents increased during network activation.

Figure 9

Open in a new tab

A, showing the frequency of GlyR‐mediated IPSCs after blockade of GABAAR with bicuculline (30 μm). Bicuculline blocked GABAAR‐mediated synaptic currents revealing synaptic events mediated by GlyR. B, showing the percentage of IPSCs mediated by GlyR and GABAAR in the presence of TTX, without TTX (aCSF only) and during 4‐AP application. Bicuculline was used to block GABAAR‐mediated events. Numbers of replications are shown in parentheses. Bar graphs represent the mean ± SEM; open circles represent individual data points; * P < 0.05, ** P < 0.01.

Results

Glycine receptor blockade in the RVLM decreased the duration of baroreflex‐mediated sympathoinhibition

Phenylephrine (5 μg/kg i.v.) produced equivalent increases in mean arterial pressure before (ΔBP: +75 ± 4 mmHg) and after (ΔBP: +78 ± 5 mmHg) bilateral RVLM microinjections of strychnine (4 mm, 100 nL) (n = 6, t 5 = 0.763, P = 0.24). Maximum baroreflex‐mediated inhibition of RSNA, achieved 15 s after phenylephrine injection, was also similar between control and RVLM strychnine trials (ΔRSNA: −98 ± 1 vs. ΔRSNA: −98 ± 1%) (n = 6, t 5 = 0.163, P = 0.438) (Fig. 1 C). However, as indicated by the recordings from one rat, the duration of inhibition of RSNA following phenylephrine was reduced by prior blockade of glycine receptors in the RVLM (Fig. 1 A). Group data (Fig. 1 B and C) show a reduction in efficiency of the baroreflex as indicated by a faster recovery of RSNA to baseline levels, (5.1 ± 1.0 vs. 11.6 ± 1.3 min, n = 6, t 5 = 3.457, P = 0.009) compared to control. Synchronous pulse‐triggered sympathetic modulation confirmed that RSNA was cardiovascular related (Fig. 1 D).

Figure 1. Time‐course of baroreflex is controlled by glycine in the RVLM.

Figure 1

Open in a new tab

A, responses to baroreceptor loading on mean arterial BP and RSNA following phenylephrine (i.v.) before and after bilateral administration of 100 nL of 4 mm (400 pmol) strychnine into the rostral ventrolateral medulla (RVLM); ‘a’ and ‘b’, corresponding raw traces. B, mean data indicate that microinjection of strychnine into the RVLM decreased the duration of the RSNA inhibition. C, group data show that the reduction in the efficiency of the baroreflex is represented by a decrease in the recovery time for RSNA, without affecting the magnitude of the inhibition (ΔRSNA). D, rectified and integrated SNA is phase‐locked with individual cardiac cycles, represented by changes in BP. E, schematic diagram showing the location of microinjection sites (100 nL). Brainstem sections were adapted from The Rat Brain in Stereotaxic Coordinates (Paxinos & Watson, 2007). Right: photomicrograph of a bainstem coronal slice showing the distribution of fluorescent red beads into the RVLM. V, trigeminal nucleus; Amb, nucleus ambiguous; iO, inferior olive; Py, pyramidal tract. Numbers of animals are shown in parentheses (n = 6). Bar graphs represent the mean ± SEM; open circles represent individual data points. * P < 0.05, ** P < 0.01.

Blockade of NTS and disinhibition of the CVLM unmasked glycinergic inhibition of the RVLM

Representative examples of responses to microinjection of strychnine and bicuculline into the RVLM in individual rats from each group are shown in Fig. 2. Mean data are shown in Fig. 3. Responses to GlyR and GABAAR blockade in the RVLM were compared among the four groups by two‐way RM ANOVA, which revealed an interaction between treatment (base, strychnine and bicuculline) and group (control, NTS‐X, CVLM‐B and CVLM‐B[NTS‐X]) for both MAP (F 6,36 = 3.781, P = 0.005) and RSNA data (F 6,36 = 6.334, P < 0.001). In inactin anaesthetized rats, bilateral blockade of glycine receptors in the RVLM did not alter MAP or RSNA in control rats, whereas subsequent blockade of GABAAR in the RVLM resulted in significant increases in both MAP and RSNA (Figs. 2 A and 3 A and B). In rats that received NTS blockade (Fig. 2 B), disinhibition of the CVLM (Fig. 2 C), or both (CVLM‐B[NTS‐X]) (Fig. 2 D), strychnine in the RVLM increased MAP and RSNA, and the addition of bicuculline resulted in further pressor and sympathoexcitatory effects (Fig. 3 A and B).

Figure 2. Examples of responses to bilateral injections of strychnine and bicuculline into the RVLM.

Figure 2

Open in a new tab

Traces from individual rats show arterial pressure (AP), heart rate (HR) and integrated renal sympathetic nerve activity (int RSNA) responses to bilateral RVLM strychnine (S1 and S2) followed by bilateral RVLM bicuculline (B1 and B2). The black line in the AP traces indicates MAP. Expanded views of raw RSNA at baseline, after strychnine and after addition of bicuculline in the RVLM are shown at the bottom. A, RVLM strychnine had minimal effects on MAP and RSNA in a control rat. B, increases in MAP and RSNA as a result of RVLM strychnine were evident in a rat with prior bilateral microinjection of muscimol into the NTS (NTS‐X). C, MAP and RSNA increased as a result of RVLM strychnine in a rat with prior disinhibition of the CVLM with bicuculline (CVLM‐B). D, MAP and RSNA increased following RVLM strychnine in a rat with both CVLM‐B and NTS‐X (CVLM‐B[NTS‐X]). In rats from all groups (A–D), RVLM bicuculline after strychnine resulted in profound further increases in MAP and RSNA. HR responses were minimal in these anaesthetized rats.

Figure 3. NTS blockade and disinhibition of the CVLM unmask glycinergic inhibition of the RVLM.

