Order-based representation in random networks of cortical neurons

Microfluidic Bi-Layer Platform to Study Functional Interaction between Co-Cultured Neural Networks with Unidirectional Synaptic Connectivity

The complex synaptic connectivity architecture of neuronal networks underlies cognition and brain function. However, studying the spiking activity propagation and processing in heterogeneous networks in vivo poses significant challenges. In this study, we present a novel two-layer PDMS chip that facilitates the culturing and examination of the functional interaction of two interconnected neural networks. We utilized cultures of hippocampal neurons grown in a two-chamber microfluidic chip combined with a microelectrode array. The asymmetric configuration of the microchannels between the chambers ensured the growth of axons predominantly in one direction from the Source chamber to the Target chamber, forming two neuronal networks with unidirectional synaptic connectivity. We showed that the local application of tetrodotoxin (TTX) to the Source network did not alter the spiking rate in the Target network. The results indicate that stable network activity in the Target network was maintained for at least 1-3 h after TTX application, demonstrating the feasibility of local chemical activity modulation and the influence of electrical activity from one network on the other. Additionally, suppression of synaptic activity in the Source network by the application of CPP and CNQX reorganized spatio-temporal characteristics of spontaneous and stimulus-evoked spiking activity in the Target network. The proposed methodology and results provide a more in-depth examination of the network-level functional interaction between neural circuits with heterogeneous synaptic connectivity.

Neural synchrony in cortical networks: mechanisms and implications for neural information processing and coding

Synchronization of neuronal discharges on the millisecond scale has long been recognized as a prevalent and functionally important attribute of neural activity. In this article, I review classical concepts and corresponding evidence of the mechanisms that govern the synchronization of distributed discharges in cortical networks and relate those mechanisms to their possible roles in coding and cognitive functions. To accommodate the need for a selective, directed synchronization of cells, I propose that synchronous firing of distributed neurons is a natural consequence of spike-timing-dependent plasticity (STDP) that associates cells repetitively receiving temporally coherent input: the “synchrony through synaptic plasticity” hypothesis. Neurons that are excited by a repeated sequence of synaptic inputs may learn to selectively respond to the onset of this sequence through synaptic plasticity. Multiple neurons receiving coherent input could thus actively synchronize their firing by learning to selectively respond at corresponding temporal positions. The hypothesis makes several predictions: first, the position of the cells in the network, as well as the source of their input signals, would be irrelevant as long as their input signals arrive simultaneously; second, repeating discharge patterns should get compressed until all or some part of the signals are synchronized; and third, this compression should be accompanied by a sparsening of signals. In this way, selective groups of cells could emerge that would respond to some recurring event with synchronous firing. Such a learned response pattern could further be modulated by synchronous network oscillations that provide a dynamic, flexible context for the synaptic integration of distributed signals. I conclude by suggesting experimental approaches to further test this new hypothesis.

Stimulus-Evoked Activity Modulation of In Vitro Engineered Cortical and Hippocampal Networks

The delivery of electrical stimuli is crucial to shape the electrophysiological activity of neuronal populations and to appreciate the response of the different brain circuits involved. In the present work, we used dissociated cortical and hippocampal networks coupled to Micro-Electrode Arrays (MEAs) to investigate the features of their evoked response when a low-frequency (0.2 Hz) electrical stimulation protocol is delivered. In particular, cortical and hippocampal neurons were topologically organized to recreate interconnected sub-populations with a polydimethylsiloxane (PDMS) mask, which guaranteed the segregation of the cell bodies and the connections among the sub-regions through microchannels. We found that cortical assemblies were more reactive than hippocampal ones. Despite both configurations exhibiting a fast (<35 ms) response, this did not uniformly distribute over the MEA in the hippocampal networks. Moreover, the propagation of the stimuli-evoked activity within the networks showed a late (35−500 ms) response only in the cortical assemblies. The achieved results suggest the importance of the neuronal target when electrical stimulation experiments are performed. Not all neuronal types display the same response, and in light of transferring stimulation protocols to in vivo applications, it becomes fundamental to design realistic in vitro brain-on-a-chip devices to investigate the dynamical properties of complex neuronal circuits.

A Theoretical Comprehensive Framework for the Process of Theories Formation

Scientists rely more and more upon computerized data mining and artificial intelligence to analyze data sets and identify association rules, which serve as the basis of evolving theories. This tendency is likely to expand, and computerized intelligence is likely to take a leading role in scientific theorizing. While the ever-advancing technology could be of great benefit, scientists with expertise in many research fields do not necessarily understand thoroughly enough the various assumptions, which underlie different data mining methods and which pose significant limitations on the association rules that could be identified in the first place. There seems to be a need for a comprehensive framework, which should present the various possible technological aids in the context of our neurocognitive process of theorizing and identifying association rules. Such a framework can be hopefully used to understand, identify, and overcome the limitations of the currently fragmented processes of technology-based theorizing and the formation of association rules in any research field. In order to meet this end, we divide theorizing into underlying neurocognitive components, describe their current technological expansions and limitations, and offer a possible comprehensive computational framework for each such component and their combination.

Consolidation of memory traces in cultured cortical networks requires low cholinergic tone, synchronized activity and high network excitability

In systems consolidation, encoded memories are replayed by the hippocampus during slow-wave sleep (SWS), and permanently stored in the neocortex. Declarative memory consolidation is believed to benefit from the oscillatory rhythms and low cholinergic tone observed in this sleep stage, but underlying mechanisms remain unclear. To clarify the role of cholinergic modulation and synchronized activity in memory consolidation, we applied repeated electrical stimulation in mature cultures of dissociated rat cortical neurons with high or low cholinergic tone, mimicking the cue replay observed during systems consolidation under distinct cholinergic concentrations. In the absence of cholinergic input, these cultures display activity patterns hallmarked by network bursts, synchronized events reminiscent of the low frequency oscillations observed during SWS. They display stable activity and connectivity, which mutually interact and achieve an equilibrium. Electrical stimulation reforms the equilibrium to include the stimulus response, a phenomenon interpreted as memory trace formation. Without cholinergic input, activity was burst-dominated. First application of a stimulus induced significant connectivity changes, while subsequent repetition no longer affected connectivity. Presenting a second stimulus at a different electrode had the same effect, whereas returning to the initial stimuli did not induce further connectivity alterations, indicating that the second stimulus did not erase the 'memory trace' of the first. Distinctively, cultures with high cholinergic tone displayed reduced network excitability and dispersed firing, and electrical stimulation did not induce significant connectivity changes. We conclude that low cholinergic tone facilitates memory formation and consolidation, possibly through enhanced network excitability. Network bursts or SWS oscillations may merely reflect high network excitability.

