2.3 Voltage protocols
The ”steady-state inactivation” (SSI ) protocol investigated the
availability of channels at different membrane potentials, the ”recovery
from inactivation” (RFI ) protocol investigated the dynamics of
channels regaining their availability after inactivation, and the
state-dependent onset (SDO ) protocol investigated the dynamics
of developing inhibition. These same protocols have been used before
(Lukacs et al., 2018); and they are illustrated in Fig. 5A.
The ”three pulse train” (3PT ) protocol (Fig. 2A and D) was
similar to the one we called by this name before (Lukacs et al., 2018);
but in this study, it was optimized for maximal time resolution. The
protocol is designed to monitor changes in both gating kinetics and
gating equilibrium. For the sake of high time resolution, it uses only a
single interpulse interval: 2 ms (instead of several interpulse
intervals as in RFI) , and only a single membrane potential: -65
mV (instead of several membrane potentials as in the ”steady-state
inactivation” protocol – SSI ). Thus we could record changes at
26 ms resolution (38.5 Hz). In each experiment, we delivered 100
consecutive trains: after 6 initial control trains we perfused riluzole
throughout 10 trains (i.e., for 260 ms), and then monitored washout for
the remaining 84 trains. Perfusion was started and stopped during the 10
ms intervals between two trains, which was amply enough for complete
exchange of the solution with the theta tube system.
Peak amplitudes of evoked currents have been determined after
subtracting capacitive and leakage artifacts. Recordings from a typical
cell throughout a 100 train experiment are shown in Fig. 2B. The 6
control traces are shown in blue, the 10 traces recorded during riluzole
perfusion are in red (dark to light red indicates consecutive traces),
and currents evoked during washout are shown in light gray to black.
Fig. 2C illustrates the same currents after subtraction of artifacts.
Fig. 2D shows the same corrected current traces are shown in sequential
order, together with the voltage protocol, as they were evoked in the
experiment (only the first 29 of the 100 trains, for the sake of
clarity). Peak amplitudes are marked by circles: blue
(1st pulses), red (2nd pulses), and
green (3rd pulses). Connecting blue, red, and green
circles we get different onset and offset characteristics, as shown for
n = 7 cells in Fig. 2E. All amplitudes were normalized to the
1st pulse-evoked amplitude at the start of the
experiment (1st, 2nd, and
3rd pulse-evoked amplitudes each to its own control,
to help comparison of inhibited fractions). Current traces for the same
seven cells, as well as for the nine recorded binding site mutant cells
are shown in supporting information Fig. S1, together with a detailed
discussion of fitted time constants.
A small fraction of channels (11.4 ± 0.1% by the end of the
100th train) underwent slow inactivation during the
test; this was a necessary tradeoff for high temporal resolution. In
order to appropriately calculate the extent of inhibition by riluzole,
we corrected for the slow inactivated fraction: A sum of an exponential
and a linear component was fit to the first 6 and last 3 points of 1st
pulse-evoked current amplitudes in each cell (supporting information
Fig. S1B and H). This fitted function represented the
non-slow-inactivated fraction fnsi(t) of the channel
population. Measured amplitude plots (AM(t) ),
then were transformed into corrected amplitude plots
(AC(t) ) by expressing it as a fraction of
non-slow-inactivated channels: AC(t) =
AM(t) / fnsi(t) . This is illustrated in
supporting information Fig. S1B and C, where panel C shows the same data
as panel B, after correction. Measurement of the extent of inhibition
and determination of onset and offset time constants were performed on
the corrected plots.