Skip to main content
NCBI home page
Journal of Neurophysiology logoLink to Journal of Neurophysiology
. 2017 Nov 15;119(3):786–795. doi: 10.1152/jn.00219.2017

Developmental changes in spinal neuronal properties, motor network configuration, and neuromodulation at free-swimming stages of Xenopus tadpoles

Stephen P Currie 1, Keith T Sillar 1,✉
PMCID: PMC5899306  PMID: 29142093

Abstract

We describe a novel preparation of the isolated brain stem and spinal cord from prometamorphic tadpole stages of the South African clawed frog (Xenopus laevis) that permits whole cell patch-clamp recordings from neurons in the ventral spinal cord. Previous research on earlier stages of the same species has provided one of the most detailed understandings of the design and operation of a central pattern generator circuit. Here we have addressed how development sculpts complexity from this more basic circuit. The preparation generates bouts of fictive swimming activity either spontaneously or in response to electrical stimulation of the optic tectum, allowing an investigation into how the neuronal properties, activity patterns, and neuromodulation of locomotor rhythm generation change during development. We describe an increased repertoire of cellular responses compared with younger larval stages and investigate the cellular-level effects of nitrergic neuromodulation as well as the development of a sodium pump-mediated ultraslow afterhyperpolarization (usAHP) in these free-swimming larval animals.

NEW & NOTEWORTHY A novel in vitro brain stem-spinal cord preparation is described that enables whole cell patch-clamp recordings from spinal neurons in prometamorphic Xenopus tadpoles. Compared with the well-characterized earlier stages of development, spinal neurons display a wider range of firing properties during swimming and have developed novel cellular properties. This preparation now makes it feasible to investigate in detail spinal central pattern generator maturation during the dramatic switch between undulatory and limb-based locomotion strategies during amphibian metamorphosis.

Keywords: development, locomotion, neuromodulation, nitric oxide, Xenopus

INTRODUCTION

The hatching-stage tadpole of the frog Xenopus laevis (stage 37/38) has one of the most completely described motor control systems of any vertebrate. This has largely been due to the development of a preparation enabling patch-clamp recordings from pairs of synaptically coupled spinal neurons, which has allowed the spinal network for locomotion to be understood in cellular and synaptic detail (Roberts et al. 2010). Furthermore, research on the early larval stages—just a day or so later in development but still before continuous free swimming begins—has provided several insights into how neural networks are modified to enable complex behavior to emerge as ontogeny progresses (Sillar et al. 1991, 1992; Zhang et al. 2009, 2011). Here we describe a novel preparation that forms the foundation for the next steps in the study of motor control in in vitro Xenopus by enabling whole cell patch-clamp recordings from brain stem and spinal cords isolated from free-swimming prometamorphic (stages 50–58; Nieuwkoop and Faber 1956) tadpoles. At these stages, the tadpoles swim almost continuously, sculling with the caudal portion of their tail in order to retain a head-down hover, facilitating their lifestyle as obligate filter feeders (Hoff and Wassersug 1986). At earlier stages of Xenopus development, swimming occurs in prolonged bouts following sensory stimulation and all neurons within the spinal network fire rhythmically throughout the swim cycle to maintain the coordinated muscle contraction underlying forward propulsion (Roberts et al. 2010). The aim of this study was to explore changes in the properties and modulation of spinal neurons that might account for the development of more flexible swimming behavior and the switch from a primarily sessile existence to one in which swimming occurs almost constantly (Currie et al. 2016). For instance, an important intrinsic property recently described in spinal neurons from earlier stages is the activity-dependent ultraslow afterhyperpolarization (usAHP) (Zhang et al. 2015; Zhang and Sillar 2012), which results from increased activation of the sodium pump. The usAHP is thought to act as a form of internal memory for previous cellular activity. At early stages of development the usAHP is detectable in ~50% of spinal locomotor neurons, but whether this is a transient feature during early development or one that persists later in development is not known. Moreover, the effects of nitric oxide (NO), which is known to potently modulate fictive locomotion at these later stages (Currie et al. 2016), have not been investigated at the cellular level. Since the effects of endogenous NO at these stages are location specific, complex, and different from earlier stages, having switched from an inhibitory to an excitatory influence (Currie et al. 2016), NO’s direct effect on spinal neurons is an important next step in understanding the role of nitrergic neuromodulation in this developing system.

METHODS

Animals and husbandry.

Experiments were performed on free-swimming, prometamorphic stages (50–58) of the South African clawed frog, Xenopus laevis (Nieuwkoop and Faber 1956). Animals were obtained by human chorionic gonadotropin hormone (1,000 U/ml; Sigma)-assisted matings of adults selected from an in-house breeding colony. Fertilized ova were collected and reared in enamel trays until the first free-feeding stages, before being transferred to standard glass aquarium tanks. All procedures complied with the UK Animals (Scientific Procedures) Act 1986 and the European Community Council directive of 24 November 1986 (86/609/EEC) and were approved by the University of St. Andrews Animal Welfare Ethics Committee.

Dissection for whole cell recording.

To make single cell patch-clamp recordings in the ventral spinal cord of prometamorphic Xenopus tadpoles, the dissection employed during extracellular experiments (Combes et al. 2004; Currie et al. 2016) was modified. Briefly, after removal and destruction of the brain under MS222 anesthesia, the remaining nervous system was cut free from the rest of the body. The caudal extent of the spinal cord was cut to completely isolate the brain stem and spinal cord, and this tissue was then pinned down securely with sharpened tungsten wire pins in a recording chamber with a rotatable Sylgard platform. With a finely etched tungsten needle, the spinal cord was then opened along the medio-lateral midline as far as the neurocoel from approximately the 12th to 15th postotic ventral roots. At the caudal extent of this first cut, a second cut was made perpendicularly, approximately as deep as the dorso-ventral midline of the spinal cord and all the way to the lateral extent of the cord. The free end of the spinal cord was then carefully peeled back toward the rostral end of the animal, removing most or all of the dorsal horn. Once at the level of the 12th postotic ventral root the dorsal portion of spinal cord was cut away with microscissors (see Fig. 1B). This modified preparation gave direct access to the ventral spinal cord, where the spinal circuits involved in motor control are presumed to lie. Moreover, exposing the ventral cord between the 12th and 15th ventral roots gave access to neurons assumed to be primarily involved in axial swimming patterns since they are located caudal to the ventral roots innervating the developing hindlimbs.

Fig 1.

Fig 1.

Open in a new tab

Whole cell patch-clamp recordings during fictive locomotor activity. A: cartoon of a stage 54/55 prometamorphic Xenopus laevis tadpole. Drawing kindly provided by Laurence D. Picton, with permission. Approximate location of patch-clamp recordings is indicated. B: schematic of isolated brain stem-spinal cord preparation for whole cell recording. The expanded section illustrates the site of recording with overlying dorsal spinal cord removed. Ci: whole cell patch-clamp and ventral root (vr) recordings of a spontaneous episode of fictive locomotion from a stage 55 tadpole. Bottom: the firing pattern on an expanded timescale to highlight the transitions between tonic and rhythmic firing during the episode. Cii: as in Ci but for an episode evoked by brief electrical stimulation of the optic tectum. Right: the waxing and waning of activity during the episode, including the recruitment and derecruitment of the spinal neuron—expanded from gray box on left. *Stimulus artifact.

Electrophysiology.

