Moesin Is a Novel Biomarker of Endothelial Injury in Sepsis

Objective Increased vascular permeability and inflammation are principal hallmark of sepsis. Moesin (MSN) is a membrane-associated cytoskeleton protein and crucial for the vascular endothelial function. This study is aimed at evaluating the role of MSN in endothelial injury during the process of sepsis. Methods Serum MSN in septic patients was measured by ELISA. BALB/c mice were injected with different doses of lipopolysaccharide (LPS) or underwent cecal ligation and single or double puncture (CLP) to mimic sublethal and lethal sepsis. After treatment, their serum MSN and PCT levels, wet to dry lung weights (W/D ratio), bronchoalveolar lavage fluid (BALF) protein concentrations, and lung injury scores were measured. The impact of MSN silencing on LPS-altered Rock1/myosin light chain (MLC), NF-κB, and inflammatory factors in human microvascular endothelial cells (HMECs), as well as monolayer HMEC permeability, was tested in vitro. Results Compared with healthy controls, serum MSN increased in septic patients and was positively correlated with SOFA scores and serum PCT levels in septic patients. LPS injection significantly increased serum the MSN and PCT expression, BALF protein levels, and W/D ratio, and the serum MSN levels were positively correlated with serum PCT, lung W/D ratio, and lung injury scores in mice. Similar results were obtained in the way of CLP modelling. LPS enhanced MSN, MLC, NF-κB phosphorylation, increased Rock1 expression, and inflammatory factors release in the cultured HMECs, while MSN silencing significantly mitigated the LPS-induced Rock1 and inflammatory factor expression, NF-κB, and MLC phosphorylation as well as the monolayer hyperpermeability in HMECs. Conclusions Increased serum MSN contributes to the sepsis-related endothelium damages by activating the Rock1/MLC and NF-κB signaling and may be a potential biomarker for evaluating the severity of sepsis.


Introduction
Sepsis is a leading cause of high morbidity and mortality worldwide, and it can lead to dysregulated host responses to infection, even acute multiple organ failure [1]. Currently, treatment for sepsis is very limited, and the high morbidity and mortality rates of sepsis result in high socioeconomic burdens worldwide [2].
Increased vascular permeability is a principal hallmark of sepsis. The endothelial integrity is crucial for the physiological organ function. During the process of infection and sepsis, endothelial cells can initiate and propagate the inflammatory response, and systemic activation of the endothelium will disrupt the endothelium barrier [3,4]. Such wide spreading vascular dysfunction finally leads to multiple organ failure and death. However, there is no reliable measure biomarker to evaluate the endothelium injury. Therefore, identification of biomarkers for evaluating endothelial activation and injury will be of significance in early management of septic patients.
Moesin, a member of the ezrin-radixin-moesin family, functions to link the plasma membranes to actin-based cytoskeleton and is crucial for the vascular endothelial function. Although three ERM proteins share high amino acid similarity, their expression is organ-and tissue-specific [2]. MSN is predominantly expressed in vascular endothelial cells. Previous studies have found that some factors, such as TNF-α, thrombin, and advanced glycation end products, can activate vascular endothelial cells and phosphorylate MSN to increase endothelial cell permeability [5][6][7]. Furthermore, lipopolysaccharide (LPS) can stimulate endothelial cells to secrete MSN [8]. In addition, MSN is required for the high-mobility group box-(HMGB-) induced endothelial cell hyperpermeability and inflammatory responses, and high levels of blood MSN are detected in septic mice and patients [9]. Hence, MSN participates in the pathogenesis of sepsis. However, little is known on whether MSN can be a useful biomarker for diagnosis and prognosis of sepsis.
In this study, we tested the hypothesis that MSN might be a potential biomarker for evaluating the severity of endothelium damages during the process of sepsis. We examined the MSN expression in septic mice and patients and explored its function in human microvascular endothelial cells (HMECs) in vitro.
end of the experiment, their lungs were harvested. The right lungs of 4 mice were immediately frozen in liquid nitrogen and stored at −80°C, while their left lungs were prepared for histological analysis. Other mice were subjected to bronchoalveolar lavage with 3 ml PBS (1 ml, 3 times) through a tracheal cannula, and the collected BALF samples were centrifuged. The BALF supernatants were stored at -80°C, and the pelleted cells were used for further analysis. The protein concentrations in individual BALF samples were determined by the bicinchoninic acid (BCA) assay using a BCA kit (Key-Gen Biotech, Nanjing, China), according to the manufacturer's instruction.
Sublethal sepsis was induced by cecal ligation and puncture (CLP), single puncture. Cecal ligation and double puncture (2CLP) were used to induce lethal sepsis. Mice underwent CLP or 2CLP with an 18-gauge needle as previously described [12,13]. Sham-operated animals served as controls. There were 20 mice in each group.

