Address correspondence and reprint requests to Dr. Hui-Lin Pan, Department of Anesthesiology and Perioperative Medicine, The University of Texas MD Anderson Cancer Center, 1515 Holcombe Blvd., Unit 110, Houston, TX 77030-4009, USA. E-mail: firstname.lastname@example.org
Barrington's nucleus (BN), commonly known as the pontine micturition center, controls micturition and other visceral functions through projections to the spinal cord. In this study, we developed a rat brain slice preparation to determine the intrinsic and synaptic mechanisms regulating pre-sympathetic output (PSO) and pre-parasympathetic output (PPO) neurons in the BN using patch-clamp recordings. The PSO and PPO neurons were retrogradely labeled by injecting fluorescent tracers into the intermediolateral region of the spinal cord at T13-L1 and S1-S2 levels, respectively. There were significantly more PPO than PSO neurons within the BN. The basal activity and membrane potential were significantly lower in PPO than in PSO neurons, and A-type K+ currents were significantly larger in PPO than in PSO neurons. Blocking A-type K+ channels increased the excitability more in PPO than in PSO neurons. Stimulting μ-opioid receptors inhibited firing in both PPO and PSO neurons. The glutamatergic EPSC frequency was much lower, whereas the glycinergic IPSC frequency was much higher, in PPO than in PSO neurons. Although blocking GABAA receptors increased the excitability of both PSO and PPO neurons, blocking glycine receptors increased the firing activity of PPO neurons only. Furthermore, blocking ionotropic glutamate receptors decreased the excitability of PSO neurons but paradoxically increased the firing activity of PPO neurons by reducing glycinergic input. Our findings indicate that the membrane and synaptic properties of PSO and PPO neurons in the BN are distinctly different. This information improves our understanding of the neural circuitry and central mechanisms regulating the bladder and other visceral organs.
The micturition reflex is regulated by a network of central neurons involving the Barrington's nucleus (BN). The BN is located bilaterally in the pontine tegumentum, ventromedial to the rostral pole of the locus coeruleus (Rizvi et al. 1994). Although the BN is likely involved in regulating various functions of abdominal visceral organs, it is commonly referred to as the pontine micturition center because of its well-documented role in the control of urinary bladder function (Valentino et al. 1996; Sasaki 2002, 2005a, b). In this regard, electrical stimulation of the BN elicits bladder contraction, inhibition of external urethral sphincter, and relaxation of the urethra in cats (Holstege et al. 1986; Sugaya et al. 1987; Noto et al. 1989), whereas bilateral lesions of the BN inhibit micturition in rats (Satoh et al. 1978). In contrast to the excitatory influence of parasympathetic preganglionic neurons on the micturition reflex, increased sympathetic activity inhibits this reflex (de Groat and Saum 1972; de Groat et al. 1993; Shefchyk 2002). Although it is well known that the BN regulates the micturition reflex by inducing the bladder detrusor muscle to contract through parasympathetic efferent nerves, it is not clear whether the BN can influence the bladder function through a direct connection to sympathetic preganglionic neurons in the spinal cord.
The BN contains at least two types of micturition-related neurons that display opposing discharge activities during the micturition contraction–relaxation cycle (Tanaka et al. 2003; Sasaki 2004, 2005a). ‘Direct’ neurons are involved in the initiation of micturition and are the source of bulbospinal axons carrying excitatory signals to parasympathetic nerves. ‘Direct’ neurons increase their firing activity during micturition contraction but are silent in the absence of a voiding reflex (de Groat et al. 1998; Tanaka et al. 2003; Sasaki 2004, 2005a). Anatomical studies suggest that these BN neurons send descending axons to parasympathetic preganglionic neurons located in the lower lumbar and sacral spinal cord to regulate bladder contraction in cats (Nadelhaft et al. 1980; Holstege et al. 1986; Blok and Holstege 1997; Sasaki 2002). On the other hand, ‘inverse’ neurons are inhibited during bladder contractions but are active during the intervals between bladder contractions in cats and rats (Sasaki 2002, 2005a; Tanaka et al. 2003). In contrast to spinally projecting ‘direct’ neurons, only a few ‘inverse’ neurons in the BN project to the spinal cord (Sasaki 2005b). Virtually nothing is known about how these two types of BN output neurons are regulated intrinsically and synaptically. The use of various anesthetics and the difficulty in identifying specific projection neurons in vivo are major confounding factors in characterizing the functional properties of the various types of output neurons in the BN.
