Azulay, Aharon et al. (2015) International Worm Meeting "The common neighbor rule as an organizing principle in the C. elegans neural network."

Itskovic, Eyal, Zaslaver, Alon, Azulay, Aharon

[International Worm Meeting, 2015]

Background:A major goal of Systems Neuroscience is to decipher the structure-function relationship in neural networks. Recently, a new organizing principle-the Common Neighbor Rule-was found in the rat cortex. According to this rule, the more common neighbors a pair of neuron has, the more likely that this pair of neurons will be connected. Same principle is found in social networks (e.g. Facebook) where two individuals sharing multiple mutual friends are more likely to be friends as well. While the emergence of this principle in social networks may be intuitive, it is not clear why would this rule prevail in neural networks.Results:To address this question we studied the C. elegans neural network for which a fully-mapped wiring diagram of its 302 neurons is available. Strikingly, we find that the CNR is a prominent feature of the C. elegans connectome. Moreover, as common neighbor sets are made of neural triads, network analyses algorithms revealed highly homogeneous structures that consist of almost only one triad. Furthermore, these homogeneous sets are embedded in the network in defined functional layers; specifically, in a set consisted of mutually synapsing neurons, these two neurons will reside on the same layer while their common neighbors will be on a different layer. Such mutually-regulating and mutually-regulated structural blocks are enriched in the sensory- and motor-neurons, respectively, and can therefore serve as integration and synchronization devices. In addition, neurons connected in a Feed-forward fashion tend to reside on different layers of the network underscoring their role in signal propagation. Finally, coarse-graining the network based on the common neighbor sets revealed previously overlooked key functionalities of the network.Conclusion:These results can explain the emergence of the common neighbor rule as an organizing principle in neural networks.

Pritz, Christian et al. (2021) International Worm Meeting "Principles for coding associative memories in a compact neural network"

Itskovits, Eyal, Menasherof, Mai, Zaslaver, Alon, Nelken, Tal, Pritz, Christian, Ruach, Rotem, Bokman, Eduard, Gritsenko, Vladimir, Azulay, Aharon

[International Worm Meeting, 2021]

A major goal in neuroscience is to elucidate the principles by which memories are stored in a neural network. Here, we have systematically studied how the four types of olfactory associative memories (short- and long-term memories, each as positive and negative associations) are encoded within the compact neural network of C. elegans worms. By combining these robust training paradigms with fast confocal calcium imaging using multi-neuron and whole-brain imaging approaches, we systematically traced experience-dependent activity changes down to the level of individual neurites. Interestingly, short-term, but not long-term, memory broadly altered memory-evoked responses of chemosensory neurons. Modulated activity in three neurons, namely AWA, AWC, and ASE, sufficed to discriminate between the different memory states. This economy in memory-coding neurons increases memory capacity and limits non-innate behavioral responses. In contrast, long-term memory was relegated to deeper layers of the network. Primary interneurons, AIY and AIA, exhibit memory-state-dependent activation signatures, allowing the sensory system to resume innate functionality. In conclusion, olfactory associative memory appears to follow a hierarchical and temporally structured encoding logic.