Introduction

Top Abstract Introduction Methods Results Discussion Conclusions References

Sympathetic neurons in prevertebral ganglia the celiac ganglia (CG), superior mesenteric ganglion (SMG) and inferior mesenteric ganglion (IMG) innervate the mesenteric vasculature and regulate intestinal activities. Functional studies have indicated that prevertebral ganglia act as complex neural networks that are capable of integrating efferent and sensory information and coordinating peripheral reflexes (Miolan and Niel, 1996).

Immunohistochemical studies demonstrate the presence of NOS-IR in nerve terminals in prevertebral ganglia arising from three distinct sources including colonic intestinofugal neurons (Furness and Anderson, 1994; Anderson et al., 1995; Mann et al., 1995), sympathetic preganglionic neurons (Furness and Anderson, 1994; Anderson et al., 1995) and primary sensory afferent neurons whose cell bodies lie in dorsal root ganglia (Zheng et al., 1997b). These distinct sources of NOS-IR in fibers innervating prevertebral ganglia may reflect involvement in the control and modulation of specific functions such as motility and secretory reflexes (Anderson et al., 1995) or blood flow (Zheng et al., 1997b).

With regard to the regulation of the mesenteric vasculature, colonic distention activates capsaicin-sensitive inhibitory reflex pathway that results in a hyperpolarization and relaxation of mesenteric arterial smooth muscle (Meehan and Kreulen, 1992). This may be mediated, at least in part, via the release of nitric oxide from dorsal root ganglion neurons. It is conceivable, therefore, that the NOS-IR primary sensory afferent neurons that provide a direct nitrergic innervation to mesenteric blood vessels also innervate prevertebral sympathetic neurons (Zheng et al., 1997b).

The action of NO on prevertebral ganglion neurons has not been fully characterized. In the mouse intact superior mesenteric ganglion, intracellular recordings have shown that the NO donor, sodium nitroprusside hyperpolarizes the majority of neurons, although a subpopulation of these showed a biphasic response, i.e., a hyperpolarization followed by a depolarization (Mazet et al., 1996). Thus, our purpose was to determine the action of NO on cultured dissociated celiac neurons using whole-cell patch-clamp recordings. Celiac neurons were chosen because of their well-characterized properties and response to dissociation and culture techniques (Coggan et al., 1991, 1994; Vanner et al., 1993). A preliminary report of this work has been published previously (Zheng et al., 1997a).

	Methods

Top Abstract Introduction Methods Results Discussion Conclusions References

Sympathetic neuronal dissociation. CG neurons were dissociated and cultured as previously described (Coggan et al., 1991). Briefly, female Duncan-Hartley guinea pigs (200-350 g) were anesthetized with halothane, decapitated and exsanguinated. The CG was excised, enzymatically dissociated (papain 9 mg/ml; collagenase 1 mg/ml; dispase 4 mg/ml) and plated as a monolayer on to poly-D-lysine coated glass-bottomed 35-mm culture dishes. Cells were maintained in feeding medium (see below) at 37°C in a 5% CO₂ humidified incubator. Two-thirds of the feeding medium was replaced every 3 days.

Electrophysiological recordings. Culture dishes containing CG neurons were transferred to the stage of an Olympus IMT-2 inverted microscope and perfused with Krebs' solution of the following composition (mM): NaCl 118.5; NaHCO₃ 23.8; KCl 4.7; MgCl₂ 1.2; CaCl₂ 2.5; KH₂PO₄ 1.2; glucose 5.5; maintained at pH 7.4 by bubbling with O₂/CO₂ (95%/5%). The temperature was held constant at 37°C. Whole cell patch recordings were made using borosilicate patch pipettes of resistance 3-8 M with the following intracellular solution composition (mM): KCl 144.5; MgCl₂ 2; EGTA (ethylene glycol-bis(-aminoethyl ether)N,N,N',N'-tetraacetic acid) 0.5; HEPES (N-[2-hydroxyethyl]piperazine-N'-[2-ethanesulfonic acid]) 5; ATP (adenosine 5'-triphosphate) 4; GTP (guanosine 5'-triphosphate) 0.25; pH adjusted to 7.35 with KOH. When cesium was used as the current carrier, the following pipette solution was used (mM): Cs-gluconate 140; HEPES 10; EGTA 1; NaCl 10; CaCl₂ 0.3; TEA 20; ATP 4; GTP 0.25; pH adjusted to 7.35 with NaOH.