In this study, we first identified pre-sympathetic output (PSO) and pre-parasympathetic output (PPO) neurons in the BN of rats by injecting fluorescent tracers into the intermediolateral region of the spinal cord at T13-L1 and S1-S2 levels, respectively. We then determined the membrane and synaptic properties of the PPO and PSO neurons using whole-cell patch-clamp recordings in anesthetic-free brain slices in vitro. Our study provides novel information that PPO and PSO neurons in the BN have different membrane properties and are differentially regulated by excitatory and inhibitory synaptic inputs.
Materials and methods
Retrograde labeling of PSO and PPO neurons in the BN
Male Sprague-Dawley rats (200–300 g) were purchased from Harlan Laboratories (Indianapolis, IN, USA). The surgical procedures and experimental protocols were approved by the Institutional Animal Care and Use Committee and conformed to the National Institutes of Health's guidelines on the ethical use of animals. We used retrograde labeling to identify PSO and PPO neurons in the BN that project to the thoracolumbar and sacral spinal cord, respectively. Rats were anesthetized using 2–3% isoflurane in O2, and a limited dorsal laminectomy was performed to expose the spinal cord at T12-L1 and S1-S2 levels. The red fluorescent microspheres (wavelength: 580 nm, 0.04 μm; Invitrogen, Eugene, OR, USA) was pressure-ejected bilaterally into the intermediolateral region of the spinal cord at S1-S2 levels (~ 200 μm from the midline and ~ 200 μm below the dorsolateral sulcus) using a glass pipette in three to four separate 50-nL injections (Li and Pan 2006; Li et al. 2010). The green fluorescent microspheres (wavelength: 505 nm, 0.04 μm, Invitrogen) were injected into the intermediolateral region of the spinal cord at T13 to L1 levels (Fig. 1a). The rat was returned to its cage for 7–9 days to permit the fluorescent tracers to be transported to the BN. After microinjection, rats were treated subcutaneously with an antibiotic (enrofloxacin; at 5 mg/kg, every 24 h for 3 days) and an analgesic (buprenorphine; at 0.5 mg/kg, every 12 h for 2 days).
To determine the distribution of PPO and PSO neurons in the BN, three rats were anesthetized with sodium pentobarbital (60 mg/kg, intraperitoneally) and were intracardially perfused with 200 mL of ice-cold normal saline containing 1000 units of heparin followed by 500 mL of 4% paraformaldehyde in 0.1 M PBS (pH 7.4) and 200 ml of 10% sucrose in 0.1 M PBS (pH 7.4). The brain was removed quickly and post-fixed for 2 h in the same fixative solution and cryoprotected in 30% sucrose in PBS for 48 h at 4°C. Then, the brainstem sections were cut to a thickness of 30 μm and mounted onto slides, dried, and coverslipped. The sections were viewed with use of a confocal microscope (Carl Zeiss, Jena, Germany) and the areas of interest were photo-documented.
