INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Trigeminal motoneurons participate in many functions. Like other quadrupeds, guinea pigs, the rodent used in the present experiments, use their jaw in delicate movements such as grooming, or as a prehensile organ for transporting their offspring as well as in aggressive behaviors such as fighting or biting. The rhythmic discharge of trigeminal motoneurons underlies masticatory and other movements. To carry out this range of radically different behaviors, trigeminal motoneurons are modulated by a variety of transmitters and neuromodulators that have been the subject of numerous morphological and electrophysiological studies. The list of neuroactive substances present in synapses in contact with somatic motoneurons includes amino acids, biogenic amines, acetylcholine, and neuropeptides (Bae et al. 1999; Del Negro and Chandler 1998; Hsiao et al. 1997; Rekling et al. 2000, Yang et al. 1997).

In recent years, nitric oxide (NO) has been identified as a novel and peculiar neurotransmitter in the CNS. This molecule has been attributed signaling, trophic, and neuroprotective functions (Brenman and Bredt 1997; Dawson et al. 1993; Estévez et al. 1998b; Garthwaite and Boulton 1995; Gross and Wolin 1995; Inglis et al. 1998; Moncada et al. 1991; Park et al. 1998; Strijbos et al. 1996).

The original evidence introduced in the present report indicates that nitrergic processes innervate trigeminal motoneurons and that NO may act as a neuromodulator on these cells by promoting the synthesis of cyclic guanosine 3,5'-monophosphate (cGMP).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Guinea pigs weighing 150-200 g (20-30 days old) were used in the present work. Seven of these animals were used for the anatomical studies and 33 for the electrophysiological studies. All experimental procedures were conducted in accord with the National Research Council Guide for the Care and Use of Laboratory Animals (7th edition, Natural Academy Press, Washington, DC, 1996).

Histological methods

INJECTION OF CTB-HRP IN JAW MUSCLES. Cholera toxin subunit bhorseradish peroxidase conjugate (CTb-HRP) was purchased from List Biological Laboratories (Campbell, CA). Guinea pigs were anesthetized with ketamine HCl (150 mg/kg im), small incisions in the skin were made to expose the masseter, digastric and temporalis muscles, and each muscle was injected with 15 µl of a 0.15% solution of CTb-HRP.

Perfusion and immunohistochemical procedures

Two to 3 days after the injections, the animals were killed with an overdose of pentobarbital sodium and perfused with heparinized saline, followed by a solution of 4% paraformaldehyde, 15% saturated picric acid, and 0.25% glutaraldehyde in 0.1 M phosphate buffer (PB) at pH 7.4. The brain stem was removed and immersed for 24 h postfixation in a solution consisting of 2% paraformaldehyde and 15% saturated picric acid in 0.1 M PB at pH 7.4. Following postfixation, the tissue was kept in a solution of sucrose (25%) in 0.1 M PB at pH 7.4 for 2 days. The brain stem was frozen, cut into 15-µm-thick sections using a cryostat; each section was then placed in a well of a multi-well tray containing a buffered solution [0.1 M PBS containing 0.3% Triton X-100 (PBST) and 0.1% sodium azide].

To immunostain retrogradely transported CTb-HRP, free-floating sections were incubated overnight in polyclonal antiserum directed against this substance (List Biological Laboratories; dilution 1:20,000). After rinsing in PBST, the tissue was incubated for 90 min in the secondary antibody at a dilution of 1:2,000. The sections were then rinsed in PBST and treated with the ABC complex (Vector standard Elite kit, Vector Laboratories, Burlingame, CA). Peroxidase activity was visualized by reacting the sections with 0.02 diaminobenzidine tetrahydrochloride (DAB) and 0.015 hydrogen peroxidase in 50 ml of 50 mM Tris-buffered saline, pH 7.6 for 15-30 min.

For nitric oxide synthase (NOS) immunocytochemistry, free-floating sections were incubated overnight in polyclonal antiserum directed against the neuronal isoform of NOS (nNOS; Accurate Chemical and Scientific, Westbury, NY; dilution: 1:500). After rinsing in PBST, the tissue was incubated for 90 min in the secondary antibody for nNOS at a dilution of 1:500. The sections were then rinsed in PBST and treated as described above. Staining for nNOS was usually combined with CTb-HRP labeling.

To detect NADPH-d chemical activity, sections were incubated in a solution of 0.1 M PBS, pH 7.4, 0.3% triton, 0.1 mg/ml p-nitroblue tetrazolium dichloride (NBT), and 1.0 mg/ml of beta-NADPH for 30-60 min. After tissue fixation with an aldehyde solution, the remaining NADPH-d activity is that associated with nNOS (Hope et al. 1991). NADPH-d histochemistry was usually combined with CTb-HRP labeling.

