Functional roles of Kv1-mediated currents in genetically identified subtypes of pyramidal neurons in layer 5 of mouse somatosensory cortex

Abstract

We used voltage-clamp recordings from somatic outside-out macropatches to determine the amplitude and biophysical properties of putative Kv1-mediated currents in layer 5 pyramidal neurons (PNs) from mice expressing EGFP under the control of promoters for etv1 or glt. We then used whole cell current-clamp recordings and Kv1-specific peptide blockers to test the hypothesis that Kv1 channels differentially regulate action potential (AP) voltage threshold, repolarization rate, and width as well as rheobase and repetitive firing in these two PN types. We found that Kv1-mediated currents make up a similar percentage of whole cell K⁺ current in both cell types, and only minor biophysical differences were observed between PN types or between currents sensitive to different Kv1 blockers. Putative Kv1 currents contributed to AP voltage threshold in both PN types, but AP width and rate of repolarization were only affected in etv1 PNs. Kv1 currents regulate rheobase, delay to the first AP, and firing rate similarly in both cell types, but the frequency-current slope was much more sensitive to Kv1 block in etv1 PNs. In both cell types, Kv1 block shifted the current required to elicit an onset doublet of action potentials to lower currents. Spike frequency adaptation was also affected differently by Kv1 block in the two PN types. Thus, despite similar expression levels and minimal differences in biophysical properties, Kv1 channels differentially regulate APs and repetitive firing in etv1 and glt PNs. This may reflect differences in subcellular localization of channel subtypes or differences in the other K⁺ channels expressed. NEW & NOTEWORTHY In two types of genetically identified layer 5 pyramidal neurons, α-dendrotoxin blocked approximately all of the putative Kv1 current (on average). We used outside-out macropatches and whole cell recordings at 33°C to show that despite similar expression levels and minimal differences in biophysical properties, Kv1 channels differentially regulate action potentials and repetitive firing in etv1 and glt pyramidal neurons. This may reflect differences in subcellular localization of channel subtypes or differences in the other K⁺ channels expressed.

Keywords: potassium channel; repetitive firing; somatosensory cortex.

Fig. 1.

Kv1-mediated currents in etv1 and glt pyramidal neurons. A: family of current traces from glt cell in control solution in response to protocol shown as inset in B (1-s steps). B: current traces from the same cell as in A, except in the presence of 100 nM α-dendrotoxin (DTX) + 30 nM margatoxin (MTX). Inset: voltage protocol for A–C. C: subtracted records (control − DTX+MTX) to show putative Kv1-mediated current. Inset: scale bars for A–C. D: summary data (means ± SE) for %block at peak current at +10 mV by 100 nM DTX, 30 nM MTX, or both toxins combined in etv1 and glt cells. E: summary data (means ± SE) for %block at steady state (500 ms, +10 mV) by 100 nM DTX, 30 nM MTX, or both toxins combined in etv1 and glt cells.

Fig. 2.

Outside-out macropatch currents and their kinetics. A–F: representative current traces of average currents sensitive to α-dendrotoxin (DTX; 100 nM), margatoxin (MTX; 20–40 nM), or DTX+MTX in glt (A–C) and etv1 (D–F) pyramidal neurons (PNs). The protocol shown in A applies to A–F. G: summary data (means ± SE) for activation time constant (τ) at +10 mV. There were no significant differences between PN types. In glt PNs, MTX-sensitive currents activated more rapidly than DTX-sensitive currents (*P < 0.05). H: summary data (means ± SE) for inactivation τ (at +10 mV). There were no significant differences between PN types. In etv1 PNs, DTX-sensitive currents inactivated more rapidly than MTX-sensitive currents (*P < 0.05). I: summary data (means ± SE) for %inactivation (at +10 mV). There were no significant differences between PN types. In etv1 PNs, MTX-sensitive currents inactivated more than DTX-sensitive currents (*P < 0.05).

Fig. 3.

Activation and inactivation time constants (tau). The rise (activation) and decay (inactivation) for the Kv1 blocker-sensitive current were fitted at several voltages (V) using the activation protocol shown in Fig. 4 (for fitting procedure, see Eq. 3 in text). A: activation tau values were not significantly different between etv1 and glt pyramidal neurons (PNs). B: inactivation tau values were not significantly different between etv1 and glt PNs and showed little voltage dependence. Data are means ± SE.

Fig. 4.

