Linear vs non-linear integration in CA1 oblique dendrites (Gómez González et al. 2011)

 Download zip file   Auto-launch 
Help downloading and running models
Accession:144450
The hippocampus in well known for its role in learning and memory processes. The CA1 region is the output of the hippocampal formation and pyramidal neurons in this region are the elementary units responsible for the processing and transfer of information to the cortex. Using this detailed single neuron model, it is investigated the conditions under which individual CA1 pyramidal neurons process incoming information in a complex (non-linear) as opposed to a passive (linear) manner. This detailed compartmental model of a CA1 pyramidal neuron is based on one described previously (Poirazi, 2003). The model was adapted to five different reconstructed morphologies for this study, and slightly modified to fit the experimental data of (Losonczy, 2006), and to incorporate evidence in pyramidal neurons for the non-saturation of NMDA receptor-mediated conductances by single glutamate pulses. We first replicate the main findings of (Losonczy, 2006), including the very brief window for nonlinear integration using single-pulse stimuli. We then show that double-pulse stimuli increase a CA1 pyramidal neuron’s tolerance for input asynchrony by at last an order of magnitude. Therefore, it is shown using this model, that the time window for nonlinear integration is extended by more than an order of magnitude when inputs are short bursts as opposed to single spikes.
Reference:
1 . Gómez González JF, Mel BW, Poirazi P (2011) Distinguishing Linear vs. Non-Linear Integration in CA1 Radial Oblique Dendrites: It's about Time. Front Comput Neurosci 5:44 [PubMed]
Citations  Citation Browser
Model Information (Click on a link to find other models with that property)
Model Type: Neuron or other electrically excitable cell;
Brain Region(s)/Organism:
Cell Type(s): Hippocampus CA1 pyramidal GLU cell;
Channel(s): I Na,p; I CAN; I Sodium; I Calcium; I Potassium; I_AHP;
Gap Junctions:
Receptor(s): NMDA;
Gene(s):
Transmitter(s):
Simulation Environment: NEURON;
Model Concept(s): Active Dendrites; Detailed Neuronal Models; Synaptic Integration;
Implementer(s):
Search NeuronDB for information about:  Hippocampus CA1 pyramidal GLU cell; NMDA; I Na,p; I CAN; I Sodium; I Calcium; I Potassium; I_AHP;
/
CA1_Gomez_2011
mechanism
x86_64
ampa.mod *
cad.mod
cal.mod
calH.mod
can.mod *
car.mod
cat.mod
d3.mod *
gabaa.mod *
gabab.mod
h.mod
hha_old.mod
hha2.mod
ican.mod
ipulse1.mod *
ipulse2.mod *
kadist.mod
kaprox.mod
kca.mod
kct.mod
KdBG.mod
km.mod
nap.mod *
netstim.mod *
netstimmm.mod *
nmda.mod *
NMDAb.mod
somacar.mod
                            
TITLE simple NMDA receptors

COMMENT
-----------------------------------------------------------------------------

Essentially the same as /examples/nrniv/netcon/ampa.mod in the NEURON
distribution - i.e. Alain Destexhe's simple AMPA model - but with
different binding and unbinding rates and with a magnesium block.
Modified by Andrew Davison, The Babraham Institute, May 2000


	Simple model for glutamate AMPA receptors
	=========================================

  - FIRST-ORDER KINETICS, FIT TO WHOLE-CELL RECORDINGS

    Whole-cell recorded postsynaptic currents mediated by AMPA/Kainate
    receptors (Xiang et al., J. Neurophysiol. 71: 2552-2556, 1994) were used
    to estimate the parameters of the present model; the fit was performed
    using a simplex algorithm (see Destexhe et al., J. Computational Neurosci.
    1: 195-230, 1994).

  - SHORT PULSES OF TRANSMITTER (0.3 ms, 0.5 mM)

    The simplified model was obtained from a detailed synaptic model that 
    included the release of transmitter in adjacent terminals, its lateral 
    diffusion and uptake, and its binding on postsynaptic receptors (Destexhe
    and Sejnowski, 1995).  Short pulses of transmitter with first-order
    kinetics were found to be the best fast alternative to represent the more
    detailed models.

