Female 6 to 8-wk-old C57BL/6 mice were obtained from the Shanghai Experimental Animal Center, Chinese Academy of Sciences. All mice were housed in specific pathogen-free conditions. All animal protocols were approved by the Laboratory Animal Ethics Committee of Xuzhou Medical University. Euthanasia was conducted in compliance with the “AVMA Guidelines for the Euthanasia of Animals”. Mice were euthanized by intraperitoneally injected 150mg/kg pentobarbital sodium (P3761, Sigma-Aldrich, St. Louis, MO, USA).

Antibodies and reagents employed in the present study were as follows: anti-p-AKT (Ser473, 4060, Cell Signaling Technology, Danvers, MA, USA), anti-AKT (4685, Cell Signaling Technology), anti-p-mTOR (Ser2448, 2971, Cell Signaling Technology), anti-mTOR (4517, Cell Signaling Technology), anti-p-STAT3 (Tyr705, BS4181, Bioworld Technology, Bloomington, MN, USA), anti-STAT3 (12640, Cell Signaling Technology), anti-glial fibrillary acidic protein (GFAP) (Rabbit, ab7260, Abcam, Cambridge, UK), anti-GFAP (Mouse, ab4648, Abcam), anti-complement component 3 (C3) (ab11862, Abcam). Alexa Fluor® 488 donkey anti-mouse IgG (A21202) and Alexa Fluor® 594 donkey anti-Rabbit IgG (A21207) antibodies were from Life Technologies (Carlsbad, CA, USA). Myelin oligodendrocyte glycoprotein (MOG) amino acids 35–55 (MOG_35-55 peptides, MEVGWYRSPFSRVVHLYRNGK) were purchased from China Peptides Co., Ltd. (051716, Shanghai, China). Met and Pio were purchased from Med Chem Express (HY-B0627, HY-13956, MCE, Monmouth Junction, NJ, USA). Recombinant mouse IL-17 was from R&D Systems (7956-ML, Minneapolis, MN, USA). Cytokine and chemokine detection kits: Interleukin‑6 (IL-6) (88-7064-88, Invitrogen, Vienna, Austria), Tumor necrosis factor-alpha (TNF-$\alpha$) (BM607-3, Invitrogen), Monocyte chemoattractant protein-1 (MCP-1) (NBP1-92659, R&D, Novus Biologicals, Centennial, CO, USA), interferon-gamma induced protein 10 (IP-10) (BMS6018, Invitrogen).

Procedures employed for EAE induction were performed as detailed in prior work [7, 27]. Female C57BL/6 mice aged 6–8 weeks were randomly assigned to groups. For EAE induction, 250 µg of MOG_35-55 was dissolved in 0.1 mL of sterile PBS, and 500 µg of inactivated Mycobacterium tuberculosis (H37Ra strain, 231141, Difco, Franklin Lakes, NJ, USA) was emulsified in 0.1 mL of complete Freund’s adjuvant (F5881, Sigma Aldrich). The two solutions were repeatedly aspirated and expelled through a three-way connector in an ice-water bath until a stable water-in-oil emulsion was formed. A total of 0.2 mL of this antigen emulsion was administered subcutaneously to each mouse via 4–6 injection sites along the back. Additionally, 200 ng of pertussis toxin (BP-225, lot: 4376920, Invitrogen) was dissolved in 0.2 mL of sterile PBS and administered intraperitoneally on days 0 and 2 post-immunization. Subsequently, neurological function scores were assessed daily on a 0–5 scale as described previously in a double-blinded way: 0, No clinical signs; 1, Loss of tail tone or limp tail; 2, Hind limb paresis (unilateral or bilateral); 3, Complete paralysis of both hind limbs; 4, Complete hind limb paralysis with forelimb paresis or paralysis; 5, Moribund state or death.

Primary astrocytes from mice were isolated and cultured following the detailed procedures reported in our previous study [6, 20]. Briefly, the cerebral cortex freed of meninges was dissected, minced, and then digested with 0.125% trypsin (T4049, Sigma-Aldrich). Following two rinses with Dulbecco’s modified Eagle’s medium/F-12 (DMEM/F-12, 11320033, Gibco, Grand Island, NY, USA) containing 10% fetal bovine serum (FBS, 10099141, Gibco), the cell suspension was passed through a sterile filter and seeded into culture flasks that had been pre-equilibrated at 37 °C with 5% CO₂. The cells were subsequently expanded for 3–4 additional passages under standard conditions. Finally, at least 95% positive cells of glial fibrillary acidic protein, as determined by the immunofluorescent assay, were applied to the experiments in vitro. Astrocytes were serum-starved for 12 h in FBS-free medium before IL-17 stimulation. All primary cells were validated for their identity by surface marker analysis and tested negative for mycoplasma.

