Astrocyte membrane responses and potassium accumulation during neuronal activity.

Older studies suggest that astrocytes act as potassium electrodes and depolarize with the potassium efflux accompanying neuronal activity. Newer studies suggest that astrocytes depolarize in response to neuronal glutamate release and the activity of electrogenic glial glutamate transporters, thus casting doubt on the fidelity with which astrocytes might sense extracellular potassium rises. Any K(+)-induced astrocyte depolarization might reflect a spatial buffering effect of astrocytes during neuronal activity. For these reasons, we studied stimulus-evoked currents in hippocampal CA1 astrocytes. Hippocampal astrocytes exhibited stimulus-evoked transient glutamate transporter currents and slower Ba(2+)-sensitive inward rectifier potassium (K(ir)) currents. In whole-cell astrocyte recordings, Ba(2+) blocked a very weakly rectifying component of the astrocyte membrane conductance. The slow stimulus-elicited current, like measurements from K(+)-sensitive electrodes under the same conditions, predicted small bulk K(+) increases (<0.5 mM) following the termination of short-stimulus trains. These currents indicate the potential for astrocyte spatial K(+) buffering. However, Ba(2+) did not significantly affect resting K(+) or the K(+) rises detected by the K(+)-sensitive electrode. To test whether local K(+) rises may be significantly higher than those detected by glial recordings or by K(+) electrodes, we assayed EPSCs and fiber volleys, two measures very sensitive to K(+) increases. We found that Ba(2+) had little effect on neuronal axonal or synaptic function during short-stimulus trains, indicating that K(ir)s do not influence local K(+) rises enough, under these conditions to affect synaptic transmission. In conclusion, our results indicate that hippocampal astrocytes are faithful sensors of K(+) rises, but we find little evidence for physiologically relevant spatial K(+) buffering during brief bursts of presynaptic activity.