Preparation of brain slices
Brainstem slices were prepared from fluorescent microsphere-injected rats. Briefly, the rats were anesthetized with 2% isoflurane and decapitated, and the brain was quickly removed and placed in ice-cold artificial cerebrospinal fluid (aCSF) saturated with a mixture of 95% O2 and 5% CO2. The aCSF solution contained (in mM) 124.0 NaCl, 3.0 KCl, 1.3 MgSO4, 2.4 CaCl2, 1.4 NaH2PO4, 10.0 glucose, and 26.0 NaHCO3. A tissue block containing the BN (stereotaxic coordinates: 9.16 and 10.04 mm caudal to the Bregma) was glued onto the stage of a vibrating microtome, and coronal slices (300 μm thick) were cut. The slices were then transferred into an incubation chamber containing aCSF continuously gassed with a mixture of 95% O2 and 5% CO2 at 34°C for at least 1 h before the electrophysiological experiments were performed. In addition, the spinal cord segments at the levels of trace injections were also removed to confirm histologically that the fluorescent tracers were limited to the vicinity of the spinal intermediolateral region (Fig. 1a).
A total of 54 rats was used for the electrophysiological recordings. Only one neuron was recorded in each brain slice, and two to three neurons were recorded per animal. Under an upright microscope, the BN is an elliptic translucent area in the lateral central gray, located medially to the locus coeruleus (Fig. 1b). Labeled PPO and PSO neurons in the BN were briefly identified with fluorescence illumination and then visualized using differential interference contrast-infrared optics (Fig. 1c). The recording electrode was pulled from borosilicate capillaries (1.2-mm outer diameter, 0.68-mm inner diameter) with use of a micropipette puller. The resistance of the pipette was 3-7 MΩ when it was filled with internal solution containing (in mM) 140.0 K+ gluconate, 2.0 MgCl2, 0.1 CaCl2, 10.0 HEPES, 1.1 EGTA, 0.3 Na2-GTP, and 2.0 Na2-ATP, adjusted to pH 7.25 with 1 M KOH (270 to 290 mOsm). The recording chamber was continuously perfused (3 mL/min) with aCSF saturated with 95% O2 and 5% CO2 at 34°C, which was maintained by an in-line solution heater. The volume of solution that was needed to fill the recording chamber was approximately 1.0 mL.
The firing activity of labeled neurons was recorded with use of a whole-cell current-clamp configuration (Li et al. 2004, 2012). Recording of the firing activity of labeled BN neurons began about 5 min after the whole-cell access to the neuron was established and the firing activity reached a steady state. The voltage-activated K+ channel currents were recorded as described previously (Sonner and Stern 2007). We used the aCSF with nominal Ca2+ (0 mM) containing (in mM) 102 NaCl, 2.5 KCl, 3 MgSO4, 26 NaHCO3, 1.25 NaH2PO4, 20 D-glucose, 0.4 ascorbic acid, 2 pyruvic acid, 3 EGTA, 0.2 CdCl2, and 30 tetraethylammonium (pH 7.4, 290–310 mosM). For recording of A-type K+ currents, Ca2+ currents were eliminated using a nominal Ca2+ aCSF containing EGTA and CdCl2. Tetrodotoxin (TTX, 0.5 μM) and tetraethylammonium were used to block voltage-activated Na+ channels and delayed rectifier K+ channels, respectively. Two separate electrophysiological protocols were used to isolate voltage-activated A-type K+ channel currents (Sonner and Stern 2007). The first one consisted a hyperpolarized conditioning pulse (−90 mV), which minimized inactivation of A-type K+ currents, followed by depolarizing command pulses (−70 to +15 mV), which resulted in activation of both A-type K+ currents and delayed rectifier K+ currents. A second protocol used a more depolarized (−40 mV) conditioning pulse to completely inactivate A-type K+ currents. Currents recorded under these two protocols were then subtracted offline to obtain A-type K+ currents.
The GABAergic spontaneous inhibitory post-synaptic currents (sIPSCs) were recorded in the presence of 50 μM (2R)-amino-5-phosphonovaleric acid (AP-5), 20 μM 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX), and 2 μM strychnine at a holding potential of 0 mV, whereas the glycinergic sIPSCs were recorded in the presence of 50 μM AP-5, 20 μM CNQX, and 20 μM bicuculline at a holding potentials of 0 mV (Jin et al. 2011; Zhou et al. 2012). The spontaneous excitatory postsynaptic currents (sEPSCs) were recorded in the presence of 20 μM bicuculline and 2 μM strychnine at a holding potential of −60 mV (Li et al. 2004). The input resistance was monitored, and the recording was abandoned if it changed more than 15%. Signals were processed with use of a Multiclamp 700B amplifier (Molecular Devices, Foster City, CA, USA), filtered at 1–2 kHz, and digitized at 20 kHz.
