Journal of Applied Physiology

Induction of long-lasting depolarization in medioventral medulla neurons by cholinergic input from the pedunculopontine nucleus

Keiko Mamiya, Kevin Bay, R. D. Skinner, E. Garcia-Rill


Stimulation of the pedunculopontine nucleus (PPN) is known to induce changes in arousal and postural/locomotor states by activation of such descending targets as the caudal pons and the medioventral medulla (MED). Previously, PPN stimulation was reported to induce prolonged responses (PRs) in intracellularly recorded caudal pontine neurons in vitro. The present study used intracellular recordings in semihorizontal slices from rat brain stem (postnatal days 12–21) to determine responses in MED neurons following PPN stimulation. One-half (40/81) of MED neurons showed PRs after PPN stimulation. MED neurons with PRs had shorter duration action potential, longer duration afterhyperpolarization, and higher amplitude afterhyperpolarization than non-PR MED neurons. PR MED neurons were significantly larger (568 ± 44 μm2) than non-PR MED neurons (387 ± 32 μm2). The longest mean duration PRs and maximal firing rates during PRs were induced by PPN stimulation at 60 Hz compared with 10, 30, or 90 Hz. The muscarinic cholinergic agonist carbachol induced depolarization in all PR neurons tested, and the muscarinic cholinergic antagonist scopolamine reduced or blocked carbachol- and PPN stimulation-induced PRs in all MED neurons tested. These findings suggest that PPN stimulation-induced PRs may be due to activation of muscarinic receptor-sensitive channels, allowing MED neurons to respond to a transient, frequency-dependent depolarization with long-lasting stable states. PPN stimulation appears to induce PRs in large MED neurons using parameters known best to induce locomotion.

  • arousal
  • locomotion
  • posture
  • reticular-activating system

the pedunculopontine nucleus (PPN), as the cholinergic arm of the reticular-activating system (RAS), has been implicated in the modulation of sleep-wake states, the startle response (SR), and posture and locomotion (reviewed in Refs. 8, 41, 52). The overall goal of our studies is the understanding of the influence of the PPN on ascending thalamocortical systems and their role in sleep-wake states, as well as the simultaneous modulation of accompanying postural and locomotor events via its influence on descending reticulospinal systems. Electrical stimulation of the PPN can induce widespread ascending effects, such as cortical EEG desynchronization (39) [better referred to as synchronization of fast cortical rhythms (53)], as well as changes in muscle tone (32) and the recruitment of stepping (10, 1217). The PPN, therefore, appears to be involved in the ascending control of transitions in state from slow-wave sleep to either waking or rapid eye movement (REM) sleep, but it also participates in descending functions, which involve changes in muscle activity from standing (extensor activation), to the SR (flexor activation, extensor inhibition), to the atonia of REM sleep (flexor and extensor inhibition), and to locomotion (flexor-extensor alternation). Such functions may be characterized as “fight-or-flight” responses.

On the one hand, PPN neurons are known to increase their firing rates during synchronization of fast rhythms in waking and REM sleep [tonically in waking, increased bursting during REM sleep and reduced activity during slow-wave sleep (3, 51, 52)]. Different types of activity patterns in relation to sleep-wake states have been defined specifically in the PPN of the behaving rat, providing evidence for the cellular mechanisms by which the PPN could be involved in the regulation of sleep-wake states (4). PPN neurons also show both tonic and rhythmic activity in relation to either the duration of stepping episodes or the rhythmic alternation of locomotor movements (14, 17).

On the other hand, the PPN sends diffuse, mostly cholinergic projections throughout the pontomedullary reticular formation (8, 12, 16, 26, 37, 43, 47). The pontine oral (PnO), gigantocellular reticular, and dorsal paragigantocellular nuclei (36, 57), along with the ventral portion of the medullary reticular formation lateral to the inferior olive (38), all appear to be involved in the decrease in muscle tone or atonia of REM sleep. PPN stimulation also alters respiratory movements and leads to acetylcholine release in the pontine reticular formation (33). Moreover, a region medial to the latter, located immediately rostral to the inferior olive and extending dorsally to the anterior third of the inferior olive, appears to be an area involved in locomotor control. This region includes the magnocellular reticular formation (pars alpha) and ventral nucleus (pars alpha) of the reticular formation (40), was termed the medioventral medulla (MED), and receives input from the PPN (12). Stimulation of the PPN at specific frequencies is known to elicit locomotion on a treadmill (15). Electrical stimulation of the MED itself, as well as injections of the cholinergic agonist carbachol into this region, were found to induce locomotion on a treadmill in decerebrate cats and rats (13, 28). These findings suggest that descending cholinergic PPN projections to the MED may underlie functions related to locomotion. In a previous study, our laboratory recorded the responses of the more posture-related anterior target of the PPN, the caudal pontine (PnC) reticular formation, and found that cholinergic PPN inputs were directed mainly at small, interneuronlike cells, rather than larger, projection-type cells (21). PPN stimulation at frequencies known to induce stepping elicited prolonged responses (PRs) outlasting stimulation by >10 s in these PnC neurons in vitro. We hypothesized that the PPN would exert a similar frequency-dependent effect on MED neurons and that the organization of PPN inputs to these locomotion-related neurons would differ from those to posture-related PnC neurons. The present studies were undertaken to determine the nature of the responses of rat MED neurons recorded in vitro following PPN stimulation using various frequencies of stimulation. Preliminary findings have been reported in abstract form (35).