Figure 3

Open in a new tab

A, mean data showing MAP in response to bilateral microinjection of strychnine followed by bicuculline (+Bic) into the RVLM of four groups of rats. Compared to preceding baseline values (base), strychnine was without effect in control rats (P = 0.187) but increased MAP in rats with prior bilateral microinjection of muscimol into the NTS (NTS‐X, P < 0.001), bicuculline into the CVLM (CVLM‐B, P = 0.018), or both (CVLM‐B[NTS‐X], P < 0.001). Bilateral microinjection of bicuculline into the RVLM at the peak response to strychnine (+Bic) resulted in further increases in MAP in all groups (P < 0.001). Within‐treatment comparisons revealed that baseline MAP was greater in CVLM‐B[NTS‐X] rats compared to rats in the control (P = 0.021) and NTS‐X (P < 0.001) groups. After RVLM strychnine, MAP was greater in CVLM‐B[NTS‐X] rats compared to the other three groups (P = 0.001 to 0.016). Following RVLM Bic, MAP was not different among the four groups. B, mean data showing RSNA (% initial baseline) in response to bilateral microinjection of strychnine followed by bicuculline into the RVLM of four groups of rats. Compared to preceding baseline values (base), strychnine was without effect in control rats (P = 0.802) but increased RSNA in rats with prior bilateral microinjection of muscimol into the NTS (NTS‐X, P < 0.001), bicuculline into the CVLM (CVLM‐B, P = 0.032), or both (CVLM‐B[NTS‐X], P = 0.002). Bilateral microinjection of bicuculline into the RVLM at the peak response to strychnine (+Bic) resulted in further increases in RSNA in all groups (P < 0.001). The number of animals is given in parentheses. Bar graphs represent the mean ± SEM and symbols represent individual data points. Treatment effects: †>base; ‡>base and strychnine; Group effects: *>NTS‐X, **>Control, NTS‐X; ***>Control, NTS‐X, and CVLM‐B (P < 0.05).

The comparisons of most interest in these experiments were responses to the blockade of RVLM GlyR and GABAAR within the groups. However, two‐way RM ANOVA also revealed group effects for MAP and, in general, MAP tended to be greater in groups with more surgical manipulation prior to RVLM injections. Although two‐way RM ANOVA on RSNA data revealed a significant interaction between group and treatment, there were no significant effects of group in the within‐treatment analysis.

To specifically evaluate the effects of manipulations in brainstem regions other than the RVLM, the effects of NTS muscimol and CVLM bicuculline were evaluated within the subgroups of animals that received these treatments. For these comparisons, RSNA was expressed as a percentage of the baseline value prior to the specific manipulation (average of 2–5 min). Muscimol was injected bilaterally into the NTS in 11 rats. Comparison of values before and after muscimol revealed that, in these anaesthetized and gallamine‐paralysed rats, NTS muscimol by itself did not change baseline MAP (+2 ± 4 mmHg, n = 11, Wilcoxon signed rank test, P = 0.21) or RSNA (+6 ± 8%, t 10 = −0.472, P = 0.434).

Recordings at the time of CVLM bicuculline injections were available in six rats without prior NTS muscimol (CVLM‐B) and five rats with CVLM‐B injections in the presence of NTS muscimol (CVLM‐B[NTS‐X]). Biphasic responses to CVLM bicuculline were compared by two‐way RM ANOVAs. For MAP, there was no effect of group (F 1,9 = 1.037, P = 0.335) and a main effect of treatment (F 2,18 = 102.6, P < 0.001), such that bicuculline in the CVLM resulted in transient decreases in MAP (P < 0.001, CVLM‐B group = −17 ± 3, CVLM‐B[NTS‐X] group = −18 ± 3 mmHg), which then stabilized at levels above baseline (P < 0.001, CVLM‐B = +12 ± 3, CVLM‐B[NTS‐X] = +18 ± 4 mmHg). For RSNA, there was an interaction between group and treatment (F 2,18 = 4.355, P = 0.029). In both the CVLM‐B (−30 ± 2%, P = 0.025) and the CVLM‐B[NTS‐X] (−38 ± 11%, P < 0.001) groups, there was a transient decrease in RSNA in response to bilateral CVLM bicuculline. In the CVLM‐B group, RSNA partially recovered and stabilized at 88 ± 8% of baseline. In the CVLM‐B[NTS‐X] group, RSNA increased to levels significantly above baseline (+22 ± 16%, P = 0.005).

During the experiments in which the NTS was blocked prior to injections into the RVLM, baroreflex inhibition of RSNA in response to a bolus i.v. injection of phenylephrine (5 μg kg−1) was tested before and 5–10 min after NTS muscimol. The ratio of %∆RSNA/∆MAP was used as an estimate of baroreflex sensitivity. After NTS muscimol, baroreflex sensitivity was significantly less after (0.01 ± 0.05) compared to before NTS‐X (−1.84 ± 0.23% RSNA mmHg−1) (paired t test, t 6 = −7.7, P = 0.001). Time control experiments in a separate group of rats (n = 6) verified that bilateral NTS muscimol eliminated baroreflex responses for a time period equivalent to the time required to complete the RVLM injection protocols following NTS muscimol (≤90 min). Compared to the control state, and consistent with attenuation of baroreflex compensation, increases in MAP as a result of repeated i.v. phenylephrine injections were greater following bilateral NTS muscimol (F 5,25 = 6.434, P < 0.001). Baroreflex‐mediated sympathoinhibition was greatly reduced for 90 min following bilateral NTS muscimol, and began to recover by 2 h (one‐way RM ANOVA) (Table 1). (RSNA, F 5,25 = 56.9, P < 0.001; ΔRSNA/ΔMAP, F 5,25 = 66.7, P < 0.001). In the current experiments, bilateral injections of bicuculline in the CVLM were performed immediately prior to RVLM injections. Previous experiments from our laboratory have verified blockade of GABAAR by bicuculline in the CVLM for a time period equivalent to that of the current experiments (Heesch et al. 2006).

Table 1.

Time course of baroreflex blockade by NTS muscimol

PE (5 μg kg−1) Con 5 min 30 min 60 min 90 min 120 min
∆MAP (mmHg) 61 ± 1 73 ± 4* 78 ± 4* 77 ± 4* 71 ± 3* 72 ± 1*
%∆ RSNA −81 ± 4 −4 ± 3†# −10 ± 5†# −10 ± 5†# −19 ± 4†# −51 ± 8†
BX ratio (%∆RSNA/∆MAP) −1.33 ± 0.08 −0.06 ± 0.04†# −0.13 ± 0.06†# −0.13 ± 0.06†# −0.28 ± 0.05†# −0.72 ± 0.12†
Open in a new tab

*>Con (P = 0.001 to 0.012); †<Con (P = 0.001); #<120 min (P = 0.001); n = 6.

GABA and glycine increased membrane conductance of RVLM neurons

The input resistance reflects all ionic currents passing through an entire cell membrane with the exception of capacity current. To demonstrate the amount of ionic current mediated by GABAAR and GlyR after application of their agonists, series of current steps were applied to PRV‐labelled RVLM neurons (Fig. 4 A–E). Bath application of GABA (100 μm) significantly decreased input resistance of RVLM neurons from 214.9 ± 31.6 to 119.4 ± 19.6 MΩ (n = 8, t 7 = 4.075, P = 0.005) (Fig. 4 B), which demonstrates increased membrane conductance as a result of activation of GABAAR. Moreover, application of GABA (100 μm) hyperpolarized RVLM neurons from −48.7 ± 3.4 mV to −56.4 ± 3.4 mV (n = 10, Wilcoxon matched‐pair signed rank test, P = 0.002) (Fig. 4 F and G).

