1. Computational and Systems Biology
  2. Neuroscience

T2N as a new tool for robust electrophysiological modeling demonstrated for mature and adult-born dentate granule cells

  1. Marcel Beining  Is a corresponding author
  2. Lucas Alberto Mongiat
  3. Stephan Wolfgang Schwarzacher
  4. Hermann Cuntz  Is a corresponding author
  5. Peter Jedlicka
  1. Ernst Strüngmann Institute (ESI) for Neuroscience in Cooperation with Max Planck Society, Germany
  2. Frankfurt Institute for Advanced Studies, Germany
  3. Goethe University, Germany
  4. Universidad Nacional del Comahue-CONICET, Argentina
https://doi.org/10.7554/eLife.26517
Share this article
Cite this article
  1. Marcel Beining
  2. Lucas Alberto Mongiat
  3. Stephan Wolfgang Schwarzacher
  4. Hermann Cuntz
  5. Peter Jedlicka
(2017)
T2N as a new tool for robust electrophysiological modeling demonstrated for mature and adult-born dentate granule cells
eLife 6:e26517.
https://doi.org/10.7554/eLife.26517
  2. Download BibTeX
  3. Download .RIS
10 figures, 3 tables and 2 additional files

Figures

Figure 1
Download asset Open asset
TREES-to-NEURON (T2N) interface linking compartmental modeling environment NEURON with morphology modeling and analysis tools of Matlab and TREES toolbox.

T2N enables fast and simple incorporation of many diverse morphologies in compartmental simulations facilitating the search for morphologically robust biophysical models. (A) Illustration of T2N …

https://doi.org/10.7554/eLife.26517.002
Figure 2
Download asset Open asset
T2N supports incorporation of realistic ion channels and synthetic morphologies.

(A) Ion channel composition of the mouse dentate granule cell (GC) model. Left: Passive and active ion channels with their specific distribution in six different regions: outer molecular layer …

https://doi.org/10.7554/eLife.26517.003
Figure 3 with 4 supplements
Download asset Open asset
Passive and active properties of the mature mouse GC model.

Comparison of electrophysiological features between experimental data (left column, grayish colors) (Mongiat et al., 2009), GC model with reconstructed morphologies (middle column, blueish colors) …

https://doi.org/10.7554/eLife.26517.005
Figure 3—figure supplement 1
Download asset Open asset
Performance of a widely used GC model with reconstructed and synthetic mouse morphologies.

This figure is analogous to Figure 3 (experimental data in left column, grayish colors) but in order to compare our model’s robustness to standard mature GC models, the biophysical model of Aradi …

https://doi.org/10.7554/eLife.26517.006
Figure 3—figure supplement 2
Download asset Open asset
Influence of morphology on electrophysiological properties in the mature mouse GC model.

Influence of dendritic (left column) and somatic (right column) surface size on electrophysiological parameters (fAHP, AP width, AP threshold and number of APs) in the model with reconstructed (blue …

https://doi.org/10.7554/eLife.26517.007
Figure 3—figure supplement 3
Download asset Open asset
Current dynamics during voltage clamp in mature mice GCs.

Currents measured during a highly hyperpolarized voltage step (−120 mV) in the experiment (left) and the models with reconstructed (middle) and synthetic (right) morphologies. The slowly activating …

https://doi.org/10.7554/eLife.26517.008
Figure 3—figure supplement 4
Download asset Open asset
Maximal rate of voltage change during an AP in the mature mouse GC model.

The maximal voltage deflection during a spike shows a sudden jump and then slow decay in the experiment when current amplitudes are increased in mature (upper row, left) and young GCs (lower row, …

https://doi.org/10.7554/eLife.26517.009
Figure 4 with 1 supplement
Download asset Open asset
Mature rat GC model.

Comparison of electrophysiological features between GC model with reconstructed morphologies (left column, blueish colors) and GC model with synthetic morphologies (right column, greenish colors) as …

https://doi.org/10.7554/eLife.26517.011
Figure 4—figure supplement 1
Download asset Open asset
Performance of the classical GC model with reconstructed and synthetic rat morphologies.

