Spike firing attenuation of serotonin neurons in learned helplessness rats is reversed by ketamine

Abstract

Animals suffering from uncontrollable stress sometimes show low effort to escape stress (learned helplessness). Changes in serotonin (5-hydroxytryptamine) signalling are thought to underlie this behaviour. Although the release of 5-hydroxytryptamine is triggered by the action potential firing of dorsal raphe nuclei 5-hydroxytryptamine neurons, the electrophysiological changes induced by uncontrollable stress are largely unclear. Herein, we examined electrophysiological differences among 5-hydroxytryptamine neurons in naïve rats, learned helplessness rats and rats resistant to inescapable stress (non-learned helplessness). Five-week-old male Sprague Dawley rats were exposed to inescapable foot shocks. After an avoidance test session, rats were classified as learned helplessness or non-learned helplessness. Activity-dependent 5-hydroxytryptamine release induced by the administration of high-potassium solution was slower in free-moving learned helplessness rats. Subthreshold electrophysiological properties of 5-hydroxytryptamine neurons were identical among the three rat groups, but the depolarization-induced spike firing was significantly attenuated in learned helplessness rats. To clarify the underlying mechanisms, potassium (K⁺) channels regulating the spike firing were initially examined using naïve rats. K⁺ channels sensitive to 500 μM tetraethylammonium caused rapid repolarization of the action potential and the small conductance calcium-activated K⁺ channels produced afterhyperpolarization. Additionally, dendrotoxin-I, a blocker of Kv1.1 (encoded by Kcna1), Kv1.2 (encoded by Kcna2) and Kv1.6 (encoded by Kcna6) voltage-dependent K⁺ channels, weakly enhanced the spike firing frequency during depolarizing current injections without changes in individual spike waveforms in naïve rats. We found that dendrotoxin-I significantly enhanced the spike firing of 5-hydroxytryptamine neurons in learned helplessness rats. Consequently, the difference in spike firing among the three rat groups was abolished in the presence of dendrotoxin-I. These results suggest that the upregulation of dendrotoxin-I-sensitive Kv1 channels underlies the firing attenuation of 5-hydroxytryptamine neurons in learned helplessness rats. We also found that the antidepressant ketamine facilitated the spike firing of 5-hydroxytryptamine neurons and abolished the firing difference between learned helplessness and non-learned helplessness by suppressing dendrotoxin-I-sensitive Kv1 channels. The dendrotoxin-I-sensitive Kv1 channel may be a potential target for developing drugs to control activity of 5-hydroxytryptamine neurons.

Keywords: 5-HT neuron; Kv1 voltage-dependent K+ channel; dendrotoxin-I; dorsal raphe nucleus; learned helplessness.

Graphical Abstract

Figure 1

Chemically stimulated 5-HT release is downregulated in the DRN of LH rats. (A) Line plots of the mean extracellular concentration of 5-HT (as represented by change in percentage from the baseline) in the DRN of non-LH (open circles, rats =11) and LH rats (closed circles, rats = 9), before and after stimulation by the high-K⁺ solution applied at time zero. (B and C) Individual data points and the mean peak amplitude (P = 0.494, Mann–Whitney U-test) (B) and latency to peak [t(18) = −2.47, P = 0.024, t-test] (C) of the extracellular 5-HT concentration in the DRN of non-LH (open) and LH rats (closed) after chemical stimulation. Data are presented as mean ± SEM. *P<0.05.

Figure 2

Spike firing of 5-HT neurons is attenuated in LH rats. (A–C) Representative voltage traces in response to 10 pA steps of hyper- and depolarizing current injections (−50 to +30 pA) in naïve (A), non-LH (B) and LH (C) rats. (D) Representative voltage traces in response to depolarizing current injections of 0 pA (upper), 30 pA (middle) and 60 pA (bottom) into naïve (left), non-LH (middle) and LH (right) rats. (E) Average spike numbers are plotted against the amplitudes of the injected currents. Green solid, blue open and blue solid symbols represent data of naïve (n=55 cells, rats =23), non-LH (n=18 cells, rats =7) and LH rats (n=25 cells, rats =13), respectively. Recordings of all neurons in non-LH (7 animals) rats and 18 out of 25 neurons in LH (8 out of 13 animals) rats were performed in a manner blinded to the phenotype. (Upper) Ratio of average spike number in conditioned rats (non-LH and LH) relative to that in naïve rats in the lower panel. (F) Individual data and average spike numbers upon 100, 110 and 120 pA current injections (spikes@100–120 pA). *P < 0.05, one-way ANOVA, post hoc, Holm–Šidák test.

