Indoxyl Sulfate Induces Leukocyte-Endothelial Interactions through Up-regulation of E-selectin*

Despite a positive correlation between chronic kidney disease and atherosclerosis, the causative role of uremic toxins in leukocyte-endothelial interactions has not been reported. We thus examined the effects of indoxyl sulfate, a uremic toxin, on leukocyte adhesion to activated endothelial cells and the underlying mechanisms. Pretreatment of human umbilical vein endothelial cells (HUVEC) with indoxyl sulfate significantly enhanced the adhesion of human monocytic cells (THP-1 cell line) to TNF-α-activated HUVEC under physiological flow conditions. Treatment with indoxyl sulfate enhanced the expression level of E-selectin, but not that of ICAM-1 or VCAM-1, in HUVEC. Indoxyl sulfate treatment enhanced the activation of JNK, p38 MAPK, and NF-κB in TNF-α-activated HUVEC. Inhibitors of JNK and NF-κB attenuated indoxyl sulfate-induced E-selectin expression in HUVEC and subsequent THP-1 adhesion. Furthermore, treatment with the NAD(P)H oxidase inhibitor apocynin and the glutathione donor N-acetylcysteine inhibited indoxyl sulfate-induced enhancement of THP-1 adhesion to HUVEC. Next, we examined the in vivo effect of indoxyl sulfate in nephrectomized chronic kidney disease model mice. Indoxyl sulfate-induced leukocyte adhesion to the femoral artery was significantly reduced by anti-E-selectin antibody treatment. These findings suggest that indoxyl sulfate enhances leukocyte-endothelial interactions through up-regulation of E-selectin, presumably via the JNK- and NF-κB-dependent pathway.

with the risk of cardiovascular events, even when the dysfunction is mild (3).
Leukocyte-endothelial interactions play an important role in the development of atherosclerosis (4). Cell adhesion molecules belonging to the immunoglobulin superfamily, such as ICAM-1 (intercellular cell adhesion molecule-1) and VCAM-1 (vascular cell adhesion molecule-1), together with members of the selectin family, including E-selectin, are upregulated to mediate monocyte/macrophage infiltration into atherosclerotic lesions (4,5).
Indoxyl sulfate is a uremic toxin synthesized in the liver from indole, a metabolite of tryptophan produced by the intestinal flora (6). In CKD patients, the serum levels of indoxyl sulfate are increased significantly compared with those in healthy individuals (7), and a number of studies have indicated that indoxyl sulfate accelerates glomerular sclerosis, whereas its accumulation promotes renal failure (8 -10).
Other studies also showed that indoxyl sulfate induces endothelial dysfunction by releasing endothelial microparticles (11) and producing reactive oxygen species (ROS) (12). However, its effect on endothelial inflammatory processes such as leukocyte recruitment to vascular endothelium has not been reported.
We report for the first time that indoxyl sulfate enhances monocyte adhesion to vascular endothelium through up-regulation of E-selectin and augmentation of oxidative stress in both in vitro and in vivo models. The underlying mechanisms seem to involve activation of JNK and NF-B. Our findings reveal a previously unrecognized molecular link between uremic toxins and cardiovascular diseases.
Cell Cultures-Human umbilical vein endothelial cells (HUVEC) were purchased from Sanko Junyaku (Tokyo, Japan) and cultured in endothelial growth medium-2 (Lonza, Walkersville, MD) at 37°C in a humidified atmosphere of 5% carbon dioxide. Plastic culture dishes were precoated with 1% (w/v) collagen, and HUVEC were used between passages 1 and 3. For use in a flow chamber apparatus, HUVEC were placed onto 22-mm fibronectin-coated glass coverslips. THP-1, a human monocytic cell line, was obtained from American Type Culture Collection, and the cells were maintained in RPMI 1640 medium supplemented with 10% FCS, 100 IU/ml penicillin, 100 g/ml streptomycin, and 2 mmol/ liter L-glutamine.
Monocyte Adhesion Assay-HUVEC monolayers on coverslips were treated with various concentrations of indoxyl sulfate for 20 h and then stimulated by the addition of TNF-␣ for 4 h. The parallel plate flow chamber and the protocol for the adhesion assay under physiological flow conditions have been described in detail previously (14). In brief, HUVEC monolayers were positioned in a flow chamber mounted on a Nikon inverted microscope. THP-1 cells (1 ϫ 10 6 /ml) were perfused through the chamber with a syringe pump (PHD2000, Harvard Apparatus Inc., Holliston, MA) for 10 min at a controlled flow rate to generate a shear stress of 1.0 dyne/cm 2 . The entire period of perfusion was recorded by videotape and then transferred to a personal computer for image analysis to determine the number of adherent cells on HUVEC monolayers in 10 randomly selected ϫ15 microscope fields.
