MG Identity Is Conserved in the Zebrafish

(A) Fluorescence of GFP (left panel) and DsRed (middle panel) in the brain of adult cd45:DsRed; mhc2dab:GFP double-transgenic fish. The right panel shows a merge of both fluorescent channels (n = 3). Scale bar: 10 μm.

(B) Flow cytometry analysis on brain cell suspensions from adult Tg(cd45:DsRed; mhc2dab:GFP) identifying cd45:DsRed^lowmhc2dab:GFP⁺ MG (red gate, MG) and cd45:DsRed^highmhc2dab:GFP⁺ CNS-associated macrophages (blue gate, MFs).

(C) Comparison of the relative expression of apoeb, p2ry12, csf1ra, C1qc, csf3r, mpeg1, pu.1/spi1b, and cd45/ptprc transcripts between MG (red bars) and MFs (blue bars) cell subsets isolated by FACS. Error bars represent mean ± SD (n = 3).

(D and E) Flow cytometry of a brain cell suspension from adult mpeg1:GFP transgenics (D) and expression of apoeb, p2ry12, csf1ra, and mpeg1 by qPCR in FACS-sorted mpeg1:GFP⁺ cell subsets (E). Error bars represent mean ± SD (n = 3).

(F) GFP immunohistochemistry on adult mpeg1:GFP brain sections. Scale bar: 25 μm.

(G) apoeb expression in mpeg1:GFP⁺ cells sorted at 1, 2, and 3 dpf. MG, microglia; MFs, macrophages.

pMFs Are the Unique Source of MG in the Vertebrate Embryo

(A–C) Expression of the MG marker apoeb (red quadrants) in the brains of cmyb^null (A), tfec^null (B), alk8^null(lost-a-fin) (C), and their sibling control animals at the indicated developmental stages. For (A), cmyb^null embryos are identified by the concomitant absence of lymphoid-specific rag1 (arrowheads) transcripts in the thymus. Scale bar: 50 μm.

(A′–C′) Quantification of apoeb⁺ MG in embryos of indicated genotypes. Each symbol represents a single embryo. Error bars represent mean ± SEM.

(D) Strategy for transient fate mapping of pMFs in vivo.

(E) Flow cytometry showing absence of kaede-Red expression in non-photoconverted embryos (left panel), whereas cells expressing both forms of kaede were present and sortable in the photoconverted embryos (right panel).

(F) qPCR analyses for apoeb and csf1ra on mpeg1:kaede-Green cells sorted from control and photoconverted embryos, and on mpeg1:kaede-Green/Red cells from photoconverted embryos. The color code matches the gates on the FACS plots. Error bars represent mean ± SD (n = 3 experiments).

(G) Sagittal sections of E10.5 Runx1^null embryos and their littermates (upper body) immunostained for CD31 (vascular) and CD45 (hematopoietic) markers, with DAPI used as a nuclear counterstain. Tile-scanned images show that Runx1^null mutants are completely devoid of CD45⁺ MG in the neuroepithelium and cephalic mesenchyme (n = 3). Scale bar: 500 μm.

See also Figures S1–S3.

Embryonic and Adult MG Have Distinct Origins

(A) Scheme of the transgenic lines used to address the MG potential of primitive versus definitive hematopoiesis.

(B and C) Whole-mount fluorescence microscopy images of Tg(kdrl:Cre; ßactin:Switch-DsRed; mpeg1:GFP) triple-transgenic embryos show absence of DsRed expression (red) in GFP⁺ pMFs (green) on the yolk ball at 30 hpf (B), and in GFP⁺ embryonic MG (green) in the brain parenchyma at 3 dpf (C) (n = 14).

(D) In contrast, in sections of adult triple-transgenic animals, DsRed labeling is observed in all GFP⁺ MG cells.

(E) Flow cytometry analysis on a brain dissected from a 12-week-old triple-transgenic animal identifies blood-derived mpeg1:GFP^high CNS-associated MFs (blue circle, MFs) and mpeg1:GFP^low MG (red circle, MG). Both populations express the DsRed reporter at high level (n = 5). Scale bar: 50 μm. MFs, macrophages; MG, microglia.

See also Figures S4 and S5.

Kinetics of Adult MG Emergence during Zebrafish Development

(A) Scheme of the transgenic lines used to discriminate pMF-derived embryonic MG from endothelial-derived adult MG.

(B) GFP (left panels) and DsRed (middle panels) immunostaining performed on brain sections from Tg(kdrl:Cre; ßactin:Switch-DsRed; mpeg1:GFP) triple transgenics at the indicated developmental stage, showing the replacement of GFP⁺ DsRed⁻ embryonic MG by GFP⁺ DsRed⁺ cells in the brain parenchyma. The right panels show a merge of both fluorescent channels. Scale bar: 50 μm.

(C) Quantification of DsRed⁺ MG in the brain parenchyma during zebrafish development, determined as percent recombination among the whole GFP⁺ MG population. Error bars represent mean ± SEM of pooled data from two experiments (n = 3). MG, microglia.

The Transcription Factor cmyb Is Required for the Establishment of Adult MG

(A) Scheme of the hematopoietic cell transplantation experiments performed to circumvent the lethality of c-myb^null animals due to the lack of definitive hematopoiesis.

(B) GFP immunostaining performed on adult brain sections from wild-type controls (upper panels) and rescued cmyb^null mutants (lower panels), demonstrating a complete absence of host-derived GFP⁺ MG in the cmyb mutants (left panels), as well as the presence of GFP⁻ donor-derived L-plastin⁺ hematopoietic cells (fuchsia, middle panels). The right panels show a merge of both fluorescent channels (n = 4). Scale bar: 50 μm.

Adult MG Derive from Embryonic HSCs

(A) Transgenic lines used to address the MG potential of embryonic HSCs.

(B) GFP (left panel) and DsRed (middle panel) immunostaining performed on quadruple-transgenic adult brain sections. The right panel shows a merge of both fluorescent channels.

(C) Pooled data from two experiments showing the percent recombination among brain MG in control (left) and gata2b:Cre adult fish (right) (n = 6). Error bars represent mean ± SEM.

(D) Correlation and regression analysis (R² = 0.85) between the chimerism in brain MG and the WKM myeloid fraction. MG, microglia; WKM, whole-kidney marrow.

See also Figure S6.