Steady-state activation of Kv1-mediated currents. A: current traces sensitive to 100 nM α-dendrotoxin (DTX) + 20 nM margatoxin (MTX) from outside-out macropatch from an etv1 pyramidal neuron (PN). Inset: voltage protocol (500- to 1,000-ms steps). B: steady-state (at 500 ms) activation curve for cell in A. C: summary activation curves (at 500 ms; means ± SE) for current sensitive to Kv1 blockers (combined data: DTX, MTX, and DTX+MTX) for etv1 (black) and glt PNs (gray). D: summary data (means ± SE) for half-activation voltage in etv1 and glt PNs for current sensitive to 100 nM DTX, 20–40 nM MTX, or their combination (DTX+MTX). There were no significant differences. E: summary data (means ± SE) for activation slope for current sensitive to 100 nM DTX, 30–40 nM MTX, or their combination (DTX+MTX). There were no significant differences. F: summary data for half-activation voltage and slope for etv1 vs. glt PNs (all Kv1 blockers combined: DTX, MTX, and DTX+MTX). There were no significant differences. G, conductance; SS, steady state; V, voltage; V_m, membrane potential.

Fig. 5.

Steady-state inactivation. The steady-state inactivation of currents sensitive to 100 nM α-dendrotoxin (DTX) + 30 nM margatoxin (MTX) was compared in outside-out macropatch recordings from 9 etv1 and 10 glt pyramidal neurons (PNs). From a holding potential of −75 mV, the current was inactivated with 5-s steps to voltages between −100 and −20 mV and then measured from a test step to +10 mV (20 s between each prepulse + test step). A: representative traces from a patch from a glt PN. Inset: voltage protocol. B: plot of normalized current vs. prepulse voltage (V) for patches from etv1 (black) and glt (gray) PNs (means ± SE). There were no significant differences between cell types. C: histograms (means ± SE) comparing etv1 and glt for the half-inactivation voltage (V₅₀). There were no significant differences between cell types. For etv1, V₅₀ was −53.8 ± 1.8 mV, and for glt it was −51 ± 1.7 mV. D: histograms (means ± SE) comparing etv1 and glt for half-inactivation slope. For etv1, the slope was 5.1 ± 0.8 mV, and for glt it was 6.1 ± 0.9 mV. There were no significant differences between cell types.

Fig. 6.

Effects of Kv1 channels on the action potential (AP). A: representative traces for single APs elicited by a 5-ms suprathreshold current injection in a etv1 pyramidal neuron (PN) in control solution (Ctl) and in the presence of 100 nM α-dendrotoxin (DTX) + 30 nM margatoxin (MTX). Note slightly hyperpolarized AP voltage threshold and broadening of the AP. B: similar traces for glt PN. Note narrower AP compared with etv1 PN in A and shift in threshold but no change in AP width in DTX+MTX. C: summary histograms (means ± SE) for AP voltage threshold for etv1 and glt PNs in Ctl vs. combined data for DTX (100 nM), MTX (20 nM), and DTX+MTX (all). There was a significant hyperpolarization of AP threshold in both cell types (*P < 0.05). D: summary histograms (means ± SE) for AP width at half-amplitude (half-width) for etv1 and glt PNs in Ctl vs. Kv1 blockers (all = combined data for DTX, MTX, and DTX+MTX). There was significant AP broadening in etv1 PNs but not in glt PNs (*P < 0.05).

Fig. 7.

Effects of Kv1 channels on rheobase and latency to the first action potential (AP). A: representative traces from a glt pyramidal neuron (PN) showing reduced current required to elicit an AP during 500-ms current injection (decreased rheobase) in the presence of 100 nM α-dendrotoxin (DTX) + 30 nM margatoxin (MTX; gray traces) and control (Ctl; black traces). B: summary histograms (means ± SE) showing that Kv1 blockers (all = combined data for DTX, MTX, and DTX+MTX) significantly reduced rheobase in both etv1 and glt PNs (*P < 0.05). C: summary histograms (means ± SE) showing that Kv1 blockers significantly reduced latency to the first AP in response to a 200-pA, 500-ms current injection in both etv1 and glt PNs (*P < 0.05).

Fig. 8.

