Analytical approaches to examine gamma-aminobutyric acid and glutamate vesicular co-packaging

Abstract

Multi-transmitter neurons, i.e., those that release more than one type of neurotransmitter, have been found in many organisms and brain areas. Given the peculiar biology of these cells, as well as the potential for diverse effects of each of the transmitters released, new tools, and approaches are necessary to parse the mechanisms and functions of synaptic co-transmission. Recently, we and others have studied neurons that project to the lateral habenula and release both gamma-aminobutyric acid (GABA) and glutamate, in some cases by packaging both transmitters in the same synaptic vesicles. Here, we discuss the main challenges with current electrophysiological approaches to studying the mechanisms of glutamate/GABA co-release, a novel statistical analysis that can identify co-packaging of neurotransmitters versus release from separate vesicle, and the implications of glutamate/GABA co-release for synapse function and plasticity.

Keywords: GABA; co-transmission; electrophysiology; glutamate; statistical analysis.

FIGURE 1

Post-synaptic currents arising from glutamate/gamma-aminobutyric acid (GABA) co-release. (A) Composite signal of glutamate/GABA co-release is biphasic. Top, simulated post-synaptic current (PSC) at reversal potential of AMPA-type glutamate receptor (AMPAR) (Vh = 0 mV). At this membrane potential, net current is inhibitory PSC (IPSC) dominant. Middle, simulated PSC at reversal potential of GABAR (V_h = –70 mV). At this membrane potential, net current is excitatory PSC (EPSC) dominant. Bottom, simulated PSC at intermediate potential of the two receptors (V_h = –35 mV). The net current is biphasic with reduced amplitudes of inward and outward currents. (B) Stochastic vesicle release causes PSC distributions to diverge between “co-packaging” and “independent” models of glutamate/GABA co-release but their normalized average trial responses are identical. Left, two types of PSCs arise from vesicles that co-package glutamate and GABA in individual vesicles/both and failure modes. Right, four types of PSCs arise from vesicles that independently package glutamate and GABA/ IPSC, EPSC, both and failure modes. Bottom, averaging across trials under the two modes of co-release result in a biphasic PSC. (C) Schematic showing the actions of co-activated pre-synaptic boutons that differentially co-release glutamate/GABA converging onto a same post-synaptic cell. Pooled synaptic response detected at the soma of a downstream neuron reflects a biphasic response although activating the “independent” site may transmit only an EPSC.

FIGURE 2

Single vesicle post-synaptic current dynamics analysis. (A) Classifications of simulated post-synaptic currents (PSCs) under a low noise condition. First, individual trials of PSCs can be sorted into failure and success modes. Next, individual success trials of PSCs can be categorized to either biphasic or uniphasic modes. Neurotransmitter co-release by segregated vesicles predicts heterogeneity of the PSCs, containing both uniphasic and biphasic events. Co-release by co-packaging vesicles predicts presence of only biphasic success events. (B1,B2) Three statistical metrics for distinguishing co-release models and their application to experimental data analysis. Data adapted from Kim et al. (2022). (B1) Top, scatterplots of the maximum and minimum amplitudes of simulated PSCs (200 trials) from independent (left) and co-packaging (middle) release models compared to that of an example synapse (98 trials) (right). Amplitudes are normalized to the average maximum (y-axis) and minimum (x-axis) amplitudes of success trials. Middle, bar graphs of probabilities of trials with failures of both PSCs (green), success of excitatory PSC (EPSC) only (pink), success of inhibitory PSC (IPSC) only (blue), and success of both PSCs (purple). Bottom, histograms of the maximum (blue) and minimum (red) amplitudes with success of release and failures of release (gray). (B2) Analysis of the three metrics are compared between simulated dataset from independent (left) and co-packaging (middle) release models and an example synapse dataset (right). Top, Metric 1 determines if glutamate and GABA currents occur in individual trials more frequently than that expected by chance. Bootstrapped probability of presence of both EPSCs and IPSCs (purple) and multiples of individual EPSC and IPSC probabilities (gray) are shown. Middle, Metric 2 determines whether conditions of presence of absence of one current influence the distribution of the other current. Cumulative distribution functions of IPSC amplitude (i_max, blue) conditional on the presence [i_max (E), solid] or the absence [i_max (no E), dashed] of EPSC. Similar analyses were performed on the EPSC amplitude [i_min, red; i_min (I), solid; i_min (no I), dashed]. Bottom, Metric 3 determines if glutamate and GABA current amplitudes are correlated trial-by-trial. Bootstrapped correlation of a pair of EPSC and IPSC amplitudes for all trials (dark green), success trials (light green), and all trials with pairing order shuffled (gray). (C) Temporal variance of biphasic PSCs. Minimum (red dot) and maximum (blue dot) amplitude peaks were extracted from the analysis time window and their timings are indicated by t_min and t_max, respectively. Δt is the time difference between t_min and t_max within a trial.