Thermoplasmonic Optical Fiber for Localized Neural Stimulation

Thermoplasmonic effect-based neural stimulation has been suggested as an alternative optical neural stimulation technology without genetic modification. Integration of near-infrared light with plasmonic gold nanoparticles has been demonstrated as a neuromodulation tool on in vitro neuronal network models. In order to further test the validity of the thermoplasmonic neural stimulation across multiple biological models (in vitro, ex vivo, and in vivo) avoiding genetic modification in optical neuromodulation, versatile engineering approaches to apply the thermoplasmonic effect would be required. In this work, we developed a gold nanorod attached optical fiber technology for the localized neural stimulation based on a thermoplasmonic effect. A simple fabrication process was developed for efficient nanoparticle coating on commercial optical fibers. The thermoplasmonic optical fiber proved that it can locally modulate the neural activity in vitro. Lastly, we simulated the spatiotemporal temperature change by the thermoplasmonic optical fiber and analyzed its applicability to in vivo animal models.

Spatial Properties of STDP in a Self-Learning Spiking Neural Network Enable Controlling a Mobile Robot

Development of spiking neural networks (SNNs) controlling mobile robots is one of the modern challenges in computational neuroscience and artificial intelligence. Such networks, being replicas of biological ones, are expected to have a higher computational potential than traditional artificial neural networks (ANNs). The critical problem is in the design of robust learning algorithms aimed at building a “living computer” based on SNNs. Here, we propose a simple SNN equipped with a Hebbian rule in the form of spike-timing-dependent plasticity (STDP). The SNN implements associative learning by exploiting the spatial properties of STDP. We show that a LEGO robot controlled by the SNN can exhibit classical and operant conditioning. Competition of spike-conducting pathways in the SNN plays a fundamental role in establishing associations of neural connections. It replaces the irrelevant associations by new ones in response to a change in stimuli. Thus, the robot gets the ability to relearn when the environment changes. The proposed SNN and the stimulation protocol can be further enhanced and tested in developing neuronal cultures, and also admit the use of memristive devices for hardware implementation.

Unstructured network topology begets order-based representation by privileged neurons

How spiking activity reverberates through neuronal networks, how evoked and spontaneous activity interacts and blends, and how the combined activities represent external stimulation are pivotal questions in neuroscience. We simulated minimal models of unstructured spiking networks in silico, asking whether and how gentle external stimulation might be subsequently reflected in spontaneous activity fluctuations. Consistent with earlier findings in silico and in vitro, we observe a privileged subpopulation of 'pioneer neurons' that, by their firing order, reliably encode previous external stimulation. We also confirm that pioneer neurons are 'sensitive' in that they are recruited by small fluctuations of population activity. We show that order-based representations rely on a 'chain' of pioneer neurons with different degrees of sensitivity and thus constitute an emergent property of collective dynamics. The forming of such representations is greatly favoured by a broadly heterogeneous connection topology-a broad 'middle class' in degree of connectedness. In conclusion, we offer a minimal model for the representational role of pioneer neurons, as observed experimentally in vitro. In addition, we show that broadly heterogeneous connectivity enhances the representational capacity of unstructured networks.

Functional Scaffolding for Brain Implants: Engineered Neuronal Network by Microfabrication and iPSC Technology

Neuroengineering methods can be effectively used in the design of new approaches to treat central nervous system and brain injury caused by neurotrauma, ischemia, or neurodegenerative disorders. During the last decade, significant results were achieved in the field of implant (scaffold) development using various biocompatible and biodegradable materials carrying neuronal cells for implantation into the injury site of the brain to repair its function. Neurons derived from animal or human induced pluripotent stem (iPS) cells are expected to be an ideal cell source, and induction methods for specific cell types have been actively studied to improve efficacy and specificity. A critical goal of neuro-regeneration is structural and functional restoration of the injury site. The target treatment area has heterogeneous and complex network topology with various types of cells that need to be restored with similar neuronal network structure to recover correct functionality. However, current scaffold-based technology for brain implants operates with homogeneous neuronal cell distribution, which limits recovery in the damaged area of the brain and prevents a return to fully functional biological tissue. In this study, we present a neuroengineering concept for designing a neural circuit with a pre-defined unidirectional network architecture that provides a balance of excitation/inhibition in the scaffold to form tissue similar to that in the injured area using various types of iPS cells. Such tissue will mimic the surrounding niche in the injured site and will morphologically and topologically integrate into the brain, recovering lost function.

Inhomogeneities in Network Structure and Excitability Govern Initiation and Propagation of Spontaneous Burst Activity

The mesoscale architecture of neuronal networks strongly influences the initiation of spontaneous activity and its pathways of propagation. Spontaneous activity has been studied extensively in networks of cultured cortical neurons that generate complex yet reproducible patterns of synchronous bursting events that resemble the activity dynamics in developing neuronal networks in vivo. Synchronous bursts are mostly thought to be triggered at burst initiation sites due to build-up of noise or by highly active neurons, or to reflect reverberating activity that circulates within larger networks, although neither of these has been observed directly. Inferring such collective dynamics in neuronal populations from electrophysiological recordings crucially depends on the spatial resolution and sampling ratio relative to the size of the networks assessed. Using large-scale microelectrode arrays with 1024 electrodes at 0.3 mm pitch that covered the full extent of in vitro networks on about 1 cm², we investigated where bursts of spontaneous activity arise and how their propagation patterns relate to the regions of origin, the network's structure, and to the overall distribution of activity. A set of alternating burst initiation zones (BIZ) dominated the initiation of distinct bursting events and triggered specific propagation patterns. Moreover, BIZs were typically located in areas with moderate activity levels, i.e., at transitions between hot and cold spots. The activity-dependent alternation between these zones suggests that the local networks forming the dominating BIZ enter a transient depressed state after several cycles (similar to Eytan et al., 2003), allowing other BIZs to take over temporarily. We propose that inhomogeneities in the network structure define such BIZs and that the depletion of local synaptic resources limit repetitive burst initiation.

