Allopregnanolone Effects on Inhibition in Hippocampal Parvalbumin Interneurons

Action potential characteristics and repetitive firing patterns in fluorescent neurons readily distinguish seven PV interneurons (IN) from 12 non-PV interneurons (Non PV IN) and six DGCs. A–C, Representative voltage traces induced by depolarizing current injection. D, Maximum action potential frequency obtained by incremental increases in current amplitude from −50 to 400 pA. One-way ANOVA showed a cell type effect (F_(2,22) = 26.49, p < 0.0001). PV interneurons had a higher maximum firing frequency than non-PV interneurons and DGC (Holm–Sidak, p < 0.0001 for both). E, Half-width of action potentials elicited by just supra-threshold current injection. One-way ANOVA showed a main effect of cell type (F_(2,22) = 7.25, p = 0.004). PV interneurons had a smaller half-width than non-PV interneurons and DGC (Holm–Sidak, p = 0.005 and p = 0.004, respectively). F, Percent accommodation of action potentials at maximum firing frequency. One-way ANOVA showed a genotype effect (F_(2,21) = 9.431, p = 0.001). PV interneurons had larger accommodation than DGC (Holm–Sidak, p = 0.001), but not non-PV interneurons (Holm–Sidak, p = 0.362). The red symbols indicate two cells in which the requisite number of action potentials was not exceeded (see Materials and Methods), so the time between the final two action potentials was taken as a minimum accommodation measure.

AlloP promotes phasic but not tonic inhibition in hippocampal PV interneurons and CA1 pyramidal neurons. A, Representative average sIPSC waveform from a PV interneuron at baseline (black trace) and after 100 nm AlloP application (red trace). B–D, sIPSC characteristics of PV interneuron (N = 9): (B) weighted decay τ (τ_w) of sIPSC (paired t test, p = 0.015; *), (C) frequency (paired t test, p = 0.392; ns), and peak amplitude (paired t test, p = 0.663; ns). E, Representative average sIPSC waveform from a CA1 pyramidal cell at baseline (black trace) and after 100 nm AlloP application (red trace). F–H, CA1 pyramidal neuron sIPSC characteristics of WT (N = 8) and δ cKO (N = 7). F, Two-way ANOVA showed a drug effect on the weighted decay τ (τ_w) of sIPSC (F_(1,13) = 20.05, p = 0.0006). AlloP increased τ_w of sIPSCs in WT and cKO (Holm–Sidak’s test, p = 0.004 and 0.024, respectively). G, Two-way ANOVA showed no drug effect on the frequency of sIPSC (F_(1,13) = 3.26, p = 0.094). AlloP had no effect on frequency of sIPSC in WT and cKO (Holm–Sidak’s test, p = 0.099 and 0.652, respectively). H, Two-way ANOVA showed no drug effect on the amplitude of sIPSC (F_(1,13) = 0.725, p = 0.41). AlloP had no effect on amplitude of sIPSCs in WT and cKO (Holm–Sidak’s test, p = 0.76 and 0.76, respectively). I, Effects of 100 nm AlloP on the holding current in a PV interneuron at −70 mV. J, Summary of PV interneuron tonic current (N = 9). There was no discernible drug effect on the tonic current (one-way ANOVA, F_{(1.307,10.45)} = 0.336, p = 0.632). There was no difference in tonic current between baseline and AlloP, AlloP and PTX, baseline and PTX (Holm–Sidak’s test, p = 0.906, respectively). K, Summary of the PV interneuron SD of current (N = 9). There was no difference in the SD between baseline and AlloP (Holm–Sidak’s test, p = 0.846). 100 μm PTX decreased the SD from baseline and AlloP (Holm–Sidak’s test, p = 0.016 and 0.028, respectively), one-way ANOVA (F_{(1.403,11.22)} = 10.20, p = 0.005). L, Effects of 100 nm AlloP on the holding current in a CA1 pyramidal cell at −70 mV. M, Summary of the tonic current in CA1 pyramidal cells. Two-way ANOVA showed no drug effect on the tonic current (F_(2,24) = 1.783, p = 0.19). AlloP had no effect on the tonic current in either WT or cKO CA1 pyramidal cells (Holm–Sidak’s test, p = 0.932 and 0.349, respectively). N, Summary of the CA1 pyramidal neuron SD deviation of currents. Two-way ANOVA showed a drug effect on the SD (F_(2,24) =18.04, p < 0.0001). AlloP had no effect on the SD in either WT or cKO (Holm–Sidak’s test, p = 0.726 and 0.240, respectively).

