Male C57BL/6J mice (6–8-week-old) were obtained from the Experimental Animal Center of Nantong University. For the duration of the experiment, the mice were housed in a controlled environment that ensured pathogen-free conditions, consistent temperature and relative humidity. Treated mice were injected with 1 mg/kg/day LPS (L8880, Solarbio, Beijing, China) intraperitoneally for 5 days, while control mice were injected intraperitoneally with an equivalent volume of physiological saline for the same duration. The mice were subjected to anesthesia by isoflurane (R510-22-10, Rayward, Shenzhen, Guangdong, China) and underwent heart perfusion using physiological saline. Brain samples were gathered for further analysis. It is important to note that the animal experiments conducted in this study received approval and strictly adhered to the regulations established by the Animal Ethics Committee of Nantong University (S20230315-005).

Human brain microvascular endothelial cells (HBMECs) were obtained from YingBioTech (Shanghai, China) with certification by short tandem repeat (STR) analysis and tested to be free of mycoplasma. HBMECs were cultured in Roswell Park Memorial Institute (RPMI) 1640 medium (A1049101, Gibco, New York, NY, USA) supplemented with 10% fetal bovine serum (FBS, Z7186FBS, Zeta Life, CA, USA) and 1% penicillin-streptomycin (C125C5, NCM Biotech, Suzhou, Jiangsu, China) at 37 °C in a 5% CO${}_2$ environment. HBMECs were exposed to 100 ng/mL LPS for 24 h.

GenPharma (Shanghai, China) synthesized the siRNAs for Sestrin2, which were transfected into HBMECs for 48 h with GP-transfect-Mate (G04008, GenePharma, Shanghai, China). Sestrin2 siRNA sequences were presented as below. Sestrin2 siRNA 1, 5${}^{\prime }$-GCGCAGAGCUCAAGGACUACC-3${}^{\prime }$; Sestrin2 siRNA 2, 5${}^{\prime }$-CAGUGUUCAAGUGCAGAAAUC-3${}^{\prime }$; Sestrin2 siRNA 3, 5${}^{\prime }$-GGCAUGUGAUGACUGUAAAUG-3${}^{\prime }$.

Sestrin2 overexpressive lentivirus was constructed by GenePharma (Shanghai, China). HBMECs were plated in a 6-well plate at the confluence of 50% and infected by Sestrin2 overexpressive lentivirus and control lentivirus with polybrene for 48 h. Then, HBMECs were exposed to LPS (100 ng/mL) for 24 h.

Mouse brain sections of 15 µm containing cortical regions were obtained by freezing section. Then, permeability of cell membrane was increased by 0.3% Triton (P0096, Beyotime, Shanghai, China) in phosphate buffer saline (PBS). Brain sections were blocked in 10% sheep serum (ZLI-9056, ZSGB-BIO, Beijing, China) at room temperature for 1 h followed by incubation in platelet endothelial cell adhesion molecule-1 (CD31) (1:100, ab24590, Abcam, Cambridge, UK) and alpha smooth muscle actin ($\alpha$-SMA) primary antibodies (1:50, 14395-1-AP, Proteintech, Wuhan, Hubei, China) at 4 °C overnight. The next day, brain sections were incubated in Alexa Fluor 594-conjugated anti-rabbit IgG (1:500; A-11037, Invitrogen, Carlsbad, CA, USA) and Alexa Fluor 488-conjugated anti-mouse IgG (1:500; A11001, Invitrogen) at room temperature for 1 h. Cell nuclei were visualized using ProLong gold anti-fade reagent containing 4${}^{\prime }$,6-diamidino-2-phenylindole (DAPI, 0100-20, Southern Biotech, Homewood, AL, USA). Immunofluorescence photos were captured by a fluorescence microscopy (BX53, Olympus, Tokyo, Japan).

