Recombinant thrombomodulin protects against LPS‐induced acute respiratory distress syndrome via preservation of pulmonary endothelial glycocalyx

Background and Purpose Disruption of the endothelial glycocalyx is causally related to microvascular endothelial dysfunction, a characteristic of sepsis‐induced acute respiratory distress syndrome (ARDS). Recombinant human thrombomodulin (rhTM) attenuates vascular endothelial injuries, but the underlying mechanism remains elusive. Here, we investigated the structural basis and molecular mechanisms of rhTM effects on vascular endothelial injury in a model of sepsis. Experimental Approach LPS (20 mg·kg−1) was intraperitoneally injected into 10‐week‐old male C57BL6 mice, and saline or rhTM was intraperitoneally injected 3 and 24 h after LPS injection. Using serum and/or lung tissue, histological, ultrastructural, and microarray analyses were performed. Key Results Survival rate of rhTM‐treated mice was significantly higher than that of control mice 48 h after LPS injection. Serum concentrations of IL‐6 and high‐mobility group box 1 were lower in the rhTM‐treated group than in the control. Injury to the endothelial glycocalyx in pulmonary capillaries was attenuated by rhTM treatment. Gene set enrichment analysis revealed up‐regulation of gene sets corresponding to cell proliferation/differentiation and anti‐inflammation, such as the TGF‐β pathway, and negative regulation of IL‐6, upon rhTM treatment. Gene expression of heparan sulfate 6‐O‐sulfotransferase 1 and endothelial cell‐specific molecule 1 (components of the endothelial glycocalyx) was significantly preserved by rhTM treatment, and their protein expression levels were maintained in endothelial cells. Conclusion and Implications Our findings show that rhTM treatment affected inflammation, cell proliferation/differentiation, and glycocalyx synthesis in serum and lung tissue, subsequently attenuating ARDS caused by endothelial injury.

Conclusion and Implications: Our findings show that rhTM treatment affected inflammation, cell proliferation/differentiation, and glycocalyx synthesis in serum and lung tissue, subsequently attenuating ARDS caused by endothelial injury.

| INTRODUCTION
Acute respiratory distress syndrome (ARDS), a clinical phenotype of sepsis, is characterised by a serious inflammatory development comprising neutrophil integration and pro-inflammatory cytokine production, thereby causing injuries to the alveolar-capillary integrity with high-permeability and non-hydrostatic pulmonary oedema (Shao, Tang, Liu, & Zhu, 2016). Under normal conditions, the vascular endothelial glycocalyx covers the surface of endothelial cells and plays a key role in microvascular and endothelial physiology (Chelazzi, Villa, Mancinelli, De Gaudio, & Adembri, 2015). The glycocalyx is an important determinant of vascular permeability (Henry & Duling, 1999;Vink & Duling, 2000). It also regulates neutrophil adhesion and lung injury during sepsis-induced ARDS (Schmidt et al., 2012), and its disruption by LPS is causally related to microvascular endothelial dysfunction (Inagawa et al., 2018).
Thrombomodulin, which exists on the surface of endothelial cells and contributes to the maintenance of vascular homeostasis (Esmon, 2005), demonstrates an anti-inflammatory effect by binding to high-mobility group box 1 (HMGB1), a nuclear architectural chromatin-binding protein with roles in DNA organisation and transcription regulation (Kadono et al., 2017). Both DNA organisation and transcription regulation play a crucial role in ARDS progression by promoting lung injury (Abeyama et al., 2005;Ito et al., 2008;Ito & Maruyama, 2011). Soluble recombinant human thrombomodulin (rhTM), consisting of all the extracellular domains of thrombomodulin, has been used to treat patients with disseminated intravascular coagulation (DIC) (Kadono et al., 2017). rhTM binds to thrombin to inhibit its procoagulant activity, promotes protein C activation (Kearon et al., 2005;Khorchidi et al., 2002;Tawara, Sakai, & Matsuzaki, 2016), and reduces the secretion of inflammatory cytokines, including IL-6 and TNF-α, under septic conditions .
However, to our knowledge, no study has yet investigated whether rhTM attenuates pulmonary endothelial glycocalyx injuries under endotoxaemic conditions. Therefore, this study aimed to evaluate the condition of the pulmonary endothelial glycocalyx after LPS administration in rhTM-treated mice. Animal studies are reported in compliance with the ARRIVE guidelines (Kilkenny, Browne, Cuthill, Emerson, & Altman, 2010) and with the recommendations made by the British Journal of Pharmacology. All efforts were made to minimise animals' suffering and reduce the number of animals used. What is the clinical significance • Recombinant human thrombomodulin could exert beneficial effects against acute respiratory distress syndrome.
anaesthesia (using a mixture of medetomidine hydrochloride 0.3 mgÁkg −1 , midazolam 4 mgÁkg −1 , and butorphanol tartrate 5 mgÁkg −1 , given i.p.). The anaesthesia was deemed sufficient if the corneal and hind-paw withdrawal reflexes were absent. Before lung specimens were obtained, the mice were killed by exsanguination from the ophthalmic artery until the righting reflex was lost.

