Participant selection criteria and definition

This study protocol was approved by the ethics committee of the Shaoxing People’s Hospital (No. 2024-072-01) and was performed in accordance with the Declaration of Helsinki. All patients had authorized family members sign an informed consent form. This single-center prospective cohort study included data from 57 patients with sepsis who were admitted between January and December 2024. Using the random number table method, participants were allocated to 1 of 2 groups: DEX (DEX in addition to standard care); and non-DEX (standard treatment). DEX was administered via continuous intravenous infusion at 0.2 to 1.5 µg per kilogram of body weight per hour for > 24 h, based on a previous study⁵⁶. The inclusion criteria: age ≥ 18 years; fulfilling the Sepsis-3 criteria; and ICU stay > 24 h and up to 100 days. The exclusion criteria: history of chronic gastrointestinal diseases or chronic renal dysfunction; pregnancy; and long-term use of hormonal drugs or undergoing immune-related treatments. Additionally, a control group comprising 30 non-sepsis ICU individuals, matched to the sepsis cohort for age and gender, was included. Inclusion criteria included ICU admission for other diagnoses and does not meet the diagnostic criteria for sepsis 3.0. Exclusion criteria included sepsis, chronic inflammatory or fibrotic diseases, immunosuppressive therapy, malignancy, or pregnancy.

Data collection

Baseline demographic data were collected for all patients, and clinical severity scores (Acute Physiologic Assessment and Chronic Health Evaluation [APACHE] II and Sequential Organ Failure Assessment [SOFA]) were assessed on enrollment. General patient information, including age, sex, infection sites (abdominal or non-abdominal infection), smoking, alcohol consumption, hypertension, and diabetes, was recorded. Baseline data and clinical indicators were collected and examined. Blood samples were obtained 3 days post-DEX treatment to analyze serum inflammatory cytokine levels and alterations in SIRT3 expression (Cat. SEE913Hu). In addition, serum levels of intestinal fatty acid binding protein (I-FABP, Cat. SEA174Hu), TNF-α (Cat. SEA133Hu), and IL-6 (Cat. SEA079Hu), and were determined using a commercially available ELISA kit (Cloud-Clone Corp., Wuhan, China).

Animal studies

All animal experiments were carried out in accordance with relevant guidelines and regulations.

All procedures involving animals complied with the ARRIVE guidelines (Reporting of In Vivo Experiments) and the NIH Guide for the Care and Use of Laboratory Animals, following approval by the Institutional Animal Care and Use Committee of Shaoxing People’s Hospital (Approval No. 2023Z047). Male C57 mice (8 weeks old; weight, 20–25 g) were housed with ad libitum access to food and water. Mice were acclimated for 7 days in the environment, with a mean room temperature of 25 ± 2 °C, humidity of 55 ± 10%, and a 12 h light–dark cycle. To determine the protective effect of DEX (Cat. HY-12719, MedChemExpress, Shanghai, China) on septic-induced intestinal injury, the mice were assigned to 4 groups (n = 10 per group): sham; CLP; CLP + low-dose DEX (40 µg/kg [DEX40]); and CLP + high-dose DEX (80 µg/kg [DEX80]). To elucidate the mechanisms underlying the protective effects of DEX in septic mice, the mice were divided into the following groups (n = 6 per group): sham; CLP; CLP + DEX; and CLP + DEX + SIRT3 inhibitor 3-TYP (Cat. HY-108331, MedChemExpress). DEX and 3-TYP were administered intraperitoneally (i.p.). The DEX concentration was determined based on preliminary experiments and previous reports, to achieve clinically relevant exposure levels^57,58,59. 3-TYP (5 mg/kg) was administered i.p. at 1 h before CLP procedure¹⁶. Surgery was performed 1 ‍h after administration of DEX and 3-TYP. Another dose of DEX was administered 6 h after surgery. The mice in the sham and CLP groups were injected i.p. with an equivalent volume of normal saline.

All experimental groups were euthanized 8 h after the final treatment. The mice were anesthetized using sevoflurane and euthanized by rapid cervical dislocation. Whole blood was collected and maintained at 4 °C for 15 min to facilitate clot formation, then centrifuged at 3,000 × g for 15 min at 4 °C. Serum were flash-frozen in liquid nitrogen and stored at − 80 °C until biomarker analysis. Intestinal segments rinsed in ice-cold PBS (pH 7.4) to remove luminal contents, and stored at -80 °C for subsequent molecular analyses.

