State-dependent bidirectional modification of somatic inhibition in neocortical pyramidal cells

Abstract

Cortical pyramidal neurons alter their responses to input signals depending on behavioral state. We investigated whether changes in somatic inhibition contribute to these alterations. In layer 5 pyramidal neurons of rat visual cortex, repetitive firing from a depolarized membrane potential, which typically occurs during arousal, produced long-lasting depression of somatic inhibition. In contrast, slow membrane oscillations with firing in the depolarized phase, which typically occurs during slow-wave sleep, produced long-lasting potentiation. The depression is mediated by L-type Ca2+ channels and GABA(A) receptor endocytosis, whereas potentiation is mediated by R-type Ca2+ channels and receptor exocytosis. It is likely that the direction of modification is mainly dependent on the ratio of R- and L-type Ca2+ channel activation. Furthermore, somatic inhibition was stronger in slices prepared from rats during slow-wave sleep than arousal. This bidirectional modification of somatic inhibition may alter pyramidal neuron responsiveness in accordance with behavioral state.

Figure 1. Direction of Action Potential-Induced Modification in Somatic Inhibition Depends on Membrane Potential

(A) The upper left figure illustrates the experimental arrangement of the recording patch pipette (r) and sharp glass pipettes for stimulation (s) and bicu-culline methiodide (BMI, 100 μM in ACSF) application. The average IPSC amplitude (mean ± SEM, n = 12) recorded from a cell is plotted against the stimulus intensity in the lower graph. Open circles represent control IPSCs, and filled circles and squares represent those recorded during ionto-phoresis of BMI (30 nA) from the electrode placed near the recording electrode (<5 μm) and that displaced toward the apical dendrite by 30–50 μm, respectively. The upper right traces represent average IPSCs evoked by stimuli at 3 μA. (B) An example of long-lasting depression of somatic IPSCs, produced by repetitive firing from a depolarized membrane potential (−60 mV). IPSC amplitudes (% of the mean baseline level) are plotted against time after repetitive firing. Traces on the upper left show average (n = 6) IPSCs before (1) and after repetitive firing (2), respectively. Action potentials are shown on the upper right. (C) Similar to (B), but long-lasting potentiation of somatic IPSC produced by repetitive firing from a hyperpolarized membrane potential (−90 mV). (D) Average time course of long-lasting depression when repetitive firing was produced from −70 (filled squares, n = 7) and −60 mV (filled circles, n = 9). In addition, control experiments are shown, in which membrane potential was merely maintained at −60 mV for 3 min from time 0 without eliciting action potentials (filled diamonds, n = 5). (E) Similar to (D), but average time course of long-lasting potentiation after repetitive firing from −90 mV (open circles, n = 9) and control experiments without action potentials (open diamonds, n = 5). Data in (D) and (E) are expressed as mean ± SEM. (F) Summary of the magnitude of change for repetitive firing from −60, −70, and −90 mV, and for control experiments in which membrane potential was maintained at −60 and −90 mV without action potentials. Symbols indicate the magnitude of change in individual cells. Horizontal bars show the mean. Asterisks indicate significant difference (p < 0.05, K-S test).

Figure 2. Direction of Membrane Oscillation Protocol-Induced Modification in Somatic Inhibition Depends on Oscillation Frequency

(A) An example of long-lasting potentiation of somatic IPSCs occurred after the membrane potential oscillated at 0.5 Hz for 10 min. (B) Similar to (A), but long-lasting depression produced by membrane oscillation at 5 Hz for 3 min. (C) Average time course of long-lasting potentiation and depression produced by membrane oscillation at 0.5 (open squares, n = 14) and 5 Hz (filled squares, n = 9), respectively. Data are expressed as mean ± SEM. (D) Summary of the magnitude of change for membrane oscillation at 0.5 and 5 Hz. *p < 0.05 (K-S test).

