Visual coding with a population of direction-selective neurons

Nonresponsive Neurons Improve Population Coding of Object Location

Understanding how heterogeneous neural populations represent sensory input to give rise to behavior remains a central problem in systems neuroscience. Here we investigated how midbrain neurons within the electrosensory system of Apteronotus leptorhynchus code for object location in space. In vivo simultaneous recordings were achieved via Neuropixels probes, high-density electrode arrays, with the stimulus positioned at different locations relative to the animal. Midbrain neurons exhibited heterogeneous response profiles, with a significant proportion (65%) seemingly nonresponsive to moving stimuli. Remarkably, we found that nonresponsive neurons increased population coding of object location through synergistic interactions with responsive neurons by effectively reducing noise. Mathematical modeling demonstrated that increased response heterogeneity together with the experimentally observed correlations was sufficient to give rise to independent encoding by responsive neurons. Furthermore, the addition of nonresponsive neurons in the model gave rise to synergistic population coding. Taken together, our findings reveal that nonresponsive neurons, which are frequently excluded from analysis, can significantly improve population coding of object location through synergistic interactions with responsive neurons. Combinations of responsive and nonresponsive neurons have been observed in sensory systems across taxa; it is likely that similar synergistic interactions improve population coding across modalities and behavioral tasks.

Long-Term Porcine Retina Explants as an Alternative to In Vivo Experimentation

Purpose

The porcine retina represents an optimal model system to study treatment approaches for inherited retinal dystrophies owing to close anatomical similarities to the human retina, including a cone enriched visual streak. The aim of this work was to establish a protocol to keep explants in culture for up to 28 days with good morphological preservation.

Methods

Two to four retina explants per eye were obtained from the central part of the retina and transferred onto a membrane insert with the photoreceptors facing down. Different medium compositions using Neurobasal-A medium containing 100 or 450 mg/dL glucose and combinations of fetal calf serum, B-27 with or without insulin and N-2 were tested. We developed a tissue quality score with robust markers for different retinal cell types (protein kinase C alpha, peanut agglutinin and 4',6-diamidino-2-phenylindol).

Results

Retinae were kept until 28 days with only little degradation. The best results were attained using Neurobasal-A medium containing 100 mg/dL glucose supplemented with B-27 containing insulin and N-2. For an easy preparation process, it is necessary to minimize transport time and keep the eyes on ice until dissected. Heat-mediated decontamination by the butcher has to be avoided.

Conclusions

Using a standardized protocol, porcine retina explants represent an easy to handle intermediate model between in vitro and in vivo experimentation. This model system is robustly reproducible and contributes to the implementation of the 3R principle to minimize animal experimentation.

Translational Relevance

This model can be used to test future therapeutic approaches for inherited retinal dystrophies.

Exploring retinal ganglion cells encoding to multi-modal stimulation using 3D microelectrodes arrays

Microelectrode arrays (MEA) are extensively utilized in encoding studies of retinal ganglion cells (RGCs) due to their capacity for simultaneous recording of neural activity across multiple channels. However, conventional planar MEAs face limitations in studying RGCs due to poor coupling between electrodes and RGCs, resulting in low signal-to-noise ratio (SNR) and limited recording sensitivity. To overcome these challenges, we employed photolithography, electroplating, and other processes to fabricate a 3D MEA based on the planar MEA platform. The 3D MEA exhibited several improvements compared to planar MEA, including lower impedance (8.73 ± 1.66 kΩ) and phase delay (-15.11° ± 1.27°), as well as higher charge storage capacity (CSC = 10.16 ± 0.81 mC/cm2), cathodic charge storage capacity (CSCc = 7.10 ± 0.55 mC/cm2), and SNR (SNR = 8.91 ± 0.57). Leveraging the advanced 3D MEA, we investigated the encoding characteristics of RGCs under multi-modal stimulation. Optical, electrical, and chemical stimulation were applied as sensory inputs, and distinct response patterns and response times of RGCs were detected, as well as variations in rate encoding and temporal encoding. Specifically, electrical stimulation elicited more effective RGC firing, while optical stimulation enhanced RGC synchrony. These findings hold promise for advancing the field of neural encoding.