All procedures were approved by the Institutional Animal Care and Use Committee. Adult timed-pregnant Sprague-Dawley rats (280–350 g) were used, and the litters were culled to 10. When the pups were 12–21 days old, individual pups were anesthetized using ketamine (70 mg/kg im) until tail pinch and corneal reflexes were absent and then were rapidly decapitated. We should note that, although ketamine is known to inhibit acetylcholine release from PPN neurons (34), the recordings took place several hours after decapitation (see below) so that any effects of the anesthetic had waned, especially given the consistency of responses following PPN stimulation. The brains were dissected free under cooled (4°C), oxygenated (95% O2–5% CO2) artificial cerebrospinal fluid (aCSF), and the brain stem containing the PPN and the MED bilaterally was blocked so slices could be cut semihorizontally. The block of tissue was glued onto a stage, and 500-μm slices were cut with a Vibroslice (Campden Institute) under cooled, oxygenated aCSF and then allowed to equilibrate for 1 h in oxygenated aCSF at room temperature before recording. The composition of the aCSF was (in mM) 122.8 NaCl, 5 KCl, 1.2 MgSO4, 2.5 CaCl2, 1.2 NaH2PO4, 25 NaHCO3, and 10 dextrose. The slices contained most MED neurons as well as the direct pathway from the PPN. Only one to two of the 500-μm slices from each brain contained the MED and PPN. In this study, we used a total of 77 pups and 81 well-studied cells for data analysis.

Recording Procedures

The recording chamber allowed the slice to be suspended on a nylon mesh so that oxygenated aCSF could flow all around the slice. The gravity-fed aCSF flowed through a sleeve of circulating warmed water so that the temperature of the aCSF in the chamber was 30 ± 1°C. The outflow was removed by suction and flow adjusted to 2–3 ml/min. Microelectrodes were pulled in a Sutter Instruments puller using Omega-Dot, thin-wall borosilicate glass, and were filled with 3 M K+ acetate and 1% biocytin, with a resistance of 60–90 MΩ. Signals were amplified with an Axoclamp 2B amplifier (Axon Institute) in these current clamp recordings. Neurons were impaled and allowed to stabilize for ∼5 min before testing. Neurons that showed a stable resting membrane potential (RMP) and that had stable, long-term recordings were accepted for data analysis. The RMPs were verified and adjusted when the electrode was withdrawn at the end of recordings (usually only 1- to 2-mV difference, sometimes >5 mV, especially after biocytin injection). In bridge mode, a series of hyperpolarizing and depolarizing current steps of 0.1 nA at RMP were applied to determine several membrane properties. These current steps also allowed the computation of a preliminary current-voltage curve during the linear range of voltage deflections using SuperScope software (GW Institute). The properties measured included membrane input resistance (Rin, determined using hyperpolarizing 300-ms-duration pulses of 0.1–0.3 nA applied at RMP), action potential (AP) amplitude and threshold (determined from the beginning of the sodium spike to its peak in APs occurring spontaneously at RMP or, if no spontaneous activity was evident, by depolarizing the membrane until individual APs were induced, i.e., at AP threshold), AP duration at threshold (determined as the duration of the AP at half amplitude in APs recorded at AP threshold), afterhyperpolarization (AHP) amplitude (determined from AP threshold to the peak of the AHP in individually occurring spikes), and AHP duration (determined from AP threshold to the return to prespike membrane potential in individually occurring spikes).

Stimulation Procedures

Electrical stimulation of the PPN was carried out by using a bipolar concentric electrode (100-μm diameter, 100-kΩ resistance, Kopf Instruments, model SNEX-100), applying currents of 100–500 μA in amplitude using pulses of 0.2- to 0.5-ms duration, at frequencies of 10–90 Hz, individually and in trains of various durations, usually 1 s. The location of the stimulating electrode was confirmed using NADPH diaphorase histochemistry, as described below. In the studies described, all stimulating electrode sites were found within the region of NADPH diaphorase-positive cells.

Neuroactive agents were applied via a manifold with six perfusion ports; hence multiple gravity-fed solutions could be applied for pharmacological characterization of neuronal properties. The concentrations of the superfused neuroactive agents in aCSF were as follows: scopolamine (SCOP; 50 μM) and tetrodotoxin (TTX; 0.3 μM). Direct effects of these agents on recorded MED neurons were confirmed before, during, and after washout and recovery from TTX superfusion. The concentrations of these agents were adjusted so that effects were evident by using superfusion times of 1–2 min. Carbachol (30 μM) was applied by micropressure adjusted to elicit a response following two to five puffs applied to the surface of the tissue when the pipette was <100 μm from the recording microelectrode. The micropressure system was set at 30 psi, 50-ms-duration puffs, and pipette resistance was designed for application of ∼100 pl/puff. The micropressure pipettes contained 2% fast green dye to visualize flow of the puffed solution across the downstream-located recording microelectrode.

Histological Procedures

At the end of the recording period, each neuron was injected with biocytin using intracellular depolarizing pulses, adjusted to elicit a train of APs (∼0.5–1.0 nA) of 500-ms duration at 1 Hz for 10–15 min. Such injections yielded well-filled neurons. All of the slices were processed for NADPH diaphorase histochemistry for selective labeling of cholinergic mesopontine (PPN) neurons around stimulation sites. Briefly, slices were fixed in 4% buffered paraformaldehyde for 1–2 h, cryoprotected in 20% sucrose, and cut in a cryostat at 50 μm. Sections were incubated in 1 mg/ml NADPH and 0.1 mg/ml nitroblue tetrazolium in PBS at 37°C for 30–60 min (12–14, 49). For intracellularly labeled MED neurons, biocytin immunocytochemistry was carried out preceding NADPH diaphorase histochemistry utilizing a Vector ABC kit using the peroxidase-antiperoxidase method with diaminobenzidine as the chromogen. Sections were mounted on gelatin-coated slides and coverslipped with Eukitt for bright-field optics.

