Sepsis is a life-threatening multiple organ failure caused by host immune response dysregulation to microbial infection, with high mortality (1, 2). For some reason, pathogens invade blood circulation, multiply (in the blood), and produce many toxins and metabolites, causing severe symptoms. Clinical presentations of sepsis are characterized by the body’s response to microbial infection, including fever, tachycardia, tachypnoea, white blood cell changes, and organ injury distant from the injection site, such as acute lung failure, acute liver injury failure, and acute kidney injury (3). During microbial infection, exogenous pathogen-associated molecular patterns and endogenous damage-associated molecular patterns disrupt most endothelial functions, leading to a procoagulant, antifibrinolytic, and proinflammatory state (4). The structure of the endothelial glycocalyx, a gel-like meshwork of glycosaminoglycans and proteoglycans surrounding the entire luminal surface of the endothelium, is disrupted early during sepsis, resulting in multi-organ failure later (5-7). Heparinase 1 is upregulated in sepsis, and is primarily responsible for glycocalyx degradation (8). Thus, the endothelium is considered a major target for sepsis therapy. However to date, no specific intervention targeting endothelial dysfunction and glycocalyx degradation exists.

Interferon-β (IFN-β) is well known for its anti-viral activity. However, increasing evidence suggests that IFN-β can benefit the host’s defense against bacterial infection (9, 10). We demonstrated that IFN-β protects against lethal endotoxic and septic shock via the upregulation of SIRT1 (11), a nicotinamide adenine dinucleotide-(NAD⁺)-dependent histone deacetylase that inhibits the immune response via the modulation of p53 and NF-κB functions (12, 13). The study showed that IFN-β increases SIRT1 expression in macrophages, inhibits proinflammatory cytokines, and protects against sepsis. The other group also showed that the intravenous treatment of IFN-β reduced the 28 d mortality by 81% in patients with acute respiratory distress syndrome, a disease with similar underlying mechanisms to sepsis (12). Many patients with sepsis do not have excessive immune activation, but do have immunosuppression (14-17). The SIRT1-mediated inhibition of immune function in macrophages cannot solely explain the protective effect of IFN-β against sepsis. Thus, in this study, we further investigated the role of IFN-β in sepsis, especially that pertaining to endothelial function modulation.

Interferon-β, in combination with nicotinamide riboside (NR), alleviates sepsis via endothelial SIRT1

Our previous study showed that the protective effect of IFN-β against sepsis was blocked by dominant negative SIRT1, indicating that SIRT1 is essential in the role of IFN-β. SIRT1 removes acetyl groups from target proteins using NAD⁺ as its substrate (18). NAD⁺ directly regulates SIRT1 activity by substrate-dependent activation (19). To enhance IFN-β efficacy by increasing intracellular NAD⁺, we combined NAD precursors, and tested the effect of the combinations on cecal ligation puncture-(CLP)-induced sepsis in mice. Intravenous administration of IFN-β (1,000 units/20 g) 6 h and 18 h after CLP challenge significantly improved survival rates by 40% on d 10, compared to that in control mice (Fig. 1A-D) (log-rank test, P < 0.01). The combination with NR, but not nicotinamide (NM) and nicotinamide mononucleotide (NMN), enhanced IFN-β-mediated improvement of the sepsis survival rate (Fig. 1A-D). These results suggest that NR is an enhancer in IFN-β-mediated sepsis protection.

To examine the effect of IFN-β plus NR on CLP-induced organ damage, lung, liver, and kidney samples were collected, and a histological examination was conducted. In CLP mice, hematoxylin and eosin (H&E) staining showed the kidney damage score, lung injury score, and liver inflammation score were increased, compared with sham mice (Supplementary Fig. 1A, B). Clinical score changes observed in IFN-β plus NR-treated mice had significantly weaker changes, compared with CLP mice (Supplementary Fig. 1B). The wet/dry weight ratio, an assessment of edema, was significantly increased in CLP mice, compared with sham mice (Supplementary Fig. 1C). IFN-β plus NR treatment significantly reduced CLP-induced increase in the wet/dry weight ratio (Supplementary Fig. 1C). These results suggest that IFN-β plus NR decreases CLP-induced multi-organ damage.

