The neocortex is composed of spiking neuronal units interconnected in a sparse, recurrent network. Neuronal networks exhibit spiking activity that transforms sensory inputs into appropriate behavioral outputs. In this study, we train biologically realistic spiking neural network (SNN) models to identify the architectural changes which enable task-appropriate computations. Specifically, we employ a binary state change detection task, where each state is defined by motion entropy. This task mirrors behavioral paradigms that mice perform in the lab. SNNs are composed of excitatory and inhibitory units randomly interconnected with connection likelihoods and strengths matched to observations from mouse neocortex. Following training, we discover that SNNs selectively adjust firing rates depending on state, and that excitatory and inhibitory connectivity between input and recurrent layers change in accordance with this rate modulation. Input channels that exhibit bias to one specific motion entropy input develop stronger connections to recurrent excitatory units during training, while channels that exhibit bias to the other input develop stronger connections to inhibitory units. Furthermore, recurrent inhibitory units which positively modulated firing rates to one input strengthened their connections to recurrent units of the opposite modulation. This specific pattern of cross-modulation inhibition emerged as the optimal solution when imposing Dale’s law throughout training of the SNNs. Removing this constraint led to the absence of the emergence of this architectural solution. This work highlights the critical role of interneurons and the specific architectural patterns of inhibition in shaping dynamics and information processing within neocortical circuits.

boRxiv

The representational drift observed in neural populations raises serious questions about how accurate decoding survives these changes. In primary visual cortex, it is hotly debated whether such variation is a direct tuning shift that would corrupt decoding or if it can be explained by changes in behavioral or internal state, which could be compensated by joint encoding of the stimulus and the state. We estimate the effects of stimulus-relevant representational drift on decoding using a publicly accessible dataset of mouse V1 responses to a natural movie. Because the only invariant component of the sensory experience across all 24 animals is that they all watch the same natural movie, we can learn a subject-invariant efficient neural representation that retains only stimulus-relevant components. We use contrastive learning between the neural response and the stimulus to learn a neural representation for stimulus-relevant features. This learned representation minimizes decoding error as quantified by Bayes risk. We show that it can be used to read out behaviorally relevant stimulus features (time, static scene, optic flow, and joint spatio-temporal features) at 33ms resolution accurately, a finer timescale than what has previously been explored. When we use the model trained on one recording session to derive feature activations on another, decoding performance is reduced by approximately . Motion encoding is most susceptible to representational drift. In addition, when we enlarge the error tolerance window to 1sec, we recover stable stimulus encoding across sessions, echoing previous findings. This shows that decoding stimulus features that vary on fast timescales may require complex computation downstream of V1 to compensate for representational drift.

arXiv

Skillful, voluntary movements are underpinned by computations performed by networks of interconnected neurons in the primary motor cortex (M1). Computations are reflected by patterns of coactivity between neurons. Using pairwise spike time statistics, coactivity can be summarized as a functional network (FN). Here, we show that the structure of FNs constructed from an instructed-delay reach task in nonhuman primates is behaviorally specific: Low-dimensional embedding and graph alignment scores show that FNs constructed from closer target reach directions are also closer in network space. Using short intervals across a trial, we constructed temporal FNs and found that temporal FNs traverse a low-dimensional subspace in a reach-specific trajectory. Alignment scores show that FNs become separable and correspondingly decodable shortly after the Instruction cue. Finally, we observe that reciprocal connections in FNs transiently decrease following the Instructioncue, consistent with the hypothesis that information external to the recorded population temporarily alters the structure of the network at this moment.

Network Neuroscience

When multiple stimuli appear together in the receptive field of a visual cortical neuron, the response is typically close to the average of that neuron’s response to each individual stimulus. The departure from a linear sum of each individual response is referred to as normalization. In mammals, normalization has been best characterized in the visual cortex of macaques and cats. Here we study visually evoked normalization in the visual cortex of awake mice using optical imaging of calcium indicators in large populations of layer 2/3 (L2/3) V1 excitatory neurons and electrophysiological recordings across layers in V1. Regardless of recording method, mouse visual cortical neurons exhibit normalization to varying degrees. The distributions of normalization strength are similar to those described in cats and macaques, albeit slightly weaker on average.

