Currents in space: understanding inhibitory field potentials

Abstract

The reports of Caton (1875) and Berger (1929) provided the earliest suggestions that the electroencephalogram (EEG) carries information about the brain's activity. However, while recordings of the EEG and the extracellular local field potential (LFP) have since become routine, the detailed mechanisms of their generation have not been completely elucidated. The LFP is created by the sum of currents flowing into and out of cells (sinks and sources respectively; see Fig. 1) and across the extracellular space. Unlike intracellularly recorded potentials, which are filtered by the plasma membrane, the LFP correlates closely with these extracellular currents. Thus, in order to understand the generation of the LFP signal, it falls to the experimentalist to identify the current sinks and sources that generate the field. Figure 1 Synaptic input generates current sinks and sources along somato-dendritic axis In a structure such as the hippocampus, the alignment of neurons causes evoked synaptic currents in a population of cells to be reflected in a field postsynaptic potential through the summation of extracellular currents. Recently, Glickfeld and colleagues (2009) reported detecting field IPSPs elicited by the firing of an individual inhibitory interneuron. They recorded such unitary field potentials using a single extracellular electrode. To gain a more complete understanding of the generation of unitary field potentials, the flow of currents in space needs to be considered. In this issue of The Journal of Physiology, Bazelot and colleagues (2010) significantly enhance our understanding of the unitary LFP by investigating the spatial flow and circuit origin of the signal. The authors recorded spontaneous activity in the CA3 region of hippocampal slices using extracellular electrodes. They report fast negative going action potentials and slow positive going events, which were identified as field IPSPs based on pharmacological manipulation and simultaneous extracellular and whole-cell recordings. Excitatory currents were not found to contribute to these slow events. The authors went on to investigate the specific network origin of these events. Whilst evoking action potentials in individual interneurons, they recorded the LFP at various positions along the CA3 stratum pyramidale concomitantly. The recordings confirmed that they were able to detect unitary events evoked by the firing of individual interneurons and revealed that the amplitude of the unitary field IPSP fell with distance. Anatomical reconstruction of the interneuron showed that events were only detected in regions spanned by the interneuronal axon. Thus, the unique axonal arborisations of interneurons should result in a unique spatial profile of unitary field IPSPs. This hypothesis was tested using multi-electrode arrays placed either along or orthogonal to the pyramidal cell somato-dendritic axis. These recordings allowed the authors to examine the changes of the unitary LFP signal in space, and thus identify sinks and sources of current. This analysis in combination with cluster analysis made it possible for the authors to distinguish the activity of individual interneurons, and to show that fields generated by perisomatic inhibition (Fig. 1B) differ from those generated by inhibition arriving at the distal dendrites (Fig. 1C). The paper highlights three important issues. Firstly, the paper shows that inhibitory, but not excitatory, synaptic currents can generate detectable unitary field potentials. It has long been thought that the EEG signal predominantly reflects excitatory currents. Circumstantial evidence has suggested that inhibitory currents contribute to the LFP signal during fast network oscillations (Trevelyan, 2009; Oren et al. 2010). Now Bazelot et al. show directly that inhibitory GABAergic currents generate a detectable current source in the hippocampus. Secondly, the paper draws attention to the importance of considering the vector flow of currents in understanding the generation of the LFP. Since the net potential at a point reflects the sum of currents flowing into and out of the point, a zero change in potential could reflect either the absence of current sinks and sources, or a zero sum of sinks and sources. Current source density analysis of multielectrode array recordings examines the changes in current over space and allows for conclusive identification of current sinks and sources. Thirdly, the successful identification of distinctive spatial profiles of the LFP generated by different types of interneuron has potential applications in the analysis of in vivo LFP recordings. Bazelot et al. have demonstrated that it is possible to identify individual interneurons by analysing inhibitory field potentials. Such analysis could enhance the analysis of the in vivo LFP by allowing for the identification of interneuron subtypes based on the location of the inhibitory current source. In summary, the work of Bazelot et al. expands both our understanding of the generation of the LFPs and the tools available for their analysis.