Excitatory GABA: How a Correct Observation May Turn Out to be an Experimental Artifact

Abstract

The concept of the excitatory action of GABA during early development is based on data obtained mainly in brain slice recordings. However, in vivo measurements as well as observations made in intact hippocampal preparations indicate that GABA is in fact inhibitory in rodents at early neonatal stages. The apparent excitatory action of GABA seems to stem from cellular injury due to the slicing procedure, which leads to accumulation of intracellular Cl(-) in injured neurons. This procedural artifact was shown to be attenuated through various manipulations such as addition of energy substrates more relevant to the in vivo situation. These observations question the very concept of excitatory GABA in immature neuronal networks.

Keywords: GABA; brain slices; energy substrates; giant depolarizing potentials; in vivo versus in vitro.

Figure 1

(A–C) GABA is depolarizing in the slice preparation and hyperpolarizing in the intact hippocampus. (A) Microelectrode recording from hippocampal neuron in a brain slice from a 4-day-old rat (KCl-containing electrode). Note that bicuculline, a GABA_A receptor antagonist, caused membrane hyperpolarization and inhibition of spontaneous synaptic activity (from Ben-Ari et al., 1989). (B) Whole-cell voltage-clamp recording with a pipette containing a K-gluconate based solution [(Cl) in the pipette was 4.2 mM] from a neuron in the intact rat hippocampus. Note that bicuculline evokes epileptiform discharges (from Khalilov et al., 1997). (C) GABAergic activities observed from isolated intact neonatal (P3) mouse hippocampus as seen by extracellular recordings from the CA3 area. Top: baseline field potentials. Note the absence of electrical activity. Bottom: note the presence of spontaneous activity and epileptiform discharges in the presence of bicuculline (blue line). To achieve better oxygenation of the preparation, a dual-side perfusion chamber and a fluid rate of 15 ml/min were used (from Wong et al., 2005). (D) Lactate without glucose maintains and even augments synaptic function. Top: local field potentials (LFPs) in response to stimulation trains when ACSF contains 10 mM glucose (red) or 10 mM lactate (blue). Bottom: examples of single LFPs at expanded time scale. Note that in the presence of lactate as the sole energy substrate, LFPs are even better maintained than under glucose-only conditions (from Ivanov et al., 2011).

Figure 2

Intracellular Cl concentration and electrical activity strongly depend on the experimental model and conditions. (A) The mean intracellular Cl concentration in neurons at different depth from the surface in the intact hippocampi () and acute hippocampal slice preparations (◯) at P5–P7. Note the highly elevated Cl concentrations in neurons from the surface layers in the slice preparation (Modified from Dzhala et al., 2012). (B) The effects of slicing conditions on intracellular Cl concentration. Mean Cl_i as a function of depth in the hippocampal slices prepared from P5–P7 mice in control ACSF and in a high sucrose solution (Modified from Dzhala et al., 2012). (C–E) Genesis of network events and amplitude of local field potentials strongly depend upon the flow rate of ACSF. (C) Spontaneous network activity recorded at a low flow rate of 1.9 ml/min (left), and a high flow rate of 5.2 ml/min (right). Note sharp wave–ripple activity only at a high flow rate. Juvenile (P14–P20) transverse hippocampal 400–450 μm thick slices from Wistar rats were used here (from Hájos et al., 2009). (D) Examples of local field potentials measured in the same slice and electrode positions at different flow rates. Note the remarkable increase in amplitude when the flow rate is increased. (E) Summary of the dependence of local field potential (LFP) amplitudes on the oxygen levels and perfusion rates. Slices 400 μm thick from P4–P7 Swiss mice (from Ivanov et al., 2011).