PLoSOne

Y. Wang, C. Wreden, M. Levy, J. Meng, Z. Marshall, J. MacLean, E. Heckscher

How circuits self-assemble starting from neuronal stem cells is a fundamental question in developmental neurobiology. Here, we addressed how neurons from different stem cell lineages wire with each other to form a specific circuit motif. In Drosophila larvae, we combined developmental genetics (twin-spot mosaic analysis with a repressible cell marker, multi-color flip out, permanent labeling) with circuit analysis (calcium imaging, connectomics, network science). For many lineages, neuronal progeny are organized into subunits called temporal cohorts. Temporal cohorts are subsets of neurons born within a tight time window that have shared circuit-level function. We find sharp transitions in patterns of input connectivity at temporal cohort boundaries. In addition, we identify a feed-forward circuit that encodes the onset of vibration stimuli. This feed-forward circuit is assembled by preferential connectivity between temporal cohorts from different lineages. Connectivity does not follow the often-cited early-to-early, late-to-late model. Instead, the circuit is formed by sequential addition of temporal cohorts from different lineages, with circuit output neurons born before circuit input neurons. Further, we generate new tools for the fly community. Our data raise the possibility that sequential addition of neurons (with outputs oldest and inputs youngest) could be one fundamental strategy for assembling feed-forward circuits.

eLife

D. Moore, J. Walker, J. MacLean, N. Hatsopoulos

To reveal the neurophysiological underpinnings of natural movement, neural recordings must be paired with accurate tracking of limbs and postures. Here, we evaluated the accuracy of DeepLabCut (DLC), a deep learning markerless motion capture approach, by comparing it with a 3D X-ray video radiography system that tracks markers placed under the skin (XROMM). We recorded behavioral data simultaneously with XROMM and RGB video as marmosets foraged and reconstructed 3D kinematics in a common coordinate system. We used the toolkit Anipose to filter and triangulate DLC trajectories of 11 markers on the forelimb and torso and found a low median error (0.228 cm) between the two modalities corresponding to 2.0% of the range of motion. For studies allowing this relatively small error, DLC and similar markerless pose estimation tools enable the study of increasingly naturalistic behaviors in many fields including non-human primate motor control.

Journal of Experimental Biology

M. Levy, J. Guo, J. MacLean

Spike trains in cortical neuronal populations vary in number and timing trial-to-trial, rendering a viable single trial coding scheme for sensory information elusive. Correlations between pairs of neocortical neurons can be segmented into either sensory or noise according to their stimulus specificity. Here we show that pairs of spikes, corresponding to reliable sensory correlations in imaged populations in layer 2/3 of mouse visual cortex are particularly informative of visual stimuli. This set of spikes is sparse and exhibits comparable levels of trial-to-trial variance relative to the full spike train. Despite this, correspondence of pairs of spikes to a specific set of sensory correlations identifies spikes that carry more information per spike about the visual stimulus than the full set or any other matched set of spikes. Moreover, this sparse subset is more accurately decoded, regardless of the decoding algorithm employed. Our findings suggest that consistent pairwise correlations between neurons, rather than first-order statistical features of spike trains, may be an organizational principle of a single trial sensory coding scheme.

bioRxiv

M. Sundiang, N. Hatsopoulos, J. MacLean

Skillful, voluntary movements are underpinned by computations performed by networks of interconnected neurons in the primary motor cortex (M1). Computations are reflected by patterns of co-activity between neurons. Using pairwise spike time statistics, co-activity can be summarized as a functional network (FN). Here, we show that the structure of FNs constructed from an instructed-delay reach task in non-human primates are behaviorally specific: low dimensional embedding and graph alignment scores show that FNs constructed from closer target reach directions are also closer in network space. Using short intervals across a trial we constructed temporal FNs and found that temporal FNs traverse a low-dimensional subspace in a reach-specific trajectory. Alignment scores show that FNs become separable and correspondingly decodable shortly after the instruction cue. Finally, we observe that reciprocal connections in FNs transiently decrease following the instruction cue consistent with the hypothesis that information external to the recorded population temporarily alters the structure of the network at this moment.

Network Neuroscience

J. Walker, F. Pirschel, M. Sundiang, M. Niekrasz, J. MacLean, N. Hatsopoulos

Marmosets are an increasingly important model system for neuroscience in part due to genetic tractability and enhanced cortical accessibility, due to a lissencephalic neocortex. However, many of the techniques generally employed to record neural activity in primates inhibit the expression of natural behaviors in marmosets precluding neurophysiological insights. To address this challenge, we have developed methods for recording neural population activity in unrestrained marmosets across multiple ethological behaviors, multiple brain states, and over multiple years. Notably, our flexible methodological design allows for replacing electrode arrays and removal of implants providing alternative experimental endpoints. We validate the method by recording sensorimotor cortical population activity in freely moving marmosets across their natural behavioral repertoire and during sleep.