This figure is analogous to Figure 4 but in order to compare our model’s robustness to standard GC models, the biophysical model of Aradi and Holmes (Aradi and Holmes, 1999) was used here with …

https://doi.org/10.7554/eLife.26517.012
Figure 5 with 1 supplement
Download asset Open asset
Backpropagating action potentials (bAPs) in mature mouse and rat GC models.

bAP characteristics at 33°C (experiment and simulation), elicited in the soma by a brief current injection. Inset: Exemplary rat and mouse GC morphology with local maximum voltage amplitudes. (A) …

https://doi.org/10.7554/eLife.26517.013
Figure 5—figure supplement 1
Download asset Open asset
Backpropagating action potentials (bAPs) in the classical GC model.

bAP characteristics at 33°C (experiment and simulation), elicited in the soma by a brief current injection. Inset: Exemplary rat and mouse GC morphology with local maximum voltage amplitudes. Note …

https://doi.org/10.7554/eLife.26517.014
Figure 6 with 2 supplements
Download asset Open asset
Dependence of the model on specific channels and parameters.

(A) Sensitivity matrix showing the relative change (color-coded) in electrophysiological parameters (y-axis) in the mature rat GC model following a 50% reduction in ion channel densities or other …

https://doi.org/10.7554/eLife.26517.015
Figure 6—figure supplement 1
Download asset Open asset
Sensitivity analysis for a doubling of parameter values in the mature rat GC model.

Sensitivity matrix analogous to Figure 6A but with increased (doubled) instead of reduced channel densities or parameters except for the cases marked with an asterix (*): the reversal potential of …

https://doi.org/10.7554/eLife.26517.016
Figure 6—figure supplement 2
Download asset Open asset
Test for resonance in the rat GC model.

GCs were injected with oscillating currents of increasing frequency to calculate their impedance. The graph shows experimental (solid curves, human GCs (Stegen et al., 2012)) and simulation (dashed …

https://doi.org/10.7554/eLife.26517.017
Figure 7
Download asset Open asset
Model of young adult-born granule cells (abGCs) in mice.

Panels are analogous to Figure 3, with comparison of electrophysiological features between experimental data (left column, grayish colors), GC model with reconstructed morphologies (middle column, …

https://doi.org/10.7554/eLife.26517.018
Figure 8
Download asset Open asset
Synaptic integration in young abGCs vs. mature GCs.

(A) Left: Scheme of the simulation configuration with 15 synapses distributed in the MML and 15 in the OML. Middle: All synapses are activated synchronously at 40 Hz. Note that young abGCs (middle …

https://doi.org/10.7554/eLife.26517.020
Appendix 2—figure 1
Download asset Open asset
Overview of the ion channel activation and inactivation kinetics in the GC model.

The illustrated voltage-dependent kinetics were automatically calculated and plotted with a function of the T2N package, which applied voltage step protocols to a single compartment comprising only …

https://doi.org/10.7554/eLife.26517.025
Appendix 2—figure 2
Download asset Open asset
Overview of the ion channel activation and inactivation kinetics in the GC model (continued from Appendix 2—figure 1).
https://doi.org/10.7554/eLife.26517.026

Tables

Table 1
Summary of all ion channel models and densities implemented in the mouse mature GC model.