Drug delivery. Drugs were applied by a U-shaped suction-ejection tube ("cloud-burst" method; Krishtal and Pidoplichko, 1980; see also Coggan et al., 1994) for time periods sufficient to allow a steady-state response to be obtained. The cloudburst method allows rapid application and removal of drugs to a discrete area, limited to the neuron being recorded from. Reduced hemoglobin (20 µM), apamin (100 nM), charybdotoxin (10 nM), 4-AP (1 mM), TEA (5-20 mM), tetrodotoxin (100 nM) and Krebs' containing cobalt (Co⁺⁺; 2 mM), cesium (Cs⁺; 2 mM) or barium (Ba⁺⁺; 2 mM) were applied by superfusion for a minimum of 10 min before the drug test.

Materials. SNAP was purchased from Molecular Probes, Inc. (Eugene, OR). Papain and collagenase were purchased from Worthington Biochemical Corp. (Freehold, NJ), dispase from Boehringer Mannheim GmbH. (Mannheim, Germany), minimal essential medium and penicillin-streptromycin from Life Technologies (Grand Island, NY), and 6,7-dimethyl-5,6,7,8-tetrahydropteridine from Calbiochem (San Diego, CA). All other cell culture reagents and chemicals were from Sigma Chemical Co. (St. Louis, MO).

Reduced hemoglobin. Hemoglobin was reduced in our laboratory with sodium dithionite following the method of Martin et al. (1985). The resulting solution was kept (20°C) for a period of not longer than 2 wk.

	Results

Top Abstract Introduction Methods Results Discussion Conclusions References

Electrophysiological recordings were made from 340 neurons from 46 guinea pigs between one and 6 days after dissociation.

Neuronal Classification

Neurons were classified on the basis of their responses to the passing of depolarizing pulses at three times rheobasic strength (holding potential 60 mV; depolarization duration 1 sec, frequency 0.2 Hz). Neurons that fired action potentials only at the onset of the depolarizing current pulses were defined as phasic, whereas tonic neurons fired action potentials throughout the duration of the depolarizing stimulus (Crowcroft and Szurszewski, 1971; Kreulen and Szurszewski, 1979). The passive and active properties of phasic and tonic neurons are summarized in table 1.

The NO donor, SNAP, induces an inward current shift in the majority of CG neurons. In neurons that were current clamped at potentials of, or more positive to, 50 mV, application of SNAP produced a 3.0 ± 1.4 mV membrane depolarization (n = 10; range 0-11 mV) that persisted throughout the drug application and decayed slowly after drug removal. Neurons voltage clamped at potentials  60 mV were found to have a standing outward current. Application of SNAP produced an inward shift in this holding current, defined as an inward current shift.

Characterization of the Response to SNAP

Concentration dependency. The steady-state inward current shifts to SNAP were concentration-dependent (fig. 1). A maximum response to SNAP could not, however, be obtained because at concentrations higher than 10 mM, SNAP-induced an irreversible current shift with a concomitant loss of seal integrity. The response to 10 mM SNAP was taken, therefore, as maximal and the other data points normalized to this point.

View larger version (12K): [in this window] [in a new window]   Fig. 1.   SNAP-induced inward current shift is concentration-dependent. A, Cloudburst applications of SNAP produced a concentration-dependent inward current shift in a neuron voltage clamped at 30 mV. The effect was reversible after wash-out of SNAP. B, Concentration-response curve for the SNAP-induced inward current shift in neurons voltage-clamped at 30 mV. Each point represents data from 4 to 24 neurons. For data to be used each neuron was tested with at least three concentrations of SNAP, one of which was 10 mM; a minimum of 10 min elapsed between successive drug applications. All points were normalized to the response produced to 10 mM SNAP.

Current/voltage relationship. The inward current shifts to SNAP were voltage dependent within the range 60 mV to 0 mV (fig. 2). No reversal of the SNAP-induced current shift was observed upon hyperpolarization, even at potentials of up to 100 mV.

