Small Number Of Neurons Research Articles

Super Resolution Imaging of Genetically Labeled Synapses in Drosophila Brain Tissue.

Understanding synaptic connectivity and plasticity within brain circuits and their relationship to learning and behavior is a fundamental quest in neuroscience. Visualizing the fine details of synapses using optical microscopy remains however a major technical challenge. Super resolution microscopy opens the possibility to reveal molecular features of synapses beyond the diffraction limit. With direct stochastic optical reconstruction microscopy, dSTORM, we image synaptic proteins in the brain tissue of the fruit fly, Drosophila melanogaster. Super resolution imaging of brain tissue harbors difficulties due to light scattering and the density of signals. In order to reduce out of focus signal, we take advantage of the genetic tools available in the Drosophila and have fluorescently tagged synaptic proteins expressed in only a small number of neurons. These neurons form synapses within the calyx of the mushroom body, a distinct brain region involved in associative memory formation. Our results show that super resolution microscopy, in combination with genetically labeled synaptic proteins, is a powerful tool to investigate synapses in a quantitative fashion providing an entry point for studies on synaptic plasticity during learning and memory formation.

Visual Learning of Arithmetic Operation

A simple Neural Network model is presented for end-to-end visual learning of arithmetic operations from pictures of numbers. The input consists of two pictures, each showing a 7-digit number. The output, also a picture, displays the number showing the result of an arithmetic operation (e.g., addition or subtraction) on the two input numbers. The concepts of a number, or of an operator, are not explicitly introduced. This indicates that addition is a simple cognitive task, which can be learned visually using a very small number of neurons. Other operations, e.g., multiplication, were not learnable using this architecture. Some tasks were not learnable end-to-end (e.g., addition with Roman numerals), but were easily learnable once broken into two separate sub-tasks: a perceptual Character Recognition and cognitive Arithmetic sub-tasks. This indicates that while some tasks may be easily learnable end-to-end, other may need to be broken into sub-tasks.

Spiking neural P systems with request rules

Spiking neural P systems, shortly called SN P systems, are a class of distributed and parallel neural-like computing models, inspired from the way of neurons spiking and communicating with each other by means of spikes. In this work, we propose a new variant of SN P systems, called SN P systems with request rules. In such a system, besides spiking and forgetting rules, a neuron can have request rules, with which the neuron can sense “stimulus” from the environment by receiving a certain number of spikes. We investigate the computation power of SN P systems with request rules. It is obtained that such systems are Turing universal, even with a small number of neurons. Specifically, (i) SN P systems with request rules having 4 neurons can compute any set of Turing computable natural numbers and (ii) with 47 neurons such systems can compute any Turing computable function.

Speaker Verification via Modeling Kurtosis Using Sparse Coding

This paper proposes a new model for speaker verification by employing kurtosis statistical method based on sparse coding of human auditory system. Since only a small number of neurons in primary auditory cortex are activated in encoding acoustic stimuli and sparse independent events are used to represent the characteristics of the neurons. Each individual dictionary is learned from individual speaker samples where dictionary atoms correspond to the cortex neurons. The neuron responses possess statistical properties of acoustic signals in auditory cortex so that the activation distribution of individual speaker’s neurons is approximated as the characteristics of the speaker. Kurtosis is an efficient approach to measure the sparsity of the neuron from its activation distribution, and the vector composed of the kurtosis of every neuron is obtained as the model to characterize the speaker’s voice. The experimental results demonstrate that the kurtosis model outperforms the baseline systems and an effective identity validation function is achieved desirably.

DTRPA1 in Non-circadian Neurons Modulates Temperature-dependent Rhythmic Activity in Drosophila melanogaster.

