Track and field athletes can either sprint for a short time or run slowly for a long time. This simple trade-off between performance and endurance is a fact of life for all moving animals. But how does an animal's motor system remember what it has done in the past and regulate future movements to avoid exhaustion? Hong-Yan Zhang and Keith Sillar at St Andrews University, UK, recently explored this question by studying motor networks in Xenopus tadpoles. They published their work in a recent edition of Current Biology.

Locomotion in Xenopus consists of swimming bouts triggered by sensory stimuli interspersed with pauses. Swimming episodes can last from a few seconds to several minutes. The team first simply controlled how long tadpoles rested after each swim bout. They found a very strong linear relationship between the interswim interval (rest time) and swim bout duration. The longer the network is active, the longer it rests afterwards. This suggests that tadpoles pace themselves. To do this, the frog locomotor system must retain a memory of what it has done in the past and dictate future output.

To examine the mechanisms underlying this network memory, the team recorded from neurons in the rhythmic circuits driving locomotion during fictive swimming (swim motor patterns evoked in a preparation isolated from sensory feedback). Following swim bouts, most neurons firing during swim rhythms showed an ‘ultra slow’ after-hyperpolarization (usAHP). For up to a minute after a swim bout, the usAHP made it harder for the network to maintain swimming. This cellular ‘hangover’ was directly and exclusively proportional to the amount of time the cell spent firing action potentials during swim bouts. The more action potentials a neuron fired during swimming, the longer the cell stayed hyperpolarized.

In other species, similar activity-dependent slow AHPs are mediated by electrogenic activity of the Na+/K+ pump. These pumps actively move K+ ions into cells and Na+ ions out. They have traditionally been considered to be unglamorous workhorses that regulate the baseline electrical potential of cellular membranes, but they have also been shown to generate activity-dependent inhibitory currents in a variety of invertebrate neurons. Zhang and Sillar next tested whether the Na+/K+ pump mediates the tadpole usAHP. Indeed, the usAHP has electrophysiological properties and pharmacological sensitivity consistent with it being mediated exclusively by a Na+/K+ pump current.

Finally, the team wanted to determine whether there is a causal relationship between Na+/K+ pump function and homeostatic regulation of swim episode and pause durations. To test this, they examined both fictive and actual swimming rhythms in the presence of a pharmacological inhibitor of Na+/K+ pump function. They found that the presence of the inhibitor increased the duration of swim bouts and decreased the duration of pauses. The delicate balance between performance and endurance was disrupted and the swim network essentially ran itself to exhaustion.

The work of Zhang and Sillar demonstrates that Na+/K+ pumps can dynamically set the excitability of cellular components in relation to previous activity. This provides the Xenopus locomotor system with a simple yet elegant cellular mechanism for controlling future behaviors in light of recent activity. Given that Na+/K+ pumps and action potentials are ubiquitous features of all brains, this mechanism might very well be a highly conserved feature of neural networks across animal phyla.

H. Y.
K. T.
Short-term memory of motor performance via activity dependent potentiation of Na+/K+ pump function
Current Biology