Developmental Biology

Pu.1/Spi1 dosage controls the turnover and maintenance of microglia in zebrafish and mammals

Yi Wu
Weilin Guo
Haoyue Kuang
Xiaohai Wu
Yuexin Wang
Shizheng Zhao
Zilong Wen
Tao Yu author has email address

Biomedical Research Institute, Shenzhen Peking University-the Hong Kong University of Science and Technology Medical Center, Shenzhen, China
Division of Life Science, the Hong Kong University of Science and Technology, Hong Kong, China
Greater Bay Biomedical Innocenter, Shenzhen Bay Laboratory, Shenzhen, China

https://doi.org/10.7554/eLife.105788.1

Open access
Copyright information

Figures and data

Adult microglia in zebrafish undergo rapid turnover and random proliferation to replenish and maintain their pool.
(A) Schematic diagram shows the workflow of EdU pulse experiment in adult TgBAC(ccl34b.1:eGFP) zebrafish and the representative images of proliferating microglia (ccl34b.1-eGFP⁺ Lcp1⁺EdU⁺) and dendritic cells (DCs) (ccl34b.1-eGFP^-Lcp1⁺EdU⁺) in the midbrain. (B) Quantification of the EdU incorporation proportions in microglia and DCs with different dosages of EdU pulses. (n=4 for each group) (C) The daily turnover rate of microglia and DCs calculated by dividing the EdU incorporation rate with EdU pulses. (D) Schematic diagram shows the experimental setup for EdU-BrdU double-pulse in adult wild-type fish and the representative images shows two BrdU⁺ microglia with or without EdU incorporation. (E) Comparison of BrdU incorporation proportions in EdU⁺eGFP⁺ and EdU^-eGFP⁺ microglia (n=6). n.s. = not significant, p>0.05.

Generation and characterization of the visible conditional knockout allele pu.1^KI.
(A) Schematic diagrams show the generation of pu.1^KI allele and the principle for pu.1 visible conditional knockout. Briefly, the donor plasmid, which contains: (1) the sgRNA target sequence in pu.1 intron 3∼4; (2) the loxP-flanked pu.1 coding sequence (exon4-6) followed by P2A-eGFP; (3) the splicing acceptor site followed by P2A-DsRed sequence was knocked into the endogenous pu.1 locus via non-homologous end joining (NHEJ) to generate pu.1^KI. In principle, the splicing event occurring between E3 and E4 In pu.1^KI would produce intact Pu.1 and eGFP concurrently. After Cre-mediated recombination, removal of pu.1 E4-6 and splicing of P2A-DsRed sequence in pu.1^CKOallele leads to the disruption of Pu.1 and fluorescent color change. (B) Co-staining of anti-eGFP and anti-Pu.1 antibodies on the yolk sac (YS) of 21-hpf pu.1^KI embryos. (C) Fluorescent imaging of the optic tectum (OT) region of 3-dpf pu.1^KI;Tg(mpeg1:LRLG) embryos. (D) Co-staining of anti-eGFP and anti-Lcp1 antibodies on the midbrain cross section of adult pu.1^KI fish. (E) Neutral red staining of pu.1^KI/+ and pu.1^KI/^Δ839 embryos at 3 dpf. (F) Quantification of NR⁺ microglia in pu.1^KI/+ (n=13) and pu.1 ^KI/^Δ839 (n=9) embryos at 3 dpf. (G) Fluorescent imaging of the YS region of 30-hpf pu.1^KI/CKO embryos. (H) Chromogram of cDNA sequence from pu.1^CKO embryos shows the precise splicing of P2A-DsRed cassette to pu.1 E3. (I) Neutral red staining of pu.1^CKO/+ and pu.1^CKO/^Δ839 embryos at 3 dpf. (J) Quantification of NR⁺ microglia in pu.1^CKO/+ (n=8) and pu.1^CKO/^Δ839 embryos at 3 dpf (n=14). n.s. = not significant, p>0.05; ****p<0.0001.

pu.1-deficient microglia were chronically eliminated in mosaic condition.
(A) The experimental setup for pu.1 conditional knockout in adult zebrafish and the representative images of midbrain cross section of pu.1^KI/+;Tg(coro1a:CreER) and pu.1^KI/^Δ839;Tg(coro1a:CreER) fish at 8 days post 4-OHT injection (dpi). (B) Quantification of the number of DsRed⁺ microglia on the midbrain cross section of pu.1^KI/+;Tg(coro1a:CreER) and pu.1^KI/^Δ839;Tg(coro1a:CreER) fish at 2 dpi (n=3) and 8 dpi (n=4). (C) Quantification of the proportion of DsRed⁺ microglia on the midbrain cross section of pu.1^KI/+;Tg(coro1a:CreER) and pu.1^KI/^Δ839;Tg(coro1a:CreER) fish at 2 dpi (n=3) and 8 dpi (n=4). (D) The experimental setup for pu.1 conditional knockout in adult zebrafish and the representative images of midbrain cross section of pu.1^KI/+;Tg(coro1a:CreER) and pu.1^KI/^Δ839;Tg(coro1a:CreER) fish at 3 months post 4-OHT injection (mpi). (E) Quantification of the number of DsRed⁺ microglia on the midbrain cross section of pu.1^KI/+;Tg(coro1a:CreER) and pu.1^KI/^Δ839;Tg(coro1a:CreER) fish at 1 mpi (n=3) and 3 mpi (n=4). (F) Quantification of the proportion of DsRed⁺ microglia on the midbrain cross section of pu.1^KI/+;Tg(coro1a:CreER) and pu.1^KI/^Δ839;Tg(coro1a:CreER) fish at 1 mpi (n=3) and 3 mpi (n=4). n.s. = not significant, p>0.05; **p<0.01; ****p<0.0001.

