Rabbit polyclonal anti-ZO-1, mouse monoclonal anti-occludin, and mouse monoclonal anti-claudin5 were obtained from Thermo Fisher Scientific (Waltham, MA, USA). Rabbit monoclonal anti-β-actin, anti-phospho-NF-κB p65, and anti-NF-κB p65 were purchased from Cell Signaling Technology (Danvers, MA, USA). Goat polyclonal anti-Axl, anti-Mertk, and rat monoclonal anti-Tyro3 antibodies were obtained from R&D Systems (Minneapolis, MN, USA). Rabbit polyclonal anti-phospho-Axl, anti-phospho-Mertk, and anti-phospho-Tyro3 were purchased from Affinity (Cincinnati, OH, USA). Horseradish peroxidase (HRP)-conjugated and Alexa Fluor-conjugated secondary antibodies were purchased from Biosharp (Hefei, China). Axl small interfering RNA (siRNA) or nonspecific control siRNA with the recommended transfection reagents were all from RiboBio (Guangzhou, China). LPS (O55:B5) was purchased from Sigma-Aldrich (Oakville, Ontario, Canada). The In Vitro Vascular Permeability Assay kit was obtained from Merck Millipore (Etobicoke, Ontario, Canada). Recombinant Mouse Gas6 protein was from R&D Systems (Minneapolis, MN, USA). The chemiluminescence (ECL) Western blot detection kit was from Wenke Biotechnology (Zhejiang, China).

Mouse aortic endothelial cells (MAECs) were derived and cultured as previously described (Wang et al., 2016). Briefly, cells were cultured in endothelial cell growth medium (Heidelberg, Germany) supplemented with 1% penicillin/streptomycin, 10% fetal bovine serum (FBS), and endothelial cell growth supplement at 37°C with 5% CO₂ and were used from passage 2–3 when grown to confluent monolayers.

Human umbilical vein endothelial cells (HUVECs) were purchased from the Shanghai Institute of Cell Biology. HUVECs were grown in endothelial cell growth medium (Heidelberg, Germany) supplemented with 1% penicillin/streptomycin, 10% FBS, and endothelial cell growth supplement. Cells were cultured in an atmosphere of 5% CO₂ at 37°C and used at passages 5–6.

Endothelial cells (ECs) were serum-starved (1% serum) for 6 h before treatment and divided into four groups randomly: 1) control group: cells without any stimulation; 2) LPS group: cells were stimulated with LPS (50 μg/L, 100 μg/L, and 200 μg/L) for 12 h; 3) Gas6 group: cells were incubated with Gas6 (200 ng/mL); and 4) Gas6 + LPS group: cells were pretreated with Gas6 (100 ng/mL, 200 ng/mL, and 400 ng/mL) for 2 h and then treated with LPS (200 μg/L) for 12 h. In some experiments, cells were transfected with Axl siRNA (siAxl) or nonspecific control siRNA (siNC) for 36 h before being treated with Gas6 and LPS as described.

In all experiments in vitro, both MAECs and HUVECs were subjected to transwell permeability assay while MAECs were used in Western blotting and HUVECs were used in immunofluorescence.

An in vitro Vascular Permeability Assay kit was used to measure the passage of fluorescein isothiocyanate (FITC)-dextran across the endothelial monolayer according to the manufacturer’s protocol. Briefly, cells were seeded at 2 × 10⁴ cells/insert on the upper chamber with collagen coating for 2–3 days to reach full confluence. ECs were incubated with Gas6 for 2 h before being stimulated with LPS for 12 h. Then, 150 μL of FITC-dextran in basal media (1:40) was added to the upper chamber and allowed to permeate through the monolayers. After 45 min, 100 μL of samples was removed from both the upper and lower chambers for fluorescence determination using a fluorescence microplate reader (SpectraMax M2, CA, USA) with an excitation wavelength of 485 nm and an emission wavelength of 535 nm. The permeability of the endothelial monolayer was evaluated by the permeability coefficient of dextran calculated as follows: Pd = [A]/t × 1/A × V/[L], where [A] is the FITC-dextran concentration in the bottom chamber and [L] is the FITC-dextran concentration in the upper chamber, A refers to the area of the membrane in cm², V is the volume of the bottom chamber, and t indicates time in minutes.