Histological sections were examined using an Olympus BX60 microscope (Olympus Optical, Tokyo, Japan). Photomicrographs were obtained by means of a digital camera attached to the microscope. The camera was connected to a microcomputer running Photoshop software. A complete description of these procedures may be found in Pose et al. (2000).

Electrophysiological studies

A detailed description of the methods of recording and data analyses can be found in Engelhardt et al. (1995); the following is an abbreviated description.

Preparation of brain stem slices

The animals were anesthetized with ether and ketamine (100 mg/kg) and decapitated. A craniotomy was then performed and the cerebellum excised. The brain stem was removed and placed in modified artificial cerebrospinal fluid (M-ACSF) at 4°C in which sucrose was substituted for NaCl to minimize hypoxic damage (Aghajanian and Rasmussen 1989).

The brain stem was glued to the platform of a Vibroslice chamber and covered with cold M-ACSF bubbled with 95% O₂-5% CO₂. Coronal slices (400 µm thick) were cut and placed in a holding chamber that contained M-ASCF at room temperature. The slices were kept in this chamber for 1 h while the M-ACSF was progressively replaced by normal-ACSF (N-ACSF). A brain stem slice that contained the motor nucleus was then transferred to an interface-type recording chamber, placed on a piece of filter paper and continuously perfused at a rate of 0.8-1 ml/min with N-ACSF at 32°C bubbled with 95% O₂-5% CO₂.

Solutions

The composition of the N-ACSF solution (in mM) was as follows: 126 NaCl, 5 KCl, 2 CaCl₂, 2 MgCl₂, 26 NaHCO₃, 2 HEPES, and 10 D-glucose, pH 7.4. In the M-ACSF 252 mM sucrose was substituted for 126 mM NaCl. The following agents were added to the N-ACSF solution. As NO-donors, we used 2 mM N-(2-aminoethyl)-N-(2-hydroxy-2-nitrosohydrazino)-1,2-ethylenediamine (spermine/nitric oxide complex, SPER/NO), 2-4 mM 2,2'-(hydroxynitrosohydrazono)bis-ethanimine-(diethylenetriamine/nitric oxide adduct, DETA/NO), and 1 mM pentakis(cyano-c)nitrosyl-ferrate(2-)disodium (sodium nitroprusside, SNP). As a membrane-permeant cGMP analogue, we used 2 mM N6,2'-0-dibutyrylguanosine-3':5'-cyclic monophosphate (db-cGMP). To block the soluble guanylyl cyclase (sGC), we used 10 µM 1H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one (ODQ).

NO released by NO-donors

The rate of NO released by DETA/NO or SPER/NO was estimated by determining the donor concentration decay monitoring the decrease of absorbance at 250 nm in a spectrophotometer. The technique was similar to that employed by Goss et al. (1997). The tests were performed on the perfusate taken from the recording chamber.

Recording and analysis

Because NO produces its effects on motoneurons by promoting the synthesis of the second intracellular messenger cGMP (sGC blockade suppressed the effects of NO-donors; see above), we used sharp electrode recording techniques in order not to affect the internal milieu of the cell during the present experiments. Intracellular recordings were obtained with glass microelectrodes (filled with KCl 3 M or KAc 2 M, tip resistance: 50-100 M) using a high-input impedance amplifier. These recordings were monitored using an oscilloscope and a microcomputer and stored for subsequent analyses on a video cassette recorder equipped with a pulse code modulation adaptor.

The electrophysiological properties of motoneurons were measured as previously described (Engelhardt et al. 1995). Briefly, the resting membrane potential was calculated as the difference between the recorded intracellular potential and that obtained after withdrawing the electrode from the cell. Action potentials were evoked by threshold current pulses. Their amplitudes were measured from the origin of the spike to its peak. The spike half-width was determined as the spike duration measured at half-amplitude. Rheobase (R_H) was the minimum stimulus intensity of a 50-ms intracellular depolarizing pulse that elicited an action potential. The input resistance (R_in) was measured from the average of 20 voltage responses to a 100-ms, 1-nA constant current pulse using the "direct" method wherein the maximal membrane potential change during the pulse is divided by the magnitude of the current (Engelhardt et al. 1995). The membrane time constant (_m) was measured using the "peeling" method to correct for the nonlinearity introduced by an I_H current (Engelhardt et al. 1995; Zengel et al. 1985).

