Function of KCNQ2 channels at nodes of Ranvier of lumbar spinal ventral nerves of rats

Linopirdine-sensitive K+ currents and KCNQ2 expression at nodes of Ranvier of lumbar spinal ventral nerves

We performed whole-cell patch-clamp recordings to investigate whether KCNQ2 may mediate outward K+ currents at NRs of lumbar spinal ventral nerves of rats. Figure 1A is an example showing a recording made at a NR of a lumbar spinal ventral nerve. In this recording, the fluorescent dye Alexa Fluor 555 was included in recording electrode internal solution to label axons. The nodal axon (narrow part) and internodal axon could be visualized under a fluorescent microscope (Fig. 1A right). Under voltage-clamp configuration, we examined membrane currents following the application of depolarizing voltage steps at NRs. Depolarizing voltage steps evoked large transient inward currents which were immediately followed by large non-inactivating outward currents (Fig. 1B). The non-inactivating outward currents were partially inhibited by 100 µM linopirdine, a potent blocker of KCNQ channels (Fig. 1B, C). Figure 1C shows current–voltage relationship (I–V curve) of the non-inactivating outward currents at the beginning (Fig. 1C left) and the end (Fig. 1C right) of the voltage steps. Both the beginning and the end parts of the non-inactivating outward currents were reduced following the application of 100 µM linopirdine. For example, at the voltage step to 18 mV, the non-inactivating outward currents at the beginning were 4.1 ± 0.2 nA (n = 10) in the control, and significantly reduced to 3.4 ± 0.1 nA (n = 10) in the presence of linopirdine (p < 0.001, t = 3.4, Fig. 1C left). The non-inactivating outward currents at the end part were 4.4 ± 0.2 nA in the control, and significantly reduced to 3.1 ± 0.1 nA (n = 10) in the presence of linopirdine (p < 0.001, t = 7.8, Fig. 1C right). Expressed as the area under the I–V curve (AUC), non-inactivating outward currents at the beginning were 40.5 ± 1.7 nA*mV in the control, and significantly reduced to 33.6 ± 1.2 nA*mV in the presence of linopirdine (n = 10, P < 0.01, t = 7.9, Fig. 1D left). The non-inactivating outward currents at the end were 42.7 ± 2.0 nA*mV in the control, and significantly reduced to 30.6 ± 1.1 nA*mV in the presence of linopirdine (n = 10, P < 0.001, t = 3.5, Fig. 1D right). Linopirdine-sensitive non-inactivating currents were isolated after subtraction of the total currents in the presence of 100 μM linopirdine from the total currents in the absence of linopirdine (Fig. 1E left). The isolated outward currents displayed slow activation and inactivation with tail currents (Fig. 1E left). Reversal potentials of the isolated outward currents were − 82.6 ± 0.9 mV (n = 10, Fig. 1E right). These results indicate that a small but significant fraction of non-inactivating outward current at NRs is mediated by linopirdine-sensitive KCNQ channels. Consistent with the above electrophysiological and pharmacological results, immunohistochemical study with anti-KCNQ2 antibody showed strong KCNQ2-immunoreactivity (KCNQ2-ir) at NRs of lumbar spinal ventral nerve fibers (Fig. 1F–H). KCNQ2-ir was sandwiched by caspr-immunoreactivity (caspr-ir), which outlined paranodal regions on myelinated nerves (Fig. 1H).

Fig. 1figure 1

Linopirdine-sensitive non-inactivating K+ currents and KCNQ2 expression at NRs of lumbar spinal ventral nerves. A Bright image (left) and fluorescent image (right) show the same region of an L5 lumbar spinal ventral nerve viewed under a 40 × objective. Arrow indicates a node of Ranvier (NR) and a patch-clamp recording electrode tip. Alexa Fluor 555 (85 μM) was included in the recording electrode internal solution for axon labelling. B Sample traces show currents recorded at a NR following voltage steps in the absence (left, control) and presence of 100 µM linopirdine (right). C I–V curves at the beginning part (left panel) and the end part (right panel) of non-inactivating outward currents in the control (n = 10, open symbols) and the presence of linopirdine (n = 10, red symbols). D Area under the I–V curve (AUC) of the beginning (left panel) and the end part (right panel) of non-inactivating outward currents in the control (n = 10, open bar) and the presence of linopirdine (n = 10, red bar). E Left panel, sample traces show linopirdine-sensitive outward currents obtained by subtraction of total outward currents in the presence of linopirdine from total currents in the control. Right panel, I–V curves of the linopirdine-sensitive outward currents at the end part (n = 10, solid circles). F, G KCNQ2-immunoreactivity (KCNQ2-ir, F) and caspr-ir (G) on lumbar spinal ventral nerves. H Overlay image of (F) and (G) shows that KCNQ2-ir is located in the space between caspr-ir, the region of NRs. Data represent mean ± SEM, **p < 0.01, ***p < 0.001, paired Student’s t-Test