Histopathology and Immunohistochemical
Staining. The collected lung tissues were fixed with 4% paraformaldehyde and paraffin-embedded. The lung tissue sections (4 μm) were routine-stained with hematoxylin-eosin (H&E). The degrees of lung injury were scored in a blinded manner as previously described [14]. To further evaluate the expression of MSN, lung tissues were used to perform immunohistochemistry. Sections were incubated with a primary antibody to MSN (1 : 100 dilution; Abcam, USA) for 1 h at room temperature and then incubated with biotinylated secondary antibodies. Horseradish peroxidase-conjugated IHC Detection Reagent was used to detect antibody-antigen complexes (Cell 2.8. Coimmunoprecipitation. HMECs were lysed with modified RIPA buffer containing protease and phosphatase inhibitor cocktail for 30 min. Cell lysates were collected and centrifuged at 10,000 g for 10 min at 4°C. The supernatants were collected, and immune complexes were captured by protein G beads slurry, according to the manufacturer's instructions. Immunoprecipitation was performed with antibodies to Rock1. The beads were eluted with RIPA buffer and boiled in 2× SDS protein loading buffer for 7 min. Then, samples were analyzed by Western blot. 2.9. Quantitative Real-Time PCR. Total RNA was extracted from HMECs using the RNAiso Plus kit (Takara, Kusatsu, Shiga, Japan), and RNA samples (1 μg each) were reversely transcribed into cDNA using the M-MLV First Strand kit (Life Technologies, Gaithersburg, MD, USA). The relative levels of MSN to control β-actin mRNA transcripts were quantified by qRT-PCR using ×2 SYBR-Green qPCR Super-Mix (High ROX) (Bimake, Houston, USA) and specific primers (Sanon Biotech, Shanghai, China) in the ABI Ste-pOne Plus RealTime PCR system (Applied Biosystems). The PCR reactions were performed in triplicate. The sequences of primers for MSN were forward 5 ′ -GATGCT GTCCTGGAATATCTGA-3′ and reverse 5′-TCTGCTCAT AGATGTTGAGACC-3′. Sequences of primers for GAPDH were forward 5′-UGACCUCAACUACAUGGUUTT-3′ and reverse 5 ′ -AACCAUGUAGUUGAGGUCATT-3 ′ . The data were analyzed by the 2 -ΔΔCt method.