Male C57BL/6 mice (20–25 g) were obtained from Shanghai Slack Laboratory Animal Center [license: SCXK (HU) 2012-0002]. Mice were acclimatized in an environment with adequate temperature and humidity and had free access to water and chow for 7 days. All experimental protocols were performed according to the National Institutes of Health Guide for the Care and Use of Laboratory Animals, and all methods were approved by the ethics committee of the Laboratory Animal Ethics Committee of Wenzhou Medical University. Mice were randomly divided into the sham group, Cecal ligation and puncture (CLP) + vehicle group, and sham + Gas6 6 μg group.

Cecal ligation and puncture were performed as described previously (Wichterman et al., 1980; Nemeth et al., 2009). In short, after anesthetization by amobarbital sodium (50 mg/kg), we ligated the cecum at the distal two-thirds by silk 4-0 and punctured it with an 18-gauge needle. A small amount of fecal content was squeezed from holes, and the cecum was returned to the abdominal cavity. After we closed the abdomen in two layers, 1 ml of sterile saline was administered. In the sham group, we located the cecum but did not perform the ligation and puncture. In the treatment group, after CLP, 6 μg of Gas6 was administered via the tail vein immediately, and in the CLP group, animals received the same volume of normal saline.

Evans blue dye extravasation assays (Xu et al., 2001) were performed 1 day after CLP. In brief, 0.2 mL of Evans blue dye (2%) was injected via the tail vein. After 2 h, the anesthetized mice were perfused with saline until no more blue dye came out of the right atrium. The lung and kidney were removed and weighed. Each tissue was then incubated in formamide at 72°C for 3 days and measured by a spectrophotometer at an excitation wavelength of 610 nm and an emission wavelength of 680 nm after the suspension was centrifuged at 10,000g for 25 min. The Evans blue extravasation was calculated as μg/g of tissue using a standard curve.

Lung and kidney tissues were fixed in 10% formalin and then embedded in paraffin for sectioning into 4-μm-thick sections. Tissues were stained with hematoxylin–eosin. The slides were reviewed in a blinded manner under an optic microscope (NIKON Eclipse Ci, Japan). The severity of lung injury was scored from 0 (normal) to 3 (severe), based on the degree of neutrophil infiltration, hemorrhage, interstitial edema, and alveolar congestion. The total score was calculated by adding up the individual scores of each category. The pathological changes in the kidney tissue were scored with a semi-quantitative scale to evaluate changes. Tubular damage was defined as tubular epithelial swelling, brush border losing, vacuolization, cellular necrosis, cast formation, and desquamation. The degree of kidney damage was determined according to the percentage of damaged tubules per all tubules by the following criteria: 0 (none), 1 (< 10%), 2 (10–25%), 3 (25–50%), 4 (50–75%), and 5 (>75%).

Cells were treated as described previously and harvested by cold Radioimmunoprecipitation Assay (RIPA) lysis buffer containing protease inhibitors and phosphatase inhibitors. After centrifugation at 12,000g for 20 min at 4°C, the protein concentration was assayed using the Bicinchoninic acid (BCA) assay. Equal amounts of protein (30 μg) were separated by 10–12% sodium dodecyl sulfate (SDS)-Polyacrylamide gel electrophoresis (PAGE) and transferred onto 0.45-μm polyvinylidene fluoride (PVDF) membranes (Millipore, CA, USA). After the membranes were blocked with 5% fat-free dry milk in Tris-buffered saline containing Tween-20 (TBST), they were incubated with anti-Tyro3 (1:1,000), anti-Axl (1:1,000), anti-claudin-5 (1:500), anti-occludin (1:1,000), anti-ZO-1 (1:500), anti-NF-κB p65, anti-phospho-NF-κB (1:1,000), phospho-Tyro3, anti-phospho-Axl, and anti-phospho-Mertk (1:500) at 4°C overnight. Membranes were washed three times and incubated with the HRP-conjugated secondary antibodies. Protein bands were visualized by using ECL reagents and performed with the ChemiDicTM XRS+ Imaging System (Hercules, CA, USA). Densitometry analysis of each band was quantified using Image Lab software and normalized against those of β-actin protein in each sample.

