Salk Institute


Our brain consists of roughly 80 billion cells, each connected through thousands of contact points adding up to a quarter of a million miles of wiring-enough to reach from here to the moon. This marvel of evolutionary engineering allows us to navigate an ever-changing environment, to learn and to remember. One approach to studying our brain is to identify how we detect relevant information in our environment and use that to drive behaviors. This task is difficult in our brains with all of its stunning complexity. Fortunately, neural circuit motifs are conserved allowing us to analyze simpler nervous systems and define basic circuit principles that underlie complex brain functions.

We use two small nervous system models- the nematode, Caenorhabditis elegans and the vertebrate, Danio rerio. The C. elegans nervous system consists of just 302 neurons that are connected by identified chemical and electrical synapses. Despite its simplicity, this animal displays a number of sophisticated behaviors providing us an ideal model to study neural circuit properties.

What is a neural circuit?

Schematic showing plasticity in chemosensory circuits. ASE neurons detect low changes in salt (a) and release INS-6 and other insulins from dense core vesicles to recruit AWC neurons to encode higher changes in salt (b).

A neural circuit is defined as a set of interconnected neurons whose activation defines a linear pathway. The C. elegans nervous system has been mapped using electron microscopy in 1986. Using this map, a number of neural circuits have been defined by neuroanatomy and function. However, we find that in the chemosensory system, neural circuit configurations are flexible. Sensory context and neuropeptide signaling can modify the configuration of a circuit. These alternate configurations represent different paths that the information could travel and potentially explain how variable behavioral output is generated by the same circuits.

How does a neural circuit integrate prior experience into a behavioral strategy?

We have developed two behavioral paradigms where prior experience modifies behavioral outputs.

  1. C. elegans when removed from food executes a local search for about 15 minutes by interrupting forward movements with turns. After 15 minutes, it leaves the local area by suppressing these turns, also termed global search. Interestingly, we find that the area of the local search is dependent on the size of the food lawn it was removed from. We observe that information about the spatial distribution of food is stored in a group of interneurons using type I dopaminergic signaling and CREB. We are planning to identify the genes whose transcription leads to the acquisition of this memory.
  2. We find that C. elegans fed different bacteria perform differently on complex behavioral tasks. We observe that neuropeptide signals play a crucial role in integrating information about this prior food experience with the behavior.

What are the neural mechanisms regulating social interactions?

We have defined a novel predator-prey relationship between two nematodes, C. elegans and an aggressive P. pacificus. We observe that a starving Pristionchus will hunt, bite and consume C. elegans in about 20 minutes. C. elegans, in turn, will avoid both Pristionchus and its secretions. We are mapping the C. elegans neural circuitry and the signaling pathways that detect the predator and drive complex behavioral and physiological changes.

Additionally, we are studying how functional neural circuits mature, how they learn and are developing new tools to manipulate them. We are also extending our analysis to include the larval zebrafish.


We use zebrafish larvae to address how neural circuits generate precise behaviors. Zebrafish is a simple vertebrate whose neural circuits share conserved anatomical, molecular, and functional homology to those in rodents and human. When subjected to different environmental stimuli, they perform well-defined movements enabling the animal to explore, forage or escape. We combine genetic, optogenetic and microscopic tools to understand how the zebrafish neural circuit generates diverse adaptive behaviors. These studies will allow us to identify the conserved cellular and molecular components driving defined behaviors.

What is the neural circuit regulating hypoxia-coping behavior?

To probe hypoxia-coping behaviors, we developed a novel microfluidic device (collaboration with Alex Groissman, UCSD Department of Physics) that exposes a larva to low oxygen conditions. Under hypoxia, zebrafish larvae enhance their rate of pectoral fin beats. We aim to trace the underlying neural circuit responsible for this robust behavioral response. We hypothesize that neuroepithelial cells in the gills are the candidate sensory tissue sensing hypoxia and drive the change in pectoral fin beats through an unknown neural circuit.

We are also interested in dissecting the neural circuits driving chemotaxis behaviors.