| Serum preparation and ELISA
Blood samples were collected from six mice, from the buccal artery, which branches off the maxillary artery, allowed to clot at 25 C for 2 h, and centrifuged (2,000 x g, 4 C for 20 min). The supernatant (serum) was then collected. Serum IL-6 was measured using ELISA quantitation kits for mouse IL-6 (Cat No. M6000B, R&D Systems).

| Histopathological examination
Whole right lobes from the lungs of six individual mice 48 h after LPS administration were fixed with PBS containing 10% formalin and embedded in paraffin. Paraffin sections (4 μm) were then deparaffinised and rehydrated. Finally, slides were counterstained with haematoxylin and eosin, and the coverslipped lung sections were scored as follows for pulmonary oedema: 1 = absent; 2 = detectable seroproteinaceous fluid in one to a few alveoli; or 3 = seroproteinaceous fluid-filled alveoli in a multifocal to coalescing pattern in the lungs. Neutrophilic infiltration was scored: 1 = absent or rare solitary neutrophils; 2 = detectable extravasated neutrophils observed as small loose cellular aggregates in one or a few airways and/or alveoli; 3 = detectable extravasated neutrophils observed as loose to compact cellular aggregates in multiple to coalescing airway and/or alveoli with some effacement of lung architecture; 4 = detectable extravasated neutrophils observed as compact cellular aggregates effacing most adjacent lung architecture (Langlois et al., 2010). These experiments were performed in a blinded manner to avoid bias.

| Western blotting
Total protein concentration in tissue lysates was measured using bicinchoninic acid protein assays, and 10 μg of protein was separated

| Electron microscopy
Electron microscopic analysis of the endothelial glycocalyx was performed as described previously (Inagawa et al., 2018;Okada et al., 2017). Briefly, mice were anaesthetised and then perfused with a solution composed of 2% glutaraldehyde, 2% sucrose, 0.  cDNA was used as a template for quantitative real-time (qRT)-PCR, which was performed using TB Green Premix Ex Taq II ( Takara Bio) according to the manufacturer's protocol, on a Thermal Cycler Dice TP 990 machine (Takara Bio). The PCR reaction conditions were 50 C for 2 min, 95 C for 10 min, followed by 40 cycles of 95 C for 15 s, plus 60 C for 1 min. The relative quantification of each transcript (HS6ST1, ESM 1, and HPSE) was determined by setting the threshold cycle (Ct) for each sample to reflect the cycle number at which the fluorescence generated within the reaction crossed the threshold level chosen as a point when the amplification was in an exponential phase.

| Microarray analysis
GAPDH was used as the loading control. The function 2ΔCt was used to determine relative abundance differences, where ΔCt was the difference in Ct values between the compared samples. Primers used in the various PCR reactions are provided in the Table S1.

| Data and statistical analysis
Data are presented as means ± SEM. Student's two-tailed t test was used for comparing the two groups, and survival data were analysed using the log-rank test; P < 0.05 was considered significant. All calculations were performed using GraphPad Prism (Ver. 7.02; La Jolla, CA).
The data and statistical analysis comply with the recommendations of the British Journal of Pharmacology on experimental design and analysis in pharmacology.

| Materials
Recombinant human thrombomodulin was provided by Asahi Kasei Pharma Corporation.