Statistical analysis

The two-tailed Wilcoxon signed-rank test was used to compare the changes in neuron properties after the application of NO-donors or other drugs. The two-tailed Mann-Whitney U test was used to compare the effects of NO-donors with and without ODQ application. The level of significance was set at P < 0.05. Values are expressed as means ± SE.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Histological results

Retrogradely labeled trigeminal motoneurons are presented in the left column of Fig. 1 (A1 and A2). The photomicrograph in A1 was taken from a coronal hemisection of the brain stem of the guinea pig that was immunostained for CTb-HRP and reacted with p-NBT to detect NADPH-d histochemical activity. Labeled motoneurons display a dark brown staining. The dashed line encompasses the boundaries of the motor nucleus reconstructed from another section counterstained using a Nissl technique. The long arrows point to motor axons containing the retrograde label. Dendritic branches that display considerable labeling are also observed. Short arrows point to labeled dendrites extending in the dorsal direction. Two trigeminal motoneurons are illustrated at high magnification (×100) in Fig. 1A2. These cells contain brown granules of labeled CTb-HRP, but do not reveal NADPH-d histochemical activity. In contrast, in Fig. 1A3, neurons of the laterodorsal tegmentum (LDT) observed in another pontine section, from the same well in the same animal, exhibited an intense blue color produced by the NBT reaction. In Fig. 1B, LDT neurons were immunolabeled using an antibody against nNOS. Laterodorsal tegmentum cells are shown to illustrate NADPH-d or nNOS positive staining in nitrergic neurons.

View larger version (95K): [in this window] [in a new window]   Fig. 1. Nitrergic fibers innervate motoneuron processes. A1: coronal section of the trigeminal motor pool, immunostained for cholera toxin subunit bhorseradish peroxidase (CTb-HRP) and reacted with p-nitroblue tetrazolium dichloride (NBT) to detect NADPH-d activity. Labeled motoneurons display a dark brown color. The dashed line indicates the boundaries of the motor nucleus. According to their location, the labeled neurons belong to the jaw closer masseter and temporal motor nuclei. Long arrows point to labeled motor axons. Short arrows point to dorsally extending dendrites. The skyblue background reflects NADPH-d activity. A2: retrogradely labeled trigeminal motoneurons are illustrated in this photomicrograph at high magnification (×100). This section was treated to detect NADPH-d histochemical activity. Note that motoneurons contained CTb-HRP deposits but did not display diaphorase activity. A3: nitrergic neurons of the laterodorsal tegmentum (LDT) displaying strong NADPH-d activity are illustrated here. The histological section illustrated in A3 was 100 µm rostral to that of the motoneuron pool shown in A1. A4: a photomicrograph taken from the trigeminal motor nucleus contralateral to the CTb-HRP injection side. The dashed circle encompasses a series of swellings from the branches of a single NADPH-d-labeled fiber. The arrow points to another fiber branching in the same section. A5: a high magnification (×100) photomicrograph taken from the trigeminal motor nucleus ipsilateral to CTb-HRP injection. The long arrow points to a NADPH-d-positive fiber that bifurcates into 2 daughter fibers that, in turn, end in bouton-like structures (arrowheads) in apposition to a retrogradely labeled dark brown dendrite. A5': a drawing of the fiber, boutons, and the dendrite illustrated in A5. B: example of the characteristic staining of nNOS containing neurons: LDT neurons that were strongly labeled for neuronal nitric oxide synthase (nNOS) are illustrated in this photomicrograph. This section was taken from the same well than the section illustrated in Fig. 1D. C1-C4: retrogradely CTb-HRP-labeled motoneuron processes appear black in these photomicrographs obtained from sections treated with nickel to enhance the staining; nNOS-immunostained fibers appear golden brown. In C1 and C2, the arrows point to nNOS-labeled fibers intermingled with dark motoneuron processes. In C3, a long nNOS containing fiber (f) branches within this section; its daughter branches are indicated by arrows. The arrowhead points to a bouton-like structure in close contact with a transversally cut fine dendrite (d"); d and d' identify a thin dendrite that is observed crossing the section. C4: the photomicrograph illustrates a nNOS-labeled fiber in golden brown color that appears to give rise to 2 bouton-like terminals in contact with the postsynaptic dendrite (d). C4': a drawing of the fiber and the dendrites illustrated in C4. Another dendrite (d') is observed, in this section. D: 2 sections of the contralateral trigeminal motor nucleus immunolabeled for nNOS. Motoneurons were negative to nNOS immunolabeling; therefore they cannot be observed. Instead, nNOS-like-labeled fibers are seen in C1 and C2. The arrows in 2 point to varicose swellings of a nNOS containing fiber. Calibration bars: 100 µm in A1; 10 µm in A2; 20 µm in A3; and 5 µm in A4 and A5. In B, 15 µm. For C1 and C2, 10 µm; for C3, 5 µm; for C4, 2 µm. For D1 and D2, 2 µm.