We next measured the effects of KCNQ channel activator retigabine on non-inactivating outward currents at NRs. At the beginning components, non-inactivating outward currents were slightly increased following the application of 10 µM retigabine (n = 6, Fig. 2A, B). AUC at the beginning part of the outward currents was 43.8 ± 4.7 nA*mV (n = 6) in the control, and significantly increased to 48.0 ± 3.2 nA*mV (n = 6, p < 0.01, t = 4.4) in the presence of retigabine (Fig. 2B right). However, we did not observe any significant difference in the end part of the non-inactivating outward currents between the control and in the presence of retigabine (Fig. 2C). Therefore, KCNQ2 channels expressed at NRs of lumbar spinal ventral nerves partially mediate non-inactivating outward K+ currents with sensitivity to linopirdine and retigabine observed.

Fig. 2figure 2

Effects of retigabine on non-inactivating outward currents at NRs. A Two sets of sample traces show currents recorded at a NR following voltage steps in the absence (left, control) and presence of 10 µM retigabine (right). Circles and triangles indicate the beginning part and the end part of the outward currents, respectively. Currents at these two parts are used to plot I–V relationship in (B) and (C). B Left panel, I–V curve of the beginning part of the outward currents in the control (n = 6, open circle) and the presence of retigabine (n = 6, blue circle). Right panel, area under the I–V curve (AUC) of the beginning part of the outward currents in the control (n = 6, open bar) and the presence of retigabine (n = 6, blue bar). C Left panel, I–V curve of the end part of the outward currents in the control (n = 6, open triangle) and the presence of retigabine (n = 6, blue triangle). Right panel, AUC of the end part of the outward currents in the control (n = 6, open bar) and the presence of retigabine (n = 6, blue bar). AUC was used for statistical comparison of non-inactivating outward currents in the control and the presence of retigabine. Data represent mean ± SEM, ns, not significantly different, **p < 0.01, paired Student’s t-Test

Role of KCNQ2 channels in regulating intrinsic electrophysiological properties at nodes of Ranvier

To determine potential role of KCNQ2 channels in regulating intrinsic electrophysiological properties of nodal membranes, we investigated whether inhibition of KCNQ2 channels by linopirdine (100 µM) could significantly alter membrane and action potential properties at NRs of the lumbar spinal ventral nerves. Under current-clamp configuration, we measured resting membrane potential (RMP) at NRs, then depolarized NRs from their RMP by injecting depolarizing current steps to evoke APs at NRs (Fig. 3A). RMP at NRs was − 84.4 ± 1.1 mV (n = 10) in the absence of linopirdine (control), and significantly depolarized to − 82.0 ± 1.7 mV (n = 10) in the presence of linopirdine (Fig. 3B, p < 0.05, t = 2.9). AP threshold was − 59.5 ± 0.6 mV (n = 10) in the control group, and − 57.4 ± 1.5 mV (n = 10) in the presence of linopirdine, and was not significantly different (Fig. 3C). AP width was 0.88 ± 0.02 ms (n = 10) in the control group, and significantly prolonged to 0.93 ± 0.03 ms (n = 10) in the presence of linopirdine (Fig. 3D, p < 0.01, t = 3.8). AP rheobase was 465.0 ± 38.8 pA (n = 10) in the control group, and significantly decreased to 390.0 ± 37.1 pA (n = 10) in the presence of linopirdine (Fig. 3E, p < 0.01, t = 3.7). AP amplitude was 104.4 ± 2.3 mV (n = 10) in the control group, and significantly decreased to 95.0 ± 2.6 mV (n = 10) in the presence of linopirdine (Fig. 3F, p < 0.001, t = 4.9). Input resistance, measured under the voltage-clamp configuration following a − 10 mV testing pulse from membrane holding voltage at − 72 mV, was 35.3 ± 2.5 MΩ (n = 10) in the control group, and significantly increased to 46.9 ± 1.9 MΩ (n = 10) in the presence of linopirdine (Fig. 3G, p < 0.001, t = 6.7).