Statistical Procedures

For comparison of data between the different groups [PR vs. nonprolonged response (NPR)] in each experiment, measures were tested using one-factor (AP or AHP duration, AHP amplitude), two-factor (group vs. property), or multifactor (10- vs. 30- vs. 60- vs. 90-Hz stimulation) ANOVA to conclude whether any of the factors had a significant effect on the magnitude of the variable and also whether the interaction of the factors significantly affects the dependent variable(s). Differences were considered significant at values of P ≤ 0.05. If a statistical significance was present, a post hoc test (Scheffé) was used to compare between groups. Each of the measures of intrinsic membrane properties and responses to various frequencies of stimulation were compared using two-factor ANOVA (e.g., response amplitude and stimulation frequency). Comparisons of duration of effect that involved repeated observations on the same neurons, e.g., responses to neuroactive agents over time, were carried out using repeated-measures ANOVA since each time sampled could be considered a different “treatment” on each cell. When statistical significance was evident, the post hoc test (Newman-Keuls) was carried out to determine differences between groups of cells across “treatment” (e.g., duration or frequency of response).



We recorded a total of 81 cells, which met the criteria of stable RMP and stable long-lasting recordings. These neurons were localized in the MED immediately anterior to the inferior olive, in a region known as the magnocellular reticular formation (pars alpha) and ventral reticular nucleus (pars alpha) in the rat (40). The cells studied were found between the lateral edge of the midline raphé medially, extending laterally to the lateral edge of the inferior olive. Figure 1 is a drawing of a representative histological section of one of the semihorizontal slices used. The localization of a selected sample of the neurons recorded, which were well filled with biocytin, reflects the distribution described. Functionally, the region studied appeared to be well posterior (>500 μm) to the PnC neurons studied previously (21). The region sampled herein was intended to be equivalent to that sampled in our previous stimulation and recording studies of the MED in decerebrate preparations (12, 13, 28).

Fig. 1.

Locations of intracellularly recorded, injected medioventral medulla (MED) cells, and pedunculopontine nucleus (PPN) stimulation sites. Semi-horizontal brain stem slices were fixed and cut at 50 μm and processed for NADPH diaphorase histochemistry (specifically labeling PPN neurons) and for avidin-biotin immunocytochemistry (labeling of biocytin-injected cells). This drawing of a representative slice shows that the section traversed the central gray (CG) anterodorsally, the superior cerebellar peduncle (SCP), the motor nucleus of V (MoV), cranial nerve VII, and the inferior olive (IO) and pyramids (P) ventroposteriorly. The small dots denote NADPH diaphorase-positive cells, within which the PPN stimulation site was located (large solid circle among dots). In the MED immediately anterior to the IO, the locations of injected prolonged response (PR; n = 17, solid medium circles) and nonprolonged response (NPR; n = 19, open medium circles) neurons are denoted. Note that the cells recorded were in a region equivalent to that identified in adult animals as the MED based on the location of locomotion-inducing stimulation sites.

Electrophysiological Properties

Of the MED neurons studied, 40 (49%) showed PRs and 41 (51%) did not show PRs (NPR). The definition of a PR requires explanation. In our laboratory's previous study using extracellular recordings in the decerebrate cat brain stem, the PR was defined as the duration of the train of APs induced in PnC neurons following PPN stimulation from the beginning of the response until firing had ceased for 1 s (9). The subsequent intracellular studies revealed that PPN stimulation induced a long-lasting depolarization in PnC neurons on which was superimposed a train of APs (21). For comparative purposes, the duration of the PR was still defined as the duration of the AP train until firing had ceased for 1 s. However, the underlying and more important mechanism at play is obviously the much longer lasting depolarization induced by PPN stimulation. In the sample MED recordings that follow, the prestimulation membrane potential is denoted by a dotted line to facilitate detection of the depolarization induced in PR cells. We will first describe the general electrophysiological characteristics of MED cells across age, then compare the properties of PR and NPR cells, and, finally, turn to detailing the features of the PRs.

Effects of age.

Of the PR cells recorded, 24 of 40 were studied in slices from pups younger than 16 days, and 16 of 40 in slices from pups aged 16–21 days. That is, we compared cells recorded during the first half of the developmental window studied with those recorded during the second half. The rationale behind this division is related to the known change at around 15 days of age in the descending control of locomotor and swimming movements in the rat (2, 25). For example, if a spinal cord transection is made before 15 days of age, rats recover the capacity for spontaneous locomotion, whereas they do not recover such capacity if the transection is performed after 15 days (50, 58). There were no statistically significant differences between PR MED cells in these age groups on the basis of mean ± SD RMP (−62 ± 6 vs. −62 ± 7 mV), AP threshold (−47 ± 8 vs. −48 ± 7 mV), AP amplitude (52 ± 7 vs. 53 ± 9 mV), AP duration at half amplitude (1.3 ± 0.5 vs. 1.1 ± 0.5 ms), AHP amplitude (14 ± 4 vs. 16 ± 3 mV), or AHP duration (161 ± 52 vs. 145 ± 56 ms), Rin (95 ± 32 vs. 91 ± 36 MΩ), duration of PR (27 ± 14 vs. 23 ± 12 s), or amplitude of PR (5.5 ± 0.8 vs. 3.4 ± 0.9 mV). NPR MED (41 of 80) cells did not show major developmental differences across 15 days either, although they did differ from PR MED cells in some respects, as follows.

PR vs. NPR cells.

The electrophysiological characteristics of PR compared with NPR cells are listed in Table 1. The main difference, of course, was the presence of a PR, along with a shorter duration AP, a longer duration AHP, and a higher AHP amplitude in PR cells, but there were no significant differences in terms of RMP, Rin, AP threshold, or AP amplitude. Figure 2, A and B, shows representative examples of the two cell types that could be divided on the basis of responsiveness to PPN stimulation. MED neurons that showed PRs (40 of 81) following PPN stimulation (using paradigms to be discussed below) had shorter duration APs compared with MED neurons without PRs (1.2 ± 0.5 vs. 1.5 ± 1.0 ms; Fig. 2C; ANOVA df = 79, F = 6.43, P < 0.01, post hoc P < 0.05), longer duration AHP (153 ± 50 vs. 126 ± 47 ms; ANOVA df = 79, F = 6.57, P < 0.01, post hoc P < 0.05), and higher amplitude AHP (15 ± 3 vs. 12 ± 4 mV; ANOVA df = 79, F = 8.95, P < 0.004, post hoc P < 0.01).