In Fig. 1A4, the circle encompasses a series of bouton-like structures containing NADPH-d activity that originated from a common fiber; the photomicrograph was taken from the trigeminal motor nucleus contralateral to the CTb-HRP muscular injection. In Fig. 1A5 a single fiber that displays NADPH-d activity branches near a CTb-HRP retrogradely labeled proximal dendrite and appear to end in two bouton-like structures in close contact with the labeled dendrite (drawing in A5').

In those sections immunostained for CTb-HRP and processed for nNOS-like immunoreactivity, labeled motoneuron processes were black; nitrergic processes displayed a golden-brown color (Fig. 1C). Neuronal NOS-like immunolabeled fibers in close proximity to CTb-HRP labeled fine dendritic processes are illustrated in Fig. 1, C1-C4. Most nNOS containing fibers within the trigeminal nucleus were very thin (~0.6-1.6 µm). According to their fine diameters, most dendrites shown in these photomicrographs likely are distal dendrites. The nitrergic fibers illustrated in Fig. 1, C1 and C2, did not branch, but appear to end in swellings, bouton-like structures in apposition to dendritic processes. In the example in Fig. 1C3 a nNOS containing fiber (f), branching within the plane of the histological section is observed; one branch appears to end in a bouton-like structure (arrowhead) making contact with a transversally cut motoneuron dendrite (d"). In Fig. 1C4 a nNOS containing fiber (arrow) is observed ending (arrowheads) in a labeled dendrite. The drawing in C4' is a reconstruction of the dendrite and the fiber, made by adjusting the focus of the microscope. The scheme illustrates that before ending, the fiber branches in two and that its branches end in bouton-like structures (arrowheads). Neuronal NOS-like immunoreactive fibers in the motorpool contralateral to CTb-HRP injections are illustrated in Fig. 1D. The example in the right panel (D2) illustrates a fiber that displays varicosities along its trajectory.

In the present study we focus on the description of the nitrergic innervation of the trigeminal motor pool because our electrophysiological experiments were carried out in these nuclei. In fact, nitrergic fibers have been observed by us within other brain stem motor pools such as the hypoglossal, facial or ambiguus nuclei, as well as in cervical and lumbar spinal motor pools. We have observed this innervation in guinea pigs, rats, and cats. In another related study motivated by the present data, we found retrogradely labeled nitrergic premotor interneurons located in the medullary reticular formation, following CTb injection in the trigeminal motorpool.

Trigeminal motoneurons did not exhibit nNOS-like immunoreactivity. The photomicrographs in Fig. 1, D1 and D2, were taken from the middle portion of the trigeminal motor nucleus. In these photomicrographs, motoneurons were not stained; instead, only golden-brown stained nNOS fibers, bouton-like structures, and fiber varicosities were observed. In contrast, the photomicrograph in B illustrates neurons of the LDT that were strongly immunostained for nNOS.

NO released by NO-donors

DETA/NO or SPER/NO release, in an aqueous buffer, two NO molecules plus two molecules of the corresponding inert molecule moiety (e.g., DETA/NO 2NO + 2DETA). The decay rate of DETA/NO or SPER/NO follows a first-order kinetics such as: [NO] = [NO donor] · e^kt. DETA/NO half life was 30 h, whereas k equaled 0.38 × 10³ min¹. SPER/NO half life was 60 min; k equaled 11.55 × 10³ min¹. Four mM DETA/NO resulted in a NO flux of 3.08 µM/min; 2 mM DETA/NO, in a flux of 1.54 µM/min, whereas 2 mM SPER/NO in a flux of 46 µM/min. In the case of the other NO donor utilized (SNP), Southam and Garthwaite (1991) estimate that a range from 1 µM to 10 mM in the perfusate corresponds to a concentration near the tisssue ranging from 0.1 nM to 1 µM. Accordingly, the dose of 1 mM used in the present work should correspond to a concentration in the recording chamber of 0.1 µM.

Electrophysiological results

Stable recordings were obtained from 45 trigeminal motoneurons with resting potentials more negative than 50 mV and action potentials larger than 50 mV. The means ± SE of these parameters are presented in Table 1. The duration of the recordings, under stable conditions, was between 20 min (minimum acceptable) and 2 h.