Feature Detection by Retinal Ganglion Cells

Retinal circuits transform the pixel representation of photoreceptors into the feature representations of ganglion cells, whose axons transmit these representations to the brain. Functional, morphological, and transcriptomic surveys have identified more than 40 retinal ganglion cell (RGC) types in mice. RGCs extract features of varying complexity; some simply signal local differences in brightness (i.e., luminance contrast), whereas others detect specific motion trajectories. To understand the retina, we need to know how retinal circuits give rise to the diverse RGC feature representations. A catalog of the RGC feature set, in turn, is fundamental to understanding visual processing in the brain. Anterograde tracing indicates that RGCs innervate more than 50 areas in the mouse brain. Current maps connecting RGC types to brain areas are rudimentary, as is our understanding of how retinal signals are transformed downstream to guide behavior. In this article, I review the feature selectivities of mouse RGCs, how they arise, and how they are utilized downstream. Not only is knowledge of the behavioral purpose of RGC signals critical for understanding the retinal contributions to vision; it can also guide us to the most relevant areas of visual feature space. Expected final online publication date for the Annual Review of Vision Science, Volume 8 is September 2022. Please see http://www.annualreviews.org/page/journal/pubdates for revised estimates.

Spatially displaced excitation contributes to the encoding of interrupted motion by a retinal direction-selective circuit

Spatially distributed excitation and inhibition collectively shape a visual neuron's receptive field (RF) properties. In the direction-selective circuit of the mammalian retina, the role of strong null-direction inhibition of On-Off direction-selective ganglion cells (On-Off DSGCs) on their direction selectivity is well-studied. However, how excitatory inputs influence the On-Off DSGC's visual response is underexplored. Here, we report that On-Off DSGCs have a spatially displaced glutamatergic receptive field along their horizontal preferred-null motion axes. This displaced receptive field contributes to DSGC null-direction spiking during interrupted motion trajectories. Theoretical analyses indicate that population responses during interrupted motion may help populations of On-Off DSGCs signal the spatial location of moving objects in complex, naturalistic visual environments. Our study highlights that the direction-selective circuit exploits separate sets of mechanisms under different stimulus conditions, and these mechanisms may help encode multiple visual features.

Versatile live-cell activity analysis platform for characterization of neuronal dynamics at single-cell and network level

Chronic imaging of neuronal networks in vitro has provided fundamental insights into mechanisms underlying neuronal function. Current labeling and optical imaging methods, however, cannot be used for continuous and long-term recordings of the dynamics and evolution of neuronal networks, as fluorescent indicators can cause phototoxicity. Here, we introduce a versatile platform for label-free, comprehensive and detailed electrophysiological live-cell imaging of various neurogenic cells and tissues over extended time scales. We report on a dual-mode high-density microelectrode array, which can simultaneously record in (i) full-frame mode with 19,584 recording sites and (ii) high-signal-to-noise mode with 246 channels. We set out to demonstrate the capabilities of this platform with recordings from primary and iPSC-derived neuronal cultures and tissue preparations over several weeks, providing detailed morpho-electrical phenotypic parameters at subcellular, cellular and network level. Moreover, we develop reliable analysis tools, which drastically increase the throughput to infer axonal morphology and conduction speed.

Current methods of neuronal network imaging cannot be used for continuous, long-term functional recordings. Here, the authors present a dual-mode high-density microelectrode array, which can simultaneously record in full-frame and high-signal-to-noise modes for label-free electrophysiological measurements.

