Vessel pruning is accompanied with endothelial cell (EC) apoptosis in the brain of larval zebrafish. (A) In vivo time-lapse confocal images showing that an EC (arrows) underwent apoptosis on a pruned brain vessel in a Tg(fli1a:nEGFP);Tg(fli1a.ep:DsRedEx) larva at 3–3.5 days post-fertilization (dpf), in which EC nuclei were labeled by both enhanced green fluorescent protein (EGFP) and DsRed (yellow). Left, projected confocal image of the whole brain vasculature; Right, time-lapse confocal images of the dashed outlined area in the left. (B) Summary of the percentages of EC apoptosis- and migration-accompanied brain vessel pruning (n = 17 larvae). (C) In vivo time-lapse confocal images showing that an EC (arrows) underwent apoptosis on a brain pruned vessel in a DAPI-injected Tg(kdrl:EGFP) larva at 3–3.5 dpf, in which EC nuclei were labeled by DAPI (blue). (D) Immunofluorescence images showing that, on a pruned vessel, an EC with typical apoptotic morphology (arrows) expressed Caspase-3. The numbers on the bars (B) represent the number of pruned vessels examined. Scales: 50 μm (left in (A), 15 μm (right in (A), 50 μm (C) and 15 μm (D).

Characterization of EC apoptosis-accompanied pruned vessels. (A) In vivo time-lapse confocal images showing the relative spatial locations of EC nuclei in pruned and adjacent brain vessels. Top: for an EC apoptosis-accompanied pruned brain vessel, the nuclei of two neighboring vessels’ ECs (red arrowheads) located at the both ends of the pruned vessel, respectively (white arrow). Bottom: for an EC migration-accompanied pruned brain vessel, the nuclei of neighboring vessels’ ECs (red arrowheads) did not occupy the ends of the pruned vessel (white arrow). (B) Summary of data showing that EC apoptosis-accompanied pruning vessels have a higher probability that neighboring vessels’ EC nuclei occupy their both ends (7 out of 9, n = 7 larvae) than EC migration-accompanied pruning vessels (3 out of 11, n = 6 larvae). (C) Length of pruned vessels with EC apoptosis (n = 9 larvae) or EC migration (n = 13 larvae). The numbers on the bars (B) or in the brackets (C) represent the number of pruned vessels examined. Data are shown as mean ± SEM. **p < 0.01 (two-tailed unpaired Student’s t-test). Scale bar: 10 μm (A).

Role of microglia in EC apoptosis-accompanied brain vessel pruning. (A) In vivo time-lapse confocal images showing that, during EC apoptosis-accompanied vessel pruning, a microglial cell (red) migrated to, engulfed and cleaned the apoptotic EC (white arrow). The Tg(coro1a:DsRedEx);Tg(kdrl:EGFP) larvae at 3–3.5 dpf were used. (B) In vivo time-lapse confocal images showing that, during EC migration-accompanied vessel pruning (white arrow), there was no obvious interaction between microglia (red) and the migrating EC (white arrowhead). (C) Representative projected confocal images (left) and summary data (right) showing that knockdown of pu.1 significantly diminished the number of microglia (red) in the brain at 3 dpf. (D) In vivo time-lapse confocal images showing that EC apoptosis-accompanied brain vessel pruning (arrow) still occurred in pu.1 morphants. (E) Summary of the percentages of EC apoptosis- and migration-accompanied brain vessel pruning from pu.1 MO-injected Tg(coro1a:DsRedEx);Tg(kdrl:EGFP) (n = 30 larvae). The numbers on the bars represent the number of animals (C) or pruned vessels (E) examined. Data are shown as mean ± SEM. ***p < 0.001 (two-tailed unpaired Student’s t-test). Scales: 20 μm (A,B), 50 μm (C) and 15 μm (D).