Cell Reports

P. Malonis, N. Hatsopoulos, J. MacLean, M. Kaufman

Motor cortex is integral to generating voluntary movement commands. However, as a dynamical system, it is unclear how motor cortical movement commands are informed by either new or sensory-driven corrective instructions. Here, we examine population activity in the primary motor cortex of macaques during a continuous, sequential arm movement task in which the movement instruction is updated several times over the course of a trial. We use Latent Factor Analysis via Dynamical Systems (LFADS) to decompose population activity into a portion explainable via dynamics, and a stream of inferred inputs required to instruct that dynamical system. The time series of inferred inputs had several surprising properties. First, input timing was more strongly locked to target appearance than to movement onset, suggesting that variable reaction times may be a function of how inputs interact with ongoing dynamics rather than variability in instruction timing. Second, inferred inputs were tuned nearly identically for both initial and corrective movements, suggesting a commonality in the structure of inputs across visually-instructed and corrective movements that was previously obscured by the complexity of the dynamical system that is M1.

bioRxiv

K. Bojanek, Y. Zhu, J. MacLean

A basic — yet nontrivial — function which neocortical circuitry must satisfy is the ability to maintain stable spiking activity over time. Stable neocortical activity is asynchronous, critical, and low rate, and these features of spiking dynamics contribute to efficient computation and optimal information propagation. However, it remains unclear how neocortex maintains this asynchronous spiking regime. Here we algorithmically construct spiking neural network models, each composed of 5000 neurons. Network construction synthesized topological statistics from neocortex with a set of objective functions identifying naturalistic low-rate, asynchronous, and critical activity. We find that simulations run on the same topology exhibit sustained asynchronous activity under certain sets of initial membrane voltages but truncated activity under others. Synchrony, rate, and criticality do not provide a full explanation of this dichotomy. Consequently, in order to achieve mechanistic understanding of sustained asynchronous activity, we summarized activity as functional graphs where edges between units are defined by pairwise spike dependencies. We then analyzed the intersection between functional edges and synaptic connectivity- i.e. recruitment networks. Higher-order patterns, such as triplet or triangle motifs, have been tied to cooperativity and integration. We find, over time in each sustained simulation, low-variance periodic transitions between isomorphic triangle motifs in the recruitment networks. We quantify the phenomenon as a Markov process and discover that if the network fails to engage this stereotyped regime of motif dominance “cycling”, spiking activity truncates early. Cycling of motif dominance generalized across manipulations of synaptic weights and topologies, demonstrating the robustness of this regime for maintenance of network activity. Our results point to the crucial role of excitatory higher-order patterns in sustaining asynchronous activity in sparse recurrent networks. They also provide a possible explanation why such connectivity and activity patterns have been prominently reported in neocortex.

PLoS Computational Biology

J. Walker, F. Pirschel, N. Gidmark, J. MacLean, N. Hatsopoulos

Generally behavioral neuroscience studies of the common marmoset employ adaptations of well-established training methods used with macaque monkeys. However, in many cases these approaches do not readily generalize to marmosets indicating a need for alternatives. Here we present the development of one such alternate: a platform for semiautomated, voluntary in-home cage behavioral training that allows for the study of naturalistic behaviors. We describe the design and production of a modular behavioral training apparatus using CAD software and digital fabrication. We demonstrate that this apparatus permits voluntary behavioral training and data collection throughout the marmoset’s waking hours with little experimenter intervention. Furthermore, we demonstrate the use of this apparatus to reconstruct the kinematics of the marmoset’s upper limb movement during natural foraging behavior.

Journal of Neurophysiology

M. Levy, O. Sporns, J. MacLean

Unbiased and dense sampling of large populations of layer 2/3 pyramidal neurons in mouse primary visual cortex (V1) reveals two functional sub-populations: neurons tuned and untuned to drifting gratings. Whether functional interactions between these two groups contribute to the representation of visual stimuli is unclear. To examine these interactions, we summarize the population partial pairwise correlation structure as a directed and weighted graph. We find that tuned and untuned neurons have distinct topological properties, with untuned neurons occupying central positions in functional networks (FNs). Implementation of a decoder that utilizes the topology of these FNs yields accurate decoding of visual stimuli. We further show that decoding performance degrades comparably following manipulations of either tuned or untuned neurons. Our results demonstrate that untuned neurons are an integral component of V1 FNs and suggest that network interactions contain information about the stimulus that is accessible to downstream elements.

Cell Reports

S. Kotekal, J. MacLean

To develop a complete description of sensory encoding, it is necessary to account for trial-to-trial variability in cortical neurons. Using a linear model with terms corresponding to the visual stimulus, mouse running speed, and experimentally measured neuronal correlations, we modeled short term dynamics of L2/3 murine visual cortical neurons to evaluate the relative importance of each factor to neuronal variability within single trials. We find single trial predictions improve most when conditioning on the experimentally measured local correlations in comparison to predictions based on the stimulus or running speed. Specifically, accurate predictions are driven by positively co-varying and synchronously active functional groups of neurons. Including functional groups in the model enhances decoding accuracy of sensory information compared to a model that assumes neuronal independence. Functional groups, in encoding and decoding frameworks, provide an operational definition of Hebbian assemblies in which local correlations largely explain neuronal responses on individual trials.