Figure 4. GABA and glycine decreased input resistance of presympathetic RVLM neurons.

Figure 4

Open in a new tab

A, current clamp recordings at −60 mV from a presympathetic RVLM neuron demonstrate a decreased voltage deflection in response to current injection after application of GABA (100 μm). B, mean data demonstrating a decrease in input resistance after GABA application. C and D, representative current clamp recordings showing responses of presympathetic RVLM neurons to current steps in control condition (C) and after application of glycine (500 μm) (D). E, mean data illustrating a decrease in input resistance following glycine application. F, current clamp recording at resting membrane potential showing that application of GABA hyperpolarized the cell. G, mean data summarizing the effect of GABA on membrane potential of RVLM neurons. H, current clamp recording at resting membrane potential showing that application of glycine hyperpolarized the cell. I, mean data summarizing the effect of glycine on membrane potential of RVLM neurons. Numbers of replications are shown in parentheses. Bar graphs represent the mean ± SEM; open circles represent individual data points; ** P < 0.01, **** P < 0.0001.

Activation of GlyR (glycine, 500 μm) also significantly decreased whole‐cell input resistance from 297.4 ± 62.3 MΩ to 73.8 ± 16.4 MΩ (n = 9, Wilcoxon matched‐pair signed rank test, P = 0.0039) (Fig. 4 C–E). Moreover, we found that application of glycine (500 μm) hyperpolarized RVLM neurons from −45.41 ± 2.12 mV to −54.42 ± 1.71 mV (n = 12, t 11 = 6.836, P < 0.0001) (Fig. 4 H and I). These data suggest that both GABAAR and GlyR are present in the RVLM, as proposed earlier (Stornetta et al. 2004; Heesch et al. 2006). To demonstrate location of PRV‐labelled and patch clamp recorded neurons, we identified some of them using the avidin‐Texas Red reaction (Fig. 5).

Figure 5. Identification and recording from PRV‐152 labelled rat RVLM neurons.

Figure 5

Open in a new tab

A, low‐magnification of whole‐mount view of a brainstem slice (300 μm thick) after fixation revealed GFP‐labelled RVLM neurons 94 h after inoculation of the cortex of the left kidney. The recorded neuron indicated (arrow) was filled with biocytin. Dotted line indicates ventral lateral edge of the slice. B, higher magnification of the same slice in (A) containing PRV‐labelled RVLM neurons. C, same slice and plane of section viewed with optics demonstrating the biocytin label (i.e. avidin–rhodamine fluorescence). The filled neuron is indicated by the arrow. D, overlay of (B) and (C). Arrow indicate the recorded neurons. D, dorsal; V, ventral; L, lateral. E, schematic diagram showing the location of patch clamp recorded neurons. Brainstem sections were adapted from The Rat Brain in Stereotaxic Coordinates (Paxinos & Watson, 2007). The diagrams of a brainstem coronal slice showing the distribution of some recorded PRV‐labelled RVLM neurons. Amb, nucleus ambiguous; Py, pyramidal tract.

GABA and glycine generated spontaneous and miniature postsynaptic currents in RVLM neurons

In the RVLM, GABAAR and GlyR are the primary sources of IPSCs (Dun & Mo, 1989; Hayar et al. 1997; Gao & Derbenev, 2013). To demonstrate active GABAergic and glycinergic neurotransmission, PRV‐labelled RVLM neurons were voltage clamped at −10 mV and sIPSCs were recorded before and after application of bicuculline (30 μm, n = 13) (Fig. 6). After application of bicuculline, the average frequency of sIPSCs was reduced from 3.3 ± 1.3 Hz to 1.1 ± 0.5 Hz (n = 8, Wilcoxon matched‐pair signed rank test, P = 0.008) (Fig. 6 F). In five neurons, application of bicuculline caused repetitive burst sIPSCs activity; therefore, these cells were not included in the analysis.

Figure 6. GABA and glycine generated spontaneous postsynaptic currents in RVLM neurons.

Figure 6

Open in a new tab

A, continuous recording of sIPSCs from a presympathetic RVLM neuron voltage clamped at −10 mV in control condition. B, same neuron shown in (A) after application of bicuculline (Bic, 30 μm). C–E, showing the cumulative distribution of inter‐event interval (C), amplitude (D) and area (E) before and after application of bicuculline (30 μm). F–I, summarizing the effects of bicuculline on frequency (F), amplitude (G), area of events (H) and I phasic (I) of presympathetic RVLM neurons. Application of bicuculline significantly decreased the amplitude, frequency, Iphasic, and area of sIPSCs and revealed bicuculline insensitive (i.e. glycine‐mediated) sIPSCs in presympathetic RVLM neurons. Numbers of replications are shown in parentheses. Bar graphs represent the mean ± SEM; open circles represent individual data points; * P < 0.05, ** P < 0.01.

The overall amplitude of sIPSCs was significantly smaller after application of bicuculline (39.7 ± 7.9 pA vs. 29.1 ± 5.8 pA, n = 8, t 7 = 3.407, P = 0.011) (Fig. 6 G). Consistently, the average area under the sIPSCs was significantly decreased following bicuculline application from 249 ± 54 pA × ms to 150 ± 34 pA × ms (n = 8, t 7 = 3.625, P = 0.009) (Fig. 6 H). Because a decreased area under the IPSCs reflects diminished I phasic and a reduction of synaptic strength, we calculated the total inhibitory I phasic. The average inhibitory phasic current was 1.0 ± 0.43 pA before and 0.19 ± 0.08 pA after bicuculline application (n = 8, Wilcoxon matched‐pair signed rank test, P = 0.008) (Fig. 6 I). These data demonstrate that, in RVLM neurons, 24.1 ± 9.0% of the total I phasic is bicuculline insensitive. The existence of bicuculline insensitive sIPSCs suggests glycine‐mediated inhibitory neurotransmission.

mIPSCs were recorded in the presence of TTX (1 μm), a voltage‐activated sodium channel blocker, which diminished action potential‐dependent neurotransmitter release. The average frequency of mIPSCs was 0.9 ± 0.2 Hz and the addition of bicuculline suppressed the frequency of mIPSCs to 0.15 ± 0.09 Hz (n = 8, Wilcoxon matched‐pair signed rank test, P = 0.008) without altering the amplitude (data not shown). This is consistent with no changes observed in the area of mIPSCs after bicuculline application.