The dissection and subsequent electrophysiological recordings were performed in HEPES saline (composition in mM: 115 NaCl, 3 KCl, 2 CaCl2, 2.4 NaHCO3, 1 MgCl2, 10 HEPES, adjusted with 4 M NaOH to pH 7.4). Whole cell patch-clamp recordings in current-clamp mode were made with microelectrodes pulled on a Sutter P97 pipette puller from borosilicate glass capillaries (Harvard Apparatus). All recordings were made from neurons at relatively ventral locations in the spinal cord, between the 11th and 16th postotic ventral roots. Patch pipettes were filled with 0.1% Neurobiotin in the intracellular solution (composition in mM: 100 K-gluconate, 2 MgCl2, 10 EGTA, 10 HEPES, 3 Na2ATP, 0.5 NaGTP, adjusted to pH 7.3 with KOH) and had resistances of ~8 MΩ. Extracellular ventral root recordings were made ipsilaterally to the exposed area for patch recording and generally caudally between the 16th and 19th postotic ventral roots. Recordings in whole cell mode were amplified with an Axoclamp 2B (Axon Instruments) amplifier and digitized with a CED power1401. All signals were displayed and saved on a PC with Spike2 software and all subsequent analysis performed in Dataview software (v8.62, courtesy of W. J. Heitler, School of Biology, University of St. Andrews).

Neuron labeling.

Neurobiotin (0.1%) in the intracellular solution was used to label neurons for anatomical identification. After electrophysiological recordings, the brain stem-spinal cord tissue was fixed in 2% glutaraldehyde in 0.1 M phosphate buffer (pH 7.2) overnight in the refrigerator (4°C). After they were rinsed with 0.1 M PBS (120 mM NaCl in 0.1 M phosphate buffer, pH 7.2), the animals were 1) washed in two changes of 1% Triton X-100 in PBS for 15 min with agitation, 2) incubated in a 1:300 dilution of extravidin peroxidase conjugate (Sigma-Aldrich) in PBS containing 0.5% Triton X-100 for 2–3 h with agitation, 3) washed again in at least four changes of PBS, 4) presoaked in 0.08% diaminobenzidine in PBS (DAB solution) for 5 min, 5) moved to a second container with 0.075% hydrogen peroxide in DAB solution for 5 min, and 6) washed in running tap water. The nervous system was then dehydrated, cleared in methyl benzoate and xylene, and mounted whole between two coverslips with DEPEX. Neuronal anatomy was observed with a Zeiss Axio Imager Ax10 at ×20 and ×40 magnification, and measurements were made with Zen Imaging Pro (v10; Zeiss) software.

Pharmacological manipulations.

Saline during electrophysiological experiments was gravity fed from one of two stock chambers. This allowed switching between control and drug conditions via a three-way tap. Saline flowed to waste and was not recirculated. Drugs used were the NO donors S-nitroso-N-acetyl-penicillamine (SNAP, 200 μM) and diethylamine NONOate (DEA-NO, 50–200 μM). No obvious differences in drug effect were seen over this concentration range of DEA-NO (cf. Currie et al. 2016). Drugs were dissolved in distilled H2O (18 MΩ), divided into aliquots, and then frozen, before being made up to final concentration in standard HEPES saline; the dilution of saline with H2O vehicle was <0.5%.

Statistical analysis.

Mean data were analyzed with a paired t-test, repeated-measures ANOVA with Bonferroni correction, or, in the case of normalized data, a Wilcoxon signed-rank test, and significance is reported at either <0.01 or <0.05. Error bars represent SD. For the cumulative probability plots of postsynaptic potential (PSP) frequency and amplitude, the Kolmogorov-Smirnov test (K-S test) was employed at a significance level of 0.05. The bin sizes were standardized throughout at 0.1 Hz for PSP frequency and 0.05 mV for PSP amplitude. Statistical analyses were performed in SPSS (version 21; IBM), graphs were produced from custom-written MATLAB scripts (MathWorks), and figures were arranged in Adobe Illustrator CC (Adobe Systems).

RESULTS

A major aim of this study was to explore developmental changes in neuronal, circuit, and neuromodulation properties that accompany the transition to more flexible and continuous tadpole swimming behavior compared with embryonic and early larval stages. To do this we have developed a new preparation that allows for whole cell patch-clamp recordings of neurons within the spinal cord at free-swimming prometamorphic stages (50–58; Nieuwkoop and Faber 1956) of Xenopus tadpoles (Fig. 1, A and B). At these stages of development, bouts of fictive swimming readily occur spontaneously in in vitro preparations of the spinal cord and brain stem (Combes et al. 2004; see Fig. 1Ci), but similar bouts can also be evoked by brief electrical stimulation of the optic tectum (Fig. 1Cii). In both cases, the activity of individual rhythmically active neurons can be recorded and assessed simultaneously with this new preparation.

Neuronal activity during swimming.

The data reported in this study derive from 104 neurons [9 identified anatomically as motor neurons (MNs); see below] recorded in 83 preparations. During spontaneous episodes of swimming, rhythmically active spinal neurons generally fired action potentials in phase with the bursts of activity recorded from ipsilateral spinal ventral roots [Fig. 1C; 50/83 (60.2%) neurons where activity in a ventral root was present for comparison]. In a typical neuron, the onset of ventral root activity coincided with or was just preceded by a depolarization of the membrane potential, which then oscillated in phase with the rhythmic network activity recorded in the ventral root. Preceding the onset of rhythmic activity, 12/83 (14.5%) neurons initially fired tonically and the ventral root recording displayed a corresponding period of tonic discharge (see Fig. 1Ci). As the ventral root began to burst rhythmically these cells also switched into a rhythmic pattern of firing, with volleys of action potentials interspersed with periods of subthreshold activity (Fig. 1, Ci, middle inset, and Cii, inset). Over the course of an episode, the activity often waxed and waned and the neuron was often derecruited but continued to receive rhythmic synaptic drive in time with the ventral root bursts (see Fig. 1, Cii, inset). Rhythmic activity was often followed by tonic ventral root discharge and neuronal firing (Fig. 1Ci) or faded away gradually with sporadic spiking activity (Fig. 1Cii). While the majority of recorded neurons fired transiently and only during periods of network activity, a subset [7/104 (6.7%)] discharged tonically at rest (Fig. 2). However, these neurons were also apparently linked to locomotor output, since during network activity their firing was often altered. The changes in firing pattern could be quite subtle, as in the modulation of tonic firing frequency seen in Fig. 2A, or they could be more dramatic, switching in and out of a rhythmic firing pattern phase-related to the locomotor cycle, as in Fig. 2B.

Fig. 2.

Fig. 2.

Open in a new tab

Neurons with a tonic firing pattern. A: ventral root and whole cell patch-clamp recordings from a stage 56 tadpole. This neuron modulates its firing rate during spontaneous episodes of motor activity (bottom). At the beginning of the episode when the ventral root activity is highest the neuron fires at 9 Hz, while just before the episode finishes this has dropped to only 5.5 Hz. B: as in A, although this cell has an even more dramatic increase in firing rate at the beginning of the evoked episode of motor activity. During the episode both neuron and ventral root burst rhythmically (see bottom) *Stimulus artifact.

Basic firing properties.