Establishment of the mouse CLP model

The CLP method, a well-established polymicrobial sepsis model validated in prior studies⁶⁰, was used to establish a mouse model of sepsis. Mice were anesthetized via i.p. injection of ketamine (80 mg/kg) and xylazine (10 mg/kg)⁶¹. A 1.5 cm midline laparotomy was performed under aseptic conditions to exteriorize the cecum. The distal cecum was ligated at 50% of its length using 6 − 0 silk suture and punctured twice with a 21-gauge needle to extrude a minimal fecal column (~ 1 mm). The cecum was gently repositioned, and the abdominal cavity was closed in two layers. Postoperatively, mice received 1 mL of pre-warmed sterile saline (37 °C) subcutaneously for resuscitation and were recovered on a heating pad (37 °C) for 4 h. Sham-operated controls underwent identical surgical procedures, excluding cecal ligation and puncture.

Histology and Immunofluorescence staining

Colonic tissue specimens were fixed in 4% paraformaldehyde and embedded in paraffin. Serial Sect. (5 μm) were deparaffinized using xylene and rehydrated using a graded ethanol series, followed by rinsing with distilled water. Hematoxylin and eosin (H&E) staining kit (Cat. G1120, Solarbio, Beijing, China) was performed for morphological evaluation. Histological changes in intestinal segments were evaluated using the Chiu et al.⁶² scoring system for hematoxylin and eosin-stained paraffin-embedded specimens. Slides were graded as follows: 0 (normal mucosa), 1 (widened subepithelial space), 2 (epithelial lifting), 3 (moderate to severe intra- or subepithelial edema), 4 (transmucosal infarction), or 5 (transmural infarction). Two independent pathologists, blinded to the experimental groups, performed the scoring, with disagreements resolved by consensus. Mean Chiu scores were computed for each group.

For immunofluorescence staining of tissue sections, slides were incubated with 0.2% Triton X-100 in PBS (pH 7.4) for 5 min and blocked with 5% bovine serum albumin (BSA) for 30 min at room temperature. Sections were incubated overnight at 4 °C with the following antibodies: anti-SIRT3 (1:400, ab217319), Anti-ZO-1 (1:200, ab221547) and anti-claudin-1 (1:200, ab211737; Abcam, Cambridge, United Kingdom). The following day, after washing with PBS, the slides were incubated with Alexa Fluor 488 or Alexa Fluor 647 secondary antibodies (1:500, ab1500077 and ab150116, respectively [Abcam]) for 30 min at room temperature in the dark. Nuclei were counterstained with DAPI (Cat. C0065, Solarbio). Images were acquired udner a confocal microscope (Leica, Wetzlar, Germany). Fluorescence intensity was measured using ImageJ software (Rasband, W.S., ImageJ, USA, ). to analyze mean fluorescence in randomly selected regions.

Terminal Deoxynucleotidyl transferase dUTP Nick end labeling (TUNEL) staining

To detect apoptotic cells in the intestinal epithelium, a commercially available kit (One-Step TUNEL Apoptosis Assay Kit, Cat. C1090, Beyotime, Jiangsu, China) was used to detect apoptotic cells, in accordance with manufacturer’s instructions. Deparaffinized tissue sections were incubated with proteinase K (20 µg/mL in 10 mM Tris-HCl, pH 7.4) for 15 min at 37 °C, followed by permeabilization with 0.1% Triton X-100 in PBS for 10 min at 4 °C. Sections were then incubated with TdT enzyme and FITC-12-dUTP reaction mix for 60 min at 37 °C in a humidified chamber, and subsequently incubated with DAPI. TUNEL-positive cells (FITC+/DAPI+) were visualized using a fluorescence microscope and apoptotic indices were calculated as the percentage of TUNEL + cells per 100 epithelial cells across five random fields per section using ImageJ.

ELISA

The supernatant of homogenized intestinal tissues (9 mL, 0.9% physiological saline and 1 g tissue) was collected after centrifugation at 12,000 rpm for 10 min at 4 °C. The supernatant of Caco-2 cells was also collected after treatment. The concentration of IL-1β (Cat. EK0392), IL-6 (Cat. EK0410) and TNF-α (Cat. EK0525) were tested in each group by using collected homogenate; the experimental method steps were performed in strict accordance with kit instructions (BOSTER Biological Technology Co. Ltd., Wuhan, China).