Effects of NO-donors on membrane properties

The typical effects of NO-donors are illustrated in Fig. 2A. The recordings in the top row of this figure were obtained from an experiment in which SPER/NO was applied. This substance induced a 9-mV depolarization and an increase in the number of action potentials elicited by the current pulse. Rheobase (R_H) decreased from 2.6 to 1.8 nA (31%). Similar findings are illustrated in the recordings in the bottom row of this figure, which are from an experiment in which DETA/NO was used. Following the application of this substance, there was a membrane depolarization of 5 mV. In this experiment, the current pulse was set equal to R_H before and during DETA/NO. A 75% decrease in R_H was observed. The effects of NO-donors were reversible after washing the preparation with the control solution (right-hand traces in top and bottom row).

View larger version (21K): [in this window] [in a new window]   Fig. 2. Electrophysiological effects of nitric oxide (NO)-donors. A: recordings obtained from a motoneuron during control conditions, during the effects of spermine/nitric oxide complex (SPER/NO), and after complete recovery (top row). A depolarizing current pulse (100 ms, 2.6 nA) evoked 2 action potentials during both control and recovery, and a train of action potentials during the effects of SPER/NO (2 mM). This motoneuron depolarized by 9 mV from 75 to 66 mV; rheobase decreased from 2.6 to 1.8 nA. A: another example of NO-donors effects this time using DETA/NO (4 mM; bottom row). In this experiment, the intensity of the depolarizing pulse was adjusted to be at threshold level in every trial. This neuron depolarized 5 mV (from 55 to 50 mV), and its rheobase decreased from 0.12 to 0.03 nA. B: top plots illustrate the changes in rheobase for each NO donor in the population of recorded neurons. Bottom plots represent the concomitant changes in membrane potential during the maximum effect of the NO-donor. The lines connect the control values (left) with maximum changes produced by the NO-donor (right).

Figure 2B illustrates the effects of SPER/NO, DETA/NO, and SNP on both R_H (top plots) and the membrane potential (bottom plots) in different motoneurons. In a total of 22 neurons, R_H decreased in all but one. In 24 motoneurons, the membrane potential depolarized in all but three. In Table 1, data from all these experiments were pooled. Following the administration of NO-donors, there was a mean depolarization of 7.5 mV and a concomitant decrease in R_H of 38%. These changes were not accompanied by modifications in the characteristics of the action potential, input resistance (R_in), or membrane time constant (_m) values.

The latency-to-onset and the time-to-peak of the membrane depolarization observed during continuous perfusion with the NO-donors are shown in Table 2. Once a maximum effect was reached, the perfusion was switched to the control solution to allow for a recovery from this effect. The latency of the effects produced by the NO-donors, SPER/NO, and DETA/NO, were on the order of 3 min. The maximum effects were reached in 8.1 and 6.2 min for SPER/NO and DETA/NO, respectively. Recovery was complete 6.0 ± 1.3 min after the perfusion was changed to the control solution. Similar latency and recovery times were found for changes in rheobase. The application of denaturalized NO-donors did not evoke any effect.

The voltage changes evoked by hyperpolarizing current pulses before and after the administration of NO-donors are shown in the averaged records illustrated in Fig. 3. As illustrated in this example, the typical voltage response reached a minimum and then gradually "sagged." At the cessation of the pulse, there was a rebound depolarization. These nonlinear responses of the motoneuron membrane are attributed to the effects of a hyperpolarization-activated cationic current (I_H) (Pape 1996). This current is suppressed by Cs+ ions (Engelhardt et al. 1995; Rekling et al. 2000).

View larger version (11K): [in this window] [in a new window]   Fig. 3. Effects of SNP on the averaged voltage response to a hyperpolarizing current pulse. The thin trace represents the response to a 1.7-nA current pulse before the administration of the NO-donor. The thicker trace is the response to the same intensity current pulse during bath application of SNP (1 mM). During the administration of this substance, there was a depolarization of 5 mV from 62 to 57 mV. The sag and the rebound depolarization were both enhanced by this substance. Each trace is the average of 20 consecutive samples.

In the example illustrated in Fig. 3, following the administration of SNP and in spite of a 5-mV depolarization, the "sag" and rebound depolarization increased, indicating an increase in I_H current. These effects, i.e., augmentation of "sag" and rebound depolarization in spite of a strong membrane potential depolarization that otherwise would have deactivated this current, were observed in 50% of the neurons tested (7 of 14).