To examine whether endothelial SIRT1 is involved in IFN-β plus NR-mediated sepsis protection, we generated mice with the endothelial cell-(EC)-specific deletion of Sirt1 by crossing Sirt1^flox/flox mice with Tek-Cre transgenic mice expressing Cre-recombinase under the control of the EC-specific Tek promoter (Fig. 1E). Figure 1D shows the genotyping results. The EC-specific Sirt1 conditional knockout (EC-Sirt1 cKO) mice were viable and fertile, with a normal appearance. To further confirm that Sirt1 was deleted, we isolated ECs from the lung tissue using anti-CD31-coated magnetic beads. The western blot analysis showed a complete deletion of SIRT1 in lung endothelial cells from EC-Sirt1 cKO mice (Fig. 1F). The EC-specific deletion of Sirt1 abolished the protective effect of IFN-β plus NR on CLP-induced sepsis (Fig. 1G). These results indicate that IFN-β protects against sepsis via endothelial SIRT1.

Interferon-β upregulates SIRT1 in endothelial cells in a protein synthesis-independent manner

We found that lung ECs express SIRT1 (Fig. 1E). To determine the effects of IFN-β on SIRT1 expression, murine yolk sac ECs (MYSECs) were treated with IFN-β in a dose and time course. As shown in Fig. 2, MYSECs were treated with 100, 200, 400, or 1,000 U/ml of IFN-β for 24 h. These treatments resulted in a dose-dependent increase in SIRT1 protein levels (Fig. 2A). Additionally, when MYSECs were treated with 1,000 U/ml of IFN-β for 0, 2, 6, or 24 h, IFN-β significantly increased SIRT1 protein expression in a time-dependent manner (Fig. 2B).

To determine the effects of IFN-β on SIRT1 transcription, MYSECs were treated with IFN-β in a dose and time course as described above, and Sirt1 mRNA levels were then detected by real-time polymerase chain reaction (RT-PCR). Following treatment with IFN-β, Sirt1 mRNA levels decreased, rather than increased, in a dose- and time-dependent manner (Fig. 2C, D).

We investigated the effect of protein synthesis inhibition on IFN-β-induced increase in SIRT1 protein levels in MYSECs. The inhibition of protein synthesis by cycloheximide (CHX) did not decrease IFN-β-induced increase in SIRT1 protein levels (Fig. 2E), indicating that IFN-β upregulates SIRT1 protein levels by protein stabilization, rather than protein synthesis.

Interferon-β plus NR preserves endothelial integrity via SIRT1 during bacterial sepsis

We performed a transwell assay to evaluate the effects of IFN-β on lipopolysaccharide-(LPS)-induced MYSEC permeability. LPS stimulation (300 ng/ml) significantly increased MYSEC permeability by 1.8-fold, compared with vehicle treatment. By contrast, IFN-β plus NR treatment significantly suppressed the hyperpermeability of MYSECs induced by LPS (Fig. 3A). The Sirt1 knockdown with Sirt1 siRNA abolished the suppression of LPS-induced MYSEC hyperpermeability by IFN-β plus NR treatment (Fig. 3B). These results suggest that IFN-β plus NR regulates endothelial permeability via SIRT1. To confirm this in vivo, we examined the kidney, lung, and liver permeability 18 h after CLP. IFN-β plus NR treatment significantly reduced CLP-induced increase in Evans blue dye (EBD) infiltration into the kidney, lung, and liver compared with the vehicle, when IFN-β plus NR was administered 6 h after CLP (Fig. 3C). IFN-β plus NR treatment increased SIRT1 levels in the lung and liver, suggesting SIRT1 has a role in endothelial permeability in vivo (Supplementary Fig. 2). EC-Sirt1 cKO abolished the suppression of CLP-induced increase in EBD infiltration into the kidney, lung, and liver by IFN-β plus NR treatment (Fig. 3C). These results indicate that IFN-β plus NR protects against sepsis via the SIRT1-mediated blockage of endothelial hyperpermeability.

Interferon-β plus NR restores CLP-induced endothelial glycocalyx damage by modulating the SIRT1/heparinase 1 pathway

Since endothelial glycocalyx is essential to maintaining vascular permeability, we examined the effect of IFN-β plus NR treatment on CLP-induced endothelial glycocalyx damage. We demonstrated that EC-Sirt1 KO decreased the expression of heparan sulfate proteoglycan (HSPG) in the lungs (Fig. 4A). CLP operation significantly reduced the expression of HSPG by 90%, compared to the vehicle treatment. In contrast, IFN-β plus NR treatment significantly blocked the decrease of HSPG expression induced by CLP (Fig. 4B). Heparinase 1 (HPA1) is a crucial enzyme involved in maintaining endothelial glycocalyx integrity by regulating degradation (20). We evaluated the effects of IFN-β plus NR on HPA1 expression upon LPS stimulation in MYSECs. LPS stimulation significantly increased the expression of HPA1 by two-fold in MYSECs, compared to the vehicle treatment. In contrast, IFN-β plus NR treatment significantly suppressed the expression of HPA1 induced by LPS (Fig. 4C). Sirtuin 1 knockdown with Sirt1 siRNA abolished the suppression of LPS-induced increase in HPA1 expression by IFN-β plus NR treatment (Fig. 4C). These results suggest that IFN-β plus NR protects endothelial damage during sepsis via the SIRT1-mediated suppression of HPA1 expression.

