Monoamines increase the excitability of spinal neurones in the neonatal rat by hyperpolarizing the threshold for action potential production

Abstract

During fictive locomotion in the adult decerebrate cat, motoneurone excitability is increased by a hyperpolarization of the threshold potential at which an action potential is elicited (V_th). This lowering of V_th occurs at the onset of fictive locomotion, is evident for the first action potential elicited and is presumably caused by a neuromodulatory process. The present study tests the hypothesis that the monoamines serotonin (5-HT) and noradrenaline (NA) can hyperpolarize neuronal V_th. The neonatal rat isolated spinal cord preparation and whole-cell recording techniques were used to examine the effects of bath-applied 5-HT and NA on the V_th of spinal ventral horn neurones. In the majority of lumbar ventral horn neurones, 5-HT (13/26) and NA (10/16) induced a hyperpolarization of V_th ranging from −2 to −8 mV. 5-HT and NA had similar effects on V_th for individual neurones. This hyperpolarization of V_th was not due to a reduction of an accommodative process, and could be seen without changes in membrane potential or membrane resistence. These data reveal a previously unknown action of 5-HT and NA, hyperpolarization of V_th of spinal neurones, a process that would facilitate both neuronal recruitment and firing.

During fictive locomotion evoked by brainstem stimulation in adult decerebrate cats, the membrane potential at which an action potential is initiated (V_th) in spinal motoneurones is hyperpolarized (Krawitz et al. 2001). This hyperpolarization of V_th occurred in every motoneurone examined and ranged from −1.8 to −26.6 mV with a mean change of −8.0 mV. It is known that V_th can be different amongst different neurones (Gustafsson & Pinter, 1984) and may become depolarized on a moment-to-moment basis because of accommodative effects on sodium channels (Kolmodin & Skoglund, 1958). However, because V_th hyperpolarization (i.e. lowering) during locomotion occurs for the very first action potential, it is not the result of a reduction in a spike accommodation process. Instead, it is likely that V_th is lowered through the actions of a neuromodulator released as part of the process initiating locomotion. This state-dependent modulation of the motoneurone threshold serves to facilitate motoneuronal recruitment and enhance motoneuronal output during fictive locomotion. Computer models suggest that rapid changes in membrane potential do not themselves account for the hyperpolarization of V_th (Dai et al. 2000). A reduction of potassium conductances and, particularly, enhancement of the activation or conductance of voltage-dependent sodium channels might mediate V_th hyperpolarization (Dai et al. 2002).

Monoamines are well known to exert multiple effects on components of the motor system in mammalian models (see Schmidt & Jordan, 2001 for review). At the cellular level, the monoamines serotonin (5-hydroxytryptophan; 5-HT) and noradrenaline (NA) produce depolarization of neurones (Connell & Wallis, 1988; Takahashi & Berger, 1990; Elliot & Wallis, 1992), facilitate the expression of plateau potentials (Hounsgaard et al. 1988), reduce the afterhyperpolarization (AHP) following an action potential (Madison & Nicoll, 1986) and enhance membrane oscillatory behaviour (MacLean et al. 1998). These cellular effects are mediated by alteration of channels and conductances such as: a facilitation of the slow inward rectifier current I_h (Wang & Dun, 1990; Takahashi & Berger, 1990; Kjaerulff & Kiehn, 2001), a low voltage-activated Ca²⁺ current (Berger & Takahashi, 1990), a persistent inward current (I_PIC; Lee & Heckman 1999), NMDA currents (MacLean & Schmidt, 2001) and an inhibition of a fast inward rectifier current I_KIR and possibly other ‘leak’ currents (Kjaerulff & Kiehn, 2001). Despite this large array of effects, monoamines have not previously been shown to modulate neuronal V_th. At the systems level, 5-HT and NA are able to elicit or facilitate locomotor output (Smith et al. 1988; Cazalets et al. 1992; Kiehn et al. 1992, 1999; Barbeau et al. 1993; Cowley & Schmidt, 1994) and alter spinal reflex activity (see, for example, Jankowska et al. 2000; Machacek et al. 2001).

