Figure 2

In vivo imaging of corpus callosal presynaptic terminals in developing neocortex: longitudinal analysis. (A) Low-magnification images of synaptophysin-tdTomato (white). Blood vessels (dashed yellow lines) served as fiduciary marks, allowing imaging of the same brain volume on successive days. Scale bar, 50 µm. Representative images are shown for two days at the start of imaging (P13 and P15) and the beginning of the third postnatal week (P21 and P23). The images shown are maximum intensity Z-projections of the imaged volume. (B) Longitudinal changes in presynaptic terminal density, normalized to the initial value on the first day of imaging. Each line represents the fold change in bouton density in one imaged region of contralateral cortex. CC synapse density dynamics were variable, even within different fields-of-view in the same brain, as illustrated by the purple and red lines, which were obtained by imaging two distinct CC fields in the same animal. In some cases, presynaptic terminal density was remarkably stable throughout this period of development (blue, from a different mouse than the purple and red lines). (C) Probability histograms of changes in bouton density as animals matured. For each imaged region, data were binned to calculate an average density over the indicated 3-day intervals, then changes in density were calculated from one bin to the next (n = 17 regions from 9 mice, with 1-2 fields imaged per mouse). For example, the top panel presents a histogram of the changes between bin 1 (P13-15) and bin 2 (P16-18) for all regions imaged. Increases in density are represented as positive changes in density, while decreases appear as negative changes in density. Wider bars correspond to wider bins, used to accurately display low probability occurrences. Across all age groups, observed age-dependent changes in density were significant (Z = −3.11, p = 0.0019, two-sided Wilcoxon signed rank test). (D) When the data for all regions were pooled and averaged, presynaptic density appeared to increase over time during the most dynamic time period (P13-17; normalized to P13; n = 10 imaged regions from 5 mice, with 2 regions per mouse; χ²(4, 38) = 5.77, p = 0.2167, Kruskal-Wallis test with Tukey-Kramer multiple comparisons correction).