Fig. 2.

Properties of PR and NPR MED cells. A: representative 19-day PR cell with a short-duration action potential (AP) and high-amplitude afterhyperpolarization (AHP) (left, calibration bars: vertical 10 mV, horizontal 10 ms) and responses to depolarizing and hyperpolarizing steps (right, calibration bars: vertical 20 mV, horizontal 100 ms), showing the presence of hyperpolarization-activated inward (Ih) and transient outward K+ (Ia) currents. B: representative 17-day NPR cell with a longer duration AP and lower amplitude AHP (left, calibration bars: vertical 10 mV, horizontal 10 ms) and responses to depolarizing steps (right, calibration bars: vertical 20 mV, horizontal 100 ms). C: mean ± SE of the AP duration of PR vs. NPR cells showing a statistically significant difference (*P < 0.05). D: responses following PPN stimulation (60 Hz, 1-s train, 60 V) for a 12-day PR cell [top, resting membrane potential (RMP) −65 mV] and for a 21-day NPR cell (bottom, RMP −65 mV). Calibration bars: vertical 10 mV, horizontal 1 s. Note that PR cells had shorter duration APs, which would allow higher firing frequencies than NPR cells.

View this table:
Table 1.

Electrophysiological properties of PR and NPR neurons in medioventral medulla

Figure 2A shows the characteristics of a PR MED cell with a shorter duration AP and higher amplitude AHP, along with those of a typical NPR MED cell (Fig. 2B) with a longer duration AP and lower amplitude AHP. In this case, the PR cell showed a hyperpolarization-activated inward (Ih) current during hyperpolarizing pulses, whereas the NPR cell did not have an Ih current, although this current was present in both cell types. In addition, both PR and NPR cells showed transient outward K+ (Ia) currents (slow repolarization, or sag, after the end of hyperpolarizing pulses, as shown in Fig. 2, A and B) or low-threshold spike (LTS) Ca2+-dependent transient outward K+ (It) currents (burst of APs on a hump after the end of hyperpolarizing pulses, not shown). As indicated above, Fig. 2C is a graph of the AP duration of PR and NPR cells showing that PR neurons in the MED had shorter duration APs. Figure 2D shows the effects of a train of stimuli at 60 Hz delivered for 1 s to the PPN on a 12-day PR MED cell at RMP (top record) and on a 21-day NPR MED cell at RMP. Note that the PR induced in the PR MED cell outlasted the stimulus by >15 s.

Figure 3A shows the responses of a representative 12-day PR MED cell with a higher amplitude, longer duration AHP than a representative NPR MED cell with a lower amplitude, shorter duration AHP. In these cases, the PR cell had an Ih current and LTS current, whereas the NPR cell showed an Ih current at greater than −80-mV membrane potential, as well as an Ia current. The distribution of PR and NPR cells of all ages plotted in terms of AHP amplitude vs. AHP duration is shown in Fig. 3C. (It should be noted that AHP duration was always measured in individually occurring APs at RMP or, if no spontaneous activity was evident, after depolarizing the membrane potential until individual APs were induced; therefore, the durations shown in this figure include AHPs in the range of −45 to −65 mV). As stated above, AHP amplitude was statistically different between these response types (PR 15 ± 3 vs. NPR 12 ± 4 mV), as was AHP duration (PR 153 ± 50 vs. 126 ± 47 ms) (Table 1). Of the PR cells recorded, 20 of 40, or 50%, showed an Ia current, whereas 25 of 40, or 63%, had an Ih current and 15 of 32, or 47%, showed an LTS current. That is, many cells had multiple currents, with only 10 cells showing only Ia and only 3 cells showing only LTS currents. NPR cells showed a similar proportion of cells with Ia (49%) and Ih (44%) but lower frequency of LTS (9%) currents. It should be noted that these cell types and their respective currents have been described previously in the pontomedullary reticular formation, and the currents have been identified pharmacologically (19, 24, 52, 54). However, no such pharmacological confirmation was undertaken in the present study.

Fig. 3.

AHP amplitude and duration in PR and NPR MED cells. A: representative 12-day PR cell showing high-amplitude AHP (left, calibration bars: vertical 10 mV, horizontal 10 ms) and responses to depolarizing and hyperpolarizing steps (right, calibration bars: vertical 20 mV, horizontal 100 ms), showing the presence of Ih and low-threshold spike currents. B: representative 12-day NPR cell showing lower amplitude AHP (left, calibration bars: vertical 10 mV, horizontal 10 ms) and responses to depolarizing and hyperpolarizing steps (right, calibration bars: vertical 10 mV, horizontal 10 ms), showing Ih and Ia currents. C: AHP amplitude vs. AHP duration of PR (solid circles) and NPR (open circles) cells showing overlapping distributions with a trend toward increasing AHP amplitudes and AHP durations for PR cells (see also Table 1). Note that the presence of higher amplitude AHPs in PR cells would tend to promote higher firing rates in the presence of such currents as Ih; however, the presence of longer duration AHPs would tend to reduce firing frequency in these cells. It is not clear how these currents interact in the intact preparation to promote locomotion.

In PR-type MED neurons, PPN stimulation using single pulses was found to induce short-latency excitatory postsynaptic potentials that decreased in latency and increased in amplitude and elicited APs as stimulating current levels were increased. Figure 4Aa shows the responses of a 12-day PR MED cell following stimulation of the PPN using single pulses of increasing amplitude, with a single trial shown at faster sweep speed (Fig. 4Ab), demonstrating a short-latency excitatory postsynaptic potential and APs. The mean ± SD latency to the beginning of the excitatory postsynaptic potential following PPN stimulation using single pulses was 3.9 ± 1.4 ms for the cells tested (n = 12). The mean threshold for inducing an excitatory postsynaptic response in both types of MED neurons following single-pulse PPN stimulation was 325 ± 82 μA. The responses of MED neurons following stimulation of the PPN were similar to those previously reported for descending PPN projections to pontomedullary locomotor areas (12, 13).

