LGMD impedance (Dewell & Gabbiani 2019)

 Download zip file 
Help downloading and running models
Accession:256024
"How neurons filter and integrate their complex patterns of synaptic inputs is central to their role in neural information processing . Synaptic filtering and integration are shaped by the frequency-dependent neuronal membrane impedance. Using single and dual dendritic recordings in vivo, pharmacology, and computational modeling, we characterized the membrane impedance of a collision detection neuron in the grasshopper, Schistocerca americana. This neuron, the lobula giant movement detector (LGMD), exhibits consistent impedance properties across frequencies and membrane potentials. Two common active conductances gH and gM, mediated respectively by hyperpolarization-activated cyclic nucleotide gated (HCN) channels and by muscarine sensitive M-type K+ channels, promote broadband integration with high temporal precision over the LGMD's natural range of membrane potentials and synaptic input frequencies. Additionally, we found that a model based on the LGMD's branching morphology increased the gain and decreased the delay associated with the mapping of synaptic input currents to membrane potential. More generally, this was true for a wide range of model neuron morphologies, including those of neocortical pyramidal neurons and cerebellar Purkinje cells. These findings show the unexpected role played by two widespread active conductances and by dendritic morphology in shaping synaptic integration."
Reference:
1 . Dewell RB, Gabbiani F (2019) Active membrane conductances and morphology of a collision detection neuron broaden its impedance profile and improve discrimination of input synchrony. J Neurophysiol [PubMed]
Model Information (Click on a link to find other models with that property)
Model Type: Neuron or other electrically excitable cell; Dendrite;
Brain Region(s)/Organism:
Cell Type(s): Locust Lobula Giant Movement Detector (LGMD) neuron;
Channel(s): I h; I M;
Gap Junctions:
Receptor(s):
Gene(s):
Transmitter(s):
Simulation Environment: NEURON;
Model Concept(s): Active Dendrites; Detailed Neuronal Models; Synaptic Integration; Membrane Properties;
Implementer(s): Dewell, Richard Burkett [dewell at bcm.edu];
Search NeuronDB for information about:  I M; I h; 
/*
 simulations testing a simple Rall neuron with branching dendrites ability to discriminate
 	synaptic input synchrony.
 simulations adapted from Migliore et. al 2004 (ModelDB: number 32992)

-RBD
*/


load_file("Rall.hoc")

Raxial = 150	// axial resistivity (Ohm-cm)
Cm = 1			// membrane capacitance (µF/cm2)
Gm = 1e-4		// membrane conductance (mS/cm2)

kam_Na=0.066
Aam_Na=20
dam_Na = 0
kbm_Na=0.06
Abm_Na=5.3
dbm_Na = -31
kah_Na=0.13
Aah_Na=2.2
dah_Na=-59
kbh_Na=0.17
Abh_Na=4.5
dbh_Na=-29

taumax_M = 25
taumin_M = 4
s1_M=11
s2_M=-10.0

e_h=-41
el = -65

zn_HH_Kdr=9.0
t2_HH_Kdr=0.3
vhalf_HH_Kdr = -36

proc ginit() {

	soma {
		cm=Cm
		insert Na
		gmax_Na = 0.18
		insert HH_Kdr
		gmax_HH_Kdr = 0.06
		t1_HH_Kdr=110
		insert M
		gmax_M = 2.2e-4
		vhalf_M = -46
		if (ismembrane("pas")) g_pas=0
	}
	forsec "Dend" {
		if (gh==1) {
			insert h
			gmax_h = 6.0e-5
			taumax_h = 300
			insert M
			gmax_M = 1.59e-4
			insert pas
			g_pas=0
		} else {
			uninsert h
			uninsert M
		}
	}

	finitialize(v_init)
	
	forsec "Dend" {
		Ra=Raxial
		cm=Cm
		if (Zstate==0) {
			uninsert Lpas2
// 			insert pas
			g_pas = Gm - Rm( 0, 1 )
			e_pas = el
		} else {
			uninsert pas
			insert Lpas2
			g0_Lpas2 = Gm - Rm( 0, 1 )
			e_Lpas2=el
			pl_Lpas2 = 0.95
			L_Lpas2 = 60
		}
	}
}

// ************* -----------------

cvode_active(1)

strdef label
synint=100	// synaptic interval (ms)
reps = 10	// number of times to repeat (each has a different randomization)
ii = 0		// current rep
nsyn =100	// number of synapses
weight=0.2	// synaptic weight (nS)
gh = 1		// whether dendrites have active conductances gh and gM

rngseed=764	// random gen seed
nfrac = 0	// noise fraction of synapse timing [0,1](1 is random, 0 is synchronous)
nfd = 10	// number of noise levels (plus 1 with no noise)
Zstate = 0	// impeedance state (0 = regular leak; 1 = inductive leak)
SaveZ = 0	// whether to save impedance properties
Zall = 1	// whether to measure impedance of all sections (1), or just potenital input sites (0)
tstop=200	// end time of simulation (ms)