Based on the drug manufacturers’ instructions (MCE), previous studies [28, 29, 30, 31], and our preliminary experiments, drug treatments are as follows:

Animal: Met 100 mg/kg/day and Pio 15 mg/kg/day were suspended in DMSO/Saline solution, respectively, and delivered individually via intraperitoneal injection, or combined and administered in a single injection of 100 µL total volume. The drug administration was performed from day 0 to day 19 post-immunization. Female C57BL/6 mice aged 6–8 weeks were randomly assigned to the following groups: normal control (NC) group, EAE group, DMSO + EAE group, Met + EAE group, Pio + EAE group, and Met + Pio + EAE group (15 mice per group).

Astrocytes: 10 mM Met is administered for 2 h, followed by stimulation with IL-17, and 10 µM Pio is administered for 1 h before IL-17 stimulation in astrocytes.

Cells or spinal cord tissues were processed for RNA extraction using TRIzol reagent (15596-026, Invitrogen). The first-strand cDNAs were developed using the Prime-Script TM RT reagent kit (RR037A, TaKaRa, Kusatsu, Shiga, Japan) from (1 µg RNA), and real-time PCR was carried out on a Roche LightCycler® 480 system (Roche Diagnostics, Basel, Switzerland) with SYBR Green qPCR Master Mix (04887352001, Roche Diagnostics). The sequences of all primers were listed in Supplementary Table 1. The specificity of each primer pair was confirmed by melting curve analysis, and the amplification efficiency was validated using a standard curve, with efficiencies ranging from 95% to 105% and R² values greater than 0.999. The gene transcription levels were quantified by the 2^{-Δ⁢Δ⁢CT} method.

ELISA was performed in accordance with the kit manufacturer’s protocol. In brief, the sera obtained from mouse blood or conditioned medium from primary mouse cultured astrocytes were harvested to quantify the secreted level of IL-6, TNF-$\alpha$, MCP-1 and IP-10.

As described previously [27], the total protein was extracted from primary astrocytes and spinal cord tissues of mice, respectively. Equivalent protein extracted from the mouse brain tissue (50 µg) or primary astrocytes (20 µg) was resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), then electro-transferred onto polyvinylidene fluoride (PVDF) membrane. After blocking with 5% non-fat milk for 2 h at room temperature, the membrane was probed with primary antibody overnight at 4 °C. Primary antibodies were applied as follows: p-mTOR (1:1000), mTOR (1:2000), p-AKT (1:1000), AKT (1:2000), p-STAT3 (1:500), STAT3 (1:1000) and $\beta$-actin (1:5000). After washing the membrane with tris-buffered saline with Tween 20 (TBST), it was incubated for 1 h at room temperature with the appropriate horseradish peroxidase (HRP)-conjugated secondary antibody, either anti-rabbit IgG (SA00001-2, 1:10,000, Proteintech, Rosemont, IL, USA) or anti-mouse IgG (Proteintech, 1:10,000). Following the TBST wash, the protein bands were identified using Clarity™ ECL Western Blot substrate (1705060, Bio-Rad, Hercules, CA, USA) and visualized with the ChemiDoc Touch imaging system (version 3.0, Bio-Rad). The protein expression in samples was normalized by $\beta$-actin.

Immunofluorescent staining of primary astrocytes was performed as described previously [7]. Briefly, frozen sections (10 µm) from the tissue of the spinal cord were incubated at room temperature for 15 minutes with 10% goat serum for blocking, followed by overnight incubation with the primary antibody at 4 °C. The primary antibodies used are as follows: anti-C3 (1:50), anti-p-STAT3 (1:100) and anti-glial fibrillary acidic protein (1:200). After a rinse, the sections were incubated at 37 °C for 1 h with Alexa488-conjugated donkey anti-mouse IgG (A21202, Life Technologies) and Alexa594-conjugated anti-Rabbit IgG (A21207, Life Technologies). Ultimately, the sections were counterstained using 4^′,6-diamidino-2-phenylindole (DAPI) (D9542, Sigma-Aldrich) and subsequently imaged with an Olympus IX51 (Olympus Corporation, Tokyo, Japan) or BX51 microscope (Olympus Corporation) (exposure times of 30 ms-100 ms). Quantitative fluorescence analysis was performed using ImageJ software (version 1.53, National Institutes of Health, Bethesda, MD, USA), where the mean fluorescence intensity is calculated by dividing the Integrated Density by the Area of the selected region.

