Summary Neural responses are noisy, and circuit structure can correlate this noise across neurons. Theoretical studies show that noise correlations can have diverse effects on population coding, but these studies rarely explore stimulus dependence of noise correlations. Here, we show that noise correlations in responses of ON-OFF direction-selective retinal ganglion cells are strongly stimulus dependent and we uncover the circuit mechanisms producing this stimulus dependence. A population model based on these mechanistic studies shows that stimulus-dependent noise correlations improve the encoding of motion direction two-fold compared to independent noise. This work demonstrates a mechanism by which a neural circuit effectively shapes its signal and noise in concert, minimizing corruption of signal by noise. Finally, we generalize our findings beyond direction coding in the retina and show that stimulus-dependent correlations will generally enhance information coding in populations of diversely tuned neurons.

In the avian inner ear, nonsensory supporting cells give rise to new sensory hair cells through two distinct processes: mitosis and direct transdifferentiation. Regulation of supporting cell behavior and cell fate specification during avian hair cell regeneration is poorly characterized. Expression of Atoh1, a proneural transcription factor necessary and sufficient for developmental hair cell specification, was examined using immunofluorescence in quiescent and regenerating hair cell epithelia of mature chickens. In untreated birds, Atoh1 protein was not detected in the auditory epithelium, which is quiescent. In contrast, numerous Atoh1-positive nuclei were seen in the utricular macula, which undergoes continual hair cell turnover.

Computation in the nervous system often relies on the integration of signals from parallel circuits with different functional properties. Correlated noise in these inputs can, in principle, have diverse and dramatic effects on the reliability of the resulting computations 1–8. Such theoretical predictions have rarely been tested experimentally because of a scarcity of preparations that permit measurement of both covariation of a neuron’s input signals and the effect of manipulating such covariation on a cell’s output. Here we introduce a new method to measure covariation of the excitatory and inhibitory inputs a cell receives. This method revealed strong correlated noise in the inputs to two types of retinal ganglion cell. Eliminating correlated noise without changing other input properties substantially decreased the accuracy with which a cell’s spike outputs encoded light inputs. Thus covariation of excitatory and inhibitory inputs can be a critical determinant of the reliability of neural coding and computation.

A common strategy by which developing neurons locate their synaptic partners is through projections to circuit-specific neuropil sublayers. Once established, sublayers serve as a substrate for selective synapse formation, but how sublayers arise during neurodevelopment remains unknown. Here, we identify the earliest events that initiate formation of the direction-selective circuit in the inner plexiform layer of mouse retina. We demonstrate that radially migrating newborn starburst amacrine cells establish homotypic contacts on arrival at the inner retina. These contacts, mediated by the cell-surface protein MEGF10, trigger neuropil innervation resulting in generation of two sublayers comprising starburst-cell dendrites. This dendritic scaffold then recruits projections from circuit partners. Abolishing MEGF10-mediated contacts profoundly delays and ultimately disrupts sublayer formation, leading to broader direction tuning and weaker direction-selectivity in retinal ganglion cells. Our findings reveal a mechanism by which differentiating neurons transition from migratory to mature morphology, and highlight this mechanism’s importance in forming circuit-specific sublayers.

Direction-selective ganglion cells (DSGCs) deliver signals from the retina to multiple brain areas to indicate the presence and direction of motion. Delivering reliable signals in response to motion is critical across light levels. Here we determine how populations of DSGCs adapt to changes in light level, from moonlight to daylight. Using large-scale measurements of neural activity, we demonstrate that the population of DSGCs switches encoding strategies across light levels. Specifically, the direction tuning of superior (upward)-preferring ON-OFF DSGCs becomes broader at low light levels, whereas other DSGCs exhibit stable tuning. Using a conditional knockout of gap junctions, we show that this differential adaptation among superior-preferring ON-OFF DSGCs is caused by connexin36-mediated electrical coupling and differences in effective GABAergic inhibition. Furthermore, this adaptation strategy is beneficial for balancing motion detection and direction estimation at the lower signal-to-noise ratio encountered at night. These results provide insights into how light adaptation impacts motion encoding in the retina.