Fig. 4.

Responses of PR cells following PPN stimulation. Aa: single-pulse stimulation of increasing (10–40 V) amplitudes showing a response at 30 V in this 12-day MED cell (left arrow). Calibration bars: vertical 10 mV, horizontal 1 s. b: Expanded sweep of response to the third pulse showing an excitatory postsynaptic potential, two APs, and a high-amplitude AHP (right arrow). Calibration bars: vertical 10 mV, horizontal 10 ms. ce: PPN stimulation using 1-s, 60-Hz, 40-V pulses and behavior of the PR at different holding potentials: −50 mV (c), −55 mV (d), and −60 mV (e). B, ad: frequency dependence of firing frequency during the PR in a 14-day MED cell showing increasing firing frequency from 10 to 30 to 60 Hz, then decreasing to 90-Hz stimulation. Calibration bars: vertical 10 mV, horizontal 1 s. C, ad: PR amplitude and firing frequency in one of the few short-duration PR cells (13 day). Note the increased depolarization as frequency of stimulation was increased from 10 to 30 to 60 Hz, and then a decrease to 90-Hz stimulation, along with lower firing frequency. Calibration bars: vertical 10 mV, horizontal 1 s. Note that the highest firing frequencies were induced at RMP by the optimal stimulation frequency of 60 Hz, the same as that used to induce locomotion on a treadmill in decerebrate animals.


If trains of stimuli were used, the brief, single AP responses described above became PRs in 40 of 81, or 49%, of MED neurons. The PR was comparable to results previously described in PnC neurons in the decerebrate cat (9) and in our laboratory's previous slice recording study (21), and these lasted ∼14 s in PnC cells. In the present study, PR duration also varied but typically lasted >20 s in MED cells (Table 1). A maximum stimulus train duration of 1 s was selected for testing because stimulation for longer durations is known to recruit locomotion in decerebrate animals (8, 10, 14, 49), whereas 1-s trains elicit only brief spinal cord activation. This also allowed comparison with previous results on PnC neurons in decerebrate animals (9) and in PnC neurons in vitro (21). Figure 4A, ce, shows the responses of a MED neuron following PPN stimulation using 40-V stimuli delivered in a 1-s train at 60 Hz and the voltage dependence of the PR. In this cell, the RMP was −60 mV, and the induced PR consisted only of a small depolarization (Fig. 4Ae). As the membrane potential was held at more depolarized levels (−55 mV, Fig. 4Ad, and −50 mV, Fig. 4Ac), the PR resulted in a long train of APs. The mean ± SD amplitude of the depolarization induced by PPN stimulation (using trains of 1 s at 60 Hz) on PR MED cells was 4.6 ± 0.4 mV (n = 24). In all of the PR MED cells tested, the membrane potential was depolarized during the stimulus train, and the PR began during or at the end of the train of stimuli (Fig. 4). Unlike the responses previously described for PnC neurons, which could show PR after a delay or arising from a hyperpolarization induced during stimulation (21), MED cells monotonically were depolarized during or at the end of stimulation.

Figure 4B shows a MED cell at RMP (−60 mV) that manifested a train of APs throughout the PR following PPN stimulation. In this case, the PRs were stimulation frequency dependent such that low-frequency (10 Hz) trains of stimulation produced lower amplitude, lower firing frequency, short-lasting membrane depolarizations, but, as stimulation train frequency was increased (30 Hz), the depolarization became prolonged, and a maximal duration PR was evident at 60 Hz (Fig. 4B). Delivery of intracellular hyperpolarizing pulses failed to reset the PR, whether the current step was of sufficient amplitude to return the membrane potential to the resting level or to significantly more negative membrane potentials (not shown). Figure 4C shows a different MED cell in which PPN stimulation induced a long-lasting depolarization without APs when stimulated at 10 Hz but increasing frequencies of stimulation-induced trains of APs and higher amplitude PRs peaking at 60 Hz and then declining at 90 Hz.

Figures 5 provides graphs of the stimulation frequency-dependent effects on the duration of the PR (A), the firing frequency induced during the second second of the PR (B), and the amplitude of the peak depolarization elicited during the PR (C). Figure 5A is a graph of the mean duration of PRs induced by specific frequencies of stimulation tested using identical stimulation parameters in a group of 27 PR MED cells (0.5-ms- duration pulses of 400-μA amplitude, each using a 1-s train). The duration of the PR, as stated above, was determined to be from the first AP following the depolarization to the last AP without a 1-s cessation of firing. The mean ± SE duration of the PR induced at 10 Hz was 9.1 ± 1.8 s, at 30 Hz it increased to 16.1 ± 2.4 s, whereas the mean duration of PRs induced by trains of 60 Hz was 26.8 ± 3.2 s, and at 90 Hz it was 20.4 ± 2.3 s. The durations of these PRs were statistically different (ANOVA df = 95, F = 10.28, P < 0.001) such that the mean duration at 30 Hz was longer than that at 10 Hz (post hoc, P < 0.05), and that at 60 Hz was longer than that at 10 Hz (post hoc, P < 0.01), 30 Hz (post hoc, P < 0.01), and 90 Hz (post hoc, P < 0.05). That is, stimulation at 60 Hz induced the longest duration PRs. Interestingly, the PRs induced by PPN stimulation in MED cells were significantly (ANOVA df = 48, F = 7.51, P < 0.009) longer (25.6 ± 3.1 s) than PRs induced in PnC neurons (14 ± 4 s) (21).

Fig. 5.