PLoS Computational Biology

B. Chambers, M. Levy, J. Dechery, J. MacLean

A promising alternative to directly measuring the anatomical connections in a neuronal population is inferring the connections from the activity. We employ simulated spiking neuronal networks to compare and contrast commonly used inference methods that identify likely excitatory synaptic connections using statistical regularities in spike timing. We find that simple adjustments to standard algorithms improve inference accuracy: A signing procedure improves the power of unsigned mutual-information-based approaches and a correction that accounts for differences in mean and variance of background timing relationships, such as those expected to be induced by heterogeneous firing rates, increases the sensitivity of frequency-based methods. We also find that different inference methods reveal distinct subsets of the synaptic network and each method exhibits different biases in the accurate detection of reciprocity and local clustering. To correct for errors and biases specific to single inference algorithms, we combine methods into an ensemble. Ensemble predictions, generated as a linear combination of multiple inference algorithms, are more sensitive than the best individual measures alone, and are more faithful to ground-truth statistics of connectivity, mitigating biases specific to single inference methods. These weightings generalize across simulated datasets, emphasizing the potential for the broad utility of ensemble-based approaches.

Network Neuroscience

A. Sederberg, J. MacLean, S. Palmer

To compensate for sensory processing delays, the visual system must make predictions to ensure timely and appropriate behaviors. Recent work has found predictive information about the stimulus in neural populations early in vision processing, starting in the retina. However, to utilize this information, cells downstream must be able to read out the predictive information from the spiking activity of retinal ganglion cells. Here we investigate whether a downstream cell could learn efficient encoding of predictive information in its inputs from the correlations in the inputs themselves, in the absence of other instructive signals. We simulate learning driven by spiking activity recorded in salamander retina. We model a downstream cell as a binary neuron receiving a small group of weighted inputs and quantify the predictive information between activity in the binary neuron and future input. Input weights change according to spike timing–dependent learning rules during a training period. We characterize the readouts learned under spike timing–dependent synaptic update rules, finding that although the fixed points of learning dynamics are not associated with absolute optimal readouts they convey nearly all of the information conveyed by the optimal readout. Moreover, we find that learned perceptrons transmit position and velocity information of a moving-bar stimulus nearly as efficiently as optimal perceptrons. We conclude that predictive information is, in principle, readable from the perspective of downstream neurons in the absence of other inputs. This suggests an important role for feedforward prediction in sensory encoding.

Proceedings of the National Academy of Sciences

J. Dechery, J. MacLean

Visual stimuli evoke activity in visual cortical neuronal populations. Neuronal activity can be selectively modulated by particular visual stimulus parameters, such as the direction of a moving bar of light, resulting in well-defined trial averaged tuning properties. However, given any single stimulus parameter, a large number of neurons in visual cortex remain unmodulated, and the role of this untuned population is not well understood. Here, we use two-photon calcium imaging to record, in an unbiased manner, from large populations of layer 2/3 excitatory neurons in mouse primary visual cortex to describe co-varying activity on single trials in neuronal populations consisting of both tuned and untuned neurons. Specifically, we summarize pairwise covariability with an asymmetric partial correlation coefficient, allowing us to analyze the resultant population correlation structure, or functional network, with graph theory. Using the graph neighbors of a neuron, we find that the local population, including both tuned and untuned neurons, are able to predict individual neuron activity on a moment to moment basis, while also recapitulating tuning properties of tuned neurons. Variance explained in total population activity scales with the number of neurons imaged, demonstrating larger sample sizes are required to fully capture local network interactions. We also find that a specific functional triplet motif in the graph results in the best predictions, suggesting a signature of informative correlations in these populations. In summary, we show that unbiased sampling of the local population can explain single trial response variability as well as trial-averaged tuning properties in V1, and the ability to predict responses is tied to the occurrence of a functional triplet motif.

PLoS Computational Biology

J. Dechery, J. MacLean

Temporal codes are theoretically powerful encoding schemes, but their precise form in the neocortex remains unknown in part because of the large number of possible codes and the difficulty in disambiguating informative spikes from statistical noise. A biologically plausible and computationally powerful temporal coding scheme is the Hebbian assembly phase sequence (APS), which predicts reliable propagation of spikes between functionally related assemblies of neurons. Here, we sought to measure the inherent capacity of neocortical networks to produce reliable sequences of spikes, as would be predicted by an APS code. To record microcircuit activity, the scale at which computation is implemented, we used two-photon calcium imaging to densely sample spontaneous activity in murine neocortical networks ex vivo. We show that the population spike histogram is sufficient to produce a spatiotemporal progression of activity across the population. To more comprehensively evaluate the capacity for sequential spiking that cannot be explained by the overall population spiking, we identify statistically significant spike sequences. We found a large repertoire of sequence spikes that collectively comprise the majority of spiking in the circuit. Sequences manifest probabilistically and share neuron membership, resulting in unique ensembles of interwoven sequences characterizing individual spatiotemporal progressions of activity. Distillation of population dynamics into its constituent sequences provides a way to capture trial-to-trial variability and may prove to be a powerful decoding substrate in vivo. Informed by these data, we suggest that the Hebbian APS be reformulated as interwoven sequences with flexible assembly membership due to shared overlapping neurons.

