Fig. 5. Workflow for automated detection and analysis of SynTagMA photoconversion.

a Green and red fluorescence is collected in alternate frames, switching between two Ti/Sapph lasers (980 nm/1040 nm). Images are registered in 3D to correct for chromatic aberration and laser alignment. b Median filtering and deconvolution is then applied to all images (both green and red channels). c To superimpose multiple time points in 3D, translation, rotation and unwarping are applied. d Synapses (ROIs) are detected as spherical objects, i.e., ‘spots’ from which fluorescence values are extracted and analyzed. e Maximum intensity projection of a CA1 pyramidal cell expressing cytosolic mCerulean (inverted gray scale) and postSynTagMA. The cell was stimulated with 50 bAPs at 100 Hz and illuminated with 395 nm (as in Fig. 4). For each identified synapse, ΔR/(G₀ + G₁) was analyzed, color-coded, and plotted at its original location. Distance from soma is indicated as concentric rings. f Photoconversion decreased exponentially from the soma with a distance constant of λ = 105 µm (median ± interquartile range, n = 1860 synapses, 1 cell, R² = 0.91). This experiment was reproduced in 4 neurons. g Spine calcium transient amplitudes during 50 bAP trains (jGCaMP7b) decreased exponentially with distance from the soma with λ = 106 µm (n = 55 synapses, 5 neurons, R² = 0.66). Scale bar: 4 µm (a, b, c), 30 µm (e).