Luciferase Reporter Gene Assay-An NF-B-firefly luciferase cDNA construct (pNF-B-Luc) and a thymidine kinase-Renilla luciferase construct (pRL-TK) were obtained from Clontech. HUVEC were cultured in 24-well plates and transiently transfected using Lipofectamine LTX transfection reagents (Invitrogen). Briefly, 500 ng of the pNF-B-Luc vector and 10 ng of the internal control pRL-TK were cotransfected into HUVEC. The culture medium was changed 4 h after transfection, and the HUVEC were incubated for another 18 h before use. The transfected HUVEC were incubated with indoxyl sulfate for 20 h and then stimulated with TNF-␣ (100 pg/ml) for 4 h. The firefly luciferase activity of the whole cell lysate was measured with the Dual-Luciferase reporter assay system (Promega, Madison, WI), according to the manufacturer's protocol, using a luminometer. Renilla luciferase activity was used to normalize the activity of firefly luciferase.
Surgical Procedure-Renal failure was induced in 9-weekold male C57BL/6J mice (Oriental Yeast, Tokyo) or BALB/c mice (Japan Crea Laboratory, Tokyo) using two-step surgical nephrectomy as reported previously (15). Briefly, under intraperitoneal anesthesia with sodium pentobarbital (Schering-Plough Corp., Kenilworth, NJ) at 65 mg/kg, two of the three branches of the left renal artery were ligated through a lateral incision. One week after the first operation, the right kidney was removed after ligation of the renal blood vessels and ureter under anesthesia as described above.
Four weeks after the procedure, blood and blood pressure were measured. Mice with blood urea nitrogen between 53 and 90 mg/dl and systolic blood pressure between 118 and 154 mm Hg were allocated to experimental groups. Seven weeks after the procedure, half of the mice were administered 0.065% indoxyl sulfate (200 mg/kg/day) in drinking water (referred to as NxϩIS (nephrectomized with indoxyl sulfate treatment); n ϭ 5), whereas the other half were given only water (Nx; n ϭ 5). Ten days later, leukocyte adhesion to the femoral artery was assessed by intravital microscopy (IVM) and image analyses as described previously (16). In brief, mice were injected via the left femoral vein with rhodamine 6G chloride (0.3 mg/kg in 200 -300 l of PBS; Molecular Probes) to label leukocytes in vivo. The femoral artery was found within 30 min after injection of rhodamine 6G chloride and visualized with an Olympus microscope (Model BX51WI) equipped with a water immersion objective. Adhesion of labeled leukocytes was clearly visualized on the anterior half of the vessels facing the objective. All images were recorded using a computer-assisted image analysis program (Meta-Morph). The number of adherent leukocytes (i.e. those that did not move for Ͼ3 s during the 1-min recording period) was counted along a region of interest.
The mice were killed, and whole blood samples were collected from the heart using heparinized syringes. After perfusion via the left ventricle with ice-cold PBS, the aorta was dissected, snap-frozen in liquid nitrogen, and stored at Ϫ80°C until RNA isolation. For E-selectin blocking study, 1 mg/kg anti-E-selectin monoclonal antibody (n ϭ 9) or isotype control antibody (control IgM; n ϭ 9) was injected into the tail veins of BALB/c mice once daily for 9 consecutive days prior to IVM analysis on day 10.
Complementary DNA Preparation and Real-time Quantitative PCR-Individual mouse aortas were homogenized, and total RNA was isolated with an RNeasy mini column kit (Qiagen, Hilden, Germany). RNA purity and concentration were determined by measuring absorbance at 260 and 280 nm, respectively. cDNA was produced from 0.5 g of RNA using a PrimeScript RT-PCR reagent kit (TAKARA BIO Inc., Kyoto) with random hexamers in 10 l of reaction solution at 37°C for 15 min.