Measurement of Endothelial
Permeability. The impact of the MSN expression on the permeability of monolayer HMECs was determined by a dextran penetration assay. Briefly, the different groups of HMECs (5 × 10 4 cells/well) were cultured in the inserted chambers (4 μm pore size and 12 mm diameter) in 12-well transwell plates for 2 days, and when the cells reached confluence, the cells in individual wells were treated in triplicate with LPS (0, 1, 2, and 4 μg/ml) for 24 h. After being washed with PBS, the cells in the inserted chambers were cultured in the bottom chamber with fresh medium and treated with 0.5 mg/ml fluorescein isothiocyanate-(FITC-) labeled dextran (40 Kd, Sigma-Aldich) at 37°C for 30 minutes. The fluorescent signals in the supernatants of the bottom chambers were determined by spectrophotometry at (Ex (λ) 490 nm; Em (λ) 520 nm). The data in the experimental groups were normalized to that of the control group.
2.11. Statistical Analysis. Data are expressed as mean ± SD or case numbers. The difference between groups was analyzed by the independent t-test or Mann-Whitney U test. The difference among groups was analyzed by one-way ANOVA, posthoc Kruskal-Wallis test (non-normal distributed parameters), and Bonferroni correction (normally distributed parameters). The relationship between two variables was analyzed by linear or logistic regression correlation analysis. All statistical analyses were performed using the SPSS version 23 and GraphPad Prism 7 Software. A p value of <0.05 was considered statistically significant.  Figure 1(a)). Furthermore, the levels of serum MSN in the septic patients were also significantly higher than that in the healthy subjects (p = 0:0278) (Figure 1(a)). Correlation analysis revealed that the levels of serum MSN in the septic patients were positively correlated with SOFA scores (r = 0:4918,p = 0:0008, Figure 1(b)) and serum PCT levels (r = 0:3455, p = 0:0232, Figure 1(c)). As expected, the levels of serum PCT in all patients were positively correlated with SOFA scores (r = 0:4807, p = 0:0010, Figure 1(d)). The results indicated that increased MSN levels were positively correlated with the severity of sepsis in this population.

Increased Serum MSN Levels Are Correlated with
Elevated Lung Permeability in a Mouse Model of LPS-Induced Septic Lung Injury. To test whether increased MSN levels could contribute to endothelium damages, a mouse model of septic lung injury was established by injecting different doses of LPS. H&E staining revealed that LPS injection significantly impaired the alveolar microstructure, enhanced inflammatory cell infiltration and lung damages, and increased lung injury scores, particularly in the mice with a higher dose LPS group (Figures 2(a) and 2(b)). There were also obvious edema, congestion, and high W/D ratios in the harvested lung tissues of the septic mice (Figures 2(c) and 2(d)). Then, we further tested the expression of phosphorylated MSN and total MSN in lung tissues by Western blotting, the results show that LPS injection can significantly increase the MSN expression in lung tissues (Figure 2(e)), and the semiquantification for WB was shown in (Figure 2(f)). We also found that LPS injection elevated the serum MSN and PCT levels, particularly in the higher dose LPS group (Figures 2(g) and 2(h)). As expected, injection with LPS also significantly increased the levels of BALF proteins, related to that in the control mice (p < 0:01,for both Figure 2(i)). To explore the relationship between MSN and sepsis correlation analysis between MSN and lung injury scores, lung W/D ratios and PCT were performed. Interestingly, the levels of serum MSN were positively correlated with PCT (r = 0:7826, p = 0:0127, Figure 2(j)), the lung injury scores (r = 0:7113, p = 0:0377, Figure 2(k)), and the lung W/D ratios (r = 0:8206, p = 0:0067, Figure 2(l)) in septic animals. Such findings indicated that increased serum MSN levels were indicatives of elevated lung permeability and vascular endothelium damage in septic mice.

Increased Serum MSN Levels Are Correlated with Lung
Injury Score and W/D Ratio in a CLP-Induced Septic Mouse Model. To further confirm the relationship between MSN and sepsis, we used sublethal and lethal CLP to construct the septic mouse model again. As expected, CLP surgery significantly aggravated the lung pathological damage, particularly in the lethal CLP group (Figure 3(a)). Correspondingly, the lethal CLP group has a high lung injury score than the sublethal CLP group (Figure 3(b), p < 0:05). Similarly, we found that the serum MSN increased significantly after CLP surgery, while this increased trend was more obvious in the lethal CLP group (Figure 3(c)). CLP surgery also significantly increased the W/D ratios related to that in the    Journal of Immunology Research control mice (Figure 3(d), p < 0:05). Then, we performed the correlation analysis between MSN and lung injury score and lung W/D ratios. We found that the lung injury scores in CLP mice were positively correlated with serum MSN levels (r = 0:6035, p = 0:0377, Figure 3(e)) and also have a positive correlation with lung W/D ratios (r = 0:6137, p = 0:0125, Figure 3(f)).

