Role of STAT3-FOXO3 Signaling in the Modulation of Neuroplasticity by PD-L1-HGF-Decorated Mesenchymal Stem Cell-Derived Exosomes in a Murine Stroke Model

Abstract

The limited therapeutic strategies available for stroke leave many patients disabled for life. This study assessed the potential of programmed death-ligand 1 (PD-L1) and hepatocyte growth factor (HGF)-engineered mesenchymal stem cell-derived exosomes (EXO-PD-L1-HGF) in enhancing neurological recovery post-stroke. EXO-PD-L1-HGF, which efficiently endocytosed into target cells, significantly diminishes the H₂O₂-induced neurotoxicity and increased the antiapoptotic proteins in vitro. EXO-PD-L1-HGF attenuates inflammation by inhibiting T-cell proliferation and increasing the number of CD8⁺CD122⁺IL-10⁺ regulatory T cells. Intravenous injection of EXO-PD-L1-HGF could target stromal cell-derived factor-1α (SDF-1α⁺) cells over the peri-infarcted area of the ischemic brain through CXCR4 upregulation and accumulation in neuroglial cells post-stroke. EXO-PD-L1-HGF facilitates endogenous nestin⁺ neural progenitor cell (NPC)-induced neurogenesis via STAT3-FOXO3 signaling cascade, which plays a pivotal role in cell survival and neuroprotection, thereby mitigating infarct size and enhancing neurological recovery in a murine stroke model. Moreover, increasing populations of the immune-regulatory CD19⁺IL-10⁺ and CD8⁺CD122⁺IL-10⁺ cells, together with reducing populations of proinflammatory cells, created an anti-inflammatory microenvironment in the ischemic brain. Thus, innovative approaches employing EXO-PD-L1-HGF intervention, which targets SDF-1α⁺ expression, modulates the immune system, and enhances the activation of resident nestin⁺ NPCs, might significantly alter the brain microenvironment and create a niche conducive to inducing neuroplastic regeneration post-stroke.

Keywords: HGF; PD‐L1; STAT3‐FOXO3 pathway; exosomes; stroke.

Figure 1

Characterization of PD‐L1‐HGF‐modified hTERT‐ADSC‐derived exosomes. a) Western blot, b) ELISA, and c) flow cytometry of hTERT‐ADSC‐PD‐L1‐HGF cells overexpressing HGF and/or PD‐L1. d) Assessment of exosome morphology by means of transmission electron microscopy. Scale bar: 100 nm. e) Particle size distribution of exosomes, as measured using nanoparticle tracking analysis. f) Western blot and g) flow cytometry of specific exosomal surface markers CD63, CD9, and CD81, as well as the transgenic markers HGF and PD‐L1, in the EXO‐ and EXO‐PD‐L1‐HGF‐treated groups. h) ELISA for the HGF concentration in the EXO‐ and EXO‐PD‐L1‐HGF‐treated groups. Data are presented as mean ± standard deviation (SD) (*p < 0.05; **p < 0.01). ELISA, enzyme‐linked immunosorbent assay; PD‐L1, programmed death‐ligand 1; CD, cluster of differentiation; HGF, hepatocyte growth factor; hTERT‐ADSC, human telomerase reverse transcriptase‐immortalized adipose tissue‐derived mesenchymal stem cell.

Figure 2

EXO‐PD‐L1‐HGF significantly attenuated the H₂O₂‐induced neuronal apoptosis in vitro. a) TUNEL staining was used to detect apoptosis in PCCs and SH‐SY5Y cells. Cell nuclei were counterstained with DAPI (blue). Quantitative estimation of the proportion of apoptotic cells in each of the following experimental groups: groups pretreated with mock, control, EXO, and EXO‐PD‐L1‐HGF, followed by 300 µm H₂O₂, as indicated. b) Annexin V‐FITC/7‐AAD double staining was used to detect cell apoptosis induced by H₂O₂ in the PCCs and SH‐SY5Y cells pretreated with mock, control, EXO, and EXO‐PD‐L1‐HGF. c) CCK‐8 assay was used to assess the cell viability of PCCs and SH‐SY5Y cells pretreated with mock, control, EXO, and EXO‐PD‐L1‐HGF overnight and then exposed to 300 µm H₂O₂. d) Western blot of proteins (e.g., Bax, caspase‐3, and Bcl‐2) involved in the H₂O₂‐induced apoptosis of PCCs and SH‐SY5Y cells. Data are presented as mean ± SD (*p < 0.05; **p < 0.01). TUNEL, terminal deoxynucleotidyl transferase dUTP nick‐end labeling; PD‐L1, programmed death‐ligand 1; HGF, hepatocyte growth factor; DAPI, 4′,6‐diamidino‐2‐phenylindole; CCK‐8, cell counting kit‐8; PCCs, primary cortical cells; H₂O₂, hydrogen peroxide.

