Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2013 Dec 12;5(5):1342-52.
doi: 10.1016/j.celrep.2013.11.004. Epub 2013 Dec 5.

An Anti-Inflammatory NOD-like Receptor Is Required for Microglia Development

Free PMC article

An Anti-Inflammatory NOD-like Receptor Is Required for Microglia Development

Celia E Shiau et al. Cell Rep. .
Free PMC article


Microglia are phagocytic cells that form the basis of the brain's immune system. They derive from primitive macrophages that migrate into the brain during embryogenesis, but the genetic control of microglial development remains elusive. Starting with a genetic screen in zebrafish, we show that the noncanonical NOD-like receptor (NLR) nlrc3-like is essential for microglial formation. Although most NLRs trigger inflammatory signaling, nlrc3-like acts cell autonomously in microglia precursor cells to suppress unwarranted inflammation in the absence of overt immune challenge. In nlrc3-like mutants, primitive macrophages initiate a systemic inflammatory response with increased proinflammatory cytokines and actively aggregate instead of migrating into the brain to form microglia. NLRC3-like requires both its pyrin and NACHT domains, and it can bind the inflammasome component apoptosis-associated speck-like protein. Our studies suggest that NLRC3-like may regulate the inflammasome and other inflammatory pathways. Together, these results demonstrate that NLRC3-like prevents inappropriate macrophage activation, thereby allowing normal microglial development.