The average resting membrane potential of all recorded neurons was −57.8 ± 6.67 mV (n = 104 from N = 83 animals). From rest, the injection of a short (2 ms) current pulse was capable of eliciting an action potential with a mean threshold for activation of 487.34 ± 390.89 pA. All but one recorded neuron (see Fig. 3Ci) was capable of firing repetitively during current injection and fired at higher frequency as the amplitude of the current pulse was increased (Fig. 3A). The frequency-current plots in Fig. 3Aiii illustrate representative examples of both a lower-threshold (also see Fig. 3Ai) and a higher-threshold (also see Fig. 3Aii) neuron. Both of these neurons were later identified as MNs (see below for details).

Fig. 3.

Fig. 3.

Open in a new tab

Firing properties of spinal neurons. Ai: example of typical low-threshold neuron response to current injection at 1.1× and 2× rheobase (Th). Aii: same as Ai but for higher-threshold neuron. Aiii: scatterplot of current pulse vs. firing frequency for the 2 neurons depicted in Ai (open circles) and Aii (filled circles). Bi: neuron displaying intrinsic oscillations following current injection at its firing threshold and at 2× rheobase. Bii: the same neuron during an episode of evoked swimming. Inset: part of trace within gray box on an expanded timescale. ***Multiple stimulation artifacts. Ci: example of a presumptive dIN firing a single spike during an episode of swimming. Cii: a spike from the same neuron evoked via current injection (black trace) with a similar evoked spike from the neuron in Aii superimposed (gray trace). Ciii: response of this neuron to current injection at its firing threshold and at 4× rheobase.

A novel property found in a small proportion of unidentified neurons in prometamorphic tadpoles was an apparently intrinsic rhythmogenic capacity (Fig. 3Bi). Thus in 2/104 neurons (2%), depolarizing current injection resulted in an oscillation of the membrane potential, with superimposed bursts of action potentials interspersed with periods where the membrane potential was repolarized below spike threshold. The intrinsic membrane oscillation was relatively slow, in the range of 0.5–1 Hz, in comparison to the membrane oscillations associated with swimming, which are typically 4–6 Hz. The firing pattern was similar to the intrinsic bursting seen in low-threshold zebrafish MNs that are recruited during the slowest swimming speeds (Gabriel et al. 2011; Menelaou and McLean 2012), but in Xenopus this has never been documented in publications based on many thousands of recordings at the embryonic and early larval stages of tadpole development. The neurons had relatively low rheobases of 110 pA and 310 pA, respectively, and both fired rhythmically during swimming (Fig. 3Bii).

As at early larval stage 42 (Sillar et al. 1991), neurons fired variably during episodes of fictive swimming and could fire multiple spikes during each motor burst (see Fig. 1Cii, inset). The exception to this was a single neuron that had a relatively depolarized resting membrane potential of −50 mV and during swimming fired one broad action potential per cycle (Fig. 3C, i and ii; for comparison Fig. 3Cii also shows the MN from Fig. 3Aii). Moreover, injection of suprathreshold current was unable to elicit repetitive firing, even at 400% of the rheobase (Fig. 3Ciii). This neuron therefore displayed physiological characteristics reminiscent of an embryonic descending interneuron (dIN; Li et al. 2006; see discussion).

Recording from motor neurons.

After patch-clamp recordings, Neurobiotin originating from within the patch solution allowed post hoc analysis of the anatomy of a subset of individual spinal neurons (see, e.g., Fig. 4). Definitive anatomical identification of spinal neurons was not possible in the majority of cases, in part because of the quality of the fills but also because of the lack of conformity of successful stainings with the well-characterized spinal neurons from earlier in Xenopus development. Nevertheless, 9/101 (8.9%) filled neurons were confirmed as MNs on the basis of their axonal projections that exited the spinal cord via a ventral root (Fig. 4A, i and iii). The MNs had a medially located soma and generally had dendrites that projected laterally into the marginal zone (Fig. 4Aiv). The primary axon exited the soma and initially ran close to the midline of the spinal cord (Fig. 4Aiv). The axon projected ipsilaterally and caudally over several spinal segments (Fig. 4A, i–iii) before turning and exiting via a ventral root (Fig. 4A, i and iii). All identified MNs projected caudally from the soma and exited via a ventral root between four and nine spinal segments away (mean = 6.33 ± 1.58 segments). Occasionally, physiological characterization of MNs was also possible (Fig. 4B). In 3/104 recorded neurons, including 1 in addition to the 9 anatomically identified MNs, each action potential following suprathreshold current injection was matched 1:1 by an impulse in the ventral root trace (Fig. 4Bi). Further confirmation of MN identity was provided by stimulating the ventral root in these preparations, which elicited antidromic spikes in the recorded neuron (Fig. 4Bii). The spikes occurred reliably after stimulation and at a very short latency (<2 ms), confirming their antidromic nature. The identified MNs had a mean resting membrane potential of −59 ± 5.55 mV and a mean rheobase of 408.57 ± 190.39 pA after a 2-ms current pulse.

Fig. 4.

Fig. 4.

Open in a new tab

Recording from motor neurons (MNs). A: Neurobiotin fills of 2 MNs from prometamorphic tadpoles. MNs were primarily identified on the basis of their long descending axon (arrow in ii and iv) that spanned several spinal segments (numbered in i) and exited the spinal cord via ventral root (* in i and iii); soma position in i is marked with arrow. Arrowhead in iv highlights dendritic arborizations around the soma, and the midline is traced with a dashed line in all panels. Unlabeled scale bars represent 50 μm. Bi: ventral root and whole cell recording of a different MN after current injection in a stage 56 tadpole. Each spike in the neuron is represented 1:1 in the ventral root trace, indicating the neuron projects to that root. Bii: an antidromic spike (bottom) from the same neuron following brief electrical stimulation of the recorded ventral root (top; *stimulation artifact).

Postspike after hyperpolarizations.

The output of most neural circuits changes dramatically during development to accommodate maturation of the behaviors they regulate. This is due partly to alterations in the synaptic connections between and the electrical properties of the constituent neurons. With regard to the latter, postspike hyperpolarizations are a defining feature of the responses of many neurons to excitatory inputs. In Xenopus tadpoles, action potentials in spinal neurons recorded at embryonic stage 37/38 are characterized by a fast (f)AHP, after which the membrane potential typically returns to rest within a few milliseconds (Sautois et al. 2007). A fAHP persists at prometamorphic stages and was found in all recorded neurons (Fig. 5Ai). In addition, a subset of neurons at prometamorphic stages [15/104 (14%)] displayed a pronounced slow (s)AHP following an action potential evoked from rest, which typically lasted 150–200 ms [Fig. 5Ai; mean 161.29 ± 59.99 ms (n = 15)]. This sAHP resembled similar responses documented in other species, which are mediated by apamin-sensitive Ca2+-dependent K+ channels (for review see Sah and Faber 2002). However, the sAHP reported here seems to be masked during rhythmic bursting, where only fAHPs were obvious (Fig. 5Aii).

Fig. 5.

Fig. 5.

Open in a new tab

Afterhyperpolarizations in spinal neurons. Ai: whole cell recording at rest with 2-ms suprathreshold current pulse. After an action potential, both fast (f) afterhyperpolarization (AHP) and slow (s) AHP are visible (see inset on expanded timescale). Aii: the same neuron during motor activity, highlighting the lack of any sAHPs despite the clear presence of fAHPs. Bi: whole cell recording at rest with 2-s suprathreshold current pulse. Following the train of action potentials there is an ultraslow (us) AHP lasting several seconds (see insets for expansion of the first few seconds). Bii: the same neuron and a ventral root recording during spontaneous motor activity. After membrane repolarization at the termination of swimming there is an usAHP lasting several seconds (see inset for expansion of first few seconds).