Journal of Neurophysiology

J. Walker, J. MacLean, N. Hatsopoulos

The common marmoset has recently gained interest as an animal model for systems and behavioral neuroscience. This is due in part to the advent of transgenic marmosets, which affords the possibility of combining genetic manipulations with physiological recording and behavioral monitoring to study neural systems. In this review, they will argue that the marmoset provides a unique opportunity to study the neural basis of voluntary motor control from an integrative perspective. First, as an intermediate animal model, the marmoset represents an important bridge in motor system function between other primates, including humans, and rodents. Second, due to the marmoset’s small brain size and lissencephalic cortex, novel electrophysiological and optical recording technologies will allow an integrative study of cortical function at multiple spatial scales beyond that afforded by other non-human primate models. Finally, as a primate expressing an ancestral state of corticospinal organization, the marmoset offers the possibility of understanding the integrative function of cortical and spinal interneuron circuitry in isolation of more recent corticomotoneuronal elaborations. If the potential of the marmoset as a model species is to be realized, they will need to learn to work with their natural behavioral repertoire. They have concluded by considering practical aspects of studying motor systems with marmosets.

Developmental Neurobiology

C. Frye, J. MacLean

Spontaneous propagation of spiking within the local neocortical circuits of mature primary sensory areas is highly nonrandom, engaging specific sets of interconnected and functionally related neurons. These spontaneous activations promise insight into neocortical structure and function, but their properties in the first 2 wk of perinatal development are incompletely characterized. Previously, we have found that there is a minimal numerical sample, on the order of 400 cells, necessary to fully capture mature neocortical circuit dynamics. Therefore we maximized our numerical sample by using two-photon calcium imaging to observe spontaneous activity in populations of up to 1,062 neurons spanning multiple columns and layers in 52 acute coronal slices of mouse neocortex at each day from postnatal day (PND) 3 to PND 15. Slices contained either primary auditory cortex (A1) or somatosensory barrel field (S1BF), which allowed us to compare sensory modalities with markedly different developmental timelines. Between PND 3 and PND 8, populations in both areas exhibited activations of anatomically compact subgroups on the order of dozens of cells. Between PND 9 and PND 13, the spatiotemporal structure of the activity diversified to include spatially distributed activations encompassing hundreds of cells. Sparse activations covering the entire field of view dominated in slices taken on or after PND 14. These and other findings demonstrate that the developmental progression of spontaneous activations from active local modules in the first postnatal week to sparse, intermingled groups of neurons at the beginning of the third postnatal week generalizes across primary sensory areas, consistent with an intrinsic developmental trajectory independent of sensory input.

Journal of Neurophysiology

B. Chambers, J. MacLean

Linking synaptic connectivity to dynamics is key to understanding information processing in neocortex. Circuit dynamics emerge from complex interactions of interconnected neurons, necessitating that links between connectivity and dynamics be evaluated at the network level. Here we map propagating activity in large neuronal ensembles from mouse neocortex and compare it to a recurrent network model, where connectivity can be precisely measured and manipulated. We find that a dynamical feature dominates statistical descriptions of propagating activity for both neocortex and the model: convergent clusters comprised of fan-in triangle motifs, where two input neurons are themselves connected. Fan-in triangles coordinate the timing of presynaptic inputs during ongoing activity to effectively generate postsynaptic spiking. As a result, paradoxically, fan-in triangles dominate the statistics of spike propagation even in randomly connected recurrent networks. Interplay between higher-order synaptic connectivity and the integrative properties of neurons constrains the structure of network dynamics and shapes the routing of information in neocortex.