Categorial values of the ion channel expression profiles: 0 = not existent or very weak, 1 = weak, 2 = moderate, 3 = strong. Conductances [mS/cm²] for each ion channel used in the model are given in …

https://doi.org/10.7554/eLife.26517.004
NameSomaAxonAISGCLIMLMMLOMLReferenceIon channel model
Nav 1.1
Nav 1.2
Nav 1.6		3
0		0
3		0
3		0
0		0
0		0
0		0
0		(Westenbroek et al., 1989; Schmidt-Hieber and Bischofberger, 2010)8-state model from (Schmidt-Hieber and Bischofberger, 2010). Inact. modified according to (Rush et al., 2005; Schmidt-Hieber and Bischofberger, 2010) (see text)
0130000(Kress et al., 2010; Schmidt-Hieber and Bischofberger, 2010)
88.12888.1280518.400----
K2Ps (passive)3112222(Lesage et al., 1997; Hervieu et al., 2001; Talley et al., 2001; Gabriel et al., 2002; Aller and Wisden, 2008; Yarishkin et al., 2014)
0.0140.0070.0070.0140.0140.0140.014
Kir 2.x3112222(Karschin et al., 1996; Miyashita and Kubo, 1997; Stonehouse et al., 1999; Prüss et al., 2003)6-state model, modification see Appendix 2.
0.14160.06740.06740.14160.14160.14160.1416
HCN1-30000222(Notomi and Shigemoto, 2004)2-state model, from (Stegen et al., 2012); activation −10 mV, added cAMP-sens. and slow comp. of act.
----0.0040.0040.004
Kv 1.10330000(Rhodes et al., 1997; Grosse et al., 2000; Monaghan et al., 2001)nh model from (Christie et al., 1989)
-0.250.25----
Kv 1.40330000(Rhodes et al., 1997; Cooper et al., 1998; Grosse et al., 2000; Monaghan et al., 2001)n4h model from (Wissmann et al., 2003)
-11----
Kv 2.13000000(Rhodes et al., 1997; Murakoshi and Trimmer, 1999)mh model, fitted using (VanDongen et al., 1990; Kramer et al., 1998; Kerschensteiner and Stocker, 1999; McCrossan et al., 2003; Gordon et al., 2006)
7.09------
Kv 3.3/3.40230000(Weiser et al., 1994; Chang et al., 2007)mh model, fitted using (Rudy et al., 1991; Schröter et al., 1991; Rettig et al., 1992; Miera et al., 1992; Riazanski et al., 2001; Desai et al., 2008)
-7.656230.7813----
Kv 4.2/3
+KChIP/DPP6		0001233(Rhodes et al., 2004; Zagha et al., 2005; Menegola and Trimmer, 2006)13-state model from (Barghaan et al., 2008); activation −20 mV according to (Barghaan et al., 2008; Figure S1A) and (Jerng et al., 1999; An et al., 2000; Bähring et al., 2001; Patel et al., 2004; Jerng et al., 2005; Rüschenschmidt et al., 2006; Kaulin et al., 2008; Kim et al., 2008)
---2.17504.354.354.35
Kv 7.2/3 (KCNQ2 and 3)0230000(Cooper et al., 2001; Klinger et al., 2011; Martinello et al., 2015)mh model from (Mateos-Aparicio et al., 2014) (η = 0.5, see Tab. S1 in that publication)
-1.34006.7000----
Cav 1.2 (L-type)3011222(Tippens et al., 2008; Leitch et al., 2009)mh1h2 model from GENESIS (Evans et al., 2013), added Ca2+-dep. inactivation (h2)
0.0200-0.01000.01000.04000.04000.0400
Cav 1.33121222(Tippens et al., 2008; Leitch et al., 2009)mh1h2 model from GENESIS (Evans et al., 2013), added Ca2+-dep. inactivation, modified after (Bell et al., 2001; Koschak et al., 2001)
0.01600.00400.00800.00400.00800.00800.0080
Cav 2.1/2 (N-/P/Q-type)3221111(Day et al., 1996; Chung et al., 2001; Li et al., 2007; Xu et al., 2007; 2010)m²h model from (Fox et al., 1987); set inact. time constant to 100 ms according to (Fox et al., 1987; Huang et al., 2010)
0.30000.05000.05000.05000.05000.05000.0500
Cav 3.2 (T-type)3112222(Craig et al., 1999; McKay et al., 2006; Martinello et al., 2015)8-state model from (Burgess et al., 2002)
0.02200.00800.00800.02200.02200.02200.0220
BK (slo1)
α
αβ		2330000(Knaus et al., 1996; Misonou et al., 2006; Sailer et al., 2006; Kaufmann et al., 2010)Model from (Jaffe et al., 2011); modification see Appendix 2
15.6
3.9		62.4
15.6		62.4
15.6		----
SK20230111(Obermair et al., 2003; Sailer et al., 2004; Maciaszek et al., 2012; Ballesteros-Merino et al., 2014)Model from (Solinas et al., 2007) based on (Hirschberg et al., 1998; 1999)
0.0010.0130.0830.0020.0040.0040.004
Table 2
Electrophysiology in mature mouse GCs – experiment vs. model.
https://doi.org/10.7554/eLife.26517.010
Intrinsic propertiesExperimentModel reconstr. morphologiesModel synth. morphologies
Rin [MΩ] (@ −82.1 mV)289.5 ± 34.9287.0 ± 14.7279.6 ± 6.9
cm [pF]48.9 ± 5.355.7 ± 2.861.2 ± 1.6
tau [ms]34.0 ± 2.031.4 ± 0.231.6 ± 0.1
Vrest [mV]−92.7 ± 0.5 *−88.7 ± 0.1−88.6 ± 0.0
Ithreshold [pA]47.5 ± 4.552.5 ± 3.750.3 ± 1.6
Vthreshold [mV]−46.3 ± 1.6 *−44.9 ± 0.3−43.8 ± 0.2
AP amplitude [mV]95.6 ± 2.196.3 ± 2.997.7 ± 1.7
AP width [ms]1.03 ± 0.021.00 ± 0.040.93 ± 0.02
fAHP [mV]15.7 ± 1.417.5 ± 1.715.8 ± 0.8
Interspike interval [ms]36.3 ± 4.936.2 ± 3.234.5 ± 1.1
Max. spike slope [V/s]450.1 ± 23.7428.0 ± 39.5519.7 ± 24.9
gKir [nS]5.46 ± 1.315.90 ± 0.895.97 ± 0.6