Fig. 3figure 3

Effects of linopirdine on intrinsic electrophysiological properties at NRs. A Sample traces illustrate action potentials (APs) recorded at a NR in the absence (control, grey) and presence of 100 µM linopirdine (red). Recordings were performed under the whole-cell current-clamp configuration and step currents were injected into the NR through recording electrodes. BG Bar graphs show, in the absence (control, open bars, n = 10) and presence of linopirdine (red bars, n = 10), intrinsic electrophysiological properties including resting membrane potential (RMP, B), action potential (AP) threshold (C), AP width (D), AP rheobase (E), AP amplitude (F), and membrane input resistance (G). Data represent mean ± SEM, ns, not significantly different, *p < 0.05, **p < 0.01, ***p < 0.001, paired Student’s t-Test

We determined effects of retigabine (10 µM) on intrinsic electrophysiological properties at NRs (Fig. 4). Input resistance was 34.7 ± 3.6 MΩ (n = 6) in the control group, and significantly decreased to 28.5 ± 2.0 MΩ (n = 6) in the presence of retigabine (Fig. 4A, p < 0.01, t = 3.4). AP rheobase was 607.1 ± 61.2 pA (n = 7) in the control group, and significantly increased to 842.9 ± 81.2 pA (n = 7) in the presence of retigabine (Fig. 4B, p < 0.01, t = 4.7). Other membrane and AP properties including RMP, AP threshold, AP width and AP amplitude were not significantly different between the control group and in the presence of 10 µM retigabine (Fig. 4C–F).

Fig. 4figure 4

Effects of retigabine on intrinsic electrophysiological properties at nodes of Ranvier. AF Bar graph shows, in the absence (control, open bars) and presence of 10 µM retigabine (Blue bars), intrinsic electrophysiological properties at NRs including membrane input resistance (A, n = 6), AP rheobase (B, n = 7), RMP (C, n = 7), AP threshold (D, n = 7), AP width (E, n = 7), and AP amplitude (F, n = 7). Data represent mean ± SEM, ns, not significantly different, **p < 0.01, paired Student’s t-Test

Role of KCNQ2 channels in controlling action potential firing patterns at nodes of Ranvier

At NRs a supra-threshold depolarizing current step usually only evoked a single AP (Fig. 5A, C). We determined whether KCNQ2 channels play a role in the AP firing pattern at NRs. This was done by investigation of AP firing patterns at NRs following the inhibition of KCNQ2 channels by linopirdine (Fig. 5B, C). In this set of experiments, APs were evoked by step currents at 3 × rheobase for 50 ms in the absence (control) and presence of 100 µM linopirdine. As shown in Fig. 5A–C, the percent of NRs showing single AP firing was 83.3% and multiple AP firing was 16.7% in the control group (n = 12, Fig. 5C left). The NRs with single AP firing significantly reduced to 41.7% and multiple AP firing significantly increased to 58.3% (n = 12) in the presence of linopirdine (Fig. 5C left, p < 0.05). Averaged AP numbers were analyzed, which were 1.2 ± 0.1 (n = 12) in the control group, and significantly increased to 3.4 ± 0.8 (n = 12) in the presence of 100 µM linopirdine (Fig. 5C right, p < 0.001, t = 2.4). We also tested effects of retigabine (10 µM) on AP firing patterns at NRs. The percent of NRs showing single AP firing was 71.5% and multiple AP firing was 28.5% (n = 7) in the control group (Fig. 5D left). The NRs showing single AP firing was 85.8% and multiple AP firing was 14.2% (n = 7) in the presence of 10 µM retigabine (Fig. 5D left), and were not significantly different from the control group. Averaged AP numbers were 1.4 ± 0.3 (n = 7) in the control group, and 1.1 ± 0.1 (n = 7) in the presence of 10 µM retigabine (Fig. 5D right), and were not significantly different.

Fig. 5figure 5

Effects of linopirdine and retigabine on AP firing patterns at nodes of Ranvier. A, B Sample traces show APs evoked at a NR by current steps in the absence (control, A) and presence of 100 µM linopirdine (B). The grey and black traces in (A) and (B) show APs evoked with rheobase (grey) and 3 × rheobase (black) stimulation, respectively. Recordings were performed under the whole-cell current-clamp configuration and step currents were injected into the NR through recording electrodes. C Left panel, percent of NRs with single AP firing (open bar) and multiple AP firing (black bar) before (control, n = 12) and following the application of linopirdine (n = 12). Right panel, Numbers of APs before (control, n = 12) and following the application of linopirdine (n = 12). D Similar to C except retigabine was tested (n = 7). Data represent mean ± SEM, ns, not significantly different, *p < 0.05, ***p < 0.001, paired Student’s t-Test

Role of KCNQ2 channels in regulating saltatory conduction

We investigated whether KCNQ2 channels at NRs may play a role in regulating saltatory conduction velocity. In this set of experiments, we performed whole-cell patch-clamp recordings at NRs while APs were elicited by electrical stimulation applied to a distal site of lumbar spinal ventral nerves (Fig. 6A). Recordings were performed in the absence (control) and presence of 100 μM linopirdine (Fig. 6B), or in the absence and presence of 10 μM retigabine (Fig. 6C). AP conduction velocity was 45.4 ± 3.2 m/s (n = 9) in the absence of linopirdine, and not changed (45.4 ± 3.2 m/s, n = 9) in the presence of 100 μM linopirdine (Fig. 6D). AP conduction velocity was 45.7 ± 5.0 m/s (n = 6) in the control group, and significantly decreased to 41.3 ± 4.2 m/s (n = 6) in the presence of 10 µM retigabine (Fig. 6E, p < 0.01, t = 5.0).

