ST2 blockade mitigates peritoneal fibrosis induced by TGF‐β and high glucose

Abstract Peritoneal fibrosis (PF) is an intractable complication of peritoneal dialysis (PD) that leads to peritoneal membrane failure. This study investigated the role of suppression of tumorigenicity (ST)2 in PF using patient samples along with mouse and cell‐based models. Baseline dialysate soluble (s)ST2 level in patients measured 1 month after PD initiation was 2063.4 ± 2457.8 pg/mL; patients who switched to haemodialysis had elevated sST2 levels in peritoneal effluent (1576.2 ± 199.9 pg/mL, P = .03), which was associated with PD failure (P = .04). Baseline sST2 showed good performance in predicting PD failure (area under the receiver operating characteristic curve = 0.780, P = .001). In mice with chlorhexidine gluconate‐induced PF, ST2 was expressed in fibroblasts and mesothelial cells within submesothelial zones. In primary cultured human peritoneal mesothelial cells (HPMCs), transforming growth factor‐β treatment increased ST2, fibronectin, β‐galactosidase and Snail protein levels and decreased E‐cadherin level. Anti‐ST2 antibody administration reversed the up‐regulation of ST2 and fibronectin expression; it also reduced fibrosis induced by high glucose (100 mmol/L) in HPMCs. Thus, high ST2 level in dialysate is a marker for fibrosis and inflammation during peritoneal injury, and blocking ST2 may be an effective therapeutic strategy for renal preservation.


| INTRODUC TI ON
Peritoneal fibrosis (PF) is a major complication of peritoneal dialysis (PD) that undermines peritoneal membrane function, eventually necessitating the discontinuation of PD. [1][2][3][4][5] Conventional glucose-based PD solutions are inexpensive, safe and effective for osmotic fluid removal, but the high glucose (HG) concentrations, glucose degradation products and acidity may be harmful. Prolonged exposure of the peritoneal membrane to a bio-incompatible dialysis solution and repeated episodes of peritonitis or haemoperitoneum can cause peritoneal injury resulting in PF. 2,6 Peritoneal fibrosis is characterized by fibroproliferative changes in the peritoneal membrane, the denudation and altered appearance of peritoneal mesothelial cells (PMCs), accumulation of extracellular matrix (ECM) molecules in submesothelial areas and vasculopathy. [7][8][9] The key mediator of this process is transforming growth factor (TGF)-β. Mesothelial cells synthesize ECM molecules including collagens I, III and IV, and fibronectin when exposed to TGF-β. 10 The TGF-β signalling pathway is transduced via its downstream effectors SNAIL, p65, β-galactosidase and connective tissue growth factor/cysteine-rich protein 61/nephroblastoma-overexpressed gene (CCN)1 to mediate fibrosis, inflammation and senescence. 11 A high concentration of glucose was shown to directly activate TGF-β and stimulate fibronectin synthesis in PMCs, leading to ECM accumulation and PF. 12,13 Suppression of tumorigenicity (ST)2 is an interleukin (IL)-1 receptor family member that exists as soluble (sST2) and trans-membrane (ST2L) isoforms. Recent studies have shown that the IL-33/ ST2 pathway plays an important role in inflammation and fibrosis in various organs such as heart, 14 lung, 15 gastrointestinal tract 16 and liver 17 ; moreover, sST2 is a useful biomarker for predicting clinical outcome in graft-vs-host disease (GVHD), 18,19 myocardial infarction 20 and heart failure 21 and kidney disorders such as IgA nephropathy, 22 lupus nephritis 23 and chronic kidney disease. [24][25][26] However, the role of ST2 in PF has not been previously reported.
To address this issue, the present study investigated the association between PD effluent levels of sST2 and PD failure (PDF) and the mechanism of action of ST2 in PF using in vitro and in vivo systems.

