High glucose-induced intestinal epithelial barrier damage is aggravated by syndecan-1 destruction and heparanase overexpression

Syndecan-1 (Sdc1) and its endo-beta-d-glucuronidase heparanase (HPSE) are implicated in maintenance of intestinal epithelial barrier (IEB), but their alterations and roles in high-glucose/hyperglycaemia (HG) conditions have not been fully investigated. This study aimed to determine the expression pattern, the possible regulation mechanism of Sdc1 and HPSE in HG conditions, and their potential effects on IEB. Therefore, diabetic mice/cell models were developed, and tissue/serum samples, cell lysate and culture supernatants were harvested. The expression of Sdc1 and HPSE in control, HG and designated interventions groups were detected. Phosphorylations of mitogen-activated protein kinase signalling pathway (MAPK), the expressions of Occludin and ZO-1, and the levels of transepithelial electrical resistance (TEER) were measured and monitored. The results showed that in HG conditions, intestinal tissue and cellular Sdc1 were significantly decreased, but the expression of HPSE, and soluble Sdc1 in serum and culture supernatants were remarkably increased. Such alterations of Sdc1 and HPSE were associated with solely p38 MAPK activation, and were correlated with the reductions of Occludin, ZO-1 and TEER. Heparin (Sdc1 analogue) and SB203580 (a p38 MAPK inhibitor), instead of insulin, alleviated Sdc1 destruction and HPSE overexpression, and effectively prevented against the reductions of tight junctions and the abnormality of intestinal permeability in HG conditions. In conclusion, we confirm the unique alterations of Sdc1 and HPSE in HG conditions, and found their interactions with p38 MAPK activation and IEB. These indicate that Sdc1/HPSE modulation can be viewed as an important complementary treatment for relieving HG-induced gastrointestinal damage.


Introduction
Diabetes mellitus (DM) is a chronic disease requiring lifelong medical attention. With hundreds of millions suffering worldwide and a rapidly rising incidence, DM poses a great burden on healthcare systems [1,2]. It is worth to note that, besides relevant cardiovascular diseases, neuropathy, and kidney failure, gastrointestinal complications also occur frequently in diabetic patients [3]. They could be manifested by complaints (such as gastroparesis, constipation and diarrhoea), or lead to morphological and functional changes of the gastrointestinal tract (raise in mucosal surface area, intestinal weight and number of goblet cells per villus, increased intestinal permeability, etc.) [3][4][5]. Nowadays, however, it is too early to claim a complete insight of the inherent links between DM and diabetes-associated enteropathy. Further studies investigating the pathogenetic and regulation/control factors are urgently needed.
Syndecan-1 (Sdc1), a predominant member of type I transmembrane heparan sulphate proteoglycans, plays important roles in inflammation, wound healing and tumour progression [6][7][8]. Its structure mainly consists of a short conserved cytoplasmic and transmembrane domain, and a long, variable ectodomain carrying heparan sulphate (HS) chains. Membrane-bound Sdc1 can constitutively shed in a soluble form to extracellular and systemic circulation. It is accelerated by a variety of extracellular stimuli and matrix proteinase, such as heparanase (HPSE, an endo-beta-D-glucuronidase) [9][10][11]. Recent studies indicate that the interplay between Sdc1 and HPSE (also named as 'Sdc1-HPSE axis') carries broad biological connotations in formation and maintenance of intestinal barrier [12,13]. Studies have also suggested that intestinal specific loss of Sdc1 and abnormal HPSE might wreck the natural barrier, causing far more susceptibility to protein-loss enteropathy and bacterial translocation [14,15]. As the diabetic state and the dysregulation of Sdc1/HPSE both lead to the gastrointestinal disorders, we wonder whether there are any certain interactions between them. However, the possible interactions have been rarely investigated, and the potential roles of Sdc1/ HPSE in diabetic enteropathy have not been explored.
Therefore, we developed diabetic mice/cell models, applied biological experiments to study the alterations of Sdc1 and HPSE in high-glucose/hyperglycaemia (HG) conditions, and then investigated the subsequent changes of barrier function of intestinal epithelium and the possible regulation mechanism. We reported here that after HG stimulation, the dramatic Sdc1 destruction with synchronous HPSE elevation were presented and were correlated with abnormities of intestinal permeability, tight junctions and the activation of p38 mitogen-activated protein kinase (MAPK) signalling pathway. Heparin (which mimics Sdc1 function) and inhibitors targeting p38 MAPK pathway, instead of insulin, effectively reduced the alterations of Sdc1 and HPSE, and improve the impaired barrier function. These results suggest a potential mechanism of diabetic enteropathy, which is depending remarkably on Sdc1 and HPSE, and indicate a feasible option for future therapeutic strategy.