LPS Upregulates the MSN Expression in HMECs in a
Time-and Dose-Dependent Manner. To understand the molecular mechanisms underlying the potential effect of LPS on human endothelial cells, we further tested the effect of LPS on the MSN expression and relevant signaling in vitro. We found that treatment with LPS increased the levels of MSN in the supernatants of cultured HMECs in a dose-and time-dependent manner (Figures 4(a) and 4(d)). Western blot analysis indicated that treatment with LPS did not significantly change the relative levels of the total MSN expression, but significantly enhanced MSN phosphorylation in HMECs, particularly with a higher dose of LPS at 24 h posttreatment (Figures 4(b) and 4(e)), and the semiquantification for WB was shown in (Figures 4(c) and 4(f)).  Journal of Immunology Research Collectively, LPS treatment increased MSN phosphorylation in HMECs in vitro. We next explore the secretion manner of MSN. After treatment with 3-MA, DMA, and BFA, respectively, MSN secretion was slightly inhibited by 3-MA and DMA, but significantly inhibited by BFA (Figures 4(h), 4(i), and 4(g)), which suggesting that MSN can be mainly secreted via the Golgi apparatus transport pathway.

MSN Silencing Mitigates the LPS-Increased HMEC
Monolayer Permeability In Vitro. To determine the role of MSN in the LPS-mediated barrier disruption, we generated MSN silencing HMECs by transfection with MSN-specific siRNA. We found that transfection with MSN-specific siRNA significantly reduced the levels of MSN mRNA transcripts and protein expression, demonstrating the efficacy of MSN silencing in HMECs (Figures 5(a) and 5(b)), and the semiquantification for WB was shown in (Figure 5(c)). Among the three siRNAs, siRNA-3 and siRNA-1 showed better knock down effect and were used in the follow-up experiments. While treatment with different doses of LPS significantly increased the amounts of penetrated FITCdextran in the control siRNA-transfected HMECs in a dosedependent manner (p < 0:001 for all), treatment with the same doses of LPS only slightly increased the levels of penetrated FITC-dextran in the MSN silencing HMECs (p < 0:05 for all, Figure 5(d)). Protein interaction analysis and coimmunoprecipitation were used to explore the possible pathway involved in MSN-mediated endothelial barrier breakdown. Protein interaction analysis showed that MSN was closely related to the rhoA/Rock pathway ( Figure 5(e)). The coimmunoprecipitation assay was performed, and endogenous interaction of p-MSN-ROCK1 complex was confirmed in HMECs ( Figure 5(f)). Western blot analyses revealed that in comparison with that    Journal of Immunology Research in the control HMECs, MSN silencing not only decreased MSN expression but also significantly mitigated or abrogated the LPS-enhanced MLC phosphorylation and Rock1 expression in HMECs ( Figure 5(g)), and the semiquantification for WB was shown in (Figure 5(h)). Such findings indicate that LPS damages the integrity of endothelium, dependent on MSN phosphorylation.

MSN Silencing Attenuates LPS-Induced Inflammatory
Cytokine Production. To further explore the proinflammatory effects of MSN in sepsis, we next performed ELISA to measure the effects of MSN silencing on the production of the inflammatory cytokines TNF-α, IL-6, IL-1β, and IL-18 in medium supernatant of HMECs. We also detected the expression of the inflammation-related signaling pathway NF-κB by Western blotting. MSN silencing reduced the levels of TNF-α, IL-6, IL-18, and IL-1β in the medium supernatant of HMECs (p < 0:05, p < 0:01, ns, p < 0:01, respectively) ( Figures 6(a)-6(d)). In addition, MSN silencing also decreased the expression of phosphorylated p65 increased by LPS (Figure 6(e)), and the semiquantification for WB was shown in (Figure 6(f)).

