A topography of behaviour

Researchers at Cornell University’s Department of Neurobiology and Behavior have discovered a simple – yet elegant and logical – organizational pattern which underlies movement in larvae of the zebrafish (Danio rerio).

Zebrafish larvae are just 4 mm long, and are a useful model because they are transparent. Developmental biologists can therefore use fluorescent dyes to investigate the structure and function of neurons. The larvae can move very quickly – when touched they initiate their escape response within 100 milliseconds. This escape behaviour, and other movements, are of course, generated by the nervous system. Fast swimming is generated by large motor neurons in the spinal cord which innervate large muscle fibres along the length of the body. When activated, these neurons elicit rhythmic bending movements of the entire body; these movements have a frequency of 80-100 Hz, and propel the larva at a high speed. Slow swimming is generated by rhythmic movements of the tail and alternating movements of the pectoral fins; these movements have a lower frequency (20-40 Hz), and are produced by the activity of other motor neurons.

Joseph Fetcho and his colleagues inserted microelectrodes into the motor neurons innervating the muscles that control swimming, to record the activity of the cells during swimming. They found that there is a linear relationship between motor neuron activity and swimming speed – neurons in the ventral region of the spinal cord are activated at slow swimming speeds, while neurons located more dorsally are recruited during faster swimming. The findings are published in the current issue of Nature.

In vivo calcium imaging was used to confirm the results. Motor neurons were filled with a dye called green dextran, which fluoresces when it binds calcium ions. Confocal microscopy was then used to image the changes in calcium ion concentration that occur in the motor neurons during swimming. The results obtained in these experiments were consistent with the electrophysiological data – when the larvae swam quickly, motor neurons in the dorsal spinal cord were recruited, and, when they swam slowly, more ventral cells were active. Also observed was an apparent relationship between the size of the neurons and the behaviour they mediate – the dorsal neurons which are active during fast swimming were larger than the ventral cells active during slow swimming.

Inhibitory spinal interneurons were then investigated using a genetic construct consisting of the gene encoding green fluorescent protein (GFP) fused with the gene encoding engrailed-1, a transcription factor which is expressed in the interneurons. In transgenic fish created using this construct, the interneurons emit green fluorescence when active. It was found that the location of interneurons was also systematically related to the behaviour they mediated. However, the relationship was the opposite of that of the motor neurons – dorsal interneurons were activated at low swimming speeds while more ventral ones were activated at fast swimming speeds.

From these observations, Fethco’s team predicted that removing ventral motor neurons would impair the movements required for slow swimming, while removing dorsal motor neurons would not. Sure enough, laser ablation of ventral cells on both sides of the spinal cord impaired the ability of the larvae to swim slowly, but had no effect on the generation of movements required for fast swimming. Conversely, ablating the dorsal motor neurons impaired the movements required for fast movements, but did not affect the ability of the larvae to swim slowly.

During zebrafish development, the first neurons to differentiate in the spinal cord are the dorsal motor neurons and interneurons; the patterning of cells in the spinal cord occurs as a result of concentration gradients of signalling molecules such as Sonic Hedgehog. It is possible that these gradients may also be responsible for the functional characteristics of cell types in the spinal cord. Although there is a high degree of evolutionary conservation of the mechanisms of spinal cord development in the vertebrate lineage, it is unclear whether the findings are applicable to humans. If they are, a better understanding of the topographical organization of motor neurons in the human spinal cord could aid researchers investigating the regeneration of these cells.


McLean, D. L., et al. (2007). A topographic map of recruitment in spinal cord. Nature 446: 71-75. [Abstract]