| Study population
The study protocol received full approval from the institutional review board of Seoul National University Hospital (no. H-1701-133-829). All procedures were performed in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards. Informed consent was obtained from study participants, and blood and peritoneal effluent samples were collected, stored and monitored by the Seoul National University Hospital Human Biobank.
A total of 75 end-stage renal disease (ESRD) patients with PD were enrolled between January 2010 and December 2016.
Demographic and clinical information at the time of diagnosis including age, gender, body mass index (BMI) and comorbidities (eg hypertension and diabetes mellitus [DM]) was extracted from electronic medical records. Samples and clinical data were collected at four time-points: baseline (defined as the first 6 months of PD) and at 1, 2 and 3 consecutive years. Exclusion criteria were age under 18 years, solid or haematological malignancies, and kidney transplantation within 3 months.

| Biochemical parameters
Biochemical parameters including blood haemoglobin and serum albumin and creatinine (Cr) levels were measured using a Modular D2400 analyser with ISE900 module (Hitachi) and a cobas 8000 modular analyser (Roche Diagnostics).

| Measurement of peritoneal effluent sST2 levels by enzyme-linked immunosorbent assay
sSuppression of tumorigenicity (sST)2 levels were measured with an ELISA kit (R&D Systems; cat. no. M3300) according to the manufacturer's instructions. The lower limit of detection for sST2 was 31.3 pg/mL.

| Isolation and primary culture of HPMCs
HPMCs from PD patients were obtained from PD effluent as previously described. 27,28 Briefly, effluent from clinically stable patients was drained and immediately processed. Bags were suspended for  Note: Values represent mean ± SD or number (percentage) unless otherwise indicated. Abbreviations: BMI, body mass index; D/P, dialysate-to-plasma ratio; ESRD, end-stage renal disease; hs-CRP, high-sensitivity C-reactive protein; Kt/V, (dialyser clearance of urea × dialysis time)/ volume of distribution of urea; NS; not significant; PDF, peritoneal dialysis failure.

TA B L E 1 Baseline characteristics of PD patients
F I G U R E 1 Effluent sST2 concentrations predict survival of PD patients. A, High baseline sST2 levels were detected in peritoneal effluent samples from patients who changed their dialysis modality to haemodialysis due to PDF (n = 10) as compared to other patients (n = 65). B, ROC curve analysis of baseline sST2 effluent concentrations in samples from patients at the start of PD treatment who did or did not develop PDF. AUC, area under the curve; CI, confidence interval and 2% Biogro-2 (Biological Industries), which contains insulin, transferrin, ethanolamine and putrescine. The medium was replaced every 2-3 days. Peritoneal leucocytes that adhered to the plates were detached after 48 hours and removed in subsequent washes.
HPMCs from the second passage were used for experiments.
Primary cultured HPMCs were seeded in six-well plates or 10-cm 2 plastic culture dishes in Dulbecco's modified Eagle's medium/F12 under normal glucose or HG conditions. When the cells reached 70%-80% confluence, they were serum-starved for 24 hours to synchronize cell growth.

| Fluorescence-activated cell sorting (FACS)
Primary cultured HPMCs were sorted using a

| Animals and treatments
Animal experiments were performed with the approval of the

| Real-time quantitative PCR analysis
Total RNA was extracted from the peritoneum, and the mRNA levels of target genes were assayed by real-time quantitative PCR. Briefly,  Real-time qPCR was conducted on an ABI PRISM 7500 sequence de-

| Statistical analysis
Results are expressed as mean ± SD or SEM.

| Patients who develop PDF show higher dialysate levels of sST2
Patient baseline characteristics are shown in Table 1 (90%) and specificity (58.5%) to predict subsequent development of PDF was 1155 pg/mL or higher (Figure 1). A multivariate analysis revealed that baseline sST2 level in effluent was independently associated with risk of PDF development after adjusting for peritoneal transport parameters (weekly Kt/V and D/P Cr) and clinical covariates such as age, sex, BMI, DM, hypertension, haemoglobin, serum albumin and Cr (Table 2).