Mice

Seven to eight-week-old male C57BL/6 mice weighing between 20 and 24 g were purchased from Nara-Biotec (Seoul, Korea). The mice were maintained in an environment with a temperature of 22 ± 2°C and a 12-h-light-dark cycle. Food and water were supplied ad libitum. All animal-related experiments were approved by the Institutional Animal Care and Use Committee of Jeonbuk National University (JBNU-2021-099). We conducted all mice-related experiments following the guidelines of the commitee. For the in vivo experiments, mice were randomly assigned to experimental or control groups, with 10 or 20 individuals for each group.

Generation of EC-Sirt1 cKO mice

The Sirt1^loxp/loxp and Tek-Cre (Tek^Cre/+) mice were purchased from the Jackson Laboratory (Bar Harbor, ME, USA). The Sirt1^loxp/loxp mice were crossed with Tek-Cre mice in which the expression of Cre-recombinase is under the control of the endothelial-specific Tek promoter. These two strains were crossed and maintained on a C57BL/6J background. We first generated double heterozygous mice for Cre and floxed Sirt1 (Tek::PKD1^fl/+); those were then bred to Sirt1^fl/fl mice via a back-mating strategy to generate excised floxed Sirt1 homozygous mice (Tek::Sirt1^fl/fl; used as conditional knock-out), and Cre-null mice (Sirt1^fl/+ and Sirt1^fl/fl); used as littermate controls (wild-type, WT) throughout the study. These mice were born at the expected Mendelian frequency. Genotyping was conducted by PCR of the tail lysate. The PCR was performed using the KAPA2G Fast HotStart Genotyping Mix (Sigma‐Aldrich, St. Louis, MO, USA) according to the manufacturer’s cycling protocol for genomic DNA. The PCR products were resolved on a 5% agarose gel containing SYBR Safe (ThermoFisher Scientific, Waltham, MA, USA) and imaged using a UV gel documentation system.

Cecal ligation and puncture

After mice were anesthetized using 150 mg/kg ketamine and 17.5 mg/kg rompun, a 1-2 cm ventral midline abdominal incision was made on the shaved and disinfected skin of the abdomen. The cecum was then exposed, ligated with 6-0 silk suture just terminal to the ileocecal valve to prevent intestinal obstruction, and punctured through with an 18-gauge needle. The punctured cecum was gently squeezed to extrude a 1-2 mm droplet of feces and then returned to the abdominal cavity. Then, the abdomen was closed, and the mice were intraperitoneally injected with pre-warmed normal saline (2.5 ml/100 g body weight) immediately after the procedure.

Survival analysis

Mice were intravenously injected with IFN-β (2 μg/kg) (R&D Systems, Minneapolis, MN, USA), NM (40 μmole/kg) (Sigma-Aldrich), NMN (40 μmole/kg) (Sigma-Aldrich), or NR (40 μmole/kg) (Medkoo Biosciences, Morrisville, NC, USA) twice daily at 6 h and 18 h after CLP. The survival time was recorded daily up to 10 days after CLP or sham surgery.

Endothelial cell isolation from mouse lung

Seven to eight-week-old WT and EC-Sirt1 cKO mice lungs were minced and digested with 3 mg/ml of collagenase II at 37°C for 45 min, stirring every 15 min. The digested tissues were passed through a 21 G cannula attached to a 20-ml syringe and filtered through a 40-μm cell strainer on top of a 50-ml Falcon tube that contained serum-containing isolation buffer for stopping digestion. After centrifugation at 300 × g at 4°C for 10 min, the ECs in the cell pellets were isolated using anti-CD31 Microbeads and the MACS system (Miltenyl Biotec, Bergisch Gladbach, North Rhine-Westphalia, Germany) according to the manufacturer’s instructions.