A Biohybrid Setup for Coupling Biological and Neuromorphic Neural Networks

Developing technologies for coupling neural activity and artificial neural components, is key for advancing neural interfaces and neuroprosthetics. We present a biohybrid experimental setting, where the activity of a biological neural network is coupled to a biomimetic hardware network. The implementation of the hardware network (denoted NeuroSoC) exhibits complex dynamics with a multiplicity of time-scales, emulating 2880 neurons and 12.7 M synapses, designed on a VLSI chip. This network is coupled to a neural network in vitro, where the activities of both the biological and the hardware networks can be recorded, processed, and integrated bidirectionally in real-time. This experimental setup enables an adjustable and well-monitored coupling, while providing access to key functional features of neural networks. We demonstrate the feasibility to functionally couple the two networks and to implement control circuits to modify the biohybrid activity. Overall, we provide an experimental model for neuromorphic-neural interfaces, hopefully to advance the capability to interface with neural activity, and with its irregularities in pathology.

Analog control with two Artificial Axons

The artificial axon is a recently introduced synthetic assembly of supported lipid bilayers and voltage gated ion channels, displaying the basic electrophysiology of nerve cells. Here we demonstrate the use of two artificial axons as control elements to achieve a simple task. Namely, we steer a remote control car towards a light source, using the sensory input dependent firing rate of the axons as the control signal for turning left or right. We present the result in the form of the analysis of a movie of the car approaching the light source. In general terms, with this work we pursue a constructivist approach to exploring the nexus between machine language at the nerve cell level and behavior.

Theta rhythm-like bidirectional cycling dynamics of living neuronal networks in vitro

The phenomena of synchronization, rhythmogenesis and coherence observed in brain networks are believed to be a dynamic substrate for cognitive functions such as learning and memory. However, researchers are still debating whether the rhythmic activity emerges from the network morphology that developed during neurogenesis or as a result of neuronal dynamics achieved under certain conditions. In the present study, we observed self-organized spiking activity that converged to long, complex and rhythmically repeated superbursts in neural networks formed by mature hippocampal cultures with a high cellular density. The superburst lasted for tens of seconds and consisted of hundreds of short (50-100 ms) small bursts with a high spiking rate of 139.0 ± 78.6 Hz that is associated with high-frequency oscillations in the hippocampus. In turn, the bursting frequency represents a theta rhythm (11.2 ± 1.5 Hz). The distribution of spikes within the bursts was non-random, representing a set of well-defined spatio-temporal base patterns or motifs. The long superburst was classified into two types. Each type was associated with a unique direction of spike propagation and, hence, was encoded by a binary sequence with random switching between the two "functional" states. The precisely structured bidirectional rhythmic activity that developed in self-organizing cultured networks was quite similar to the activity observed in the in vivo experiments.

Stimulation triggers endogenous activity patterns in cultured cortical networks

Cultures of dissociated cortical neurons represent a powerful trade-off between more realistic experimental models and abstract modeling approaches, allowing to investigate mechanisms of synchronized activity generation. These networks spontaneously alternate periods of high activity (i.e. network bursts) with periods of quiescence in a dynamic state which recalls the fluctuation of in vivo UP and DOWN states. Network bursts can also be elicited by external stimulation and their spatial propagation patterns tracked by means of multi-channel micro-electrode arrays. In this study, we used rat cortical cultures coupled to micro-electrode arrays to investigate the similarity between spontaneous and evoked activity patterns. We performed experiments by applying electrical stimulation to different network locations and demonstrated that the rank orders of electrodes during evoked and spontaneous events are remarkably similar independently from the stimulation source. We linked this result to the capability of stimulation to evoke firing in highly active and “leader” sites of the network, reliably and rapidly recruited within both spontaneous and evoked bursts. Our study provides the first evidence that spontaneous and evoked activity similarity is reliably observed also in dissociated cortical networks.

Mesoscale Architecture Shapes Initiation and Richness of Spontaneous Network Activity

Spontaneous activity in the absence of external input, including propagating waves of activity, is a robust feature of neuronal networks in vivo and in vitro The neurophysiological and anatomical requirements for initiation and persistence of such activity, however, are poorly understood, as is their role in the function of neuronal networks. Computational network studies indicate that clustered connectivity may foster the generation, maintenance, and richness of spontaneous activity. Since this mesoscale architecture cannot be systematically modified in intact tissue, testing these predictions is impracticable in vivo Here, we investigate how the mesoscale structure shapes spontaneous activity in generic networks of rat cortical neurons in vitro In these networks, neurons spontaneously arrange into local clusters with high neurite density and form fasciculating long-range axons. We modified this structure by modulation of protein kinase C, an enzyme regulating neurite growth and cell migration. Inhibition of protein kinase C reduced neuronal aggregation and fasciculation of axons, i.e., promoted uniform architecture. Conversely, activation of protein kinase C promoted aggregation of neurons into clusters, local connectivity, and bundling of long-range axons. Supporting predictions from theory, clustered networks were more spontaneously active and generated diverse activity patterns. Neurons within clusters received stronger synaptic inputs and displayed increased membrane potential fluctuations. Intensified clustering promoted the initiation of synchronous bursting events but entailed incomplete network recruitment. Moderately clustered networks appear optimal for initiation and propagation of diverse patterns of activity. Our findings support a crucial role of the mesoscale architectures in the regulation of spontaneous activity dynamics.SIGNIFICANCE STATEMENT Computational studies predict richer and persisting spatiotemporal patterns of spontaneous activity in neuronal networks with neuron clustering. To test this, we created networks of varying architecture in vitro Supporting these predictions, the generation and spatiotemporal patterns of propagation were most variable in networks with intermediate clustering and lowest in uniform networks. Grid-like clustering, on the other hand, facilitated spontaneous activity but led to degenerating patterns of propagation. Neurons outside clusters had weaker synaptic input than neurons within clusters, in which increased membrane potential fluctuations facilitated the initiation of synchronized spike activity. Our results thus show that the intermediate level organization of neuronal networks strongly influences the dynamics of their activity.

