Kinetic features dictate sensorimotor alignment in the superior colliculus.

The execution of goal-oriented behaviours requires a spatially coherent alignment between sensory and motor maps. The current model for sensorimotor transformation in the superior colliculus relies on the topographic mapping of static spatial receptive fields onto movement endpoints1-6. Here, to experimentally assess the validity of this canonical static model of alignment, we dissected the visuo-motor network in the superior colliculus and performed in vivo intracellular and extracellular recordings across layers, in restrained and unrestrained conditions, to assess both the motor and the visual tuning of individual motor and premotor neurons. We found that collicular motor units have poorly defined visual static spatial receptive fields and respond instead to kinetic visual features, revealing the existence of a direct alignment in vectorial space between sensory and movement vectors, rather than between spatial receptive fields and movement endpoints as canonically hypothesized. We show that a neural network built according to these kinetic alignment principles is ideally placed to sustain ethological behaviours such as the rapid interception of moving and static targets. These findings reveal a novel dimension of the sensorimotor alignment process. By extending the alignment from the static to the kinetic domain this work provides a novel conceptual framework for understanding the nature of sensorimotor convergence and its relevance in guiding goal-directed behaviours.

Synaptic architecture of leg and wing premotor control networks in Drosophila.

Animal movement is controlled by motor neurons (MNs), which project out of the central nervous system to activate muscles1. MN activity is coordinated by complex premotor networks that facilitate the contribution of individual muscles to many different behaviours2-6. Here we use connectomics7 to analyse the wiring logic of premotor circuits controlling the Drosophila leg and wing. We find that both premotor networks cluster into modules that link MNs innervating muscles with related functions. Within most leg motor modules, the synaptic weights of each premotor neuron are proportional to the size of their target MNs, establishing a circuit basis for hierarchical MN recruitment. By contrast, wing premotor networks lack proportional synaptic connectivity, which may enable more flexible recruitment of wing steering muscles. Through comparison of the architecture of distinct motor control systems within the same animal, we identify common principles of premotor network organization and specializations that reflect the unique biomechanical constraints and evolutionary origins of leg and wing motor control.

A Continuous Attractor Model with Realistic Neural and Synaptic Properties Quantitatively Reproduces Grid Cell Physiology.

Computational simulations with data-driven physiological detail can foster a deeper understanding of the neural mechanisms involved in cognition. Here, we utilize the wealth of cellular properties from Hippocampome.org to study neural mechanisms of spatial coding with a spiking continuous attractor network model of medial entorhinal cortex circuit activity. The primary goal is to investigate if adding such realistic constraints could produce firing patterns similar to those measured in real neurons. Biological characteristics included in the work are excitability, connectivity, and synaptic signaling of neuron types defined primarily by their axonal and dendritic morphologies. We investigate the spiking dynamics in specific neuron types and the synaptic activities between groups of neurons. Modeling the rodent hippocampal formation keeps the simulations to a computationally reasonable scale while also anchoring the parameters and results to experimental measurements. Our model generates grid cell activity that well matches the spacing, size, and firing rates of grid fields recorded in live behaving animals from both published datasets and new experiments performed for this study. Our simulations also recreate different scales of those properties, e.g., small and large, as found along the dorsoventral axis of the medial entorhinal cortex. Computational exploration of neuronal and synaptic model parameters reveals that a broad range of neural properties produce grid fields in the simulation. These results demonstrate that the continuous attractor network model of grid cells is compatible with a spiking neural network implementation sourcing data-driven biophysical and anatomical parameters from Hippocampome.org. The software (version 1.0) is released as open source to enable broad community reuse and encourage novel applications.

Mapping model units to visual neurons reveals population code for social behaviour.