At earlier stages of Xenopus development (Zhang and Sillar 2012), intense firing activity in spinal neurons triggers an even longer-duration AHP (~60 s), termed the ultraslow (us)AHP. At prometamorphic stages an usAHP was evident in response to the injection of suprathreshold current pulses (Fig. 5Bi) and also after the termination of episodes of rhythmic swimming (Fig. 5Bii). With the same stimulus paradigm as Zhang and Sillar (2012)—a train of increasing suprathreshold current pulses (Fig. 6Ai)—a direct comparison of the responses of spinal neurons at stages 37/38–42 with those at stages 50–58 was possible. The usAHP was detectable in a far higher percentage of recorded neurons [81/93 (87%), including 8/9 (89%) identified MNs] at prometamorphic stages (50–58) compared with stages 37/38–42 [87/202 (43%), including 39/67 (58%) identified MNs; Fig. 6Aii and see Zhang and Sillar 2012]. The amplitude of the usAHP, measured as the change in membrane potential between rest and the peak slow hyperpolarization following current injection, was similar between the two stages of development. On average, a hyperpolarization of 4.84 ± 2.64 mV occurred at stages 37/38–42 (N = 20; Zhang and Sillar 2012), while during prometamorphosis the hyperpolarization was 4.66 ± 2.6 mV (N = 25; not significant; Fig. 6Aii). The duration of the usAHP, measured as the time between the end of the stimulus train and the point where the membrane potential returned to rest, was significantly shorter at prometamorphic stages. On average, the usAHP duration measured 19.33 ± 24.50 s in neurons from prometamorphic stages (N = 25; stages 50–58), while at embryonic and early larval stages the usAHP duration was 50.78 ± 34.21 s (N = 20; P < 0.01; Fig. 6Aii and see Zhang and Sillar 2012).

The usAHP described in stage 37/38–42 tadpole spinal neurons is thought to be mediated solely via an increase in the activity of the Na+-K+ pump, and, as such, the membrane hyperpolarization is not associated with a detectable change in membrane input resistance (IR) (Zhang and Sillar 2012). During prometamorphosis the IR of spinal neurons was reduced, but only during the first few seconds of an usAHP (Fig. 6B). At 500 ms and 2,000 ms after the end of the stimulus the IR was reduced significantly to 89.45 ± 8.61% and 94.30 ± 7.35% of control, respectively (N = 9; Fig. 6Bii). The change in IR cannot account for the complete recovery of the membrane potential to rest, however, since it returned to control levels (100%) in just 6.67 ± 7.35 s, which was significantly shorter than the duration of the usAHP to the same stimulus, which measured 18.58 ± 15.33 s (N = 9; P < 0.05, t-test; Fig. 6Biii). This suggests that the usAHP comprises both a long-duration Na+ pump-based event lasting the duration of the usAHP and a superimposed event, likely caused by the opening of a membrane ion channel that is active in the early stages of the usAHP.

Fig. 6.

Fig. 6.

Open in a new tab

Development of the usAHP. Ai: whole cell recording during 10-pA current steps from −50 pA to 260 pA (protocol was stopped after 20th suprathreshold current step). The protocol (originally used in Zhang and Sillar 2012) drives an usAHP in the neuron, and basic parameters can be measured. Aii: direct comparison of usAHP between stages 37/38–42 (N = 25; Zhang and Sillar 2012) and prometamorphic stages 50–58 (N = 20). Bi: whole cell recording with hyperpolarizing steps both before and after driving an usAHP (note that action potentials are truncated; see expansion for details of protocol). Bii: graph of input resistance (IR) at successive hyperpolarizing steps following the usAHP relative to IR at rest. Biii: graph of mean time taken for IR to return to baseline levels and mean duration of the usAHP to the same protocol (N = 9) Ci: whole cell recording showing a typical response to successive hyperpolarizing steps from rest. Membrane sag and postinhibitory rebound (PIR) are highlighted. Cii: whole cell recording from the same cell showing how PIR often leads to rebound action potentials after membrane repolarization. Di: whole cell recording during successive supratheshold current steps highlighting the loss of neuronal excitability during the usAHP (note that action potentials are truncated). Dii: comparison of neuronal response to short suprathreshold current injection before (black trace) and after (gray trace) driving an usAHP. Diii: mean data for the same protocol showing the latency to the first spike (N = 17) and instantaneous firing frequency of the second spike (N = 6) after the usAHP relative to before it. *P < 0.05, ***P < 0.01 from a paired t-test (Aii, Biii) or a Wilcoxon signed-rank test (Bii, Diii).

One candidate current to mediate such a response is the hyperpolarization-activated Ih current. Although not reported in embryonic spinal neurons (stage 37/38), there is evidence that Ih currents emerge by larval stage 42 (Picton 2017), and prometamorphic Xenopus spinal neurons show strong evidence of possessing Ih channels (Fig. 6C). In 42/104 (40%) neurons recorded in the present study, hyperpolarization of the membrane potential caused a characteristic depolarizing sag potential, while termination of hyperpolarizing pulses caused a depolarizing overshoot of the membrane potential (Fig. 6Ci). On some occasions, this postinhibitory rebound was large enough to cause firing in the neuron (Fig. 6Cii). Ih has not previously been reported in Xenopus embryo spinal neurons and is thus another example of a change in the cellular properties that occurs during larval development.

On the basis of evidence from earlier stages of Xenopus development (Zhang et al. 2015; Zhang and Sillar 2012), we predicted that neurons will be less likely to fire if excited within the period when the usAHP is active. To test this hypothesis the relative excitability of neurons before and immediately after the onset of the usAHP was investigated. Neurons with a detectable usAHP showed a reduction in excitability during the membrane hyperpolarization (Fig. 6D). The latency to first spike following suprathreshold current injection significantly increased by 35% from 13.87 ± 7.55 ms at rest to 17.70 ± 8.58 ms during the trough of the hyperpolarization (Fig. 6Diii; N = 17; P < 0.01). Furthermore, the instantaneous spike frequency of a second spike to the current pulse was significantly reduced to 96% of that in control: 126.17 ± 47.42 Hz at rest to 121.73 ± 48.88 Hz during the usAHP (Fig. 6Diii; N = 6; P < 0.05). In many cases, this led to the same stimulus eliciting fewer spikes during the usAHP than before (Fig. 6Dii).

Modulation by nitric oxide.

The behavioral repertoire expressed by an individual depends upon its developmental, ecological, and arousal states at any given moment in time, and neuromodulation plays an important, determinant role in sculpting behavior to prevailing conditions. NO is known to be a potent inhibitory modulator of locomotion in embryonic and early larval stages of Xenopus development, where it has been shown to enhance both GABAergic and glycinergic inhibition within the spinal cord (McLean and Sillar 2002, 2004). In contrast, at later prometamorphic stages the effects of NO on the occurrence of spontaneous locomotor activity are excitatory (Currie et al. 2016). Moreover, the excitatory effects are thought to be mediated primarily via the brain stem in these older animals. Together these findings highlight the need to investigate the effects, if any, of NO on spinal neurons during fictive locomotion in prometamorphic tadpoles.