4‐AP application promoted release of glycine

In the RVLM, application of 4‐AP potentiated active GABAergic and glycinergic neurotransmission. 4‐AP was used to prolong the depolarization of neurons and increase their neurotransmission. An example of recordings from an RVLM neuron is shown in Fig. 7. Bath application of 4‐AP significantly increased the frequency of sIPSCs from 2.8 ± 1.9 Hz to 8.6 ± 2.6 (n = 9, Friedman, Dunn, P = 0.0005) (Fig. 8 A). To demonstrate that 4‐AP promotes the release of GABA and/or glycine, bicuculline was used to block GABAAR‐mediated synaptic currents. Bicuculline application suppressed the frequency of sIPSCs to 5.3 ± 1.8 Hz (n = 9, Friedman, Dunn, P = 0.029) (Fig. 8 A).

Figure 7. 4‐AP increased the frequency and amplitude of GlyR‐mediated synaptic currents.

Figure 7

Open in a new tab

A, continuous recording of sIPSCs from a presympathetic RVLM neuron voltage clamped at −10 mV. B, same neuron as in (A) after application of 4‐AP (2 mm). C, same neuron following application of bicuculline (30 μm). Bicuculline decreased the frequency and amplitude of sIPSCs evoked by 4‐AP application.

The amplitude of sIPSCs was also affected by 4‐AP application. In the presence of 4‐AP, the amplitude of sIPSCs was significantly increased from 30.8 ± 4.9 pA to 44.8 ± 4.9 pA (n = 9, ANOVA, F 2,16 = 5.964, Tukey, P = 0.01) and application of bicuculline significantly decreased it to 35.1 ± 6 pA (n = 9, F 2,16 = 5.964, Tukey, P = 0.039) (Fig. 8 B). Application of 4‐AP also increased the area of sIPSCs. The mean area of sIPSCs was increased from 167.7 ± 27.9 pA × ms to 385.3 ± 62.2 pA × ms after 4‐AP application (n = 9, ANOVA, F 1.512,12.1 = 6.361, Tukey, P = 0.046) (Fig. 8 C). When the GABAAR mediated component of sIPSC was blocked by bicuculline, there was no significant change in the mean area of the sIPSCs 263.5 ± 44.9 pA × ms (n = 9, ANOVA, F 1.512,12.1 = 6.361, Tukey, P = 0.176). Consistent with the increase in sIPSC area, the total inhibitory I phasic was increased by 4‐AP. In the control condition, the average I phasic was 0.59 ± 0.43 pA and increased to 3.51 ± 1.15 after 4‐AP application (n = 9, Friedman, Dunn, P = 0.001). Application of bicuculline in the presence of 4‐AP significantly decreased I phasic to 1.72 ± 0.61 pA (n = 9, Friedman, Dunn, P = 0.014) (Fig. 8 D). These results suggest that the majority of the I phasic is generated by the action potential‐dependent synaptic release of GABA and blockade of GABAAR reduces I phasic.

To evaluate whether the sIPSCs that remained after GABAAR blockade were generated by activation of GlyR, strychnine (1 μm), a competitive GlyR antagonist, was applied in the presence of 4‐AP and bicuculline. Figure 8 E shows that application of strychnine after bicuculline blocked the remaining sIPSCs in a PRV‐labelled RVLM neuron. Responses were similar in three other neurons tested, such that the addition of strychnine almost completely eliminated IPSCs (0.09 ± 0.05 Hz, n = 4).

Discussion

We identified a potentially important and previously undescribed mechanism of GlyR‐dependent inhibition of RVLM neurons and the role of GlyR in baroreflex function. Our data demonstrate that blockade of GlyR in the RVLM shortened the recovery time of RSNA following an increase in BP. In addition, a baroreflex independent glycinergic inhibition of the RVLM was evident following blockade of the NTS. Disinhibition of the CVLM unmasked glycinergic inhibition of the RVLM, which was evident with or without intact outputs from the NTS, suggesting that inputs from the CVLM to the RVLM drive glycinergic inhibition of the RVLM. In brain slice experiments, we showed that the release of glycine requires the potentiation of active synaptic inputs. In the steady‐state condition, GABAergic sIPSCs were dominant. By contrast, when active synaptic inputs were increased, glycinergic (bicuculline non‐sensitive) sIPSCs became predominant. We conclude that GABAAR in the RVLM are sufficient to provide initial and maximum baroreflex‐mediated inhibition of sympathetic outflow, whereas GlyR are required to control time course and strength of inhibition in the RVLM neurons.

Glycine and baroreflex function

It is well established that presympathetic neurons in the RVLM represent the major descending input to preganglionic sympathetic neurons in the intermediolateral cell column of the spinal cord. Thus, presympathetic neurons in the RVLM are critical for determining the level of sympathetic vasomotor tone and short‐term control of BP (Llewellyn‐Smith & Verberne, 2011). The activity of RVLM neurons depends on the intrinsic properties of the neurons and the balance of inhibitory and excitatory synaptic inputs (Bowman et al. 2013). In the RVLM, GABA and glycine are often co‐expressed in the same presynaptic terminals (Ottersen et al. 1988; Llewellyn‐Smith et al. 1995; Llewellyn‐Smith & Weaver, 2001; Llewellyn‐Smith et al. 2001; Stornetta et al. 2004), suggesting that both of these inhibitory transmitters may contribute to the control of presympathetic neurons in the RVLM. In addition, GABA and glycine share the same presynaptic transporter (Wojcik et al. 2006) and GABAAR and GlyR cluster together on the postsynaptic membrane (Levi et al. 1999; Fischer et al. 2000; Kneussel & Betz, 2000). Microinjection of GABA or glycine into the RVLM decreases SNA and BP (Guertzenstein & Silver, 1974; Dampney et al. 2003a). However, in intact animals, blockade of GABAAR but not GlyR in the RVLM increases BP and SNA (Amano & Kubo, 1993; Heesch et al. 2006). In our experiments, transient increases in BP after bolus injection of phenylephrine resulted in inhibition of RSNA followed by recovery to initial baseline levels. Strychnine in the RVLM produced no changes in baseline BP and RSNA by itself. However, following blockade of RVLM GlyR, RSNA recovered to initial baseline levels more rapidly. These data suggest that glycinergic transmission in the RVLM normally contributes to arterial baroreflex‐mediated sympathoinhibition. Additional experiments revealed that blockade of the NTS, which provides excitatory inputs to both the CVLM and RVLM (Dampney et al. 2003b; Llewellyn‐Smith & Verberne, 2011; Guyenet et al. 2013), and disinhibition of the CVLM, which comprises a major inhibitory input to the RVLM (Dampney et al. 2003b; Llewellyn‐Smith & Verberne, 2011; Guyenet et al. 2013), unmasked glycinergic inhibition in the RVLM.