The present study examined whether bath application of monoamines could alter V_th in the absence of locomotion. The goal of the present study was to determine the effect of bath applied 5-HT and NA on neuronal V_th in the isolated neonatal rat spinal cord preparation, and compare these effects to the hyperpolarization of V_th previously seen during fictive locomotion in the in vivo decerebrate cat. Portions of this work have been presented in preliminary form (Fedirchuk, 2001).

Methods

Experiments were conducted on spinal cords isolated from neonatal (postnatal days 1–5) Sprague-Dawley rats. Experiments were conducted in accordance with guidelines for the ethical treatment of animals issued by the Canadian Council on Animal Care and with the approval of the institutional protocol review committee. Once animals were anaesthetized with halothane in a chamber, they were rapidly decapitated, eviscerated and placed in an artificial cerebrospinal (aCSF) solution that was oxygenated with 95% O₂, cooled to 4°C and contained 125 mm NaCl, 2.5 mm KCl, 26 mm NaHCO₃, 1.25 mm NaH₂PO₄, 25 mmd-glucose, 1 mm MgCl₂ and 2 mm CaCl₂. Dorsal and ventral spinal laminectomy was done with fine scissors and the spinal cord caudal to the lower thoracic segments was isolated. The spinal cord was then moved to a recording chamber coated on the bottom with Sylgard, where it was initially pinned ventral side up, hemisected down the midline using an etched tungsten needle, and then pinned with the medial surfaces up. The spinal cord was then allowed to slowly warm to room temperature.

The ventral horns of the lower lumbar segments were then targeted for single-cell recording using glass microelectrodes filled with a solution containing 140 mm potassium gluconate, 0.2 mm EGTA and 10 mm Hepes, with KOH to bring the pH to 7.3. The filled electrodes had resistances ranging from 3 to 6 MΩ. The microelectrode was introduced from the medial surface of the ventral horn, because preliminary experiments showed that this arrangement was favourable for obtaining stable recordings from ventral horn neurones. A whole-cell single cell recording arrangement was obtained using the ‘blind patch’ technique. An Axopatch 1D microelectrode amplifier controlled with pCLAMP 7 software (Axon Instruments) was used for recording. Series resistance was monitored, was usually <30 MΩ, and was compensated only when in current-clamp mode.

V_th was measured in two ways. In current clamp, the V_th for eliciting an action potential could be directly measured from the voltage record as the membrane potential at the point of maximal change of voltage (inflection point) which was visually determined at the onset of an action potential evoked by a depolarizing ramp of current injection. In voltage-clamp mode, neurones were depolarized from an initial holding potential of −60 mV by applying 100 ms depolarizing steps that increased in 2 mV increments (except where noted below). Steps were delivered at a repetition rate of 2 Hz. Fast inward currents which would have mediated action potentials in current-clamp mode were evident on the recorded current trace, and the potential of the smallest depolarizing step capable of inducing a fast inward current was considered to be V_th. The initial holding potential of −60 mV was chosen in order to approximate the resting membrane potential of the neurone and reduce the possibility of the initial holding potential itself inducing activation of voltage-sensitive conductances.

Serotonin (5-HT) and/or noradrenaline (NA) were applied individually from 10 mm stock solutions to the small-volume (< 3 ml) stationary bath in concentrations that ranged from 2 to 50 μm (usually 10 or 12 μm). Washout of the drugs was accomplished by perfusing the bath with oxygenated aCSF at a rate of 5–10 ml min⁻¹. A low volume gas flow of 95% O₂−5% CO₂ was directed on to the recording chamber throughout the experiment.

Results

Measurement of V_th

Initial experiments compared current-clamp and voltage-clamp records from the same neurones. The monoaminergic effects on V_th were consistent in direction between cells in which V_th was measured both in current clamp and voltage clamp. However, V_th values determined using the voltage-clamp mode were more stable and faster to determine than those obtained in current-clamp mode, so the voltage-clamp technique was used to determine all V_th values reported in this study. Variability in the determination of the V_th using current clamp might be attributable to the known limitations of using a head stage primarily designed for voltage-clamp recordings in ‘current-clamp’ mode. Voltage traces are approximated and fast voltage transients, like the onset of an action potential, might be subject to distortion (see Magistretti et al. 1996). For 34 neurones recorded in 26 preparations, the absolute value of V_th determined using the voltage-clamp protocol ranged from −30 to −48 mV (mean =−38 mV; s.d.= 6 mV). There was no relation between the input resistance of the neurone (range 40–720 MΩ) and its absolute V_th (linear regression r² < 0.01, P = 0.97). Stopping the perfusion of the recording bath for periods up to 6 min did not induce a change in V_th(n = 3).