Mice were anesthetized and perfused with 4% paraformaldehyde (P6148, Sigma-Aldrich) in 0.1 M sodium phosphate buffer (pH 7.4). Spinal cords were then rapidly extracted and fixed with the same fixation solution overnight at 4 °C. To evaluate inflammation and demyelination of EAE mice, 4 µm paraffin-embedded spinal cord sections were performed by hematoxylin and eosin (H&E) and luxol fast blue (LFB) staining, respectively [7]. Moreover, the ultrastructure of spinal cords was examined using an electron microscope (EM, HT7800, Hitachi High-Technologies, Tokyo, Japan).

Data are given as mean $\pm$ SEM and analyzed by two-tailed unpaired Student’s t test for two groups comparison or one-way multiple range analysis of variance (ANOVA) for multiple columns comparison. A Mann-Whitney test was used for nonparametric data (EAE scoring), and the normality of the data for all groups was assessed using the Shapiro-Wilk test. p values were determined for at least three independent experiments in triplicate using GraphPad Prism version 8.5 software (GraphPad Software, San Diego, CA, USA). p values 0.05 were considered significant. The assumption of homogeneity of variances was verified using Levene’s test. When the assumption of equal variances was violated, corrected analytical methods (Welch’s ANOVA) were employed. Upon a statistically significant effect being identified in the one-way ANOVA, post-hoc tests were conducted.

Studies on MS patients have confirmed that Th17 cells are abundant in peripheral blood, cerebrospinal fluid and focal lesions. Furthermore, the number of Th17 cells and the levels of Th17 cells-associated inflammatory mediators are significantly increased during the relapse phase of MS. Therefore, IL-17, as the key cytokine predominantly secreted by effector Th17 cells, plays an important role in the process of MS [31, 32]. We firstly detected the activation of neurotoxic astrocytes and the production of inflammatory cytokines in mouse primary astrocytes stimulated by IL-17 in vitro. The results of real-time PCR showed that the mRNA transcription levels of the related markers of A1-like astrocytes, including UDP glucuronosyltransferase 1 family, polypeptide a (Ugt1a), Histocompatibility 2, D region locus 1 (H2D1), Serpin peptidase inhibitor, clade G, member 1 (Serping1), Histocompatibility 2, T region locus 23 (H2T23), serglycin (Srgn) and C3, were significantly increased at 6 h by IL-17 stimulation and persisted until 12 h, compared with 0 h group (Fig. 1A). Meanwhile, the mRNA level of inflammatory cytokines and chemokines, including IL-6, TNF-$\alpha$, MCP-1 and IP-10, peaked at 6 h with IL-17 treatment and continued for 24 h (Fig. 1B). It is suggested that IL-17 induced the activation of neurotoxic A1-like reactive astrocytes and the production of pro-inflammatory cytokines.

Fig. 1.

IL-17 induces the activation of A1-like reactive astrocytes and AKT/mTOR/STAT3 signals in vitro. (A) The mRNA transcription levels of the related markers of A1 astrocytes (Ugt1a, H2D1, Serping1, H2T23, Srgn and C3) and (B) of inflammatory cytokines and chemokines (IL-6, TNF-$\alpha$, MCP-1 and IP-10) were quantified through real-time PCR analysis in primary mouse astrocytes treated with IL-17 at different time points. (C) The levels of p-AKT and p-mTOR were evaluated using a Western blotting assay in primary mouse astrocytes exposed to IL-17. (D,E) Relative density was used to analyze the protein expression of p-AKT and p-mTOR. (F) The levels of p-STAT3 were analyzed using a Western blotting assay in primary mouse astrocytes treated with IL-17. (G) Relative density was applied to analyze the protein expression of p-STAT3. *p 0.05, **p 0.01 and ***p 0.001 versus 0 h group (n = 3–5 per group). Data are represented as the means $\pm$ SEM. Ugt1a, UDP glucuronosyltransferase 1 family, polypeptide a; H2D1, histocompatibility 2, D region locus 1; Serping1, peptidase inhibitor, clade G, member 1; H2T23, histocompatibility 2, T region locus 23; Srgn, serglycin; C3, complement component 3; IL-17, interleukin 17; TNF-$\alpha$, tumor necrosis factor-alpha; MCP-1, monocyte chemoattractant protein 1; IP-10, interferon gamma-induced protein 10; AKT, protein kinase B; mTOR, mammalian target of rapamycin; STAT3, signal transducer and activator of transcription 3.

