Distinct laminar origins of high-gamma and low-frequency ECoG signals revealed by optogenetics

Electrocorticography (ECoG) provides a high-spatiotemporal-resolution measure of cortical activity (cortical surface electrical potentials, CSEPs) in humans and animals. The CSEP high-gamma band (H{gamma}, 65-170 Hz) correlates with neuronal firing rates at the columnar spatial scale and is widely used as a biomarker of local activity. Whether H{gamma} reports all stages of columnar processing, intermediate processing in L2/3 (close to the ECoG electrode), or the main columnar output in L5, is unknown. We disentangled the laminar origins of H{gamma} and other ECoG bands by optogenetically suppressing L2/3 or L5 pyramidal cells during micro-ECoG recording in mouse somatosensory cortex. Whisker deflections evoked transient, topographically localized CSEPs. L5 optogenetic suppression most strongly reduced 65-450 Hz (H{gamma}-uH{gamma}) bands in sensory-evoked ECoG signals, whereas L2/3 suppression most strongly reduced 4-30 Hz ({theta}-{beta}) bands. Thus, different CSEP frequency bands reflect layer-specific activity and are biomarkers of distinct stages of columnar processing. Significance StatementElectrocorticography (ECoG) is widely used as a mesoscale measure of cortical activity in human and animals, providing a unique methodological from basic neuroscience discovery to understanding the human brain in health and disease. However, the cell types and laminar sources that generate ECoG signals are unknown, which impedes interpretation of ECoG findings in both the clinic and basic research. We disentangled the laminar origins of ECoG frequency bands by optogenetically suppressing L2/3 or L5 pyramidal cells during ECoG recording in mouse somatosensory cortex. We found that L5 optogenetic suppression most strongly reduced high frequencies whereas L2/3 suppression most strongly reduced lower frequencies. Thus, different ECoG frequency bands reflect layer-specific activity and are biomarkers of distinct stages of columnar processing.