SiMEA: a framework for simulating neurons on multi-electrode array

A Multi-Electrode Array (MEA) is a practical device for recording the extracellular activity of in-vitro biological culture. Such culture - for instance neurons - is prone to mistakes leading to irrelevant recordings or no recording at all. Additionally, with the expenses generated by in-vitro culture, minimizing risks is a must. This paper proposes a framework designed and implemented for simulating the spatial positioning of neuronal cultures on a MEA. The framework serves as a sandbox for researchers to simulate the model of their MEA experiments before its eventual in-vitro implementation. The framework enables simulating the density of the plated culture, the death of cells over time, choosing diverse reconstructed morphologies of cells, and simulating their spiking activity in interaction with Brian2 simulator.

Selectivity of stimulus induced responses in cultured hippocampal networks on microelectrode arrays

Sensory information can be encoded using the average firing rate and spike occurrence times in neuronal network responses to external stimuli. Decoding or retrieving stimulus characteristics from the response pattern generally implies that the corresponding neural network has a selective response to various input signals. The role of various spiking activity characteristics (e.g., spike rate and precise spike timing) for basic information processing was widely investigated on the level of neural populations but gave inconsistent evidence for particular mechanisms. Multisite electrophysiology of cultured neural networks grown on microelectrode arrays is a recently developed tool and currently an active research area. In this study, we analyzed the stimulus responses represented by network-wide bursts evoked from various spatial locations (electrodes). We found that the response characteristics, such as the burst initiation time and the spike rate, can be used to retrieve information about the stimulus location. The best selectivity in the response spiking pattern could be found for a small subpopulation of neurones (electrodes) at relatively short post-stimulus intervals. Such intervals were unique for each culture due to the non-uniform organization of the functional connectivity in the network during spontaneous development.

A Possible Common Neurophysiologic Basis for MDD, Bipolar Disorder, and Schizophrenia: Lessons from Electrophysiology

There is ample electrophysiological evidence of attention dysfunction in the EEG/ERP signal of major depressive disorder (MDD), bipolar disorder, and schizophrenia. The reduced attention-related ERP waves show much similarity between MDD, bipolar disorder, and schizophrenia, raising the question whether there are similarities in the neurophysiologic process that underlies attention dysfunction in these pathologies. The present work suggests that there is such a unified underlying neurophysiologic process, which results in reduced attention in the three pathologies. Naturally, as these pathologies involve different clinical manifestations, we expect differences in their underlying neurophysiology. These differences and their subtle manifestation in the ERP marker for attention are also discussed. MDD, bipolar disorder, and schizophrenia are just three of multiple neuropsychiatric disorders, which involve changes in the EEG/ERP manifestations of attention. Further work should expand the basic model presented here to offer comprehensive modeling of these multiple disorders and to emphasize similarities and dissimilarities of the underlying neurophysiologic processes.

Repeated stimulation of cultured networks of rat cortical neurons induces parallel memory traces

During systems consolidation, memories are spontaneously replayed favoring information transfer from hippocampus to neocortex. However, at present no empirically supported mechanism to accomplish a transfer of memory from hippocampal to extra-hippocampal sites has been offered. We used cultured neuronal networks on multielectrode arrays and small-scale computational models to study the effect of memory replay on the formation of memory traces. We show that input-deprived networks develop an activity⇔connectivity balance where dominant activity patterns support current connectivity. Electrical stimulation at one electrode disturbs this balance and induces connectivity changes. Intrinsic forces in recurrent networks lead to a new equilibrium with activity patterns that include the stimulus response. The new connectivity is no longer disrupted by this stimulus, indicating that networks memorize it. A different stimulus again induces connectivity changes upon first application but not subsequently, demonstrating the formation of a second memory trace. Returning to the first stimulus does not affect connectivity, indicating parallel storage of both traces. A computer model robustly reproduced experimental results, suggesting that spike-timing-dependent plasticity and short time depression suffice to store parallel memory traces, even in networks without particular circuitry constraints.

Universality, complexity and the praxis of biology: Two case studies

The phenomenon of biology provides a prime example for a naturally occurring complex system. The approach to this complexity reflects the tension between a reductionist, reverse-engineering stance, and more abstract, systemic ones. Both of us are reductionists, but our observations challenge reductionism, at least the naive version of it. Here we describe the challenge, focusing on two universal characteristics of biological complexity: two-way microscopic-macroscopic degeneracy, and lack of time scale separation within and between levels of organization. These two features and their consequences for the praxis of experimental biology, reflect inherent difficulties in separating the dynamics of any given level of organization from the coupled dynamics of all other levels, including the environment within which the system is embedded. Where these difficulties are not deeply acknowledged, the impacts of fallacies that are inherent to naive reductionism are significant. In an era where technology enables experimental high-resolution access to numerous observables, the challenge faced by the mature reductionist-identification of relevant microscopic variables-becomes more demanding than ever. The demonstrations provided here are taken from two very different biological realizations: populations of microorganisms and populations of neurons, thus making the lesson potentially general.

Can functionality in evolving networks be explained reductively?

Philosophers of biology disagree about an adequate explication of the concept of function. Instead of perpetuating the debate on the level of in principle-arguments, this paper aims first at reconstructing functional talk in the biological research papers of Marom and Braun, which focus on two different kinds of evolving networks, and in discussing the ontological consequences which the authors draw from their results. Marom investigates evolving neural networks controlling Braitenberg vehicles. Braun observes the evolutionary rearrangement or "rewiring" of the genetic network of genetically modified yeast on a short time scale. In both cases, the parameters under investigation are defined in functional terms. However, both authors report striking differences in the structures that realize one and the same function, as well as striking differences in the function of identical structures. From this, they construct an argument against reductionism. The second aim of my paper is an inquiry into the epistemic legitimacy of this conclusion. This requires addressing critically several concepts on which Marom and Braun's argument is built.

