There are more connections in the human brain than there are stars in the Milky Way, so scientists use simple organisms to search for universal neural-circuit motifs. Their latest find is a neuron for social behaviour.
The nematode worm Caenorhabditis elegans has two distinct foraging styles: social and solitary1. Social strains move rapidly on a patch of bacterial food, favour the borders of the patch (bordering), and cluster in tight little swarms when feeding (aggregation). Solitary strains, by contrast, move slowly on food, show no spatial preferences and feed alone. Whether or not a worm swarms might not seem a propitious place to dig for universal principles of neuronal function. On the contrary, however: by tracing the genetic basis of foraging style, neurobiologists have unearthed several unexpected findings2,3. On page 1171 of this issue, Macosko et al.4 revisit social feeding and discover a 'hub-and-spoke' circuit for regulation of behaviour by a cluster of diverse sensory neurons.
Differences in C. elegans foraging styles can be explained by naturally occurring variations in the gene npr-1. This gene encodes the NPR-1 receptor, which resides on the surface of neurons and binds to protein fragments called neuropeptides, which are released as signals by other neurons. Solitary strains express a high-activity version of npr-1, whereas social feeders express a low-activity version1; mutants with defective npr-1 are likewise social. Nine types of neuron express npr-1 (refs 4, 5), but, for technical reasons, it has been difficult to pinpoint which of these mediate social feeding.
Macosko et al.4 solve this problem by using, for the first time in C. elegans, the Cre–lox system — a powerful genetic technique for restricting the cellular expression pattern of distinct genes. They find that restoring the expression of normal NPR-1 protein specifically in a pair of neurons, known as RMG, converts npr-1 mutants into solitary feeders. Importantly, killing these neurons in npr-1 mutants also converts worms into solitary feeders, so RMG is probably turned off in solitary strains.
To determine the neural circuit within which RMG functions, the authors inspected the essentially complete anatomical wiring diagram of the C. elegans nervous system6. They find that this pair of neurons is connected by gap junctions — hollow proteins that allow electrical currents to flow between cells — to seven other neurons, six of which are sensory (Fig. 1a). This pattern of connectivity immediately suggested a hub-and-spoke arrangement with RMG at the centre. But how does this circuit function in social feeding?
In search of an answer, Macosko et al.4 chose the chemosensory neuron ASK, which integrates food and pheromone signals emitted by hermaphrodites7. In npr-1-defective social feeders, inactivating ASK and its sister chemosensory neurons reversed the effect of npr-1 mutation, causing the animals to be solitary. Moreover, restoring the function of ASK, and just one other sensory neuron, in these animals caused aggregation, bordering and locomotion speed to return to nearly the levels seen in social strains, suggesting that ASK can promote social feeding.
The next obvious question was how the ASK–RMG spoke of the circuit might act to regulate foraging style. ASK is required to attract males to hermaphrodites8, and Macosko and colleagues show that it signals the presence of pheromones by a reduction in its activity. These clues prompted the authors to test whether social and solitary strains might differ in their behavioural responses to pheromones. Whereas pheromones repelled solitary animals (as previously shown8,9), they attracted the npr-1-defective social strains. This striking observation has two key implications. First, it points to pheromones as the missing, long-range sensory cue that attracts worms into feeding groups. Second, it brings C. elegans aggregation closer to the orbit of behaviours that are more than just metaphorically social.
Macosko et al. also find that a pulse of pheromones reduces ASK activity to a much greater extent in the npr-1-defective social strain than in the solitary strain. This observation provides a clue as to how the RMG–ASK spoke of the circuit might function. In one model, RMG resides in either a low (solitary feeding) or a high (social feeding) activity state, depending on the activity of NPR-1 and, possibly, on the concentration of neuropeptides (Fig. 1b). The state of RMG activity is communicated electrically to ASK through gap junctions, with the result that membrane potential in ASK is set respectively low or high. When ASK is in its low state, there is little activity to inhibit. But when it is in its high state, a pheromone pulse can more strongly inhibit it. Thus, RMG switches the dynamic range of ASK's pheromone response.
In support of this model, the researchers4 show that pheromone-induced synaptic responses in the interneuron AIA — the main postsynaptic target of ASK — are small in the solitary strain and large in the social strain. Thus, from the point of view of the neurons in the social strain that get their news about the proximity of other nematodes from ASK, it would seem as though the animal has encountered a bigger change in pheromone concentration, presumably leading to a stronger orientation response. It would be interesting to determine whether AIA is required for aggregation responses.
An attractive feature of the hub-and-spoke motif is that, if its gap junctions pass ionic currents bidirectionally (as many do), then the circuit ought to operate simultaneously in opposing directions. In the centrifugal direction, RMG should broadcast to sensory neurons of various modalities its assessment of the animal's internal state, encoded by neuropeptides, thereby redirecting the attentions of the animal according to its needs. In the centripetal direction, by simple laws of current flow in circuits, RMG's membrane voltage should be a weighted average of the state of the sensory neurons, providing a balanced assessment of the local environment. We do not yet know whether the hub-and-spoke circuit as a whole functions as proposed. The evidence that one of the seven spokes (RMG–ASK) functions centrifugally now provides the impetus to test other spokes as well.
As neuroscientists contemplate large-scale anatomical reconstructions of other nervous systems10, it can be argued that costly wiring diagrams are not worth the expense because of unavoidable ambiguities in the final result. Macosko and colleagues' work4 provides a bracing example to the contrary, for without the guidance of the C. elegans wiring diagram6, one might have lost sight of this rich vein of inquiry.
de Bono, M. & Bargmann, C. I. Cell 94, 679–689 (1998).
Gray, J. M. et al. Nature 430, 317–322 (2004).
Persson, A. et al. Nature 10.1038/nature 07820 (2009).
Macosko, E. Z. et al. Nature 458, 1171–1175 (2009).
Coates, J. C. & de Bono, M. Nature 419, 925–929 (2002).
White, J. G., Southgate, E., Thomson, J. N. & Brenner, S. Phil. Trans. R. Soc. Lond. B 314, 1–340 (1986).
Schackwitz, W. S., Inoue, T. & Thomas, J. H. Neuron 17, 719–728 (1996).
Srinivasan, J. et al. Nature 454, 1115–1118 (2008).
White, J. Q. et al. Curr. Biol. 17, 1847–1857 (2007).
Helmstaedter, M., de Kock, C. P. J., Feldmeyer, D., Bruno, R. M. & Sakmann, B. Brain Res. Rev. 55, 193–203 (2007).
About this article
Annals of Operations Research (2018)
IUBMB Life (2011)