Real-time quantitative RT-PCR was performed with a LightCycler (Roche Applied Science) to quantitate the mRNA expression of ICAM-1, VCAM-1, E-selectin, p47 phox , p22 phox , and GAPDH in mouse aortas. All primers were obtained from Qiagen. Quantitative RT-PCR was carried out using a QuantiTect TM SYBR Green RT-PCR kit (Qiagen) with GAPDH as an internal control.
Statistical Analysis-Data are expressed as means Ϯ S.E. One-way analysis of variance with Tukey's post hoc test or two-tailed unpaired t test was used to estimate statistical sig-nificance. A p value Ͻ0.05 was considered to be statistically significant.

Indoxyl Sulfate Enhances Monocyte Adhesion to TNF-␣activated Vascular
Endothelium-First, we examined the effects of indoxyl sulfate on leukocyte-endothelial interactions under physiological flow conditions (shear stress of 1.0 dyne/ cm 2 ). Although indoxyl sulfate did not induce significant monocytic THP-1 cell adhesion to non-stimulated HUVEC, it significantly enhanced THP-1 cell adhesion to TNF-␣-stimulated HUVEC (Fig. 1A). The effect of indoxyl sulfate became significant when HUVEC were treated with 100 pg/ml TNF-␣ (Fig. 1B). Considering the relatively low blood levels of TNF-␣ even in atherogenic conditions, our finding may indicate pathophysiologically relevant inflammatory situations in vivo. Indoxyl sulfate enhanced monocytic THP-1 cell adhesion to TNF-␣-activated HUVEC in a dose-dependent manner (Fig.  1C). Furthermore, indoxyl sulfate also enhanced the adhesion of freshly isolated human peripheral blood monocytes to HUVEC (supplemental Fig. S1).
To identify the adhesion molecules responsible for this effect of indoxyl sulfate, Western blot analysis was carried out. As shown in Fig. 1D, indoxyl sulfate enhanced TNF-␣-induced E-selectin expression in HUVEC. The effect of indoxyl sulfate (2.0 mmol/liter) was observed at a TNF-␣ concentration as low as 25 pg/ml and became saturated at 250 pg/ml (Fig. 1D), whereas as little as 0.2 mmol/liter indoxyl sulfate significantly increased TNF-␣-induced E-selectin expression (Fig. 1E). The expression of ICAM-1 and VCAM-1 was not significantly enhanced by indoxyl sulfate treatment (Fig. 1, D  and E).
To confirm the importance of indoxyl sulfate-enhanced E-selectin expression, we examined adhesion assays using functional blocking antibodies against E-selectin, ICAM-1, and VCAM-1. As shown in supplemental Fig. S3, anti-E-selectin antibody, but not antibodies to ICAM-1 and VCAM-1, inhibited indoxyl sulfate-enhanced leukocyte adhesion, whereas all antibodies inhibited base-line leukocyte adhesion induced by TNF-␣.
NF-B Signaling Pathway Is Involved in Indoxyl Sulfateenhanced Leukocyte-Endothelial Interactions-We also examined the potential contribution of NF-B. Indoxyl sulfate potentiated the activation of NF-B in TNF-␣-activated HUVEC, as indicated by the phosphorylation of the p65 component of NF-B (Fig. 3A). To further confirm the direct involvement of an NF-B-dependent pathway, HUVEC were transiently transfected with an NF-B promoter-reporter construct. Indoxyl sulfate enhanced TNF-␣-induced luciferase activity (Fig. 3B). An inhibitor of the NF-B signaling pathway, BAY11-7082, blocked indoxyl sulfate-enhanced THP-1 cell adhesion (Fig. 3C) and E-selectin expression (Fig.  3D). Indoxyl sulfate-enhanced endothelial activation was inhibited when NF-B was knocked down by siRNA against the NF-B p65 subunit (supplemental Fig. S4). Although these  DECEMBER 10, 2010 • VOLUME 285 • NUMBER 50 inhibitors reduced base-line THP-1 adhesion to activated HUVEC, they significantly inhibited indoxyl sulfate-enhanced adhesion (supplemental Fig. S5).

Indoxyl Sulfate Induces Leukocyte-Endothelial Interactions
We also examined whether indoxyl sulfate modifies TNF-␣ receptor expression in HUVEC. We documented that indoxyl sulfate did not change the mRNA expression levels of TNFR1 and TNFR2 even after treatment with TNF-␣ (supplemental Fig. S6).