View larger version (19K): [in this window] [in a new window]   Fig. 2.   SNAP-induced inward current shift is dependent on voltage and neuronal conductance. A, Steady-state current/voltage plot obtained from a ramp depolarization from 100 to 0 mV. In the presence of 10 mM SNAP, the amplitude of the steady-state current was reduced. The net SNAP-induced current shift was obtained by subtracting the control current from the SNAP current (see Ba). All subsequent current/voltage plots of the SNAP-induced current shift were obtained with this subtraction method. Ba, Net inward current shift induced by SNAP using a ramp depolarization protocol. Bb, Inward current shift induced by SNAP at different holding potentials. The traces have been obtained from the same neuron as in A. Note that the amplitude of the current shifts obtained with the ramp depolarization protocol (Ba) are similar to the amplitudes obtained with the cloudburst applications at different holding potentials (Bb). C, Neuronal conductance (nS; calculated between 60 and 30 mV in control conditions) plotted against the amplitude of the inward current induced by 10 mM SNAP applied to neurons voltage clamped at 30 mV. First order regression correlation co-efficient r2 = 0.89.

SNAP effect is prevented by hemoglobin. Pretreatment with the NO-scavenger, reduced hemoglobin (20 µM), attenuated the SNAP-induced inward current shift (n = 8; fig. 3A) implying that the SNAP-induced response is mediated via NO generation. The average reduction of the inward current shift induced to application of 10 mM SNAP was 72 ± 9% (n = 6; range 62-100%).

View larger version (22K): [in this window] [in a new window]   Fig. 3.   The SNAP-induced inward current shift is attenuated by reduced hemoglobin and prevented by pretreatment with cesium. Aa, Concentration-response curve for the inward current shift induced by SNAP (VH = 30 mV) in the presence and absence of reduced hemoglobin 20 µM. Responses are normalized to the inward current shift produced in response to 10 mM SNAP under control conditions (n = 5). Ab, Traces illustrating the inward current shift (VH = 30 mV) induced by cloudburst application of SNAP (upper trace). After 10 min superfusion with the NO-scavenger reduced hemoglobin (20 µM), the inward current response to SNAP is abolished (lower trace). Ba, The inward current shift induced by 10 mM SNAP was abolished by prior exposure to 5 mM cesium in the extracellular solution (n = 6). Bb, The inward current shift induced by cloudburst application of SNAP (VH = 30 mV) was abolished when K+ was replaced by Cs+ as the intracellular ion carrier. *Significantly different from control; P < .05.

SNAP-induced current is mediated by alterations of neuronal potassium currents. Decreasing the extracellular potassium from 5.9 to 2.5 mM increased the amplitude of the maximum inward current shift at 0 mV from 277 ± 18 pA to 10 mM SNAP in control (5.9 mM extracellular potassium concentration; n = 3; range 242 to 300 pA) to 394 ± 5.0 pA in 2.5 mM extracellular potassium (n = 3; range 384 to 400 pA). Conversely, increasing the potassium concentration to 10 mM decreased the inward current shift to 168 ± 21.1 pA (n = 3; range 143 to 210 pA). Addition of cesium (5 mM) to the superfusing Krebs' solution abolished the SNAP-induced current shift (fig. 3Ba). This indicated that the SNAP-induced inward current was mediated via inhibition of neuronal potassium conductance(s).

Characterization of SNAP effects on ionic currents. To determine the ionic currents involved in the SNAP-induced inward current shift, the effects of SNAP were studied under both current- and voltage-clamp conditions.

View larger version (30K): [in this window] [in a new window]   Fig. 4.   SNAP reduces the after-hyperpolarization currents. A, Perfusion of SNAP (10 mM) reduced the duration and amplitude of the after-hyperpolarization following bursts of action potential firing. Previous exposure to apamin (100 nM) abolished this after-burst hyperpolarization identifying it as the intermediate after-hyperpolarization current (data not shown). The depolarization observed after SNAP application was offset by injection of DC current necessary to return the holding voltage to the control value (50 mV). Control = black trace; SNAP = gray trace. B, Expanded trace from A. Perfusion of SNAP (10 nM) reduced the amplitude of the hyperpolarization following individual action potentials, an effect mediated by charybdotoxin-sensitive BK channels, as well as reducing the inter-spike interval, an effect mediated by apamin-sensitive SK channels. Control = black trace; SNAP = gray trace. C, Previous exposure to TEA (5 mM) reduced but did not abolish the inward current shift induced by SNAP (10 mM; n = 7). The magnitude of this reduction in the SNAP-induced current shift by TEA was similar to that in the presence of charybdotoxin (10 nM; n = 7). D, Previous exposure to apamin (100 nM) reduced but did not abolish the inward current shift induced by SNAP (10 mM; n = 5). A combination of apamin (100 nM) and TEA (5 mM) was required to completely abolish the SNAP-induced inward current shift (n = 4). *Significantly different from control; P < .05.