A radio-immunoprecipitation assay buffer (P0013C, Beyotime, Shanghai, China) was used to lyse mouse cortex tissues and HBMECs followed by the measurement of protein concentration using a bicinchoninic acid (BCA) analysis kit (P0012S, Beyotime). After separation on 10% polyacrylamide gels, proteins were transferred to polyvinylidene fluoride (PVDF) (IPVH00010, Millipore, Boston, MA, USA) membranes, which were then incubated with primary antibodies overnight at 4 °C after blocking in 5% nonfat milk. The next day, PVDF membranes were incubated in goat anti-rabbit secondary antibody (1:2000, SA00001-2, Proteintech) and goat anti-mouse secondary antibody (1:2000, SA00001-1, Proteintech) labeled by horseradish peroxidase at room temperature for 1 h. Blot images were detected using a chemiluminescence system and the band signal was analyzed using ImageJ software 1.51j8 (National Institutes of Health, Bethesda, MD, USA). Primary antibodies were purchased from Proteintech and details were presented as inducible nitric oxide synthase (iNOS) (1:1000, 18985-1-AP), Sestrin2 (1:1000, 21346-1-AP), glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (1:5000, 60004-1-Ig), $\alpha$-SMA (1:1000, 14395-1-AP), light chain 3 (LC3) (1:1000, 14600-1-AP), Occludin (1:1000, 27260-1-AP), zona occludens protein 1 (ZO-1) (1:1000, 21773-1-AP), T-mTOR (1:1000, 66888-1-Ig), p-mTOR (1:1000, 67778-1-Ig), interleukin (IL)-1$\beta$ (1:500, 16806-1-AP), IL-6 (1:500, 21865-1-AP), IL-18 (1:500, 10663-1-AP) and tumor necrosis factor $\alpha$ (TNF-$\alpha$) (1:1000, 17590-1-AP).

Total RNA was extracted using Trizol reagent (T9108, Takara, Tokyo, Japan). First, RNA was purified with gDNA wiper solution and reverse transcribed into cDNA with a HiScript III 1st Strand cDNA Synthesis Kit (R312-01, Vazyme, Nanjing, China). Then, target mRNA was detected by real-time polymerase chain reaction (PCR) with Taq Pro Universal SYBR qPCR Master Mix (Q712-02, Vazyme). The PCR conditions were 95 °C for 30 s, followed by 40 cycles at 95 °C for 10 s and 60 °C for 30 s. Target RNA levels were normalized to GAPDH. The quantitative expression level was analyzed using the 2${}^{-\mathrm{\Delta }\mathrm{\Delta }\text{Ct}}$ method. Details of primer sequence were provided as below. Mouse IL-1$\beta$ forward primer: 5${}^{\prime }$-GAAATGCCACCTTTTGACAGTG-3${}^{\prime }$; reverse primer: 5${}^{\prime }$-TGGATGCTCTCATCAGGACAG-3${}^{\prime }$. Mouse IL-6 forward primer: 5${}^{\prime }$-TAGTCCTTCCTACCCCAATTTCC-3${}^{\prime }$; reverse primer: 5${}^{\prime }$-TTGGTCCTTAGCCACTCCTTC-3${}^{\prime }$. Mouse GAPDH forward primer: 5${}^{\prime }$-AGGTCGGTGTGAACGGATTTG-3${}^{\prime }$; reverse primer: 5${}^{\prime }$-GGGGTCGTTGATGGCAACA-3${}^{\prime }$.

HBMECs were digested with 0.25% trypsin (C100C1, NCM Biotech, Suzhou, Jiangsu, China) containing ethylene diamine tetraacetic acid (EDTA) followed by centrifugation at 1200 rpm for 5 min. Cell pellet was resuspended with fixation fluid for electron microscope (G1102, Servicebio, Wuhan, Hubei, China) after discarding supernatant. Astrocytes were washed using sucrose-sodium cacodylate buffer and soaked in osmium tetroxide-sodium cacodylate for 2 h. Then, astrocytes were stained using 2% aqueous uranyl acetate after washing in water. Subsequently, samples were dehydrated in an ethanol series and embedded in Epon 812 (90529-77-4, SPI Science, West Chester, PA, USA). Ultrathin sections of 80 nm were obtained using a ultramicrotome with a diamond knife. Grids were poststained with 2% saturated uranyl acetate in 1% lead citrate and 50% ethanol. Photographs were captured using an electron microscope camera (TECNAI G2 20 TWIN, Thermo fisher, Waltham, MA, USA).