Cells were fixed in 4% paraformaldehyde for 15 min at room temperature and then washed three times with phosphate-buffered saline (PBS). After the cells were blocked with 1% bovine serum albumin (BSA) for 60 min at room temperature, they were incubated with antibodies against NF-κB (1:400), anti-ZO-1 (1:50), anti-claudin-5 (1:50), and occludin (1:100) at 4°C overnight. After the samples were washed three times, they were incubated with goat anti-rabbit Alexa Fluor 584-conjugated secondary antibody (1:400) for 2 h at room temperature. The cells were then counterstained with 4′6-diamidino2-phenylindole (DAPI) (Solarbio, Beijing, China) for 5 min and washed three times. Finally, the cell images were visualized using a Nikon A1Rsi confocal microscope (Nikon, Japan).

According to the manufacturer’s recommendations, cells were grown to 50% confluency and then washed with PBS twice before transfection with siRNA. Cells were transfected with 10 nM mouse Axl siRNA (siAxl) or nonspecific control siRNA (siNC) using riboFECT^™ buffer. Cells were transfected for 36 h, and the silencing efficiency was confirmed by Western blotting.

Data are expressed as the mean ± SD. All data analyses were carried out using GraphPad Prism 7 (GraphPad Software, La Jolla, CA, USA). Statistical significance between different groups was assessed by one-way ANOVA followed by Dunnett’s multiple comparison tests. Comparisons between two groups were made using Student’s t test. P < 0.05 was considered statistically significant.

To confirm the effect of Gas6 on the acute kidney and lung injury caused by CLP, histological staining was performed, and the representative histological microphotographs are shown in Figure 1A . Compared to the sham group, lung tissues showed marked pathological lesions characterized by edema, hemorrhage, and inflammatory infiltration in the CLP group, while treatment with Gas6 could mitigate the severity of pathological damages in a certain degree. Pathological scores were correspondingly elevated in the CLP group and significantly decreased in the Gas6-treated groups as quantified in Figure 1B . In kidney tissues, pathological lesions of tubular epithelial swelling, brush border injury, vacuolization, cellular necrosis, and cast formation were enhanced in the CLP group as compared with the sham group, whereas it was attenuated by Gas6 treatment. As quantified in Figure 1C , Gas6 treatment reduced the histological damage scores of the kidney.

To determine whether Gas6 ameliorated vascular hyperpermeability caused by sepsis, we performed an Evans blue dye extravasation assay. As shown in Figure 1D , 24 h after CLP, the mean Evans Blue dye accumulation in the lung and kidney was increased fourfold in the CLP group compared with the sham group. Remarkably, compared to the CLP group, the Gas6-treated group showed significantly reduced Evans blue dye accumulation. Gas6 alone did not affect Evans blue dye extravasation.

The effects of LPS and Gas6 on endothelial permeability in MAECs and HUVECs were examined by analysis of Pd with FITC-dextran, an indicator of endothelial cell permeability. As shown in Figure 1E , LPS significantly increased the endothelial permeability in a dose-dependent manner within 12 h. Meanwhile, as shown in Figure 1F , LPS at a concentration of 200 μg/L significantly increased the HUVECs endothelial permeability within 12 h. In contrast, when MAECs and HUVECs were pretreated with Gas6 for 2 h and then challenged with LPS for another 12 h, Gas6 reduced permeability, and this reduction was substantial when Gas6 was administered at 400 ng/mL ( Figure 1G and H ).

To confirm whether Gas6 alleviated the endothelial barrier by regulating TJ proteins, we examined the expression of ZO-1, occludin, and claudin5 by Western blotting. After treatment with different concentrations of LPS for 12 h, the expression levels of ZO-1 and occludin were markedly decreased in a dose-dependent manner. Occludin was significantly decreased when the LPS concentration was 200 µg/L, while the expression of ZO-1 was reduced at all concentrations. Interestingly, claudin5 was increased after treatment with LPS ( Figure 2A and C ).