| Nomenclature of targets and ligands
Key protein targets and ligands in this article are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMA-COLOGY (Harding et al., 2018), and are permanently archived in the  (Figure 1a).
In control mice receiving LPS only, IL-6 levels were significantly raised at 6 h after LPS injection and peaked at 12 h after injection.
Thereafter, IL-6 levels returned to baseline within 48 h after LPS injection ( Figure 1b). In rhTM-treated mice, IL-6 concentrations were significantly decreased at 6 and 12 h after LPS administration compared with those in control mice. Moreover, serum HMGB1 levels were decreased in rhTM-treated mice compared with that in untreated mice 48 h after LPS administration (Figure 1c).
To assess pulmonary injury 48 h after LPS injection, we used the previously reported clinical scoring system (Figure 1d

| Pulmonary endothelial glycocalyx injury was attenuated in rhTM-treated mice
Syndecan-1, a component of the glycocalyx, was found to be degraded after LPS administration. However, syndecan-1 degradation was markedly attenuated in rhTM-treated mice compared with that in saline-treated mice (Figure 2a,b). To quantitatively assess endothelial glycocalyx injury, we performed intensity measurements using WGA staining, which enabled visualisation of endothelial glycocalyx (Kataoka et al., 2016).
WGA intensity was reduced in saline-injected mice after LPS administration compared with that in sham mice (Figure 2c,d). After LPS injection, WGA intensity was increased in rhTM-treated mice compared with that in saline-treated mice. These results together suggest that endothelial glycocalyx injury in lung capillaries was attenuated by rhTM treatment.

| rhTM maintained endothelial glycocalyx structure under endotoxaemic conditions
The ultrastructure of the endothelium and endothelial glycocalyx was analysed using electron microscopy. Conventional scanning electron microscopy results showed the pulmonary capillaries to be of the con- Conversely, in the rhTM-treated group, the continuous structure of the glycocalyx was maintained, and endothelial glycocalyx injury was attenuated (Figure 4h,i,k,l). These results indicate that rhTM treatment attenuated endothelial glycocalyx injury under endotoxaemia.

| GSEA following rhTM treatment in the lungs
Because rhTM treatment was found to significantly improve survival rate following LPS injection, we next sought to comprehensively determine which genes were affected by rhTM treatment. To significantly up-regulated in the rhTM-treated group compared with those in the control group (P < 0.01). Conversely, the expression of genes associated with glycosaminoglycan degradation was not found to differ significantly between the rhTM-treated and control mice ( Figure S2). Additionally, the qRT-PCR results revealed no significant differences in the expression of HPSE (heparinase), an enzyme that degrades polymeric heparan sulfate molecules, between rhTM-treated and saline-injected mice ( Figure S3).
Together, these results suggest that rhTM treatment may functionally affect anti-inflammatory, cell proliferation or differentiation and glycocalyx synthesis pathways. 3.5 | Ki67, HS6ST1, and ESM 1 expression in pulmonary capillaries Next, to determine the cell proliferation and differentiation capacity, as well as the cellular distribution of HS6ST1 and ESM 1 proteins, immunohistochemical analysis was performed. Specifically, Ki67 expression, a cell proliferation and differentiation marker, was found to be increased in rhTM-treated mice, compared with the sham and saline-treated mice (Figure 6a,b). Moreover, HS6ST1 expression was observed in several cell types including inflammatory and endothelial cells. However, its expression in endothelial cells was particularly increased in rhTM-treated mice compared with sham and saline-treated mice (Figure 6c,d). To provide further confirmation, double immunostaining for HS6ST1 and CD31, an endothelial cell marker, was performed and HS6ST1 and CD31 were found to be co-localised in rhTM-treated mice (Figure 6e).
Further, ESM 1 was also expressed in endothelial cells of all experimental groups and in several cell types including inflammatory cells.
However, the highest number of ESM 1-positive endothelial cells was observed in rhTM-treated mice (Figure 7a,b); meanwhile, both ESM 1 and CD31 were found to be co-localised in lungs of mice treated with rhTM ( Figure 7c). Taken together, these results suggest that rhTM treatment may increase cell proliferation and differentiation and that glycocalyx synthesis may be promoted by HS6ST1 and ESM 1.