In fruit flies Drosophila melanogaster, environmental cycles of light and temperature are known to influence behavioral rhythms through dedicated sensory receptors. But the thermosensory pathways and molecular receptors by which thermal cycles modulate locomotor activity rhythms remain unclear. Here, we report that neurons expressing warmth-activated ion channel Drosophila Transient Receptor Potential-A1 (dTRPA1) modulate distinct aspects of the rhythmic activity/rest rhythm in a light-dependent manner. Under light/dark (LD) cycles paired with constantly warm ambient conditions, flies deficient in dTRPA1 expression are unable to phase morning and evening activity bouts appropriately. Correspondingly, we show that electrical activity of a few neurons targeted by the dTRPA1(SH)-GAL4 driver modulates temperature-dependent phasing of activity/rest rhythm under LD cycles. The expression of dTRPA1 also affects behavior responses to temperature cycles combined with constant dark (DD) or light (LL) conditions. We demonstrate that the mid-day "siesta" exhibited by flies under temperature cycles in DD is dependent on dTRPA1 expression in a small number of neurons that include thermosensory anterior cell neurons. Although a small subset of circadian pacemaker neurons may express dTRPA1, we show that CRY-negative dTRPA1(SH)-GAL4 driven neurons are critical for the suppression of mid-thermophase activity, thus enabling flies to exhibit siesta. In contrast to temperature cycles in DD, under LL, dTRPA1 is not required for exhibiting siesta but is important for phasing of evening peak. Our studies show that activity/rest rhythms are modulated in a temperature-dependent manner via signals from dTRPA1(SH)-GAL4 driven neurons. Taken together, these results emphasize the differential influence of thermoreceptors on rhythmic behavior in fruit flies in coordination with light inputs.

QRFP and Its Receptors Regulate Locomotor Activity and Sleep in Zebrafish.

The hypothalamus is thought to play a key role in regulating sleep in vertebrate animals, but few sleep-promoting signaling pathways that function in the hypothalamus have been identified. Here we use the zebrafish, a diurnal vertebrate, to functionally and anatomically characterize the neuropeptide QRFP. We show that QRFP is exclusively expressed in a small number of neurons in the larval zebrafish hypothalamus that project widely in the brain. We also show that QRFP overexpression reduces locomotor activity, whereas animals that lack QRFP signaling are more active and sleep less. These results suggest that QRFP signaling participates in the hypothalamic regulation of sleep.

Commentary: Burst Firing in a Motion-Sensitive Neural Pathway Correlates with Expansion Properties of Looming Objects That Evoke Avoidance Behaviors.

Commentary: Burst Firing in a Motion-Sensitive Neural Pathway Correlates with Expansion Properties of Looming Objects That Evoke Avoidance Behaviors.

Identifying and tracking simulated synaptic inputs from neuronal firing: insights from in vitro experiments

Accurately describing synaptic interactions between neurons and how interactions change over time are key challenges for systems neuroscience. Although intracellular electrophysiology is a powerful tool for studying synaptic integration and plasticity, it is limited by the small number of neurons that can be recorded simultaneously in vitro and by the technical difficulty of intracellular recording in vivo. One way around these difficulties may be to use large-scale extracellular recording of spike trains and apply statistical methods to model and infer functional connections between neurons. These techniques have the potential to reveal large-scale connectivity structure based on the spike timing alone. However, the interpretation of functional connectivity is often approximate, since only a small fraction of presynaptic inputs are typically observed. Here we use in vitro current injection in layer 2/3 pyramidal neurons to validate methods for inferring functional connectivity in a setting where input to the neuron is controlled. In experiments with partially-defined input we inject a single simulated input with known amplitude on a background of fluctuating noise. In a fully-defined input paradigm we then control the synaptic weights and timing of many (1024) simulated presynaptic neurons. By analyzing the firing of neurons in response to these artificial inputs, we ask 1) How does functional connectivity inferred from spikes relate to simulated synaptic input? and 2) What are the limitations of connectivity inference? Using likelihood-based models of functional connectivity, we find that individual current-based synaptic inputs are detectable over a broad range of amplitudes and conditions. Detectability depends on input amplitude and output firing rate, and excitatory inputs are detected more readily than inhibitory (Figure ​(Figure1).1). Moreover, as we model increasing numbers of presynaptic inputs, we are able to estimate connection strengths more accurately and detect the presence of connections more quickly. We find that model-based methods outperform previous methods for detecting synaptic connections (nonparametric tests on the cross-correlation), and simulation results (adaptive exponential integrate-and-fire) suggest that the model results hold even when synaptic input is conductance-based rather than current-based. Together these results illustrate the possibilities and outline the limits of inferring synaptic input from spikes. Figure 1 A) Detectability of synaptic inputs of different strengths in a fully-defined input paradigm (data for N = 4 Layer 2/3 pyramidal cells) for 1Hz, 5Hz, and 10Hz post-synaptic firing rates. Note that detection of inputs improves with higher firing rates ...

