Synchronized bursts of miniature inhibitory postsynaptic currents

Abstract

Spike-independent miniature postsynaptic currents are generally stochastic and are therefore not thought to mediate information relay in neuronal circuits. However, we recorded endogenous bursts of IPSCs in hypothalamic magnocellular neurones in the presence of TTX, which implicated a coordinated mechanism of spike-independent GABA release. IPSC bursts were identical in the absence of TTX, although the burst incidence increased 5-fold, indicating that IPSC bursts were composed of miniature IPSCs (mIPSCs), and that the probability of burst generation increased with action potential activity. IPSC bursts required extracellular calcium, although they were not dependent on calcium influx through voltage-gated calcium channels or on calcium mobilization from intracellular stores. Current injections simulating IPSC bursts were capable of triggering and terminating action potential trains. In 25% of dual recordings, a subset of IPSC bursts were highly synchronized in onset in pairs of magnocellular neurones. Synchronized IPSC bursts displayed properties that were consistent with simultaneous release at GABA synapses shared between pairs of postsynaptic magnocellular neurones. Synchronized bursts of inhibitory synaptic inputs represent a novel mechanism that may contribute to the action potential burst generation, termination and synchronization responsible for pulsatile hormone release from neuroendocrine cells.

Figure 1. mIPSC bursts in magnocellular neurones

A, a representative example of an mIPSC burst recorded at −60 mV with a high [Cl⁻] patch solution illustrating the typical abrupt onset with summated mIPSCs and the typical gradual termination. a1 and a2, higher sweep speed details of the onset and termination of the burst in A. The external solution in A contained 1 μm TTX, 50 μm APV and 10 μm CNQX. B, mIPSC bursts were recorded from both putative VP and putative OT neurones (GFP+ and GFP−, respectively). The fluorescence micrograph demonstrates the GFP-specific labelling of VP neurones with an overlay of GFP-expressing VP neurones (green) and immunofluorescence-labelled OT neurones (red) in the same section of SON (monoclonal PS38 antibody to OT-neurophysin kindly supplied by Dr H. Gainer, National Institutes of Health, Bethesda, MD, USA). There was no detectable double labelling of neurones with the two markers. OC, optic chiasm.

Figure 2. IPSC bursts and mIPSC bursts are identical, except in incidence

A–D, representative examples, from 4 cells, of IPSC and mIPSC bursts recorded in normal aCSF (left) and after >10 min in TTX (right) in the same cells. In A and B, holding potential (V_H) was 0 mV. In C and D, V_H was −60 mV and a high [Cl⁻] patch solution was used. Lower traces are higher sweep speed details of segments underlined above. The IPSC burst duration (E) and intra-burst IPSC frequency (F) were the same in control aCSF (C) and aCSF containing TTX (TTX). The incidence of the IPSC bursts (G), on the other hand, was 5-fold greater in aCSF lacking TTX. H, the distribution of IPSC bursts in the cells represented in E–G (x-axis bin size is 2 bursts h⁻¹).

Figure 3. The Ca²⁺ dependence of mIPSC bursts and extra-burst mIPSCs

Exchanging the normal extracellular medium with Ca²⁺-free aCSF abolished all mIPSC bursts in a within-cell comparison (A) and in the same experiment also reduced the frequency of stochastic mIPSCs (B). Inhibiting the SERCA pump with 10 μm thapsigargin (C) or blocking VGCCs with 200 μm Cd²⁺ (D) did not abolish the mIPSC bursts. E, the amplitude of background mIPSCs was not affected by Cd²⁺. F, representative traces from one of the cells in A showing an mIPSC burst in normal aCSF (left, TTX) and the presence of only stochastic mIPSCs after 10 min of perfusion with Ca²⁺-free medium (right, ‘TTX, Ca-free’). G, representative traces from another of the cells in A illustrating the mIPSC frequency in normal aCSF (left) and after 20 min perfusion with Ca²⁺-free medium (right).