A:PR duration following PPN stimulation at 10 Hz, which increased significantly at 30 Hz (P < 0.05), further increasing at 60 Hz (P < 0.01), and then decreasing at 90 Hz (P < 0.05 compared with 60 Hz). B: firing frequency during the second second of the PR following PPN stimulation at 10 Hz, which increased numerically [nonsignificant (NS)] at 30 Hz, further increasing at 60 Hz (P < 0.01), and then decreasing numerically (NS) at 90 Hz. C: amplitude of the depolarization (in mV) induced during the PR following PPN stimulation at 10 Hz, which increased significantly (P < 0.01) at 30 Hz, further increasing at 60 Hz (P < 0.01), and then decreasing at 90 Hz (P < 0.05 compared with 60 Hz). Note that 60 Hz was the optimal stimulation frequency for eliciting the longest duration PR, the highest firing frequency, and the highest PR amplitude. That is, a transient, powerful input from the PPN to MED neurons can induce a long-lasting change in state lasting many seconds. Significant difference: *P < 0.05; **P < 0.01.

There was a difference between the effects of trains of different frequencies and the firing frequency induced in MED neurons during the PRs (ANOVA df = 75, F = 8.75, P < 0.001). Figure 5B is a graph of the mean firing frequency induced during the second second of the PR (activity during the first second of the PR varied considerably from subsequent activity; therefore, the second second was considered to be a steady-state firing frequency). The mean ± SE of the firing frequency of MED neurons following stimulation using trains of 10 Hz (1.8 ± 0.4/s) increased numerically following 30-Hz trains [3.0 ± 0.5/s, post hoc, nonsignificant (NS)] and then increased markedly following 60-Hz trains (5.1 ± 0.4/s, post hoc, P < 0.01 vs. 10 and 30 Hz). This effect appeared to peak at 60 Hz, since firing frequency decreased numerically using 90-Hz trains (3.9 ± 0.5/s, post hoc, NS). That is, the maximal firing frequency induced in MED neurons during PRs following PPN stimulation was effected when using trains of 60 Hz.

Figure 5C is a graph of the mean ± SE of the peak depolarization induced by different frequencies of PPN stimulation using a 1-s train (ANOVA df = 117, F = 12.12, P < 0.001). At 10 Hz, the depolarization induced was 1.6 ± 0.3 mV, increasing significantly (post hoc, P < 0.01) at 30 Hz (3.1 ± 0.3 mV), further increasing at 60 Hz (4.6 ± 0.4 mV, post hoc, P < 0.01), but decreasing again at 90 Hz (3.7 ± 0.4 mV, post hoc, P < 0.05 compared with 60 Hz). These data show that the peak depolarization induced in MED cells was after stimulation using 60-Hz trains.

Pharmacological Properties

Because many PPN neurons are cholinergic and project to the pontomedullary reticular formation, we tested the effects of cholinergic agents on the manifestation of PRs in MED neurons. Micropressure application of the cholinergic agonist carbachol (30 μM) on MED neurons was found to induce long-lasting depolarization in all neurons tested (n = 29). Figure 6A shows the responses of a MED neuron (RMP −54 mV) following PPN stimulation (which elicited a PR) (6Aa) and after carbachol application (depolarization and prolonged activation) (6Ab), which lasted >30 s. Following superfusion with TTX to block sodium channels and the generation of APs, PPN stimulation, as expected, failed to induce a postsynaptic response on this MED cell (6Ac), but micropressure application of carbachol during TTX superfusion again induced a slow depolarization in this cell (6Ad), demonstrating that the effect of carbachol was directly on the membrane of this MED cell and not indirectly generated. Figure 6B shows that the PR induced in another MED cell (RMP −65 mV) by PPN stimulation (Ba) was similar to the long-lasting response induced by carbachol application (Bb), but superfusion (50 μM) of the muscarinic cholinergic antagonist SCOP before carbachol application (Bc) blocked the carbachol-induced depolarization in all cells tested (n = 8).

Fig. 6.

Pharmacological responses in PR MED cells. Aa: response of a 12-day MED cell following PPN stimulation (1 s, 60 Hz, 50 V) at RMP (−54 mV). b: Response of the same cell following micropressure application of carbachol (CAR) that lasted >30 s. c: Response of the same cell following PPN stimulation as in a, except following superfusion of tetrodotoxin (TTX) to block sodium channels. d: Response of the same cell following CAR application as in b, except after superfusion of TTX, showing that the CAR-induced depolarization was directly on the cell recorded. Ba: response of a 13-day MED cell after stimulation of the PPN showing a long-lasting depolarization and train of APs (PR). b: PR induced in the same cell following CAR application at RMP (−65 mV). c: Pretreatment with scopolamine (SCOP) superfusion blocked the CAR-induced PR shown in b. Calibration bars: vertical 10 mV, horizontal 1 s. Note that PPN stimulation-induced PRs may be due to activation of muscarinic cholinergic receptors on MED neurons, inducing long-lasting stable firing in these presumed, locomotion-related neurons.

A total of 21 NPR MED neurons also were tested using carbachol. Interestingly, only 1 of 21, or 5%, of NPR cells was depolarized by carbachol (as in Fig. 6), although this cell did not show a PR. On the other hand, 8 of 21, or 38%, of NPR MED neurons were hyperpolarized by carbachol, suggesting that cholinergic projections to this region exercise a push (PR depolarization)-pull (NPR hyperpolarization) on at least some MED neurons. PPN stimulation hyperpolarized 4 of 41, or 10%, of NPR cells, showing that superfusion may be more effective than stimulation in inducing responses in NPR neurons, perhaps due to a paucity of receptors that are more efficiently activated by superfusion. Most NPR neurons (12 of 21, or 57%) were not affected by carbachol or by PPN stimulation (37 of 41, or 90%).