Figure 3

EXO‐PD‐L1‐HGF suppressed the activation of cytotoxic T cells, promoted the activation of Tregs, and enhanced the proliferation of NPCs. a) Purified CD3‐expressing T cells were cultured with an anti‐CD3/28 activator, stained with CFSE dye, and then treated with control, EXO, and EXO‐PD‐L1‐HGF for 3 d, following which the proliferation of the cells was assessed by means of flow cytometry. b) ELISA was used to detect the expression of the proinflammatory cytokine IFN‐γ in the control, EXO, and EXO‐PD‐L1‐HGF groups. c) Purified CD3‐expressing T cells were cultured with an anti‐CD3/28 activator and treated with control, EXO, and EXO‐PD‐L1‐HGF for 3 d. Flow cytometry showed that EXO‐PD‐L1‐HGF promoted the production of IL‐10 by CD8⁺CD122⁺ Treg cells. d,e) NPCs were seeded in an ultralow well plate and treated with control, EXO, and EXO‐PD‐L1‐HGF for 7 d. The size and number of NPCs were increased in the EXO‐PD‐L1‐HGF treatment group, as compared to that in the other groups. f) The EXO‐PD‐L1‐HGF group promoted the proliferation of NPCs, as shown using the CCK‐8 assay. NPCs, neural progenitor cells; CCK‐8, cell counting kit‐8; PD‐L1, programmed death‐ligand 1; Tregs, regulatory T cells; CFSE, carboxyfluorescein succinimidyl ester; CD, cluster of differentiation.

Figure 4

Identification of differentially expressed genes in NPCs treated with EXO and EXO‐PD‐L1‐HGF. a) The clustering heatmaps showed that several neurogenesis‐related genes, including MET, VEGF, bFGF, CXCR4, STAT3, and FOXO3, displayed upregulation in the EXO‐PD‐L1‐HGF treatment group. b) Comparison of the expression profiles of cells revealed that 311 genes were upregulated upon EXO‐PD‐L1‐HGF treatment, as compared with that upon the EXO‐PD‐L1‐HGF and control treatments, using fold‐change >1.5 and an adjusted p‐value < 0.05 as criteria. c) GO annotation revealed that these differentially expressed genes were involved in functions related to cell development, cell differentiation, response to stimulus, and cell motility. d) GSEA using both GO and KEGG pathway databases indicated that genes mediated by EXO‐PD‐L1‐HGF treatment were enriched in the pathways of neurogenesis, regulation of cell development, positive regulation of cell differentiation, and chemokine signaling. e) Stat3 and Foxo3 were significantly expressed in the four significantly enriched gene sets. f) Gene concept network analysis of the significantly differentially expressed genes associated with the four enriched functions detailed that both Stat3 and Foxo3 were linked to the GSEA‐enriched functional terms. g,h) Validation of the mRNA expression of STAT3 and FOXO3 revealed a significant increase upon EXO‐PD‐L1‐HGF treatment in NPCs. Data are presented as mean ± SD (*p < 0.05; **p < 0.01). GO, Gene Ontology; GSEA, Gene Set Enrichment Analysis; KEGG, Kyoto Encyclopedia of Genes and Genomes; NPCs, neural progenitor cells; PCCs, primary cortical cells.

Figure 5

EXO‐PD‐L1‐HGF positively regulated the STAT3/FOXO3 pathway involved in neuroprotection and neuroregeneration. a) Expression of phospho‐STAT3 and phospho‐FOXO3 in NPCs and PCCs upon control, EXO, and EXO‐PD‐L1‐HGF treatments. Actin was used as the loading control. b) The protein expression levels of phospho‐STAT3 and phospho‐FOXO3 in PCCs were decreased upon addition of a stat3 inhibitor, as observed in the western blot. c) The western blot result showed that knockdown of FOXO3 expression reduced the increased levels of phospho‐FOXO3 and Bcl‐2 protein expression induced by EXO‐PD‐L1‐HGF, and enhanced the suppression of cleaved caspase‐3 levels upon treatment with 300 µm H₂O₂. d) In vitro apoptosis results showed that knockdown of FOXO3 expression suppressed the attenuation of cell death induced by EXO‐PD‐L1‐HGF upon treatment with 300 µm H₂O₂. e) The number of neurospheres formed by FOXO3^‐/‐ mice pretreated with control, EXO, and EXO‐PD‐L1‐HGF was determined. FOXO3^‐/‐ repressed the promotion of neurosphere formation by EXO‐PD‐L1‐HGF. f) The size of neurospheres formed by FOXO3^‐/‐ mice pretreated with control, EXO, and EXO‐PD‐L1‐HGF was measured. FOXO3^‐/‐ mice suppressed the increase in size promoted by EXO‐PD‐L1‐HGF. Quantified data are shown next to each graph. Data are presented as mean ± SD (*p < 0.05; **p < 0.01).