  - ANALYTIC EXPRESSION

    The first-order model can be solved analytically, leading to a very fast
    mechanism for simulating synapses, since no differential equation must be
    solved (see references below).



References

   Destexhe, A., Mainen, Z.F. and Sejnowski, T.J.  An efficient method for
   computing synaptic conductances based on a kinetic model of receptor binding
   Neural Computation 6: 10-14, 1994.  

   Destexhe, A., Mainen, Z.F. and Sejnowski, T.J. Synthesis of models for
   excitable membranes, synaptic transmission and neuromodulation using a 
   common kinetic formalism, Journal of Computational Neuroscience 1: 
   195-230, 1994.

-----------------------------------------------------------------------------
ENDCOMMENT



NEURON {
	POINT_PROCESS NMDAb
	RANGE g, Alpha, Beta, e, gmax
	NONSPECIFIC_CURRENT  iNMDA            : i
	GLOBAL Cdur, mg, Cmax
}
UNITS {
	(nA) = (nanoamp)
	(mV) = (millivolt)
	(umho) = (micromho)
	(mM) = (milli/liter)
}

PARAMETER {
	Cmax	= 1	 (mM)           : max transmitter concentration
:	Cdur	= 30	 (ms)		: transmitter duration (rising phase)
	Cdur	= 1.1	 (ms)		: transmitter duration (rising phase)
:	Alpha	= 0.072	 (/ms /mM)	: forward (binding) rate
	Alpha	= 10	 (/ms /mM)	: forward (binding) rate
:	Beta	= 0.0066 (/ms)		: backward (unbinding) rate
	Beta	= 0.0125 (/ms)		: backward (unbinding) rate
:	e	= 45	 (mV)		: reversal potential
	e	= 0	 (mV)		: reversal potential
        mg      = 1      (mM)           : external magnesium concentration

}


ASSIGNED {
	v		(mV)		: postsynaptic voltage
	iNMDA 		(nA)		: current = g*(v - e)
	g 		(umho)		: conductance
	Rinf				: steady state channels open
	Rtau		(ms)		: time constant of channel binding
	synon
        B 
	gmax                              : magnesium block

}

STATE {Ron Roff}

INITIAL {
	Rinf = Cmax*Alpha / (Cmax*Alpha + Beta)
	Rtau = 1 / (Cmax*Alpha + Beta)
	synon = 0
}

BREAKPOINT {
	SOLVE release METHOD cnexp
        B = mgblock(v)
	g = (Ron + Roff)*1(umho) * B
	iNMDA = g*(v - e)

}

DERIVATIVE release {
	Ron' = (synon*Rinf - Ron)/Rtau
	Roff' = -Beta*Roff
}

FUNCTION mgblock(v(mV)) {
        TABLE 
        DEPEND mg
        FROM -140 TO 80 WITH 1000

        : from Jahr & Stevens

        mgblock = 1 / (1 + exp(0.062 (/mV) * -v) * (mg / 3.57 (mM)))
}

: following supports both saturation from single input and
: summation from multiple inputs
: if spike occurs during CDur then new off time is t + CDur
: ie. transmitter concatenates but does not summate
: Note: automatic initialization of all reference args to 0 except first

			
NET_RECEIVE(weight, on, nspike, r0, t0 (ms)) {
	: flag is an implicit argument of NET_RECEIVE and  normally 0
        if (flag == 0) { : a spike, so turn on if not already in a Cdur pulse
		nspike = nspike + 1
		if (!on) {
			r0 = r0*exp(-Beta*(t - t0))
			t0 = t
			on = 1
			synon = synon + weight
			state_discontinuity(Ron, Ron + r0)
			state_discontinuity(Roff, Roff - r0)
		}
:		 come again in Cdur with flag = current value of nspike
		net_send(Cdur, nspike)
       }
	if (flag == nspike) { : if this associated with last spike then turn off
		r0 = weight*Rinf + (r0 - weight*Rinf)*exp(-(t - t0)/Rtau)
		t0 = t
		synon = synon - weight
		state_discontinuity(Ron, Ron - r0)
		state_discontinuity(Roff, Roff + r0)
		on = 0
	}
gmax = weight
}