Interaction of electrically evoked activity with intrinsic dynamics of cultured cortical networks with and without functional fast GABAergic synaptic transmission

The modulation of neuronal activity by means of electrical stimulation is a successful therapeutic approach for patients suffering from a variety of central nervous system disorders. Prototypic networks formed by cultured cortical neurons represent an important model system to gain general insights in the input–output relationships of neuronal tissue. These networks undergo a multitude of developmental changes during their maturation, such as the excitatory–inhibitory shift of the neurotransmitter GABA. Very few studies have addressed how the output properties to a given stimulus change with ongoing development. Here, we investigate input–output relationships of cultured cortical networks by probing cultures with and without functional GABA_Aergic synaptic transmission with a set of stimulation paradigms at various stages of maturation. On the cellular level, low stimulation rates (<15 Hz) led to reliable neuronal responses; higher rates were increasingly ineffective. Similarly, on the network level, lowest stimulation rates (<0.1 Hz) lead to maximal output rates at all ages, indicating a network wide refractory period after each stimulus. In cultures aged 3 weeks and older, a gradual recovery of the network excitability within tens of milliseconds was in contrast to an abrupt recovery after about 5 s in cultures with absent GABA_Aergic synaptic transmission. In these GABA deficient cultures evoked responses were prolonged and had multiple discharges. Furthermore, the network excitability changed periodically, with a very slow spontaneous change of the overall network activity in the minute range, which was not observed in cultures with absent GABA_Aergic synaptic transmission. The electrically evoked activity of cultured cortical networks, therefore, is governed by at least two potentially interacting mechanisms: A refractory period in the order of a few seconds and a very slow GABA dependent oscillation of the network excitability.

Slow dynamics in features of synchronized neural network responses

In this report trial-to-trial variations in the synchronized responses of neural networks are explored over time scales of minutes, in ex-vivo large scale cortical networks. We show that sub-second measures of the individual synchronous response, namely-its latency and decay duration, are related to minutes-scale network response dynamics. Network responsiveness is reflected as residency in, or shifting amongst, areas of the latency-decay plane. The different sensitivities of latency and decay durations to synaptic blockers imply that these two measures reflect aspects of inhibitory and excitatory activities. Taken together, the data suggest that trial-to-trial variations in the synchronized responses of neural networks might be related to effective excitation-inhibition ratio being a dynamic variable over time scales of minutes.

Stimulus information stored in lasting active and hidden network states is destroyed by network bursts

In both humans and animals brief synchronizing bursts of epileptiform activity known as interictal epileptiform discharges (IEDs) can, even in the absence of overt seizures, cause transient cognitive impairments (TCI) that include problems with perception or short-term memory. While no evidence from single units is available, it has been assumed that IEDs destroy information represented in neuronal networks. Cultured neuronal networks are a model for generic cortical microcircuits, and their spontaneous activity is characterized by the presence of synchronized network bursts (SNBs), which share a number of properties with IEDs, including the high degree of synchronization and their spontaneous occurrence in the absence of an external stimulus. As a model approach to understanding the processes underlying IEDs, optogenetic stimulation and multielectrode array (MEA) recordings of cultured neuronal networks were used to study whether stimulus information represented in these networks survives SNBs. When such networks are optically stimulated they encode and maintain stimulus information for as long as one second. Experiments involved recording the network response to a single stimulus and trials where two different stimuli were presented sequentially, akin to a paired pulse trial. We broke the sequential stimulus trials into encoding, delay and readout phases and found that regardless of which phase the SNB occurs, stimulus-specific information was impaired. SNBs were observed to increase the mean network firing rate, but this did not translate monotonically into increases in network entropy. It was found that the more excitable a network, the more stereotyped its response was during a network burst. These measurements speak to whether SNBs are capable of transmitting information in addition to blocking it. These results are consistent with previous reports and provide baseline predictions concerning the neural mechanisms by which IEDs might cause TCI.

Controlling neural network responsiveness: tradeoffs and constraints

In recent years much effort is invested in means to control neural population responses at the whole brain level, within the context of developing advanced medical applications. The tradeoffs and constraints involved, however, remain elusive due to obvious complications entailed by studying whole brain dynamics. Here, we present effective control of response features (probability and latency) of cortical networks in vitro over many hours, and offer this approach as an experimental toy for studying controllability of neural networks in the wider context. Exercising this approach we show that enforcement of stable high activity rates by means of closed loop control may enhance alteration of underlying global input-output relations and activity dependent dispersion of neuronal pair-wise correlations across the network.

Adaptive enhancement of learning protocol in hippocampal cultured networks grown on multielectrode arrays

Learning in neuronal networks can be investigated using dissociated cultures on multielectrode arrays supplied with appropriate closed-loop stimulation. It was shown in previous studies that weakly respondent neurons on the electrodes can be trained to increase their evoked spiking rate within a predefined time window after the stimulus. Such neurons can be associated with weak synaptic connections in nearby culture network. The stimulation leads to the increase in the connectivity and in the response. However, it was not possible to perform the learning protocol for the neurons on electrodes with relatively strong synaptic inputs and responding at higher rates. We proposed an adaptive closed-loop stimulation protocol capable to achieve learning even for the highly respondent electrodes. It means that the culture network can reorganize appropriately its synaptic connectivity to generate a desired response. We introduced an adaptive reinforcement condition accounting for the response variability in control stimulation. It significantly enhanced the learning protocol to a large number of responding electrodes independently on its base response level. We also found that learning effect preserved after 4–6 h after training.

Short-term memory in networks of dissociated cortical neurons

Short-term memory refers to the ability to store small amounts of stimulus-specific information for a short period of time. It is supported by both fading and hidden memory processes. Fading memory relies on recurrent activity patterns in a neuronal network, whereas hidden memory is encoded using synaptic mechanisms, such as facilitation, which persist even when neurons fall silent. We have used a novel computational and optogenetic approach to investigate whether these same memory processes hypothesized to support pattern recognition and short-term memory in vivo, exist in vitro. Electrophysiological activity was recorded from primary cultures of dissociated rat cortical neurons plated on multielectrode arrays. Cultures were transfected with ChannelRhodopsin-2 and optically stimulated using random dot stimuli. The pattern of neuronal activity resulting from this stimulation was analyzed using classification algorithms that enabled the identification of stimulus-specific memories. Fading memories for different stimuli, encoded in ongoing neural activity, persisted and could be distinguished from each other for as long as 1 s after stimulation was terminated. Hidden memories were detected by altered responses of neurons to additional stimulation, and this effect persisted longer than 1 s. Interestingly, network bursts seem to eliminate hidden memories. These results are similar to those that have been reported from similar experiments in vivo and demonstrate that mechanisms of information processing and short-term memory can be studied using cultured neuronal networks, thereby setting the stage for therapeutic applications using this platform.