Due to the association of Akt/mTOR/STAT3 pathways with Th17 response and inflammation [23, 25, 33], we further explored the possible molecular mechanisms of IL-17-induced activation of A1-like astrocytes. Western blotting assay was used to detect the activation of the AKT/mTOR/STAT3 pathway. The results showed that the phosphorylation of AKT and mTOR reached their peak at 5 min after IL-17 stimulation (Fig. 1C–E; The original Western blotting images are provided in the Supplementary Material-WB image). Meanwhile, the level of STAT3 phosphorylation (p-STAT3) was increased at 3 h and maintained until 12 h (Fig. 1F,G). Overall, the results suggested that IL-17 induced the activation of neurotoxic astrocytes and the AKT/mTOR/STAT3 signaling pathway.

To determine the effect of Met and Pio on the activation of A1-like astrocytes, primary mouse astrocytes were pre-treated with Met and Pio prior to being stimulated by IL-17. As shown in Fig. 2A, the mRNA transcription levels of the related markers of A1-like astrocytes, including Ugt1a, H2D1, Serping1, H2T23, Srgn and C3, were significantly decreased by Met and Pio, compared with IL-17 stimulation. C3 is one of the most important markers of A1-like astrocytes [9]. Furthermore, the results of immunofluorescence saining sssay (IFA) showed that the protein expression of C3 was decreased in astrocytes treated with Met and Pio prior to IL-17 stimulation (Fig. 2B,C). These results suggested that Met and Pio may downregulate the activation of A1-like reactive astrocytes.

Fig. 2.

Met and Pio suppress the A1-like phenotype of astrocytes induced by IL-17. Primary mouse astrocytes were pretreated with Metformin (Met) and Pioglitazone (Pio), prior to being stimulated with IL-17 for 6 h. (A) The mRNA transcription levels of the related markers of A1 astrocytes were detected by real-time PCR assay. (B) Immunofluorescent staining for glial fibrillary acidic protein (GFAP), C3 and nuclear staining of 4^′,6-diamidino-2-phenylindole (DAPI) in primary astrocytes treated with IL-17. Scale bars, 20 µm. (C) The number of C3-positive cells was analyzed. Data are derived from three independent experiments and are presented as the means $\pm$ SEM. **p 0.01 and ***p 0.001 versus Dulbecco’s modified Eagle’s medium (DMEM) group. ^#p 0.05, ^#⁢#p 0.01 and ^#⁢#⁢#p 0.001 versus DMSO + IL-17 group.

To explore the possible mechanism of Met and Pio in downregulating A1-like astrocytes activation, the activation of AKT/mTOR/STAT3 pathway was measured in primary mouse astrocytes treated with IL-17, prior to pre-treatment with Met and Pio. The results presented that the phosphorylation level of AKT and mTOR was suppressed by Met and Pio, compared to the IL-17 + DMSO treatment group (Fig. 3A–C; The original Western blotting images are provided in the Supplementary Material-WB image). And then, IFA was employed to observe the phosphorylation level of STAT3. The results displayed that the level of p-STAT3 was dramatically decreased and the nuclear translocation was significantly reduced in pre-treatment of primary mouse astrocytes with Met and Pio, compared with the IL-17 + DMSO treatment group (Fig. 3D,E).

Fig. 3.

Met and Pio downregulate AKT/mTOR/STAT3 signals in astrocytes activated by IL-17. Primary mouse astrocytes were pretreated with Met and Pio, prior to being stimulated by IL-17. (A) The levels of p-AKT and p-mTOR were evaluated by Western blotting assay in primary mouse astrocytes stimulated by IL-17 for 5 min and 3 h, respectively. (B,C) Relative density was employed to evaluate the protein expression of p-AKT and p-mTOR. (D) Immunofluorescent staining for GFAP (green), p-STAT3 (red) and nuclear staining of DAPI (blue) in primary astrocytes treated with IL-17 for 6 h. Scale bars, 20 µm. (E) The number of p-STAT3-positive cells was analyzed. The data are derived from three separate experiments and are presented as the means $\pm$ SEM. ***p 0.001 versus DMEM; ^#⁢#p 0.01 and ^#⁢#⁢#p 0.001 versus DMSO + IL-17 group.