PLoS Computational Biology

B. Chambers, J. MacLean

Structured multineuronal activity patterns within local neocortical circuitry are strongly linked to sensory input, motor output, and behavioral choice. These reliable patterns of pairwise lagged firing are the consequence of connectivity since they are not present in rate-matched but unconnected Poisson nulls. It is important to relate multineuronal patterns to their synaptic underpinnings, but it is unclear how effectively statistical dependencies in spiking between neurons identify causal synaptic connections. To assess the feasibility of mapping function onto structure we used a network model that showed a diversity of multineuronal activity patterns and replicated experimental constraints on data acquisition. Using an iterative Bayesian inference algorithm, we detected a select subset of monosynaptic connections substantially more precisely than correlation-based inference, a common alternative approach. We found that precise inference of synaptic connections improved with increasing numbers of diverse multineuronal activity patterns in contrast to increased observations of a single pattern. Surprisingly, neuronal spiking was most effective and precise at revealing causal synaptic connectivity when the lags considered by the iterative Bayesian algorithm encompassed the timescale of synaptic conductance and integration (∼10 ms), rather than synaptic transmission time (∼2 ms), highlighting the importance of synaptic integration in driving postsynaptic spiking. Last, strong synaptic connections were detected preferentially, underscoring their special importance in cortical computation. Even after simulating experimental constraints, top down approaches to cortical connectivity, from function to structure, identify synaptic connections underlying multineuronal activity. These select connections are closely tied to cortical processing.

Journal of Neurophysiology

A. Sederberg, S. Palmer, J. MacLean

A behavioral response appropriate to a sensory stimulus depends on the collective activity of thousands of interconnected neurons. The majority of cortical connections arise from neighboring neurons, and thus understanding the cortical code requires characterizing information representation at the scale of the cortical microcircuit. Using two-photon calcium imaging, we densely sampled the thalamically evoked response of hundreds of neurons spanning multiple layers and columns in thalamocortical slices of mouse somatosensory cortex. We then used a biologically plausible decoder to characterize the representation of two distinct thalamic inputs, at the level of the microcircuit, to reveal those aspects of the activity pattern that are likely relevant to downstream neurons. Our data suggest a sparse code, distributed across lamina, in which a small population of cells carries stimulus-relevant information. Furthermore, we find that, within this subset of neurons, decoder performance improves when noise correlations are taken into account.

Journal of Neurophysiology

S. Gururangan, A. Sadovsky, J. MacLean

Correlations in local neocortical spiking activity can provide insight into the underlying organization of cortical microcircuitry. However, identifying structure in patterned multi-neuronal spiking remains a daunting task due to the high dimensionality of the activity. Using two-photon imaging, we monitored spontaneous circuit dynamics in large, densely sampled neuronal populations within slices of mouse primary auditory, somatosensory, and visual cortex. Using the lagged correlation of spiking activity between neurons, we generated functional wiring diagrams to gain insight into the underlying neocortical circuitry. By establishing the presence of graph invariants, which are label-independent characteristics common to all circuit topologies, our study revealed organizational features that generalized across functionally distinct cortical regions. Regardless of sensory area, random and -nearest neighbors null graphs failed to capture the structure of experimentally derived functional circuitry. These null models indicated that despite a bias in the data towards spatially proximal functional connections, functional circuit structure is best described by non-random and occasionally distal connections. Eigenvector centrality, which quantifies the importance of a neuron in the temporal flow of circuit activity, was highly related to feedforwardness in all functional circuits. The number of nodes participating in a functional circuit did not scale with the number of neurons imaged regardless of sensory area, indicating that circuit size is not tied to the sampling of neocortex. Local circuit flow comprehensively covered angular space regardless of the spatial scale that we tested, demonstrating that circuitry itself does not bias activity flow toward pia. Finally, analysis revealed that a minimal numerical sample size of neurons was necessary to capture at least 90 percent of functional circuit topology. These data and analyses indicated that functional circuitry exhibited rules of organization which generalized across three areas of sensory neocortex.

PLoS Computational Biology

M. Runfeldt, A. Sadovsky, J. MacLean

Sensory information is processed and transmitted through the synaptic structure of local cortical circuits, but it is unclear how modulation of this architecture influences the cortical representation of sensory stimuli. Acetylcholine (ACh) promotes attention and arousal and is thought to increase the signal-to-noise ratio of sensory input in primary sensory cortices. Using high-speed two-photon calcium imaging in a thalamocortical somatosensory slice preparation, we recorded action potential activity of up to 900 neurons simultaneously and compared local cortical circuit activations with and without bath presence of ACh. We found that ACh reduced weak pairwise relationships and excluded neurons that were already unreliable during circuit activity. Using action potential activity from the imaged population, we generated functional wiring diagrams based on the statistical dependencies of activity between neurons. ACh pruned weak functional connections from spontaneous circuit activations and yielded a more modular and hierarchical circuit structure, which biased activity to flow in a more feedforward fashion. Neurons that were active in response to thalamic input had reduced pairwise dependencies overall, but strong correlations were conserved. This coincided with a prolonged period during which neurons showed temporally precise responses to thalamic input. Our results demonstrate that ACh reorganizes functional circuit structure in a manner that may enhance the integration and discriminability of thalamic afferent input within local neocortical circuitry.

