Repetitive stimuli result in use-dependent accumulation of Nav channels in slow-recovering states. (A) The standard paired pulse protocol was used to elicit Nav current with a 5-ms step to 0 mV (pulse 1, P1), with recovery intervals at −80 mV from 0.3 ms to 3 s preceding another 5-ms step (pulse 2, P2) to 0 mV. P1 currents are on the left, while P2 currents on the right followed recovery durations of 1, 3, 10, 20, 50, 100, 200, 500, and 3,000 ms. Horizontal colored lines and colored traces highlight traces following 10, 100, and 3,000 ms. (B) A train of five 5-ms steps to 0 mV was used as the P1 stimulus, with 10-ms intervals between steps to 0. Currents on the right are as in A. Traces at 10 (blue) and 100 (red) ms exhibit marked diminution compared with that in A. Dotted lines reflect the amplitudes of the recoveries for 10, 100, and 3,000 ms from A. (C) Traces are as in B, but for a train of twenty 5-ms steps to 0 mV, with 10 ms between each step in the train. Dotted lines on the right again correspond to fractional recoveries observed in A for 10, 100, and 3,000 ms. (D) Following stimulation with the trains, the standard single inactivation pulse was again applied, as in A, showing that, following a single inactivation step, fast recovery returns to its initial amplitude. (E) Fractional recovery from inactivation is plotted for the cell in A–D, with each curve fit with a two-exponential function with the following values: for initial control recovery, Af = 0.49 ± 0.03, τf = 34.5 ± 3.0 ms, As = 0.52 ± 0.02, and τs = 481.3 ± 52.5; following the 5-pulse train, Af = 0.40 ± 0.03, τf = 59.7 ± 6.0 ms, As = 0.60 ± 0.03, and τs = 515.6 ± 46.1 ms; following a 10-pulse train, Af = 0.31 ± 0.05, τf = 68.4 ± 12.4 ms, As = 0.68 ± 0.05, and τs = 534.8 ± 58.3 ms; following a 20-pulse train, Af = 0.29 ± 0.04, τf = 84.0 ± 12.5 ms, As = 0.73 ± 0.04, and τs = 562.8 ± 44.0 ms; after recovery, Af = 0.47 ± 0.02, τf = 36.2 ± 3.2 ms, As = 0.54 ± 0.23, and τs = 475.3 ± 48.5 ms. (F) Averaged values (±SD) for fractional recovery are plotted for a set of 19 cells (recovery runs were only obtained for 16 of the cells). For the initial control protocol, recovery parameters (mean ± 90% confidence limit) are Af = 0.50 ± 0.02, τf = 19.6 ± 1.4 ms, As = 0.51 ± 0.02, and τs = 404.3 ± 39.4 ms; for the 5-pulse P1 stimulus, Af = 0.30 ± 0.02, τf = 33.6 ± 3.3 ms, As = 0.70 ± 0.02, and τs = 413.6 ± 22.9 ms; for the 10-pulse P1 stimulus, Af = 0.24 ± 0.03, τf = 45.2 ± 6.4 ms, As = 0.75 ± 0.02, and τs = 418.6 ± 24.3 ms; for the 20-pulse P1 stimulus, Af = 0.26 ± 0.04, τf = 67.6 ± 12.0 ms, As = 0.76 ± 0.04, and τs = 493.3 ± 38.0 ms; following return to the single pulse protocol, Af = 0.48 ± 0.02, τf = 20.5 ± 1.7 ms, As = 0.53 ± 0.02, and τs = 436.3 ± 44.4 ms. (G) Mean values (±SD) of the fast recovery amplitude for each condition are plotted along with the best fit values from each individual cell in the set of 19 cells. Control (1P) and recovery (1P(rec)) fast amplitude differed from all other protocols (5P, 10P, 20P) at P = 0.000 (Kolmogorov–Smirnov test). 1P versus 1P(rec): 0.085; 5P versus 10P: 0.049; 5P versus 20P: 0.462; 10P versus 20P: 0.742. (H) Fast and slow time constants (±SD) are plotted for each of the recovery protocols, along with the individual determinations for each cell (small symbols). Kolmogorov–Smirnov P values for comparisons of fast time constant values were as follows: for 1P versus 5P, 0.000; 1P versus 10P, 0.000; 1P versus 20P, 0.000; 1P versus 1P(rec), 0.957; 5P versus 10P, 0.018; 5P versus 20P, 0.000; 10P versus 20P, 0.002; 20P versus. 1P(rec), 0.000. For comparisons of slow time constants: 1P differed from 5P, 10P, and 20P at 0.000; 1P versus 1P(rec), 0.354; 5P versus 10P, 0.116; 10P versus 20P, 0.956; 20P versus 1P(rec), 0.000.