Furthermore, we asked whether Met and Pio disturbed Akt/mTOR/STAT3 pathways to regulate the inflammatory reaction. Thus, the mRNA transcription and secretion levels of inflammatory cytokines and chemokines (IL-6, TNF-$\alpha$, MCP-1, and IP-10) were assessed. The results showed that the production and release of inflammatory cytokines and chemokines were significantly downregulated by Met and Pio, compared to IL-17 + DMSO treatment (Fig. 4A,B). The above results suggested that Met and Pio downregulated the activation of Akt/mTOR/STAT3 pathways in A1-like astrocytes and the production of inflammatory cytokines.

Fig. 4.

Met and Pio decrease the production of pro-inflammatory cytokines in astrocytes treated with IL-17. (A) The mRNA transcription levels and (B) the secreted levels of inflammatory cytokines and chemokines (IL-6, TNF-$\alpha$, MCP-1 and IP-10) were examined by real-time PCR assay and ELISA in primary mouse astrocytes pretreated with Met and Pio, prior to being stimulated with IL-17 for 6 h. Results were represented as mean $\pm$ SEM. **p 0.01 and ***p 0.001 versus DMEM; ^#p 0.05, ^#⁢#p 0.01 and ^#⁢#⁢#p 0.001 versus DMSO + IL-17 group.

To determine whether to affect the activation of A1-like astrocytes in vivo, Met and Pio were injected into mice constituted by MOG_35-55on day 0 post immunization (dpi) till 19th dpi. Firstly, the clinical score revealed that Met and Pio not only delayed the onset time of EAE but also alleviated pathogenesis (Fig. 5A, Supplementary Fig. 1). Meanwhile, the production and release of IL-6, TNF-$\alpha$, MCP-1 and IP-10 were dramatically reduced in the Met and Pio treatment group (Fig. 5B,C). H&E and LFB staining exhibited that EAE mice had a large number of inflammatory cell infiltrations and more severe demyelination lesions in the spinal cords, in comparison with NC mice. In contrast, the EAE mice treated with Met and Pio were characterized by less inflammatory cell infiltration and demyelination lesions in the spinal cords than those of EAE mice (Fig. 5D,E). In the EAE group, the myelin sheath appeared disintegrated under electron microscopy, while only mild loosening of the sheath was shown in the Met and Pio treatment groups (Fig. 5F). Furthermore, the phosphorylation level of AKT, mTOR, and STAT3 was decreased in the Met and Pio treatment groups, compared to EAE mice (Fig. 6A–D; The original Western blotting images are provided in the Supplementary Material-WB image). And then, the mRNA transcription levels of the related markers of A1-like astrocytes, including Ugt1a, H2D1, Serping1, H2T23, Srgn and C3, were significantly decreased in the Met and Pio treatment group, compared with the EAE group (Fig. 6E). Thus, these data suggest that Met and Pio may reduce A1-like astrocytes activation, in turn alleviating the inflammatory response and myelin damage.

Fig. 5.

Met and Pio alleviate experimental autoimmune encephalomyelitis (EAE) pathogenesis and reduce the production of inflammatory cytokine in mice. (A) The EAE clinical scores were assessed daily on 15 mice per group. (n = 15 mice per group). (B,C) The concentrations of IL-6, TNF-$\alpha$, MCP-1, and IP-10 in the spinal cords and peripheral blood of mice were evaluated using real-time PCR and ELISA assay, respectively. **p 0.01, ***p 0.001 versus normal control (NC) group; ^#p 0.05, ^#⁢#p 0.01 and ^#⁢#⁢#p 0.001 versus DMSO + EAE group (n = 4~6 mice per group). (D) Hematoxylin and eosin (H&E), (E) luxol fast blue (LFB) staining and (F) electron microscope were employed to evaluate the presence of infiltrating inflammatory cells, the myelin sheath damage in the spinal cords of mice (n = 5 mice per group). Scale bar = 50 μm, for (D,E); Scale bar = 1 μm for (F). Red arrows indicate the areas of myelin damage.

Fig. 6.

Met and Pio inhibit AKT/mTOR/STAT3 signals and A1 astrocytes activity in EAE mice. (A) The levels of p-AKT, p-mTOR and p-STAT3 were evaluated by Western blotting assay. (B–D) Relative density was used to analyze the protein expression of p-AKT, p-mTOR and p-STAT3. (E) The mRNA transcription levels of the related markers of A1 astrocytes (Ugt1a, H2D1, Serping1, H2T23, Srgn and C3) were detected by real-time PCR assay. ***p 0.001 versus NC group; ^#p 0.05, ^#⁢#p 0.01 and ^#⁢#⁢#p 0.001 versus DMSO + EAE group (n = 4~6 mice per group). Data are represented as the means $\pm$ SEM.