As well as increasing the occurrence of spontaneous locomotor activity (Currie et al. 2016), bath application of the NO donors SNAP (200 μM; N = 4) or DEA-NO (50–200 μM; N = 7) caused a depolarization of spinal neurons (Fig. 7, Ai and Bi). The membrane potential depolarized significantly by 8.34 ± 7.99% relative to control during NO donor application and subsequently reduced to 6.88 ± 7.56% of control upon washout (Fig. 7Bi; N = 11, P < 0.01). On average, the depolarization was 4.03 ± 3.21 mV relative to control. The same drug application caused a reversible decrease in IR to 94.76 ± 7.50% of control (Fig. 7Bii; N = 9; 5 in DEA-NO and 4 in SNAP, P < 0.05), and upon washout the IR returned to control levels, 99.16 ± 7.23%.

Fig. 7.

Fig. 7.

Open in a new tab

Nitrergic modulation of prometamorphic spinal neurons. A: ventral root and whole cell recordings during bath application of 200 μM DEA-NO. A reversible increase in resting membrane potential (RMP) is visible in the whole cell record (Ai, bottom). Aii: examples of spontaneous episodes motor activity are shown on expanded timescales. B: graphs of mean RMP (Bi) and input resistance (IR; Bii) relative to control during NO donor application (200 μM SNAP, N = 4; 200 μM DEA-NO, N = 7) and during washout. *P < 0.05, ***P < 0.01 from a Wilcoxon signed-rank test.

During quiescent periods, frequent depolarizing PSPs were observed and NO donors were found to significantly increase their frequency (Fig. 8, A and Ci). The mean frequency of PSPs increased from 1.40 ± 0.85 Hz in control to 2.66 ± 1.68 Hz during NO donor application (Fig. 8Ci; N = 10, 4 in SNAP and 6 in DEA-NO; P < 0.05). During washout, the PSP frequency returned toward control levels, to 1.43 ± 0.85 Hz. The cumulative probability of PSP interevent interval shifted to the left, indicating a significant increase, in 5/10 recorded neurons (see example in Fig. 8Bi; P < 0.05). The mean amplitude of PSPs was not significantly altered during NO donor application (Fig. 8, A and Cii); however, the cumulative probability of PSP amplitude shifted to the right in 5/10 recorded neurons, indicating that the proportion of large-amplitude PSPs was increased significantly compared with control in these cases (Fig. 8Bii; P < 0.05).

Fig. 8.

Fig. 8.

Open in a new tab

NO increases PSPs in spinal neurons. A: whole cell recording from 3 quiescent periods during control, DEA-NO application, and washout showing depolarizing postsynaptic potentials (PSPs). B: cumulative frequency plots for PSP interevent interval (Bi) and amplitude (Bii) for the cell recorded in A. C: average data for PSP frequency (Ci) and amplitude (Cii) during control, NO donor application (200 μM SNAP, N = 4; 200 μM DEA-NO, N = 6), and washout. *P < 0.05 from a repeated-measures ANOVA with Bonferroni correction.

As well as these more general effects, bath application of DEA-NO (50–200 μM) also specifically reduced or, in some neurons, completely abolished the usAHP following depolarizing current injection (Fig. 9A). Overall, the usAHP amplitude reduced significantly to 33.21 ± 13.40% of the value in control; on average the amplitude was −3.42 ± 1.54 mV in control and −1.04 ± 0.42 mV during DEA-NO application (Fig. 9, A–C; N = 6, P < 0.01). Upon washout, the usAHP amplitude recovered partially to −1.25 ± 0.24 mV or 43.08 ± 20.13% of control (Fig. 9, C and D). The effects of DEA-NO on the usAHP appear to be independent of the depolarization during NO donor application, since these effects were evident when the membrane potential was the same or even hyperpolarized vs. control (e.g., Figure 9B). This result is particularly interesting since it suggests that the Na+-K+ pump is modulable in Xenopus spinal neurons and therefore must be considered when investigating the neuromodulation of spinal central pattern generator (CPG) networks.

Fig. 9.

Fig. 9.

Open in a new tab

Nitric oxide blocks the usAHP in prometamorphic neurons. A–C: whole cell recording of response to suprathreshold current steps before (A), during (B), and after (C) bath application of 200 μM DEA-NO (note that action potentials are truncated). D: average data of usAHP amplitude relative to control during 200 μM DEA-NO application and during washout (N = 6). ***P < 0.01 from a Wilcoxon signed-rank test.

DISCUSSION

Cellular correlates of spontaneous swimming activity.

We have described a new preparation that enables patch-clamp recordings from spinal neurons of prometamorphic tadpoles during bouts of fictive swimming. In contrast to earlier stages of development, rhythmically active neurons display a range of firing patterns during swimming activity and in response to depolarizing current pulses. One firing pattern seen in a proportion of neurons that may relate to recruitment and derecruitment during swimming is illustrated in Fig. 1C. These neurons fired tonically when the ventral root was discharging low-amplitude tonic activity and then switched into a rhythmic firing pattern when the ventral root was bursting. Three of twelve neurons displaying this activity pattern were identified morphologically as MNs (see Fig. 4), and therefore this pattern of firing may underlie the low-amplitude tonic activity recorded in ventral root recordings. Moreover, this pattern of activity would be suitable to provide a basal tone of muscle activation just before rhythmic contraction during locomotion. A bilateral stiffening of the muscles immediately rostral to those engaged in propulsive locomotion could be important to generate thrust, without causing unwanted lateral movement of the more rostral regions of the body. Other common features of these neurons were a short period of tonic ventral root discharge and neuronal firing following the end of an episode of rhythmic activity (Fig. 1Ci) or activity diminishing gradually with sporadic spiking activity (Fig. 1Cii). This suggests a switch from the relatively abrupt termination of episodes in earlier stages of Xenopus development, often coincident with a barrage of GABAergic potentials (Reith and Sillar 1999). The ability to fire in a graded fashion may be a general feature of more mature locomotor networks. In the embryonic tadpole, neurons have two basic states: quiescent and rhythmically active. This may be sufficient for a lifestyle in which movement is solely a means of escape but would be of little use to an animal that needs to move constantly and dynamically, as in free-swimming Xenopus larvae. Instead, the larvae require a greater ability to change the direction and speed of locomotion as well as to selectively recruit different parts of the tail. At the cellular level, this flexibility must necessarily involve differential activation of neurons but might also involve different firing patterns such as those described here. Understanding how the firing patterns of neurons in the spinal network map onto an episode of spontaneous swimming will be an important future step in understanding how such a well-coordinated behavior is controlled.

Another novel finding in prometamorphic tadpoles is that a proportion of spinal neurons fire tonically from rest (Fig. 2). No such neurons have been reported at earlier stages (37/38–42) of development. The spiking of this type of neuron is modulated during spontaneous motor activity (Fig. 2A) and can switch between periods of tonic activity and rhythmic bursting (Fig. 2B). These neurons appear to represent a subpopulation of spinal neurons that are continuously active, which is presumably the result of a different set of intrinsic properties compared with other more typical CPG neurons. If a proportion of MNs were tonically active, this type of activity might contribute to a resting tone in the axial muscle of these animals that is maintained in parts of the tail that are not contributing to ongoing locomotion. This type of postural control may be important in Xenopus swimming to facilitate their stereotypical “hovering” during feeding, especially since there will be reduced trimming forces because of the lack of significant forward propulsion during this behavior (Hoff and Wassersug 1986; Webb 2002). However, the fact that ventral roots are not continually active (i.e., there is no evidence for continuous low-amplitude activity) may be an argument against these neurons being MNs. These tonically active neurons appear to represent a newly reported phenotype of ventral spinal neuron in Xenopus tadpoles that may be integral components of the neural circuitry required to generate spontaneous motor output. However, it cannot be completely ruled out that removing the dorsal half of the spinal cord during the dissection could contribute to this firing profile.

