Skip to content

Advertisement

  • Poster presentation
  • Open Access

Prediction of ion channel parameter differences between groups of young and aged pyramidal neurons using multi-stage compartmental model optimization

  • 1Email author,
  • 2,
  • 3,
  • 1 and
  • 2
BMC Neuroscience201516 (Suppl 1) :P282

https://doi.org/10.1186/1471-2202-16-S1-P282

  • Published:

Keywords

  • Differential Evolution
  • Pyramidal Neuron
  • Latin Hypercube Sampling
  • Optimization Protocol
  • Target Neuron

Electrophysiological recording and morphological reconstruction of pyramidal neurons from layer 3 prefrontal cortex of young and aged adult rhesus monkeys reveals higher firing rates and structural differences in aged neurons relative to young [1]. Prior computational modeling of these neurons has demonstrated that morphology alone is insufficient to account for functional differences, predicting that passive and active biophysical parameters differ between age groups [2]. Here we use compartmental models featuring several independent ion channels to provide insight into the precise electrophysiological parameters that underlie excitability differences between these neurons.

We fit ion channel parameters in a neuron model using a rapid, multi-stage, semi-automated approach based on our previous optimization protocol [3], resulting in a population of models representing the target neuron. First, we use a stereotypical pyramidal cell morphology [4] with size scaled to match the target neuron surface area. Second, we use differential evolution (DE) to tune passive and H-current parameters that shape the model's response to subthreshold somatic current injection. Third, we add 7 voltage- and calcium-dependent ion channels with free maximal conductance and kinetic parameters, and set up an optimization that will fit model output to empirical responses to current injections that evoke repetitive action potential firing. We manually select error functions for the suprathreshold optimization, then use a Latin hypercube sampling (LHS) to generate a space-filling set of parameter combinations. We then simulate each point in the LHS, and use the results to calculate weights for error functions. This novel weighting method is fully automated, and scales error functions based on their expected contribution to the total error during the suprathreshold optimization. Fourth, we perform suprathreshold optimization using DE, with the free parameters and weighted error functions chosen in stage three. The result of this optimization is a population of models that fit the target neuron. DE allows for fully parallel optimization, so we generate 256 well-fitting models within 80 hours of optimization through the Neuroscience Gateway [5].

We have used this method to generate populations of models for 1 aged and 3 young neurons. Linear discriminant analysis reveals that these four populations are readily distinguished within the parameter space. Models of the three young neurons have different ion channel parameters despite similarities in electrophysiological responses. Principal component analysis (PCA) reveals that all models of the aged neuron can be separated from models of the young neurons along the first principal component. Preliminary differences between parameters fitting these four neurons reveal, for example, that in the aged neuron the L-type calcium channel activates more slowly and has greater maximal conductance, and voltage dependence of the persistent sodium channel is lower. Generation of populations of models for other empirically characterized young and aged neurons is underway with our optimization protocol. Analytical techniques such as PCA will help generate predictions about intracellular changes during normal aging.

Authors’ Affiliations

(1)
Fishberg Department of Neuroscience and Friedman Brain Institute, Icahn School of Medicine at Mount Sinai, New York, NY 10029, USA
(2)
Department of Mathematics, Franklin and Marshall College, Lancaster, PA 17604, USA
(3)
Department of Anatomy and Neurobiology, Boston University School of Medicine, Boston, MA 02118, USA

References

  1. Chang YM, Rosene DL, Killiany RJ, Mangiamele LA, Luebke JI: Increased action potential firing rates of layer 2/3 pyramidal cells in the prefrontal cortex are significantly related to cognitive performance in aged monkeys. Cereb Cortex. 2005, 15 (4): 409-418.PubMedView ArticleGoogle Scholar
  2. Coskren P, Luebke JI, Kabaso D, Wearne SL, Yadav A, Rumbell T, Hof PR, Weaver CM: Functional consquences of age-related morphologic changes to pyramidal neurons of the rhesus monkey prefrontal cortex. J Comput Neurosci. 2014, Dec 20 [Epub ahead of print]Google Scholar
  3. Rumbell T, Draguljić D, Luebke J, Hof P, Weaver CM: Automatic fitness function selection for compartmental model optimization. BMC Neuroscience. 2014, 15 (Suppl 1): O5-PubMed CentralView ArticleGoogle Scholar
  4. Traub RD, Gloveli T, Whittington MA: Fast rhythmic bursting can be induced in layer 2/3 cortical neurons by enhancing persistent Na+ conductance or by blocking BK channels. J Neurophysiol. 2003, 89: 909-921.PubMedView ArticleGoogle Scholar
  5. Sivagnanam S, Majumdar A, Yoshimoto K, Astakhov V, Bandrowski A, Martone ME, Carnevale NT: Introducing the Neuroscience Gateway. IWSG, CEUR Workshop Proceedings. 2013, [http://CEUR-WS.org]Google Scholar

Copyright

© Rumbell et al. 2015

This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Advertisement