Global Motion Processing by Populations of Direction-Selective Retinal Ganglion Cells

Simple stimuli have been critical to understanding neural population codes in sensory systems. Yet it remains necessary to determine the extent to which this understanding generalizes to more complex conditions. To examine this problem, we measured how populations of direction-selective ganglion cells (DSGCs) from the retinas of male and female mice respond to a global motion stimulus with its direction and speed changing dynamically. We then examined the encoding and decoding of motion direction in both individual and populations of DSGCs. Individual cells integrated global motion over ∼200 ms, and responses were tuned to direction. However, responses were sparse and broadly tuned, which severely limited decoding performance from small DSGC populations. In contrast, larger populations compensated for response sparsity, enabling decoding with high temporal precision (<100 ms). At these timescales, correlated spiking was minimal and had little impact on decoding performance, unlike results obtained using simpler local motion stimuli decoded over longer timescales. We use these data to define different DSGC population decoding regimes that use or mitigate correlated spiking to achieve high-spatial versus high-temporal resolution.SIGNIFICANCE STATEMENT ON-OFF direction-selective ganglion cells (ooDSGCs) in the mammalian retina are typically thought to signal local motion to the brain. However, several recent studies suggest they may signal global motion. Here we analyze the fidelity of encoding and decoding global motion in a natural scene across large populations of ooDSGCs. We show that large populations of DSGCs are capable of signaling rapid changes in global motion.

Early Visual Motion Experience Improves Retinal Encoding of Motion Directions

Altered sensory experience in early life often leads to altered response properties of the sensory neurons. This process is mostly thought to happen in the brain, not in the sensory organs. We show that in the mouse retina of both sexes, exposed to a motion-dominated visual environment from eye-opening, the ON-OFF direction selective ganglion cells (ooDSGCs) develop significantly stronger direction encoding ability for motion in all directions. This improvement occurs independent of the motion direction used for training. We demonstrated that this enhanced ability to encode motion direction is mainly attributed to increased response reliability of ooDSGCs. Closer examination revealed that the excitatory inputs from the ON bipolar pathway showed enhanced response reliability after the motion experience training, while other synaptic inputs remain relatively unchanged. Our results demonstrate that retina adapts to the visual environment during neonatal development.SIGNIFICANCE STATEMENT We found that retina, as the first stage of visual sensation, can also be affected by experience dependent plasticity during development. Exposure to a motion enriched visual environment immediately after eye-opening greatly improves motion direction encoding by direction selective retinal ganglion cells (RGCs). These results motivate future studies aimed at understanding how visual experience shapes the retinal circuits and the response properties of retinal neurons.

Large-Scale, High-Resolution Microelectrode Arrays for Interrogation of Neurons and Networks

High-density microelectrode arrays (HD-MEAs) are increasingly being used for the observation and manipulation of neurons and networks in vitro. Large-scale electrode arrays allow for long-term extracellular recording of the electrical activity from thousands of neurons simultaneously. Beyond population activity, it has also become possible to extract information of single neurons at subcellular level (e.g., the propagation of action potentials along axons). In effect, HD-MEAs have become an electrical imaging platform for label-free extraction of the structure and activation of cells in cultures and tissues. The quality of HD-MEA data depends on the resolution of the electrode array and the signal-to-noise ratio. In this chapter, we begin with an introduction to HD-MEA signals. We provide an overview of the developments on complementary metal-oxide-semiconductor or CMOS-based HD-MEA technology. We also discuss the factors affecting the performance of HD-MEAs and the trending application requirements that drive the efforts for future devices. We conclude with an outlook on the potential of HD-MEAs for advancing basic neuroscience and drug discovery.

Activity Correlations between Direction-Selective Retinal Ganglion Cells Synergistically Enhance Motion Decoding from Complex Visual Scenes

Neurons in sensory systems are often tuned to particular stimulus features. During complex naturalistic stimulation, however, multiple features may simultaneously affect neuronal responses, which complicates the readout of individual features. To investigate feature representation under complex stimulation, we studied how direction-selective ganglion cells in salamander retina respond to texture motion where direction, velocity, and spatial pattern inside the receptive field continuously change. We found that the cells preserve their direction preference under this stimulation, yet their direction encoding becomes ambiguous due to simultaneous activation by luminance changes. The ambiguities can be resolved by considering populations of direction-selective cells with different preferred directions. This gives rise to synergistic motion decoding, yielding more information from the population than the summed information from single-cell responses. Strong positive response correlations between cells with different preferred directions amplify this synergy. Our results show how correlated population activity can enhance feature extraction in complex visual scenes.

