Clec7a Worsens Long-Term Outcomes after Ischemic Stroke by Aggravating Microglia-Mediated Synapse Elimination

Abstract

Ischemic stroke (IS) is a leading cause of morbidity and mortality globally and triggers a series of reactions leading to primary and secondary brain injuries and permanent neurological deficits. Microglia in the central nervous system play dual roles in neuroprotection and responding to ischemic brain damage. Here, an IS model is employed to determine the involvement of microglia in phagocytosis at excitatory synapses. Additionally, the effects of pharmacological depletion of microglia are investigated on improving neurobehavioral outcomes and mitigating brain injury. RNA sequencing of microglia reveals an increase in phagocytosis-associated pathway activity and gene expression, and C-type lectin domain family 7 member A (Clec7a) is identified as a key regulator of this process. Manipulating microglial Clec7a expression can potentially regulate microglial phagocytosis of synapses, thereby preventing synaptic loss and improving neurobehavioral outcomes after IS. It is further demonstrat that microglial Clec7a interacts with neuronal myeloid differentiation protein 2 (MD2), a key molecule mediating poststroke neurological injury, and propose the novel hypothesis that MD2 is a ligand for microglial Clec7a. These findings suggest that microglial Clec7a plays a critical role in mediating synaptic phagocytosis in a mouse model of IS, suggesting that Clec7a may be a therapeutic target for IS.

Keywords: Clec7a; ischemic stroke; microglia; phagocytosis; synapses.

Figure 1

Synaptic phagocytosis by microglia after ischemic stroke. A) CD68⁺ lysosome content (green) in Iba‐1⁺ microglia (red) in the ischemic penumbra of sham and tMCAO mice. Scale bars, 10 µm. B) Quantification of the lysosomal area relative to the microglial area. C) Representative single‐plane images of PSD95 and synaptophysin (Syn) (excitatory synapses) in the ischemic penumbra of sham and tMCAO mice. Dotted circles highlight colocalized synapses. Scale bars, 5 µm. D) Quantification of excitatory synapses. E‐H. Confocal images and quantification of synaptic puncta engulfed by microglia showing the colocalization of PSD95 or Syn with CD68 within microglia in the ischemic penumbra of sham and tMCAO mice. The white arrows indicate PSD95⁺CD68⁺ E) or Syn⁺CD68⁺ G) puncta within microglia. Mice aged 8–10 weeks were used for the experiments shown in this figure. Scale bars, 10 µm. Statistics were derived from 18 slices, n = 6 mice for each group. In B, D, F, and H, significance was calculated using two‐tailed unpaired Student's t‐test. Data are presented as mean ± SD. ***p < 0.001.

Figure 2

Microglia eliminate excitatory synapses after ischemic stroke. A) Schematic illustration of reporters for synapse elimination. B–E) Representative images of sham and tMCAO microglia (Iba‐1⁺, white) with only mCherry puncta (red) derived from ExPre (B), ExPost (C), InhiPre (D) or InhiPost (E) reporters in the ischemic penumbra. Microglia are highlighted with white outlines. F,G) Quantification of the volume of synapses containing mCherry alone within microglia. Scale bars, 10 µm. ExPre, excitatory presynapse; ExPost, excitatory postsynapse; InhiPre, inhibitory presynapse; InhiPost, inhibitory postsynapse. Mice aged 8–10 weeks were used for the experiments shown in this figure. The statistical data were derived from 20 slices, n = 4 mice per group. AUs, arbitrary units. In F and G significance was calculated using two‐tailed unpaired Student's t‐test. Data are presented as mean ± SD. ***p < 0.001; ns indicates no significant difference.

Figure 3

Microglia are essential for long‐term neurological dysfunction and cognitive impairment after ischemic stroke. A) Overview of the experimental timeline. PLX5622 treatment was initiated 2 weeks before surgery (after surgery, all mice returned to a normal diet), neurobehavioral assessments were conducted 1, 3, 7, 14, 21, and 28 days after surgery, and Morris water maze experiments were performed 22–27 days after surgery. B) Representative confocal images and quantification of microglial density in the whole brain confirming the loss of Iba‐1⁺ microglia (green) in PLX5622‐treated mice after 14 days. Scale bars, 1 mm. n = 3 mice per group. C) Representative images of TTC‐stained coronal brain sections (left) from mice after tMCAO. The right panel shows the quantification of the relative infarct volume. n = 4 mice per group. D–G) Long‐term neurological dysfunction after treatment with PLX5622 was assessed with the mNSS (D), rotarod test (E), adhesive contact test (F), and adhesive removal test (G). n = 9 mice per group. H–M) The results of the Morris water maze test: schematic diagram and representative navigation trajectories for learning and memory (H); swimming speed (I); summary data for escape latency (J); time spent in the four zones (K); number of entries into the four zones (L); and number of entries onto the platform (M). n = 12–14 mice per group. Mice aged 8–10 weeks were used for the experiments shown in this figure. Significance was calculated using either two‐tailed unpaired Student's t‐test (B‐C), one‐way ANOVA (I, M), or two‐way ANOVA (D‐G, J‐L), followed by Tukey's multiple comparisons test. Data are presented as mean ± SD. *p < 0.05, **p < 0.01, and ***p < 0.001 versus Sham; ^# p < 0.05, ^## p < 0.01, and ^### p < 0.001 versus tMCAO, ns indicates no significant difference.