The drugs were freshly prepared and delivered using syringe pumps at the final concentrations indicated. AP-5, CNQX, and bicuculline were obtained from Tocris Bioscience (Ellisville, MO, USA). Strychnine, 4-aminopyridine (4-AP), and [D-Ala2,N-MePhe4,Gly-ol]-enkephalin (DAMGO) were purchased from Sigma-Aldrich (St. Louis, MO, USA).
Data are presented as means ± SEM. The input resistance of neurons was determined by quantifying changes in membrane potentials produced by episodic injection of incremental quantities of negative current (500 ms duration). The amplitude of action potentials was measured from the beginning of the fastest rising phase to the peak. Half-width was defined as the duration between the rising and decay phases measured at the level of 50% amplitude (Vydyanathan et al. 2005). sIPSCs and sEPSCs were detected by setting a threshold above the noise level and analyzed with use of MiniAnalysis software (Synaptosoft Inc., Leonia, NJ, USA). Unpaired Student's t-test was used to compare the difference in the electrophysiological parameters between PPO and PSO neurons. The effects of drugs on the firing rate and the amplitude and frequency of sIPSCs and sEPSCs were determined using the non-parametric Wilcoxon signed–rank test or Kruskal–Wallis test with Dunn's post hoc test. p < 0.05 was considered statistically significant.
Distribution of PPO and PSO neurons in the BN
The BN contains a cluster of spinally projecting neurons located ventral to the fourth ventricle between 9.16 and 10.04 mm caudal to the Bregma. Confocal images show that labeled PPO and PSO neurons in the BN were distributed through the rostra-caudal levels of the BN (Fig. 1d). To compare the relative distribution of PPO and PSO neurons in the BN, we randomly selected three brain sections per animal. The number of PPO neurons (8.2 ± 2.1 cells/optical section) was significantly greater than that of PSO neurons (2.7 ± 1.6 cells/optical section) within the BN. Among a total of 74 PPO neurons labeled by the red tracer and 27 PSO neurons labeled by the green tracer, only four neurons were labeled by both tracers (Fig. 1d).
Membrane properties of PPO and PSO neurons in the BN
The resting membrane potentials of PPO neurons were significantly hyperpolarized (−50.9 ± 2.4 mV, n = 20) compared with that of PSO neurons (−43.3 ± 2.2 mV, n = 16; p < 0.05, Fig. 2a–c). However, the PPO and PSO neurons had similar input resistance (689.6 ± 56.8 vs. 765.3 ± 38.9 MΩ, p > 0.05) and cell capacitance (45.6 ± 5.4 vs. 48.5 ± 4.6 pF; p > 0.05). Most PPO neurons (18 of 20, 90%) and PSO neurons (15 of 16, 93.8%) displayed spontaneous firing activity. The basal firing rate was significantly lower in PPO neurons than in PSO neurons (Fig. 2b). Also, the amplitude of action potentials was significantly larger in PPO neurons than in PSO neurons (Fig. 2d), and the threshold of action potentials was significantly higher in PPO neurons than in PSO neurons (Fig. 2e). There were no significant differences in the action potential half-width between PSO and PPO neurons (2.0 ± 0.1 vs. 1.95 ± 0.1 ms).