•

Direction-selective ganglion cells respond to motion as well as luminance changes

•

This obscures the readout of direction from single cells under complex texture motion

•

Population decoding improves direction readout supralinearly over individual cells

•

Strong spike correlations further enhance readout through increased synergy

Saliency model based on a neural population for integrating figure direction and organizing Border Ownership

Attentional selection is a function of the brain that allocates computational resources momentarily to the most important part of a visual scene. Saliency map models have been used to predict the location of attentional selection and gaze. Border Ownership (BO) indicates the direction of the figure with respect to the border. I here propose a biologically plausible saliency model based on neural population for integrating the activities of intermediate-level visual areas with neurons selective to BO. A variety of BO organizations produces a population of model neurons that represent the grouping structure. In the model I propose, the interactions and the population responses of these model neurons underlie the determination of saliency and the accurate prediction of gaze location. I tested 100 patterns for BO organizations and found that the proposed saliency model not only reproduced the characteristics of perceptual organization but also captured object locations in natural images. Furthermore, the saliency model based on the population responses of the BO organization significantly improved the gaze prediction accuracy compared with previous saliency-based models. These results suggest a crucial role for a wide variety of BO organizations and neural population coding to determine saliency mediating attentional selection and to predict gaze location.

Gap Junctions Contribute to Differential Light Adaptation across Direction-Selective Retinal Ganglion Cells

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.

Neural Mechanisms of Motion Processing in the Mammalian Retina

Visual motion on the retina activates a cohort of retinal ganglion cells (RGCs). This population activity encodes multiple streams of information extracted by parallel retinal circuits. Motion processing in the retina is best studied in the direction-selective circuit. The main focus of this review is the neural basis of direction selectivity, which has been investigated in unprecedented detail using state-of-the-art functional, connectomic, and modeling methods. Mechanisms underlying the encoding of other motion features by broader RGC populations are also discussed. Recent discoveries at both single-cell and population levels highlight the dynamic and stimulus-dependent engagement of multiple mechanisms that collectively implement robust motion detection under diverse visual conditions.

Formation of retinal direction-selective circuitry initiated by starburst amacrine cell homotypic contact

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.

Visual Neurons in the Superior Colliculus Innervated by Islet2+ or Islet2- Retinal Ganglion Cells Display Distinct Tuning Properties

Throughout the visual system, different subtypes of neurons are tuned to distinct aspects of the visual scene, establishing parallel circuits. Defining the mechanisms by which such tuning arises has been a long-standing challenge for neuroscience. To investigate this, we have focused on the retina’s projection to the superior colliculus (SC), where multiple visual neuron subtypes have been described. The SC receives inputs from a variety of retinal ganglion cell (RGC) subtypes; however, which RGCs drive the tuning of different SC neurons remains unclear. Here, we pursued a genetic approach that allowed us to determine the tuning properties of neurons innervated by molecularly defined subpopulations of RGCs. In homozygous Islet2-EphA3 knock-in (Isl2^EA3/EA3) mice, Isl2⁺ and Isl2⁻ RGCs project to non-overlapping sub-regions of the SC. Based on molecular and anatomic data, we show that significantly more Isl2⁻ RGCs are direction-selective (DS) in comparison with Isl2⁺ RGCs. Targeted recordings of visual responses from each SC sub-region in Isl2^EA3/EA3 mice revealed that Isl2⁻ RGC-innervated neurons were significantly more DS than those innervated by Isl2⁺ RGCs. Axis-selective (AS) neurons were found in both sub-regions, though AS neurons innervated by Isl2⁺ RGCs were more tightly tuned. Despite this segregation, DS and AS neurons innervated by Isl2⁺ or Isl2⁻ RGCs did not differ in their spatial summation or spatial frequency (SF) tuning. Further, we did not observe alterations in receptive field (RF) size or structure of SC neurons innervated by Isl2⁺ or Isl2⁻ RGCs. Together, these data show that innervation by Isl2⁺ and Isl2⁻ RGCs results in distinct tuning in the SC and set the stage for future studies investigating the mechanisms by which these circuits are built.

