Nicholas T. Carnevale

The NEURON simulation environment

The moment-to-moment processing of information by the nervous system involves the propagation and interaction of electrical and chemical signals that are distributed in space and time. Biologically realistic modeling is needed to test hypotheses about the mechanisms that govern these signals and how nervous system function emerges from the operation of these mechanisms. The NEURON simulation program provides a powerful and flexible environment for implementing such models of individual neurons and small networks of neurons. It is particularly useful when membrane potential is nonuniform and membrane currents are complex. We present the basic ideas that would help informed users make the most efficient use of NEURON.

Neuron: A Tool for Neuroscientists

NEURON is a simulation environment for models of individual neurons and networks of neurons that are closely linked to experimental data. NEURON provides tools for conveniently constructing, exercising, and managing models, so that special expertise in numerical methods or programming is not required for its productive use. This article describes two tools that address the problem of how to achieve computational efficiency and accuracy.

ModelDB: A Database to Support Computational Neuroscience

Wider dissemination and testing of computational models are crucial to the field of computational neuroscience. Databases are being developed to meet this need. ModelDB is a web-accessible database for convenient entry, retrieval, and running of published models on different platforms. This article provides a guide to entering a new model into ModelDB.

Expanding NEURON’s Repertoire of Mechanisms with NMODL

Neuronal function involves the interaction of electrical and chemical signals that are distributed in time and space. The mechanisms that generate these signals and regulate their interactions are marked by a rich diversity of properties that precludes a one size fits all approach to modeling. This article presents a summary of how the model description language NMODL enables the neuronal simulation environment NEURON to accommodate these differences.

Compartmentalization of GABAergic Inhibition by Dendritic Spines

Dendritic Precision Strikes The effects of excitatory synaptic inputs are considered to be highly compartmentalized because of the biophysical properties of dendritic spines. Individual inhibitory synapses, however, are thought to affect dendritic integration in a more extended spatial region. Combining optogenetic stimulation of dendrite-targeting γ-aminobutyric acid—mediated interneurons with two-photon calcium imaging in postsynaptic pyramidal cell dendrites, Chiu et al. (p. 759) challenge this latter view. The findings suggest that the effect of an inhibitory synapse can be as compartmentalized as that of an excitatory synapse, provided that the synapses are localized on spine heads. Inhibitory synapses can control individual dendritic spines independently from their neighbors. γ-aminobutyric acid–mediated (GABAergic) inhibition plays a critical role in shaping neuronal activity in the neocortex. Numerous experimental investigations have examined perisomatic inhibitory synapses, which control action potential output from pyramidal neurons. However, most inhibitory synapses in the neocortex are formed onto pyramidal cell dendrites, where theoretical studies suggest they may focally regulate cellular activity. The precision of GABAergic control over dendritic electrical and biochemical signaling is unknown. By using cell type-specific optical stimulation in combination with two-photon calcium (Ca2+) imaging, we show that somatostatin-expressing interneurons exert compartmentalized control over postsynaptic Ca2+ signals within individual dendritic spines. This highly focal inhibitory action is mediated by a subset of GABAergic synapses that directly target spine heads. GABAergic inhibition thus participates in localized control of dendritic electrical and biochemical signaling.

Electrical Advantages of Dendritic Spines

Many neurons receive excitatory glutamatergic input almost exclusively onto dendritic spines. In the absence of spines, the amplitudes and kinetics of excitatory postsynaptic potentials (EPSPs) at the site of synaptic input are highly variable and depend on dendritic location. We hypothesized that dendritic spines standardize the local geometry at the site of synaptic input, thereby reducing location-dependent variability of local EPSP properties. We tested this hypothesis using computational models of simplified and morphologically realistic spiny neurons that allow direct comparison of EPSPs generated on spine heads with EPSPs generated on dendritic shafts at the same dendritic locations. In all morphologies tested, spines greatly reduced location-dependent variability of local EPSP amplitude and kinetics, while having minimal impact on EPSPs measured at the soma. Spine-dependent standardization of local EPSP properties persisted across a range of physiologically relevant spine neck resistances, and in models with variable neck resistances. By reducing the variability of local EPSPs, spines standardized synaptic activation of NMDA receptors and voltage-gated calcium channels. Furthermore, spines enhanced activation of NMDA receptors and facilitated the generation of NMDA spikes and axonal action potentials in response to synaptic input. Finally, we show that dynamic regulation of spine neck geometry can preserve local EPSP properties following plasticity-driven changes in synaptic strength, but is inefficient in modifying the amplitude of EPSPs in other cellular compartments. These observations suggest that one function of dendritic spines is to standardize local EPSP properties throughout the dendritic tree, thereby allowing neurons to use similar voltage-sensitive postsynaptic mechanisms at all dendritic locations.

Translating network models to parallel hardware in NEURON.

The increasing complexity of network models poses a growing computational burden. At the same time, computational neuroscientists are finding it easier to access parallel hardware, such as multiprocessor personal computers, workstation clusters, and massively parallel supercomputers. The practical question is how to move a working network model from a single processor to parallel hardware. Here we show how to make this transition for models implemented with NEURON, in such a way that the final result will run and produce numerically identical results on either serial or parallel hardware. This allows users to develop and debug models on readily available local resources, then run their code without modification on a parallel supercomputer.

Discrete event simulation in the NEURON environment

Abstract The response of many types of integrate and fire cells to synaptic input can be computed analytically and their threshold crossing either computed analytically or approximated to high accuracy via Newton approximation. The NEURON simulation environment simulates networks of such artificial spiking neurons using discrete event simulation techniques in which computations are performed only when events are received. Thus, computation time is proportional only to the number of events delivered and is independent of the number of cells or problem time.

Efficient mapping from neuroanatomical to electrotonic space

Previous studies documented the importance of electrotonic structure in single-neuron computations. Here we elaborate a new approach to electrotonic theory and analysis. We begin with a more versatile measure Lij of the electrotonic distance between any two locations i and j on a neuron. If Vi is the voltage at the origin and Vj is the voltage at Some other point, the electrotonic distance is Lij=ln(Vi/Vj). Voltage decays e-fold per unit of L for any two points on the neuron, regardless of its electrotonic architecture. Lij increases as the sinusoidal frequency of Vi increases. If j lies on the direct path between i and k, then Lik=Lij+Ljk. This relation enables electrotonic transforms of the neuron—graphical mappings from neuroanatomical to electrotonic space. For each neuron, there exists an infinite number of such transforms, which can be done from any reference location on the neuron, as a function of voltage transfer to or from that location, and for any frequency of input signal. Sets of these trans...

Electrical behaviour of dendritic spines as revealed by voltage imaging

Thousands of dendritic spines on individual neurons process information and mediate plasticity by generating electrical input signals using a sophisticated assembly of transmitter receptors and voltage-sensitive ion channel molecules. Our understanding, however, of the electrical behaviour of spines is limited because it has not been possible to record input signals from these structures with adequate sensitivity and spatiotemporal resolution. Current interpretation of indirect data and speculations based on theoretical considerations are inconclusive. Here we use an electrochromic voltage-sensitive dye which acts as a transmembrane optical voltmeter with a linear scale to directly monitor electrical signals from individual spines on thin basal dendrites. The results show that synapses on these spines are not electrically isolated by the spine neck to a significant extent. Electrically, they behave as if they are located directly on dendrites.