Figure 5
gNa density variability between neurons decorrelates spiking in response to fluctuating currents in a computational model across a range of conditions. (A) Spike patterns of neurons with several levels of maximal sodium conductance (gNa) in response to the same time varying noisy current injection with a mean (μ) of 90 pA, standard deviation (σ) of 22 pA, and time constant (τ) of 3 ms. (B) Histograms of the gNa assigned to neurons used in the simulations in C–H, generated from data in Fig. 1. CV-gNa denotes the coefficient of variation in gNa between different cells. Distribution includes 3 values > 180 pS/µm2 not shown on the histogram for visibility. (C) Raster plots of neuronal spike times of 240 neurons generated from fluctuation currents as in A with gNa sampled from distributions shown in B. (D) Resultant mean pairwise correlation (top left), mean phase coherence (top right), firing rate (bottom left), and coefficient of variation in firing rate (bottom right) between neurons as current μ is varied and σ = 22 pA and τ of 3 ms. (E) Results as in D while varying current σ with μ = 90 pA and τ = 3 ms. (F) Results as in D while varying current τ and accompanying variation σ which gives a membrane potential standard deviation of 3.7 mV and μ = 90 pA. (G) Results as in D while varying gNa μ with current μ = 90 pA, σ = 22 pA, and τ = 3 ms. (H) Pairwise correlation between neurons vs their geometric mean rate for 1000 neurons simulated as in C. Black dashed line indicates best fit line between 10 and 30 Hz. Dashed blue line indicates mean correlation over entire CV-gNa = 0.53 population in C. Dashed red line indicates mean level of correlation over entire CV-gNa = 0.1 population in C.