Effects of blocking sGC and/or membrane-permeant c-GMP

The consequence of blocking the sGC on the effects of NO-donors are presented in Fig. 4A. The top traces are recordings obtained during the passage of a 50-ms depolarizing current pulse that was just suprathreshold for eliciting an action potential in every trial. Perfusion with SPER/NO produced a 7-mV membrane depolarization accompanied by a decrease (45%) in threshold current. Seven minutes after switching to the control solution, there was an almost complete recovery from these effects. At this time, ODQ, a sGC inhibitor, was applied (10 µM). This substance per se had no effect on membrane potential and/or excitability of this neuron. During the continuous perfusion with the solution containing ODQ, SPER/NO was applied. In the presence of ODQ, there was almost no effect of SPER/NO. This motoneuron depolarized only 1 mV, and the threshold current decreased only by 16%. In contrast, during ODQ application, db-cGMP produced qualitatively similar effects to those of NO-donors in normal conditions. The last recording in Fig. 4A is an example of the effects of db-cGMP on the same cell. During perfusion with this compound in the presence of ODQ, the membrane potential depolarized by 3 mV, and there was a decrease in threshold current (37%).

View larger version (29K): [in this window] [in a new window]   Fig. 4. Blockade of soluble guanylyl cyclase (sGC) by 1H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one (ODQ) suppresses the effects of SPER/NO. A: before perfusion with ODQ, the NO-donor evoked a depolarization of 7 mV (from 55 to 48 mV) and a 45% decrease in rheobase (from 0.25 to 0.11 nA). ODQ by itself did not evoke any detectable effect (4th trace from the left). During ODQ perfusion, SPER/NO (2 mM) evoked a 1-mV depolarization and a 16% decrease in rheobase (5th trace from the left). db-cGMP (2 mM), however, produced a 3-mV depolarization and a 37% decrease in rheobase. B: bar diagrams for the population of examined motoneurons. The bars on the left represent the decrease in rheobase observed during the perfusion with the NO-donor alone (empty bar) and during the perfusion with NO donor in the presence of 10 µM ODQ (filled bar). The bars on the right represent the membrane depolarization induced by the NO donor alone and that observed during the perfusion with NO donor plus ODQ (empty and filled bars, respectively). ODQ suppressed both the changes in RH and membrane potential elicited by NO-donors. C: intracellular record showing the depolarizing and excitatory effects of db-cGMP. This substance evoked a 13-mV (from 83 to 70 mV) depolarization and a 33% decrease in rheobase (from 3.9 to 2.7 nA). In 6 neurons membrane potential changes were from 62.17 ± 5.4 mV (control) to 51 ± 3.9 mV (during db-cGMP), these changes were statistically significant (P < 0.03). Mean rheobase, in control conditions was 1.73 ± 0.56 nA, whereas during db-cGMP application was to 1.36 ± 0.36; these values were not statistically significant (P < 0.3).

The blocking action of ODQ on the NO donor induced depolarization and changes in R_H are summarized in the bar histograms of Fig. 4B. In Fig. 4C, the effects of db-cGMP on another neuron that had not been previously treated with ODQ, are shown. Dibutyryl-cGMP elicited a membrane depolarization of 13 mV and a concomitant 33% decrease in R_H which, therefore mimicked the effects of NO-donors.

Synaptic activity

High-gain membrane potential recordings obtained before and during perfusion with DETA/NO are illustrated in Fig. 5, A1 and A2. These recordings were obtained from two different motoneurons. In control conditions, discrete, small (<1 mV) synaptic potentials were observed only occasionally. Following the administration of DETA/NO, there occurred a large increase in both the number and the amplitude of synaptic potential activity. These potentials displayed a fast rise and slow decay time, which resembled the shape of stimulus-evoked synaptic potentials. The largest potentials were in the order of 4-5 mV in amplitude. In the case of the recordings illustrated in Fig. 5, A2, following DETA/NO, there were also large hyperpolarizing potentials. A cluster of these hyperpolarizing potentials can be observed in the third trace from the top in Fig. 5A2. The increase in synaptic potential activity following NO-donors was seen in 22 neurons. In all of these neurons, only occasionally was potential synaptic activity observed during control conditions. Bath application of tetrodotoxin (5 µM) that blocked motoneuron action potential suppressed the synaptic activity produced by application of DETA/NO 2 mM (Fig. 5C). In 20 other neurons, the application of NO-donors did not elicit any detectable synaptic activity.

View larger version (49K): [in this window] [in a new window]   Fig. 5. NO-donors, membrane-permeant db-cGMP, and tetrodotoxin effects on synaptic potential activity. A1 and A2: recordings obtained from 2 different motoneurons. In A1, DETA/NO (4 mM) evoked mainly depolarizing potentials. In A2, this substance evoked both depolarizing and hyperpolarizing potentials. B1: db-cGMP evoked large synaptic potentials that on occasions triggered action potential activity. In B2, the synaptic activity was characterized by both depolarizing and hyperpolarizing potentials. K-acetate-filled microelectrodes were used for these experiments. In C, DETA/NO evoked synaptic potential activity, whereas TTX supressed the effects of DETA/NO.