We also compared the responses of PPO and PSO neurons to injection of depolarizing currents. The membrane was first hyperpolarized to −90 mV, and then incremental steps of depolarizing currents were applied for a duration of 500 ms (currents injected from 0 to +100 pA in an incremental step of 20 pA and interval of 1 s). Stepwise injection of depolarizing currents elicited fewer action potentials in PPO neurons than in PSO neurons at each level of current injection. The PPO neurons (77.2 ± 13.7 ms, n = 15) displayed a longer delayed onset latency of the action potential compared with PSO neurons (42.8 ± 7.3 ms, n = 14, p < 0.05%; Fig. 3a). Furthermore, PPO neurons (62.6 ± 6.7 ms) showed a delayed return of the membrane potential to baseline after the repolarizing pulse compared with that in PSO neurons (41.0 ± 4.5 ms, p < 0.05%; Fig. 3a).
Difference in voltage-activated K+ channel currents between PPO and PSO neurons in the BN
Because the delayed onset of firing activity and return of the membrane potentials to baseline suggest the presence of a transient outward current such as the A-type K+ current (Tasker and Dudek 1991; Vydyanathan et al. 2005), it is possible that more A-type K+ channel currents are present in PPO neurons than in PSO neurons. We first determined the effect of blocking A-type K+ channels with 4-AP on the firing activity of labeled PPO and PSO neurons in the BN. Because 4-AP not only blocks voltage-activated K+ channels but also stimulates voltage-activated Ca2+ channels (Wu et al. 2009), we examined the effect of 4-AP on the firing activity in the presence of Cd2+, a voltage-activated Ca2+ channel blocker. Cd2+ (50 μM) had no significant effect on the firing rate of either PPO or PSO neurons. In the presence of Cd2+, bath application of 5 mM 4-AP significantly reduced the onset latency and increased the firing rate of 8 PPO neurons and 9 PSO neurons (Fig. 3b and c). The magnitude of increase in the firing rate was significantly greater in PPO neurons than in PSO neurons (Fig. 3d).
We next recorded the total voltage-activated K+ channel currents in both PPO and PSO neurons after blocking voltage-activated Na+ and Ca2+ channels using TTX and Cd2+, respectively. The peak amplitude of the total K+ currents was significantly greater in 7 PPO neurons than in 8 PSO neurons at holding potentials from −30 to +15 mV (Fig. 4). Furthermore, A-type K+ currents, characterized as having strong voltage dependency and rapid activation and inactivation kinetics, were present in both PPO and PSO neurons. However, the peak amplitude of A-type K+ currents was significantly larger in 7 PPO neurons than in 8 PSO neurons tested (Fig. 4).
Differences in the control of the excitability of PPO and PSO neurons by excitatory and inhibitory synaptic inputs in the BN
We first determined the influence of GABAergic input on the excitability of PPO and PSO neurons in the BN. Bath application of 20 μM bicuculline, a selective GABAA receptor antagonist, similarly increased the firing rate of 8 PSO neurons and 9 PPO neurons (Fig. 5a and b).
We then assessed the role of glycinergic input in the control of the excitability of PPO and PSO neurons. Bath application of 2 μM strychnine, a specific glycine receptor antagonist, significantly increased the firing activity of 8 PPO neurons tested (Fig. 5c and d). However, strychnine did not significantly change the firing activity of 7 PSO neurons examined (Fig. 5c and d).
We also determined the role of glutamatergic input in the control of the firing rate of PSO and PPO neurons. Blocking AMPA and NMDA receptors with specific antagonists (20 μM CNQX and 50 μM AP-5, respectively) significantly decreased the firing rate of PSO neurons (n = 7, Fig. 5e and f). Strikingly, CNQX and AP-5 caused a significant increase in the firing rate of PPO neurons (n = 7, Fig. 5e and f).
Our results above (Fig. 5c and d) indicate that unlike PSO neurons, PPO neurons are tonically inhibited by glycinergic input. It is possible that blocking ionotropic glutamate receptors may reduce glycinergic inputs to PPO neurons and indirectly disinhibit PPO neurons. To test this hypothesis, we determined the effect of CNQX and AP-5 on the firing activity of 8 PPO and 7 PSO neurons after blocking the glycine receptor with strychnine. Strychnine alone increased the firing rate of PPO neurons but did not significantly affect the firing of PSO neurons. Bath application of CNQX and AP-5 failed to significantly alter the firing rate of PPO neurons in the presence of strychnine, although it significantly reduced the firing activity of PSO neurons (Fig. 5g and h).
