Simultaneous dendritic voltage and calcium imaging and somatic recording from Purkinje neurons in awake mice.

Animals and surgery

All animal procedures were conducted in accordance with guidelines of the Okinawa Institute of Science and Technology Institutional Animal Care and Use Committee in an Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC International)-accredited facility. Cerebellar chronic cranial window surgeries were performed on 2-month-old male C57/BL6 mice, using a 5 mm glass cover slip with silicone access port 31 . The window was positioned to allow imaging within lobule V of the cerebellar vermis and the access port was positioned to allow access to the imagining area via a micropipette for PN labelling and electrical recording. Silver wires were attached to the skull using wire glue and dental cement on either side of and anterior to the cranial window for EEG recording.

Microscope setup

We used a custom-built combined wide-field, two-photon microscope (MOM, Sutter Instruments) with either a ×5/N.A. 0.13 air objective (Zeiss) or a ×25/N.A. 1.05 water immersion objective with 2 mm working distance (Olympus) with ScanImage software 62 . Bright field imaging was performed using a sCMOS camera (PCO.edge, PCO). A femtosecond-pulsed Ti:sapphire laser (Vision II, Coherent), circularly polarized and under-filling the back focal plane of the ×25 objective, was used to excite fluorescence (typical laser power at 1020 nm: 60 mW), which was detected by two GaAsP photomultiplier tubes (Hamamatsu) in the spectral range of 490–550 nm (green: GCaMP6f) and 550–750 nm (red: ANNINE-6plus). The mice were headfixed on a platform that consisted of a vertically rotating treadmill, head-plate stage and micromanipulator tower (Sutter Instruments), all mounted on a horizontally rotating stage (8MR190-90-4247, Standa). The sCMOS camera and infrared light source were used to record behavioural activity during recording at 37.3 fps. A second infrared video camera (Sony) was used to monitor mouse behaviour throughout the experiment. An EEG amplifier (Sigmann Elektronik) was mounted on the platform and connected to silver wires attached to the mouse skull during surgery to monitor EEG activity during awake and anaesthetized conditions. The micromanipulator (M-285, Sutter instruments) was used for GCaMP6f virus (UPenn Vector Core) and ANNINE-6plus dye (www.sensitivefarbstoffe.de, Dr. Hinner and Dr.

Hübener Sensitive Dyes

GbR) injection and also for electrophysiological recording.

AAV injections

One week following surgery mice were anaesthetized (1−2% Isoflurane) and head mounted for two-photon guided injection of the adeno-associated viral vectors (AAVs) into the PN layer approximately 150 μm below the dura. For this beveled quartz electrodes (0.7 mm ID, pulled and beveled to 10−20 μm tip diameter) containing AAV1.hSyn. Cre (2E13 GC/ml), AAV1.CAG.Flex. GCaMP6f (1.3E13 GC/ml), and 50 μM FITC in PBS at a ratio of 1:1:1 were used to specifically target PNs and visually control the position and size of the viral injection, while 12 h). After ~20 h, the brightest labelled cell was selected for imaging experiments, which were performed the day following PN labelling. This also guaranteed that the PN was healthy and had not been damaged by the labelling procedure. Where possible, several PNs were labelled in the same mouse (up to 5) and used for simultaneous dendritic voltage and calcium imaging recordings. Between neuron labelling sessions and following surgery, mice were returned to their cages and allowed to recover. Simultaneous dendritic voltage and calcium imaging Approximately 20 h after ANNINE-6plus labelling, extracellular recordings were performed on double labelled PNs with simultaneous imaging from distal dendrites (40 s of awake recording in 10 s epochs, during which time we saw no photo-bleaching of ANNINE-6plus. After at least 40 s of data had been collected and well before any photo-damage was observed (maximum recording time for a single PN was about 4 min), mice were lightly anaesthetized and a z-stack image reconstruction was made with the extracellular recording electrode still in place to reconstruct the labelled PN. In 15 PNs (7 with somatic recording and 8 without) recordings were repeated while the mouse was anaesthetized (1% isoflurane) (Supplementary Fig. 1 ) and 3D reconstructions of labelled PNs were made (Fig. 1e , Supplementary Figs. 5 - 7 ). Pharmacology In a separate set of experiments (11 PNs), the extracellular recording electrode was replaced by a micropipette used for pharmacological manipulation. In this case, the micropipette: a beveled quartz micropipette (0.7 mm ID, Sutter Instruments) was placed ~50 μm below the PN soma and the drug (0.2% lidocaine (Sigma); to block APs, or 100 μM CNQX disodium salt (Tocris) dissolved in saline; to block excitatory synaptic input), was applied by pressure injection (