Murine yolk sac endothelial cell culture and siRNA transfection

The MYSECs were purchased from ATCC (Manassas, VA, USA) and cultured in 5% CO₂ at 37°C in Dulbecco’s Modified Eagle Medium supplemented with 10% fetal bovine serum, 0.2% glutamax, and 0.5% penicillin-streptomycin. For the Sirt1 knockdown, negative control siRNA (Invitrogen, Carlsbad, CA, USA) or Sirt1-siRNA were transfected at a concentration of 50 nm with Transfection Reagent (Invitrogen) diluted in siRNA Transfection Medium (Invitrogen). The sequences of the Sirt1-siRNAs were as follows: sense (5’-TCGTGGAGACATTTTTAATCAGG-3’) and antisense (5’-GCTTCATGATGGCAA GTGG -3’).

Western blot

Tissue and cell lysates (20 μg) were loaded, run on 7.5% sodium dodecyl sulfate-polyacrylamide gels, and transferred to polyvinylidene fluoride membranes as previously described (21). Membranes were blocked with 3% nonfat milk in Tris-buffered saline (pH 7.4) with 0.1% Tween 20 for 1 h at room temperature. The membranes were incubated with primary antibodies against SIRT1 (B-7, sc-74465, 1:1000) (Santa Cruz Biotechnology, Dallas, TX, USA), heparinase 1 (ab228660, 1:500) (Abcam, Cambridge, Cambridgeshire, UK), and β-actin (A5441, 1:5000) (Sigma) overnight at 4°C. After washing, the membranes were incubated with horseradish peroxidase-(HRP)-conjugated secondary antibody for 1 h at room temperature. Protein bands were visualized using an electrochemical luminescence system (Intron Biotechnology, Inc., Seongnam-si, South Korea) and a Fusion Fx7 Spectra (Vilber Lourmat, Collégien, France). Protein expression was quantified using Fusion-Capt software (version 16.08; Vilber Lourmat).

In vitro endothelial monolayer permeability assay

The MYSECs (2.5 × 10⁵ cells/well) were cultured as described above on the top chamber of the transwell inserts (0.4 μm; Corning B.V. Life Sciences, The Netherlands) for 24 h with 300 ng/ml of LPS in the presence of 40 ng/ml of IFN-β and/or 800 μM NR. The media in the upper chambers were exchanged with 300 μl of serum-free media containing 5 μl of streptavidin-HRP (R&D Systems). After 5 min, 20 μl of the media in the lower chambers were transferred to new 96-wells and analyzed for HRP activities by adding 100 μl of 3,3’,5,5’-tetramethylbenzidine substrate (R&D Systems). The absorption at 450 nm was acquired with a microplate reader (Molecular Devices, Sunnyvale, CA, USA).

In vivo vascular permeability assay

We evaluated vascular leakage using EBD (Sigma-Aldrich) extravasation. Eighteen hours after CLP surgery, 200 μl of 0.5% EBD was administered via the caudal vein. After the dye circulated for 1 h, the organs were excised and imaged. After imaging, the organ tissues were blotted dry, weighed, and homogenized in formamide (Sigma-Aldrich). Following 48-h extraction at 55°C, the tissue fluid was centrifuged at 12,000 × g for 25 min. A microplate reader (Molecular Devices, San Jose, CA, USA) evaluated the EBD concentration of the supernatant at 620 nm absorbance.

Real-time polymerase chain reaction

For the RT-PCR analysis, the total RNA was extracted from cells or tissues using RNeasy Plus Mini Kit (Qiagen, Dusseldorf, Germany). High-fidelity cDNA was generated from each RNA sample using the Superscript III cDNA amplification system (Invitrogen). Quantitative RT-PCR samples were prepared as a mixture using the Quantitect SYBR Green PCR kit (Takara Bio, Shiga, Japan) following the manufacturer’s instructions. The reaction was performed at 65°C for 5 min, followed by 42°C for 1 h and 95°C for 5 min using a MyCycler^TM Thermal Cycler (Bio-Rad, Hercules, CA, USA).

Immunofluorescence

The lung sections were incubated with 5% bovine serum albumin (BSA) + 5% goat serum in phosphate-buffered saline (PBS) with 0.05% Tween-20 for 1 h at room temperature. The slides were then incubated with an anti-HSPG antibody (Invitrogen) at 4°C overnight. After washing, the slides were incubated with Fluorescein isothiocyanate (FITC) conjugated anti-rat antibody (Invitrogen) for 1 h at 37°C. Slides were counterstained with 4’,6-diamidino-2-phenylindole (DAPI).

Statistical analysis

Data were presented as the mean ± standard error. A student’s test was performed to confirm the significance of values, and P < 0.05 was considered significant.

This work was supported by the National Research Foundation of Korea [grant number 2017R1A5A2015061].

S. D. and M.-K. H are inventors of a patent related to this study. The authors have no conflicting interests.