Figure 3

Contributions of voltage-dependent and Ca²⁺-activated K⁺ channels to the spike firing of DRN 5-HT neurons in naïve rats. (A) (Left) Representative waveforms in normal Ringer solution (control) and in the presence of apamin (100 nM), DTX-I (100 nM), HpTx2 (200 nM) or TEA (500 µM). Afterhyperpolarization was suppressed in the presence of apamin (arrowhead). The width of the action potential was prolonged by 500 µM TEA (arrow). (Right) The time scale of the left traces around the Na⁺-spike was expanded. (B) Representative spike trains in response to a current injection of 120 pA. (C–F) Average spike numbers in response to depolarizing current injections (10–120 pA with 10 pA intervals, 1 s duration) in normal Ringer solution (open symbols, n = 55 cells, rats = 23) and in the presence of apamin (C, closed symbols, n = 9, rats = 3), TEA (D, n = 14, rats = 3), DTX-I (E, closed symbols, n = 8, rats = 2) or HpTx2 (F, closed symbols, n = 13, rats = 5) are plotted against the amplitudes of the injected currents. (Upper) The ratio of the average spike numbers in the presence of individual blockers relative to those in the normal external solution in individual lower panels. (G) Frequency distribution histogram of the average spike numbers upon 100 − 120 pA current injections (spikes@100–120 pA) (n = 55, rats = 23 for control and n = 8, rats = 2 for DTX-I). (H) Frequency distributions of the proportion of 5-HT neurons with spikes@100–120 pA higher than or equal to/lower than the median spikes@100–120 pA (7.67) in normal external solution (control). The proportion of 5-HT neurons with average spike numbers higher than the median was significantly increased in the presence of DTX-I [P = 0.029, Fisher’s exact test (one-sided)]. (I) Representative waveforms of a single spike and spike trains upon the injection of 120 pA current. DTX-I enhanced the spike firing even in the presence of 500 μM TEA. (J) Similar to E, but data in the presence of TEA alone (same as data in D) or TEA plus DTX-I (n = 11, rats = 3).

Figure 4

Firing attenuation in LH rats is reversed in the presence of DTX-I. (A) Representative voltage traces in response to depolarizing current injections of 0 pA (upper), 30 pA (middle) and 60 pA (bottom) in naïve (left), non-LH (middle) and LH (right) rats in the presence of DTX-I (100 nM). (B) The average spike numbers in the presence of DTX-I are plotted against the amplitudes of the depolarizing currents. Blue open and blue solid symbols represent data of non-LH (n=8 cells, rats = 2) and LH (n=13 cells, rats = 4) rats, respectively. The naïve rats’ data (green closed squares) are the same as in Fig. 3E (green closed circles). (Upper) The ratio of the average spike numbers of conditioned rats (non-LH and LH) relative to that of naïve rats in the lower panel. (C) Individual data and the average spikes@100–120 pA. There were no statistically significant differences [F(2, 26)=2.118, P=0.141, one-way ANOVA].

Figure 5

DTX-I-sensitive Kv1 channels are expressed in DRN neurons. (A–C) Fluorescent in situ hybridization combined with immunolabeling of TPH2 (red) and NeuN (blue) showing expression of Kcna1 (Kv1.1, A), Kcna2 (Kv1.2, B) and Kcna6 (Kv1.6, C) mRNA (green) in the DRN. A(i)–C(i), all merged images; A(ii)–C(ii), Kv1 signals. (D, G and J) Magnified images of the square areas in A(i), B(i) and C(i). (E–L) Magnified images of the square areas in D, G and J. E(i)–L(i), all merged images; E(ii)–L(ii), merged images of Kv1 and NeuN signals. Arrows indicate TPH2-positive neurons. Scale bars, A(i), 100 μm; D, 20 μm.

Figure 6

Ketamine (Ket) facilitates spike firing of 5-HT neurons via suppression of DTX-I-sensitive Kv1 channels. (A) Representative waveforms in response to current injections of 120 pA in normal Ringer solution (naïve) and in the presence of Ket (100 μM), DTX-I (100 nM) or DTX-I and Ket (DTX+Ket) in naïve rats. (B) The average spike numbers in response to depolarizing current injections (10–120 pA with 10 pA intervals, 1 s duration) in normal Ringer solution (green symbols, n=17 cells, rats =9) and in the presence of Ket (pink symbols, n=8, rats =3) in naïve rats. (C) Dose–response plots of the spikes@100–120 pA in control solution (green symbols) and in the presence of 1, 10 and 100 μM Ket (pink symbols) in naïve rats (n=10, rats = 6 for control and 100 μM Ket; n=6, rats = 4 for 1 and 10 μM Ket). Data were fitted by the Hill equation (red line). (D) Similar to B, but average spike numbers in the presence of DTX-I (orange symbols, n=7, rats =4) or DTX-I and Ket (purple symbols, n=7, rats =4). (E) Frequency distributions of the proportion of 5-HT neurons with spikes@100–120 pA higher than or equal to/lower than the median average spikes@100–120 pA (11) of the normal external solution. The proportion of 5-HT neurons with spikes@100–120 pA larger than the median was significantly increased in the presence of Ket [P=0.006, Fisher’s exact test (one-sided)] in naïve rats. (F) Similar to E, but comparing between DTX-I and the co-application of DTX-I and Ket. These distributions were not significantly different (P=0.5). (G) Representative waveforms in response to injection of a current of 120 pA in non-LH and LH rats in the presence or absence of Ket. (H) Similar to B, but average spike numbers in non-LH and LH rats in the presence (pink square symbols; non-LH, n = 10, rats = 5; LH, n = 16, rats = 5) or absence (blue circle symbols; non-LH, n=10, rats =5; LH, n=8, rats =5) of Ket. (I) Individual data and average spikes@100–120 pA. *P <0.05, one-way ANOVA, post hoc, Holm–Šidák test.