Discussion
Sepsis is a life-threatening organ dysfunction caused by the dysregulated host response to infection, sepsis-3 definition of sepsis with the clinical criteria based on the SOFA score. Septic shock is a special form of sepsis, which refers to a state of acute circulatory failure [15]. Rapid diagnosis is crucial for effective treatment of septic patients. However, there is still no ideal biomarker for diagnosis of sepsis. Although microbiological examination remains as the gold standard for the infection diagnosis, it requires 24 to 72 hours or longer to complete [16]. Serum PCT and C-reactive protein (CRP) levels are widely used as specific biomarkers for diagnosis of bacterial infection-related inflammation, but they are limited in their ability to distinguish sepsis from other inflammatory conditions [17]. The sepsis 3.0 criteria confirm the advantage of the SOFA score in diagnosis and prediction of sepsis. In this study, we observed that increased serum MSN and PCT levels in septic patients were positively correlated with SOFA scores, and both of them have a similar correlation coefficient to SOFA scores. Furthermore, increased serum MSN levels were also correlated positively with W/D ratios and lung pathological damage scores in septic mice. Such findings suggest that serum MSN levels may be a potential biomarker for evaluating the severity of sepsis. MSN is a protein linker between cellular membranes and the actin-based cytoskeleton in cells. MSN is always maintained in a resting stage, in which its N-terminal and Cterminal domains connect each other to form a closed conformation, leading to its functional sites hidden inside the molecule. MSN phosphorylation changes its conformation into an active conformation by separating its N-terminal and C-terminal domains to expose its functional sites that attach to the cell membrane and cytoskeleton, respectively. The phosphorylated MSN is closely associated with cytoskeletal rearrangement [18][19][20]. MSN is highly expressed in endothelial cells, including pulmonary vascular and lymphocytes [21]. Both of high vascular endothelium permeability and aberrant macrophage activation are critical events in the pathogenesis of sepsis [22,23]. We found that LPS increased MSN levels in the supernatants of cultured HMECs, MSN phosphorylation, and the monolayer HMEC permeability. Conversely, MSN silencing by transfection with a specific siRNA mitigated or abrogated the LPS-increased monolayer HMEC permeability in vitro. Our data support the notion that phosphorylated MSN is crucial for high endothelium permeability. Given that MSN can interact with a variety of effector molecules through different mechanisms, increasing endothelium permeability [24], MSN may    Journal of Immunology Research contribute to the pathogenesis of sepsis and be a potential target for design of therapies for preserving the endothelium integrity. Therefore, our findings may provide new insights into the pathogenesis of sepsis. It is well known that MSN is important for the innate immune response. A previous study has shown that MSN may act as a receptor of LPS on monocytes [25]. LPS can bind to the carboxyl-terminus of MSN to promote the MSN expression and phosphorylation. The activated MSN is associated with TLR4 and CD14, leading to the LPS-related signaling and proinflammatory cytokine production [26]. Hence, MSN is crucial for the maintenance of vascular endothelial barrier and innate immune cell activation. Consistently, we found that (1) LPS increased MSN levels in the   Figure 6: MSN silencing reversed LPS-induced inflammatory cytokine production and p65 phosphorylation in HMECs. HMECs were, respectively, transfected with siNC and siMSN RNA for 48 h, then exposed to LPS for 24 h, and the supernatants were collected for the ELISA test. MSN silencing reduced the production of TNF-α (a, p < 0:05), IL-6 (b, p < 0:01), IL-18 (c, ns), and IL-1β (d, p < 0:01) in the medium supernatant of HMECs. In addition, MSN silencing also decreased the expression of phosphorylated p65 increased by LPS (e), and the semiquantification for WB was shown in (f).