Spiking neural network model of reinforcement learning in the honeybee implemented on the GPU

Honeybees can learn and perform complex behavioral tasks despite their small brains that contain less than a million neurons. At the same time they are accessible to physiological experiments and the relatively small number of neurons in their brain lends itself to quite detailed numerical simulations. Bees therefore are a good model system for studying sensory cognition and reinforcement learning. We have shown in earlier work [1] that the anatomy and known electrophysiological properties of the olfactory pathway of insects in combination with spike-timing dependent plasticity (STDP) and lateral inhibition lend themselves to an unsupervised self-organization of synaptic connections for the recognition of odors. Here we extend this model by adding mechanisms of reinforcement learning, as suggested by [2] (see Figure ​Figure1).1). We employ a three factor learning rule where plasticity is governed by pre-synaptic and post-synaptic activity and a global octopaminergic/dopaminergic reinforcement signal, triggered by a reward. We investigated the role of feed-forward and feedback mechanisms, as well as the role of the connectivity initially achieved by unsupervised STDP. Figure 1 Network diagram for the hypothesized model of reinforcement learning in the honeybee olfactory system. Excitatory connections are shown in black, inhibitory connections in blue and learning synapses in red. Grey arrows represent the abstractions modeled ... Our model is implemented in the GeNN [3] framework, which facilitates the use of GPUs for spiking neural network simulations using a code generation framework. Because of the massive parallelism provided by GPUs, we can simulate tens of thousands of neurons in real time in the sparse firing regime relevant here. We investigated optimization strategies and neuron and synapse model choices for a better performance on the GPU. The model presented here is a stepping-stone to more sophisticated learning models and multi-sensory integration in the Green Brain Project [4], in which we aim to control a flying robot with a simulation of learning and decision making mechanisms in the honeybee related both to the olfactory and visual pathways.

A computational model of cell culture dynamics: the role of connectivity and synaptic receptors in the appearance of synchronized bursting events

How an ensemble of neurons wires together to form a functional unit is a fundamental problem in neuroscience. The architecture of neuronal wiring, in fact, determines how neurons communicate and may be important for information processing performed by neuronal networks. However current knowledge is mainly limited to networks consisting of a small number of neurons, while the topological structure of biological networks remains still unknown. Primary neuronal cultures represent an ideal preparation to investigate the basic principles of network dynamics. At the mature stage, they display network-wide synchronous bursting events (SBEs) sharing similar spatiotemporal patterns of firing [1]. Interestingly, high-density MEA recordings have shown that SBEs actually correspond to propagating activities through the network. Typically, the simulated SBEs originated from a few and specific sites as in experiments [2], but the nature and the role of such events is still under debate. In order to investigate the determinants of such dynamics, we developed a computational model that mimics the main features of the recordings obtained by a high density multi-electrode-array device (4096 electrodes inter-spaced by 20um, [3]). With only a few topological constraints, the model expressed realistic SBEs along time that can be well clustered into only a few groups differing for their ignition sites and propagation directions, similarly to what it is observed experimentally. Furthermore, we used the model together with experimental datasets to investigate the effects of synaptic blockers of the AMPA, NMDA and GABA synapses on the network activity. In particular, we showed that NMDA receptors can be among the principal mechanisms involved in triggering a sequence of SBEs, a firing regime that is typically observed in for mature neuronal cultures. Such regime is characterized by a principal SBE recruiting a great percentage of neurons and followed by a sequence of several weaker SBEs interleaved by hundreds of milliseconds. Altogether, the results obtained with our neural network computational model show that this model can replicate most of the salient firing properties observed experimentally in cultured neuronal networks and that it can serve for exploring the properties of signals and responses observed in neuronal networks properties.