Long-term optical imaging of neurovascular coupling in mouse cortex using GCaMP6f and intrinsic hemodynamic signals

Cerebral hemodynamics are modulated in response to changes in neuronal activity, a process termed neurovascular coupling (NVC), which can be disrupted by neuropsychiatric diseases (e.g., stroke, Alzheimer's disease). Thus, there is growing interest to image long-term NVC dynamics with high spatiotemporal resolutions. Here, by combining the use of a genetically-encoded calcium indicator with optical techniques, we develop a longitudinal multimodal optical imaging platform (MIP) that enabled time-lapse tracking of NVC over a relatively large field of view in the mouse somatosensory cortex at single cell and single vessel resolutions. Specifically, GCaMP6f was used as marker of neuronal activity, which along with MIP allowed us to simultaneously measure the changes in neuronal [Ca²⁺]_i fluorescence, cerebral blood flow velocity (CBFv) and hemodynamics longitudinally for more than eight weeks. We show that [Ca²⁺]_i fluorescence was detectable one week post viral injection and the damage to local microvasculature and perfusion recovered two weeks after injection. By three weeks post viral injection, maximal neuronal and CBFv responses to hindpaw stimulations were observed. Moreover, single neuronal activation in response to hindpaw stimulation was consistently recorded, followed by ∼2 s delayed dilation of contiguous microvessels. Additionally, resting-state spontaneous neuronal and hemodynamic oscillations were detectable throughout the eight weeks of study. Our results demonstrate the capability of MIP for longitudinal investigation of the organization and plasticity of the neurovascular network during resting state and during stimulation-evoked neuronal activation at high spatiotemporal resolutions.

Loss of Neuroligin3 specifically downregulates retinal GABAAα2 receptors without abolishing direction selectivity

The postsynaptic adhesion proteins Neuroligins (NLs) are essential for proper synapse function, and their alterations are associated with a variety of neurodevelopmental disorders. It is increasingly clear that each NL isoform occupies specific subsets of synapses and is able to regulate the function of discrete networks. Studies of NL2 and NL4 in the retina in particular have contributed towards uncovering their role in inhibitory synapse function. In this study we show that NL3 is also predominantly expressed at inhibitory postsynapses in the retinal inner plexiform layer (IPL), where it colocalizes with both GABAA- and glycinergic receptor clusters in a 3:2 ratio. In the NL3 deletion-mutant (knockout or KO) mouse, we uncovered a dramatic reduction of the number of GABAAα2-subunit containing GABAA receptor clusters at the IPL. Retinal activity was thereafter assessed in KO and wild-type (WT) littermates by multi-electrode-array recordings of the output cells of retina, the retinal ganglion cells (RGCs). RGCs in the NL3 KO showed reduced spontaneous activity and an altered response to white noise stimulation. Moreover, upon application of light flashes, the proportion of cells firing at light offset (OFF RGCs) was significantly lower in the NL3 KO compared to WT littermates, whereas the relative number of cells firing at light onset (ON RGCs) increased. Interestingly, although GABAAα2-bearing receptors have been related to direction-selective circuits of the retina, features of direction selective-retinal ganglion cells recorded remained unperturbed in the NL3 KO. Together our data underscore the importance of NL3 for the integrity of specific GABAAergic retinal circuits and identifies NL3 as an important regulator of retinal activity.