  1. *after subtraction of a calculated liquid junction potential of 12.1 mV.

Table 3
Ion channels or currents that were reported to be less expressed in immature GCs and were downregulated in the young GC model
https://doi.org/10.7554/eLife.26517.019
Channel nameCell type and ReferenceDownregulation in the model [%]
Kir 2.xYoung adult-born GCs (Mongiat et al., 2009)73
Kv1.4Young postnatal GCs (Maletic-Savatic et al., 1995; Guan et al., 2011)0
Kv 2.1Young postnatal GCs (Maletic-Savatic et al., 1995; Antonucci et al., 2001; Guan et al., 2011)50
Kv3.4Young postnatal GCs (Riazanski et al., 2001)0
Kv4.2/4.3
+KChIP/DPP6		Young postnatal GCs (Maletic-Savatic et al., 1995; Riazanski et al., 2001)50
Kv 7.2 and 7.3 (KCNQ2 and 3)Young postnatal GCs (Tinel et al., 1998; Smith et al., 2001; Geiger et al., 2006; Safiulina et al., 2008)50
Nav1.2/6Young postnatal GCs (Liu et al., 1996; Pedroni et al., 2014)25
Cav1.2Young postnatal GCs (Jones et al., 1997)0
Cav1.3 (L-type)Young postnatal GCs (Kramer et al., 2012)50
BK-α/BK-β4Young postnatal GCs (MacDonald et al., 2006; Xu et al., 2015)40/100

Additional files

Supplementary file 1

Analogous to Table 2, this table compares electrophysiological properties of experimental data and simulations performed with the biophysical model of Aradi and Holmes (Aradi and Holmes, 1999) and reconstructed (middle column) or synthetic (right column) rat morphologies.

https://doi.org/10.7554/eLife.26517.021
Transparent reporting form
https://doi.org/10.7554/eLife.26517.022

Download links