Serotonergic effects on V_th

The effect of bath-applied 5-HT of V_th was assessed for 23 lumbar ventral horn neurones. The minimum depolarizing step which elicited a fast inward current remained stable prior to application of 5-HT to the bath, and the membrane resistance and access resistance were monitored. Figure 1A shows an example of one cell in which a step to −40 mV from the initial holding potential of −60 mV was the smallest depolarizing step able to elicit a fast inward current (left traces, denoted by arrow) and was V_th for this neurone. Steps to more depolarized holding potentials invariably elicited a fast inward current. Repeated trials were obtained prior to application of 5-HT to ensure that the control V_th reflected a stable, repeatable value (not illustrated). Within 3 min of the addition of 12 μm of 5-HT to the bath, the first fast inward current was induced by a smaller depolarizing voltage step (to −46 mV). The difference indicates a −6 mV hyperpolarization of the V_th. This change of V_th was reversed by washing out the 5-HT (n = 7/7 cells; see Fig. 1C, left traces). A second administration of 5-HT produced the same change in the V_th (Fig. 1C, right traces).

Figure 1. Bath-applied 5-HT hyperpolarizes V_th.

Prior to application of 5-HT (A), depolarizing steps that increased in 2 mV increments were applied to a ventral horn neurone and the resulting whole-cell current recorded. The successive current records (upper traces) and voltage command steps (lower traces, 1–14 in A) are shown overlaid. A step to −40 mV (thicker voltage line) from a holding potential of μ60 mV was needed to elicit a fast inward current (indicated by arrow). After 3 min with 5-HT in the bath, a step to −46 mV (B, thicker voltage line) was sufficient to activate the fast inward current. This −6 mV hyperpolarization of V_th was reversed with the washout of 5-HT (C) and was repeatable with re-application of 5-HT (D). The scaling in A applies to all panels; leak currents are not subtracted.

In order to test whether the successive subthreshold depolarizing steps present in the incremental depolarizing step protocol might have induced an accommodative process that is 5-HT sensitive, in two cells the protocol was modified (see Fig. 2). The V_th was first determined using incremental depolarizing steps (as in Fig. 1; not illustrated in Fig. 2), then the protocol was customized so that the first depolarizing step delivered would equal the control V_th value for that cell. Successive steps decreased in amplitude by 2 mV decrements (see Fig. 2A, lower traces). As shown in Fig. 2B, 5-HT induced a hyperpolarization of V_th using the decreasing step protocol, in the absence of subthreshold depolarizing steps. The same 5-HT-induced hyperpolarization of V_th was observed for these neurones using the decreasing step as was seen using the incremental step protocol (not illustrated). This indicates that the V_th hyperpolarization is not caused by a serotonergic reduction of an experimentally induced accommodation.

Figure 2. V_th hyperpolarization is not caused by 5-HT reducing an experimentally induced accommodation.

In order to address the possibility that subthreshold depolarizing voltage steps may induce an accommodation which is 5-HT sensitive, an alternate protocol was used in which successive depolarizing voltage steps were of decreasing amplitude. The first step delivered was the minimum voltage step needed to elicit a fast inward current in the absence of 5-HT (denoted by arrow, step 1 in A). In the presence of 5-HT, a −44 mV depolarizing step was also able to evoke a rapid inward current (B), demonstrating a −4 mV hyperpolarization of V_th. The time scale in A also applies to B.

5-HT elicited a hyperpolarization of V_th in 16/23 neurones ranging from −2 to −8 mV (mean −4 mV). This V_th hyperpolarization was not accompanied by (1) a consistent change the amplitude of the fast inward current, (2) changes in the access resistance during the experiment, or (3) consistent changes in membrane resistance during 5-HT administration. V_th hyperpolarization with 5-HT administration was seen for neurones exhibiting small-amplitude fast inward currents (<500 pA) and neurones with large-amplitude fast inward currents (>1 nA). There was no relation between the absolute value of the control V_th and the effect of 5-HT on V_th (linear regression; r²= 0.01, P = 0.88), or the initial holding current required to maintain the cell at −60 mV and the effect of 5-HT on V_th (linear regression; r²= 0.02, P = 0.65).