Role of Oxidative Stress in Monocytic Cell Adhesion and E-selectin Expression-Because indoxyl sulfate has been
shown to cause oxidative stress in HUVEC (12), we evaluated the role of intracellular ROS in indoxyl sulfate-mediated endothelial activation. Staining for ROS using dihydroethidium revealed a marked increase in ROS production in indoxyl sulfate-treated HUVEC (Fig. 4A). Both the NAD(P)H oxidase inhibitor apocynin and the glutathione donor N-acetylcysteine reduced indoxyl sulfate-induced THP-1 cell adhesion to HUVEC (Fig. 4B) and E-selectin expression (Fig. 4C). In contrast, allopurinol (xanthine oxidase inhibitor) and rotenone (mitochondrial electron transport inhibitor) failed to change either indoxyl sulfate-induced E-selectin expression or THP-1 cell adhesion (data not shown).
Probenecid Inhibits Indoxyl Sulfate-enhanced Monocytic Cell Adhesion and E-selectin Expression-A recent study suggested a role of cell-surface anion transporters such as organic anion transporters (OATs) in indoxyl sulfate-mediated cell activation (17). In our experiment, we confirmed the existence of OAT1 and OAT3 in HUVEC by RT-PCR (data not shown). We then examined the effect of an inhibitor of OATs, probenecid, on indoxyl sulfate-mediated endothelial activation. Probenecid inhibited the indoxyl sulfate-mediated increases in THP-1 adhesion (Fig. 5A) and E-selectin expression (Fig. 5B) but did not suppress the expression levels of ICAM-1 and VCAM-1 (data not shown).

FIGURE 2. Effects of indoxyl sulfate on TNF-␣-induced MAPK pathways.
A, Western blot detection of phosphorylated (p) and total MAPK family members. HUVEC were incubated in the presence of the indicated concentrations of indoxyl sulfate (IS) for 20 h and then with (ϩ) or without (Ϫ) TNF-␣ (100 pg/ml) for 15 min. B and C, adhesion assay and Western blotting for E-selectin, respectively. HUVEC were incubated with (ϩ) or without (Ϫ) 0.2 mmol/liter indoxyl sulfate for 20 h and then treated with the JNK inhibitor SP600125 (10 mol/liter), the p38 MAPK inhibitor SB203580 (5 mol/liter), or the ERK1/2 inhibitor U0126 (10 mol/liter) for 30 min, followed by incubation with (ϩ) or without (Ϫ) TNF-␣ (100 pg/ml) for 4 h. The cells were subjected to adhesion assays. The data from the adhesion assay are means Ϯ S.E. (n ϭ 10). HPF, high-power field. *, p Ͻ 0.01 versus without TNF-␣; †, p Ͻ 0.01 versus with TNF-␣ and indoxyl sulfate; NS, not significant. The data shown are representative of three independent experiments. Indoxyl Sulfate Induces Leukocyte-Endothelial Interactions through Up-regulation of E-selectin in Vivo-Finally, we investigated whether indoxyl sulfate enhances leukocyte-endothelial interactions in vivo using a novel IVM system (16,18). We generated nephrectomized CKD mice and allocated them to indoxyl sulfate-treated (NxϩIS) and non-treated (Nx) groups. Systolic blood pressure was elevated and plasma urea was high in both Nx and NxϩIS groups compared with normal control mice. As expected, the NxϩIS group exhibited significantly higher plasma indoxyl sulfate than the Nx group, whereas other plasma parameters, body weight, and blood pressure were not significantly different between the two groups (Table 1). IVM analysis showed that the number of leukocytes that adhered to the femoral artery was significantly increased in the NxϩIS group compared with the Nx group (Fig. 6, A and B). Simultaneous measurements of mRNA of adhesion molecules in the aorta revealed that E-selectin, but not ICAM-1 or VCAM-1, was significantly up-regulated in the NxϩIS group compared with the Nx group (Fig. 6C). The gene expression levels of the NAD(P)H oxidase subunits p47 phox and p22 phox were also up-regulated in the NxϩIS group (Fig. 6D).
To confirm the role of E-selectin in vivo, we utilized a functional blocking antibody against E-selectin to modulate leukocyte adhesion to the femoral artery. We injected anti-E-selectin antibody or control antibody (control IgM) into NxϩIS mice and conducted IVM analysis. Urea, lipid profile, and blood pressure did not change significantly after anti-E-selectin antibody treatment (supplemental Table S1). The number of adherent leukocytes decreased in the anti-E-selectin antibody group compared with the control IgM group (Fig. 6E).

