Computational protocol: Dendritic NMDA spikes are necessary for timing-dependent associative LTP in CA3 pyramidal cells

Similar protocols

Protocol publication

[…] CA3 pyramidal neurons were loaded with Fluo-5F (100 μM) and Alexa Fluor 495 (10 μM; Molecular Probes) through the recording pipette for at least 20 min before two-photon imaging. Neurons were imaged using a two-photon microscope (Scientifica) equipped with a Ti:sapphire laser (Tsunami, Spectra Physics) tuned to 840 nm and a 40 × water-immersion objective lens (0.8 NA, Olympus). Laser power under the objective was typically between 10 and 15 mW. Fluorescence was detected using two photomultiplier tubes using 525/50 nm (green channel) and 620/60 nm (red channel) emission filters. Scanning and image acquisition were controlled by HelioScan software. Time series of Fluo-5F fluorescence images were acquired at 10 Hz with 100 × 100 pixel resolution across imaging fields encompassing neuronal dendrites. The dimensions of a field of view (FOV) were 30 × 30 μm.For the imaging experiments, the ITDP protocol consisted of 60 pairings. As it was of interest to compare responses between different FOVs, images were obtained sequentially from each FOV resulting in data sets consisting of 30 responses when 2 FOVs were selected or 20 responses when 3 FOVs were selected. Simultaneous data acquisition from more than 1 FOV would have resulted in images with insufficient resolution for detailed analysis. In this respect, it should be noted that we previously found no change in the probability of evoking an NMDA spike over the time course of the 60 pairings ( in Brandalise and Gerber, 2014).At the end of each experiment, a z-stack of the fluorescently labelled CA3 pyramidal cell was acquired. Dye which inadvertently leaked from the recording pipette was sometimes taken up by surrounding glia cells. Data were analysed with NIH ImageJ and Igor Pro (WaveMetrics) software. In ImageJ, dendritic segments were manually selected as regions of interest (ROIs). Ca2+ signals were expressed as ΔF/F=(F−F0)/F0 where F and baseline F0 represent mean fluorescence values in an ROI. A Ca2+ transient was accepted as a signal when its amplitude was greater than two times the s.d. of the noise. The onset of the evoked Ca2+ transients was defined by the timing of the electrical stimulation (0.1 Hz). Ca2+ transient integrals for ROIs were calculated in MATLAB as temporal integrals over a 2 s post-stimulus time window (units of ‘%s'; after subtraction of the mean ΔF/F in a 1 s pre-stimulus baseline window). Collapsed z-stack overview images were used to measure the distance from dendritic branches to the recording electrode in the soma. For the analysis of the frequency of Ca2+ transients associated with NMDA spikes, as well as their amplitudes and integrals, ROIs were selected such that they enclosed an entire dendritic segment, hence the area covered by ROIs was variable ( and ). For detailed analysis of the spatial extent of the Ca2+ transients, a series of uniformly sized ROIs (∼1 × 1 μm) were positioned along dendrites exhibiting responses as identified by heat maps ( and ). […]

Pipeline specifications

Software tools HelioScan, ImageJ
Applications Laser scanning microscopy, Microscopic phenotype analysis
Chemicals Calcium, N-Methylaspartate