Importance of Adult Dmbx1 in Long-Lasting Orexigenic Effect of Agouti-Related Peptide.

Dmbx1 is a brain-specific homeodomain transcription factor expressed primarily during embryogenesis, and its systemic disruption (Dmbx1(-/-)) in the ICR mouse strain resulted in leanness associated with impaired long-lasting orexigenic effect of agouti-related peptide (AgRP). Because spatial and temporal expression patterns of Dmbx1 change dramatically during embryogenesis, it remains unknown when and where Dmbx1 plays a critical role in energy homeostasis. In the present study, the physiological roles of Dmbx1 were examined by its conditional disruption (Dmbx1(loxP/loxP)) in the C57BL/6 mouse strain. Although Dmbx1 disruption in fetal brain resulted in neonatal lethality, its disruption by synapsin promoter-driven Cre recombinase, which eliminated Dmbx1 expression postnatally, exempted the mice (Syn-Cre;Dmbx1(loxP/loxP) mice) from lethality. Syn-Cre;Dmbx1(loxP/loxP) mice show mild leanness and impaired long-lasting orexigenic action of AgRP, demonstrating the physiological relevance of Dmbx1 in the adult. Visualization of Dmbx1-expressing neurons in adult brain using the mice harboring tamoxifen-inducible Cre recombinase in the Dmbx1 locus (Dmbx1(CreERT2/+) mice) revealed Dmbx1 expression in small numbers of neurons in restricted regions, including the lateral parabrachial nucleus (LPB). Notably, c-Fos expression in LPB was increased at 48 hours after AgRP administration in Dmbx1(loxP/loxP) mice but not in Syn-Cre;Dmbx1(loxP/loxP) mice. These c-Fos-positive neurons in LPB did not coincide with neurons expressing Dmbx1 or melanocortin 4 receptor but did coincide with those expressing calcitonin gene-related peptide. Accordingly, Dmbx1 in the adult LPB is required for the long-lasting orexigenic effect of AgRP via the neural circuitry involving calcitonin gene-related peptide neurons.

A new implementation of the algorithm of adaptive construction of hierarchical neural network classifiers

The article presents development of the algorithm of adaptive construction of hierarchical neural network classifiers based on automatic modification of the desired output of perceptrons with a small number of neurons in the single hidden layer. The conducted testing of the new program implementation of this approach demonstrated that the considered algorithm was more computationally efficient and provided higher quality of solution of multiple classification problems in comparison with standard multi-layer perceptron.

Convolutional Neural Networks with Fused Layers Applied to Face Recognition

In this paper, we propose an effective convolutional neural network (CNN) model to the problem of face recognition. The proposed CNN architecture applies fused convolution/subsampling layers that result in a simpler model with fewer network parameters; that is, a smaller number of neurons, trainable parameters, and connections. In addition, it does not require any complex or costly image preprocessing steps that are typical in existing face recognizer systems. In this work, we enhance the stochastic diagonal Levenberg–Marquardt algorithm, a second-order back-propagation algorithm to obtain faster network convergence and better generalization ability. Experimental work completed on the ORL database shows that a recognition accuracy of 100% is achieved, with the network converging within 15 epochs. The average processing time of the proposed CNN face recognition solution, executed on a 2.5 GHz Intel i5 quad-core processor, is 3 s per epoch, with a recognition speed of less than 0.003 s. These results show that the proposed CNN model is a computationally efficient architecture that exhibits faster processing and learning times, and also produces higher recognition accuracy, outperforming other existing work on face recognizers based on neural networks.

