Background and motivationLocal field potentials (LFPs) are compound signals comprising synaptic currents and several transmembrane currents from active structures, which represent the dynamic flow of information across the brain. Although LFP analyses have remained largely limited to chemical synaptic inputs, neurons and other cell types also receive gap junctional inputs that play essential roles in neuronal and network physiology. Gap junctional inputs have been historically excluded from LFP analyses because, unlike synaptic receptors, these inputs are not mediated by transmembrane currents that involve the extracellular space. However, the voltage response to gap junctional inputs onto active compartments triggers several transmembrane currents across the neuron. Therefore, two fundamental questions required for enhanced accuracy of LFP interpretations are: (i) Do gap junctional inputs onto active compartments contribute to LFPs? (ii) Are there differences in extracellular signatures associated with gap junctionalvs. chemical synaptic inputs onto active compartments?MethodologyWe built morphologically realistic conductance-based neuronal models and placed a 3D array of extracellular electrodes spanning the somato-dendritic stretch. We employed different types of inputs: (i) synchronous; (ii) random; and (iii) rhythmic (1–128 Hz). We computed LFPs at all electrodes and analyzed the spatiotemporal profiles of intra- and extra-cellular voltages for several model configurations, involving different input types, with activevs. passive dendrites, with gap junctionsvs. chemical synapses, and in the presencevs. absence of different ion channels.ResultsWe demonstrate a striking reversal in the polarity of extracellular potentials associated with synchronous inputs through chemical synapsesvs. gap junctions onto active dendrites. Whereas synchronous inputs through chemical synapses yielded a negative deflection in proximal electrodes, those onto gap junctions manifested a positive deflection. Importantly, we observed extracellular dipoles only when inputs arrived through chemical synapses, but not with gap junctions. Remarkably, the slow hyperpolarization-activation cyclic nucleotide-gated (HCN) channels, which typically conduct inward currents, mediated outward currents triggered by the fast voltage transition caused by synchronous inputs. With random inputs, extracellular potentials in proximal electrodes were largely negative with chemical synapses but were biphasic with gap junctional inputs. Finally, with rhythmic inputs arriving through gap junctions, we found strong suppression of LFP power at higher frequencies. There were frequency-dependent differences in the spike phase associated with the LFP, depending on whether inputs arrived through gap junctions or chemical synapses. LFP differences across all input types were mediated by the relative dominance of synaptic currentsvs. voltage-driven transmembrane currents with chemical synapsesvs. gap junctions, respectively.ImplicationsOur analyses unveil a prominent role for gap junctional connections in shaping the spatiotemporal and spectral profiles of extracellular potentials, with critical implications for polarities and spatial spread of LFPs. The stark dichotomies in extracellular signatures associated with gap junctionalvs. chemical synaptic inputs imply that conclusions could be erroneous if cells were incorrectly assumed to be exclusively receiving chemical synaptic inputs.