Journal of Neurophysiology

A. Sadovsky, J. MacLean

Spiking correlations between neocortical neurons provide insight into the underlying synaptic connectivity that defines cortical microcircuitry. Here, using two-photon calcium fluorescence imaging, we observed the simultaneous dynamics of hundreds of neurons in slices of mouse primary visual cortex (V1). Consistent with a balance of excitation and inhibition, V1 dynamics were characterized by a linear scaling between firing rate and circuit size. Using lagged firing correlations between neurons, we generated functional wiring diagrams to evaluate the topological features of V1 microcircuitry. We found that circuit connectivity exhibited both cyclic graph motifs, indicating recurrent wiring, and acyclic graph motifs, indicating feedforward wiring. After overlaying the functional wiring diagrams onto the imaged field of view, we found properties consistent with Rentian scaling: wiring diagrams were topologically efficient because they minimized wiring with a modular architecture. Within single imaged fields of view, V1 contained multiple discrete circuits that were overlapping and highly interdigitated but were still distinct from one another. The majority of neurons that were shared between circuits displayed peri-event spiking activity whose timing was specific to the active circuit, whereas spike times for a smaller percentage of neurons were invariant to circuit identity. These data provide evidence that V1 microcircuitry exhibits balanced dynamics, is efficiently arranged in anatomical space, and is capable of supporting a diversity of multineuron spike firing patterns from overlapping sets of neurons.

Journal of Neuroscience

F. Neubauer, A. Sederberg, J. MacLean

During the generalization of epileptic seizures, pathological activity in one brain area recruits distant brain structures into joint synchronous discharges. However, it remains unknown whether specific changes in local circuit activity are related to the aberrant recruitment of anatomically distant structures into epileptiform discharges. Further, it is not known whether aberrant areas recruit or entrain healthy ones into pathological activity. Here we study the dynamics of local circuit activity during the spread of epileptiform discharges in the zero-magnesium in vitro model of epilepsy. We employ high-speed multi-photon imaging in combination with dual whole-cell recordings in acute thalamocortical (TC) slices of the juvenile mouse to characterize the generalization of epileptic activity between neocortex and thalamus. We find that, although both structures are exposed to zero-magnesium, the initial onset of focal epileptiform discharge occurs in cortex. This suggests that local recurrent connectivity that is particularly prevalent in cortex is important for the initiation of seizure activity. Subsequent recruitment of thalamus into joint, generalized discharges is coincident with an increase in the coherence of local cortical circuit activity that itself does not depend on thalamus. Finally, the intensity of population discharges is positively correlated between both brain areas. This suggests that during and after seizure generalization not only the timing but also the amplitude of epileptiform discharges in thalamus is entrained by cortex. Together these results suggest a central role of neocortical activity for the onset and the structure of pathological recruitment of thalamus into joint synchronous epileptiform discharges.

Frontiers in Neural Circuits

P. Kruskal, L. Li, J. MacLean

Circuit reactivations involve a stereotyped sequence of neuronal firing and have been behaviorally linked to memory consolidation. Here we use multiphoton imaging and patch-clamp recording, and observe sparse and stereotyped circuit reactivations that correspond to UP states within active neurons. To evaluate the effect of the circuit on synaptic plasticity, we trigger a single spike-timing-dependent plasticity (STDP) pairing once per circuit reactivation. The pairings reliably fall within a particular epoch of the circuit sequence and result in long-term potentiation. During reactivation, the amplitude of plasticity significantly correlates with the preceding 20–25 ms of membrane depolarization rather than the depolarization at the time of pairing. This circuit-dependent plasticity provides a natural constraint on synaptic potentiation, regulating the inherent instability of STDP in an assembly phase-sequence model. Subthreshold voltage during endogenous circuit reactivations provides a critical informative context for plasticity and facilitates the stable consolidation of a spatiotemporal sequence.