Differences in the excitatory and inhibitory synaptic inputs between PPO and PSO neurons in the BN
We determined whether there is a difference in the excitatory glutamatergic input to PSO and PPO neurons in the BN by recording glutamatergic sEPSCs. The frequency of sEPSCs was much lower in PPO neurons (n = 12) than in PSO neurons (n = 15, Fig. 6). However, the amplitude of sEPSCs did not differ significantly between PPO and PSO neurons (Fig. 6). Bath application of 20 μM CNQX abolished sEPSCs of all the PPO and PSO neurons examined at the end of the experiments.
To assess whether PPO and PSO neurons receive different amounts of inhibitory GABAergic and glycinergic inhibitory inputs, we recorded GABAergic and glycinergic sIPSCs in labeled PPO and PSO neurons in the BN. The baseline frequency and amplitude of GABAergic sIPSCs did not differ significantly between 9 PPO and 8 PSO neurons (Fig. 7a–c). However, the baseline frequency of glycinergic sIPSCs was significantly higher in 8 PPO neurons than in 7 PSO neurons (Fig. 7d–f).
Because removing glutamatergic input had an opposite effect on the firing activity of PPO and PSO neurons, we next determined how glutamatergic input controls GABAergic and glycinergic inputs to PPO and PSO neurons. Bath application of CNQX and AP-5 significantly decreased the frequency, but not the amplitude, of GABAergic sIPSCs in 9 PPO and 8 PSO neurons (Fig. 7a–c). However, application of CNQX and AP-5 caused a large reduction in the frequency of glycinergic sIPSCs in PPO (n = 8), but not in PSO (n = 7), neurons (Fig. 7d–f).
Differences in levels of μ-opioid receptor–mediated inhibition of excitability of PPO and PSO neurons in the BN
Activation of μ-opioid receptors in the dorsal pontine tegmentum inhibits the micturition reflex (Willette et al. 1988). To determine whether there is a difference in the control of PPO and PSO neurons by μ-opioid receptors in the BN, we tested the effect of DAMGO, a highly specific μ-opioid receptor agonist, on the firing activity of these two types of neurons. Bath application of 1 μM DAMGO markedly reduced the firing rate of PPO neurons (n = 8, Fig. 8a and b). The membrane potential was significantly hyperpolarized from −52.6 ± 1.5 to −55.4 ± 1.6 mV in these PPO neurons. DAMGO also significantly decreased the firing rate of 9 PSO neurons (Fig. 8c and d) and only slightly hyperpolarized the membrane potentials of these neurons (from −45.6 ± 2.4 to −47.8 ± 1.9 mV, p > 0.05).
Our study provides new information for the understanding of the neural circuitry and central mechanisms regulating the micturition reflex. Because of the difficulty in identifying PPO and PSO neurons in vivo and the potential confounding effects of anesthetics in anesthetized animals, we developed a brain slice preparation to determine the membrane and synaptic properties of PPO and PSO neurons in the BN of rats. To retrogradely label these two populations of neurons that can be later identified for electrophysiological recording in brain slices, we injected green and red fluorescent tracers into the T13-T1 and S1-S2 levels of the spinal cord, respectively. Fluorescent microspheres were used as retrograde tracers because they remain at the injection site for a long period and have no detectable effect on the neuronal properties (Katz et al. 1984; Li et al. 2004; Ye et al. 2011). We found that there were many more PPO neurons than PSO neurons within the BN, which is consistent with the concept that the BN has a predominantly excitatory role in inducing bladder contraction and micturition reflex (Tanaka et al. 2003; Sasaki 2004, 2005b). It should be recognized that because the retrograde tracers were injected into the spinal cord, some labeled PPO and PSO neurons in the BN may be also involved in the control of the visceral organs other than the bladder. For the sake of simplicity, we limit our discussion of the functional properties of PPO and PSO neurons in the BN to the micturition reflex. Another limitation of this preparation is that the physiological responses of PPO and PSO neurons in the BN associated with micturition reflexes cannot be directly assessed because the normal afferent inputs from the visceral organs and the neuronal connections to other brain regions are removed in brain slices.