| rhTM accelerates glycocalyx synthesis
We found that rhTM treatment up-regulated the gene set associated with placental blood vessel development, including the heparan sulfate 6-O-sulfotransferase 1 (HS6ST1) gene. This gene is a member of the heparan sulfate biosynthetic enzyme family (Habuchi, Habuchi, & Kimata, 2004) and, as such, is essential for heparan sulfate synthesis, one of the primary components of the endothelial glycocalyx (Reitsma et al., 2007). As HS6ST1 deficiency is lethal in mice primarily during the later embryonic stages, resulting in abnormal angiogenesis in the labyrinthine zone of the placenta along with aberrant lung morphology similar to that observed in pulmonary emphysema (Habuchi et al., 2007), HS6ST1 is considered to be essential for angiogenesis also in the pulmonary circulation. Furthermore, cell surface heparan sulfate proteoglycans interact with a myriad of growth factors, receptors and extracellular matrix proteins, resulting in regulation of many receptor-ligand interactions (Habuchi et al., 2004). Thus, heparan sulfate has important functions in a variety of developmental, morphogenetic and pathogenic processes (Habuchi et al., 2007).
Heparan sulfate biosynthetic enzymes play important roles in producing numerous distinct heparan sulfate fine structures that perform multiple biological activities. Heparan sulfate is also a component of the endothelial glycocalyx and has been recently reported to be associated with endothelial glycocalyx regeneration (Mensah et al., 2017), Our current study also indicated that ESM 1 expression was increased in rhTM-treated mice. ESM 1 is a dermatan sulfate proteoglycan that is principally expressed in pulmonary microcirculation as one of the glycocalyx components (Bechard et al., 2000;Lassalle et al., 1996;Orbegozo et al., 2017;Zhang et al., 2012). ESM 1 expression increases in the presence of pro-angiogenic growth factors, such as FGF-2 or VEGF (Aitkenhead et al., 2002;Rennel et al., 2007). It also contributes to the pathophysiology of ARDS (Bull,

| rhTM inhibits glycocalyx injuries through its anti-inflammatory effects
The pulmonary endothelial glycocalyx becomes injured by inflammatory conditions such as sepsis (Inagawa et al., 2018). Therefore, inhibition of inflammation may protect the endothelial glycocalyx structure (Fukuta et al., 2019;Suzuki et al., 2019). Moreover, excessive secretion of IL-6 causes cellular injury during sepsis (Tanaka, Narazaki, & Kishimoto, 2014). The present study revealed that serum IL-6 concentration was reduced by rhTM treatment, consistent with the gene set analysis results of negative regulation of IL-6 production by rhTM.
Hence, this phenomenon may be involved in decreasing vascular permeability via glycocalyx protection induced by rhTM treatment.
Moreover, the current study showed that rhTM can inhibit serum HMGB1 concentration during endotoxaemia. Because HMGB1 induces the secretion of IL-1β, TNF-α, and IL-6 from several types of inflammatory cells, including neutrophils and macrophages, it plays an important role in initiating and maintaining the amplification of the inflammatory cascade (Fu et al., 2016;Takahashi et al., 2016), consistent with previous reports that suggested that the N-terminus of thrombomodulin binds to and decomposes HMGB1, subsequently exerting an anti-inflammatory effect (Abeyama et al., 2005;Ito et al., 2008;Ito & Maruyama, 2011).

| Study limitations
Sepsis is an extremely complex condition in humans compared with simple endotoxaemia in an experimental model. Because our focus in this study was to investigate the direct relationship between endothelial glycocalyx injury and septic vasculitis, we used an endotoxaemia model that does not reflect certain typical septic conditions such as bacterial infection. Thus this would represent a limitation of our current study. In this study, rhTM was administered at a high dose of 30 mgÁkg −1 , because the rhTM used was the human, rather than murine, recombinant form. In addition, the route of administration was i.p. and not i.v., as would be typical in a clinical setting.
Although it was shown here that rhTM up-regulates genes involved in glycocalyx synthesis in the endothelium, which translates to improved glycocalyx coverage and overall survival in LPS-induced ARDS, further confirmatory studies, including the use of knockout mice, are necessary to fully elucidate the precise mechanisms that allow rhTM to promote endothelial glycocalyx synthesis under conditions of sepsis.
In conclusion, our data has shown that rhTM treatment protected the endothelial glycocalyx from endotoxaemia-induced lung injury in mice. This mechanism may reduce the damage associated with inflammation and acceleration of the biosynthesis of the glycocalyx itself. As rhTM is currently used in clinical applications, it may be considered as a novel strategy against septic vasculitis, through its ability to protect the endothelial glycocalyx.