Hong-Yan Zhang, Wen-Chang Li, William J Heitler and Keith T Sillar
School of Biology, University of St Andrews, St Andrews KY16 9TS, UK

https://doi.org/10.36866/pn.80.19

Gap junctions provide low resistance paths for the diffusion of ions and small molecules between neighbouring cells. They underlie the electrical synapses that couple neuronal activity in some networks. Until recently the presence of electrical coupling via gap junctions was thought to prevail mainly in invertebrates and at early stages of vertebrate development (Walton & Navarette, 1991), but more recently the presence of electrical synapses in both juvenile and mature networks throughout the vertebrate CNS has significantly increased interest in the role of electrical transmission in neural circuit function.

Shaping the swimming motor programme in hatching Xenopus tadpoles

At the time of hatching from their egg membranes (stage 37/38), tadpoles of the South African clawed frog, Xenopus laevis, are normally stationary but are able to swim when touched. They propel themselves through the water by generating side-to-side oscillations of the body at frequencies of 10 to 20 Hz. The swimming muscles, or myotomes, are segmented blocks of somitic origin which contract in a precisely coordinated sequence with strict alternation between the left and right sides, and a head-to-tail propagation with a brief delay between adjacent segments. Intracellular recordings from myotomal motoneurons during swimming show that the population supplying a given muscle segment all fire synchronously, and that each motoneuron normally fires just a single action potential per cycle. This results in bursts of ventral root discharge which, when recorded with extracellular suction electrodes, last only about 5 ms (Fig. 1A). The majority of the population of 12 to 15 motoneurons per segment discharge on almost every cycle of swimming, irrespective of the swimming frequency, constraining motor output flexibility at this early stage of development.

The motor system of Xenopus laevis tadpoles changes its output dramatically in just 24 hours after hatching, with a large increase in the duration of ventral root bursts (stage 42, Fig. 1B). This rapid developmental change presents an excellent opportunity to study the function of electrical coupling in a simple, developing model system.

The presence of electrical connections between neighbouring motoneurons at stage 37/38 (Perrins & Roberts, 1995a,b) suggests a causal role for gap junctions in the synchronization of activity, since a change in transmembrane voltage in one member of an electrically coupled syncytium will ‘drag’ the membrane potentials of its neighbours, with minimal delay, in the same direction. Furthermore, since recent patch clamp recordings of the same motoneurons have shown that when depolarized by injected current they are capable of firing multiple action potentials (Zhang et al. 2009), the fact that they only fire once per swim cycle might also be partially due to electrical coupling. It is not known if there is a particular advantage to the embryo of synchronizing and limiting firing in motoneurons serving each myotome, but perhaps this optimizes muscle contraction.

Electrical coupling in the spinal cord appears to be limited to within the same neuron classes: motoneuron–motoneuron (e.g. Xenopus tadpole, Perrins & Roberts, 1995a), or interneuron–interneuron interactions (e.g. neonatal mouse, Hinckley & Ziskind-Conhaim, 2006). Recently it has been established that a population of excitatory interneurons in the tadpole swimming network, called dINs, also form an electrically coupled network (Li et al. 2009). These glutamatergic interneurons are critical for the maintenance of the swimming rhythm; rostrally located dINs form positive feedback connections amongst each other such that they drive the spinal swimming rhythm for many tens or hundreds of consecutive cycles (Li et al. 2006). This suggested that the electrical synapses amongst them might be important in sustaining rhythmic locomotor activity once it has started.

Pharmacology of gap junctions and electrical synapses

Two perennial problems in studying the role of electrical synapses in neural network activity have been: (i) the difficulty in making paired recordings from coupled neurons; and (ii) the paucity of specific drugs that block gap junctions. Although certain pharmacological reagents that block gap junctions have been available for a long time, they have a reputation for being non-specific and capricious. The prevalence of electrical coupling between the dINs, pairs of which can be recorded simultaneously with two patch electrodes, provided an excellent opportunity to test the specificity of some commercially available gap junction blockers (Li et al. 2009). When applied at concentrations sufficient to block electrical coupling between dINs, all except one drug, 18β-glycyrrhetinic acid (18β-GA), had serious unwanted side effects. 18β-GA was therefore the drug of choice for further detailed investigation into the consequences of gap junction blockade.

The most striking effect of 18β-GA on the ventral root bursts was to cause a gradual increase in duration from circa 5 ms up to 20 ms (Fig. 2), similar to the increase that occurs naturally by stage 42 of development (Fig. 1Aii and Bii).