In Fig. 5, B1 and B2, high-gain membrane potential recordings obtained from two different neurons, both before and during the application of db-cGMP, are shown. In both experiments, the membrane-permeant compound mimicked the effects of NO-donors on synaptic activity. In the recording illustrated in B1, the depolarizing synaptic potentials triggered action potentials. In the recordings presented in B2, both depolarizing and hyperpolarizing synaptic potentials are present following the application of db-cGMP. This increase in synaptic potential activity following db-cGMP was observed in six of seven motoneurons tested (86%).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In the present study, we described the results of anatomical and electrophysiological experiments that indicate that nitrergic processes innervate trigeminal motoneurons and that NO modulates these cells via the synthesis of cGMP. The present, that is the initial study on this type of motor innervation, indicates the existence of a system/s of premotor nitrergic neurons with direct motor functions and provides a foundation for future studies designed to illustrate the physiological significance of our findings.

Motoneurons did not display nNOS-like immunoreactivity and were negative to the chemical reaction to detect NADPH-d activity. These techniques are based on completely different biochemical principles, and the results, which indicate that these cells do not express nNOS, support each other (Dawson et al. 1991; Hope et al. 1991). These data agree with previous data from spinal and other brain stem motoneurons (Mariotti and Bentivoglio 1996; Yu 1994) and with those of Vázquez et al. (1999), who found a transient expression of nNOS in hypoglossal motoneurons in the maturing postnatal rat, but not in rats older than 21 days of age.

Nitrergic innervation of the motor pool

The retrograde labeling of motoneurons resulted in a dark staining of somas and dendrites. CTb-HRP reached even distal dendrites because dark granular reaction product could be observed inside very thin fibers (on the order of 1 µm). Most contacts established by nitrergic fibers were with motoneuron dendrites. Neuronal NOS containing fibers running alongside fine, labeled dendrites was a common observation. The proximity of these elements is important because NO is not only produced at nerve terminals, but also throughout the entire nerve cell, including its axons (Park et al. 1998; Wiklund et al. 1997). NO is produced following activity-dependent Ca²⁺ entry into nNOS containing processes, and NO, once produced, diffuses freely from these processes (Brenman and Bredt 1997; Garthwaite and Boulton 1995). Because the nNOS containing fiber varicosities may be interpreted as synapses "en passant" and the bouton-like structures as standard synaptic terminals (Brown et al. 1977; Maxwell et al. 1982; Rethelyi et al. 1982), it is possible that a classical neurotransmitter is colocalized with nNOS in these terminal-like structures (Maqbool et al. 1995).

NO synthesized by synapses en passant, boutons, and fibers within the motor pool would be expected to diffuse to modulate postsynaptic targets. How far this molecule diffuses at an efficacious concentration depends on the size of the neuronal element producing it and on the amount produced (Park et al. 1998; Philippides et al. 2000). It would be expected that, since their size is small, each individual fiber or terminal would not produce as much NO as a nitrergic soma and effective NO diffusion from fibers and terminals would not be as extensive as that associated with nitrergic cell bodies (Park et al. 1998; Philippides et al. 2000). Therefore it is possible that the amount of NO synthesized by a single process in the motor pool would only suffice to act on those neural elements closest to them that, as shown here, are predominantly dendritic. If, however, many nNOS containing processes were activated at the same time, larger amounts of NO with a greater effective diffusion distance would be produced in the motorpool; under these circumstances trigeminal motoneurons could be simultaneously modulated. In this regard, NO innervation of the motor nucleus may serve the function of synchronizing the activity of these cells.

Effects of NO on motoneuron electrophysiological properties

The three different NO-donors that we used gave essentially the same results in spite of their different molecular structure. This suggests that the effects of their bath application were due to the release of NO. Millimolar concentrations of NO-donors in N-ACSF solution resulted in micromolar concentrations of NO in the recording chamber (see RESULTS). Because NO is subject to rapid inactivation by tissue-derived factors (Brovkovych et al. 1999; Southam and Garthwaite 1991), its actual concentration in the extracellular milieu in contact with cells within the brain stem slice should have been significantly lower, conceivably, in the order of tenths of micromoles (Brovkovych et al. 1999; Shibushi and Okada 1991; Southam and Garthwaite 1991). Physiological measurements of NO in brain tissue have resulted in disparate values. Malinski et al. (1993) reported values <10 nM for baseline NO levels in brain, whereas Suzuki et al. (1998) found basal NO levels averaging 1.6 µM; Buerk (2001) found in parietal cortex basal NO levels of 1.1 µM. If the estimations by Suzuki and Buerk (Buerk 2001; Suzuki et al. 1998) are correct, then the dosages used in the present work may have been close to physiological concentrations. However, it is important to note that the amounts of NO produced in situ by the activity of any given set of nitrergic synapses in any given brain structure is hitherto ignored.