Based on the above results, LPS at a concentration of 200 μg/L was used for the following studies. Then, MAECs were incubated with different concentrations of Gas6 for 2 h before LPS stimulation. As shown in Figure 2B and D , we found that ZO-1 and occludin protein levels were completely improved at all concentrations of Gas6, and there was no difference among the three concentrations. Notably, the expression of claudin5 was also obviously increased after treatment with Gas6. Therefore, we used LPS at 200 μg/L and Gsa6 at 400 ng/mL in the following experiments. Additionally, Gas6 did not change the TJ protein expression level when incubated alone. To clarify further the mechanism of TJ injury, immunofluorescence shows that the location and distribution of ZO-1 and occludin in the cell–cell junction areas were disrupted under LPS (200 μg/L) stimulation compared with the control, while Gas6 (400 ng/mL) attenuated the disruption obviously. Meanwhile, claudin5 was restored as well after treatment with Gas6 (400 ng/mL) ( Figure 3A–C ).

To determine the influence of Gas6 on NF-κB activation induced by LPS, we performed Western blotting of NF-κB. First, the phosphorylation of NF-κB p65 (p-NF-κB p65) and total NF-κB p65 was examined after different concentrations of LPS stimulation for 30 min. As shown in Figure 4A and C , LPS increased the phosphorylation of NF-κB p65 in a dose-dependent manner. As shown in Figure 4B and D , phosphorylation of NF-κB p65 was abrogated when cells were pretreated with Gas6 (100 ng/mL, 200 ng/mL, and 400 ng/mL) for 2 h prior to LPS exposure and co-cultured for 30 min. In addition, treatment with Gas6 alone did not affect the NF-κB activity.

The expression of TAM receptors (Tyro3, Axl, and Mertk) in MAECs was determined by Western blotting. Cardiac muscle cell H9C2 was used as a positive control. We found that MAECs expressed all three receptors simultaneously ( Figure 5A and C ). However, only Axl was activated when MAECs were incubated with Gas6. We detected obvious Axl phosphorylation within 15 min ( Figure 5B and D ). There was no significant change in the other two receptors (data not shown). To further confirm whether Gas6 attenuated LPS-induced vascular endothelial hyperpermeability via Axl/NF-κB signaling pathways in vitro, we transfected endothelial cells with siAxl or siNC molecules. The transfection efficiency was examined by Western blotting. Compared with cells transfected with siNC, cells transfected with siAxl showed a 39.5% reduction in Axl expression ( Figure 5E and F ). The reduction in Axl expression significantly inhibited the protective effect of Gas6 in permeability ( Figure 6A and B ) when compared to that of the siNC group. Real-time PCR was also used to examine the Axl siRNA transfection efficiency after MAECs transfected with siAxl (Supplementary Figure 1). Furthermore, Gas6 significantly increased ZO-1, occludin, and claudin5 expression, which was inhibited when Axl expression was downregulated using siAxl compared with that of the siNC group (Figure 6C and D ). As shown in Figure 7A and B , the breakdown of ZO-1 and occludin was markedly restored after Gas6 (400 ng/mL) treatment compared to the LPS group (200 μg/L), and the protection of ZO-1 and occludin was decreased when transfected with siAxl. Meanwhile, the fluorescence density of claudin5 was increased after treatment with Gas6, while it was decreased when transfected with siAxl ( Figure 7C ). In addition, knocking down Axl markedly reduced Gas6’s inhibitive effect on phosphorylation levels of NF-κB stimulated by LPS ( Figure 8A and B ). To further confirm the NF-κB activation induced by LPS, we detected NF-κB p65 translocation in HUVECs by immunofluorescence. As shown in Figure 8C , LPS obviously induced the nuclear localization of NF-κB p65, while Gas6 strongly mitigated the NF-κB p65 nuclear transfer induced by LPS, and the effect was attenuated with siAxl transfected.