A small subset of neurons (2/104) in the prometamorphic spinal cord displayed intrinsic bursting in response to depolarizing current injection (Fig. 3B). The membrane oscillations underlying these bursts were slow relative to fictive swim frequency, on the order of 0.5–1 Hz. This is in contrast to zebrafish, where a subset of low-threshold MNs display very similar, but much faster, intrinsic bursting to depolarizing current injection, which is thought to contribute to their propensity to be recruited at the lowest swimming speeds (Gabriel et al. 2011; Menelaou and McLean 2012). In the lamprey (Wallén and Grillner 1987) and neonatal rat (Hochman et al. 1994), similar but conditional bistability of the membrane potential is expressed in the presence of NMDA. Despite their relatively slow cycle period, the oscillations are proposed to contribute to the rising phase of locomotor cycles in the lamprey, since their frequency is modulated by current injection, mimicking the effects of intrinsic membrane currents during locomotion (Wallén and Grillner 1987). In early larval stages of Xenopus, a similar slow oscillation of spinal neuron membrane potential is found to be dependent on both NMDA and 5-HT (Scrymgeour-Wedderburn et al. 1997) and mediates a slow modulation of swimming activity over several consecutive cycles (Reith and Sillar 1998). As in the lamprey and zebrafish, the intrinsically bursting neurons described here could contribute to the fast oscillations of the membrane potential during swimming, reducing the reliance on fast synaptic inhibition for burst termination, as is the case in earlier stages of Xenopus development (Dale 1985; Soffe 1987; Soffe et al. 1984). Another possibility is that the intrinsic oscillations may contribute to a slower modulation of swimming that alters the intensity of motor output and in these animals could be involved in the “waxing and waning” of activity along the rostro-caudal axis of the body. This would be analogous to the NMDA-dependent modulation seen at earlier stages of development, controlling the relative intensity of motor bursts over the course of multiple cycles (Reith and Sillar 1998).

Development of AHPs and their role in spontaneous network activity.

In addition to fAHPs, which are found at early embryonic stages of development (Sautois et al. 2007), 14% of recorded prometamorphic spinal neurons display a pronounced slow sAHP, which typically lasts 150–200 ms (Fig. 5Ai). A very similar sAHP, primarily mediated by apamin-sensitive Ca2+-dependent K+ channels, is found in spinal neurons within the lamprey (el Manira et al. 1994; Hill et al. 1992; Wallén et al. 1989). This additional property of prometamorphic spinal neurons may offer additional opportunities for neuromodulation since the sAHP is a target for, for example, 5-HT in lamprey spinal neurons (for reviews see Grillner et al. 2001) and acetylcholine in mammalian MNs (Miles et al. 2007).

Nearly half of recorded spinal neurons (43%) at embryonic and early larval stages of Xenopus development display an usAHP that is dependent on increased Na+-K+ pump activity following intense periods of activity (Zhang and Sillar 2012). A similar phenomenon has also been described in MNs regulating crawling behavior in Drosophila larvae (Pulver and Griffith 2010) and spinal CPG neurons in mammalian locomotor networks (Picton et al. 2017). The usAHP is proposed to act as a simple mechanism for short-term memory of cellular activity that dynamically sets the excitability of neurons based on previous activity and that regulates the duration of locomotor bouts in light of previous network output. As we show here, the basic phenomenon of the usAHP persists into later, prometamorphic stages of Xenopus development (Fig. 5B, Fig. 6). However, there are several key differences. Although we cannot be sure we are comparing like with like in terms of cell type, since the majority of neurons recorded in the present study were not identified anatomically, the usAHP occurs in a far higher proportion of rhythmically active spinal neurons (Fig. 6Aii; 87% compared with 43%). In MNs, the one cell type reliably identified in the present study, the usAHP occurred in 89% of recordings, compared with only 58% of MNs at stages 37/38–42 (Zhang and Sillar 2012), suggesting they are representative of the overall trend. Possible explanations for this more widespread expression of the usAHP are the differential expression of Na+-K+ pump subunits (Azarias et al. 2013) or their accessory proteins (Cornelius and Mahmmoud 2003) or that Na+-K+ pump activity is regulated via second messenger pathways (reviewed in Therien and Blostein 2000) and influenced by one of the many neuromodulators known to act on the Xenopus spinal CPG (Sillar et al. 2014). NO donors, for instance, were found to reduce the usAHP in these older animals (Fig. 9), and this may contribute to the overall excitatory effect of NO at these stages of development (Fig. 7, Fig. 8, and see Currie et al. 2016). At earlier stages of development, NO has net inhibitory effects on Xenopus locomotion (McLean and Sillar 2002, 2004), and this may be partially due to fewer network neurons being susceptible to NO’s inhibitory effects on the usAHP (Zhang, Picton, and Sillar, unpublished observation). A role for NO modulating the Na+-K+ pump has been suggested in the rat midbrain, where it was shown to enhance NMDA-induced oscillations, mimicking the effects of increasing pump activity (Cox and Johnson 1998; Johnson et al. 1992). In the vasculature, NO donors have been shown to activate the pump (Gupta et al. 1996); however, these results are confounded by reports in both the kidney (Meffert et al. 1994) and cerebral cortex (Sato et al. 1995), where NO donors have been shown to inhibit pump activity. In fact, there is evidence that NO-mediated cell death may be partly via S-nitrosation of the sodium pump, which is one of many metabolic membrane proteins targeted by NO (Jaffrey et al. 2001).

As well as occurring in a higher percentage of recorded neurons in free-swimming Xenopus larvae, the usAHP is also shorter in duration (Fig. 6Aii) and associated with an initial reduction in IR (Fig. 6B) during prometamorphosis. The reduction in IR is only associated with the first few seconds of the usAHP and recovers significantly earlier (~5 s) than the duration of the usAHP (~20 s; Fig. 6B). Since the change in IR is almost certainly associated with the opening of an ion channel, and the usAHP causes a hyperpolarization, one plausible candidate is Ih. We found evidence for the development of Ih in prometamorphic Xenopus spinal neurons, since 42/104 (40%) displayed a prominent sag response and postinhibitory rebound following membrane hyperpolarization (Fig. 6C). It is possible that the usAHP interacts with Ih, which is now prevalent in spinal neurons, speeding up the recovery of the membrane potential. A similar interaction has recently been reported in leech CPG neurons (Kueh et al. 2016). Another possible explanation for the reduction in IR during the initial part of the usAHP is the opening of Shal-type IA channels. Like Ih channels, these are also activated by hyperpolarization and have been suggested to contribute to the usAHP-like responses in Drosophila larvae (Pulver and Griffith 2010).

Nitrergic modulation of spinal neurons.