We propose that the NTS and CVLM are part of the network that control glycine release in the RVLM. It is possible that inhibition of NTS neurons withdraws excitation of a pathway that normally suppresses glycine release in the RVLM. Because our data revealed that disinhibition of neurons in the CVLM alone caused tonic glycine release in the RVLM, the CVLM is a probable source of glycine in the RVLM.

GABAergic and glycinergic inhibition of RVLM neurons

Our electrophysiological data confirmed functional expression of GABAAR and GlyR on PRV‐labelled RVLM neurons. Previous studies showed that focal electrical stimulation of the RVLM triggered evoked GABAergic and glycinergic synaptic currents, and GABAAR and GlyR‐mediated sIPSCs were observed in the RVLM (Dun & Mo, 1989). It has been shown that GABAAR mediate the majority of IPSCs under steady‐state conditions (Hayar et al. 1996; Gao & Derbenev, 2013), supporting the established view of GABA as the major inhibitory neurotransmitter in the RVLM. On the other hand, the robust decrease of BP after microinjection of glycine and the existence of GlyR‐mediated IPSCs suggest that glycine plays a substantial but ill‐defined role in the control of RVLM excitability. Araujo et al. (1999) identified concentration‐dependent effects of glycine microinjected into the RVLM of conscious rats. It was proposed that microinjection of high concentrations of glycine increased BP via modulation of NMDA receptors; however, a decrease in BP was observed during microinjection of low concentrations of glycine (Araujo et al. 1999). It is possible that the increase in BP seen with high concentrations of glycine in the RVLM is associated with the diffusion of glycine to the CVLM and inhibition of RVLM‐projecting GABAergic neurons in the CVLM. The result would be diminished GABAergic inhibition of RVLM neurons followed by an increase of BP. The decrease in BP observed with a lower concentration of glycine in the RVLM would be a result of the direct effects on RVLM presympathetic neurons. In support of this interpretation, our electrophysiological data did not reveal depolarization or increased action potential firing of RVLM neurons after application of glycine (Fig. 4 H and I). Thus, our data suggest that activation of GlyR in the RVLM inhibits PRV‐labelled RVLM neurons.

We have identified the mechanism of glycinergic and GABAergic co‐inhibition in the RVLM. First, we segregated GABAergic IPSCs from glycinergic IPSCs using bicuculline and strychnine, which are selective blockers of GABAAR and GlyR, respectively. It has been reported that 10–30 μm of bicuculline is the optimal concentration to block GABAAR‐mediated currents in slices from the hypothalamus, brainstem and spinal cord (Takazawa & MacDermott, 2010; Gao & Smith, 2010b; Gao & Derbenev, 2013). Strychnine was used to block glycine‐mediated IPSCs, although the use of higher concentration of strychnine is not ideal because it also blocks GABAAR‐mediated current (Braestrup & Nielsen, 1980). Recording of sIPSCs from PRV‐labelled RVLM neurons in the presence of bicuculline revealed that a small portion of sIPSCs is generated by GlyR. The low frequency of glycinergic IPSCs is probably associated with the low probability of release of glycine from presynaptic inputs. Our experimental design did not allow us to identify and selectively stimulate inhibitory inputs containing GABA and/or glycine. To overcome this limitation, we used a non‐selective voltage‐dependent potassium channel blocker (4‐AP) to increase the release of neurotransmitters from active synapses in the RVLM. Llewellyn‐Smith & Weaver (2001) showed that, in the RVLM, GABA and glycine are often found in the same presynaptic terminals and in cell bodies. We demonstrated that PRV‐labelled RVLM neurons exhibit both GABAergic and glycinergic synaptic events. We also showed that the release of glycine from presynaptic inputs requires the potentiation of active synaptic inputs. However, it is essential to note that increased activity of synaptic inputs decreased the percentage of GABAergic and increased the percentage of glycinergic sIPSCs. As shown in Fig. 9, in the presence of TTX, 92% of IPSCs were mediated by GABAAR and only 8% were mediated by GlyR. Increased activity of synaptic inputs with 4‐AP decreased the proportion of IPSCs mediated by GABAAR (53%) and increased GlyR‐mediated IPSCs to 47%, suggesting that glycinergic inhibition is recruited by potentiation of active synaptic inputs. In summary, the results of the present study suggest that GABA is a major inhibitory neurotransmitter under steady‐state conditions, whereas release of glycine requires the potentiation of active synaptic inputs.

Physiological significance

Redundancy in CNS pathways for the inhibition of efferent sympathetic nerve activity is well established. For example, afferent inputs from the carotid sinus, aortic and cardiopulmonary stretch receptors all terminate in the NTS and follow a similar medullary path from the NTS to the CVLM, and then the RVLM. When input from one group of afferent fibres is compromised, the other sympathoinhibitory reflexes compensate to maintain tonic inhibition of the RVLM. Complex synaptic and/or dendritic interactions among various inputs contribute to the final integrated output from each of these brainstem nuclei (Llewellyn‐Smith et al. 2001; Llewellyn‐Smith & Verberne, 2011). Functionally, interactions among these feedback systems provide ‘back‐up’ for the control of sympathetic nerve activity. In addition, when these traditional sympathoinhibitory reflexes are compromised chronically, separate CNS mechanisms may emerge to limit increases in sympathetic outflow. After sinoaortic denervation or lesion of the NTS, the initial increase in arterial pressure subsides within 1–2 weeks, possibly as a result of the engagement of sympathoinhibitory pathways distinct from arterial and cardiopulmonary receptor inputs to the NTS (Schreihofer & Sved, 1992). Even within a single CNS pathway, co‐release from presynaptic nerve terminals of neuromodulators and classical transmitters may contribute to the final postsynaptic response. For example, during more intense stimulation, co‐release of peptides along with traditional fast‐acting transmitters may serve to prolong and amplify final responses (Nusbaum et al. 2017).

