Computational protocol: An amplified promoter system for targeted expression of calcium indicator proteins in the cerebellar cortex

Similar protocols

Protocol publication

[…] Microscopy was performed using a custom built two-photon microscope (MOM, Sutter) based on the design of Winfried Denk and running ScanImage (Pologruto et al., ). The excitation wavelength was 850 nm of a Ti:sapphire laser (MaiTai, Spectra Physics). A 20×/1.0 N.A. water immersion objective (Zeiss) and two GaAsP photomultiplier tubes (Hamamatsu) were used. Fluorescence was detected simultaneously in the green (520–600 nm bandpass filter, Chroma) and blue (460–500 nm bandpass filter, Chroma) wavelength range of cpVenus and cyan fluorescent protein (CFP), respectively. Laser power for functional imaging of PCs and molecular layer interneurons was up to 50 mW, and granular layer interneurons up to 100 mW. The ratio of cpVenus and CFP fluorescence was used to analyze functional signals. Ratiometric imaging was done to eliminate movement artifacts when recording from awake animals. At the beginning of the imaging session, animals were anesthetized with 1.5% isoflurane in O2 and then headfixed under the microscope. Stacks for three-dimensional reconstruction were acquired under anesthesia (1.0–1.5% isoflurane) with 512 by 512 pixels and 4 or 9 averages. Animals for analysis of expression time course were not used for functional imaging to avoid potential damage by bleaching or phototoxicity. For the time courses stacks were acquired every 2 and 3 days for interneurons and PCs, respectively. Functional imaging was done under awake conditions sitting on a platform with isoflurane/O2 flow off for more than 5 min. Behavior was recorded with an infrared camera (DCR-DVD308 DVD Handycam, Sony) and infrared illumination (HVL-HIRL Video/IR Combo Light, Sony). Two-photon movies were acquired with 128 by 128 pixels and 2 ms per line (256 ms/frame, a rate of 3.9 Hz) and data were analyzed with MATLAB and ImageJ.Masks of PC dendrites were found using spatial independent component analysis (sICA) implemented with the FastICA algorithm (Hyvarinen and Oja, ). The movies were first converted to ΔR/R movies by calculating and registering time dependent ΔR/R values for each pixel. sICA was run on these ΔR/R movies with 10 (for movies with up to 5 dendrites) or 50 (for movies more than 10 dendrites) spatial components where the contrast function was a Gaussian approximation to negentropy (Hyvarinen and Oja, ). This procedure separated individual dendrites to individual sICA components. However, not all the sICA components carried a dendrite. sICA components carrying a dendrite were determined by visual inspection. Usually there was no ambiguity in this procedure since such sICA components consisted of a clear group of high intensity pixels in the shape of a dendrite (Ozden et al., ) over a low intensity background. To extract the mask of a dendrite from its sICA component, the sICA component was thresholded at a level (determined manually) so that bright pixels belonging to the dendrite is set to 1 (above threshold) and the rest to 0 (below threshold). The threshold values varied significantly between movies and depended on the signal-to-noise ratio. However, the final dendrite masks obtained with this method had good separation for individual dendrites with little overlap between them. […]

Pipeline specifications

Software tools ScanImage, ImageJ, fastICA
Applications Miscellaneous, Laser scanning microscopy, Microscopic phenotype analysis
Organisms Canine parvovirus, Mus musculus, Homo sapiens
Diseases Drug-Related Side Effects and Adverse Reactions
Chemicals Calcium, Mannitol, Tetracycline