Given NO’s excitatory effect on the occurrence of spontaneous locomotor activity during prometamorphosis (Currie et al. 2016), it is perhaps not surprising that NO donors were found to depolarize spinal neurons (Fig. 7) and increase PSP frequency (Fig. 8). However, evidence published previously strongly suggests that NO mediates its excitatory effects on spontaneous activity via the brain stem and has little effect directly on the spinal network itself (Currie et al. 2016). Furthermore, at embryonic and early larval stages NO also depolarizes spinal neurons despite having a potent inhibitory effect on motor output (McLean and Sillar 2002, 2004). One possible explanation to reconcile these apparent anomalies is that presynaptic facilitation of depolarizing PSPs reflects the potentiation of input synapses coming from brain stem neurons involved in the descending activation of the swim CPG.

The increase in PSPs after NO donor application is not surprising given that NO is known to facilitate synaptic transmission in neurons throughout the nervous system (for review see Garthwaite and Boulton 1995). More specifically, in earlier stages of Xenopus development, NO is known to increase both GABAergic inhibitory PSPs onto MNs, prematurely terminating swim episodes, and norepinephrine release onto glycinergic neurons, slowing swim frequency (McLean and Sillar 2002, 2004). NO’s highly diffusible nature allows it to act on multiple neurons and synapses simultaneously, and as such any increase in NO concentration would be expected to facilitate local synaptic transmission. An important next step in these experiments is to deduce the origin of the PSPs.

In summary, we describe a novel preparation that enables patch-clamp recordings from spinal neurons at later stages in the development of a well-characterized motor control system for swimming in the Xenopus tadpole. This preparation will now make it feasible to investigate in detail the maturation of a spinal CPG that first appeared during embryonic development and to explore the role of attendant modulatory pathways after the switch to a free-swimming lifestyle, as well as how the new circuitry controlling the limbs becomes incorporated.

GRANTS

This work was supported by a PhD scholarship from the Biotechnology and Biological Sciences Research Council (S. P. Currie).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

S.P.C. and K.T.S. conceived and designed research; S.P.C. performed experiments; S.P.C. analyzed data; S.P.C. and K.T.S. interpreted results of experiments; S.P.C. prepared figures; S.P.C. and K.T.S. drafted manuscript; S.P.C. and K.T.S. edited and revised manuscript; S.P.C. and K.T.S. approved final version of manuscript.

ACKNOWLEDGMENTS

Present address of S. P. Currie: Centre for Discovery Brain Sciences, The University of Edinburgh, Hugh Robson Building, Edinburgh, United Kingdom.