Induction of diabetic mice
The Ethics Committee of Nanfang Hospital, Southern Medical University, approved all of the protocols and procedures using animals (approval number: NFEC-201112-K6). A total of 14 female C57BL/6J mice (6-8 weeks old, weighing about 20 g) were obtained from the Animal Center of Southern Medical University (Guangzhou, China). All were housed in specified pathogens-free conditions and fed standard rodent chow and fresh distilled water. After 1-week quarantine, mice were divided randomly into normal control (NC) group and DM group, composed of seven mice each. Diabetes was induced with streptozotocin (STZ, 40 mg/kg) injected into abdominal cavity for five consecutive days by the method previously reported [16]. In NC group, mice were administrated with equal volume of citrate acid buffer as the solvent of STZ. All mice were killed at week 10 after modelling. Tissue samples removed from small intestines and serum samples were used to measure the expression of the interest proteins (Sdc1, HPSE, occludin, ZO-1, total/phosphorylated MAPK, etc.).

Cell culture and in vitro diabetes models
Normal rat small intestine crypt cell line (intestinal epithelial cell 6, IEC-6) was obtained from American Type Culture Collection (Rockville, MD, USA) and was maintained in DMEM (Gibco, Cambridge, MA, USA) supplemented with 10% foetal bovine serum (Gibco) at 37°C with an atmosphere of 5% CO 2 . Cells were grown on polyester membranes in Transwell inserts (6.5 mm, pore size 0.4 lm; Costar, Cambridge, MA, USA), glass or culture plates to be the adequate model for further study. Cells were conventionally grown for 96 hrs (48 hrs for immunofluorescence assay) before subsequent stimulations. To assess DM model, IEC-6 cells were exposed to normal (NG, 12.5 mM) or high concentration of D-glucose (HG, 50 mM); the corresponding control groups were exposed to L-glucose (LG) in normal medium with D-glucose (12.5 mM D-glucose plus 37.5 mM L-glucose) to account for medium hyperosmolarity. In addition, to evaluate the therapeutic effect, IEC-6 cells were cultured in the presence of high glucose alone (50 mM D-glucose, for 24 hrs, NC group), or high glucose with insulin (0.01 unit/ml, for 24 hrs, Ins group), or with heparin (0.5 lg/ml, for 24 hrs, Hep group), or with SB203580 (10 lg/ml, for 90 min., MI group) [17,18]. Culture supernatants and whole cell lysate were harvested at designated time and were stored for subsequent evaluations.

Measurement of transepithelial electrical resistance
Exactly, 2.0 9 10 6 IEC-6 cells per well were seeded on the collagencoated membrane Transwell inserts with 200 ll culture medium added to the apical chamber and 600 ll to the basolateral chamber. The electrical resistance of confluent polarized IEC-6 monolayers was measured by transepithelial electrical resistance (TEER) with an electrical resistance system (EVOM; World Precision Instruments, Berlin, Germany). A pair of chopstick electrodes was placed at each of the apical and basolateral chambers of three different points to evaluate TEER. Readings were taken every 24 hrs until the net TEER had 1367 risen steadily above 250 Ω cm 2 (at days 5-7). At this point, regulatory factors (PBS, D-/L-glucose, heparin, insulin, HPSE mAb, p38 MAPK inhibitors, etc.) were added to both the apical and the basolateral chamber, and TEER value was recorded at the appointed time after these interventions.

ELISA
Target proteins in culture medium and serum were quantitative determined by ELISA kits according to the manufacturer's instructions. Briefly, samples, standards and diluted biotinylated antibody were added into precoated wells and incubated for 1 hr at room temperature. After 3 washes, horseradish peroxidasestreptavidin conjugate was added, and the plate was incubated for 30 min. at room temperature, then substrate was added and the colour was allowed to develop for 10-15 min. The reaction was stopped with sulphuric acid, and the absorbance was read at 450 nm in an ELISA plate reader (Bio-Rad, Hercules, CA, USA).
The concentrations of target proteins were calculated based on the standard curve.