The minimum concentration of 5-HT able to reproducibly hyperpolarize V_th was 2 μm, but titration of the 5-HT concentration (n = 4) showed that concentrations of 8–10 μm were required to elicit the maximal hyperpolarization of V_th. Therefore, for the majority of experiments concentrations of 10–12 μm were used. The preparations that exhibited a 5-HT-induced hyperpolarization of V_th ranged from postnatal day (P) 1 to P5, and there was no relation between the effect of 5-HT on V_th and the age of the neonatal rat from which the spinal cord was harvested (linear regression; r²= 0.11, P = 0.22).

Not every neurone recorded exhibited a 5-HT-induced hyperpolarization of V_th. Four of 23 ventral horn neurones showed no change in V_th in the presence of 5-HT (12–26 μm). The remaining 3/23 neurones showed a 2 or 4 mV depolarization of V_th in the presence of 5-HT (24 or 27 μm). One of the three preparations in which a neurone showed no change in V_th in the presence of 5-HT and one of the three preparations in which the V_th of a neurone was seen to depolarize also yielded recordings from a different neurone in which the V_th hyperpolarized in the presence of 5-HT. Therefore the responses of neurones that showed either no change or a depolarization of V_th were characteristic of the particular neurones rather than being determined by the experimental preparation.

Application of 5-HT to the extracellular solution did not produce consistent effects on the input resistance amongst different neurones. For individual cells, the input resistance could decrease, increase or remain unchanged with the application of 5-HT. The baseline current required to maintain the cell at the initial −60 mV holding potential became more negative during 5-HT application in 13/23 neurones (a change from −10 to −30 pA). For these neurones, a membrane depolarization would have occurred had the membrane potential not been held at −60 mV, which is consistent with previous reports documenting a 5-HT-induced depolarization of spinal motoneurones (see Connell & Wallis, 1988; Takahashi & Berger, 1990; Elliot & Wallis, 1992). It is notable that cells showing hyperpolarization of V_th in the presence of 5-HT did not necessarily show concomitant changes in either their membrane resistance or baseline holding current. In addition, one of the neurones that had no change in V_th in the presence of 5-HT did show a −30 pA change in holding current and a reduction in membrane resistance during the drug application. Rhythmic fluctuations in membrane current during voltage-clamp protocols, or of membrane potential during current-clamp protocols, which would have denoted rhythmic network activity, were not observed for any neurone.

Noradrenergic effects on V_th

The effect of bath-applied NA on neuronal V_th was assessed in 16 ventral horn neurones. Of these, 10/16 showed a hyperpolarization of V_th (range −2 to −6 mV), 4/16 neurones showed a depolarization of V_th of either 2 or 4 mV, and two neurones showed no change in V_th when NA was present in the bath (concentrations ranged from 6 to 50 μm). The time course of the NA effect on V_th was the same as for 5-HT. An alteration of V_th occurred within 2–3 min and could be washed out within several minutes (n = 9/9). The effects of 5-HT and NA on V_th were compared in five neurones. As in the example shown in Fig. 3, the effect on V_th of the first drug was assessed (in this case 5-HT), the preparation was washed and it was confirmed that the V_th returned to its control value, and the second drug was applied. For 3/5 cells, the change in V_th induced by NA was identical to that produced by 5-HT (−4, −2 and 2 mV changes). Of the other two cells, one showed a 4 mV depolarization of V_th with 5-HT and a 2 mV depolarization with NA and the other showed no change in V_th with 5-HT and a −2 mV hyperpolarization of V_th with NA. In one cell, after NA and 5-HT were each applied separately and their effects on V_th assessed (both caused a −4 mV change), they were re-applied simultaneously. The effect on V_th was the same as that produced by either agonist on its own (i.e. −4 mV change).

Figure 3. Both 5-HT and NA can hyperpolarize V_th.

Application of 5-HT induced a −4 mV hyperpolarization of V_th (B) compared to control (A). Following washout of the 5-HT effect (C, 6 min of wash), bath-applied NA also elicited a −4 mV hyperpolarization of V_th. Simultaneous application of 5-HT and NA also produced a −4 mV hyperpolarization of V_th (not illustrated). The scaling in A applies to all panels.