DISCUSSION
In this study, we found that pretreatment of HUVEC with indoxyl sulfate prior to incubation with TNF-␣ or IL-1␤ increased the adhesion of monocytic THP-1 cells or human peripheral blood monocytes to HUVEC in a dose-dependent manner ( Fig. 1 and supplemental Fig. S1). Considering that the concentrations of indoxyl sulfate (0.2 mmol/liter) (7,8) and TNF-␣ and IL-1␤ (19) were comparable with their blood levels observed in CKD patients, our findings indicate an important molecular background underlying the inflammatory process in CKD patients with high circulating uremic toxins. Our data may explain a recent observation by Barreto et al. (20) that serum indoxyl sulfate is associated with the incidence of vascular diseases and total mortality in CKD patients.
Indoxyl Sulfate and Endothelial Activation-We also found that indoxyl sulfate enhanced TNF-␣and IL-1␤-activated E-selectin expression in HUVEC, which plays a dominant role in monocyte adhesion during inflammation (21,22). Furthermore, anti-E-selectin antibody significantly inhibited leukocyte adhesion enhanced by indoxyl sulfate (supplemental Fig.  S3). Thus, the present results raise the possibility that indoxyl sulfate exacerbates CKD-related vascular endothelial inflammation through up-regulation of E-selectin.
In CKD patients, monocyte/macrophage accumulation has been observed in the extracapillary areas of glomeruli (23,24). Intense expression of E-selectin (24,25) and activated glomerular endothelial cells (26) have been observed in the kidneys of CKD animal models. Furthermore, circulating E-selectin is a strong predictor of death and cardiovascular events in patients with end-stage renal disease (27,28). These observations suggest an important role of E-selectin in renal deterioration.
Indoxyl Sulfate Enhances E-selectin Expression via JNK-Our results indicate that indoxyl sulfate mediates E-selectin overexpression through enhancement of the JNK and NF-B signaling pathways (Figs. 2 and 3). The importance of JNK in E-selectin expression in comparison with other MAPK family members has been repeatedly emphasized (29 -31). Although Kuldo et al. (31) demonstrated that inhibition of p38 MAPK diminished E-selectin expression in HUVEC activated by TNF-␣ for 24 h, we failed to detect any effect of a p38 MAPK inhibitor on E-selectin expression stimulated by TNF-␣ for 4 h.
We confirmed the expression of OAT1 and OAT3, which are molecules responsible for the incorporation of indoxyl sulfate, in HUVEC (32). Moreover, probenecid, a chemical inhibitor of OATs, suppressed indoxyl sulfate-mediated Eselectin expression in HUVEC and subsequent THP-1 cell adhesion. These results suggest that OAT1-and/or OAT3mediated intracellular transport may be necessary for the exhibition of the indoxyl sulfate effects in HUVEC.
Role of E-selectin in Indoxyl Sulfate-induced Inflammation in Vivo-As demonstrated previously by others, nephrectomy per se does not induce significant vascular inflammatory reactions in mice unless they are apolipoprotein E-deficient (knock-out) (33,34). However, indoxyl sulfate induced E-selectin expression as well as leukocyte adhesion to the femoral artery in the Nx mice. Furthermore, anti-E-selectin antibody significantly blocked leukocyte accumulation in NxϩIS mice, suggesting a dominant role of E-selectin in the leukocyte adhesion observed in the Nx mice (35). These results suggest that indoxyl sulfate induces leukocyte-endothelial interactions through up-regulation of E-selectin in vivo.
Previous studies demonstrated that indoxyl sulfate produced oxidative stress in vivo and in vitro through NAD(P)H oxidase activity (12,36). Consistent with these observations, we documented up-regulation of p47 phox and p22 phox in Nx mice treated with indoxyl sulfate, suggesting a role of NAD(P)H oxidase in indoxyl sulfate-mediated oxidative stress in vivo. The therapeutic efficacy of quenching uremic toxins by oral adsorbents such as AST-120 (Kremezin, Kureha Corp., Tokyo) in the reduction of oxidative stress in CKD has been reported previously (36 -39).
In conclusion, indoxyl sulfate enhanced leukocyte-endothelial interactions via increased selectin expression and oxidative stress in vitro and in vivo. Our data suggest a novel causative role of uremic toxins in producing vascular inflammation in patients with renal dysfunction.