The rich variety of behaviours observed in animals arises through the interplay between sensory processing and motor control. To understand these sensorimotor transformations, it is useful to build models that predict not only neural responses to sensory input1-5 but also how each neuron causally contributes to behaviour6,7. Here we demonstrate a novel modelling approach to identify a one-to-one mapping between internal units in a deep neural network and real neurons by predicting the behavioural changes that arise from systematic perturbations of more than a dozen neuronal cell types. A key ingredient that we introduce is 'knockout training', which involves perturbing the network during training to match the perturbations of the real neurons during behavioural experiments. We apply this approach to model the sensorimotor transformations of Drosophila melanogaster males during a complex, visually guided social behaviour8-11. The visual projection neurons at the interface between the optic lobe and central brain form a set of discrete channels12, and prior work indicates that each channel encodes a specific visual feature to drive a particular behaviour13,14. Our model reaches a different conclusion: combinations of visual projection neurons, including those involved in non-social behaviours, drive male interactions with the female, forming a rich population code for behaviour. Overall, our framework consolidates behavioural effects elicited from various neural perturbations into a single, unified model, providing a map from stimulus to neuronal cell type to behaviour, and enabling future incorporation of wiring diagrams of the brain15 into the model.

From pixels to connections: exploring in vitro neuron reconstruction software for network graph generation.

Digital reconstruction has been instrumental in deciphering how in vitro neuron architecture shapes information flow. Emerging approaches reconstruct neural systems as networks with the aim of understanding their organization through graph theory. Computational tools dedicated to this objective build models of nodes and edges based on key cellular features such as somata, axons, and dendrites. Fully automatic implementations of these tools are readily available, but they may also be purpose-built from specialized algorithms in the form of multi-step pipelines. Here we review software tools informing the construction of network models, spanning from noise reduction and segmentation to full network reconstruction. The scope and core specifications of each tool are explicitly defined to assist bench scientists in selecting the most suitable option for their microscopy dataset. Existing tools provide a foundation for complete network reconstruction, however more progress is needed in establishing morphological bases for directed/weighted connectivity and in software validation.

Statistically inferred neuronal connections in subsampled neural networks strongly correlate with spike train covariances.

Statistically inferred neuronal connections from observed spike train data are often skewed from ground truth by factors such as model mismatch, unobserved neurons, and limited data. Spike train covariances, sometimes referred to as "functional connections," are often used as a proxy for the connections between pairs of neurons, but reflect statistical relationships between neurons, not anatomical connections. Moreover, covariances are not causal: spiking activity is correlated in both the past and the future, whereas neurons respond only to synaptic inputs in the past. Connections inferred by maximum likelihood inference, however, can be constrained to be causal. However, we show in this work that the inferred connections in spontaneously active networks modeled by stochastic leaky integrate-and-fire networks strongly correlate with the covariances between neurons, and may reflect noncausal relationships, when many neurons are unobserved or when neurons are weakly coupled. This phenomenon occurs across different network structures, including random networks and balanced excitatory-inhibitory networks. We use a combination of simulations and a mean-field analysis with fluctuation corrections to elucidate the relationships between spike train covariances, inferred synaptic filters, and ground-truth connections in partially observed networks.

Is Learning in Biological Neural Networks Based on Stochastic Gradient Descent? An Analysis Using Stochastic Processes.

In recent years, there has been an intense debate about how learning in biological neural networks (BNNs) differs from learning in artificial neural networks. It is often argued that the updating of connections in the brain relies only on local information, and therefore a stochastic gradient-descent type optimization method cannot be used. In this note, we study a stochastic model for supervised learning in BNNs. We show that a (continuous) gradient step occurs approximately when each learning opportunity is processed by many local updates. This result suggests that stochastic gradient descent may indeed play a role in optimizing BNNs.

Network-level encoding of local neurotransmitters in cortical astrocytes.

Astrocytes, the most abundant non-neuronal cell type in the mammalian brain, are crucial circuit components that respond to and modulate neuronal activity through calcium (Ca2+) signalling1-7. Astrocyte Ca2+ activity is highly heterogeneous and occurs across multiple spatiotemporal scales-from fast, subcellular activity3,4 to slow, synchronized activity across connected astrocyte networks8-10-to influence many processes5,7,11. However, the inputs that drive astrocyte network dynamics remain unclear. Here we used ex vivo and in vivo two-photon astrocyte imaging while mimicking neuronal neurotransmitter inputs at multiple spatiotemporal scales. We find that brief, subcellular inputs of GABA and glutamate lead to widespread, long-lasting astrocyte Ca2+ responses beyond an individual stimulated cell. Further, we find that a key subset of Ca2+ activity-propagative activity-differentiates astrocyte network responses to these two main neurotransmitters, and may influence responses to future inputs. Together, our results demonstrate that local, transient neurotransmitter inputs are encoded by broad cortical astrocyte networks over a minutes-long time course, contributing to accumulating evidence that substantial astrocyte-neuron communication occurs across slow, network-level spatiotemporal scales12-14. These findings will enable future studies to investigate the link between specific astrocyte Ca2+ activity and specific functional outputs, which could build a consistent framework for astrocytic modulation of neuronal activity.