Figure 3. Amplitude of Somatic IPSCs and mIPSCs Is Larger in Slices Prepared from Rats during Slow-Wave State than Arousal

(A) Example EEGs during arousal (top) and slow-wave sleep-like states induced with urethane anesthesia (bottom). (B) Average minimal-stimulation evoked IPSC traces for all of the responses (not including failures) from all of the recorded neurons from four pairs of litter-mate rats during arousal (upper, 32 cells) and slow-wave sleep-like state (lower, 31 cells). No significant difference (p > 0.6, t test) was found in 10%–90% rise time of IPSCs between slices prepared during arousal (1.3 ± 0.11 ms, n = 32) and slow-wave sleep-like state (1.3 ± 0.10 ms, n = 31). (C) Comparison of IPSC amplitudes (holding potential, −70 mV) averaged for cells sampled from each pair of littermates. In slices from four pairs of rats, the number of cells was 4, 6, 14, and 7 for the slow-wave sleep-like state group and 5, 4, 9, and 14 for the arousal group, respectively. *p < 0.05 (paired t test). (D–F) Comparison of overall average IPSC amplitude (D), series resistance (E), and input resistance (F) for neurons sampled from slices prepared during arousal (n = 32) and slow-wave sleep-like state (n = 31). *p < 0.05 (t test). (G) Example mIPSC traces recorded from two cells held at −70 mV in the presence of tetrodotoxin (1 μM). The top and bottom cells were sampled from slices prepared from a pair of littermates during arousal and urethane-induced slow-wave sleep-like state, respectively. Cumulative probability histograms for amplitude (left) and interevent interval of mIPSCs (right) were plotted for the two cells (solid line, slow-wave sleep-like state; interrupted line, arousal). (H) Similar to (C), but for comparison of mIPSC amplitudes averaged for cells sampled from each pair of littermates (black circles). In slices from three pairs of littermates, the number of sampled cells was 9, 10, and 10 for the slow-wave sleep-like state group and 9, 10, and 10 for the arousal group, respectively. In addition, data from four pairs of littermates, from which slices were prepared during arousal and natural slow-wave sleep, were shown in red. The number of sampled cells was 14, 11, 9, and 6 for the natural slow-wave sleep group and 11, 14, 10, and 8 for the arousal group, respectively. (I) Comparison of overall average mIPSC amplitude for neurons sampled from slices prepared during arousal (n = 29) and urethane-induced slow-wave sleep-like state (n = 29, black bars) and a similar comparison (red bars) between slices prepared during arousal (n = 43) and natural slow-wave sleep (n = 40). (J and K) Similar to (H) and (I), but for interevent interval of mIPSCs. (L and M) Similar to (K), but for series resistance (L) and input resistance (M). Because no significant difference was found in these two parameters between arousal and slow-wave state for rats anesthetized with either urethane or isoflurane, these data were combined in these figures. Error bars indicate SEM.

Figure 4. Somatic IPSCs Show Similar Long-Lasting Changes Due to Voltage Pulse Stimuli

(A) Depolarizing voltage pulses (upper left) from −70 and −90 mV produced long-lasting depression (traces 1 and 2, filed circles, n = 11) and potentiation (traces 3 and 4, open circles, n = 11) of IPSCs, respectively. (B) Intracellular loading of BAPTA (10 mM) via patch pipette completely abolished the production of long-lasting depression (traces 1 and 2, filled squares, n = 8) and potentiation (traces 3 and 4, open squares, n = 7). Data of the time course graphs in (A) and (B) are expressed as mean ± SEM. (C) Summary of the magnitude of change induced by voltage pulse stimuli from −70 mV (filled circles), −90 mV (open circles), −70 mV with BAPTA (filled squares), and −90 mV with BAPTA (open squares). *p < 0.05 (K-S test).

