Rhythmic movements such as locomotion are produced by oscillatory circuits in the central nervous system. Work in fruitflies shows that the neural circuitry for such movements develops without peripheral sensory feedback.
Human babies kick, turn and wriggle before birth. Likewise, chick, frog, fish, lobster and fly embryos move before hatching1,2,3. In adults, rhythmic patterns of movement, such as flying, swimming, crawling and breathing, are produced by central pattern generators, groups of neurons in the central nervous system that can generate rhythmic motor patterns in the absence of input from sensory neurons at the body's periphery, although sensory input modifies the centrally generated motor patterns to the needs of the animal4. The precocious movements seen in embryos are thought to result from early activity in these pattern-generating neural circuits which will eventually develop to produce behaviourally relevant rhythmic movements after birth or hatching5.
On page 174 of this issue6, Suster and Bate describe how they have used genetics to construct Drosophila embryos and larvae that have either no peripheral sensory function or almost no peripheral sensory neurons. Nonetheless, these animals move. The results show that the basic structure of the central circuits that produce rhythmic movements can develop and function in the absence of sensory inputs.
Drosophila larvae move forwards and backwards by peristaltic (wavelike) crawling produced by waves of muscle contraction. The basic central circuitry that produces these larval movements is present in late-stage embryos, as such embryos also show peristaltic movements (Fig. 1). Suster and Bate6 used time-lapse videomicroscopy to monitor peristaltic movements in late embryos as an assay of the effect of genetically altering sensory input during the development of the circuitry that produces these movements.
In their first set of experiments, the authors created flies in which the gene for tetanus toxin was expressed in the peripheral sensory neurons that give the embryo information about its body position during normal peristalsis. The presence of tetanus toxin prevented the neurons from releasing neurotransmitter at the synapse, or junction, between neurons. It does this by interfering with the process by which vesicles containing the transmitter fuse with the neuronal cell membrane to release their contents. In contrast, non-toxin-containing motor neurons and interneurons (neurons that are neither sensory nor motor but are part of the central brain circuits) released transmitter normally. The embryos without sensory synapses generated both forward and backward, albeit abnormal, peristalsis, and hatched successfully from their egg cases.
In these first experiments, the sensory neurons were still present, although not releasing neurotransmitter from vesicles. Therefore, it was possible that the neurons were still exerting a role in shaping the construction of the motor circuits, either by acting as a substrate or scaffold for growth or by releasing signalling molecules not contained in vesicles. To address this possibility, the authors produced animals lacking the protein encoded by the senseless gene. The senseless protein is necessary for the normal development of peripheral sensory neurons, and in the absence of this gene almost all of the body's sensory neurons fail to develop7. Suster and Bate6 found that these embryos also continued to make rhythmic peristaltic movements, although they did not hatch successfully.
What can Suster and Bate's experiments tell us about the roles of sensory neurons in embryonic peristalsis? Both sets of genetically altered animals move, but neither set moves normally. Because a few sensory neurons persist in the absence of the senseless gene product, it is gratifying that essentially the same 'take-home messages' were obtained by two independent methods.
First, these experiments are an elegant use of genetics to make the point that embryonic and larval movements are produced by central pattern-generating circuits, revisiting early experiments in other systems in which workers surgically cut sensory nerves to show that rhythmic movements are centrally generated8. Second, it has long been recognized that sensory feedback has a crucial role in shaping the motor patterns produced by central pattern-generating circuits9. Therefore, it is not surprising that there were differences in the movements produced in normal and genetically altered animals lacking sensory function. In both sets of altered animals, forward peristalsis was less frequent than in control animals. So even at this early stage of development, sensory feedback may be helping the animal to control the direction of its movement, which is essential if the animal is to adapt its movements to the world's demands.
Suster and Bate6 demonstrate that sensory transmission is not required for the development of a circuit that is adequate for producing rhythmic movements. But they do not demonstrate that the circuit that forms in the absence of sensory information or most sensory neurons is identical to that which forms in their presence. Theoretical and experimental work shows that similar activity patterns can be produced by different underlying mechanisms at the level of both single neurons and neuronal networks10,11. Moreover, previous work from Bate's laboratory has shown that voltage-dependent currents in Drosophila neurons are altered when they develop in the absence of synaptic inputs from other neurons12.
Therefore, it seems likely that the circuit that forms in the absence of sensory input or sensory neurons is altered to compensate in some way for that absence. Because movement is essential for survival, there may be many complementary mechanisms that enable the animals to construct circuits that can generate 'good-enough' movements, even when deprived of some of their most important neuronal inputs and in the absence of some of the usual molecular signals.
It will be fascinating to see how the neuronal properties and synaptic connectivity of the circuits that control movement in these genetically altered animals compare with those that develop normally. We can also look forward to genetic manipulations in other organisms that are better suited to the study of reflex pathways in the control of movement, so that we can discover in detail the developmental consequences of altered sensory signalling.
Fénelon, V. S., Casasnovas, B., Simmers, J. & Meyrand, P. Curr. Opin. Neurobiol. 8, 705–709 (1998).
Bate, M. Curr. Opin. Neurobiol. 9, 670–675 (1999).
O'Donovan, M. J. Curr. Opin. Neurobiol. 9, 94–104 (1999).
Marder, E. Nature 410, 755 (2001).
Bekoff, A., Nusbaum, M. P., Sabichi, A. L. & Clifford, M. J. Neurosci. 7, 2320–2330 (1987).
Suster, M. L. & Bate, M. Nature 416, 174–178 (2002).
Nolo, R., Abbott, L. A. & Bellen, H. J. Cell 102, 349–362 (2000).
Marder, E. & Bucher, D. Curr. Biol. 11, R986–R996 (2001).
Pearson, K. G. Curr. Opin. Neurobiol. 5, 786–791 (1995).
Goldman, M. S., Golowasch, J., Marder, E. & Abbott, L. F. J. Neurosci. 21, 5229–5238 (2001).
Golowasch, J., Casey, M., Abbott, L. F. & Marder, E. Neural Comput. 11, 1079–1096 (1999).
Baines, R. A., Uhler, J. P., Thompson, A., Sweeney, S. T. & Bate, M. J. Neurosci. 21, 1523–1531 (2001).
About this article
Vestibular Lesion-Induced Developmental Plasticity in Spinal Locomotor Networks during Xenopus laevis Metamorphosis
PLoS ONE (2013)
Gustatory feedback affects feeding related motor pattern generation in starved 3rd instar larvae of Calliphora vicina
Journal of Insect Physiology (2011)
Feel the heat: The effect of temperature on development, behavior and central pattern generation in 3rd instar Calliphora vicina larvae
Journal of Insect Physiology (2011)
The cephalic and pharyngeal sense organs of Calliphora vicina 3rd instar larvae are mechanosensitive but have no profound effect on ongoing feeding related motor patterns
Journal of Insect Physiology (2010)
A New Bio‐Inspired Robot Based on Senseless Motion; Theoretical Study and Preliminary Technological Results
Multidiscipline Modeling in Materials and Structures (2008)