Figure 4. Intra-burst mIPSCs have faster rise times and larger amplitudes than extra-burst mIPSCs

A, recording of mIPSCs at −60 mV with high [Cl⁻] patch solution. B,10–90% rise time; C, amplitude; and D, inter-event interval (IEI) of sequential mIPSCs immediately preceding, during and immediately following the mIPSC burst shown in A. Arrows indicate the same time points (the onset and end of the mIPSC burst). x-axis in B–D is event number. B and C contain only events with a flat baseline, whereas D contains the IEIs of all mIPSCs. B–D cover the same time periods. E, the cumulative fraction histogram for the amplitude of burst and background mIPSCs of all 6 cells analysed. F, the amplitude distribution of mIPSCs in the burst (top) and an equal number of mIPSCs preceding the burst illustrated in A–D did not reveal clear multiquantal events.

Figure 7. Modulation of spiking activity by spontaneous IPSP bursts

A, example of a spontaneous IPSP burst (arrow) that immediately precedes and appears to trigger a rebound burst of APs—a detail of which (of boxed area) is shown in (a1). B, example of a spontaneous IPSP burst that appears to terminate an AP burst-like afterdischarge. The afterdischarge of APs was triggered with a 0.5 s depolarizing pulse (as in Fig. 6). b1, detail of boxed area in B showing that the IPSP burst occurred at the end of the AP afterdischarge and appeared to contribute to its termination. C, another example of a spontaneous IPSP burst appearing to terminate a spontaneous train of APs. c1, detail of the boxed area in C, showing the IPSP burst between two spontaneous AP bursts and appearing to terminate the first of the two bursts.

Figure 6. Intracellular current injections mimicking an mIPSC burst terminated AP afterdischarges

A–C, sequential records from one cell showing termination of a burst-like AP afterdischarge by an mIPSC burst current injection (see Methods). A, an intracellular DC depolarizing pulse (40 pA, 0.5 s, up-arrow) triggered 15 APs followed by a prolonged afterdischarge of APs. B, the same pulse was delivered and triggered an AP afterdischarge, but the burst was terminated by the injection of an mIPSC burst current injection during the AP afterdischarge. Inset: expanded view of boxed area. C, the depolarizing pulse was delivered again without the mIPSC burst current injection and again resulted in a prolonged AP afterdischarge. Upper traces are V_m and lower traces I_m. Calibration bars in A also apply to B and C (except inset).

Figure 5. Intracellular current injections mimicking an mIPSC burst triggered APs and AP afterdischarges

A, an mIPSC burst current injection (lower trace) (‘mIPSC burst playback’, see Methods) triggered several APs. B and C, examples of cells responding to mIPSC burst current injections with AP afterdischarges. Upper traces are V_m and lower traces I_m.

Figure 9. Homogeneity of synchronized IPSC bursts

A, example of synchronized bursts (double arrow, onset time difference 0.9 ms) recorded in a pair of MNCs at 0 mV with low [Cl⁻] patch solution and no TTX. Three sequential bursts are given for each cell, illustrating the similarity in duration for the synchronized bursts relative to the unsynchronized bursts. For cell 2, the first burst shown was the first burst of the recording. Unsynchronized bursts are indicated by single vertical arrows. B, the onsets of the synchronized bursts in A shown at higher sweep speed. C, detail of boxed area in B at higher sweep speed shows the IPSC resolution and also lack of synchronization of individual IPSCs.

Figure 8. Synchronized bursts of IPSCs in pairs of MNCs

A, running histogram of frequency of IPSCs recorded simultaneously in 2 cells, spaced ∼100 μm apart, at −60 mV with high [Cl⁻] patch solution. Vertical dashed lines denote synchronized bursts. B, raw traces for portion of histogram in A, containing the last 2 episodes of synchronized bursts. C and D, expanded views of the left and right boxed areas in B, respectively.