Mitochondrial transplantation confers protection against the effects of ischemic stroke by repressing microglial pyroptosis and promoting neurogenesis

Abstract

Transferring healthy and functional mitochondria to the lateral ventricles confers neuroprotection in a rat model of ischemia-reperfusion injury. Autologous mitochondrial transplantation is also beneficial in pediatric patients with cardiac ischemia-reperfusion injury. Thus, transplantation of functional exogenous mitochondria may be a promising therapeutic approach for ischemic disease. To explore the neuroprotective effect of mitochondria transplantation and determine the underlying mechanism in ischemic stroke, in this study we established a photo-thrombosis-induced mouse model of focal ischemia and administered freshly isolated mitochondria via the tail vein or to the injury site (in situ). Animal behavior tests, immunofluorescence staining, 2,3,5-triphenyltetrazolium chloride (TTC) staining, mRNA-seq, and western blotting were used to assess mouse anxiety and memory, cortical infarct area, pyroptosis, and neurogenesis, respectively. Using bioinformatics analysis, western blotting, co-immunoprecipitation, and mass spectroscopy, we identified S100 calcium binding protein A9 (S100A9) as a potential regulator of mitochondrial function and determined its possible interacting proteins. Interactions between exogenous and endogenous mitochondria, as well as the effect of exogenous mitochondria on recipient microglia, were assessed in vitro. Our data showed that: (1) mitochondrial transplantation markedly reduced mortality and improved emotional and cognitive function, as well as reducing infarct area, inhibiting pyroptosis, and promoting cortical neurogenesis; (2) microglial expression of S100A9 was markedly increased by ischemic injury and regulated mitochondrial function; (3) in vitro, exogenous mitochondria enhanced mitochondrial function, reduced redox stress, and regulated microglial polarization and pyroptosis by fusing with endogenous mitochondria; and (4) S100A9 promoted internalization of exogenous mitochondria by the microglia, thereby amplifying their pro-proliferation and anti-inflammatory effects. Taken together, our findings show that mitochondrial transplantation protects against the deleterious effects of ischemic stroke by suppressing pyroptosis and promoting neurogenesis, and that S100A9 plays a vital role in promoting internalization of exogenous mitochondria.

Figure 1

Mortality and behavioral disorders induced by ischemia in a mouse model were rescued by mitochondrial transplantation. (A) The experimental scheme for extracting and isolating mitochondria, as well as injecting mitochondria via the tail vein intravenously (i.v.) or to the injury site (in situ) of wild-type mice subjected to stroke. (B) Exogenous mitochondria labeled with MitoTracker Red were observed in microglia (ionized calcium-binding adapter molecule 1 [Iba-1]⁺) in the area of cerebral infarct on the 3^rd day after stroke. (C) Mortality of mice between the ischemic (Isc) and i.v. mitochondrial transplantation after ischemia (Iv + Isc) groups on the 3^rd day after 5 consecutive times of induction of ischemic stroke. (D) Mouse movements and percentage of time spent in the center in the OFT (n = 15 in each group) and in the open arm in the EPM test (n = 15) on the 3^rd day after stroke. (E) Mouse movements and percentage of time spent around the new object in the NOR test on the 3^rd day after stroke, n = 10. *P < 0.05, **P < 0.01 (Student’s t-test for C or one-way analysis of variance followed by Bonferroni’s post hoc test for D and E). OFT: Open field test; EPM: elevated plus maze; NOR: novel object recognition.

Figure 2

Intravenous mitochondrial transplantation reduced cortical infarct area, inhibited pyroptosis, and promoted neurogenesis. (A) Immunostaining for Caspase-3 (green)/neuronal nuclei (NeuN; red) of cortical tissue from wild-type mice in the Isc and Iv + Isc groups on the 3^rd day after stroke. (B) 2,3,5-Triphenyltetrazolium chloride staining on the 3^rd day after stroke, and quantitative analysis of the cortical infarct area (non-red region, indicated with blue arrows), n = 4. (C) Immunostaining for NOD-like receptor thermal protein domain associated protein 3 (NLRP3; red)/ionized calcium-binding adapter molecule 1 (Iba-1; green) on the 1^st, 3^rd, 5^th, and 7^th day after stroke. (D) Total gasdermin D (GSDMD), cleaved GSDMD (N-GSDMD), interleukin (IL)-1β, and active interleukin (IL)-1β protein expression levels in bilateral cortex tissues from mice in the Isc and Iv + Isc groups on the 3^rd day after stroke, as well as the quantitative analysis, n = 4. (E) Cyclin D1 protein expression levels in bilateral cortex tissues from mice in the Isc and Iv + Isc groups on the 3^rd day after stroke, as well as the quantitative analysis, n = 4. (F) Immunostaining for 5-bromo-2′-deoxyuridine (BrdU; green)/doublecortin (DCX; red) on the 7^th day after stroke. Note that the BrdU⁺DCX⁺/BrdU⁺ cell ratio was increased in the Iv + Isc group compared with Isc group, n = 4. *P < 0.05, **P < 0.01 ****P < 0.0001 (one-way analysis of variance followed by Bonferroni’s post hoc test). Isc: Ischemia; Iv: mitochondria administered intravenously.