// ginit()

objref nc[nsyn], grph[2], spkrec,timevec, rsyn[nsyn], s[nsyn], rc, rd

use_mcell_ran4()
lowindex = mcell_ran4_init()
rc = new Random()
rc.uniform(0,nDend-1)
rd = new Random()
rd.uniform(0,1)

// create synapse objects
for i=0, nsyn-1 {
	rsyn[i] = new Exp2Syn(0.5)
	rsyn[i].e=0
	rsyn[i].tau1 = 0.3
	rsyn[i].tau2 = 3
}

timevec = new Vector()
spkrec = new NetCon(&v(0.5), nil)
spkrec.record("spkproc()")

// procedure to update spike vector with each spike
proc spkproc() {
	stoprun = 1
	timevec.append(t,gh,Zstate,nfrac*synint,rngseed,ii)
	continuerun(t+1e-6)
}

// set new set of random times for synapses
proc setsyntime() {
	for i=0, nsyn-1 {
		s[i] = new NetStim(0.5)
		s[i].interval=synint// mean time between inputs
		s[i].number = 1		// mean number of inputs (at each location)
		s[i].start = 20		// time of first input
		s[i].noise=nfrac		// randomness of timing (0 is not random, 1 is jitter = interval)
		s[i].seed(rngseed+3)	// seed for rng
		nc[i] = new NetCon(s[i],rsyn[i],0,0,weight*1e-3)
// 		nc = new NetCon(source, target (pnt), threshold (mV), delay (ms), weight)
	}
}

// set new set of random locations for synapses
proc setsynloc() {local sect, seg

	rc.MCellRan4(rngseed+1)
	rd.MCellRan4(rngseed+2)
	for i=0, nsyn-1 {
		sect=int(rc.repick())
		seg=rd.repick()
		Dendrite[sect] { rsyn[i].loc(seg) }
	}
}

// create plot to display simulations
grph[0] = new Graph(0)
graphList[0].append(grph[0])
{grph[0].view(0, -70, tstop, 70, 950, 400, 800, 350)}
grph[0].addvar("v(0.5)",1,1)
grph[0].begin()
grph[0].family(1)

// grph[1] = new Graph(0)
// graphList[0].append(grph[1])
// {grph[1].view(0, 0, tstop, 2e-4, 100, 400, 800, 350)}
// grph[1].begin()
// grph[1].family(1)


// run simulations
proc runsynchro() {localobj fobj

	for ii=1,reps {
// 		active model
		Zstate=0
		gh=1
		ginit()
		grph[0].color(4)
		label = "active_dend"
		grph[0].label(0.65,0.75,label)
// 		grph[1].addvar("Dendrite[0].g_pas(0.5)",1,1)
		runc()
		
// 		passive model
		gh=0
		Zstate=0
		ginit()
		grph[0].color(3)
		label = "passive_dend"
		grph[0].label(0.65,0.8,label)
		runc()
		
// 		inductive model
		gh=0
		Zstate=1
		ginit()
// 		grph[1].addvar("Dendrite[0].g_Lpas2(0.5)",1,1)
		grph[0].color(2)
		label = "passive_dend_Zpas"
		grph[0].label(0.65,0.7,label)
		runc()
		
		rngseed+=4
	}
	

	// save simulation data
	nspkval = timevec.size()
	sprint( datafile, "%s/%s_synchro_spks_%g_%g", DATADIR, model.s, rngseed,reps)
	fobj = new File()
	fobj.wopen(datafile)
	fobj.printf("Data is spike time, gh, Z state, noise tau (ms), rng seed, rep\n")
	fobj.printf("%g spikes recorded\n", nspkval/6)
	fobj.printf("%g values\n", nspkval)
	n = timevec.printf(fobj, "%-1.8g\t")
	fobj.close()
	
}


proc runc() { local n,i

	setsynloc()

	for n=0,nfd {
		nfrac = 1.0/nfd*n
		setsyntime()
		
		run()
	}
	if (SaveZ==1 && ii==1) {
		sprint(label, "%s%g_Z_%s",model.s, nBranch, label)
		if (Zall) {
			SaveImpedanceProfile( label, "", 50, 0, 0.2, 1, 1)
		} else {
			SaveImpedanceProfile( label, distlist, 50, 0, 0.2, 1, 1)
		}
	}

}


runsynchro()

Loading data, please wait...