Mechanical stimulation and electrophysiological monitoring at subcellular resolution reveals differential mechanosensation of neurons within networks.

A growing consensus that the brain is a mechanosensitive organ is driving the need for tools that mechanically stimulate and simultaneously record the electrophysiological response of neurons within neuronal networks. Here we introduce a synchronized combination of atomic force microscopy, high-density microelectrode array and fluorescence microscopy to monitor neuronal networks and to mechanically characterize and stimulate individual neurons at piconewton force sensitivity and nanometre precision while monitoring their electrophysiological activity at subcellular spatial and millisecond temporal resolution. No correlation is found between mechanical stiffness and electrophysiological activity of neuronal compartments. Furthermore, spontaneously active neurons show exceptional functional resilience to static mechanical compression of their soma. However, application of fast transient (∼500 ms) mechanical stimuli to the neuronal soma can evoke action potentials, which depend on the anchoring of neuronal membrane and actin cytoskeleton. Neurons show higher responsivity, including bursts of action potentials, to slower transient mechanical stimuli (∼60 s). Moreover, transient and repetitive application of the same compression modulates the neuronal firing rate. Seemingly, neuronal networks can differentiate and respond to specific characteristics of mechanical stimulation. Ultimately, the developed multiparametric tool opens the door to explore manifold nanomechanobiological responses of neuronal systems and new ways of mechanical control.

Structural and functional asymmetry of the neonatal cerebral cortex.

Features of brain asymmetry have been implicated in a broad range of cognitive processes; however, their origins are still poorly understood. Here we investigated cortical asymmetries in 442 healthy term-born neonates using structural and functional magnetic resonance images from the Developing Human Connectome Project. Our results demonstrate that the neonatal cortex is markedly asymmetric in both structure and function. Cortical asymmetries observed in the term cohort were contextualized in two ways: by comparing them against cortical asymmetries observed in 103 preterm neonates scanned at term-equivalent age, and by comparing structural asymmetries against those observed in 1,110 healthy young adults from the Human Connectome Project. While associations with preterm birth and biological sex were minimal, significant differences exist between birth and adulthood.

Structured cerebellar connectivity supports resilient pattern separation.

The cerebellum is thought to help detect and correct errors between intended and executed commands1,2 and is critical for social behaviours, cognition and emotion3-6. Computations for motor control must be performed quickly to correct errors in real time and should be sensitive to small differences between patterns for fine error correction while being resilient to noise7. Influential theories of cerebellar information processing have largely assumed random network connectivity, which increases the encoding capacity of the network's first layer8-13. However, maximizing encoding capacity reduces the resilience to noise7. To understand how neuronal circuits address this fundamental trade-off, we mapped the feedforward connectivity in the mouse cerebellar cortex using automated large-scale transmission electron microscopy and convolutional neural network-based image segmentation. We found that both the input and output layers of the circuit exhibit redundant and selective connectivity motifs, which contrast with prevailing models. Numerical simulations suggest that these redundant, non-random connectivity motifs increase the resilience to noise at a negligible cost to the overall encoding capacity. This work reveals how neuronal network structure can support a trade-off between encoding capacity and redundancy, unveiling principles of biological network architecture with implications for the design of artificial neural networks.

Sparse balance: Excitatory-inhibitory networks with small bias currents and broadly distributed synaptic weights.

