The Journal of Neuroscience, June 7, 2006, ():
Slow State Transitions of Sustained Neural Oscillations by Activity-Dependent Modulation of Intrinsic Excitability
J. Neurosci. Fröhlich et al.
26: 6153
Supplemental data
Files in this Data Supplement:
- supplemental material
-
Supplementary Figure 1
Membrane voltage traces for depolarizing current step injection into model PY cell for three different current intensities (200%, 150%, and 100%) from top to bottom.
- supplemental material
-
Supplementary Figure 2
Time courses of membrane voltage Vm, calcium-activated potassium current IKCa, high-threshold calcium current ICa, and (in-)activation of ICa for K+o = 6.0 mM. Tonic firing (left column) and slow bursting (right column). Note the different time-scales for the two columns. Bursts consisted of about 8 spikes at the onset, a sustained depolarization with spike inactivation lasting about 100 ms and a pronounced after-hyperpolarization at the end (top panels). Dynamic interplay of high-threshold calcium current ICa and calcium-activated potassium current IKCa caused bursting (middle four panels). In comparison to the tonic firing mode, IKCa was stronger by an order of magnitude and terminated the burst. Similarly, ICa was more pronounced during bursting. Activation variable m and inactivation variable h of the high-threshold calcium conductance GCa are shown in bottom panels. During tonic firing, high firing frequency and lack of pronounced after-hyperpolarization prevented GCa from fully deinactivating (inactivation variable h » 0.34, bottom left subpanel). This contrasted with the bursting regime where the lower burst frequency and the after-hyperpolarization caused more complete deinactivation of GCa between bursts (h » 0.57, bottom right subpanel). As a consequence, the depolarizing effect of ICa was stronger and prevented the membrane voltage from returning to rest after the first spike in the burst. The resulting gradual depolarization after each spike maintained the depolarized state of the burst. The different levels of deinactivation of the high-threshold calcium conductance GCa therefore mediate the bistability between tonic firing and bursting.
- supplemental material
-
Supplementary Figure 3
Robustness analysis to variation of intrinsic conductances (parameter sensitivity analysis). Dependence of single cell hysteresis for transition between slow bursting and fast run on GCa (A), GK(Ca) (B), and GNaP (C) without (top row) and with (bottom row) hyperpolarization-activated depolarizing conductance Gh. The width of the hysteresis greatly depended on the choice of GCa. The hysteresis was robust to variations in GK(Ca). Increasing GNaP shifted the hysteresis to lower values of K+o. In all three cases, introducing Gh = 0.1 mS/cm2 decreased the value of K+o for which the neuron became active and narrowed the width of the hysteresis.
- supplemental material
-
Supplementary Figure 4
Transition from slow bursting to fast run in a PY cell for weak and strong excitatory drive of IN. (A) The inhibitory interneuron follows firing in the PY cell for gAMPA(PY-IN) = 0.010 mS and gNMDA(PY-IN) = 0.014 mS. PY firing is interrupted by IPSCs. PY cells fire in groups of spikes (fast bursts). (B) The IN interneuron is silent during fast run for gAMPA(PY-IN) = 0.007 mS and gNMDA(PY-IN) = 0.008 mS. Fast run in the PY cell is tonic firing at high frequency.
- supplemental material
-
Supplementary Figure 5
Controlling inactivation, h, of high-threshold calcium conductance GCa induced switching between slow bursting and fast run. (A) An unperturbed time-course of a PY cell in a model of small cortical network (top panel). Enforcing inactivation h = 0.53 for 1 sec caused the network to switch to slow bursting (middle panel). Similarly, a transition from slow bursting to fast run was induced by setting h = 0.36 for 1 sec (bottom panel). Horizontal bars delimit intervals where h was held constant. (B) Phase plane plots of switching between fast run and slow bursting (top panel) and vice versa (bottom panel) by controlling inactivation h.
- supplemental material
-
Supplementary Figure 6
Effect of synaptic coupling. PY membrane voltage time course (left column) and time-dependent autocorrelation computed every 100 msec for a window of 400 msec (right column). (A) Transition between slow bursting and fast run for gAMPA(PY-PY) = 0.20 mS, gNMDA(PY-PY) = 0.013 mS, gAMPA(PY-IN) = 0.010 mS, gNMDA(PY-IN) = 0.014 mS, gGABA(IN-PY) = 0.05 mS. (B) Weak excitation and strong inhibition (scaling: PY-PY = 0.8 and IN-PY = 1.2). PY cells remained in fast run. No transition to slow bursting due to decreased excitatory drive. (C) Strong excitation and weak inhibition (scaling: PY-PY = 1.2 and IN-PY = 0.8). Network directly returned to silence after a single epoch of bursting.