Complex object motion represented by context-dependent correlated activity of visual interneurones

Accurate and adaptive encoding of complex, dynamic visual information is critical for the survival of many animals. Studies across a range of taxa have investigated behavioral and neuronal responses to objects that represent a threat, such as a looming object approaching along a direct collision course. By investigating neural mechanisms of avoidance behaviors through recording multineuronal activity, it is possible to better understand how complex visual information is represented in circuits that ultimately drive behaviors. We used multichannel electrodes to record from the well-studied locust nervous system to explore how object motion is reflected in activity of correlated neural activity. We presented locusts (Locusta migratoria) with objects that moved along one of 11 unique trajectories and recorded from descending interneurons within the ventral nerve cord. Spike sorting resulted in 405 discriminated units across 20 locusts and we found that 75% of the units responded to some form of object motion. Dimensionality reduction through principal component (PCA) and dynamic factor (DFA) analyses revealed population vector responses within individuals and common firing trends across the pool of discriminated units, respectively. Population vector composition (PCA) varied with the stimulus and common trends (DFA) showed unique tuning related to changes in the visual size and trajectory of the object through time. These findings demonstrate that this well-described collision detection system is more complex than previously envisioned and will drive future experiments to explore fundamental principles of how visual information is processed through context-dependent dynamic ensembles of neurons to initiate and control complex behavior.

Representation of Multidimensional Stimuli: Quantifying the Most Informative Stimulus Dimension from Neural Responses

A common way to assess the function of sensory neurons is to measure the number of spikes produced by individual neurons while systematically varying a given dimension of the stimulus. Such measured tuning curves can then be used to quantify the accuracy of the neural representation of the stimulus dimension under study, which can in turn be related to behavioral performance. However, tuning curves often change shape when other dimensions of the stimulus are varied, reflecting the simultaneous sensitivity of neurons to multiple stimulus features. Here we illustrate how one-dimensional information analyses are misleading in this context, and propose a framework derived from Fisher information that allows the quantification of information carried by neurons in multidimensional stimulus spaces. We use this method to probe the representation of sound localization in auditory neurons of chinchillas and guinea pigs of both sexes, and show how heterogeneous tuning properties contribute to a representation of sound source position that is robust to changes in sound level.SIGNIFICANCE STATEMENT Sensory neurons' responses are typically modulated simultaneously by numerous stimulus properties, which can result in an overestimation of neural acuity with existing one-dimensional neural information transmission measures. To overcome this limitation, we develop new, compact expressions of Fisher information-derived measures that bound the robust encoding of separate stimulus dimensions in the context of multidimensional stimuli. We apply this method to the problem of the representation of sound source location in the face of changes in sound source level by neurons of the auditory midbrain.

Hierarchical differences in population coding within auditory cortex

Most models of auditory cortical (AC) population coding have focused on primary auditory cortex (A1). Thus our understanding of how neural coding for sounds progresses along the cortical hierarchy remains obscure. To illuminate this, we recorded from two AC fields: A1 and middle lateral belt (ML) of rhesus macaques. We presented amplitude-modulated (AM) noise during both passive listening and while the animals performed an AM detection task ("active" condition). In both fields, neurons exhibit monotonic AM-depth tuning, with A1 neurons mostly exhibiting increasing rate-depth functions and ML neurons approximately evenly distributed between increasing and decreasing functions. We measured noise correlation (r_noise) between simultaneously recorded neurons and found that whereas engagement decreased average r_noise in A1, engagement increased average r_noise in ML. This finding surprised us, because attentive states are commonly reported to decrease average r_noise We analyzed the effect of r_noise on AM coding in both A1 and ML and found that whereas engagement-related shifts in r_noise in A1 enhance AM coding, r_noise shifts in ML have little effect. These results imply that the effect of r_noise differs between sensory areas, based on the distribution of tuning properties among the neurons within each population. A possible explanation of this is that higher areas need to encode nonsensory variables (e.g., attention, choice, and motor preparation), which impart common noise, thus increasing r_noise Therefore, the hierarchical emergence of r_noise-robust population coding (e.g., as we observed in ML) enhances the ability of sensory cortex to integrate cognitive and sensory information without a loss of sensory fidelity.NEW & NOTEWORTHY Prevailing models of population coding of sensory information are based on a limited subset of neural structures. An important and under-explored question in neuroscience is how distinct areas of sensory cortex differ in their population coding strategies. In this study, we compared population coding between primary and secondary auditory cortex. Our findings demonstrate striking differences between the two areas and highlight the importance of considering the diversity of neural structures as we develop models of population coding.