Cortical circuits generate excitatory currents that must be cancelled by strong inhibition to assure stability. The resulting excitatory-inhibitory (E-I) balance can generate spontaneous irregular activity but, in standard balanced E-I models, this requires that an extremely strong feedforward bias current be included along with the recurrent excitation and inhibition. The absence of experimental evidence for such large bias currents inspired us to examine an alternative regime that exhibits asynchronous activity without requiring unrealistically large feedforward input. In these networks, irregular spontaneous activity is supported by a continually changing sparse set of neurons. To support this activity, synaptic strengths must be drawn from high-variance distributions. Unlike standard balanced networks, these sparse balance networks exhibit robust nonlinear responses to uniform inputs and non-Gaussian input statistics. Interestingly, the speed, not the size, of synaptic fluctuations dictates the degree of sparsity in the model. In addition to simulations, we provide a mean-field analysis to illustrate the properties of these networks.

Global organization of neuronal activity only requires unstructured local connectivity.

Modern electrophysiological recordings simultaneously capture single-unit spiking activities of hundreds of neurons spread across large cortical distances. Yet, this parallel activity is often confined to relatively low-dimensional manifolds. This implies strong coordination also among neurons that are most likely not even connected. Here, we combine in vivo recordings with network models and theory to characterize the nature of mesoscopic coordination patterns in macaque motor cortex and to expose their origin: We find that heterogeneity in local connectivity supports network states with complex long-range cooperation between neurons that arises from multi-synaptic, short-range connections. Our theory explains the experimentally observed spatial organization of covariances in resting state recordings as well as the behaviorally related modulation of covariance patterns during a reach-to-grasp task. The ubiquity of heterogeneity in local cortical circuits suggests that the brain uses the described mechanism to flexibly adapt neuronal coordination to momentary demands.

Electrophysiological responses to seizurogenic compounds dependent on E/I balance in human iPSC-derived cortical neural networks.

Construction of in vitro functional assay systems using human-induced pluripotent stem cells (iPSCs) as indicators for evaluating seizure liability of compounds has been anticipated. Imbalance of excitation/inhibition (E/I) inputs triggers seizure; however, the appropriate ratio of E/I neurons for evaluating seizure liability of compounds in a human iPSC-derived neural network is unknown. Here, five neural networks with varying E/I ratios (88/12, 84/16, 74/26, 58/42, and 48/52) were constructed by altering the ratios of glutamatergic (E) and GABA (I) neurons. The responsiveness of each network against six seizurogenic compounds and two GABA receptor agonists was then examined by using six representative parameters. The 52% GABA neuron network, which had the highest ratio of GABA neurons, showed the most marked response to seizurogenic compounds, however, it suggested the possibility of producing false positives. Moreover, analytical parameters were found to vary with E/I ratio and to differ for seizurogenic compounds with different mechanism of action (MoA) even at the same E/I ratio. Clustering analysis using six parameters showed the balance of 84/16, which is the closest to the biological balance, was the most suitable for detection of concentration-dependent change and classification of the MoA of seizurogenic compounds. These results suggest the importance of using a human-iPSC-derived neural network similar to the E/I balance of the living body in order to improve the prediction accuracy in the in vitro seizure liability assessment.

Engineering brain assembloids to interrogate human neural circuits.

The development of neural circuits involves wiring of neurons locally following their generation and migration, as well as establishing long-distance connections between brain regions. Studying these developmental processes in the human nervous system remains difficult because of limited access to tissue that can be maintained as functional over time in vitro. We have previously developed a method to convert human pluripotent stem cells into brain region-specific organoids that can be fused and integrated to form assembloids and study neuronal migration. In contrast to approaches that mix cell lineages in 2D cultures or engineer microchips, assembloids leverage self-organization to enable complex cell-cell interactions, circuit formation and maturation in long-term cultures. In this protocol, we describe approaches to model long-range neuronal connectivity in human brain assembloids. We present how to generate 3D spheroids resembling specific domains of the nervous system and then how to integrate them physically to allow axonal projections and synaptic assembly. In addition, we describe a series of assays including viral labeling and retrograde tracing, 3D live imaging of axon projection and optogenetics combined with calcium imaging and electrophysiological recordings to probe and manipulate the circuits in assembloids. The assays take 3-4 months to complete and require expertise in stem cell culture, imaging and electrophysiology. We anticipate that these approaches will be useful in deciphering human-specific aspects of neural circuit assembly and in modeling neurodevelopmental disorders with patient-derived cells.

