Animals exhibit different behavioral responses to the same sensory cue depending on their internal state at a given moment. How and where in the brain are sensory inputs combined with state information to select an appropriate behavior? Here, we investigate how food deprivation affects olfactory behavior in Drosophila larvae. We find that certain odors repel well-fed animals but attract food-deprived animals and that feeding state flexibly alters neural processing in the first olfactory center, the antennal lobe. Hunger differentially modulates two output pathways required for opposing behavioral responses. Upon food deprivation, attraction-mediating uniglomerular projection neurons show elevated odor-evoked activity, whereas an aversion-mediating multiglomerular projection neuron receives odor-evoked inhibition. The switch between these two pathways is regulated by the lone serotonergic neuron in the antennal lobe, CSD. Our findings demonstrate how flexible behaviors can arise from state-dependent circuit dynamics in an early sensory processing center.

Many animals perceive odorant molecules by collecting information from ensembles of olfactory neurons. These neurons employ receptors that are tuned to recognize odorant molecules by chemical binding affinity. Olfactory systems are able, in principle, to detect and discriminate large numbers of odorants by using combinatorial coding strategies. Multineuronal imaging, combined with high-throughput stimulus delivery, allow for the comprehensive measurement of ensemble-level sensory representations. Here, we used microfluidics and multineuronal imaging to study olfactory representations at the sensory periphery of the nematode C. elegans. The collective activity of chemosensory neurons in C. elegans reveals high-dimensional representations of olfactory information across a broad space of odorant molecules. We reveal diverse tuning properties and dose-response curves across chemosensory neurons and across odorant molecules. We describe the unique contribution of each sensory neuron to an ensemble-level olfactory code, and show how the encoding of a set of natural stimuli, nematode pheromones, differs from the encoding of small volatile organic molecules. The integrated activity of the sensory periphery of C. elegans contains sufficient information to robustly encode the intensity and identity of a broad panel of odorants.

There are numerous different odorant molecules in nature but only a relatively small number of olfactory receptor neurons (ORNs) in brains. This “compressed sensing” challenge is compounded by the constraint that ORNs are nonlinear sensors with a finite dynamic range. Here, we investigate possible optimal olfactory coding strategies by maximizing mutual information between odor mixtures and ORNs’ responses with respect to the bipartite odor-receptor interaction network (ORIN) characterized by sensitivities between all odorant–ORN pairs. For ORNs without spontaneous (basal) activity, we find that the optimal ORIN is sparse—a finite fraction of sensitives are zero, and the nonzero sensitivities follow a broad distribution that depends on the odor statistics. We show analytically that sparsity in the optimal ORIN originates from a trade-off between the broad tuning of ORNs and possible interference. Furthermore, we show that the optimal ORIN enhances performances of downstream learning tasks (reconstruction and classification). For ORNs with a finite basal activity, we find that having inhibitory odor–receptor interactions increases the coding capacity and the fraction of inhibitory interactions increases with the ORN basal activity. We argue that basal activities in sensory receptors in different organisms are due to the trade-off between the increase in coding capacity and the cost of maintaining the spontaneous basal activity. Our theoretical findings are consistent with existing experiments and predictions are made to further test our theory. The optimal coding model provides a unifying framework to understand the peripheral olfactory systems across different organisms.

No abstract

Long-term memories and learned behavior are conventionally associated with stable neuronal representations. However, recent experiments showed that neural population codes in many brain areas continuously change even when animals have fully learned and stably perform their tasks. This representational "drift" naturally leads to questions about its causes, dynamics, and functions. Here, we explore the hypothesis that neural representations optimize a representational objective with a degenerate solution space, and noisy synaptic updates drive the network to explore this (near-)optimal space causing representational drift. We illustrate this idea in simple, biologically plausible Hebbian/anti-Hebbian network models of representation learning, which optimize similarity matching objectives, and, when neural outputs are constrained to be nonnegative, learn localized receptive fields (RFs) that tile the stimulus manifold. We find that the drifting RFs of individual neurons can be characterized by a coordinated random walk, with the effective diffusion constants depending on various parameters such as learning rate, noise amplitude, and input statistics. Despite such drift, the representational similarity of population codes is stable over time. Our model recapitulates recent experimental observations in hippocampus and posterior parietal cortex, and makes testable predictions that can be probed in future experiments.