Figure 6

EXO and EXO‐PD‐L1‐HGF prominently migrated into the ischemic brain, owing to the upregulated expression of CXCR4. a) Control, DiD‐EXO, and DiD‐EXO‐PD‐L1‐HGF were intravenously injected into ischemic stroke mice. The fluorescence intensity of EXO‐PD‐L1‐HGF was observed at 4, 24, and 72 h after injection, using an IVIS system. b) Immunofluorescence showed that SDF‐1α was expressed in the ischemic injured area, and DiD‐EXO‐PD‐L1‐HGF migrated significantly to the injured site. c) Surface expression of CXCR4 was higher in the EXO‐PD‐L1‐HGF and EXO groups. d) Knockdown of FOXO3 expression suppressed the EXO‐PD‐L1‐HGF‐mediated promotion of the migration of nestin‐GFP‐expressing NPCs to the injured site. e) BrdU and Ki‐67 were used as cell proliferation markers to observe the effect of EXO‐PD‐L1‐HGF treatment on the proliferation of nestin‐GFP‐expressing cells. Quantified data are shown next to each graph. Data are presented as mean ± SD (*p < 0.05; **p < 0.01). PD‐L1, programmed death‐ligand 1; HGF, hepatocyte growth factor; GFP, green fluorescent protein; SDF‐1α, stromal cell‐derived factor‐1 alpha.

Figure 7

Activation of FOXO3 modulated the EXO‐PD‐L1‐HGF‐enhanced functional recovery and infarct volume reduction poststroke. a) Immunofluorescence showed that EXO‐PD‐L1‐HGF treatment promoted the differentiation of NPCs. NG2 is an oligodendrocyte marker, while GFAP is an astrocyte marker, and DCX is an immature neuron marker. b) Rotarod and c) beam walking tests result showed that knockout of FOXO3 expression (FOXO3 ^–/– ) suppressed the neurological function improvement induced by EXO‐PD‐L1‐HGF. d) TTC staining results showed that FOXO3 ^–/– mice repressed the attenuation of infarction size over the ischemia hemisphere. The quantified data are shown next to each graph. e) TUNEL staining was performed to detect apoptosis in the ischemia region. Cell nuclei were counterstained with DAPI (blue). Quantitative estimation of the proportion of apoptotic cells in each experimental group: LV‐sh‐control and LV‐shFOXO3 pretreatment followed by treatment with control, EXO, and EXO‐PD‐L1‐HGF, as indicated. Data are presented as mean ± SD (*p < 0.05; **p < 0.01). FOXO3, forkhead box 3; TTC, triphenyltetrazolium chloride; PD‐L1, programmed death‐ligand 1; HGF, hepatocyte growth factor; TUNEL, terminal deoxynucleotidyl transferase dUTP nick‐end labeling; DAPI, 4′,6‐diamidino‐2‐phenylindole; NPCs, neural progenitor cells; GFAP, glial fibrillary acidic protein; NG2, neural/glial antigen 2; DCX, doublecortin.

Figure 8

Intravenous injection of EXO‐PD‐L1‐HGF reduces proinflammatory responses via regulation of anti‐inflammatory leukocytes in the ischemic brain and spleen. a) The number of CD3‐expressing T cells was higher in the ischemic hemisphere, as compared to that in either hemisphere. In the b) ischemic hemisphere and c) spleen, there was a significantly reduced percentage of CD11c⁺CD80⁺ and CD11c⁺CD86⁺ (activated dendritic cells), CD3⁺CD8⁺IFN‐γ⁺ (cytotoxic T cells), CD3^–NK1.1⁺ (NK cells), and CD11b⁺CD80⁺ cells (M1 macrophage cells) in the EXO‐PD‐L1‐HGF treatment group. In the d) ischemic hemisphere and e) spleen, there was an enhanced percentage of CD‐19⁺IL‐10⁺ (regulatory B cells), CD8⁺CD122⁺IL‐10⁺ (Treg cells), and CD11b⁺CD206⁺F4/80⁺ cells (M2 macrophage cells) in the EXO‐PD‐L1‐HGF treatment group. Data are presented as mean ± SD (*p < 0.05; **p < 0.01). PD‐L1, programmed death‐ligand 1; HGF, hepatocyte growth factor; CD, cluster of differentiation; IFN‐γ, interferon gamma.