The urinary bladder is innervated by both sympathetic and parasympathetic efferent nerves. Increased sympathetic nerve activity, in response to a modest increase in bladder pressure from the urine accumulation, closes the internal sphincter and inhibits the contraction of the bladder wall musculature, allowing the bladder to fill (de Groat et al. 1993; Khadra et al. 1995). Also, moderate bladder distension inhibits parasympathetic efferent activity. When the bladder is full, increased afferent activity centrally reduces sympathetic activity and augments parasympathetic tone, which leads to contraction of the bladder and opening of the internal sphincter to cause bladder emptying (de Groat and Saum 1976). Although the BN can influence visceral functions through the sympathetic efferent nerves (Cano et al. 2000), it remains unclear whether some BN neurons project directly to the sympathetic preganglionic neurons in the spinal cord. Although the retrograde tracers we used in this study generally are not taken up by fibers in passage (Katz et al. 1984), we cannot rule out the possibility that some severed axons in the injection region of the spinal cord may be labeled and might result in labeling of some spinally projecting neurons in the BN. However, this problem seems less likely because very few PPO and PSO neurons in the BN were double labeled by the two tracers. The BN may inhibit the activity of sympathetic preganglionic neurons to the bladder muscle during micturition contraction. We showed that very few neurons in the BN were double labeled by tracers injected into both thoraco-lumbar and sacral cords, indicating that sacral cord-projecting PPO neurons do not directly suppress sympathetic preganglionic neurons via their axon collaterals in the thoraco-lumbar segments.
Various types of BN neurons can be distinguished on the basis of their firing properties in response to the bladder contraction–relaxation cycle. For instance, ‘direct’ neurons display low basal activity and increased firing patterns during bladder contraction (Sasaki 2004, 2005a, b). The ‘inverse’ neurons in the BN are inhibited during bladder contractions but are more active during the intervals between bladder contractions in cats (Tanaka et al. 2003; Sasaki 2004, 2005a). We found that in the brain slice, the basal activity and membrane potential were lower in PPO than in PSO neurons. Also, PPO neurons displayed a longer delayed onset of action potentials at the beginning of the pulse in response to a depolarizing current and a delayed return of the membrane potential to baseline after the repolarization pulse. Furthermore, we found that blocking A-type K+ currents with 4-AP produced a larger increase in the firing activity of PPO neurons than that in PSO neurons and that PPO neurons expressed larger A-type K+ currents than those in PSO neurons. In addition to regulating the action potential waveforms and inter-spike intervals (Segal et al. 1984; Vydyanathan et al. 2005), A-type K+ currents may also contribute to lower resting membrane potentials, hyperpolarized threshold, and larger amplitude of action potentials in PPO neurons. Therefore, the presence of large A-type K+ channel currents in PPO neurons may play a major role in keeping the basal activity of these neurons low in the absence of excitatory input from the bladder.
The neuronal firing activity is controlled by both intrinsic and synaptic inputs, which finely tune neuronal excitability. Because glycine and γ-aminobutyric acid (GABA) are the major inhibitory neurotransmitters in the brainstem, we determined the role of these excitatory and inhibitory inputs in the control of the firing activity of the PPO and PSO neurons in the BN. We found that blocking the GABAA receptors significantly increased the firing activity of both PPO and PSO neurons. However, blocking glycine receptors increased firing activity in PPO, but not in PSO, neurons. Our findings suggest that although the excitability of PPO and PSO neurons is tonically inhibited by GABAergic input, the firing activity of PPO neurons is also subject to tonic inhibition by glycinergic inputs. Consistent with this notion, we found that the basal frequency and amplitude of GABAergic sIPSCs were similar in PPO and PSO neurons. However, the frequency of basal glycinergic sIPSCs was significantly greater in PPO than in PSO neurons. It has been shown that systemic administration of strychnine can evoke reflex bladder contractions in cats (Shefchyk et al. 1998). Thus, the presence of a large glycinergic input is probably another important factor in the low basal excitability of PPO neurons in the BN.