Figure 5. Somatic GABA Responses Show Similar Long-Lasting Changes Due to Voltage Pulse Stimuli

(A) Upper drawing shows the experimental configuration. Lower traces show that a current (Control) evoked in a pyramidal neuron by iontophoretic application of GABA (short horizontal bar) to its soma was abolished by bath application of 20 μM BMI (BMI), indicating that it was mediated by GABA_A receptors. (B) GABA responses demonstrated long-lasting depression for voltage pulse stimuli from −70 mV (traces 1 and 2, filled squares, n = 13) and long-lasting potentiation for those from −90 mV (traces 3 and 4, open squares, n = 9), as were the cases in somatic IPSCs. Data of the time course graph are expressed as mean ± SEM. (C) Summary of the magnitude of change in GABA responses for stimuli from −70 mV and −90 mV. *p < 0.05 (K-S test).

Figure 6. L- and R-Type Ca²⁺ Channels Mediate Long-Lasting Depression and Potentiation

(A) Voltage pulse stimuli from −70 mV produced long-lasting potentiation instead of depression of GABA responses (traces 1 and 2, open circles, n = 7) in the presence of the L-type Ca²⁺ channel blocker nifedipine (20 μM). In the presence of a mixture of nifedipine and SNX-482 (500 nM), an R-type Ca²⁺ channel blocker (Mixture), the same stimuli failed to induce long-lasting changes (traces 3 and 4, open squares, n = 7). (B) Voltage pulse stimuli from −90 mV produced long-lasting depression instead of potentiation (traces 1 and 2, filled squares, n = 5) in the presence of SNX-482 (500 nM). In the presence of the mixture of two blockers, the same stimuli never produced long-lasting changes (traces 3 and 4, filled circles, n = 6). Data of the time course graphs in (A) and (B) are expressed as mean ± SEM. (C) Summary of the magnitude of change for stimuli started from −70 mV in the presence of nifedipine and the mixture of nifedipine and SNX-482 and those from −90 mV in the presence of SNX-482 and the same mixture. *p < 0.05 (K–S test).

Figure 7. Long-Lasting Depression and Potentiation of GABA Responses Result from Endocytosis and Exocytosis of GABAA Receptors

(A) Voltage pulse stimuli from −70 mV produced long-lasting depression and an increase in CV (filled bars), whereas those from −90 mV produced long-lasting potentiation and a decrease in CV (open bars). *p < 0.05 (paired t test). Data are expressed as mean ± SEM. (B) Voltage pulse stimuli from −70 mV produced long-lasting potentiation in P4-loaded neurons (traces 1 and 2, open squares, n = 5), but long-lasting depression in scrambled P4-loaded neurons (traces 3 and 4, open circles, n = 9). (C) Voltage pulse stimuli from −90 mV produced long-lasting depression in BoNT/D-loaded neurons (traces 1 and 2, filled squares, n = 9), but long-lasting potentiation in heat-inactivated BoNT/D-loaded neurons (traces 3 and 4, filled circles, n = 9). Data of the time course graphs in (B) and (C) are expressed as mean ± SEM. (D) Summary of the magnitude of change in GABA responses for P4, scrambled P4, BoNT/D, and heat-inactivated BoNT/D. *p < 0.05 (K–S test).

Figure 8. Schematic Drawing of the Molecular Mechanisms for State-Dependent Bidirectional Modification of Somatic Inhibition

During arousal, pyramidal neurons maintain a depolarized membrane potential and fire repetitively, which activates L-type voltage-gated Ca²⁺ channels (VGCCs) effectively, but R-type Ca²⁺ channels only slightly, because the latter channels are inactivated in this condition. Calcium influx through L-type Ca²⁺ channels initiates endocytosis of GABA_A receptors, which removes those receptors from the postsynaptic membrane and results in long-lasting depression of somatic inhibition. During slow-wave sleep, pyramidal neurons undergo slow membrane potential oscillations and fire on the depolarizing phase of the oscillations, which activates R-type Ca²⁺ channels in addition to L-type Ca²⁺ channels. Calcium influx through R-type Ca²⁺ channels initiates exocytosis of GABA_A receptors, which inserts those receptors into the postsynaptic membrane. In this condition, R-type Ca²⁺ channel-mediated insertion of GABA_A receptors may surpass L-type channel-mediated removal of those receptors, leading to long-lasting potentiation of somatic inhibition.