Our current experiments demonstrate redundancy in inhibitory transmission within the RVLM. Assessment of the functional significance of GABAergic and glycinergic co‐inhibition of neuronal circuits is challenging. In addition to activating postsynaptic receptors, neurotransmitters also modulate presynaptic receptors and affect the packaging of other neurotransmitters into synaptic vesicles (Hnasko & Edwards, 2012). Several lines of evidence suggest that the excitability of RVLM neurons is largely determined by GABA (Cravo and Morrison 1993; Schreihofer et al. 2000). Intriguingly, our results indicate that GABA and glycine probably work together to enhance the temporal resolution of inhibition. In regard to arterial baroreflex function, glycine released in the RVLM prolongs the duration of sympathoinhibition. The relative low level of glycinergic IPSCs in control conditions suggests that glycine is released on demand. Moreover, application of 4‐AP increased the frequency and amplitude of glycinergic as well as GABAergic IPSCs, increased the total inhibitory I phasic current, and thus increased the strength of inhibition. Thus, our results suggest that GABA controls the threshold excitability of presympathetic RVLM neurons, whereas glycine increases the strength of inhibition, if needed.

Methodological considerations

Our approaches have certain possible limitations. We used PRV‐152 to identify RVLM neurons. The spread of PRV‐152 is strictly retrograde and transsynaptic (Strack & Loewy, 1990) and GFP labelling was used to identify presympathetic RVLM neurons as described previously (Gao & Derbenev, 2013). However, we cannot completely exclude the possibility that some of the PRV‐152‐labelled cells were interneurons labelled via synaptic contact with RVLM presympathetic neurons. Our in vivo studies were conducted in anaesthetized rats and it is possible that the central effects of anaesthesia could have direct or indirect influence on sympathetic and cardiovascular responses. However, the interpretation of our results is strengthened by the fact that we found evidence of glycinergic inhibitory influences in the RVLM in experiments using different anaesthetic and neuromuscular blocking agents.

In the current experiments, we demonstrated that bilateral muscimol injections into the NTS produced prolonged inhibition of the arterial baroreflex, although muscimol alone did not significantly change arterial pressure or RSNA. Because pathways from the NTS mediate both excitatory and inhibitory cardiovascular responses (Llewellyn‐Smith & Verberne, 2011), this is not necessarily unexpected. However, using concentrations similar to ours, other studies have reported pressor and sympathoexcitatory responses to bilateral injections of muscimol into the NTS of anaesthetized rats (Schreihofer & Sved, 1992; Mueller & Hasser, 2006). In our experiments, rats received a constant infusion of the commonly used neuromuscular blocking agent, gallamine triethiode, and it is possible that non‐specific effects of circulating gallamine may have dampened the tonic response to muscimol in the NTS. Applied directly to nerve fibres, gallamine has been shown to affect K+ and possibly Na+ conductances (Smith & Schauf, 1981). CNS access of i.v. administered gallamine triethiodide is normally limited by the blood–brain barrier, although it is possible that surgical exposure of the NTS and possible damage to fibres in the region could have provided access to circulating gallamine. Regardless, our data demonstrating long‐term inhibition of the arterial baroreflex following NTS muscimol are consistent with generalized neuronal blockade within the NTS. It is less probable that circulating gallamine would have gained access to the RVLM, and the major focus of the the present study was the evaluation of inhibitory transmission within the RVLM.

It is recognized that microinjections into brainstem regions or bath application of drugs probably affect neurons involved in multiple modalities. However, taken together with evidence that RVLM strychnine modified the arterial baroreflex response, our data suggest that brainstem cardiovascular pathways include glycinergic neurotransmission. Our experiments used young normotensive male rats with sympathetic vasomotor drive presumably at near normal levels. Male sex was probably not a factor because previous experiments provided evidence of glycinergic inhibition in the RVLM of female rats of a similar age (Heesch et al. 2006). Future experiments will be needed to determine whether the role of glycinergic inhibition in the RVLM is changed in hypertensive animals, as well as to evaluate possible age‐related changes in plasticity.

Additional information

Competing interests

The authors declare that they have no competing interests.

Author contributions

All in vivo experiments were conducted at the Department of Biomedical Sciences, Florey Institute of Neuroscience and Mental Health, University of Melbourne and Dalton Cardiovascular Research Centre, University of Missouri. WSK and STY performed the experiments and the analysis and interpretation of data, as well as manuscript drafting, editing and revising. CMH conceived and designed the work, and was responsible for the assembly, analysis and interpretation of data, as well as manuscript drafting, editing and revising. All of the in vitro experiments were conducted at the Department of Physiology, School of Medicine, Tulane University. HG performed the experiments and the assembly and analysis of data, as well as manuscript drafting, editing and revising. AVD conceived and designed the work, and was responsible for the assembly, analysis and interpretation of data, as well as manuscript drafting, editing and revising. All authors approved the final version of the manuscript submitted for publication, and agree to be accountable for all aspects of the work. All qualified authors are listed.

Funding

This work was supported by National Institute of Health R01 HL122829 (A. V. Derbenev) and R01 HL098602 (C. M. Heesch); Australian Research Council Future Fellowship FT170100363 and the National Health and Medical Research Council of Australia GNT1079680 (S. T. Yao), High Blood Pressure Research Council of Australia; the Rebecca L Cooper Medical Foundation; and the Victorian Government through the Operational Infrastructure Scheme (W. S. Korim). PRV‐152 was obtained from the Centre for Neuroanatomy with Neurotropic Viruses (P40 OD10996).

Acknowledgements

We thank Mr J. Glenn Phaup (Dalton Cardiovascular Research Centre, University of Missouri) and Dr Charles D. Nichols (Department of Pharmacology and Experimental therapeutics, Louisiana State University) for their excellent technical assistance.

Biography

Hong Gao received her PhD from the Department of Cell and Molecular Biology at Tulane University in 2008 where she studied the inhibitory regulation of neurons of the rat dorsal motor nucleus of vagus. Currently, she works with Dr Andrei V. Derbenev as a research scientist at Tulane University School of Medicine in New Orleans, LA, USA. Her studies focus on synaptic regulation of neurons in the rostral ventrolateral medulla and the paraventricular nucleus of the hypothalamus, which are brain regions critical for regulating sympathetic output.

