This site uses cookies, so that our service can work better. I understand

Andrea Volterra

Closing lecture

Research Group of Andrea Volterra, Department of Fundamental Neuroscience, University of Lausanne, Lausanne, Switzerland




Towards decoding the language of astrocyte-neuron-vascular communication via 1D-to-3D Ca2+ imaging

Astrocytes sense neuronal inputs and in turn can modulate synapses and blood vessels via intracellular Ca2+ signaling. Despite their importance, the properties of astrocytic Ca2+ signals are not well understood and led in the last decade to controversial observations and a hot debate as to the real role of astrocytes in synaptic and vascular functions. One of the key problems is that astrocytic Ca2+ dynamics so far have been studied with conventional 2D imaging, which we find monitors at maximum ~5% of an astrocyte volume and ~10% of its activity. To overcome this problem, we developed 3D Ca2+ imaging of entire individual astrocytes in adult hippocampal slices and in vivo in awake mice using the genetically-encoded Ca2+ indicator GCaMP6f (Bindocci et al., Science, 2017). We were able to routinely acquire two full z-stacks of ∼60 focal planes/sec, comprising the whole volume of an individual astrocyte, and captured all Ca2+ events lasting ≥1.5 sec (FWHM). We also visualized faster events in 10Hz acquisitions on smaller volumes (8 planes/sec). Concerning distribution, 80% of an astrocyte Ca2+ activity was localized in the optically sub-resolved part of the cell (the so-called gliapil), where most of the intermingling with synapses is present, with the remaining 20% being present in the structural core (soma, stem processes, end-feet). Core activity was mostly in processes and end-feet, whereas somatic activity was infrequent and not representative of the astrocyte Ca2+ dynamics. This is an important observation because the astrocyte role in synaptic and vascular functions has been classically interpreted based on recordings of just somatic Ca2+ activity. Moreover, 10Hz acquisitions showed that activity in processes is mainly fast (FWHM:~0.7 sec) and local (~40 um3, ~12% of the process volume), resembling in size the events produced by minimal axonal stimulations (~60 um3), therefore possibly containing a large component generated by synaptic inputs. Likewise, Ca2+ activity in end-feet is mostly confined to the end-foot domain itself, is fast (FWHM: 0.75 sec) and often occurs asynchronously in different end-feet of an individual astrocyte, even when they cover contiguous segments of the same vessel. These data imply that astrocytes possess distinct signaling domains and that local signaling is a primary component of their processing, with most of the activity remaining highly compartmentalized under normal conditions. Further, combining fast 1-to-3D imaging and several synthetic and genetically encoded Ca2+ indicators, we developed imaging conditions allowing us to consistently visualize and study the population of fast and local Ca2+ “events” in astrocytic processes. We found that these small and fast events represent the vast majority of all Ca2+ transients detected in a 3D astrocyte. In IP3R2ko mice, originally described as fully devoid of astrocytic Ca2+ events, the fast astrocytic Ca2+ activity was reduced (in both number and amplitude) but not at all abolished, in line with the more recent view that Ca2+ dynamics in astrocytes depend on multiple and mutually interactive Ca2+ sources. This observation is important in view of the “negative” data in IP3R2ko concerning synaptic, vascular and behavioral relevance of astrocytic Ca2+activity. Therefore, all the past IP3R2ko data need revision in light of the new findings. Overall, our work provides a novel state-ofthe-artfor correctly studying the biology of astrocytes and their communications with neurons and the vascular cells in physiology and disease.

Supported by grants: ERC Advanced "Astromnesis" and SNSF 31003A-173124 to AV.