Remote controlling the behavior of the worm [with video]

We combined optogenetics and new optical tools to control chemotactic behavior of Caenorhabditis elegans to understand the neural circuit that animals use to track chemo attractive gradients. We discovered that controlling the dynamics of activity in just one interneuron pair was sufficient to force the animal to locate, turn towards and track virtual light gradients.

HFSP Cross-Disciplinary Fellow Askin Kocabas and colleagues
authored on Thu, 17 January 2013

Can we understand the dynamics of the nervous system of any animal well enough to be able to control all of its behaviors? What are the key circuits in the underlying neural networks that we need to control to hijack the animal’s behaviors? This is a difficult question to answer even in the nematode, Caenorhabditis elegans, which has only 302 neurons interconnected through 7000 synapses.

Figure: Snapshots of the animal making a gradual turn to orient itself to the virtual gradient direction generated by light.

To find food, the animal uses a combination of reversals and turns to detect and track attractive signals in its environment. It then moves up the gradient of these signals in the hope of finding food. We asked if we could directly control chemotactic behavior in Caenorhabditis elegans by driving appropriate electrical activity patterns in the nervous system.

The nervous system of this animal consists of roughly one hundred interneurons that process the information they receive from the hundred or so sensory neurons to drive motor neurons and thus control the animal’s muscles. Using laser ablation and mutant analyses, previous studies identified several neurons necessary for the different locomotory behaviors underlying chemotaxis. However, these studies do not tell us whether there exists either a small or diffuse circuit of neurons that processes environmental signals to control and coordinate the different locomotory behaviors leading to chemotaxis. Neither does it tell us the patterns of electrical activity in such a circuit that might control chemotactic behavior.

To find the key neurons of the network controlling chemotaxis we answered two intricately linked questions: which neurons do we control and what activity patterns do we generate in them? Even if we could answer these questions, how do we drive such precise electrical activity patterns in a freely moving animal? In our paper, we tackled this challenge. By combining optogenetics and real time imaging, we developed a new imaging system that could visualize, identify and specifically illuminate the neuron(s) of interest, all within 25 milliseconds, to drive any pattern of electrical activity in the nervous system of freely moving animals. By modeling the sensory signals during chemotaxis, we identified the classes of temporal activity patterns that might be stimulated in the nervous system during chemotaxis. By stimulating single neurons with these temporal patterns we identified the key interneurons that could drive chemotaxis.

Surprisingly, we discovered that controlling the dynamics of activity in just one interneuron pair, AIY, was sufficient to force the animal to locate, turn towards and track virtual light gradients. In a virtual environment, we could directly control the dynamics of activity in the interneurons of the nematode with light to evoke chemotactic behavior (see movie below). The animals search, turn towards and track the direction of the virtual light gradient robustly even if we suddenly rotate the direction of the gradient.

Our work has resulted in fundamental discoveries about chemotactic behavior in C.elegans, and importantly, our approach opens new avenues of research in two ways.  Firstly, it provides a framework for discovering how different neurons control complex behaviors. Secondly, it describes novel technologies that will allow the scientific community to accurately perturb neural activity in specific neurons of freely moving animals.






Controlling interneuron activity in Caenorhabditis elegans to evoke chemotactic behaviour. Kocabas A, Shen CH, Guo ZV, Ramanathan S.  Nature 490, 273–277.

Pubmed link