AlloP promotes tonic inhibition in the presence of 5 μm exogenous GABA. A, Representative effects of 5 μm GABA, 100 nm AlloP, and 100 μm PTX on the holding current in a PV interneuron. B, Summary of changes in mean current relative to pre-GABA baseline in WT, δ* KI, and δ cKO PV interneurons. In WT PV interneurons (N = 8), one-way ANOVA showed a drug effect (F_(1.942,13.6) = 12.79, p = 0.001). GABA/AlloP potentiated a tonic current of −24.8 pA, which is increased over GABA-induced current (Holm–Sidak, p = 0.012). There is a decrease of tonic current by 100 μm PTX (Holm–Sidak, p = 0.012). In δ* KI PV interneurons (N = 5), GABA/AlloP potentiated a tonic current of −23.8 pA. 100 μm PTX decreased GABA/AlloP potentiated tonic current (paired t test, p = 0.004). In δ cKO PV interneurons (N = 5), GABA/AlloP potentiated an average of −17.7 pA tonic current, which was suppressed by PTX (paired t test, p = 0.009). C, Summary of the SD of current. In WT (N = 8), one-way ANOVA showed a drug effect (F_{(1.330,9.311)} = 15.98, p = 0.002). There is a decrease of SD by PTX (Holm–Sidak’s test, p = 0.016). In δ* KI PV interneurons (N = 7), one-way ANOVA showed a drug effect (one-way ANOVA, F_{(1.444,8.633)} = 11.10, p = 0.006). PTX decreased SD (Holm–Sidak’s test, p = 0.016). In δ cKO PV interneurons (N = 6), one-way ANOVA showed a drug effect (one-way ANOVA, F_{(1.642,8.212)} = 18.28, p = 0.001). PTX decreased SD (Holm–Sidak’s test, p = 0.008).

AlloP promotes inhibition of DGC firing but not PV interneuron firing. A, Representative action potential patterns of DGC baseline with 200 pA current injection and following application of 100 nm AlloP. B, Average number of action potentials in DGCs elicited at the indicated current amplitudes (N = 8). Upper panel, Two-way ANOVA on number of action potentials showed a drug effect (F_(1,6) = 10.87, p = 0.017). AlloP decreased the number of action potentials with 100 pA current step (Holm–Sidak, p < 0.001), 150 pA (Holm–Sidak’s test, p < 0.0001), 200 pA (Holm–Sidak’s test, p < 0.0001), and 250 pA (Holm–Sidak’s test, p = 0.0004). Lower panel, 10 μm gabazine prevented the effect of AlloP on excitability [two-way ANOVA (F_(6,30) = 1.716, p = 0.152)]. C, Upper panel, Effect of AlloP on the DGC membrane resistance. A hyperpolarization was induced by injecting −50 pA current for 500 ms. Membrane resistance was calculated by dividing the final change in voltage by the current (N = 7, paired t test, p = 0.018). Lower panel, In the presence of gabazine (10 μm), there was no consistent change in the input resistance of DGCs (N = 6, paired t test, p = 0.277). D, Representative firing of a PV interneuron at baseline and after addition of 100 nm AlloP. E, Number of action potentials in PV interneurons elicited current injection (N = 6). Two-way ANOVA on number of action potentials showed no drug effect (F_(1,5) = 0.227, p = 0.654). F, No effects of AlloP on the PV interneuron membrane resistance (N = 7, paired t test, p = 0.618). In D–F, four PV interneurons were from CA1 and three were from dentate gyrus.

AlloP potentiation of phasic inhibition does not affect the firing frequency of DGCs or PV interneurons. A, Representative action potential patterns for DGC with (red traces) or without (black traces) electrical synaptic stimulation (red arrows) during baseline or (B) 100 nm AlloP. C, Representative action potential patterns of PV interneurons with or without electrical stimulation during baseline or (D) 100 nm AlloP. The first action potential latency showed no genotype difference between DGCs and PV interneurons [two-way ANOVA (F_(1,11) = 1.097, p = 0.3175)]. E, Normalized AlloP effect on action potentials of DGCs (N = 11) or PV interneurons (N = 6). Two-way ANOVA showed no genotype difference (F_(1,15) = 0.1632, p = 0.6919). There is no difference for normalized action potentials between DGCs and PV INs with electrical stimulation (+), AlloP (+), or both conditions (Holm–Sidak’s, p = 0.955, p > 0.1, or p = 0.961, respectively).