Figure 3

Transplanting mitochondria i.v. and in situ inhibited pyroptosis and promoted neurogenesis. (A) KEGG enrichment analysis of DEGs between the control vs. ischemic groups on the 3^rd day after stroke. (B) KEGG enrichment analysis of DEGs between the Iv + Isc vs. ischemic groups. (C) KEGG enrichment analysis of DEGs between the In situ + Isc vs. ischemic groups. (D) Venn diagram of the Iv + Isc vs. ischemic group DEGs and the In situ + Isc vs. ischemic group DEGs related to pyroptosis, as well as the expression levels of representative mRNAs, n = 4, 4, 4, and 5, respectively. (E) Venn diagram of the Iv + Isc vs. ischemic group DEGs and the In situ + Isc vs. ischemic group DEGs related to neurogenesis, as well as the expression levels of representative mRNAs, n = 4, 4, 4, and 5, respectively. DEGs: Differential expression genes; Isc: ischemia; Iv: mitochondria administered intravenously; KEGG: Kyoto Encyclopedia of Genes and Genomes.

Figure 4

Mitochondrial transplantation induced upregulation of S100A9, crucial link between neurogenesis and pyroptosis. (A) Venn diagram showing that 20 DEGs in the ischemic cortex on the 3rd day after stroke between the WT (WT-isc) and interleukin (IL)-17 KO (IL-17 KO-isc) groups were enriched in genes related to mitochondria, neurogenesis, and pyroptosis, n = 4. (B) The expression of some of these DEGs, such as S100A9, changed when IL-17 was knocked out. (C) Heatmap of DEGs related to S100A9 among the control, ischemic, Iv + Isc, and In situ + Isc groups. (D) S100A9 protein expression levels in bilateral cortex tissues on the 1^st, 3^rd, 5^th and 7^th days after stroke. Note the significantly increase in S100A9 expression in the ischemic cortex in the first three days after stroke. n = 3. *P < 0.05, ***P < 0.001, ****P < 0.0001 (one-way analysis of variance followed by Bonferroni’s post hoc test). (E) Double-immunostaining for S100A9 (green) and neuronal nuclei (NeuN), ionized calcium-binding adapter molecule 1 (Iba-1), or glial fibrillary acidic protein (GFAP; red) in the ischemic cortex on the 3rd day after stroke. DEGs: Differential expression genes; Isc: ischemia; Iv: mitochondria administered intravenously; KO: knock out; S100A9: S100 calcium binding protein A9; WC: the control cortex of WT; WI: the ischemic cortex of WT; WT: wild-type.

Figure 5

S100A9 mediated the beneficial effects of mitochondrial transplantation on pyroptosis and neurogenesis. (A) Double-immunostaining for S100A9 (green)/Iba-1 (red) in ischemic cortex tissues from WT mice left untreated or treated with the S100A9 inhibitor paquinimod, and in IL-17 KO mice on the 3^rd day after stroke. (B) S100A9 protein expression in control cortex (Con), ischemic cortex (Isc), ischemic cortex after administering mitochondria intravenously (Iv + Isc), and ischemic cortex after administering mitochondria and paquinimod intraperitoneally (i.p.) (Iv+Isc+P), n = 4. (C) S100A9 protein expression in the cortical infarct area in the WT and IL-17 KO groups, n = 4. (D) NOD-like receptor thermal protein domain associated protein 3 (NLRP3) and gasdermin D (GSDMD) protein expression levels in the four groups, n = 4. (E) β-catenin, p-Akt, and S100A9 protein expression levels in the four groups, n = 4. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 (one-way analysis of variance followed by Bonferroni’s post hoc test). (F) S100A9 immunoprecipitation and mass spectrometry analysis. Iba-1: Ionized calcium-binding adapter molecule 1; Isc: ischemia; KO: knock out; S100A9: S100 calcium binding protein A9; WT: wild-type.