Glutamate is the predominant excitatory neurotransmitter in the brain. We found that blocking ionotropic glutamate receptors reduced the firing of PSO neurons in the BN, suggesting that glutamatergic input contributes to the basal activity of these neurons. However, blocking ionotropic glutamate receptors paradoxically increased the firing activity of PPO neurons. This finding is striking and unexpected for the role of glutamatergic input in regulating PPO neurons. Direct recording of glutamatergic sEPSCs revealed that the frequency of sEPSCs was much lower in PPO neurons than in PSO neurons. These results suggest that compared with PSO neurons, PPO neurons receive much less excitatory input directly from glutamatergic neurons. We showed that PPO neurons received major inhibitory input from glycinergic interneurons and that removal of glycinergic inputs excited PPO, but not PSO, neurons. It is thus possible that the excitability of glycinergic interneurons is predominantly regulated by glutamatergic input and that blocking ionotropic glutamate receptors may inhibit glycinergic neurons and subsequently lead to disinhibition of PPO neurons. This hypothesis is supported by our finding that blocking ionotropic glutamate receptors failed to excite PPO neurons in the presence of the glycine receptor antagonist. Furthermore, we found that blocking ionotropic glutamate receptors decreased the frequency of glycinergic sIPSCs much more in PPO neurons than in PSO neurons. Therefore, glycinergic terminals that make direct synaptic contact with PPO neurons are tonically regulated by glutamatergic input. Our findings in brain slices are also consistent with the in vivo study showing that microinjection of glutamate receptor agonists into the pontine micturition center can inhibit bladder reflexes likely through stimulation of inhibitory interneurons in cats (Mallory et al. 1991). In contrast, PSO neurons seem to receive major excitatory input directly from glutamatergic neurons. This could explain why blocking ionotropic glutamate receptors still reduced the excitability of PSO neurons, although it also decreased GABAergic and glycinergic inputs to these neurons. The differences in synaptic contacts and neural circuitry further illustrate the distinct roles of inhibitory and excitatory input in regulating the firing activity of PPO and PSO neurons in the BN.
Activation of μ-opioid receptors in the brain inhibits the micturition reflex in cats and rats (Hisamitsu and de Groat 1984; Willette et al. 1988; Noto et al. 1991). We found that stimulation of μ-opioid receptors with DAMGO reduced the firing activity of both PPO and PSO neurons in the BN. Stimulation of μ-opioid receptors causes membrane hyperpolarization in neurons through activation of G protein-coupled inwardly rectifying potassium channels (North et al. 1987). Because PPO neurons are essential for initiation of the micturition reflex and bladder contraction, our finding is consistent with previous studies showing that the urinary retention induced by μ-opioid receptor agonists that was observed in patients may be mediated centrally (Kontani and Kawabata 1988; Verhamme et al. 2008).
In summary, our study provides novel information about the distinct membrane and synaptic properties of spinally projecting PPO and PSO neurons in the BN. Compared with PSO neurons, the basal activity of PPO neurons is low because these neurons are endowed with large A-type K+ channels and receive predominant inhibitory glycinergic inputs. Also, the glycinergic interneurons that control the excitability of PPO neurons are tonically influenced by excitatory glutamatergic input. These findings could serve as a basis for further studies of central mechanisms involved in micturition disorders, such as urinary control problems caused by neurological diseases, stress and anxiety.
This study was supported by grants from the National Institutes of Health (GM064830 and NS073935) and by the N.G. and Helen T. Hawkins Endowment (to H.-L.P.). The authors declare that they have no conflict of interest.