Fig. 6figure 6

Effects of linopirdine and retigabine on saltatory conduction velocity. A Experimental setting for recording of AP propagation through NRs of lumbar spinal ventral nerves. APs were evoked from the distal site of the nerve bundle. B Two overlay sample traces show APs recorded from a NR following electrical stimulation. The recordings were performed in the absence (control, black) and presence of 100 µM linopirdine (red). C Two overlay sample traces show APs recorded from a NR following electrical stimulation. The recordings were performed in the absence (control, black) and presence of 10 µM retigabine (blue). In both (B) and (C), arrow indicates stimulation artifact. D Summary data of AP conduction velocity determined from the recordings exemplified in (B) in the control (n = 9) and the presence of linopirdine (n = 9). E Summary data of AP conduction velocity determined from the recordings exemplified in (C) in the control (n = 6) and the presence of retigabine (n = 6). Data represent mean ± SEM, ns, not significantly different, **p < 0.01, paired Student’s t-Test

We then investigated the potential role of KCNQ2 channels in regulating AP firing frequency at NRs. In this set of experiments, trains of stimuli at frequencies of 1, 10, 100, 200, 500, and 1000 Hz were applied to a distal end of nerves for a duration of 20 s at each stimulation frequency. Figure 7A shows sample traces of nodal APs elicited by 500 Hz stimulation for 20 s in the absence (Fig. 7A top) and presence of 100 μM linopirdine (Fig. 7A bottom). Plot of AP success rate vs stimulation frequency shows that AP success rate was 100% at stimulation frequency up to 100 Hz but reduced at stimulation frequency of 200 Hz or higher stimulation frequencies in control (Fig. 7B). However, AP success rate at high frequencies was increased in the presence of linopirdine (Fig. 7A–C). For example, at the stimulation frequency of 500 Hz, AP success rate was 32.5 ± 7.7% (n = 6) in the control, and significantly increased to 43.1 ± 9.1% (n = 6) in the presence of linopirdine (Fig. 7C, p < 0.001, t = 7.3). In Fig. 7D, AP success rate at 500 Hz stimulation is plotted for a prolonged period of 20 s. AP success rate in the control group gradually reduced over time and reaching steady state level after 12 s (Fig. 7D). In contrast, AP success rate was reduced to a lesser degree in the presence of linopirdine (Fig. 7D). We used the area under the curve (AUC) in Fig. 7D to compare AP success rate at 500 Hz stimulation, which shows a significantly higher AUC following the application of linopirdine (p < 0.001, t = 7.3, Fig. 7E). We also examined whether retigabine may affect AP success rate. Tested at stimulation frequency of 1, 10, 100, 200, 500, and 1000 Hz, there was no significant difference in AP success rate between control and following the application of 10 μM retigabine (n = 5, Fig. 7F, G).

Fig. 7figure 7

Effects of linopirdine and retigabine on high-frequency saltatory conduction. A Two sets of sample trances show APs recorded at a NR following a train of electrical stimulation at 500 Hz in control (top panel) and the presence of 100 µM linopirdine (bottom panel). Stimulation was applied for a period of 20 s. Only a short period of APs at initial time point (0 s), 5 s, 10 s, 15 s and 20 s time point are presented to illustrate changes of AP success rates over the time period of 20 s. B AP success rate at NRs with distal stimulation at frequency of 1, 10, 100, 200, 500 and 1000 Hz in the absence (n = 6) and presence of linopirdine (n = 6). Recording duration at each frequency was 20 s. C Comparison between averaged AP success rates at NRs at 500 Hz stimulation (n = 6). D Time course of AP success rate at NRs during a train of 500 Hz stimulation for 20 s. Black circles, control (n = 6); red circles, in the presence of linopirdine (n = 6). Time bin, 0.2 s. E Area under the curve of D for the control (open bar, n = 6) and the presence of linopirdine (red bar, n = 6). F Similar to (B) except retigabine was tested (n = 5). G Similar to C except retigabine was tested (n = 5). Data represent mean ± SEM, ns, not significantly different, ***p < 0.001, paired Student’s t-Test

留言 (0)

沒有登入
gif