Deep learning human actions from video via sparse filtering and locally competitive algorithms

Physiological and psychophysical evidence suggest that early visual cortex compresses the visual input on the basis of spatial and orientation-tuned filters. Recent computational advances have suggested that these neural response characteristics may reflect a `sparse coding' architecture--in which a small number of neurons need to be active for any given image--yielding critical structure latent in natural scenes. Here we present a novel neural network architecture combining a sparse filter model and locally competitive algorithms (LCAs), and demonstrate the network's ability to classify human actions from video. Sparse filtering is an unsupervised feature learning algorithm designed to optimize the sparsity of the feature distribution directly without having the need to model the data distribution. LCAs are defined by a system of differential equations where the initial conditions define an optimization problem and the dynamics converge to a sparse decomposition of the input vector. We applied this architecture to train a classifier on categories of motion in human action videos. Inputs to the network were small 3D patches taken from frame differences in the videos. Dictionaries were derived for each action class and then activation levels for each dictionary were assessed during reconstruction of a novel test patch. Overall, classification accuracy was at ? 97 %. We discuss how this sparse filtering approach provides a natural framework for multi-sensory and multimodal data processing including RGB video, RGBD video, hyper-spectral video, and stereo audio/video streams.

Tracking the fear memory engram: discrete populations of neurons within amygdala, hypothalamus, and lateral septum are specifically activated by auditory fear conditioning.

Memory formation is thought to occur via enhanced synaptic connectivity between populations of neurons in the brain. However, it has been difficult to localize and identify the neurons that are directly involved in the formation of any specific memory. We have previously used fos-tau-lacZ (FTL) transgenic mice to identify discrete populations of neurons in amygdala and hypothalamus, which were specifically activated by fear conditioning to a context. Here we have examined neuronal activation due to fear conditioning to a more specific auditory cue. Discrete populations of learning-specific neurons were identified in only a small number of locations in the brain, including those previously found to be activated in amygdala and hypothalamus by context fear conditioning. These populations, each containing only a relatively small number of neurons, may be directly involved in fear learning and memory.

Chemosensory apparatus of Drosophila larvae.

Many insects, including Drosophila melanogaster, have a rich repertoire of olfactory behavior. Combination of robust behavioral assays, physiological and molecular tools render D. melanogaster as highly suitable system for olfactory studies. The small number of neurons in the olfactory system of fruit flies, especially the number of sensory neurons in the larval stage, makes the exploration of sensory coding at all stages of its nervous system a potentially tractable goal, which is not possible in the foreseeable future in any mammalian preparation. Advances in physiological recordings, olfactory signaling and detailed analysis of behavior, can place larvae in a position to ask previously unanswerable questions.

Identifying and tracking simulated synaptic inputs from neuronal firing: insights from in vitro experiments.