The statistical analyses were conducted using GraphPad Prism 6.0 (GraphPad Software, San Diego, CA, USA). A Student’s t-test was used to compare the difference between two groups and one-way analyses of variance (ANOVA) was used to compare the differences among multiple groups. The statistical difference was considered significant if p 0.05. At least three replicates of each experiment were conducted, and all data were presented as mean $\pm$ standard deviation (SD).

To investigate the mechanism of EndoMT in the inflammatory response, LPS was administered intraperitoneally to mice several times to establish a chronic inflammatory model. In comparison to the control group, the intraperitoneal injection of LPS significantly enhanced the expressions of iNOS and TNF-$\alpha$ in the mouse cortex (Fig. 1A). LPS treatment also induced the expressions of inflammatory factor IL-1$\beta$, IL-6 and IL-18 (Fig. 1B). Additionally, the mRNA expression levels of IL-1$\beta$ (Fig. 1C) and IL-6 (Fig. 1D) were notably increased in the LPS-injected mice.

Fig. 1.

Lipolysaccharide (LPS) induced inflammatory reaction in mouse cortex. (A) Representative blots showed the expression of inducible nitric oxide synthase (iNOS) and tumor necrosis factor $\alpha$ (TNF-$\alpha$) in the cortex of control and LPS-injected mouse. (B) Representative blots showed the expression of interleukin (IL)-1$\beta$, IL-6 and IL-18 in the cortex of control and LPS-injected mouse. (C,D) The expressions of IL-1$\beta$ mRNA (C) and IL-6 mRNA (D) in the cortex of control and LPS-injected mouse. n = 5 in each group. **p 0.01 and ***p 0.001. Con, control; GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

After determining the inflammation induced by LPS, immunofluorescence staining of CD31 and $\alpha$-SMA was performed to assess the level of EndoMT. As shown in Fig. 2A, a notable increase in the co-localization of CD31 and $\alpha$-SMA was observed in the cortex of LPS-treated mice. In comparison to the control group, the expression levels of Occludin and ZO-1 were observed to decrease, while the expression level of $\alpha$-SMA increased in the LPS-treated group (Fig. 2B). Additionally, injection of LPS significantly enhanced the expression of Sestrin2 in the cortex of mice (Fig. 2C). These findings suggested that LPS treatment induced endothelial cell dysfunction and increased protein expression consistent with early EndoMT differentiation, accompanied by upregulated Sestrin2 expression.

Fig. 2.

LPS induced Endothelial-to-mesenchymal transition (EndoMT) in mouse cortex. (A) Immunofluorescence staining of CD31 (green), alpha smooth muscle actin ($\alpha$-SMA) (red) and DAPI (blue) in the cortex of control and LPS-injected mouse. Scale bar: 100 µm. (B) Representative blots showed the expressions of Occludin, zona occludens protein 1 (ZO-1) and $\alpha$-SMA in the cortex of control and LPS-injected mouse. (C) Representative blots showed the expression of Sestrin2 in the cortex of control and LPS-injected mouse cortex. n = 4 in each group. *p 0.05, **p 0.01 and ***p 0.001. Con, control; DAPI, 4${}^{\prime }$,6-diamidino-2-phenylindole; GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