Discussion

The results of this study show that bath-applied 5-HT or NA can alter the V_th of spinal ventral horn neurones in the neonatal rat. The most common effect, a hyperpolarization of V_th ranging in amplitude from −2 to −8 mV, was seen for the majority of neurones tested. This change was reversible and repeatable, was not due to a monoaminergic reduction of an experimentally induced accommodation, and was not accompanied by consistent changes in the amplitude of the fast inward current underlying spiking.

The hyperpolarization of V_th induced by 5-HT or NA was similar to the hyperpolarization of V_th seen during fictive locomotion in the cat (Krawitz et al. 2001) in that it was evident within minutes of application of the drug to the bath, and recovered within minutes of washout of the monoamine. The onset of V_th hyperpolarization was slower in this study than that seen during fictive locomotion in the cat, where hyperpolarization of V_th occurred within seconds of electrical brainstem stimulation and was evident at the onset of locomotor activity. It is probable that the time required for diffusion of the drug into the spinal cord following bath application, and the delay associated with clearing effective doses of 5-HT or NA from the spinal tissue during washout, account for the slower onset and recovery observed in the present study. In addition, the relatively lower incidence and smaller amplitude of V_th hyperpolarization seen in the present study might be due to the inability of the exogenously applied monoamines to selectively activate the receptors mediating the change in V_th. We have seen that activation of endogenous serotonergic systems in neonatal rat brainstem/spinal cord preparations can induce both a higher incidence of V_th hyperpolarization, and larger hyperpolarizations of V_th than reported here (Gilmore & Fedirchuk, 2002).

The fact that hyperpolarization of V_th was not limited to neurones exhibiting particular postsynaptic responses in the presence of the monoamine (e.g. induction of negative holding current at −60 mV) suggests that the V_th hyperpolarization does not depend on a neuronal depolarization. In our previous computer modelling study, the putative modulatory process that was most effective in inducing a hyperpolarization of V_th without concomitant changes in action potential shape was the modulation of the activation profile of the fast sodium current underlying action potentials (Dai et al. 2002). The modulation of the amplitude and inactivation profile of sodium channels via phosphorylation has been documented (West et al. 1991) and a role for this modulatory process in mediating neuronal plasticity has been suggested (see Cantrell & Catterall, 2001). It is therefore possible that a modulatory process facilitating activation of Na⁺ channels might underlie the monoaminergic hyperpolarization of neuronal V_th observed in the present study.

In addition to modulation of the fast sodium current underlying spiking, it is also possible that other channel types could be involved in the monoamine-induced V_th hyperpolarization. Reducing a potassium conductance could also hyperpolarize V_th, although to a lesser degree than direct manipulation of sodium channels (Dai et al. 2002). In addition, persistent inward currents mediated by calcium and sodium channels are activated at membrane potentials near or even below spike threshold (Lee & Heckman, 2001; Li et al. 2004). Therefore monoaminergic facilitation of persistent inward currents (Lee & Heckman, 1999), or the NMDA current (MacLean & Schmidt, 2001), might cause a contribution of these currents to spike initiation and also contribute to V_th hyperpolarization.

Although the cells recorded in the present study were unidentified lumbar ventral horn neurones, the fact that hyperpolarization of V_th could be seen in cells having either low or high membrane resistence values suggests that the V_th of both larger and smaller ventral horn cells can be hyperpolarized by monoamines. The hyperpolarization of V_th of interneurones would facilitate their activation and might contribute to the previously described ability of monoamines to initiate locomotor activity and facilitate spinal reflexes in the neonatal rat. Monoamines did not evoke rhythmic activity in the present study, probably because of disruption of spinal networks by hemisection. However, it is possible that neural elements presynaptic to the recorded neurone may have been activated by the applied monoamines, and therefore other transmitter systems might contribute to the 5-HT- and NA-induced effects on V_th that were observed. The present study demonstrates the ability of 5-HT and NA to induce V_th hyperpolarization, and the utility of the in vitro neonatal rat preparation for examining the mechanisms underlying this modulatory process. Future studies will assess the potential involvement of other transmitter systems in the monoamine-induced V_th modulation, and the relative susceptibility of motoneuronal and interneuronal populations to V_th hyperpolarization.