Accurately describing synaptic interactions between neurons and how interactions change over time are key challenges for systems neuroscience. Although intracellular electrophysiology is a powerful tool for studying synaptic integration and plasticity, it is limited by the small number of neurons that can be recorded simultaneously in vitro and by the technical difficulty of intracellular recording in vivo. One way around these difficulties may be to use large-scale extracellular recording of spike trains and apply statistical methods to model and infer functional connections between neurons. These techniques have the potential to reveal large-scale connectivity structure based on the spike timing alone. However, the interpretation of functional connectivity is often approximate, since only a small fraction of presynaptic inputs are typically observed. Here we use in vitro current injection in layer 2/3 pyramidal neurons to validate methods for inferring functional connectivity in a setting where input to the neuron is controlled. In experiments with partially-defined input, we inject a single simulated input with known amplitude on a background of fluctuating noise. In a fully-defined input paradigm, we then control the synaptic weights and timing of many simulated presynaptic neurons. By analyzing the firing of neurons in response to these artificial inputs, we ask 1) How does functional connectivity inferred from spikes relate to simulated synaptic input? and 2) What are the limitations of connectivity inference? We find that individual current-based synaptic inputs are detectable over a broad range of amplitudes and conditions. Detectability depends on input amplitude and output firing rate, and excitatory inputs are detected more readily than inhibitory. Moreover, as we model increasing numbers of presynaptic inputs, we are able to estimate connection strengths more accurately and detect the presence of connections more quickly. These results illustrate the possibilities and outline the limits of inferring synaptic input from spikes.

An abnormal GABAergic system in the inferior colliculus provides a basis for audiogenic seizures in genetically epilepsy-prone rats.

In this review of neuroanatomical studies of the genetically epilepsy-prone rat (GEPR), three main topics will be covered. First, the number of GABAergic neurons and total neurons in the inferior colliculus of GEPRs will be compared to those of the nonepileptic Sprague-Dawley rat. Next, the number of small neurons in the inferior colliculus will be described in both developmental and genetic analyses of GEPRs and their backcrosses. Last, results from two types of studies on the propagation pathways for audiogenic seizures in GEPRs will be shown. Together, these studies demonstrate a unique GABAergic, small neuron defect in the inferior colliculus of GEPRs that may play a vital role in the initiation and spread of seizure activity during audiogenic seizures. This article is part of a Special Issue entitled "Genetic and Reflex Epilepsies, Audiogenic Seizures and Strains: From Experimental Models to the Clinic".

Neural Coding: Sparse but On Time

To code information efficiently, sensory systems use sparse representations. In a sparse code, a specific stimulus activates only few spikes in a small number of neurons. A new study shows that the temporal pattern across sparsely activated neurons encodes information, suggesting that the sparse code extends into the time domain.

1D-3D hybrid modeling-from multi-compartment models to full resolution models in space and time.

Investigation of cellular and network dynamics in the brain by means of modeling and simulation has evolved into a highly interdisciplinary field, that uses sophisticated modeling and simulation approaches to understand distinct areas of brain function. Depending on the underlying complexity, these models vary in their level of detail, in order to cope with the attached computational cost. Hence for large network simulations, single neurons are typically reduced to time-dependent signal processors, dismissing the spatial aspect of each cell. For single cell or networks with relatively small numbers of neurons, general purpose simulators allow for space and time-dependent simulations of electrical signal processing, based on the cable equation theory. An emerging field in Computational Neuroscience encompasses a new level of detail by incorporating the full three-dimensional morphology of cells and organelles into three-dimensional, space and time-dependent, simulations. While every approach has its advantages and limitations, such as computational cost, integrated and methods-spanning simulation approaches, depending on the network size could establish new ways to investigate the brain. In this paper we present a hybrid simulation approach, that makes use of reduced 1D-models using e.g., the NEURON simulator—which couples to fully resolved models for simulating cellular and sub-cellular dynamics, including the detailed three-dimensional morphology of neurons and organelles. In order to couple 1D- and 3D-simulations, we present a geometry-, membrane potential- and intracellular concentration mapping framework, with which graph- based morphologies, e.g., in the swc- or hoc-format, are mapped to full surface and volume representations of the neuron and computational data from 1D-simulations can be used as boundary conditions for full 3D simulations and vice versa. Thus, established models and data, based on general purpose 1D-simulators, can be directly coupled to the emerging field of fully resolved, highly detailed 3D-modeling approaches. We present the developed general framework for 1D/3D hybrid modeling and apply it to investigate electrically active neurons and their intracellular spatio-temporal calcium dynamics.