NEURON, SenseLab ModelDB, NeuroMorpho.org, and Signal Processing in Brain Microcircuitry

Each neuron in the brain may carry out an enormous amount of computation. A neuron may receive thousands of different signal inputs through chemical synapses which the neuron can integrate and modify through the neuron’s shape (cell morphology), membrane conductance properties, and intracellular biochemical processes, including moment-by-moment changes in gene expression. The convergence of neuroinformatics tools (including online data repositories), simulation tools, new research techniques, and especially the sharing of these tools, data, knowledge, and techniques through the Internet is enabling us to make more progress in understanding the mechanics of computation where most of it is really taking place in the brain. In the neuropil composed of axon terminals, synapses, and dendrites.

One of two digital reconstructions of CA1 pyramidal neurons from the NeuroMorpho.org data repository used in the paper being discussed.
Figure 1. One of two digital reconstructions of CA1 pyramidal neurons from the NeuroMorpho.org data repository used in the paper being discussed. Neuron pc2b is shown here. The neuron ri06 was also used (not shown).

The authors of the paper “Local Control of Postinhibitory Rebound Spiking in CA1 Pyramidal Neuron Dendrites” (published May 5, 2010 in the Journal of Neuroscience) use today’s tools to great effect to investigate the possibility of a previously undiscovered signal processing mechanism at work in microcircuits localized in the neuropil. The focus is on hippocampus Cornu Ammonis area 1 (CA1) pyramidal neurons, postinhibitory rebound spikes, and two membrane channels: 1) hyperpolarization-activated inward current (I-h) and 2) A-type potassium channels (K-A). To test their hypothesis they use the NEURON simulator, previous standard simulations of CA1 pyramidal neurons from SenseLab’s ModelDB repository, reconstructions of CA1 pyramidal neurons from the NeuroMorpho.org data repository, and animal experiments.

Note: Postinhibitory rebound spikes are action potentials elicited when a neuron’s membrane potential returns (rebounds) to a more electrically positive value after being hyperpolarized (made more electrically negative) by inhibitory input.

I-h is expressed in a gradient in Hippocampus Cornu Ammonis area 1 (CA1) pyramidal neurons, with the lowest density in dendrites next to the soma and the highest density in the distal dendrites. I-h can participate in the mechanism underlying the generation of postinhibitory rebound spikes. Furthermore, the dendritic distribution of I-h, increasing with the distance from the soma, is paralleled by K-A. The authors hypothesized that the relatively high concentration of I-h and K-A on CA1 pyramidal neuron distal dendrites provides a localized substrate for signal processing in the neuropil. The increase and decrease in the number of K-A, known to be regulated by intracellular second messenger and phosphorylation cascades, in distal dendritic membranes would regulate the local generation of rebound spikes. Down regulation of K-A would increase the likelihood of a postinhibitory rebound spike. Up regulation of K-A would decrease the likelihood.

Note: The NEURON model associated with the paper being discussed is “CA1 pyramidal neuron: rebound spiking (Ascoli et al.2010)” and may be downloaded from the SenseLab ModelDB repository. Also, the following previously existing model was used: “CA1 pyramidal neuron: signal propagation in oblique dendrites (Migliore et al 2005).”

The results showed that K-A reduction could enable the initiation of local dendritic postinhibitory rebound spikes. Generation of postinhibitory rebound spikes depends on a combination of the number and the spatial extend of K-A reduction. In some configurations this may even lead to the spike’s propagation to the soma.


Other related blog posts:

Brain Modeling Using NEURON, Interneurons, and Resonant Circuits

Brain Modeling Using NEURON: Superficial Pyramidal, Deep Pyramidal, Aspiny, and Stellate Neurons

Brain Modeling Using NEURON: Neural Activity Underlying Magnetoencephalography