The most consistent effect of NO-donors was a motoneuron depolarization (mean: 7.5 mV), accompanied by a mean 38% decrease in R_H. Membrane potential depolarization attributed to NO activation of either an I_H or a persistent Na⁺ inward current is common in neurons in the CNS of mammals and invertebrates (Bains and Ferguson 1997; Hammarstrom and Gage 1999; Pape and Mager 1992; Park et al. 1998; Pineda et al. 1996; Sawada et al. 1995; Travagli and Gillis 1994; Zhi-Qing et al. 1998). These effects are usually accompanied by an increase in membrane conductance. However, if the depolarization were due to an increase in I_H or to an increase in a persistent Na⁺ current membrane conductance changes may not result in detectable changes in R_in. This is because, given the large net driving force of Na⁺, only a small change in membrane permeability for this ion would be needed to produce a few millivolts of depolarization. Only one study reports that depolarization may be evoked by closing a K⁺ conductance; in this case there was a concomitant increase in R_in (Koh and Jacklet 1999). Somatic motoneurons show clear evidence of an I_H current near the membrane resting potential (Chandler et al. 1994; Engelhardt et al. 1995; Morales et al. 1987; Nishimura et al. 1989; Rekling et al. 2000). The present results indicate that I_H is increased by NO in spite of a strong background depolarization, which would tend to deactivate I_H channels. The existence of a Na⁺-dependent persistent current has been described in trigeminal motoneurons by Chandler et al. (1994) and in facial motoneurons by Nishimura et al. (1989). This current is considered to be an important factor in determining the excitability of these neurons near voltage threshold and, in principle, could also be modulated by NO.

Synaptic potential activity

A remarkable increase in the activity of both depolarizing and hyperpolarizing synaptic potentials occurred under the influence of NO-donors and membrane-permeant cGMP, indicating that NO is capable of acting on both excitatory and inhibitory presynaptic terminals in contact with trigeminal motoneurons. In addition, NO may have activated premotor interneurons whose soma were included in the brain stem slice. Indeed, many neurons discharge following NO application (Bains and Ferguson 1997; Pape 1992; Pineda et al. 1996; Travagli and Gillis 1994). Tetrodotoxin effects are consistent with this possibility.

Biochemical pathways for the effects of NO

The effects of NO-donors, but not those of db-cGMP, were suppressed by previous administration of ODQ. This indicates that NO-induced changes in motoneuron properties and activity were dependent of an activation of sGC and of a subsequent synthesis of cGMP. Consistent with this interpretation is the fact that db-cGMP mimicked the effects of NO. Motoneurons are known to contain sGCs, and cGMP (Furuyama et al. 1993; Southam and Garthwaite 1993). In that case, under normal conditions, these cells should be postsynaptic targets of the NO produced by nNOS containing processes. We did not examine the metabolic pathways involved beyond cGMP synthesis. It remains to be determined whether cGMP synthetized by NO-mediated sGC stimulation activates cyclic nucleotide-gated channels expressed by these cells (Kingston et al. 1999) or acts indirectly via the activation of cGMP-dependent protein kinases (Ruth 1999).

Relevance of NO to physio- and pathophysiological mechanisms

NO had an overall excitatory effect on motoneurons. Therefore this substance should now be regarded as a physiological neuromodulator in the trigeminal motor nucleus. In addition, NO is believed to play a role in the development of motoneuron dendrites (Inglis et al. 1998). These authors propose that NOS containing neurons adjacent to spinal cord motor nuclei are the source of NO. Our findings suggest that a direct nitrergic innervation of trigeminal motoneurons may also produce this putatively trophic molecule. Although it appears that NO may function as a protective molecule, if produced in excess, it may lesion normal cells (Dawson et al. 1993; Estévez et al. 1998a,b; Gross and Wolin 1995; Inglis et al. 1998; Moncada et al. 1991; Strijbos et al. 1996). NO has been considered a factor in the pathogenesis of motoneuron degeneration in amiotrophic lateral sclerosis (Abe et al. 1995; Beckman et al. 1993; Chou et al. 1996a,b). From the present results it is clear that the nitrergic fibers and terminals of medullary origin that we describe in the trigeminal motor pool provide an anatomical substrate for physiological, either signaling or neuroprotective, and/or pathological effects attributed to NO.