The motor cortex consists of millions of neurons that are active in complex spatial and temporal patterns to generate behaviour. But what contribution, if any, do action potentials in individual neurons make to movement? Brecht and colleagues addressed this question directly by making intracellular recordings from vibrissae motor cortex pyramidal neurons in anaesthetized rats. In technically challenging experiments, the authors injected small amounts of current into the recorded neurons while simultaneously monitoring the position of individual whiskers.

Upper panel shows the forward (f) and backward (b) position of whisker E1 (wE1) during stimulation of a single layer-6 pyramidal cell. Scale bar, 10 mm. Lower panel is a plot of the position of whisker E1 over time during control and stimulation trials. The black trace corresponds to the control trial, in which subthreshold or hyperpolarizing pulses were used. Dashed line shows the onset of the action potential (AP) train. Image courtesy of T. Margrie, Department of Physiology, University of London.

In about a quarter of the neurons, trains of action potentials — triggered by the intracellular current injection — were associated with small back and forth motions in local groups of whiskers. The latency for whisker deflection was quite long, and movements often outlasted the stimulation considerably. This indicates that the action potentials that were triggered in the recorded neuron influenced a downstream circuit that was then responsible for pattern generation. This influence varied with action potential frequency, and, interestingly, with the laminar location of the recorded neuron. For example, the phase of the whisking movements relative to the action potential stimulus train was constant in the case of layer 5 neurons, but not layer 6 neurons.

Activity in the motor cortex has previously been correlated with behaviour using electrophysiological recordings. Furthermore, extracellular microstimulation techniques in behaving animals have enabled researchers to perturb ongoing activity to test theories about cortical representations of movement. The intracellular stimulation technique used by Brecht and colleagues has the advantage of exquisite control over membrane potential and action potential generation coupled with high spatial resolution — the currents used are so low that there is little danger of stimulating fibres of passage.

However, the trade-off is that animals have to be anesthetized to maintain stable recordings, and the effects of anaesthesia are likely to be significant. In fact, the authors provide some preliminary evidence that bursts of action potentials in single neurons are even more effective in awake animals: in a few cases, they were able to maintain a stable intracellular recording while they antagonized the anaesthesia pharmacologically. As the animals were waking up, the whisker responses to intracellular stimulation were larger and faster than in any of the cases where the animals were more deeply anaesthetized.

So, is it appropriate to conclude that every spike conveys a message? The results are provocative, but more work needs to be done to determine how well the observations generalize to awake animals and to more complex motor systems.