graphic file with name TJP-597-283-g001.gif

Edited by: Ian Forsythe & Gregory Funk

References

  1. Amano M & Kubo T (1993). Involvement of both GABAA and GABAB receptors in tonic inhibitory control of blood pressure at the rostral ventrolateral medulla of the rat. Naunyn Schmiedebergs Arch Pharmacol 348, 146–153. [DOI] [PubMed] [Google Scholar]
  2. Araujo GC, Lopes OU & Campos RR (1999). Importance of glycinergic and glutamatergic synapses within the rostral ventrolateral medulla for blood pressure regulation in conscious rats. Hypertension 34, 752–755. [DOI] [PubMed] [Google Scholar]
  3. Blessing WW (1988). Depressor neurons in rabbit caudal medulla act via GABA receptors in rostral medulla. Am J Physiol Heart Circ Physiol 254, H686–H692. [DOI] [PubMed] [Google Scholar]
  4. Bowman BR, Kumar NN, Hassan SF, McMullan S & Goodchild AK (2013). Brain sources of inhibitory input to the rat rostral ventrolateral medulla. J Comp Neurol 521, 213–232. [DOI] [PubMed] [Google Scholar]
  5. Boychuk CR & Smith BN (2016). Glutamatergic drive facilitates synaptic inhibition of dorsal vagal motor neurons after experimentally induced diabetes in mice. J Neurophysiol 116, 1498–1506. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Braestrup C & Nielsen M (1980). Strychnine as a potent inhibitor of the brain GABA‐benzodiazepine receptor complex. Brain Res Bull 5, 681–684. [Google Scholar]
  7. Chan RK & Sawchenko PE (1998). Organization and transmitter specificity of medullary neurons activated by sustained hypertension: implications for understanding baroreceptor reflex circuitry. J Neurosci 18, 371–387. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Cravo SL & Morrison SF (1993). The caudal ventrolateral medulla is a source of tonic sympathoinhibition. Brain Res 621, 133–136. [DOI] [PubMed] [Google Scholar]
  9. Dampney RA, Horiuchi J, Tagawa T, Fontes MA, Potts PD & Polson JW (2003a). Medullary and supramedullary mechanisms regulating sympathetic vasomotor tone. Acta Physiol Scand 177, 209–218. [DOI] [PubMed] [Google Scholar]
  10. Dampney RA, Polson JW, Potts PD, Hirooka Y & Horiuchi J (2003b). Functional organization of brain pathways subserving the baroreceptor reflex: studies in conscious animals using immediate early gene expression. Cell Mol Neurobiol 23, 597–616. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Derbenev AV, Stuart TC & Smith BN (2004). Cannabinoids suppress synaptic input to neurones of the rat dorsal motor nucleus of the vagus nerve. J Physiol 559, 923–938. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Deuchars SA, Spyer KM & Gilbey MP (1997). Stimulation within the rostral ventrolateral medulla can evoke monosynaptic GABAergic IPSPs in sympathetic preganglionic neurons in vitro. J Neurophysiol 77, 229–235. [DOI] [PubMed] [Google Scholar]
  13. Dufour A, Tell F, Kessler JP & Baude A (2010). Mixed GABA‐glycine synapses delineate a specific topography in the nucleus tractus solitarii of adult rat. J Physiol 588, 1097–1115. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Dun NJ & Mo N (1989). Inhibitory postsynaptic potentials in neonatal rat sympathetic preganglionic neurones in vitro. J Physiol 410, 267–281. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Fischer F, Kneussel M, Tintrup H, Haverkamp S, Rauen T, Betz H & Wassle H (2000). Reduced synaptic clustering of GABA and glycine receptors in the retina of the gephyrin null mutant mouse. J Comp Neurol 427, 634–648. [DOI] [PubMed] [Google Scholar]
  16. Gao H & Derbenev AV (2013). Synaptic and extrasynaptic transmission of kidney‐related neurons in the rostral ventrolateral medulla. J Neurophysiol 110, 2637–2647. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Gao H & Smith BN (2010a). Tonic GABAA receptor‐mediated inhibition in the rat dorsal motor nucleus of the vagus. J Neurophysiol 103, 904–914. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Gao H & Smith BN (2010b). Zolpidem modulation of phasic and tonic GABA currents in the rat dorsal motor nucleus of the vagus. Neuropharmacology 58, 1220–1227. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Grundy D (2015). Principles and standards for reporting animal experiments in The Journal of Physiology and Experimental Physiology. J Physiol 593, 2547–2549. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Guertzenstein PG & Silver A (1974). Fall in blood pressure produced from discrete regions of the ventral surface of the medulla by glycine and lesions. J Physiol 242, 489–503. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Guyenet PG, Darnall RA & Riley TA (1990). Rostral ventrolateral medulla and sympathorespiratory integration in rats. Am J Physiol Regul Integr Comp Physiol 259, R1063–R1074. [DOI] [PubMed] [Google Scholar]
  22. Guyenet PG, Stornetta RL, Bochorishvili G, Depuy SD, Burke PG & Abbott SB (2013). C1 neurons: the body's EMTs. Am J Physiol Regul Integr Comp Physiol 305, R187–R204. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Hayar A, Feltz P & Piguet P (1997). Adrenergic responses in silent and putative inhibitory pacemaker‐like neurons of the rat rostral ventrolateral medulla in vitro. Neuroscience 77, 199–217. [DOI] [PubMed] [Google Scholar]
  24. Hayar A, Piguet P & Feltz P (1996). GABA‐induced responses in electrophysiologically characterized neurons within the rat rostro‐ventrolateral medulla in vitro. Brain Res 709, 173–183. [DOI] [PubMed] [Google Scholar]
  25. Heesch CM, Laiprasert JD & Kvochina L (2006). RVLM glycine receptors mediate GABAA and GABAB independent sympathoinhibition from CVLM in rats. Brain Res 1125, 46–59. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Hnasko TS & Edwards RH (2012). Neurotransmitter corelease: mechanism and physiological role. Annu Rev Physiol 74, 225–243. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Jonas P, Bischofberger J & Sandkuhler J (1998). Corelease of two fast neurotransmitters at a central synapse. Science 281, 419–424. [DOI] [PubMed] [Google Scholar]
  28. Kneussel M & Betz H (2000). Receptors, gephyrin and gephyrin‐associated proteins: novel insights into the assembly of inhibitory postsynaptic membrane specializations. J Physiol 525, 1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Korim WS, Bou Farah L, McMullan S & Verberne AJ (2014). Orexinergic activation of medullary premotor neurons modulates the adrenal sympathoexcitation to hypothalamic glucoprivation. Diabetes 63, 1895–1906. [DOI] [PubMed] [Google Scholar]
  30. Korim WS, Egwuenu E, Fong AY, McMullan S, Cravo SL & Pilowsky PM (2012). Noxious somatic stimuli diminish respiratory‐sympathetic coupling by selective resetting of the respiratory rhythm in anaesthetized rats. Exp Physiol 97, 1093–1104. [DOI] [PubMed] [Google Scholar]