Western blotting
Total cellular proteins were extracted and separated in 10% SDS-PAGE gels and then transferred to Polyvinylidenefluoride (PVDF) membrane. PVDF blots were blocked with a solution of 5% (w/v) skim milk in Trisbuffered saline containing 0.1% Tween-20 (TBS-T) for 1 hr at room temperature. The primary antibodies were diluted and added, and the blots were incubated overnight at 4°C. After washed in TBS-T, the blots were incubated in relevant secondary antibodies at a dilution of 1:5000 for 1 hr. Immunoreactive bands were visualized using Immobilon Western HRP Substrate (Merck Millipore, Billerica, MA, USA) and analysis with the Bio-Image Analysis System (Syngene, Frederick, MD, USA). Band intensity was normalized to b-actin and quantitated by densitometry using Image J software (National Institutes of health). Data represent the average of three separate experiments.

Immunofluorescence
Monolayers on chamber slides or plates were washed three times with PBS and fixed in 4% paraformaldehyde for 10-15 min. at room temperature. After being made permeable with 0.5% Triton X-100 in PBS for 5 min., cells were then blocked with 5% PBS-diluted bovine serum albumin (blocking solution) for 30 min. at room temperature, and were incubated with the interest primary antibodies overnight at 4°C. After washes with PBS, the monolayers were incubated with the Alexa fluor-594 conjugated antibody (1:500) at room temperature for 1 hr in the dark. Cells were washed with PBS and nuclei stained with DAPI (Invitrogen, Carlsbad, CA, USA) and examined under the Olympus BX51 microscope (Tokyo, Japan).

Statistical analysis
All statistics were determined by SPSS software (Version 13.0, Chicago, IL, USA). Descriptive statistics were calculated with means and standard errors, then independent samples t-test for comparing means between two groups and one-way ANOVA test for among three or more groups (with LSD method for post hoc multiple comparisons) were used. Comparisons of ranked data were determined by Mann-Whitney U-test (between two independent groups) or Kruskal-Wallis H-test (among multiple independent groups). Spearman's correlation analysis was employed to define associations. A P < 0.05 was considered significant.

Dramatic Sdc1 destruction with synchronous HPSE elevation in diabetic mice
In vivo levels of Sdc1 and HPSE under control and diabetic conditions were detected by Western blotting, qRT-PCR and ELISA. In the diabetic group (DM), tissue Sdc1 from small-intestinal samples was significantly lower (1.002 AE 0.076 versus 0.510 AE 0.065, P = 0.002, Fig. 1A), but qRT-PCR showed no significant changes of Sdc1 mRNA when compared with NC (1.062 AE 0.106 versus 1.350 AE 0.149, P = 0.159, Fig. 1B), which indicated the decreases of Sdc1 protein derived from the destruction after its synthesis. Correspondingly, both of tissue protein and mRNA of HPSE were synchronously increased (0.184 AE 0.032 versus 0.402 AE 0.028, P = 0.003; 1.121 AE 0.172 versus 2.203 AE 0.236, P = 0.009) ( Fig. 1A and C). Meanwhile, the average of serum Sdc1 in NC and DM group were 12.2 AE 0.788 ng/ml and 22.2 AE 1.01 ng/ml (P = 4.97 9 10 À6 , Fig. 1D), serum HPSE were 364.3 AE 43.3 mU/ml and 961.8 AE 73.0 mU/ml (P = 1.35 9 10 À5 , Fig. 1E) respectively, both exhibiting notable elevations in DM groups. Significant correlations were found, not only between serum and tissue Sdc1 (r = À0.701, P = 0.005) but also between serum and tissue HPSE (r = À0.837, P = 1.86 9 10 À4 ). obtained from diabetic mice, high glucose led to obvious Sdc1 destruction (Figs 2A and C, 3A, B and D) and HPSE elevation (Fig. 3A, C and E). But it is also worth noting that no statistical differences were observed between NC (with normal culture medium) and LG group (with L-glucose culture medium) (Figs 2 and  3). These results also indicated the above alterations were not derived from the hyperosmolarity, but from some specific intracellular mediators and (/or) pathways after abnormal D-glucose absorption.