REFERENCES

  1. Azarias G, Kruusmägi M, Connor S, Akkuratov EE, Liu XL, Lyons D, Brismar H, Broberger C, Aperia A. A specific and essential role for Na,K-ATPase α3 in neurons co-expressing α1 and α3. J Biol Chem 288: 2734–2743, 2013. doi: 10.1074/jbc.M112.425785. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Combes D, Merrywest SD, Simmers J, Sillar KT. Developmental segregation of spinal networks driving axial- and hindlimb-based locomotion in metamorphosing Xenopus laevis. J Physiol 559: 17–24, 2004. doi: 10.1113/jphysiol.2004.069542. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Cornelius F, Mahmmoud YA. Functional modulation of the sodium pump: the regulatory proteins “Fixit”. News Physiol Sci 18: 119–124, 2003. [DOI] [PubMed] [Google Scholar]
  4. Cox BA, Johnson SW. Nitric oxide facilitates N-methyl-d-aspartate-induced burst firing in dopamine neurons from rat midbrain slices. Neurosci Lett 255: 131–134, 1998. doi: 10.1016/S0304-3940(98)00737-X. [DOI] [PubMed] [Google Scholar]
  5. Currie SP, Combes D, Scott NW, Simmers J, Sillar KT. A behaviorally related developmental switch in nitrergic modulation of locomotor rhythmogenesis in larval Xenopus tadpoles. J Neurophysiol 115: 1446–1457, 2016. doi: 10.1152/jn.00283.2015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Dale N. Reciprocal inhibitory interneurones in the Xenopus embryo spinal cord. J Physiol 363: 61–70, 1985. doi: 10.1113/jphysiol.1985.sp015695. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. El Manira A, Tegnér J, Grillner S. Calcium-dependent potassium channels play a critical role for burst termination in the locomotor network in lamprey. J Neurophysiol 72: 1852–1861, 1994. doi: 10.1152/jn.1994.72.4.1852. [DOI] [PubMed] [Google Scholar]
  8. Gabriel JP, Ausborn J, Ampatzis K, Mahmood R, Eklöf-Ljunggren E, El Manira A. Principles governing recruitment of motoneurons during swimming in zebrafish. Nat Neurosci 14: 93–99, 2011. doi: 10.1038/nn.2704. [DOI] [PubMed] [Google Scholar]
  9. Garthwaite J, Boulton CL. Nitric oxide signaling in the central nervous system. Annu Rev Physiol 57: 683–706, 1995. doi: 10.1146/annurev.ph.57.030195.003343. [DOI] [PubMed] [Google Scholar]
  10. Grillner S, Wallén P, Hill R, Cangiano L, El Manira A. Ion channels of importance for the locomotor pattern generation in the lamprey brainstem-spinal cord. J Physiol 533: 23–30, 2001. doi: 10.1111/j.1469-7793.2001.0023b.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Gupta S, Phipps K, Ruderman NB. Differential stimulation of Na+ pump activity by insulin and nitric oxide in rabbit aorta. Am J Physiol Heart Circ Physiol 270: H1287–H1293, 1996. doi: 10.1152/ajpheart.1996.270.4.H1287. [DOI] [PubMed] [Google Scholar]
  12. Hill R, Matsushima T, Schotland J, Grillner S. Apamin blocks the slow AHP in lamprey and delays termination of locomotor bursts. Neuroreport 3: 943–945, 1992. doi: 10.1097/00001756-199210000-00032. [DOI] [PubMed] [Google Scholar]
  13. Hochman S, Jordan LM, Schmidt BJ. TTX-resistant NMDA receptor-mediated voltage oscillations in mammalian lumbar motoneurons. J Neurophysiol 72: 2559–2562, 1994. doi: 10.1152/jn.1994.72.5.2559. [DOI] [PubMed] [Google Scholar]
  14. Hoff K, Wassersug R. The kinematics of swimming in larvae of the clawed frog, Xenopus laevis. J Exp Biol 122: 1–12, 1986. [Google Scholar]
  15. Jaffrey SR, Erdjument-Bromage H, Ferris CD, Tempst P, Snyder SH. Protein S-nitrosylation: a physiological signal for neuronal nitric oxide. Nat Cell Biol 3: 193–197, 2001. doi: 10.1038/35055104. [DOI] [PubMed] [Google Scholar]
  16. Johnson SW, Seutin V, North RA. Burst firing in dopamine neurons induced by N-methyl-d-aspartate: role of electrogenic sodium pump. Science 258: 665–667, 1992. doi: 10.1126/science.1329209. [DOI] [PubMed] [Google Scholar]
  17. Kueh D, Barnett WH, Cymbalyuk GS, Calabrese RL. Na+/K+ pump interacts with the h-current to control bursting activity in central pattern generator neurons of leeches. eLife 5: e19322, 2016. doi: 10.7554/eLife.19322. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Li WC, Soffe SR, Wolf E, Roberts A. Persistent responses to brief stimuli: feedback excitation among brainstem neurons. J Neurosci 26: 4026–4035, 2006. doi: 10.1523/JNEUROSCI.4727-05.2006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. McLean DL, Sillar KT. Nitric oxide selectively tunes inhibitory synapses to modulate vertebrate locomotion. J Neurosci 22: 4175–4184, 2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. McLean DL, Sillar KT. Metamodulation of a spinal locomotor network by nitric oxide. J Neurosci 24: 9561–9571, 2004. doi: 10.1523/JNEUROSCI.1817-04.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Meffert MK, Premack BA, Schulman H. Nitric oxide stimulates Ca2+-independent synaptic vesicle release. Neuron 12: 1235–1244, 1994. doi: 10.1016/0896-6273(94)90440-5. [DOI] [PubMed] [Google Scholar]
  22. Menelaou E, McLean DL. A gradient in endogenous rhythmicity and oscillatory drive matches recruitment order in an axial motor pool. J Neurosci 32: 10925–10939, 2012. doi: 10.1523/JNEUROSCI.1809-12.2012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Miles GB, Hartley R, Todd AJ, Brownstone RM. Spinal cholinergic interneurons regulate the excitability of motoneurons during locomotion. Proc Natl Acad Sci USA 104: 2448–2453, 2007. doi: 10.1073/pnas.0611134104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Nieuwkoop PD, Faber J. Normal Table of Xenopus laevis (Daudin). Amsterdam: North Holland, 1956. [Google Scholar]
  25. Picton LD. The Roles of Dopamine and the Sodium Pump in the Spinal Control of Locomotion (PhD thesis). St Andrews, UK: Univ. of St Andrews, 2017. [Google Scholar]
  26. Picton LD, Nascimento F, Broadhead MJ, Sillar KT, Miles GB. Sodium pumps mediate activity-dependent changes in mammalian motor networks. J Neurosci 37: 906–921, 2017. doi: 10.1523/JNEUROSCI.2005-16.2016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Pulver SR, Griffith LC. Spike integration and cellular memory in a rhythmic network from Na+/K+ pump current dynamics. Nat Neurosci 13: 53–59, 2010. doi: 10.1038/nn.2444. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Reith CA, Sillar KT. A role for slow NMDA receptor-mediated, intrinsic neuronal oscillations in the control of fast fictive swimming in Xenopus laevis larvae. Eur J Neurosci 10: 1329–1340, 1998. doi: 10.1046/j.1460-9568.1998.00144.x. [DOI] [PubMed] [Google Scholar]
  29. Reith CA, Sillar KT. Development and role of GABAA receptor-mediated synaptic potentials during swimming in postembryonic Xenopus laevis tadpoles. J Neurophysiol 82: 3175–3187, 1999. doi: 10.1152/jn.1999.82.6.3175. [DOI] [PubMed] [Google Scholar]
  30. Roberts A, Li WC, Soffe SR. How neurons generate behavior in a hatchling amphibian tadpole: an outline. Front Behav Neurosci 4: 16, 2010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Sah P, Faber ES. Channels underlying neuronal calcium-activated potassium currents. Prog Neurobiol 66: 345–353, 2002. doi: 10.1016/S0301-0082(02)00004-7. [DOI] [PubMed] [Google Scholar]
  32. Sato T, Kamata Y, Irifune M, Nishikawa T. Inhibition of purified (Na+,K+)-ATPase activity from porcine cerebral cortex by NO generating drugs. Brain Res 704: 117–120, 1995. doi: 10.1016/0006-8993(95)01165-X. [DOI] [PubMed] [Google Scholar]
  33. Sautois B, Soffe SR, Li WC, Roberts A. Role of type-specific neuron properties in a spinal cord motor network. J Comput Neurosci 23: 59–77, 2007. doi: 10.1007/s10827-006-0019-1. [DOI] [PubMed] [Google Scholar]
  34. Scrymgeour-Wedderburn JF, Reith CA, Sillar KT. Voltage oscillations in Xenopus spinal cord neurons: developmental onset and dependence on co-activation of NMDA and 5HT receptors. Eur J Neurosci 9: 1473–1482, 1997. doi: 10.1111/j.1460-9568.1997.tb01501.x. [DOI] [PubMed] [Google Scholar]
  35. Sillar KT, Combes D, Simmers J. Neuromodulation in developing motor microcircuits. Curr Opin Neurobiol 29: 73–81, 2014. doi: 10.1016/j.conb.2014.05.009. [DOI] [PubMed] [Google Scholar]
  36. Sillar KT, Wedderburn JF, Simmers AJ. The development of swimming rhythmicity in post-embryonic Xenopus laevis. Proc Biol Sci 246: 147–153, 1991. doi: 10.1098/rspb.1991.0137. [DOI] [PubMed] [Google Scholar]
  37. Sillar KT, Wedderburn JF, Simmers AJ. Modulation of swimming rhythmicity by 5-hydroxytryptamine during post-embryonic development in Xenopus laevis. Proc Biol Sci 250: 107–114, 1992. doi: 10.1098/rspb.1992.0137. [DOI] [PubMed] [Google Scholar]
  38. Soffe SR. Ionic and pharmacological properties of reciprocal inhibition in Xenopus embryo motoneurones. J Physiol 382: 463–473, 1987. doi: 10.1113/jphysiol.1987.sp016378. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Soffe SR, Clarke JD, Roberts A. Activity of commissural interneurons in spinal cord of Xenopus embryos. J Neurophysiol 51: 1257–1267, 1984. doi: 10.1152/jn.1984.51.6.1257. [DOI] [PubMed] [Google Scholar]
  40. Therien AG, Blostein R. Mechanisms of sodium pump regulation. Am J Physiol Cell Physiol 279: C541–C566, 2000. doi: 10.1152/ajpcell.2000.279.3.C541. [DOI] [PubMed] [Google Scholar]
  41. Wallén P, Buchanan JT, Grillner S, Hill RH, Christenson J, Hökfelt T. Effects of 5-hydroxytryptamine on the afterhyperpolarization, spike frequency regulation, and oscillatory membrane properties in lamprey spinal cord neurons. J Neurophysiol 61: 759–768, 1989. doi: 10.1152/jn.1989.61.4.759. [DOI] [PubMed] [Google Scholar]
  42. Wallén P, Grillner S. N-methyl-d-aspartate receptor-induced, inherent oscillatory activity in neurons active during fictive locomotion in the lamprey. J Neurosci 7: 2745–2755, 1987. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Webb PW. Control of posture, depth, and swimming trajectories of fishes. Integr Comp Biol 42: 94–101, 2002. doi: 10.1093/icb/42.1.94. [DOI] [PubMed] [Google Scholar]
  44. Zhang HY, Issberner J, Sillar KT. Development of a spinal locomotor rheostat. Proc Natl Acad Sci USA 108: 11674–11679, 2011. doi: 10.1073/pnas.1018512108. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Zhang HY, Li WC, Heitler WJ, Sillar KT. Electrical coupling synchronises spinal motoneuron activity during swimming in hatchling Xenopus tadpoles. J Physiol 587: 4455–4466, 2009. doi: 10.1113/jphysiol.2009.173468. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Zhang HY, Picton L, Li WC, Sillar KT. Mechanisms underlying the activity-dependent regulation of locomotor network performance by the Na+ pump. Sci Rep 5: 16188, 2015. doi: 10.1038/srep16188. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Zhang HY, Sillar KT. Short-term memory of motor network performance via activity-dependent potentiation of Na+/K+ pump function. Curr Biol 22: 526–531, 2012. doi: 10.1016/j.cub.2012.01.058. [DOI] [PubMed] [Google Scholar]

Articles from Journal of Neurophysiology are provided here courtesy of American Physiological Society

RESOURCES