The effect was dependent on the concentration of blocker applied, suggesting that the gradual removal of the influence of electrical coupling allowed motoneuron firing in each cycle to de-synchronize. Two plausible and non-exclusive consequences of gap junction blockage might account for this change. First, the motoneuron population might de-synchronize, and second, individual motoneurons might start to fire multiple spikes per cycle. Patch clamp recordings from motoneurons confirmed the former but not the latter hypothesis (Zhang et al. 2009). The motoneurons still only fired a single spike per cycle, but the temporal coordination of their firing relative to each other was disrupted, producing a de-synchronized ventral root burst (Fig. 3).

An additional dose-dependent effect of the gap junction blocker was a reduction in the duration of the swim episode, an effect not readily explicable at the level of the motoneurons. The most likely reason for this effect is that the reduced coupling between dINs reduces the firing level of their population, thus weakening the rhythm-generating mechanism (Fig. 4), which is based on positive feedback among dINs (Li et al. 2009).

What happens during development?

Gap junction coupling at the time of hatching therefore synchronizes the single spiking of motoneurons in each cycle. However, just one day later in development Xenopus larvae generate a very different motor pattern in which the duration of each ventral root burst increases to around 20 ms (Fig. 1Aii and Bii), remarkably similar to the effect of 18β-GA on embryo swimming bursts (Fig. 2). This led to the hypothesis that during development there is a reduction in electrical coupling, allowing motoneurons to escape from synchrony. The outcome is a motor pattern that is inherently more flexible because it is no longer constrained by the synchronizing influence of gap junctions.

We tested this hypothesis by applying the gap junction blocker at the highest concentration tested at the hatching stage to larval stage 42 tadpoles and found that the drug had no effect on either ventral root burst or swim episode durations. This suggests that gap junctions no longer play a significant role in coordinating motoneuron activity or sustaining larval swimming and that the effects on the hatching stage are indeed a direct result of the drug’s effect on electrical coupling. A further implication of these findings is that the electrical coupling between motoneurons does not play a crucial role in the developmental transition from single to multiple spikes per cycle in individual motoneurons because gap junction block at the hatching stage does not affect the neurons’ single spike per cycle pattern during swimming.

Motor system development – cause or effect of gap junctions?

Perhaps the simplest explanation for the change in the motor output pattern during the development of the tadpole swimming system is that the gene for electrical synapse proteins, presumably connexins of some sort, is turned off around the time of hatching, allowing motoneurons to de-synchronize their firing. A more complex hypothesis, which still needs to be tested, is that the gap junctions are still present but the coupling strength is merely turned down, for example by an intrinsic modulator, so that they may be available at a later date for some unknown function. There is good evidence in other systems that electrical synapses are subject to neuromodulation, for example by biogenic amines like dopamine or serotonin (Lasater & Dowling, 1985; Roerig & Sutor, 1996). The developmental transition in tadpole swimming from a simple, brief ventral root burst to a more robust and flexible burst in larvae is under serotonergic control from the raphe nucleus whose projections invade the spinal cord over precisely the same period of development. Therefore it is conceivable that the growth of raphe axons and their release of serotonin regulates the developmental expression of swimming output by decreasing gap junction coupling between motoneurons.

References

Hinckley CA & Ziskind-Conhaim L (2006). Electrical coupling between locomotor-related excitatory interneurons in the mammalian spinal cord. J Neurosci 26, 8477–8483.

Lasater EM & Dowling JE (1985). Dopamine decreases conductance of the electrical junctions between cultured retinal horizontal cells. Proc Natl Acad Sci 82, 3025–3029.

Li WC, Roberts A & Soffe SR (2009). Locomotor rhythm maintenance: electrical coupling among premotor excitatory interneurons in the brainstem and spinal cord of young Xenopus tadpoles. J Physiol 587, 1677–1693.

Li WC, Soffe SR, Wolf E & Roberts A (2006). Persistent responses to brief stimuli: feedback excitation among brainstem neurons. J Neurosci 26, 4026–4035

Perrins P & Roberts A (1995a). Cholinergic and electrical motoneuron-to-motoneuron synapses contribute to on-cycle excitation during swimming in Xenopus embryos. J Neurophysiol 73, 1005–1012.

Perrins P & Roberts A (1995b). Cholinergic contribution to excitation in a spinal locomotor central pattern generator in Xenopus embryos. J Neurophysiol 73, 1013–1019.

Roerig B & Sutor B (1996). Serotonin regulates gap junction coupling in the developing rat somatosensory cortex. Eur J Neurosci 8, 1685–1695.

Walton KD & Navarette R (1991). Postnatal changes in motoneurone electrotonic coupling studied in the in vitro rat lumbar spinal cord. J Physiol 433, 283–305.

Zhang H-Y, Li W-C, Heitler WJ & Sillar KT (2009). Electrical coupling synchronises spinal motoneuron activity during swimming in hatchling Xenopus tadpoles. J Physiol 587, 4455–4466. http://jp.physoc.org/content/587/18/4455.long