Onset doublets. A: traces from an etv1 pyramidal neuron (PN) showing regular spiking at onset of response to direct current (DC) injection (300 pA, 2 s) in control solution. Scale bars apply to A–C. B: traces from the same etv1 PN in control solution in response to a larger (400 pA, 2 s) current injection. Note the initial doublet of action potentials (APs) at the onset of the response. C: traces from a glt PN showing regular spiking at the onset of a DC injection (250 pA, 2 s) in control solution (black trace). After 100 nM α-dendrotoxin (DTX) + 30 nM margatoxin (MTX; red), there was an initial doublet of APs at the onset of the current step. D: plot of instantaneous firing frequency [f; 1/interspike interval (ISI)] vs. injected current (I) for the etv1 PN from A and B. Plotted are the first 3 ISIs as well as average firing over the entire 2-s epoch. Note the sharp deviation of the curve for the first ISI vs. the others at currents >300 pA. This deviation was diagnostic for the presence of an initial doublet and served as our working criteria for the presence of a doublet. E: similar plot for a glt PN. As in the etv1 PN, there was a sharp deviation of the first ISI curve above 200 pA, reflecting onset doublet in response to currents >200 pA. F: plot of the percentage of cells showing an onset burst at a given level of injected current using the criteria shown in D and E. In control solution (black), virtually all cells could fire with an onset burst in response to currents >500 pA. After Kv1 block (red), the curve was shifted leftward, indicating a reduction in the current required to initiate an onset doublet. Inset: histogram (means ± SE) indicating that Kv1 block results in a significant reduction in the minimal current to elicit an onset doublet (*P < 0.05). G: plot of instantaneous firing frequency (1/ISI; means ± SE) for the first vs. the second ISI vs. the injected current for a population of 37 etv1 PNs in control (black) and Kv1 blockers (red). Kv1 block increased the frequency for both the first and second intervals but had a greater effect on the first interval, resulting in deviation from the first vs. second ISI at lower currents. H: similar plot and similar effects of Kv1 block for a population of 39 glt PNs. I: plot of the percentage of cells showing an onset burst at a given level of injected current using the criteria shown in D and E. In control solution (black), virtually all cells could fire with an onset burst in response to currents >500 pA. After Kv1 block (red), the curve was shifted leftward. This indicates a reduction in the current required to initiate an onset doublet. Inset: histogram (means ± SE) indicating that Kv1 block results in a significant reduction in the minimal current to elicit an onset doublet in glt PNs (*P < 0.05).

Fig. 9.

Onset burst firing and acceleration during firing in glt pyramidal neurons (PNs). A: trace from a glt PN firing in response to a 2-s direct current (DC) injection (250 pA). There is an initial burst of 3 action potentials riding on a slow depolarization at the onset of the cell’s response. In control solutions, 8/83 glt PNs fired with an onset burst (and no etv1 PNs). B: trace from a different glt PN that exhibited firing that accelerated in rate during a 2-s DC injection. This pattern was observed in 4/52 glt PNs tested (and no etv1 PNs). C: plot of instantaneous frequency [Instant Freq; 1/interspike interval (ISI)] vs. time for the glt PN in B, showing acceleration (increased frequency with time) during the 2-s firing in control solution (black). In 30 nM margatoxin (MTX; gray), the cell fired faster and acceleration was replaced with spike frequency adaptation (SFA). In 3 of 4 glt PNs that showed acceleration in control solutions, the acceleration was lost after Kv1 block.

Fig. 10.

Effects of Kv1 channels on gain of repetitive firing. A1: traces from a glt neuron in response to 2-s current step designed to elicit firing at ~10 Hz (100 pA and 9.5 Hz in this example) in control solution (Ctl). Scale bars apply to A1 and A2 (mV scale also applies to A3). A2: traces for same cell and same stimulus as in A1, except in the presence of 100 nM α-dendrotoxin (DTX) + 20 nM margatoxin (MTX). A3: initial portions of traces in A1 and A2 at an expanded time base to show advanced firing in DTX+MTX. B: summary data (means ± SE) for etv1 and glt cells in response to Kv1 blockers (all = 100 nM DTX, 20 nM MTX, and DTX+MTX). *P < 0.05, significant difference. C: summary data (means ± SE) for etv1 and glt cells for initial slope from linear fit to frequency-current (f-I) data in response to Kv1 blockers (all). *P < 0.05, significant difference. D: average f-I plot for 38 etv1 cells in control solution (black) and Kv1 blockers (DTX, MTX, and DTX+MTX; gray). The f-I slope was significantly steeper in the Kv1 toxins (P < 0.001). E: average f-I plot for 42 glt cells in control solution (black) and Kv1 blockers (DTX, MTX, and DTX+MTX; gray). The f-I slope was significantly steeper in the Kv1 toxins (P < 0.001).

Fig. 11.

Effects of Kv1 blockers on spike frequency adaptation (SFA). %SFA was defined as the ratio [(last ISI − 3rd ISI)/last ISI] × 100%, where ISI is interspike interval (see text). A: plot of instantaneous frequency (1/ISI) vs. time (f-t plot) for an etv1 PN pyramidal neuron (PN) in control solution (black) and after 100 nM α-dendrotoxin (DTX) + 20 nM margatoxin (MTX; gray). Note smooth, slow reduction in firing rate with time in both conditions. In Kv1 blockers, firing rate was increased and SFA reduced. In this cell, %adaptation was 74% in control and 53.1% in Kv1 blockers. B: similar f-t plot for a glt PN. There was an initial fast SFA (ISIs 1 and 2) and very little slow SFA in control (11%). In Kv1 blockers, firing rate was increased and SFA slightly increased (to 19.1%). C: plot of %adaptation vs. injected current (I) for etv1 and glt PNs in control (black) and Kv1 blockers (gray). In etv1 PNs, %adaptation was reduced by Kv1 blockers at high currents, which are associated with faster firing. In glt PNs, there was a small but statistically significant increase in %adaptation with DTX+MTX at the lowest current (and firing rate).