SNAP-induced outward current. The addition of the nonselective voltage-dependent calcium channel blocker, cobalt (2 mM), to the superfusing Krebs' solution was expected to mimic the response seen with TEA and apamin but, instead, in 26/26 neurons to which cobalt was applied, the application of SNAP resulted in the induction of an outward current (fig. 5A). In the presence of cobalt, at 0 mV, the amplitude of the outward current shift was 211 ± 40 pA (range 62 to 558 pA; n = 26).

View larger version (20K): [in this window] [in a new window]   Fig. 5.   In the presence of cobalt, SNAP induces an outward current shift that is mediated by NO and gK. A, Left panel: Inward current shift induced by application of 10 mM SNAP. Right panel: In the same neuron, in the presence of extracellular cobalt (2 mM), cloudburst application of SNAP induced an outward current shift (VH = 30 mV). B, The outward current shift to SNAP (10 mM) was voltage-sensitive and abolished by perfusion of extracellular cesium (5 mM; n = 3); C, The outward current shift to SNAP (10 mM) in the presence of cobalt was abolished by prior exposure to reduced hemoglobin (20 µM; n = 5). *Significantly different from control; P < .05.

SNAP increases the delayed rectifier current. The delayed rectifier current was studied in the presence of tetrodotoxin (100 nM) and cobalt (1 mM). Neurons were voltage clamped at 50 mV and then stepped (250 msec) between 60 mV and +20 mV before being returned to the holding voltage of 50 mV.

View larger version (19K): [in this window] [in a new window]   Fig. 6.   SNAP activates the delayed rectifier current. A, SNAP (10 mM) increased the outward delayed rectifier current. The stimulus artifact has been removed. B, In the absence of cobalt, the inward current shift induced by SNAP (10 mM) was increased in the presence of 4-AP (1 mM). The outward current induced by SNAP (10 mM) in the presence of cobalt was reduced by prior exposure to 4-AP (1 mM). The amplitude of the 4-AP-induced increase in the inward current shift was similar in magnitude to the 4-AP-induced reduction in the outward current shift.

	Discussion

Top Abstract Introduction Methods Results Discussion Conclusions References

We report that the NO-donor SNAP produced an inward current shift in the majority (89%) of cultured dissociated guinea pig celiac ganglion neurons. This action of SNAP was mediated via NO because the response was abolished by prior exposure to the NO-scavenger, reduced hemoglobin and via a decrease in neuronal potassium conductances because of its block by intra- and extracellular cesium and its sensitivity to the extracellular potassium concentration. The SNAP-induced inward current shift was mediated through an inhibition of after-hyperpolarization currents because apamin, charybdotoxin or TEA attenuated the response; a combination of apamin and TEA was required to block the response. Blockade of I_AHP by cobalt unmasked a SNAP-induced outward current. Abolition by hemoglobin indicated that this outward current was similarly mediated via NO, and sensitivity to extracellular cesium implied an involvement of potassium conductances. Application of 4-AP (1 mM) reduced this outward current. Low concentrations of 4-AP inhibit both the A-current (Belluzzi et al., 1985) and the delayed rectifier (Marsh and Brown, 1991). The A-current, however, is unlikely to have any role in the SNAP-induced responses since SNAP induced an inward current shift in conditions in which I_A would be inactivated. Conversely, the outward current induced by SNAP in the presence of cobalt could be mediated through activation of I_KV. In the absence of cobalt, 4-AP increased the SNAP-induced inward current shift implying that SNAP simultaneously inhibits I_AHP and activates I_KV.

