Neuronal activity elicits metabolic and vascular responses, during which oxygen is first consumed and then supplied to the tissue via an increase in cerebral blood flow. Understanding the spatial and temporal dynamics of blood and tissue oxygen () responses following neuronal activity is crucial for understanding the physiological basis of functional neuroimaging signals. However, our knowledge is limited because previous measurements have been made at low temporal resolution (>100 ms). Here we recorded at high temporal resolution (1 ms), simultaneously with co-localized field potentials, at several cortical depths from the whisker region of the somatosensory cortex in anaesthetized rats and mice. Stimulation of the whiskers produced rapid, laminar-specific changes in . Positive responses (i.e. increases) were observed in the superficial layers within 50 ms of stimulus onset, faster than previously reported. Negative responses (i.e. decreases) were observed in the deeper layers, with maximal amplitude in layer IV, within 40 ms of stimulus onset. The amplitude of the negative, but not the positive, response correlated with local field potential amplitude. Disruption of neurovascular coupling, via nitric oxide synthase inhibition, abolished positive responses to whisker stimulation in the superficial layers and increased negative responses in all layers. Our data show that responses occur rapidly following neuronal activity and are laminar dependent.

The medial entorhinal cortex (mEC) is a key structure which controls the communication between the hippocampus and the neocortex. During slow-wave sleep, it stands out from other cortical regions by exhibiting persistent activity that outlasts neocortical Up states, decoupling the entorhinal cortex-hippocampal interaction from the neocortex. Here, we compared the mechanisms involved in the maintenance of the Up state in the barrel cortex (BC) and mEC using whole cell recordings in acute mouse brain slices. Bath application of an NMDA receptor antagonist abolished Up states in the BC, and reduced the incidence but not the duration of Up states in the mEC. Conversely, blockade of kainate receptors decreased Up state duration in the mEC, but not in the BC. Voltage clamp recordings demonstrated the presence of a non-NMDA glutamate receptor-mediated slow excitatory postsynaptic current, sensitive to the selective kainate receptor antagonist UBP-302, in layer III neurons of the mEC, which was not observed in the BC. Moreover, we found that kainate receptor-mediated currents assist in recovery back to the Up state membrane potential following a current-induced hyperpolarisation of individual cells in the mEC. Finally, we were able to generate Up state activity in a network model of exponential integrate-and-fire neurons only supported by AMPA and kainate receptor-mediated currents. We propose that synaptic kainate receptors are responsible for the unique properties of mEC Up states.