| TGF-β induces fibrosis in primary cultured HPMCs
To confirm that ST2 is expressed during PF in HPMCs, the cells were treated with rTGF-β (1, 2 and 4 ng/mL) and ST2 protein expression was evaluated by Western blotting. ST2 was up-regulated by rTGF-β treatment; this was accompanied by increases in fibronectin, β-galactosidase and Snail and a decrease in E-cadherin expression (Figure 2A,B). These results indicate that TGF-β induces fibrosis in HPMCs, with up-regulation of ST2.

| ST2 blockade attenuates TGF-β-induced epithelial-to-mesenchymal transition (EMT) in HPMCs
To evaluate the role of ST2 in fibrosis, rTGF-β-stimulated HPMCs were pre-treated with ST2 blocking Ab. As expected, ST2 blockade suppressed the rTGF-β-induced increase in sST2 expression in a dose-dependent manner, with a concomitant decrease in fibronectin level ( Figure 2C,D). Thus, TGF-β induces EMT in HPMCs, an effect that is abrogated by blocking ST2 function.

| HG concentration induces fibrosis in primary cultured HPMCs
We investigated whether ST2 plays a similar role in HG-stimulated HPMCs by FACS analysis. The number of ST2-positive HPMCs was increased upon treatment with HG solution (100 and 200 mmol/L), which is similar to commercial 2.5% and 4.25% PD solutions (Figure 3).

| ST2 blockade reverses PF in HPMCs cultured under HG conditions
To determine whether ST2 blockade affects PF, we investigated changes in cell morphology and expression of fibrosis-related molecules in HG-stimulated HPMCs following treatment with ST2 mAb.
HG (200 mmol/L) induced a fibroblast-like morphology characterized by a spindle-like shape, in contrast to the cuboidal form of unstimulated HPMCs ( Figure 5A). Interestingly, administration of sST2 mAb (0.5 and 1.0 μg/mL) restored a normal cell morphology. Consistent with these results, sST2 mAb abrogated the up-regulation of the mesenchymal markers fibronectin, collagen 4 and p-p65 and down-regulation of the epithelial marker E-cadherin induced by HG ( Figure 5B,C).
Moreover, FACS analysis revealed that the proportion of αSMA-positive HPMCs-which was increased under HG conditions-was reduced by ST2 blockade ( Figure 5D). These results suggest that ST2 mediates HG-induced fibrosis in HPMCs, which can be mitigated by ST2 inhibition.

| ST2 expression in a mouse model of PF
To establish the expression profile of ST2 during PF progression, we carried out an immunohistochemical analysis of a mouse model of CG-induced PF. In control mice, peritoneal tissues showed no thickening of the submesothelial area. However, CG-treated mice showed elevated levels of collagen 1, fibronectin and ST2 in IHC staining ( Figure 6A). Moreover, the mRNA expression of fibronectin and ST2 was also increased in the peritoneal tissue of CG-treated mice ( Figure 6B). In Western blot analysis, the protein expression of fibronectin, ST2 and IL-33 was also significantly increased, which is in agreement with our in vitro findings ( Figure 6C,D).

| Amelioration of PF in ST2 knockout mice
We finally examined whether deletion of ST2 could ameliorate PF in CG-treated mice. There was no histological difference in the perito- In summary, the results of this study demonstrate that high baseline sST2 level in PD effluent is associated with increased risk of PDF and that ST2 blockade can ameliorate PF. Thus, effluent sST2 level was found to be a promising biomarker for peritoneal damage, and therapeutic strategies that inhibit ST2 function may be effective in preventing PF development and progression.

ACK N OWLED G EM ENTS
The biospecimens for this study were provided by the Seoul National University Hospital Human Biobank, a member of the Korea Biobank Network, which is supported by the Ministry of Health and Welfare.
All samples derived from the National Biobank of Korea were obtained with informed consent under institutional review board-approved protocols. (IRB 1506-097-681).

CO N FLI C T O F I NTE R E S T
No conflicts of interest, financial or otherwise, are declared by the authors.