Identification of a conserved and acute neurodegeneration-specific microglial transcriptome in the zebrafish

Abstract

Microglia are brain resident macrophages important for brain development, connectivity, homeostasis and disease. However, it is still largely unclear how microglia functions and their identity are regulated at the molecular level. Although recent transcriptomic studies have identified genes specifically expressed in microglia, the function of most of these genes in microglia is still unknown. Here, we performed RNA sequencing on microglia acutely isolated from healthy and neurodegenerative zebrafish brains. We found that a large fraction of the mouse microglial signature is conserved in the zebrafish, corroborating the use of zebrafish to help understand microglial genetics in mammals in addition to studying basic microglia biology. Second, our transcriptome analysis of microglia following neuronal ablation suggested primarily a proliferative response of microglia, which we confirmed by immunohistochemistry and in vivo imaging. Together with the recent improvements in genome editing technology in zebrafish, these data offer opportunities to facilitate functional genetic research on microglia in vivo in the healthy as well as in the diseased brain. GLIA 2016;65:138-149.

Keywords: RNA sequencing; microglia; neuronal cell death; proliferation; transcriptome; zebrafish.

Figure 1

Sequencing of the zebrafish microglia transcriptome. (A) Schematic representation of acute isolation of zebrafish microglia from mpeg1‐promoter driven GFP transgenic zebrafish (mpeg1‐GFP). (B) FACS plot showing isolated populations for RNAseq in zebrafish microglia (green) and other brain cells (blue). (C) Differential gene expression (Volcano plot) showing genes significantly higher expressed in microglia (green) and other brain cells (blue). FDR <0.01, LogFC > |2|, n = 3. (D) Heatmap showing Z‐score values of all genes differentially expressed between microglia and other brain cells (6511 genes) (FDR <0.01, LogFC > |2|). (E, F) Expression values (CPM) for known microglial, neuronal, radial glial and oligodendrocyte genes in GFP⁺ (microglia) and GFP⁻ (brain) cells. Values in (E) and (F) represent means of three independent experiments; Error bars in (E,F) represent standard deviation. FDR, False discovery rate; CPM, counts per million.

Figure 2

Conserved microglia functions in the zebrafish. (A) Ingenuity pathway analysis (IPA) canonical pathways most significantly enriched in zebrafish microglia. (B) Heatmap showing expression Z‐scores of immune genes significantly higher expressed in microglia than other brain cells (FDR <0.01 and logFC >1). (C) CPM values of zebrafish orthologs of genes involved in synaptic pruning in microglia and brain samples. Values in (C) represent mean of three independent experiments; Error bars in (C) represent standard deviation. TLR, Toll‐like receptors; PR, purinergic receptors; AP, antigen presentation; IL, interleukins + interleukin receptors; CP, complement.

Figure 3

Conserved microglia gene expression in the zebrafish. (A) Volcano plot showing expression of zebrafish orthologs of genes found to be enriched in microglia compared with Zhang et al. (2014), Butovsky et al. (2014), and Hickman et al. (2013), respectively. All zebrafish genes are shown in grey. Differentially expressed orthologs (FDR <0.01, LogFC > |2|) are shown in green (microglia) or blue (other brain cells). (B) Expression values (CPM) of orthologs for genes found by one or more of the above mentioned studies. (C) Venn diagram showing overlap of microglia‐specific orthologs of genes found in three transcriptomic studies (Butovsky et al., 2014; Hickman et al., 2013; Zhang et al., 2014). Values in (B) represent means of three independent experiments; Error bars in (B) represent standard deviation.

Figure 4

Identification of gene expression changes upon neuronal cell death. (A) TUNEL and L‐plastin staining showing increased neuronal cell death upon treatment with MTZ for 48 hours accompanied by microglia activation. mCherry signal represents engulfed neurons. (B) Schematic representation of cells isolated for RNA sequencing on activated microglia. (C) Volcano plot showing differentially expressed genes upon the activation of microglia (FDR <0.01; LogFC > |2|). (D) Heatmap showing Z‐score values of all genes differentially expressed between activated microglia and control microglia (367 genes) (FDR <0.01, LogFC > |2|).(E) IPA canonical pathway analysis on significantly upregulated genes in microglia upon NTR‐mediated ablation (FDR <0.01;LogFC > |2|). Scale bar = 40 µm in (A). For quantification in (A) cells were counted in 3 selected volumes in the olfactory bulb (4.0 × 10 ^{− 4} mm³) per fish (n = 3). Error bars represent standard deviation, *P < 0.05, ***P < 0.001 (Student's t‐test).

Figure 5

Microglia proliferation upon the induction of neuronal cell death. (A) Immunofluorescence staining in the olfactory bulbs of 3‐month‐old treated MTZ‐treated and DMSO‐treated (control) fish. (B) Intravital imaging in 7 dpf zebrafish larvae with Apoe‐driven GFP undergoing NTR‐mediated neuronal cell death, showing the presence of dividing zebrafish microglia upon neuronal death. n = 3. Scale bar = 40 µm in (A). For quantification in (A) cells were counted in 3 selected volumes in the olfactory bulb (4.0 × 10 ^{− 4} mm³) per fish (n = 3). Error bars represent standard deviation, *P < 0.05 (Student's t‐test).