Sdc1-HPSE and p38 MAPK synergistically exacerbate barrier destruction
Compared to NC and LG group, the relative expression of Occludin in HG group was remarkably decreased (v 2 = 6.489, P = 0.039, Fig. 6A). Though the overall difference among groups was not statistically significant (v 2 = 5.600, P = 0.061), the relative expression of ZO-1 in HG group had shown much lower than NC and LG group (versus 0.220 AE 0.025 versus 0.802 AE 0.101, 0.647 AE 0.072) (Fig. 6A). Meanwhile, the IEC-6 cell monolayers in high concentration of D-glu-cose, instead of normal concentration of D-glucose or isotonic L-glucose, had an obviously lower TEER values (104.7 AE 6.0 versus 212.1 AE 10.9 and 195.4 AE 12.8 Ω cm 2 , P = 0.001), indicating a serious damage to the barrier structure and function (Fig. 6B). Specifically, correlation analyses suggested that there were apparently correlations among Sdc1, HPSE and TEER (all P < 0.05, Fig. 6B).
HPSE blockade and p38 stimulation were applied to further investigate the interplay among p38, Sdc1/HPSE alteration and barrier disorder. Compared with normal culture conditions (with 12.5 mM D-glucose, NC group), the expressions of Occludin, ZO-1 (Fig. 7A) and the level of TEER (Fig. 7B) in cells treated with Anisomycin (p38 activator) were remarkably decreased. However, when Anisomycin and HPSE mAb were both added, dramatic reversals of the decreases of Occludin, ZO-1 and TEER were performed (Fig. 7). These results indicated that under high-glucose condition, HPSE was essential and acted synergistically with p38 to trigger barrier destruction.

The effects of insulin, Sdc1 analogue and p38 inhibitor on barrier functions
As high-glucose cultivation led to dysregulations of Sdc1/HPSE, p38 activation and barrier damage, we consequently investigated whether insulin uptake or interventions targeted at Sdc1/HPSE and p38 could provide therapeutic effect. As complements to the above results, it was found that insulin caused the increased cleavage of Sdc1 (HG versus Ins group, 255.2 AE 8.9 versus 281.4 AE 7.1 pg/ml, P = 0.022) and higher expression of HPSE (197.6 AE 9.1 versus 226.6 AE 7.4 mU/ml, P = 0.019) in HG culture supernatants, while heparin (Sdc1 analogue) effectively reduced both soluble Sdc1 (200.2 AE 4.6 pg/ml, P = 0.0003) and HPSE (164.2 AE 5.67 mU/ml, P = 0.01) ( Fig. 8B and C). The subsequent Western blot and TEER measurement showed that insulin could not relieve the reduction in Occludin and ZO-1 (Fig. 8A); the co-incubation with insulin and high glucose even reduced TEER by about 26.8% than high glucose alone ( Fig. 8D; HG versus Ins group, P = 0.007). On the contrary, heparin and SB203580 effectively increased the expressions of Occludin and ZO-1, and TEER (P = 0.016, 0.025, 2.7 9 10 À4 , respectively; Fig. 8A and D).