  31. Levi S, Chesnoy‐Marchais D., Sieghart W & Triller A (1999). Synaptic control of glycine and GABA(A) receptors and gephyrin expression in cultured motoneurons. J Neurosci 19, 7434–7449. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Liu Q & Wong‐Riley MT (2013). Postnatal development of brain‐derived neurotrophic factor (BDNF) and tyrosine protein kinase B (TrkB) receptor immunoreactivity in multiple brain stem respiratory‐related nuclei of the rat. J Comp Neurol 521, 109–129. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Llewellyn‐Smith IJ, Minson JB, Pilowsky PM, Arnolda LF & Chalmers JP (1995). The one hundred percent hypothesis: glutamate or GABA in synapses on sympathetic preganglionic neurons. Clin Exp Hypertens 17, 323–333. [DOI] [PubMed] [Google Scholar]
  34. Llewellyn‐Smith IJ, Schreihofer AM & Guyenet PG (2001). Distribution and amino acid content of enkephalin‐immunoreactive inputs onto juxtacellularly labelled bulbospinal barosensitive neurons in rat rostral ventrolateral medulla. Neuroscience 108, 307–322. [DOI] [PubMed] [Google Scholar]
  35. Llewellyn‐Smith IJ & Verberne AJM (2011). Central Regulation of Autonomic Functions. Oxford University Press, New York. [Google Scholar]
  36. Llewellyn‐Smith IJ & Weaver LC (2001). Changes in synaptic inputs to sympathetic preganglionic neurons after spinal cord injury. J Comp Neurol 435, 226–240. [DOI] [PubMed] [Google Scholar]
  37. McMenamin CA, Anselmi L, Travagli RA & Browning KN (2016). Developmental regulation of inhibitory synaptic currents in the dorsal motor nucleus of the vagus in the rat. J Neurophysiol 116, 1705–1714. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Mueller PJ & Hasser EM (2006). Putative role of the NTS in alterations in neural control of the circulation following exercise training in rats. Am J Physiol Regul Integr Comp Physiol 290, R383–R392. [DOI] [PubMed] [Google Scholar]
  39. Nabekura J, Katsurabayashi S, Kakazu Y, Shibata S, Matsubara A, Jinno S, Mizoguchi Y, Sasaki A & Ishibashi H (2004). Developmental switch from GABA to glycine release in single central synaptic terminals. Nat Neurosci 7, 17–23. [DOI] [PubMed] [Google Scholar]
  40. Nusbaum MP, Blitz DM & Marder E (2017). Functional consequences of neuropeptide and small‐molecule co‐transmission. Nat Rev Neurosci 18, 389–403. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. O'Brien JA & Berger AJ (1999). Cotransmission of GABA and glycine to brain stem motoneurons. J Neurophysiol 82, 1638–1641. [DOI] [PubMed] [Google Scholar]
  42. Ottersen OP, Storm‐Mathisen J & Somogyi P (1988). Colocalization of glycine‐like and GABA‐like immunoreactivities in Golgi cell terminals in the rat cerebellum: a postembedding light and electron microscopic study. Brain Res 450, 342–353. [DOI] [PubMed] [Google Scholar]
  43. Park JB, Skalska S & Stern JE (2006). Characterization of a novel tonic gamma‐aminobutyric acidA receptor‐mediated inhibition in magnocellular neurosecretory neurons and its modulation by glia. Endocrinology 147, 3746–3760. [DOI] [PubMed] [Google Scholar]
  44. Paxinos G & Watson C (2007). The Rat Brain in Stereotaxic Coordinates. Academic Press/Elsevier, Amsterdam. [Google Scholar]
  45. Rahman J, Latal AT, Besser S, Hirrlinger J & Hulsmann S (2013). Mixed miniature postsynaptic currents resulting from co‐release of glycine and GABA recorded from glycinergic neurons in the neonatal respiratory network. Eur J Neurosci 37, 1229–1241. [DOI] [PubMed] [Google Scholar]
  46. Russier M, Kopysova IL, Ankri N, Ferrand N & Debanne D (2002). GABA and glycine co‐release optimizes functional inhibition in rat brainstem motoneurons in vitro. J Physiol 541, 123–137. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Schreihofer AM, Stornetta RL & Guyenet PG (2000). Regulation of sympathetic tone and arterial pressure by rostral ventrolateral medulla after depletion of C1 cells in rat. J Physiol 529, 221–236. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Schreihofer AM & Guyenet PG (1997). Identification of C1 presympathetic neurons in rat rostral ventrolateral medulla by juxtacellular labeling in vivo. J Comp Neurol 387, 524–536. [DOI] [PubMed] [Google Scholar]
  49. Schreihofer AM & Sved AF (1992). Nucleus tractus solitarius and control of blood pressure in chronic sinoaortic denervated rats. Am J Physiol Regul Integr Comp Physiol 263, R258–R266. [DOI] [PubMed] [Google Scholar]
  50. Smith KJ & Schauf CL (1981). Effects of gallamine triethiodide on membrane currents in amphibian and mammalian peripheral nerve. J Pharmacol Exp Ther 217, 719–726. [PubMed] [Google Scholar]
  51. Stornetta RL, McQuiston TJ & Guyenet PG (2004). GABAergic and glycinergic presympathetic neurons of rat medulla oblongata identified by retrograde transport of pseudorabies virus and in situ hybridization. J Comp Neurol 479, 257–270. [DOI] [PubMed] [Google Scholar]
  52. Strack AM & Loewy AD (1990). Pseudorabies virus: a highly specific transneuronal cell body marker in the sympathetic nervous system. J Neurosci 10, 2139–2147. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Sun MK & Guyenet PG (1985). GABA‐mediated baroreceptor inhibition of reticulospinal neurons. Am J Physiol Regul Integr Comp Physiol 249, R672–R680. [DOI] [PubMed] [Google Scholar]
  54. Takazawa T & MacDermott AB (2010). Glycinergic and GABAergic tonic inhibition fine tune inhibitory control in regionally distinct subpopulations of dorsal horn neurons. J Physiol 588, 2571–2587. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Thompson SH (1977). Three pharmacologically distinct potassium channels in molluscan neurones. J Physiol 265, 465–488. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Wojcik SM, Katsurabayashi S, Guillemin I, Friauf E, Rosenmund C, Brose N & Rhee JS (2006). A shared vesicular carrier allows synaptic corelease of GABA and glycine. Neuron 50, 575–587. [DOI] [PubMed] [Google Scholar]
  57. Xu H & Smith BN (2015). Presynaptic ionotropic glutamate receptors modulate GABA release in the mouse dorsal motor nucleus of the vagus. Neuroscience 308, 95–105. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Yao Y, Hildreth CM, Farnham MM, Saha M, Sun QJ, Pilowsky PM & Phillips JK (2015). The effect of losartan on differential reflex control of sympathetic nerve activity in chronic kidney disease. J Hypertens 33, 1249–1260. [DOI] [PubMed] [Google Scholar]

Articles from The Journal of Physiology are provided here courtesy of The Physiological Society

RESOURCES