Structure of Long-Range Direct and Indirect Spinocerebellar Pathways as Well as Local Spinal Circuits Mediating Proprioception.

Proprioception, the sense of limb and body position, generates a map of the body that is essential for proper motor control, yet we know little about precisely how neurons in proprioceptive pathways are wired. Defining the anatomy of secondary neurons in the spinal cord that integrate and relay proprioceptive and potentially cutaneous information from the periphery to the cerebellum is fundamental to understanding how proprioceptive circuits function. Here, we define the unique anatomic trajectories of long-range direct and indirect spinocerebellar pathways as well as local intersegmental spinal circuits using genetic tools in both male and female mice. We find that Clarke's column neurons, a major contributor to the direct spinocerebellar pathway, has mossy fiber terminals that diversify extensively in the cerebellar cortex with axons terminating bilaterally, but with no significant axon collaterals within the spinal cord, medulla, or cerebellar nuclei. By contrast, we find that two of the indirect pathways, the spino-lateral reticular nucleus and spino-olivary pathways, are in part, derived from cervical Atoh1-lineage neurons, whereas thoracolumbar Atoh1-lineage neurons project mostly locally within the spinal cord. Notably, while cervical and thoracolumbar Atoh1-lineage neurons connect locally with motor neurons, no Clarke's column to motor neuron connections were detected. Together, we define anatomic differences between long-range direct, indirect, and local proprioceptive subcircuits that likely mediate different components of proprioceptive-motor behaviors.SIGNIFICANCE STATEMENT We define the anatomy of long-range direct and indirect spinocerebellar pathways as well as local spinal proprioceptive circuits. We observe that mossy fiber axon terminals of Clarke's column neurons diversify proprioceptive information across granule cells in multiple lobules on both ipsilateral and contralateral sides, sending no significant collaterals within the spinal cord, medulla, or cerebellar nuclei. Strikingly, we find that cervical spinal cord Atoh1-lineage neurons form mainly the indirect spino-lateral reticular nucleus and spino-olivary tracts and thoracolumbar Atoh1-lineage neurons project locally within the spinal cord, whereas only a few Atoh1-lineage neurons form a direct spinocerebellar tract.

When shared concept cells support associations: Theory of overlapping memory engrams.

Assemblies of neurons, called concepts cells, encode acquired concepts in human Medial Temporal Lobe. Those concept cells that are shared between two assemblies have been hypothesized to encode associations between concepts. Here we test this hypothesis in a computational model of attractor neural networks. We find that for concepts encoded in sparse neural assemblies there is a minimal fraction cmin of neurons shared between assemblies below which associations cannot be reliably implemented; and a maximal fraction cmax of shared neurons above which single concepts can no longer be retrieved. In the presence of a periodically modulated background signal, such as hippocampal oscillations, recall takes the form of association chains reminiscent of those postulated by theories of free recall of words. Predictions of an iterative overlap-generating model match experimental data on the number of concepts to which a neuron responds.

Distributed sensory coding by cerebellar complex spikes in units of cortical segments.

Sensory processing is essential for motor control. Climbing fibers from the inferior olive transmit sensory signals to Purkinje cells, but how the signals are represented in the cerebellar cortex remains elusive. To examine the olivocerebellar organization of the mouse brain, we perform quantitative Ca2+ imaging to measure complex spikes (CSs) evoked by climbing fiber inputs over the entire dorsal surface of the cerebellum simultaneously. The surface is divided into approximately 200 segments, each composed of ∼100 Purkinje cells that fire CSs synchronously. Our in vivo imaging reveals that, although stimulation of four limb muscles individually elicits similar global CS responses across nearly all segments, the timing and location of a stimulus are derived by Bayesian inference from coordinated activation and inactivation of multiple segments on a single trial basis. We propose that the cerebellum performs segment-based, distributed-population coding that represents the conditional probability of sensory events.