Parametric models to relate spike train and LFP dynamics with neural information processing

Spike trains and local field potentials (LFPs) resulting from extracellular current flows provide a substrate for neural information processing. Understanding the neural code from simultaneous spike-field recordings and subsequent decoding of information processing events will have widespread applications. One way to demonstrate an understanding of the neural code, with particular advantages for the development of applications, is to formulate a parametric statistical model of neural activity and its covariates. Here, we propose a set of parametric spike-field models (unified models) that can be used with existing decoding algorithms to reveal the timing of task or stimulus specific processing. Our proposed unified modeling framework captures the effects of two important features of information processing: time-varying stimulus-driven inputs and ongoing background activity that occurs even in the absence of environmental inputs. We have applied this framework for decoding neural latencies in simulated and experimentally recorded spike-field sessions obtained from the lateral intraparietal area (LIP) of awake, behaving monkeys performing cued look-and-reach movements to spatial targets. Using both simulated and experimental data, we find that estimates of trial-by-trial parameters are not significantly affected by the presence of ongoing background activity. However, including background activity in the unified model improves goodness of fit for predicting individual spiking events. Uncovering the relationship between the model parameters and the timing of movements offers new ways to test hypotheses about the relationship between neural activity and behavior. We obtained significant spike-field onset time correlations from single trials using a previously published data set where significantly strong correlation was only obtained through trial averaging. We also found that unified models extracted a stronger relationship between neural response latency and trial-by-trial behavioral performance than existing models of neural information processing. Our results highlight the utility of the unified modeling framework for characterizing spike-LFP recordings obtained during behavioral performance.

Fast coding of orientation in primary visual cortex

Understanding how populations of neurons encode sensory information is a major goal of systems neuroscience. Attempts to answer this question have focused on responses measured over several hundred milliseconds, a duration much longer than that frequently used by animals to make decisions about the environment. How reliably sensory information is encoded on briefer time scales, and how best to extract this information, is unknown. Although it has been proposed that neuronal response latency provides a major cue for fast decisions in the visual system, this hypothesis has not been tested systematically and in a quantitative manner. Here we use a simple ‘race to threshold’ readout mechanism to quantify the information content of spike time latency of primary visual (V1) cortical cells to stimulus orientation. We find that many V1 cells show pronounced tuning of their spike latency to stimulus orientation and that almost as much information can be extracted from spike latencies as from firing rates measured over much longer durations. To extract this information, stimulus onset must be estimated accurately. We show that the responses of cells with weak tuning of spike latency can provide a reliable onset detector. We find that spike latency information can be pooled from a large neuronal population, provided that the decision threshold is scaled linearly with the population size, yielding a processing time of the order of a few tens of milliseconds. Our results provide a novel mechanism for extracting information from neuronal populations over the very brief time scales in which behavioral judgments must sometimes be made.

How can humans and animals make complex decisions on time scales as short as 100 ms? The information required for such decisions is coded in neural activity and should be read out on a very brief time scale. Traditional approaches to coding of neural information rely on the number of electrical pulses, or spikes, that neurons fire in a certain time window. Although this type of code is likely to be used by the brain for higher cognitive tasks, it may be too slow for fast decisions. Here, we explore an alternative code which is based on the latency of spikes with respect to a reference signal. By analyzing the simultaneous responses of many cells in monkey visual cortex, we show that information about the orientation of visual stimuli can be extracted reliably from spike latencies on very short time scales.

Detecting multineuronal temporal patterns in parallel spike trains

We present a non-parametric and computationally efficient method that detects spatiotemporal firing patterns and pattern sequences in parallel spike trains and tests whether the observed numbers of repeating patterns and sequences on a given timescale are significantly different from those expected by chance. The method is generally applicable and uncovers coordinated activity with arbitrary precision by comparing it to appropriate surrogate data. The analysis of coherent patterns of spatially and temporally distributed spiking activity on various timescales enables the immediate tracking of diverse qualities of coordinated firing related to neuronal state changes and information processing. We apply the method to simulated data and multineuronal recordings from rat visual cortex and show that it reliably discriminates between data sets with random pattern occurrences and with additional exactly repeating spatiotemporal patterns and pattern sequences. Multineuronal cortical spiking activity appears to be precisely coordinated and exhibits a sequential organization beyond the cell assembly concept.

Emergence of a small-world functional network in cultured neurons

The functional networks of cultured neurons exhibit complex network properties similar to those found in vivo. Starting from random seeding, cultures undergo significant reorganization during the initial period in vitro, yet despite providing an ideal platform for observing developmental changes in neuronal connectivity, little is known about how a complex functional network evolves from isolated neurons. In the present study, evolution of functional connectivity was estimated from correlations of spontaneous activity. Network properties were quantified using complex measures from graph theory and used to compare cultures at different stages of development during the first 5 weeks in vitro. Networks obtained from young cultures (14 days in vitro) exhibited a random topology, which evolved to a small-world topology during maturation. The topology change was accompanied by an increased presence of highly connected areas (hubs) and network efficiency increased with age. The small-world topology balances integration of network areas with segregation of specialized processing units. The emergence of such network structure in cultured neurons, despite a lack of external input, points to complex intrinsic biological mechanisms. Moreover, the functional network of cultures at mature ages is efficient and highly suited to complex processing tasks.

Enhancement of neural representation capacity by modular architecture in networks of cortical neurons

Biological networks are ubiquitously modular, a feature that is believed to be essential for the enhancement of their functional capacities. Here, we have used a simple modular in vitro design to examine the possibility that modularity enhances network functionality in the context of input representation. We cultured networks of cortical neurons obtained from newborn rats in vitro on substrate-integrated multi-electrode arrays, forcing the network to develop two well-defined modules of neural populations that are coupled by a narrow canal. We measured the neural activity, and examined the capacity of each module to individually classify (i.e. represent) spatially distinct electrical stimuli and propagate input-specific activity features to their downstream coupled counterpart. We show that, although each of the coupled modules maintains its autonomous functionality, a significant enhancement of representational capacity is achieved when the system is observed as a whole. We interpret our results in terms of a relative decorrelation effect imposed by weak coupling between modules.