An action of NO to inhibit calcium-dependent potassium currents has been reported in hippocampal slices, with a concomitant enhancement of neuronal excitability (Erdemli and Krnjevi, 1995). Actions of nitric oxide to activate the potassium delayed rectifier current have been reported in the rat pulmonary artery (Zhao et al., 1997). To our knowledge, however, this is the first report of nitric oxide mediating a simultaneous activation (of I_KV) and inhibition (of I_AHP) of potassium conductances in the same neuron or tissue.

Both inhibitory and excitatory actions of NO-donors have been reported in spinal thermosensitive neurons where NO causes an inhibitory effect, observed as a decrease in firing rate, in the majority of neurons recorded from laminae I and II. In contrast, however, the majority of thermosensitive neurons in lamina X respond to NO-donors with an increase in excitability (Schmid and Pehl, 1996; Schmid et al., 1997). This biphasic effect is likely to be due to that the presence (laminae I and II) or absence (lamina X) of I_AHP.

Few studies have been undertaken to assess the actions of NO in sympathetic ganglionic neurons. Intracellular recordings were made from the mouse superior mesenteric ganglion in vitro where NO is proposed to act to modulate slow synaptic transmission inducing a membrane hyperpolarization in the majority (64%) of neurons, although a small number of neurons (8%) respond with a membrane hyperpolarization followed by a depolarization (Mazet et al., 1996). The results of our study differ from these intracellular recordings that may reflect the influence that the synaptic inputs present in intact ganglia may have on neuronal excitability and behavior, reflected in the differing neuromodulatory actions of NO. A similar neuromodulatory role for NO has been proposed in pancreatic ganglia of the cat, where both NO and the NO donor, sodium nitroprusside, evoked a neuronal membrane hyperpolarization as well as an initiating fast excitatory postsynaptic potentials in the majority of neurons recorded (Sha et al., 1995).

Our experiments indicated that SNAP exerted little influence upon neurons at resting membrane potentials. The mean resting membrane potential (E_m) for all dissociated neurons, both phasic and tonic, was 50.1 ± 0.1 mV. This compares favorably with a value for E_m of 54 ± 0.9 mV for intact celiac ganglion neurons (Kreulen and Szurszewski, 1979) implying that E_m does not alter significantly after dissociation and culture. SNAP-induced an inward current shift only when neurons were depolarized to potentials less negative than -50 mV. At the average resting membrane potential (50.1 ± 0.9 mV), the NO donor, SNAP, produced only a 24 ± 5.4 pA inward current shift. This would suggest that in the absence of any other depolarization, release of nitric oxide would have little effect on the prevertebral neurons.

If, however, the neuron was depolarized by any means other than inhibition of I_AHP, then the nitrergic input would inhibit the after-hyperpolarization currents. These calcium-dependent potassium currents hyperpolarize the neuron after either a burst of action potentials or a single action potential of duration sufficient to raise intracellular calcium levels, and act to limit action potential firing frequency by increasing the refractory period (for review see Adams and Harper, 1995). Inhibition of these currents by NO decreased the after-hyperpolarization after bursts of action potentials as well as after individual action potentials, and decreased the refractory period resulting in an increased ability of the neuron to fire action potentials as well as an increased frequency of action potential burst firing.

If the prevertebral sympathetic neuron was depolarized but the I_AHP already inhibited, then the predominant effect of the nitrergic input would be to activate I_KV. Such a delayed outward rectifying potassium current terminates the action potential and may make a significant contribution to the after-hyperpolarization after the action potential (Marsh and Brown, 1991; for review see Adams and Harper, 1995). Activation of this current by NO would tend to offset the existing CG neuronal depolarization, thereby reducing its excitability.

	Conclusions

Top Abstract Introduction Methods Results Discussion Conclusions References

The overall effect of NO on sympathetic postganglionic neurons will depend, therefore, on the excitability state of the neuron at the time it receives the nitrergic input that, in turn, may be dependent on the principal neurotransmitters involved in synaptic transmission in these neurons. In particular, an excitatory input of an neurotransmitter such as acetylcholine (which does not decrease I_AHP) would be amplified by a concomitant nitrergic input, although an excitatory neurotransmitter such as substance P, which can decrease I_AHP, (Vanner et al., 1993) would be antagonized.