Signals throughout the nervous system diverge into parallel excitatory and inhibitory pathways that later converge on downstream neurons to control their spike output. Converging excitatory and inhibitory synaptic inputs can exhibit a variety of temporal relationships. A common motif is feedforward inhibition, in which an increase (decrease) in excitatory input precedes a corresponding increase (decrease) in inhibitory input. The delay of inhibitory input relative to excitatory input originates from an extra synapse in the circuit shaping inhibitory input. Another common motif is push-pull or “crossover” inhibition, in which increases (decreases) in excitatory input occur together with decreases (increases) in inhibitory input. Primate On midget ganglion cells receive primarily feedforward inhibition and On parasol cells receive primarily crossover inhibition; this difference provides an opportunity to study how each motif shapes the light responses of cell types that play a key role in visual perception. For full-field stimuli, feedforward inhibition abbreviated and attenuated responses of On midget cells, while crossover inhibition, though plentiful, had surprisingly little impact on the responses of On parasol cells. Spatially structured stimuli, however, could cause excitatory and inhibitory inputs to On parasol cells to increase together, adopting a temporal relation very much like that for feedforward inhibition. In this case, inhibitory inputs substantially abbreviated a cell’s spike output. Thus inhibitory input shapes the temporal stimulus selectivity of both midget and parasol ganglion cells, but its impact on responses of parasol cells depends strongly on the spatial structure of the light inputs.

In chickens, nonsensory supporting cells divide and regenerate auditory hair cells after injury. Anatomical evidence suggests that supporting cells can also transdifferentiate into hair cells without dividing. In this study, we characterized an organ culture model to study auditory hair cell regeneration, and we used these cultures to test if direct transdifferentiation alone can lead to significant hair cell regeneration. Control cultures (organs from posthatch chickens maintained without streptomycin) showed complete hair cell loss in the proximal (high-frequency) region by 5 days. In contrast, a 2-day treatment with streptomycin induced loss of hair cells from all regions by 3 days. Hair cell regeneration proceeded in culture, with the time course of supporting cell division and hair cell differentiation generally resembling in vivo patterns. The degree of supporting cell division depended upon the presence of streptomycin, the epithelial region, the type of culture media, and serum concentration. On average, 87% of the regenerated hair cells lacked the cell division marker BrdU despite its continuous presence, suggesting that most hair cells were regenerated via direct transdifferentiation. Addition of the DNA polymerase inhibitor aphidicolin to culture media prevented supporting cell division, but numerous hair cells were regenerated nonetheless. These hair cells showed signs of functional maturation, including stereociliary bundles and rapid uptake of FM1-43. These observations demonstrate that direct transdifferentiation is a significant mechanism of hair cell regeneration in the chicken auditory after streptomycin damage in vitro.

Animals often encounter large increases in odor intensity that can persist for many seconds. These increases in the background odor are often accompanied by increases in the variance of the odor stimulus. Previous studies have shown that a persistent odor stimulus (odor background) results in a decrease in the response to brief odor pulses in the olfactory receptor neurons (ORNs). However, the contribution of adapting mechanisms beyond the ORNs is not clear. Thus, it is unclear how adaptive mechanisms are distributed within the olfactory circuit and what impact downstream adaptation may have on the encoding of odor stimuli. In this study, adaptation to the same odor stimulus is examined at multiple levels in the well studied and accessible Drosophila olfactory system. The responses of the ORNs are compared to the responses of the second order, projection neurons (PNs), directly connected to them. Adaptation in PN spike rate was found to be much greater than adaptation in the ORN spike rate. This greater adaptation allows PNs to encode odor contrast (ratio of pulse intensity to background intensity) with little ambiguity. Moreover, distinct neural mechanisms contribute to different aspects of adaptation; adaptation to the background odor is dominated by adaptation in spike generation in both ORNs and PNs, while adaptation to the odor pulse is dominated by changes within olfactory transduction and the glomerulus. These observations suggest that the olfactory system adapts at multiple sites to better match its response gain to stimulus statistics.