Introduction: Astrocytes in the sensory cortices have access to both sensory and internal (e.g. arousal) signals. The relative contribution of these inputs to astrocytic calcium signalling remains unclear. We therefore sought to establish how calcium signals of astrocytes in the primary visual cortex (V1) of awake mice respond to changes in visual stimulus, and changes in locomotion or arousal.

Methods: We measured population-level astrocytic calcium responses from V1 in 6–12-month-old mice (C57/BJ6 background with humanised MAPT gene), of either sex. Animals were anaesthetized with isoflurane (3% induction, 1.5% maintenance) and prepared for aseptic surgery, a craniotomy was made above V1, and a bar was attached to the skull for subsequent head-fixation. In 8 animals, AAV8-GFAP-GCaMP6f was injected, and a fibre-photometry cannula was implanted above the dura. In 13 animals an electrode was instead implanted in layer 4 to measure visually evoked local field potentials (VEPs). At least 3 weeks after recovery from anaesthesia, astrocytic calcium or neural VEPs were measured in awake head-fixed animals, free to locomote on a wheel, in the presence and absence of brief (0.5 – 1.5s) flashed visual stimuli. Calcium signals (ΔF/F) were calculated as the peak change in fluorescence in the 2-10s following onset of spontaneous locomotion bouts, or appearance of a visual stimulus, relative to that in the 1s prior. All procedures were performed under appropriate project and personal licenses from the UK Home Office.

Results:  Astrocytic calcium increased rapidly at the onset of locomotion (mean±sem: 12.3%±3.8, n=8). A high-contrast visual stimulus also elicited an increase in calcium levels relative to pre-stimulus baseline, but response amplitude depended on the behavioural state of the animal. Response was larger if the animals were already locomoting (movement speed exceeding 1 cm/s; 8.0%±1.9, n=7) than if they were stationary (1.8%±0.4; p =0.016, Wilcoxon Signed Rank). Subsequent analyses were therefore confined to trials when the animal was locomoting. Astrocytic responses increased with stimulus contrast (12% contrast: 2.2%±0.8; 100% contrast: 7.8%±2.2, n=6). Astrocytic responses were visuotopic: small stimuli presented in ipsilateral visual field evoked much lower responses (1.1%±0.6, n=8) than those in an appropriate location in contralateral visual field (6.8%±3.0; p = 0.023, Wilcoxon Signed Rank). Responses to small, flashed squares in the contralateral hemifield suggested presence of receptive fields that were similar in size to those obtained for neural VEPs. 

Conclusion: Astrocytes in mouse V1 respond with increase in calcium levels to the onset of locomotion, or the change in arousal that accompanies locomotion. Astrocytes also respond on presentation of visual stimuli, but that response depends on behavioural state. When an animal is locomoting, astrocytic population response increases with the strength of a visual stimulus and is tuned to the location of that stimulus in the visual field.