Explorative Data Analysis of In-Vitro Neuronal Network Behavior Based on an Unsupervised Learning Approach

In the present chapter authors want to expose new insights in the field of Computational Neuroscience at regard to the study of neuronal networks grown in vitro. Such kind of analyses can exploit the availability of a huge amount of data thanks to the use of Multi Electrode Arrays (MEA), a multi-channel technology which allows capturing the activity of several different neuronal cells for long time recordings. Given the possibility of simultaneous targeting of various sites, neuroscientists are so applying such recent technology for various researches. The chapter begins by giving a brief presentation of MEA technology and of the data produced in output, punctuating some of the pros and cons of MEA recordings. Then we present an overview of the analytical techniques applied in order to extrapolate the hidden information from available data. Then we shall explain the approach we developed and applied on MEAs prepared in our cell culture laboratory, consisting of statistical methods capturing the main features of the spiking, in particular bursting, activity of various neuron, and performing data dimensionality reduction and clustering, in order to classify neurons according to their spiking properties having showed correlated features. Finally the chapter wants to furnish to neuroscientists an overview about the quantitative analysis of in-vitro spiking activity data recorded via MEA technology and to give an example of explorative analysis applied on MEA data. Such study is based on methods from Statistics and Machine Learning or Computer Science but at the same time strictly related to neurophysiological interpretations of the putative pharmacological manipulation of synaptic connections and mode of firing, with the final aim to extract new information and knowledge about neuronal networks behavior and organization.

Spiking signatures of spontaneous activity bursts in hippocampal cultures

Dense dissociated hippocampal cultures are known to generate spontaneous bursting electrical activity which can be recorded by multielectrode arrays. We have analyzed spatio-temporal profiles of the distribution of spikes in the bursts recorded after 2 weeks in vitro. We have found a statistically significant similarity between the spiking patterns in sequential bursting events, we refer to these spiking patterns as spiking signatures. Such spiking signatures may appear in different parts of the bursts, including the activation patterns - the first spike times in the bursts, and deactivation patterns - the last spike times in the bursts. Moreover, these patterns may display apparent time scaling, e.g., they may be replayed in the subsequent bursts at different speeds, while preserving the spiking order. We discuss how such properties of the bursts may be associated with the formation of repeatable signaling pathways in cultured networks in vitro.

Representation and learning in neuronal networks: a conceptual nervous system approach

The work presented in this review describes the use of large cortical networks developing ex vivo, in a culture dish, to study principles underlying synchronization, adaptation, learning, and representation in neuronal assemblies. The motivation to study neuronal networks ex vivo is outlined together with a short description of recent results in this field. Following a short description of the experimental system, a set of basic results will be presented that concern self-organization of activity, dynamical and functional properties of neurons and networks in response to external stimulation. This short review ends with an outline of future questions and research directions.

Revealing ensemble state transition patterns in multi-electrode neuronal recordings using hidden Markov models

In order to harness the computational capacity of dissociated cultured neuronal networks, it is necessary to understand neuronal dynamics and connectivity on a mesoscopic scale. To this end, this paper uncovers dynamic spatiotemporal patterns emerging from electrically stimulated neuronal cultures using hidden Markov models (HMMs) to characterize multi-channel spike trains as a progression of patterns of underlying states of neuronal activity. However, experimentation aimed at optimal choice of parameters for such models is essential and results are reported in detail. Results derived from ensemble neuronal data revealed highly repeatable patterns of state transitions in the order of milliseconds in response to probing stimuli.

Role of homeostasis in learning sparse representations

Neurons in the input layer of primary visual cortex in primates develop edge-like receptive fields. One approach to understanding the emergence of this response is to state that neural activity has to efficiently represent sensory data with respect to the statistics of natural scenes. Furthermore, it is believed that such an efficient coding is achieved using a competition across neurons so as to generate a sparse representation, that is, where a relatively small number of neurons are simultaneously active. Indeed, different models of sparse coding, coupled with Hebbian learning and homeostasis, have been proposed that successfully match the observed emergent response. However, the specific role of homeostasis in learning such sparse representations is still largely unknown. By quantitatively assessing the efficiency of the neural representation during learning, we derive a cooperative homeostasis mechanism that optimally tunes the competition between neurons within the sparse coding algorithm. We apply this homeostasis while learning small patches taken from natural images and compare its efficiency with state-of-the-art algorithms. Results show that while different sparse coding algorithms give similar coding results, the homeostasis provides an optimal balance for the representation of natural images within the population of neurons. Competition in sparse coding is optimized when it is fair. By contributing to optimizing statistical competition across neurons, homeostasis is crucial in providing a more efficient solution to the emergence of independent components.

Dynamics of excitability over extended timescales in cultured cortical neurons

Although neuronal excitability is well understood and accurately modeled over timescales of up to hundreds of milliseconds, it is currently unclear whether extrapolating from this limited duration to longer behaviorally relevant timescales is appropriate. Here we used an extracellular recording and stimulation paradigm that extends the duration of single-neuron electrophysiological experiments, exposing the dynamics of excitability in individual cultured cortical neurons over timescales hitherto inaccessible. We show that the long-term neuronal excitability dynamics is unstable and dominated by critical fluctuations, intermittency, scale-invariant rate statistics, and long memory. These intrinsic dynamics bound the firing rate over extended timescales, contrasting observed short-term neuronal response to stimulation onset. Furthermore, the activity of a neuron over extended timescales shows transitions between quasi-stable modes, each characterized by a typical response pattern. Like in the case of rate statistics, the short-term onset response pattern that often serves to functionally define a given neuron is not indicative of its long-term ongoing response. These observations question the validity of describing neuronal excitability based on temporally restricted electrophysiological data, calling for in-depth exploration of activity over wider temporal scales. Such extended experiments will probably entail a different kind of neuronal models, accounting for the unbounded range, from milliseconds up.

