Two-photon calcium imaging reveals long-term changes in cerebellar granule cell responsiveness following high-frequency mossy fibers stimulation

Recent years have witnessed major improvements in our understanding of the brain thanks to technological developments and a new multidisciplinary approach. Single-cell studies have allowed researchers to gain insights into the detailed characteristics of different neuronal types, while techniques such as functional magnetic resonance imaging (fMRI) enabled an investigation of the global activity of the brain. However, very little is known about the microcircuits function and the way in which each single cell contributes to the final output. In order to investigate the spatio-temporal organization of neuronal activity in local microcircuits, a simultaneous and fast recording from selected multiple single neurons with a single-cell resolution is required. Optical techniques fit this purpose, but a parallel and fast signal detection requires the presence of multiple confocal excitation volumes that cannot be achieved through traditional confocal and two-photon microscopy. The use of a Spatial Light Modulator (SLM) to divide a coherent excitation light in multiple diffraction limited focal points, which can be configured in different patterns, enabled us to perform optical recordings from different neurons simultaneously. We recently developed an SLM-two photon microscope (SLM-2PM) able to resolve the spatio-temporal organization of activity in acute cerebellar slices through simultaneous calcium signals acquisition from multiple granule cells (GrCs) (Gandolfi2014, Pozzi2015). In this work the SLM-2PM was optimized to investigate the effect of a mossy fibers (mfs) high-frequency stimulation protocol (HFS), that is known to induce long-term potentiation and depression (LTP and LTD) at the mossy fiber-granule cell (mf-GrC) synapses (Armano2000, Gall2005, D’Errico2009). After the induction of plasticity we observed long-term changes of GrCs calcium responses, both as potentiation and as depression (CaR-P and CaR-D), with impressive variations in signal amplitude (CaR-P +317.2 ± 28.7% n=30, CaR-D -66.4 ± 1.6% n=22, 30 minutes after HFS). In particular, CaR-P amplitude changes were several times larger than those reported for synaptic currents measurements (D’Angelo1999, Sola2004). To elucidate the underlying mechanisms we performed patch-clamp experiments in loose cell-attached (LCA) configuration. Also in these experiments, neuronal activity showed prolonged changes comparable to those observed in SLM-2PM recordings. We then assessed GrCs electroresponsiveness by comparing results from both SLM-2PM and LCA recordings with those previously obtained in our lab through patch-clamp whole-cell recordings (WCRs) and local field potentials measurements (Armano2000, Sola2004, Gall2005, Mapelli2007, D’Errico2009). The amplitude changes observed in CaR-P resulted compatible to a combination of changes in GrCs intrinsic excitability and in synaptic transmission leading conjointly to a large increase in the probability of action potential (AP) generation (Armano2000). We performed preliminary SLM-2PM experiments on IB2 KO mice (a recent animal model of autism spectrum disorders) to investigate neurotransmission alterations in the granular layer; SLM-2PM results were in agreement with those from WCRs of our lab (Soda et al., manuscript in preparation). Finally, the comparison between the SLM-2PM and the modeling data sets allowed to validate the network model of the cerebellar granular layer activity developed in our lab. The results obtained with the SLM-2PM further assess the Ca2+ role in the expression of long-term plasticity at the mf-GrC synapses and its critical involvement in determining the sign of plasticity. The consistency of SLM-2PM data with previous results obtained through different techniques indicates that this system can be used to investigate the whole activity of the cerebellar cortex network.