Anatomical Characteristics

The present studies include only limited morphometric analysis of somatic size in well-injected neurons. These cells were measured as follows: dendrites were truncated at the base, and the area of each cell was measured by using an image analysis system (NIH Image 163). There was a statistically significant difference (ANOVA, df = 34, F = 11.59, P < 0.001) between well-injected PR (n = 17) and NPR (n = 19) MED cells in terms of cell area (PR cells, 568 ± 44 vs. NPR cells, 387 ± 32 μm2). There was a relatively small variability in cell area measures, suggesting the presence of a limited range of cell sizes within each population. Of course, recordings were carried out over an age range during which there is considerable growth; however, the population of injected PR and NPR cells spanned the full age range (12–21 days), and both populations were evenly distributed across these ages. When the measured neurons were divided according to age, PR cells <16 days (n = 9) had a mean area of 549 ± 36 μm2 compared with those ≥16 days, which had an area of 574 ± 42 μm2 (n = 8) (NS). A difference across age was also absent in NPR neurons (<16 days, n = 10, 352 ± 29 μm2 vs. ≥16 days, n = 9, 401 ± 27 μm2) (NS). Although there was a small numerical increase across age for both cell types, the difference was not statistically different. The variability in cell sizes suggests the presence of groups of cells of similar mean areas throughout this developmental period.

Although there was a difference in mean cell area between PR and NPR MED cells, Rin was not statistically different between cell types (PR 93 ± 50 MΩ vs. NPR 88 ± 49 MΩ, NS, Table 1). The variability in Rin was considerable within each population of cells, perhaps leading to the lack of statistical difference. Therefore, the marked difference in cell area was not reflected in a lower Rin in PR cells (in fact, Rin was numerically higher in PR than in NPR cells). Figure 7 shows representative examples of intracellularly injected PR (A) and NPR (B) cells. Figure 7C is a graph of mean cell areas for PR (n = 17) and NPR (n = 19) cells, showing the statistically significant difference noted above (P < 0.01).

Fig. 7.

Morphometry of PR and NPR MED cells. A: representative injected 15-day PR cell in NADPH diaphorase and avidin-biotin processed section. B: representative injected 18-day NPR cell in similarly processed section. Calibration bar for A and B: 50 μm. C: significantly different (**P < 0.01) mean cell areas of 17 PR (∼600 μm2) and 19 NPR (∼400 μm2) cells. D: PR MED cell input resistance (Rin) before and after PPN stimulation (left), and before and after CAR application under TTX (right). Note that Rin decreased following PPN stimulation and after CAR application in all cells tested. Note that PPN muscarinic inputs are preferentially directed at large, perhaps output, MED neurons, and channels in those neurons are opened (decreased Rin) by both PPN stimulation and CAR application. The effects of CAR were effected directly on the MED neurons, not indirectly, suggesting that PPN output to the powerfully modulates MED neurons.

A prominent feature of PRs in PnC neurons was an increase in Rin, suggesting that channels were closed during PRs (and were tentatively identified as potassium leak channels) (21). In the present study of MED cells, hyperpolarizing pulses were applied intracellularly to measure conductance before PPN or before carbachol application. After the stimulation or drug application, the membrane potential was depolarized, but the amplitude of hyperpolarizing pulses increased in amplitude when the membrane potential was manually restored to RMP by current injection, indicative of a decrease in Rin during PRs. For a population (n = 8) of cells tested, the Rin observed before PPN stimulation was 30 ± 2 MΩ compared with the Rin observed in the same cells after the membrane potential had been restored manually, 16 ± 2 MΩ. This difference was statistically significant (ANOVA, df = 15, F = 31.99, P < 0.001) (Fig. 7C). In a group of cells tested (n = 8) before carbachol application (note that all of these tests were performed in the presence of TTX to determine direct effects on the membrane of the MED cells recorded), Rin was 162 ± 25 MΩ, but it decreased to 122 ± 16 MΩ after carbachol application and during restoration of the membrane to RMP. This difference was statistically significant (ANOVA, df = 15, F = 32.59, P < 0.001). We should note that, in cells in which both PPN stimulation and carbachol application were determined, both treatments elicited a decrease in Rin, suggesting that channels were opened by both treatments in MED neurons. As mentioned above, there was wide disparity in the distribution of Rin values across MED cells.


Major Results

Our hypothesis that the PPN would influence MED neurons by inducing long-lasting PRs and that the organization of inputs to the MED would differ from that of inputs to the PnC was aptly demonstrated. While activation of PPN inputs to PnC and MED neurons both elicit PRs, the input to the MED is directed at large, perhaps output neurons, whereas the input to the PnC is directed at smaller, perhaps interneurons. Moreover, PRs in PnC neurons are mediated by the closing of potassium leak currents (leading to increased Rin), whereas PRs in MED neurons are induced by direct channel opening (leading to decreased Rin). These findings suggest that the PPN exerts a differential control over different descending targets, both of which are induced to respond with long-lasting stable firing rates. The implications of the main findings described herein are as follows: 1) PPN stimulation using medium-frequency stimulation (60 Hz) induced PRs with longer duration in about one-half (49%) of MED neurons compared with stimulation at lower or higher frequencies. These results are similar to those previously described in the decerebrate cat and rat slices and suggest the presence of a frequency-dependent effect of PPN projections to PnC (9, 21), as well as MED neurons. 2) PPN stimulation at ∼60 Hz was the most effective frequency range for eliciting the maximum firing rates (∼5 Hz) in PR cells following PPN stimulation (compared with 10, 30, or 90 Hz). This effect indicates that the maximal firing frequency induced following PPN stimulation was at frequencies known best to induce locomotion, suggesting a possible explanation for the long-known, but little understood need to use such frequencies when stimulating the mesopontine region to induce locomotion (8, 46). Recent studies on unanesthetized, freely moving rats revealed that most neurons in the region of the MED were more active during waking states, and most of these cells showed discharge bursts during movement (7). These authors concluded that MED neurons modulate spinal processes during phasic motor activity. Our results suggest that the PPN may modulate MED firing during waking states that, in turn, trigger locomotor events, but the activity induced is tonic, not phasic. Recordings within the region described herein as the MED may reveal such activity in the freely moving animal in response to changes in state induced by PPN activation. 3) The lack of significant developmental changes across the period studied indicates that these cells have reached a stable maturational level. However, PRs were induced preferentially on large, presumably reticulospinal MED neurons (∼600 μm2), not on smaller, presumably interneuron, MED neurons (∼400 μm2). 4) PRs were voltage dependent, being present at a range between −65 and −45 mV, as was evident in PnC neurons (21), suggesting a similar working range for the PR in both regions, despite the differences described herein. 5) PR cells did not show induction of the PR by depolarizing pulses or resetting of the PR by hyperpolarizing pulses, so that these responses are unlikely to be plateau potentials (22), again, similar to what was observed in PnC neurons, despite the fact that those PRs manifested channel closing rather than opening (21). 6) However, PRs could be reduced or blocked in all neurons by the muscarinic antagonist SCOP in neurons that were also depolarized by superfusion of the muscarinic agonist carbachol, suggesting that a muscarinic receptor is involved in generating PRs, especially in large MED neurons. 7) In MED cells, Rin decreased during PRs and during carbachol application, suggesting that these treatments opened, rather than closed (as in PnC neurons), membrane channels. 8) Some NPR MED neurons were hyperpolarized by carbachol superfusion, suggesting that descending PPN projections have differential effects on the population of smaller MED neurons.