Figure 1
Figure 1. Lack of microglia in nlrc3-like mutants
(A-B) Live 5 dpf zebrafish larvae stained by neutral red (Herbomel et al., 2001) to visualize microglia in the brain. (A) Heterozygous st73/+ larvae (arrow, microglia around optic tectum). (B) Homozygous st73 larvae have no microglia (arrow). (C,D) Microglial marker apoe expression at 5 dpf in the brain. (E-F) Neutral red stained heterozygous (E) and homozygous mutant (F) showing normal whole animal morphologies at 5 dpf. (G) Genetic and physical map of the st73 locus. (H) Sequence chromatogram showing the lesion (asterisk) in the coding sequence of nlrc3-like (LOC100538217). (I) The st73 mutation introduces a premature stop codon near the 5′ end of the open reading frame of nlrc3-like. (A, B) Dorsal views; anterior is to the top. (E,F) Lateral views; anterior is to the left. See also Figure S1.
Figure 2
Figure 2. Aberrant migration of primitive macrophages in nlrc3-like−/− mutants
(A) Diagram showing the stereotypical migration route of macrophages from yolk sac to embryo proper (arrows). (B) Fluorescent images showing abnormal active coalescence of yolk sac macrophages (mpeg1:EGFP+ in green, arrows) that overlap with apoptotic marker TUNEL in red in mutants (bottom panel; n=12/12), which does not occur in wildtype (top panel; n=11/11). Region of imaging is indicated by yellow box in A. (C) Trajectories of individual yolk sac macrophages are shown using the MTrackJ cell tracking tool, where the end point is indicated by a solid triangle marker. See Movies S1 and S2 for the time-lapse series. (D) Time-lapse imaging of the head region (red box in A) where macrophages are expected to migrate into the brain, using transient transgenesis of the macrophage mpeg1:EGFP construct in green and the stable transgene kdrl:mCherry-CAAX to visualize the vasculature in red. Imaging shows macrophages entering the brain wildtype (top panels) but not in mutants (bottom panels). The numbers indicate the time in minutes. See Movies S3 and S4. (E) Whole mount in situ hybridization at 2.5 dpf showing macrophage mfap4+ and microglial apoe+ cells in wildtype heterozygous embryos (arrows) but not in mutants (arrows). Aberrant macrophage clusters form near cranial vasculature in mutants (arrowheads). MB, midbrain; HB, hindbrain; dpf, day postfertilization. See also Figure S2.
Figure 3
Figure 3. Inappropriate inflammatory activation of macrophages in nlrc3-like−/− mutants
High magnification images showing striking cell morphology differences between yolksac macrophages of wildtype (A) and mutant (B). Macrophages in the mutants have multiple, large vacuoles (arrows). (C) Plot showing significantly higher percentages of primitive macrophages (mpeg1-EGFP+) exhibiting vacuolation (≥ 1 large cytoplasmic vacuole; p=0.047) and cell death (overlap of TUNEL staining with macrophage DAPI staining; p=0.037) in nlrc3-like−/− mutants (n=5 embryos) compared with siblings (n= 6 embryos). Statistical significance was determined by the two-tailed student’s t-test. Examples of dying mutant macrophages exhibiting apoptosis (D) and cellular breakdown (E). (F) Plot of macrophage number over time; diagram indicates the region of quantification (colored boxes); n=5 for each genotype at 1 dpf and 2.5 dpf, p = 0.0009 at 2.5 dpf; at 3–4 dpf, n=7 siblings and n= 3 mutants, p= 2E–05. (G) Graph shows relative mRNA levels of pro-inflammatory cytokines by qPCR comparing mutants (n=6) to wildtype siblings (n=6) at 3 dpf; p= 0.70 (β-actin); 0.002 (il-1β); 0.004 (il-8); 0.008 (il-12a); 0.03 (tnfα). n, number of independent embryos analyzed; dpf, day postfertilization. Error bars show s.e.m.; *, p < 0.05; **, p<0.01, ***, p < 0.001, all p-values are two-tailed.
Figure 4
Figure 4. Systemic inflammation in nlrc3-like−/− mutants in the absence of infection or injury
Neutrophil-specific marker mpo shows no brain neutrophils in wildtype siblings (arrow, A) but abundant infiltration in mutants (arrow, B). Lateral views, anterior to the left. (C,D) Representative frames from time-lapse imaging in the intact 2.5 dpf embryos (see Movies S5 and S6) show wandering neutrophils (lyz:EGFP) in the brain of mutants (D) but none in wildtype (C), with reference to vasculature (kdrl:mCherry-CAAX). Dorsal views, anterior to the top. (E,F) Higher magnification showing GFP channel alone (dotted box region in C and D, respectively). (G) Plot shows that high numbers of neutrophil infiltrate the brain in mutants (n=15) but nearly zero in wildtype (n=16, p = 0.0018). (H,I) Representative images from live imaging of neutrophils at 1.5 dpf show a large number of neutrophils in brain (asterisk) and circulation (arrows) in mutants (I) (see Movies S7 and S8). Lateral views, anterior to the top. (J) Plot shows significantly higher numbers of circulating neutrophils in mutants (n=4) compared with wildtype (n=5; p = 0.0083). (K) Neutrophils (lyz:EGFP as pseudo-colored in red) are closely intermixed with the aberrantly coalescing macrophages (L-plastin expression only as shown in green) in the mutants, whereas wildtype neutrophils are dispersed and not clustered with macrophages. n, number of embryos analyzed. Error bars show s.e.m.; **, p < 0.01; all p-values are two-tailed; ey, eye; mb, midbrain; ot, otic vesicle. See also Figure S3.
Figure 5
Figure 5. nlrc3-like has an autonomous and continuous function in macrophages that become microglia
(A) Top, diagram of transgenic constructs driving either control mCherry or nlrc3-like expression by tissue-specific gene promoters in skin (keratin 4/krt4), neurons (elavl3/huC), neutrophils (lyz), or macrophages (mpeg1). Bottom, representative images showing rescue of neutral red+ microglia in nlrc3-like−/− mutants when expression of wildtype nlrc3-like is restored in macrophages (right) but not by control mCherry expression (left). (B) Plot quantifying the percentage of microglia rescue using different tissue-specific expression vectors. Ctrl, control mCherry injected; NLR, nlrc3-like construct injected. (C) Images and plot showing microglia by neutral red staining at 6 dpf, after nlrc3-like mRNA injection at the 1-cell stage. Injected nlrc3-like−/− mutants show no microglia or partial rescue (> 5) at 6 dpf, but no mutants are rescued to the wildtype level. In contrast, RNA injection does fully rescue microglia in some mutants at earlier stages (Fig. S1). (D) Graph showing relative pro-inflammatory transcript levels after transient rescue of nlrc3-like−/− mutants by injection of wildtype nlrc3-like mRNA at 1-cell stage. Injected mutants with no microglia at 6 dpf have elevated levels of pro-inflammatory cytokine transcripts that are similar to uninjected mutants. In contrast, mutants with partial rescue of microglia at 6 dpf have lower levels of il-8 and tnf-α expression, indicating that the number of microglia correlates inversely to the extent of inflammatory cytokine expression. Individual plots are normalized to their corresponding wildtype siblings, either uninjected or injected. Numbers below bar graphs represent n, number of embryos analyzed. Error bars show s.e.m. *, p < 0.05, one-tailed Student’s t-test; n.s., not significant. See also Figures S4 and S5.
Figure 6
Figure 6. NLRC3-like negatively regulates inflammatory signaling through macrophages, and may interact with the inflammasome component ASC and other pyrin and NACHT containing proteins
(A) Macrophage mfap4 expression showing complete macrophage ablation by irf8 MO injection at 2 dpf (arrow). Injected wildtype embryos show complete loss of neutral red+ microglia (n=29/29, data not shown). (B) Graph showing relative il-1β mRNA levels, with pair-wise comparisons indicated by the dotted lines. Depletion of macrophages in nlrc3-like mutants reduces il-1β and other pro-inflammatory cytokine levels similar to wildtype levels. (C) 4–12% SDS-PAGE analyses of reciprocal pull-down assays show binding of zebrafish ASC with NLRC3-like (lanes 2 and 4), and minimal binding to tag-alone controls (lanes 1 and 3). MBP-tagged full-length NLRC3-like runs at ~160kDa (blue arrowhead) with several smaller processed forms; GST-ASC at ~50kDa; MBP alone from pMAL-c2X vector at ~50kDa; GST alone at ~30kDa. Blot shows 0.4% of the input prey protein per lane. (D) Top, schematic of the wildtype NLRC3-like protein and deletion versions: NLRC3-like-Δpyrin with deletion at amino acids (aa) 230–269 and NLRC3-like-ΔNACHT with deletion at aa 357–518. Bottom, bar graph showing large percentage of rescue of st73 mutants (87.5%, n=16) by mRNA injection encoding wildtype NLRC3-like but none using the mutant constructs. Numbers at left edge of bar graphs represent n, number of embryos analyzed. Error bars show s.e.m. *, p < 0.05, **, p < 0.01, ***, p < 0.001; n.s., not significant; MO, morpholino; WT, wildtype; sib, siblings; mut, mutants.
Figure 7
Figure 7. Exposure of nlrc3-like heterozygous embryos to systemic LPS challenge disrupts microglia development
(A) Neutral staining at 3 dpf showing normal formation of microglia in control uninjected nlrc3-like heterozygous embryos. In contrast, many nlrc3-like heterozygous embryos injected with LPS at 1 dpf and 2 dpf have strongly reduced (6–16 microglia) or eliminated (0–5 microglia) microglia. (B) Bar graph showing frequency of loss of microglia in nlrc3-like heterozygous fish injected with LPS, compared with little effect on the wildtype siblings.

Similar articles

See all similar articles

Cited by 33 articles

See all "Cited by" articles

Publication types

MeSH terms

LinkOut - more resources