Figure 6

Internalized exogenous mitochondria fused with endogenous mitochondria, improving microglial viability. (A) Mitochondria, labeled with MitoTracker Red, were internalized by primary microglia in vitro (Iba-1⁺). (B) Endogenous (green) and exogenous (red) mitochondria were observed in live microglia. (C) Mitofusin-2 (MFN2) protein expression levels in HT22 cells, microglia (MG), and N9 cells in the control and mitochondria-transplanted groups, n = 4. (D) MFN2 protein expression in control microglia and microglia co-cultured with mitochondria at different concentrations, n = 3. (E) Microglial viability, as assessed by CellTiter-Glo® Luminescent Cell Viability Assay, was improved by co-culturing the cells with mitochondria at three different concentrations for 24 hours, n = 4. (F) Microglial ATP levels were also elevated by co-culturing with exogenous mitochondria for 24 hours, n = 4. (G) The reactive oxygen species (ROS) level decreased when microglia were co-cultured with mitochondria at three different concentrations for 24 hours, n = 4. (H) Mitochondria membrane potential in control microglia, CCCP-treated microglia, and microglia co-cultured with mitochondria at three different concentrations was measured by JC-1 assay. Note that the JC-1 polymer/monomer ratio was increased in all groups by co-culturing with mitochondria isolated from Neuro-2a cells, mouse liver cells, and even human umbilical cord mesenchymal stem cells, n = 4. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 (one-way analysis of variance followed by Bonferroni’s post hoc test or Student’s t-test). CCCP: Carbonyl cyanide 3-chlorophenylhydrazone; Iba-1: ionized calcium-binding adapter molecule 1.

Figure 7

Exogeneous mitochondria regulated microglial polarization and pyroptosis. (A) Heatmap and Venn diagram of RNA-seq data from N9 cells showing DEGs among the control (Con), half dose (M1/2), and full dose of mitochondria (M1) groups. The heatmap showing DEGs between every pair of groups in these three groups. (B) Overlapping genes related to both mitochondria and inflammation, mitochondria and pyroptosis, mitochondria and survival, and mitochondria and neurogenesis, among the M1 vs. Con DEGs. (C) KEGG enrichment analysis of the M1/2 vs. Con DEGs. (D) KEGG enrichment analysis of M1 vs. Con DEGs. (E) The relative mRNA expression levels of representative genes involved in microglia activation, such as nitric oxide synthase 2 (NOS2), interleukin 6 (IL-6), arginase-1 (Arg-1), and vascular endothelial growth factor (VEGFα), n = 3. (F) The relative mRNA expression levels of representative genes involved in inflammation, such as Caspase-1 and gasdermin D (GSDMD), n = 3. (G) The relative mRNA expression levels of representative genes involved in neurogenesis, such as β-catenin, JUN, and MYC, n = 3. DEGs: Differentially expressed genes; KEGG: Kyoto Encyclopedia of Genes and Genomes.

Figure 8

S100A9 promoted microglial internalization of exogenous mitochondria. (A) NLRP3, gasdermin D (GSDMD), and cleaved GSDMD (N-GSDMD) expression levels in control microglia or microglia co-cultured with mitochondria for 24 hours, with or without 5 hours of oxygen-glucose deprivation (OGD), n = 4. (B) Mitofusin-2 (MFN2) and IL-1β protein expression levels in control microglia or microglia co-cultured with mitochondria for 24 hours, with or without 5 hours of OGD, n = 4. (C) β-catenin and arginase-1 (Arg-1) protein expression levels in microglia subjected to OGD for 5 hours, with or without mitochondria co-culturing for 24 hours, n = 4. (D) MFN2 and S100A9 protein expression levels in microglia transfected with S100A9 RNA interference (RNAi) and S100A9 overexpression viruses, co-cultured with mitochondria for 24 hours, as well as quantitative data, n = 4. (E) GSDMD and Arg-1 protein expression levels in microglia transfected with S100A9 RNAi and S100A9 overexpression viruses, co-cultured with mitochondria for 48 hours, as well as quantitative data, n = 4. (F) Microglia transfected with S100A9 RNAi and S100A9 overexpression viruses (green fluorescent protein (GFP)⁺, green), subjected to OGD 5 hours, and co-cultured with mitochondria (red) for 24 hours. (G) Internalization of exogenous mitochondria (red) by N9 cells with S100A9 overexpressed (green) or S100A9 knock down (green) was compared after co-culturing with mitochondria for 6 hours, n = 4. *P < 0.05, **P < 0.01, ***P < 0.001 (one-way analysis of variance followed by Bonferroni’s post hoc test). O5M24: Cells under OGD for 5 hours, recovered and treated with mitochondria for 24 hours; O5R24: microglia under OGD for 5 hours and recovered for 24 hours; OGD: oxygen glucose deprivation.