Nature Communications

C. Piochon, P. Kruskal, J. MacLean, C. Hansel

Spike-timing-dependent plasticity (STDP) provides a cellular implementation of the Hebb postulate, which states that synapses, whose activity repeatedly drives action potential firing in target cells, are potentiated. At glutamatergic synapses onto hippocampal and neocortical pyramidal cells, synaptic activation followed by spike firing in the target cell causes long-term potentiation (LTP)—as predicted by Hebb—whereas excitatory postsynaptic potentials (EPSPs) evoked after a spike elicit long-term depression (LTD)—a phenomenon that was not specifically addressed by Hebb. In both instances the action potential in the postsynaptic target neuron is an instructive signal that is capable of supporting synaptic plasticity. STDP generally relies on the propagation of Na+ action potentials that are initiated in the axon hillhock back into the dendrite, where they cause depolarization and boost local calcium influx. However, recent studies in CA1 hippocampal pyramidal neurons have suggested that local calcium spikes might provide a more efficient trigger for LTP induction than backpropagating action potentials. Dendritic calcium spikes also play a role in an entirely different type of STDP that can be observed in cerebellar Purkinje cells. These neurons lack backpropagating Na+ spikes. Instead, plasticity at parallel fiber (PF) to Purkinje cell synapses depends on the relative timing of PF-EPSPs and activation of the glutamatergic climbing fiber (CF) input that causes dendritic calcium spikes. Thus, the instructive signal in this system is externalized. Importantly when EPSPs are elicited before CF activity, PF-LTD is induced rather than LTP. Thus, STDP in the cerebellum follows a timing rule that is opposite to its hippocampal/neocortical counterparts. Regardless, a common motif in plasticity is that LTD/LTP induction depends on the relative timing of synaptic activity and regenerative dendritic spikes which are driven by the instructive signal.

Frontiers in Neural Circuits

A. Sadovsky, J. MacLean

Mapping the flow of activity through neocortical microcircuits provides key insights into the underlying circuit architecture. Using a comparative analysis we determined the extent to which the dynamics of microcircuits in mouse primary somatosensory barrel field (S1BF) and auditory (A1) neocortex generalize. We imaged the simultaneous dynamics of up to 1126 neurons spanning multiple columns and layers using high-speed multiphoton imaging. The temporal progression and reliability of reactivation of circuit events in both regions suggested common underlying cortical design features. We used circuit activity flow to generate functional connectivity maps, or graphs, to test the microcircuit hypothesis within a functional framework. S1BF and A1 present a useful test of the postulate as both regions map sensory input anatomically, but each area appears organized according to different design principles. We projected the functional topologies into anatomical space and found benchmarks of organization that had been previously described using physiology and anatomical methods, consistent with a close mapping between anatomy and functional dynamics. By comparing graphs representing activity flow we found that each region is similarly organized as highlighted by hallmarks of small world, scale free, and hierarchical modular topologies. Models of prototypical functional circuits from each area of cortex were sufficient to recapitulate experimentally observed circuit activity. Convergence to common behavior by these models was accomplished using preferential attachment to scale from an auditory up to a somatosensory circuit. These functional data imply that the microcircuit hypothesis be framed as scalable principles of neocortical circuit design.

Journal of Neuroscience

A. Luczak, J. MacLean

Even in the absence of sensory stimuli, cortical networks exhibit complex, self-organized activity patterns. While the function of those spontaneous patterns of activation remains poorly understood, recent studies both in vivo and in vitro have demonstrated that neocortical neurons activate in a surprisingly similar sequential order both spontaneously and following input into cortex. For example, neurons that tend to fire earlier within spontaneous bursts of activity also fire earlier than other neurons in response to sensory stimuli. These “default patterns” can last hundreds of milliseconds and are strongly conserved under a variety of conditions. In this paper, we will review recent evidence for these default patterns at the local cortical level. We speculate that cortical architecture imposes common constraints on spontaneous and evoked activity flow, which result in the similarity of the patterns.

Frontiers in Integrative Neuroscience

A. Sadovsky, P. Kruskal, J. Kimmel, J. Ostmeyer, F. Neubauer, J. MacLean

Population dynamics of patterned neuronal firing are fundamental to information processing in the brain. Multiphoton microscopy in combination with calcium indicator dyes allows circuit dynamics to be imaged with single-neuron resolution. However, the temporal resolution of fluorescent measures is constrained by the imaging frequency imposed by standard raster scanning techniques. As a result, traditional raster scans limit the ability to detect the relative timing of action potentials in the imaged neuronal population. To maximize the speed of fluorescence measures from large populations of neurons using a standard multiphoton laser scanning microscope (MPLSM) setup, we have developed heuristically optimal path scanning (HOPS). HOPS optimizes the laser travel path length, and thus the temporal resolution of neuronal fluorescent measures, using standard galvanometer scan mirrors. Minimizing the scan path alone is insufficient for prolonged high-speed imaging of neuronal populations. Path stability and the signal-to-noise ratio become increasingly important factors as scan rates increase. HOPS addresses this by characterizing the scan mirror galvanometers to achieve prolonged path stability. In addition, the neuronal dwell time is optimized to sharpen the detection of action potentials while maximizing scan rate. The combination of shortest path calculation and minimization of mirror positioning time allows us to optically monitor a population of neurons in a field of view at high rates with single-spike resolution, ∼125 Hz for 50 neurons and ∼8.5 Hz for 1,000 neurons. Our approach introduces an accessible method for rapid imaging of large neuronal populations using traditional MPLSMs, facilitating new insights into neuronal circuit dynamics.

Journal of Neurophysiology