A generic framework for real-time multi-channel neuronal signal analysis, telemetry control, and sub-millisecond latency feedback generation

Distinct modules of the neural circuitry interact with each other and (through the motor-sensory loop) with the environment, forming a complex dynamic system. Neuro-prosthetic devices seeking to modulate or restore CNS function need to interact with the information flow at the level of neural modules electrically, bi-directionally and in real-time. A set of freely available generic tools is presented that allow computationally demanding multi-channel short-latency bi-directional interactions to be realized in in vivo and in vitro preparations using standard PC data acquisition and processing hardware and software (Mathworks Matlab and Simulink). A commercially available 60-channel extracellular multi-electrode recording and stimulation set-up connected to an ex vivo developing cortical neuronal culture is used as a model system to validate the method. We demonstrate how complex high-bandwidth (>10 MBit/s) neural recording data can be analyzed in real-time while simultaneously generating specific complex electrical stimulation feedback with deterministically timed responses at sub-millisecond resolution.

Wrestling model of the repertoire of activity propagation modes in quadruple neural networks

The spontaneous activity of engineered quadruple cultured neural networks (of four-coupled sub-networks) exhibits a repertoire of different types of mutual synchronization events. Each event corresponds to a specific activity propagation mode (APM) defined by the order of activity propagation between the sub-networks. We statistically characterized the frequency of spontaneous appearance of the different types of APMs. The relative frequencies of the APMs were then examined for their power-law properties. We found that the frequencies of appearance of the leading (most frequent) APMs have close to constant algebraic ratio reminiscent of Zipf's scaling of words. We show that the observations are consistent with a simplified “wrestling” model. This model represents an extension of the “boxing arena” model which was previously proposed to describe the ratio between the two activity modes in two coupled sub-networks. The additional new element in the “wrestling” model presented here is that the firing within each network is modeled by a time interval generator with similar intra-network Lévy distribution. We modeled the different burst-initiation zones’ interaction by competition between the stochastic generators with Gaussian inter-network variability. Estimation of the model parameters revealed similarity across different cultures while the inter-burst-interval of the cultures was similar across different APMs as numerical simulation of the model predicts.

Tradeoffs and constraints on neural representation in networks of cortical neurons

Neural representation is pivotal in neuroscience. Yet, the large number and variance of underlying determinants make it difficult to distinguish general physiologic constraints on representation. Here we offer a general approach to the issue, enabling a systematic and well controlled experimental analysis of constraints and tradeoffs, imposed by the physiology of neuronal populations, on plausible representation schemes. Using in vitro networks of rat cortical neurons as a model system, we compared the efficacy of different kinds of "neural codes" to represent both spatial and temporal input features. Two rate-based representation schemes and two time-based representation schemes were considered. Our results indicate that, by large, all representation schemes perform well in the various discrimination tasks tested, indicating the inherent redundancy in neural population activity; Nevertheless, differences in representation efficacy are identified when unique aspects of input features are considered. We discuss these differences in the context of neural population dynamics.

Contribution of GABAergic interneurons to the development of spontaneous activity patterns in cultured neocortical networks

Periodic synchronized events are a hallmark feature of developing neuronal networks and are assumed to be crucial for the maturation of the neuronal circuitry. In the developing neocortex, the early network oscillations coincide with an excitatory action of the neurotransmitter gamma-aminobutyric acid (GABA). A relationship between the emerging inhibitory action of GABA and the gradual disappearance of early synchronized network activity has been previously suggested. Therefore we investigate the interplay between the action of GABA and spontaneous activity in cultured networks of the lateral or dorsal embryonic rat neocortex, which show considerable difference in the content of GABAergic neurons. Here we present the results of long-term monitoring of spontaneous electrical activity of cultured networks growing on microelectrode arrays and the time course of changes in GABA action using calcium imaging. All cultures studied displayed stereotyped synchronized burst events at the end of the first week in vitro. As the GABA_A depolarizing action decreases, naturally or after bumetanide treatment, network activity in lateral cortex cultures changed from stereotypic bursting to more clustered and asynchronous activity patterns. Dorsal cortex cultures and cultures lacking GABA_A-receptor mediated synaptic transmission, retained an immature synchronous firing pattern, but developed prominent intraburst oscillations (∼3–10 Hz). Large, mostly parvalbumin positive, GABAergic neurons dominate the GABAergic population in lateral cortex cultures. These large interneurons were virtually absent in dorsal cortex cultures. Based on these results, we suggest that the richly interconnected large GABAergic neurons contribute to desynchronize and temporally differentiate the spontaneous activity of cultured cortical networks.

Long-term, multisite, parallel, in-cell recording and stimulation by an array of extracellular microelectrodes

Here we report on the development of a novel neuroelectronic interface consisting of an array of noninvasive gold-mushroom-shaped microelectrodes (gMmicroEs) that practically provide intracellular recordings and stimulation of many individual neurons, while the electrodes maintain an extracellular position. The development of this interface allows simultaneous, multisite, long-term recordings of action potentials and subthreshold potentials with matching quality and signal-to-noise ratio of conventional intracellular sharp glass microelectrodes or patch electrodes. We refer to the novel approach as "in-cell recording and stimulation by extracellular electrodes" to differentiate it from the classical intracellular recording and stimulation methods. This novel technique is expected to revolutionize the analysis of neuronal networks in relations to learning, information storage and can be used to develop novel drugs as well as high fidelity neural prosthetics and brain-machine systems.

Functional identification of biological neural networks using reservoir adaptation for point processes

The complexity of biological neural networks does not allow to directly relate their biophysical properties to the dynamics of their electrical activity. We present a reservoir computing approach for functionally identifying a biological neural network, i.e. for building an artificial system that is functionally equivalent to the reference biological network. Employing feed-forward and recurrent networks with fading memory, i.e. reservoirs, we propose a point process based learning algorithm to train the internal parameters of the reservoir and the connectivity between the reservoir and the memoryless readout neurons. Specifically, the model is an Echo State Network (ESN) with leaky integrator neurons, whose individual leakage time constants are also adapted. The proposed ESN algorithm learns a predictive model of stimulus-response relations in in vitro and simulated networks, i.e. it models their response dynamics. Receiver Operating Characteristic (ROC) curve analysis indicates that these ESNs can imitate the response signal of a reference biological network. Reservoir adaptation improved the performance of an ESN over readout-only training methods in many cases. This also held for adaptive feed-forward reservoirs, which had no recurrent dynamics. We demonstrate the predictive power of these ESNs on various tasks with cultured and simulated biological neural networks.