Figure 4

Transcriptome sequencing reveals microglial responses to ischemic stroke. A) Schematic of RNA‐seq analysis of microglia isolated from the ischemic brain 7 days after tMCAO and from sham mice, n = 3 per group, each n consists of 30 mice. B) Principal component analysis (PCA) plot. The two groups of samples were separated when plotting the first (PC1) versus the second component (PC2). C) Scatterplot showing the comparison of log₂CPM (counts per million) in sham microglia versus tMCAO microglia; red dots represent DAM genes. The orange dashed line indicates a fold change of 1. D) GO analysis of DEGs identified key biological processes in tMCAO microglia compared with sham microglia, which are graphically displayed in bubble plots according to Z scores and significance (log10[adjusted p value]). E–H) Heatmap showing genes related to neuronal death (E), immune response (F), homeostatic state (G) and DAM state (H) in tMCAO microglia versus sham microglia. I) qRT‒PCR analysis showing the expression of phagocytosis‐related genes in isolated microglia from the ischemic penumbra of tMCAO and sham mice, n = 6 per group, each n consists of 3 mice. J) Protein‒protein interaction (PPI) network for microglial phagocytosis‐related genes. K,L). Representative single‐plane images (K) and quantitative analyses (L) of Clec7a (red) in Iba‐1⁺ microglia (green) in the ischemic penumbra from sham and tMCAO mice. Scale bars, 100 µm. Mice aged 8–10 weeks were used for the experiments shown in this figure. Statistics were derived from 12 slices, n = 4 mice per group. In I and L significance was calculated using two‐tailed unpaired Student's t‐test. Data are presented as mean ± SD. *p < 0.05, **p < 0.01, and ***p < 0.001.

Figure 5

Clec7a promotes microglia‐mediated synaptic phagocytosis after ischemic stroke. A) Schematic of the experimental design. B) PCR genotyping showed that Clec7a^i∆MG (lanes 1, 3, 4, and 6) and WT littermate control (lane 7) mice were successfully generated by crossing Cx3cr1^CreERT2 transgenic mice with Clec7a^fl/fl mice. Heterozygotes (Ht) that contained both Clec7a WT and Clec7a^flox bands are shown. C) Representative photographs of TTC staining (left) after tMCAO and quantification of the relative infarct volume (right). n = 7 mice per group. D) T2 images at 24 h after the onset of ischemia in both groups. n = 8 mice per group. E) DW images of both groups at 24 h after ischemic stroke. n = 8 mice per group. F) CD68⁺ lysosome content (green) in Iba‐1⁺ microglia (red) in the ischemic penumbra from Clec7a^i∆MG and Clec7a^fl/fl mice subjected to tMCAO. Scale bars, 5 µm. G) Quantification of the lysosomal area relative to the microglial area. H) Representative single‐plane images of Syn and PSD95 staining in the ischemic penumbra. Dotted circles highlight colocalized synapses. Scale bars, 5 µm. I) Quantification of excitatory synapses. J–M) Confocal images and quantification of synaptic puncta engulfed by microglia showing the colocalization of PSD95 (J, K) or Syn (L, M) with CD68 within microglia in the ischemic penumbra from sham and tMCAO mice. The white arrows indicate PSD95⁺CD68⁺ (J) or Syn⁺CD68⁺ (L) puncta within microglia. Scale bars, 10 µm. Statistics were derived from 18 slices, n = 6 mice per group. Mice aged 8–10 weeks were used for the experiments shown in this figure. Significance was calculated using either two‐tailed unpaired Student's t‐test (C–E, G, and I) or one‐way ANOVA, Tukey's multiple comparisons test (K and M). Data are presented as mean ± SD. ***p < 0.001 versus Sham; ^## p < 0.01 and ^### p < 0.001 versus tMCAO.