Discussion
In the present study, dramatic destruction of Sdc1, elevation of HPSE expression, p38 MAPK activation and synchronous damage of intestinal barrier are performed after HG stimulation. These alterations are effectively overturned by exogenous additions of heparin (Sdc1 analogue) and SB203580 (p38 MAPK inhibitor), but not insulin. Our study therefore suggests the implications of Sdc1/HPSE abnormities, and therapeutical value of the relevant modulations towards intestinal barrier dysfunction in HG conditions. Reduced Sdc1 and elevated HPSE have been commonly reported in malignant tumours, bacterial infections and chronic inflammatory diseases [13,22,23]. The abnormities of Sdc1 and HPSE under HG Fig. 4 Diabetic condition increases the phosphorylation of p38 MAPK, but not ERK1/2 or JNK1/2. Western blots were probed with antibodies against phosphorylation-specific (p-) and total (t-) p38, ERK1/2 and JNK1/2. Compared to NC group, p/t-p38 was significantly increased in DM group (P = 0.002), while phospho-ERK and phospho-JNK showed no statistically significant changes (P = 0.141, 0.338, respectively). Statistically significant correlations were discovered between p/t-p38 and tissue Sdc1 (r = À0.785, P = 0.001), and p/t-p38 and tissue HPSE (r = 0.829, P = 2.5 9 10 À4 ). NC: normal control group, DM: diabetic group. Molecular weight: p/t-p38, 40 kD; p/t-JNK1/2, 46 and 54 kD; ERK1/2, 42 and 44 kD.  5 The activation of p38 MAPK under high-glucose cultivation. IEC-6 cells were cultured as mentioned in 'Materials and methods'. Western blots were applied with antibodies against phosphorylation-specific (p-) and total (t-) p38. Compared to the NC and LG group, p/t-p38 MAPK of HG group was increased remarkably (P = 0.039). Notably, significant correlations were also showed between p/t-p38 and cellular Sdc1 protein (r = À0.867, P = 0.002), soluble Sdc1, HPSE protein and soluble HPSE (all r = 0.867, P = 0.002). conditions might provide an alternative perspective to better interpret HG induced/related enteropathy. For example, previous studies have showed Sdc1 participates in modulating extracellular growth factors binding/activities and relevant signalling pathways, the pathological loss of Sdc1 therefore partly accounts for abnormal proliferation and death/apoptosis raised in HG condition [9,24]. Reduced Sdc1 and elevated HPSE presented in HG condition and after insulin treatment also offer some interpretations for the phenomenon that patients with type 2 diabetes have higher incidence of colorectal carcinoma, while insulin therapy may even increase this risk [25]. Recently, studies have demonstrated that Sdc1 and HPSE are important endogenous regulators of energy balance and nutrient metabolism. HPSE overexpression and subsequent HS chains cleavage would not only cause diarrhoea, reduce reflex hyperphagia and food intake [26][27][28], but also alter lipoprotein lipase and clearance of triglyceride [29,30].
These emphasize yet again the diverse functions of Sdc1 and HPSE, and suggest their pathological effects of Sdc1/HPSE imbalance on systemic and intestinal homoeostasis under HG condition.
The exact mechanism of Sdc1 alteration in HG group is not identified. Our study has showed that HG stimulation does not reduce Sdc1 mRNA level. In this view, HG stimulation should not been the direct effector of Sdc1 reduction and the decrease in Sdc1 protein arises from the destruction after its synthesis. Owing to the physiological functions and well correlation with Sdc1 alteration, enhanced HPSE should be one of the most essential components causing Sdc1 destruction. A relevant study conducted by Yang et al. [11] has also showed that HPSE regulates both the level and location of Sdc1 within the tumour microenvironment by enhancing its synthesis and subsequent shedding from the tumour cell surface. From this perspective, HPSE is an important target of HG stimulation. The damages of TJs are still thought to be important manifestations of intestinal morphological and functional alterations [3]. Our study, however, has showed Sdc1/HPSE can play a unique role on TJs expressions. ZO-1 and Occludin decrease synchronously with Sdc1 destruction triggered by HG stimulation, but increase after adding Sdc1-HPSE regulators (heparin, HPSE mAb). Nowadays, the interactions between Sdc1/HPSE and TJs have been paid increasing attention. Smith et al. has also reported increased Claudin-2 expression begins in early HIV-1 infection and is co-expressed with Sdc1 in the intestinal epithelium [31]. These results highlight the capacity of Sdc1/HPSE on TJs regulation, which are indispensable in barrier formation and maintenance.
Increased p38 phosphorylation after HG stimulation might be essential for the abnormalities of many cellular processes [32]. For example, p38 activation promotes the production of fibronectin by the mesangial cells, and facilitates epithelial to mesenchymal transdifferentiation [33]. Notably, it is supposed that there are complex relationships between HPSE and p38 MAPK pathway. Previous studies have showed HPSE may enhance three main MAPK members (including p38, ERK and JNK, solely or simultaneously) by phosphorylation [34][35][36]; certain pathophysiology processes induced by HPSE could be mediated after p38 regulation [37,38]. In our study, however, the concurrent of HPSE up-regulation and sole p38 activation is a unique feature of HG induction and the mechanism for sole p38 activation is still obscure. Meanwhile, our study revealed p38 MAPK inhibitor reduced HPSE expression and the barrier destruction triggered by p38 activation was alleviated after HPSE blockade. Therefore, there is not only regulatory relationship but also mutual cooperation/interaction between HPSE and p38 MAPK.
Hypoglycaemic therapy remains to be the major therapeutical method reducing damage to intestinal epithelial cells and barrier. The features of Sdc1 and HPSE alterations, however, might provide a new rationale for therapeutic approach. Though studies showed HS mimetics/HPSE inhibitors (heparin, PI-88, SST001, etc.) fortify Sdc1 content of vascular endothelial cells and prevent the progression of diabetic angiopathy [39,40], there are few studies investigating whether they protect against HG-induced gastrointestinal damage. Our study here has presented heparin effectively maintains tight junctions and barrier function in HG condition. Therefore, preservation of Sdc1/HS on epithelial cells may be a strategy to prevent and treat DM-associated gastrointestinal complications. Deserve to note that HS mimetics, including heparin and modified heparin, would disturb or inhibit the coagulation cascade. Hypo-/non-anticoagulant heparin and anti-HPSE drugs may become promising options for clinical application.
In conclusion, we have found the unique alterations of Sdc1 and HPSE in HG condition, and confirmed their interactions with p38 MAPK activation and maintenance of intestinal barrier. The results indicate that Sdc1/HPSE modulation can be viewed as a potential treatment for relieving HG-induced gastrointestinal damage.