Extended Data Fig. 8: Additional data related to key findings in Fig. 3.

a, UMAP visualization of all cells coloured by wounded status from Extended Data Fig. 7a. Blow-up of the muLEC cluster boxed. b, UMAP visualization of all cells coloured by the mean mixing metric score per cell phenotype. c, Boxplot showing log-transformed mixing metric scores for each cell-phenotype. Cell number analysed: n = 1862 (AEC), n = 1088 (LEC), n = 644 (VEC), n = 935 (Glia), n = 1025 (muLEC), n = 598 (Neuron), n = 782 (Epithelial), n = 791 (Fibroblast), n = 322 (Macrophage), n = 224 (Neural_crest), n = 160 (Myeloid_progenitors), n = 228 (Photoreceptor), n = 84 (Thrombocyte), n = 55 (Adipocyte), n = 54 (Oligodendrocytes), n = 27 (Neutrophil), total n = 8879 cells. Bounds of box = 25^th - 75^th quantile, centre = median, whiskers extend to +/−1.5 times interquartile range. ANOVA for multiple comparisons, two-sided T-test for comparison to muLEC. corrected for multiple testing using holm method. (ANOVA p value = 0), Comparison to muLEC adjusted p-values: 1.1e-36 (Adipocytes), 8.6e-132 (AEC), 1.6e-114 (Epithelial), 2e-122 (Fibroblast), 1.7e-50 (Glia), 1.7e-117 (LEC), 4.8e-49 (Macrophages), 3.8e-86 (Myeloid_progenitors), 3.3e-129 (Neural_crest), 1.6e-62 (Neuron), 1.1e-70 (Oligodendrocyte), 1.3e-138 (Photoreceptor), 4.2e-25 (Thrombocyte), 3.4e-101 (VEC). ****adjusted p = <0.0001. d, Stacked bar plot showing composition of wounded status for each cell phenotype for Level 01 dataset. e, Lollipop plot showing the number of DEG (adjusted p-val<0.05 and |log2FC | > 0.1) between wounded and control cells for each cell phenotype (muLEC, n = 350). Two-sided Wilcox test, adjusted for multiple correction using Bonferroni method. f, Barplot of GO terms upregulated in muLEC vs LEC in wounded sample, coloured by GO category and sorted by number of genes overlapping with the gene set. Top 10 most overlapping terms are shown (full list in Supplementary Data Table 4). muLECs show the most change of any cell type due to increased endocytic and metabolic gene expression. One-sided Fisher’s exact test, adjusted for multiple correction using Benjamini-Hochberg method (employed through enrichR). padj <0.05 & log2FC > 0.5 (top10). g, Network visualisation of gene ontology (GO) terms (biological processes) upregulated in wounded compared to control muLEC. Each node represents a GO term, and edges reflect the pairwise similarity between nodes. Connected nodes annotated and coloured manually to highlight biological themes. h, Dotplot showing expression of select genes driving GO term enrichment, split by wounded status. I, Confocal projection of muLECs (Tg(lyve1b:ERK-KTRClover)^uom117, green), macrophages (Tg(mpeg1:Gal4ff)^gl25; Tg(14xUAS:Nfsb-mCherry)^c264, red) and 10 kDa Dextran Alexa405 (blue) in empty-liposome (top) and clodronate-liposome injected (bottom) 7 dpf zebrafish heads (lateral view). j, Quantification of total number of macrophages and muLECs in empty control-liposome (green) and clodronate-liposome injected (orange) 7 dpf zebrafish (n = 13 control and n = 21 clodronate) from i. * 0.0235, **** <0.0001. k, Confocal projection of muLECs (Tg(−5.2lyve1b:DsRed)^nz101, orange) in 7 dpf siblings (left) and osr2 mutants (right) zebrafish brains, 24 h post wounding (dotted white circle) to the left midbrain. l, Quantification of average length of muLECs of siblings (green) and osr2 mutants (blue) (n = 39 sibling and n = 21 mutant embryos) in both uninjured and injured hemispheres (as indicated) from k. **** <0.0001. m, Confocal projections of meningeal blood vessels (Tg(kdrl:EGFP)^s843, green) in flt4^−/− mutants. muLECs (Tg(−5.2lyve1b:DsRed)^nz101, magenta) shown inset. n, Quantification of meningeal vessel density in control, flt4 mutants and ccbe1 morphants from Fig. 3g and m. (n = 7 control, n = 8 flt4 mutant and n = 12 ccbe1 MO brains). o, Variance of vascular density per animal (see methods) in control and flt4^um203 mutants (flt4^−/−). Two-sided Levene’s test *p = 0.015 for flt4^−/− (n = 7 control, n = 8 flt4 mutant brains). p, Quantification of brain width in control, flt4 mutants and ccbe1 morphants from (Fig. 3i) (n = 22 control brains, n = 19 flt4 mutant brains and n = 9 ccbe1 morphants). q, Quantification of brain width in control, flt4 mutants and ccbe1 morphants from (Fig. 3k) (n = 8 control brains, n = 6 flt4 mutant brains and n = 9 ccbe1 morphants). r, Confocal projection of muLECs muLECs (Tg(−5.2lyve1b:DsRed)^nz101, magenta) and neutrophils (Tg(lyz:BFP), green) in sibling (top left), flt4^um203 mutants (flt4^−/−, top right) and control morphants (control MO, bottom right), ccbe1 morphants (ccbe1 MO, bottom left) 7 dpf zebrafish brains. s, Quantification of total number of neutrophils populating the zebrafish leptomeninges in siblings, flt4 mutants and control morphants, ccbe1 morphants from (h) (n = 18 siblings brains, n = 25 flt4 mutant brains and n = 14 control morphants, n = 12 ccbe1 morphants). i, k, m, r, Scale bar 100 μm. For 8c: ANOVA for multiple comparisons, two-sided T-test for comparison to muLEC. corrected for multiple testing using holm method. (ANOVA p value = 0), ****p = <0.0001. One-way ANOVA *p = 0.0235 and ****p = <0.0001 for control versus clodronate-injected in j. One-way Analysis of Variance (ANOVA) ****p = <0.0001 for siblings versus osr2 mutants in l. For 8n, p, q: One way ANOVA; multiple comparison made between means of each condition (mean control vs mean flt4 vs mean ccbe1). For 8 s: unpaired two-tailed t-test (siblings vs flt4) and (control MO vs ccbe1 MO). For all dot-plots; size of dots represents proportion of expressing cells and colour represents average log-normalised expression.

Source data