Cell culture and treatments

The human intestinal Caco-2 cell line (Cell Bank of the Chinese Academy of Sciences, Shanghai, China) was cultured in DMEM supplemented with 10% fetal bovine serum (FBS) (Cat. A5256701, Gibco, Waltham, MA, USA), 100 U/ml penicillin and 100 mg/ml streptomycin at 37 °C in a humidified atmosphere of 5% CO₂. After a 24-hour post-seeding adherence period, Caco-2 cells were treated with DEX at concentrations of 0, 2, 5, 10, 20, and 40 µM for 24 h. The optimal DEX concentration was determined using the CCK-8 assay based on the cell viability results. To model sepsis-induced intestinal barrier injury, cells pre-treated with DEX for 2 h were co-stimulated with 100 ‍µg/ml LPS (Cat. L5418, Sigma) for 24 h. LPS concentrations were obtained from a previous study.^3,63.

Transfection of plasmids and small interfering RNA

Small interfering RNA (siRNA) (Obio Technology Corp. Ltd., Shanghai, China) was used to generate si-SIRT3 or a negative control (si-NC), which was then transfected into Caco-2 cells using Lipofectamine 3000 (Cat. L3000015, ThermoFisher, Waltham, MA, USA), in accordance with manufacturer’s instructions.

Cell proliferation analysis

Caco-2 cells were placed in 96-well plates at a density of 5 × 10³ cells/well. The cells were treated with LPS or DEX. The cells were then cultured using DMEM supplemented with 10% FBS for 48 h at 37 °C in a 5% CO₂ environment. Following the incubation period, 10 µL CCK-8 reagent (Cat. HY-K0301, MedChemExpress) was added to each well, incubated for 2 h at 37 °C. Cell viability was assessed by measuring optical density at 450 nm using a microplate reader.

Western blot

Proteins were extracted from cellular and tissue samples using RIPA buffer (Beyotime, Jiangsu, China). An assay using BSA was used to determine protein concentrations. An equal amount of proteins from each sample was added to a Millipore-provided 10% sodium-dodecyl polyacrylamide gels (SDS–PAGE) to perform electrophoresis. The resolved proteins were then electro-transferred to polyvinylidene difluoride (PVDF) membranes, which were subsequently blocked using 5% non-fat milk in TBST. For antibody incubation, the blocked membranes were incubated at 4 °C overnight with primary antibodies that targeted SIRT3 (dilution 1:1000, ab217319), ZO-1 (1:500, ab221547), Bax (1:1000, ab32503) occludin (1:2000, ab216327), claudin-1 (1:1000, ab211737). Bcl-2 (1:1000; #3498; Cell Signaling Technology, Danvers, MA, USA), JAK2 (1:1000; ab108596), p-JAK2 (1:1000; ab32101), and STAT3 (Cat.10253, Proteintech, Rosemont, IL, USA) p- STAT 3 (Cat. 80199, Proteintech) and GAPDH (Cat.60004, Proteintech). After washing 3 times with PBS 3, the membranes were incubated with HRP-conjugated goat anti-rabbit IgG secondary antibody (1:5000 dilution) for 30 min at room temperature. Protein bands were visualized using the BeyoECL Plus chemiluminescence substrate (Cat. P0018FM, Beyotime), and the luminescence signals were measured using ImageJ software.

Measurement of transepithelial electrical resistance

Transepithelial electrical resistance (TEER) was quantified as previously described⁶⁴. Briefly, Caco-2 cells were seeded into a 6-well transwell culture plate (0.4 μm, Corning, Corning NY, USA) at a density of 1 × 10⁵ cells/ml. Cells were maintained until 80% confluency, 1 µg/mL LPS, 5 mmol/L ATP, 1 µmol/L parthenolide (PTL), or nano PTL were treated. Transepithelial electrical resistance was quantified within 12 h using a voltohmmeter (EVOM2, World Precision Instruments, Sarasota, FL, USA) in accordance with the EVOM2 operating manual. Measurements were taken at baseline (0 h) and within 12 h post-treatment. Blank inserts (cell-free) were used to determine background resistance TEER was calculated as (units, Ω·cm²: (measured resistance value – blank resistance value) × transwell effective membrane area.

Statistical analysis

Data were analyzed using Prism version 10.5 (GraphPad Inc., San Diego, CA, USA). Continuous data are presented as mean ± standard deviation (SD). Normality was assessed using the Shapiro-Wilk test, and homogeneity of variance was confirmed via Brown-Forsythe test. Statistical analyses between tow groups were performed using the unpaired Student’s t test (two-tailed) and one-way analysis of variance (ANOVA) with Tukey post-hoc test was used for the comparison among multi-groups. For non-normal continuous or ordinal data, the Kruskal-Wallis test was used, followed by Dunn’s test with Bonferroni correction. Categorical variables were compared using the chi-square test or Fisher’s exact test, as appropriate. All experiments included biological triplicates with technical triplicates per condition. Differences with P