pu.1-deficient microglia in mosaic condition were eliminated by cell competition.
(A) The experimental setup for BrdU incorporation and TUNEL assays in adult pu.1^KI/+;Tg(coro1a:CreER) and pu.1^KI/^Δ839;Tg(coro1a:CreER) fish at 1 mpi. (B) Quantification of the percentage of TUNEL⁺ cells in eGFP⁺ and DsRed⁺ microglia in pu.1^KI/+;Tg(coro1a:CreER) (n=4) and pu.1^KI/^Δ839;Tg(coro1a:CreER) (n=5) fish at 1 mpi. (C) Quantification of the percentage of BrdU⁺ cells in eGFP⁺ and DsRed⁺ microglia in adult pu.1^KI/+;Tg(coro1a:CreER) (n=4) and pu.1^KI/^Δ839;Tg(coro1a:CreER) fish at 1 mpi. (D) The experimental setup for BrdU incorporation and TUNEL assays in adult pu.1^Δ839/+;TgBAC(ccl34b.1:eGFP) and pu.1^Δ839/Δ839;TgBAC(ccl34b.1:eGFP) fish. (E) Quantification of the percentage of TUNEL⁺eGFP⁺ microglia in adult pu.1^Δ839/+;TgBAC(ccl34b.1:eGFP) and pu.1^Δ839/Δ839;TgBAC(ccl34b.1:eGFP) fish. (F) Quantification of the percentage of BrdU⁺eGFP⁺ microglia in adult pu.1^Δ839/+;TgBAC(ccl34b.1:eGFP) (n=6) and pu.1^Δ839/Δ839;TgBAC(ccl34b.1:eGFP) (n=5) fish. n.s. = not significant, p>0.05; *p<0.05; **p<0.01.

Inactivation of Tp53 largely restored the number of pu.1-deficient microglia in mosaic condition.
(A) The experimental setup for the isolation of eGFP⁺ and DsRed⁺ microglia from pu.1^KI/^Δ839;Tg(coro1a:CreER) adult brain at 3 mpi for transcriptomic analysis. (B) The volcano plot of differential expressed genes (DEG) between eGFP⁺ and DsRed⁺ microglia at 3 mpi. (C) Relative expression (tpm) of tp53 in eGFP⁺ (n=3) and DsRed⁺ (n=3) microglia at 3 mpi. (D) The experimental setup for pu.1 conditional knockout in wild-type and tp53^-/-adult zebrafish, and the representative images of midbrain cross section of pu.1^KI/^Δ839;Tg(coro1a:CreER) and pu.1^KI/^Δ839;tp53^-/-;Tg(coro1a:CreER) fish at 3 mpi. (E) Quantification of the number of DsRed⁺, eGFP⁺ and total (DsRed + eGFP) microglia in pu.1^KI/^Δ839;Tg(coro1a:CreER) (n=7) and pu.1^KI/^Δ839;tp53^-/-;Tg(coro1a:CreER) (n=6) fish at 3 mpi. (F) Quantification of the proportion of DsRed⁺ microglia in pu.1^KI/^Δ839;Tg(coro1a:CreER) (n=7) and pu.1^KI/^Δ839;tp53^-/-;Tg(coro1a:CreER) (n=6) fish at 90 dpi. n.s. = not significant, p>0.05; *p<0.05.

Dosage-dependent regulation of microglia maintenance by Pu.1/Spi1 is evolutionary conserved in mice.
(A) The experimental setup for Pu.1 conditional knockout in adult mice. (B) Representative images of IBA1 and DAPI co-staining in the cortex, hippocampus and thalamus of Pu.1^Fl/+;Cx3cr1^CreER and Pu.1^Fl/Fl;Cx3cr1^CreERmice at 7 dpi. (C) Quantification of the density of IBA1⁺ microglia in the cortex, hippocampus and thalamus of Pu.1^Fl/+;Cx3cr1^CreER(n=3) and Pu.1^Fl/Fl;Cx3cr1^CreER (n=3) mice at 7 dpi. (D) The experimental setup for conditional knockout of Pu.1 in Pu.1^Fl/+;Cx3cr1^CreER mice, and the subsequent PCR detection and T-clone quantification of Pu.1^Fl and Pu.1^KO alleles in sorted YFP⁺ microglia. (E) Gel image shows the relative intensity of amplified DNA bands of Pu.1^Fl and Pu.1^KOalleles in microglia sorted from Pu.1^Fl/+;Cx3cr1^CreERmice at different stages post TAM injection. (F) Quantification of the percentage of Pu.1^KO allele at 3 dpi (n=4) and 3.5 mpi (n=4) by T-clone assay. *p<0.05; **p<0.01; ***p<0.001; ****p<0.0001.

Sign up for email alerts