Figure 6

Inducible knockdown of microglial Clec7a rescues impaired neurological function after ischemic stroke. A–D) Long‐term neurological dysfunction after tMCAO was assessed with the mNSS (A), rotarod test (B), adhesive contact test (C), and adhesive removal test (D). E) Schematic diagram and representative navigation trajectories for learning and memory. F) The swimming speed is presented. G) The average escape latencies. H) The number of entries onto the platform. I) The duration in each zone. J) The number of entries into each zone. n = 10‐13 mice per group. Mice aged 8–10 weeks were used for the experiments shown in this figure. Significance was calculated using either one‐way ANOVA (F and H) or two‐way ANOVA (A‐D, G, and I,J), followed by Tukey's multiple comparisons test. Data are presented as mean ± SD. *p < 0.05, **p < 0.01, and ***p < 0.001 versus Sham; ^# p < 0.05, ^## p < 0.01, and ^### p < 0.001 versus tMCAO; ns indicates no significant difference.

Figure 7

Microglial Clec7a knockout restores neuronal activity and synaptic transmission after ischemic stroke. A) Representative traces of the action potentials recorded from ischemic penumbra neurons in +180 pA current injection of Clec7a^fl/fl and Clec7a^i∆MG mice subjected to tMCAO and sham operation. Scale bar, 20 mV, 50 ms. B) Statistics showed that as the injection current increased (from 0 to +360 pA), the number of spikes was significantly increased in tMCAO Clec7a^fl/fl mice compared to tMCAO Clec7a^i∆MG and sham mice (Clec7a^fl/fl group, n = 16 neurons in 3 mice, Clec7a^i∆MG group, n = 16 neurons in 3 mice; Clec7a^fl/fl + tMCAO group, n = 22 neurons in 5 mice, Clec7a^i∆MG + tMCAO group, n = 24 neurons in 6 mice). C) Representative PPR traces of Clec7a^fl/fl and Clec7a^i∆MG mice subjected to tMCAO and sham operation after two consecutive stimuli. Scale bar, 100pA, 10 ms. D) Statistical analysis showed that the ratios of tMCAO Clec7a^fl/fl mice was increased compared to tMCAO Clec7a^i∆MG and sham mice (Clec7a^fl/fl group, n = 10 neurons in 3 mice, Clec7a^i∆MG group, n = 12 neurons in 3 mice; Clec7a^fl/fl + tMCAO group, n = 12 neurons in 5 mice, Clec7a^i∆MG + tMCAO group, n = 16 neurons in 6 mice). Statistical significance was determined by two‐way ANOVA (B) and one‐way ANOVA (D) followed by Tukey's multiple comparisons test. Data are presented as mean ±  SD. *p < 0.05 and ***p < 0.001 versus Sham; ^### p < 0.001 versus tMCAO.

Figure 8

Clec7a interacts with MD2 in the ischemic brain. A) Immunoprecipitates were subjected to SDS‐PAGE and silver staining. B) Annotated LC–MS/MS spectra of 150‐LGAHLLKIDNSKEFEFIESQTSSHR‐174 from Clec7a and 85‐IELPKRKEVLCHGHDDDYSFCRALK‐109 from MD2 in anti‐Clec7a immunoprecipitates from tMCAO mice. C) Normalized binding curve of Clec7a and MD2. The binding curve yielded a Kd of 2.96 ± 1.11 µm. The concentration of Clec7a was kept constant at 50 nm while the MD2 concentration was varied from 10 µM to 1.2 nM. n = 3 per group. D) Endogenous Clec7a was coimmunoprecipitated with MD2 after ischemic stroke. The experiments were repeated independently three times. E) The interaction between Clec7a (cyan) and MD2 (pink). The yellow dashed line indicates hydrogen bonding. F) Representative images and analysis of the ischemic penumbra showing the colocalization of MD2 with PSD95, as indicated by the white arrowheads; scale bars, 5 µm. G) Representative images and analysis of the ischemic penumbra showing the colocalization of MD2 with Syn as indicated by the white arrowheads; scale bars, 5 µm. H) Schematic illustrating the in vitro phagocytosis assay in which primary microglia isolated from Clec7a^i∆MG or Clec7a^fl/fl mice phagocytosed MD2^ko/ko or MD2^wt/wt synaptosomes tagged with pHrodo. I) Representative images of cultured Clec7a^i∆MG or Clec7a^fl/fl microglia engulfing pHrodo‐conjugated synaptosomes (top: engulfed pHrodo; bottom: merged image of Iba‐1 and pHrodo). Scale bars, 20 µm. Statistics were derived from 18 slices, n = 6 per group. Mice aged 8–10 weeks were used for the experiments shown in this figure. Significance was calculated using either two‐tailed unpaired Student's t‐test (F and G) or one‐way ANOVA, Tukey's multiple comparisons test (I). Data are presented as mean ± SD. ***p < 0.001.