Functional Implications

The PPN sends widespread projections throughout the pontomedullary reticular formation (41, 42, 45), including the anterior pontine region (PnO) (37). Injections of cholinergic agonists into a region called the “pontine inhibitory area,” located in the dorsal oral pontine reticular formation, induce a REM sleep-like state (1, 59), an effect thought to be mediated via muscarinic blockade of an outward, G protein-coupled, K+ current (48). Lesion of this area can produce REM sleep without atonia (20, 27), but these lesions may also damage neurons and axons responsible for REM sleep initiation and maintenance (44). Comprehensive electrophysiological studies on pontine reticular neurons have been performed (reviewed in Ref. 52), which found LTS burst and nonbursting neurons, some with a transient outward A current (similar to those reported herein for the MED). These neurons receive excitatory cholinergic input from the laterodorsal tegmental nucleus (23) and are depolarized by nicotinic and muscarinic agonists (18, 54), although some cells are hyperpolarized by muscarinic agonists (19).

Descending PPN projections to the posterior pontine region (PnC) may be involved in, for example, modulation of the SR (see comprehensive reviews on the SR for more information: Refs. 5, 29, 55). Briefly, the physiological significance of the SR is that it evolved as a defensive response because of the protective nature of the behavior pattern elicited, basically consisting of 1) eyelid closure (measured in the laboratory as the eye-blink reflex); 2) limb flexion (measured as increased activation of flexors and inhibition of extensors, an effect that allows the organism to begin locomotion by inactivating anti-gravity muscles); and 3) a transient increase in sympathetic output (measured as an increase in heart rate and respiration with the goal of providing more oxygen to muscles that would be activated if fight or flight is required) (5). The pathway for the SR, which includes giant neurons in the PnC, has been worked out (5). These giant cells make up only ∼1% of PnC neurons (29), but, when lesioned by excitotoxic agents injected into this region, the amplitude of the SR is reduced (31). PPN lesions have been reported to reduce prepulse inhibition of the SR (56), and auditory-responsive PnC neurons appear to be inhibited by cholinergic agonists (30), suggesting that the PPN modulates SR gating.

Descending projections from the PPN orthodromically activate almost one-half (47%) of MED cells at short latencies, most of which project to the spinal cord, i.e., are reticulospinal and are most likely the larger cells in the region (13). Electrical stimulation of the region of the MED, as well as injection of cholinergic agonists into this region, induced locomotion on a treadmill in decerebrate cats (12) and rats (28). The present results show that PPN stimulation induced PRs on the larger MED neurons, while inhibiting or not affecting smaller MED cells. Moreover, all MED neurons exhibiting PRs following PPN stimulation were depolarized by application of a cholinergic agonist (and the effect was blocked by a muscarinic antagonist). These results suggest that cholinergic PPN input to the MED may activate large reticulospinal neurons via PRs, thereby eliciting a lasting change in state in these cells as well. Because the MED is directly involved in inducing locomotion, the PPN input to this region is presumed to trigger arousal-related locomotor events.

In general, it is well accepted that the RAS, while controlling sleep-wake cycles and arousal, also modulates posture and locomotion, perhaps related to the execution of fight-or-flight responses. On the one hand, stimulation of various points in the posterior mesencephalon is known to induce controlled locomotion on a treadmill in the decerebrate cat and rat (46, 49). Such locomotion-inducing sites include the PPN (reviewed in Ref. 8). Their stimulation induces short-latency excitation of hindlimb motoneurons at disynaptic latencies (6, 45). On the other hand, stimulation of the PPN has been found to suppress muscle tone but induce stepping (i.e., initial suppression followed by activation) (32). The present findings of induction of PRs by 60-Hz PPN stimulation and similar activation of MED cells by carbachol suggest the possibility of a firing frequency-dependent mechanism, which can explain all of these observations, demonstrating that 1) with the use of amplitudes that are slowly ramped up at medium frequencies of stimulation (∼60 Hz), reticulospinal pathways for locomotion are “recruited” to produce alternation of flexor and extensor hindlimb nerves; and 2) using sudden-onset trains of high frequencies (>90 Hz), locomotion can be inhibited and muscle tone suppressed (15). The initial effect may be equated with a “reset” function, which interrupts ongoing activity, allowing the motor system to trigger or recruit subsequent activity without interference (8, 41). Further research is required before more conclusive statements can be made regarding the frequency dependency of such complicated functions; however, this type of control system would have great survival value when involved in fight vs. flight responses.


This study was supported by National Institute of Neurological Disorders and Stroke Grant NS-20246 and National Science Foundation Award 0237314.


  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.


View Abstract