Subsequently, HBMECs were subjected to LPS exposure in vitro. This exposure led to an upregulation of iNOS and Sestrin2 expression in HBMECs, as depicted in Fig. 3A. To investigate the impact of Sestrin2 on EndoMT, specific Sestrin2 siRNAs were designed and transfected into HBMECs. Among these siRNAs, Sestrin2 siRNA 1 exhibited the most effective knockdown efficiency in terms of Sestrin2 expression, as shown in Supplementary Fig. 1A. Consequently, Sestrin2 siRNA 1 was employed in subsequent experiments. Notably, knockdown of Sestrin2 exacerbated the reduced expression levels of Occludin and ZO-1 in the LPS-treated group (Fig. 3B). Simultaneously, knockdown of Sestrin2 facilitated the enhanced expression of $\alpha$-SMA in the LPS-exposed group (Fig. 3B). Then, Senstrin2 overexpressive lentivirus was constructed and elevated the expression of Senstrin2 in HBMECs (Supplementary Fig. 1B). Sestrin2 overexpression rescued the decreased expression levels of Occludin and ZO-1 in the LPS-treated group (Fig. 3C). Meanwhile, overexpression of Sestrin2 inhibited the enhanced expression of $\alpha$-SMA in the LPS-treated group (Fig. 3C). These results suggested that Sestrin2 overexpression attenuated EndoMT in HBMECs.

Fig. 3.

Sestrin2 overexpression attenuated EndoMT in human brain microvascular endothelial cells (HBMECs). (A) Representative blots showed iNOS and Sestrin2 expression in the control and LPS-exposed HBMECs. (B) Representative blots showed Occludin, ZO-1 and $\alpha$-SMA expression in the control and LPS-exposed HBMECs with/without Sestrin2 siRNA treatment. (C) HBMECs infected by Sestrin2 overexpressive lentivirus and the efficiency of Sestrin2 overexpression was detected. Representative blots showed Occludin, ZO-1 and $\alpha$-SMA expression in the control and LPS-exposed HBMECs with/without Sestrin2 overexpressive lentivirus infection. *p 0.05, **p 0.01 and ***p 0.001. Con, control; OE, overexpression.

Subsequently, we aimed to investigate the downstream effects of Sestrin2 in the LPS-induced EndoMT process. As shown in Fig. 4A, the expression of microtubule-associated protein1 LC3-II was observed to be elevated in the cortex of mice treated with LPS. Similarly, an increase in LC3B-II expression was observed in LPS-exposed HBMECs (Fig. 4B). LPS treatment increased the autolysosomes (black arrows) in HBMECs (Fig. 4C). Overexpression of Sestrin2 facilitated the LC3-II expression in HBMECs (Fig. 4D), whereas knockdown of Sestrin2 suppressed the expression of LC3-II in HBMECs (Fig. 4E). Furthermore, knockdown of Sestrin2 resulted in reduced expression levels of Occludin and ZO-1, which were initially inhibited by treatment with the autophagy inducer rapamycin (Fig. 4F). Additionally, the increased expression of $\alpha$-SMA in the Sestrin2 siRNA-treated group was inhibited by rapamycin (Fig. 4F). Knockdown of Sestrin2 activated phosphorylation of mTOR, which was inhibited by rapamycin (Supplementary Fig. 2). These results suggested that Sestrin2 regulated EndoMT depending on mTOR-mediated autophagy.

Fig. 4.

Sestrin2 regulated EndoMT depending on autophagy in HBMECs. (A) Representative blots showed the LC3-II expression in the cortex of control and LPS-injected mouse. (B) Representative blots showed the expression of LC3-II in the control and LPS-exposed HBMECs. (C) Transmission electron microscope photograph showed the autolysosomes (black arrows). Scale bar: 5 µm. (D) Representative blots showed the LC3-II expression in the control and LPS-exposed HBMECs with/without Sestrin2 overexpressive lentivirus infection. (E) Representative blots showed the LC3-II expression in the control and LPS-exposed HBMECs with/without Sestrin2 siRNA treatment. (F) Representative blots showed the LC3-II expression in the control and rapamycin-exposed HBMECs with/without Sestrin2 siRNA treatment. *p 0.05 and **p 0.01. Con, control; Rapa, rapamycin; LC3, light chain 3.