Structures of Neural Correlation and How They Favor Coding

The neural representation of information suffers from "noise"-the trial-to-trial variability in the response of neurons. The impact of correlated noise upon population coding has been debated, but a direct connection between theory and experiment remains tenuous. Here, we substantiate this connection and propose a refined theoretical picture. Using simultaneous recordings from a population of direction-selective retinal ganglion cells, we demonstrate that coding benefits from noise correlations. The effect is appreciable already in small populations, yet it is a collective phenomenon. Furthermore, the stimulus-dependent structure of correlation is key. We develop simple functional models that capture the stimulus-dependent statistics. We then use them to quantify the performance of population coding, which depends upon interplays of feature sensitivities and noise correlations in the population. Because favorable structures of correlation emerge robustly in circuits with noisy, nonlinear elements, they will arise and benefit coding beyond the confines of retina.

Role for Visual Experience in the Development of Direction-Selective Circuits

Visually guided behavior can depend critically on detecting the direction of object movement. This computation is first performed in the retina where direction is encoded by direction-selective ganglion cells (DSGCs) that respond strongly to an object moving in the preferred direction and weakly to an object moving in the opposite, or null, direction (reviewed in [1]). DSGCs come in multiple types that are classified based on their morphologies, response properties, and targets in the brain. This study focuses on two types-ON and ON-OFF DSGCs. Though animals can sense motion in all directions, the preferred directions of DSGCs in adult retina cluster along distinct directions that we refer to as the cardinal axes. ON DSGCs have three cardinal axes-temporal, ventral, and dorsonasal-while ON-OFF DSGCs have four-nasal, temporal, dorsal, and ventral. How these preferred directions emerge during development is still not understood. Several studies have demonstrated that ON [2] and ON-OFF DSGCs are well tuned at eye-opening, and even a few days prior to eye-opening, in rabbits [3], rats [4], and mice [5-8], suggesting that visual experience is not required to produce direction-selective tuning. However, here we show that at eye-opening the preferred directions of both ON and ON-OFF DSGCs are diffusely distributed and that visual deprivation prevents the preferred directions from clustering along the cardinal axes. Our findings indicate a critical role for visual experience in shaping responses in the retina.

Cortical Axons, Isolated in Channels, Display Activity-Dependent Signal Modulation as a Result of Targeted Stimulation

Mammalian cortical axons are extremely thin processes that are difficult to study as a result of their small diameter: they are too narrow to patch while intact, and super-resolution microscopy is needed to resolve single axons. We present a method for studying axonal physiology by pairing a high-density microelectrode array with a microfluidic axonal isolation device, and use it to study activity-dependent modulation of axonal signal propagation evoked by stimulation near the soma. Up to three axonal branches from a single neuron, isolated in different channels, were recorded from simultaneously using 10-20 electrodes per channel. The axonal channels amplified spikes such that propagations of individual signals along tens of electrodes could easily be discerned with high signal to noise. Stimulation from 10 up to 160 Hz demonstrated similar qualitative results from all of the cells studied: extracellular action potential characteristics changed drastically in response to stimulation. Spike height decreased, spike width increased, and latency increased, as a result of reduced propagation velocity, as the number of stimulations and the stimulation frequencies increased. Quantitatively, the strength of these changes manifested itself differently in cells at different frequencies of stimulation. Some cells' signal fidelity fell to 80% already at 10 Hz, while others maintained 80% signal fidelity at 80 Hz. Differences in modulation by axonal branches of the same cell were also seen for different stimulation frequencies, starting at 10 Hz. Potassium ion concentration changes altered the behavior of the cells causing propagation failures at lower concentrations and improving signal fidelity at higher concentrations.