Dentate gyrus network model pattern separation and granule cell scaling in epilepsy (Yim et al 2015)

 Download zip file 
Help downloading and running models
Accession:185355
The dentate gyrus (DG) is thought to enable efficient hippocampal memory acquisition via pattern separation. With patterns defined as spatiotemporally distributed action potential sequences, the principal DG output neurons (granule cells, GCs), presumably sparsen and separate similar input patterns from the perforant path (PP). In electrophysiological experiments, we have demonstrated that during temporal lobe epilepsy (TLE), GCs downscale their excitability by transcriptional upregulation of ‘leak’ channels. Here we studied whether this cell type-specific intrinsic plasticity is in a position to homeostatically adjust DG network function. We modified an established conductance-based computer model of the DG network such that it realizes a spatiotemporal pattern separation task, and quantified its performance with and without the experimentally constrained leaky GC phenotype. ...
Reference:
1 . Yim MY, Hanuschkin A, Wolfart J (2015) Intrinsic rescaling of granule cells restores pattern separation ability of a dentate gyrus network model during epileptic hyperexcitability. Hippocampus 25:297-308 [PubMed]
Citations  Citation Browser
Model Information (Click on a link to find other models with that property)
Model Type: Realistic Network; Neuron or other electrically excitable cell;
Brain Region(s)/Organism: Dentate gyrus;
Cell Type(s): Dentate gyrus granule GLU cell; Dentate gyrus mossy cell; Dentate gyrus basket cell; Dentate gyrus hilar cell; Dentate gyrus MOPP cell;
Channel(s): I Chloride; I K,leak; I Cl, leak; Kir; Kir2 leak;
Gap Junctions:
Receptor(s): GabaA; AMPA;
Gene(s): IRK; Kir2.1 KCNJ2; Kir2.2 KCNJ12; Kir2.3 KCNJ4; Kir2.4 KCNJ14;
Transmitter(s): Gaba; Glutamate;
Simulation Environment: NEURON;
Model Concept(s): Activity Patterns; Spatio-temporal Activity Patterns; Intrinsic plasticity; Pathophysiology; Epilepsy; Homeostasis; Pattern Separation;
Implementer(s): Yim, Man Yi [manyi.yim at googlemail.com]; Hanuschkin, Alexander ; Wolfart, Jakob ;
Search NeuronDB for information about:  Dentate gyrus granule GLU cell; GabaA; AMPA; I Chloride; I K,leak; I Cl, leak; Kir; Kir2 leak; Gaba; Glutamate;
TITLE BK channel (big conductance, calcium-activated potassium channel)
COMMENT

Original Mod File:
Original name 'CaBK.mod'
Santhakumar V, Aradi I, Soltesz I (2005) J Neurophysiol 93:437-53 
https://senselab.med.yale.edu/modeldb/showModel.cshtml?model=51781&file=/dentategyrusnet2005/CaBK.mod
Original name 'cagk.mod'
Migliore M, Cook EP, Jaffe DB, Turner DA, Johnston D (1995) J Neurophysiol 73:1157-68 
https://senselab.med.yale.edu/modeldb/showModel.cshtml?model=3263&file=/ca3_db/cagk.mod

Current version by A. Hanuschkin <AH, 2011> for:
Yim MY, Hanuschkin A, Wolfart J (2015) Hippocampus 25:297-308.
http://onlinelibrary.wiley.com/doi/10.1002/hipo.22373/abstract

Further Mod File history:
Moczydlowski and Latorre (1983) J. Gen. Physiol. 82

ENDCOMMENT


UNITS {
	(molar) = (1/liter)
}

UNITS {
	(mV) =	(millivolt)
	(mA) =	(milliamp)
	(mM) =	(millimolar)
}


NEURON {
	SUFFIX bk
	USEION nca READ ncai VALENCE 2
	USEION lca READ lcai VALENCE 2
	USEION tca READ tcai VALENCE 2
	USEION k READ ek WRITE ik
	RANGE gkbar,gkca, ik
	GLOBAL oinf, otau
}

UNITS {
	FARADAY = (faraday)  (kilocoulombs)
	R = 8.313424 (joule/degC)
}

PARAMETER {
	celsius		(degC)
	v		(mV)
	gkbar=.01	(mho/cm2)	: Maximum Permeability
	cai = 5.e-5	(mM)
	ek		(mV)

	d1 = .84
	d2 = 1.
	k1 = .48e-3	(mM)
	k2 = .13e-6	(mM)
	abar = .28	(/ms)
	bbar = .48	(/ms)
        st=1            (1)
	lcai		(mV)
	ncai		(mV)
	tcai		(mV)
}

ASSIGNED {
	ik		(mA/cm2)
	oinf
	otau		(ms)
        gkca          	(mho/cm2)
}

INITIAL {
	cai= ncai + lcai + tcai
        rate(v,cai)
        o=oinf
}

STATE {	o }		: fraction of open channels

BREAKPOINT {
	SOLVE state METHOD cnexp
	gkca = gkbar*o^st
	ik = gkca*(v - ek)
}

DERIVATIVE state {	: exact when v held constant; integrates over dt step
: Please note that cai was not assiged here in the original Santhakumar (2005) version (which we used). It should be cai = ncai + lcai + tcai, as noted by
: Morgan RJ, Santhakumar V, Soltesz I (2007) Prog Brain Res 163:639-58
: See
: https://senselab.med.yale.edu/modeldb/showModel.cshtml?model=124513&file=/dentate_gyrus/CaBK.mod
	rate(v, cai)
	o' = (oinf - o)/otau
}

FUNCTION alp(v (mV), c (mM)) (1/ms) { :callable from hoc
	alp = c*abar/(c + exp1(k1,d1,v))
}

FUNCTION bet(v (mV), c (mM)) (1/ms) { :callable from hoc
	bet = bbar/(1 + c/exp1(k2,d2,v))
}

FUNCTION exp1(k (mM), d, v (mV)) (mM) { :callable from hoc
	exp1 = k*exp(-2*d*FARADAY*v/R/(273.15 + celsius))
}

PROCEDURE rate(v (mV), c (mM)) { :callable from hoc
	LOCAL a
	a = alp(v,c)
	otau = 1/(a + bet(v, c))
	oinf = a*otau
}