Indoxyl sulfate impairs valsartan-induced neovascularization

Abstract

Studies revealed that the use of renin-angiotensin-aldosterone system antagonism is not associated with a statistically significant reduction in the risk of cardiovascular events in patients with chronic kidney disease (CKD) compared with that in the general population. We tested the hypothesis that indoxyl sulfate (IS) can interfere with the protective effect of valsartan-mediated on endothelial function in vitro and neovascularization in mice underwent subtotal nephrectomy. In human aortic endothelial cells, we first demonstrated that IS impaired the valsartan-mediated phosphorylation of eNOS^Thr495, nitric oxide production and tube formation via NADPH oxidase (NOX) and protein kinase C (PKC) phosphorylation, but this effect was suppressed by cotreatment with apocynin and calphostin C. In vivo, IS attenuated valsartan-induced angiogenesis in Matrigel plugs in mice. Moreover, in subtotal nephrectomy mice who underwent hindlimb ischemic surgery, valsartan significantly increased the mobilization of endothelial progenitor cells in circulation as well as the reperfusion of blood flow and density of CD31⁺ capillaries in ischemic limbs. However, IS attenuated the protective effect of valsartan-induced neovascularization and increased the expression of p-PKCα^Ser657 and p-eNOS^Thr497 in ischemic limbs. Cotreatment of apocynin and calphostin C reversed the IS impaired-neovascularization and decreased the expression of p-PKCα^Ser657 and p-eNOS^Thr497 in ischemic limbs. Our study suggests that the NOX/PKC/eNOS signaling pathway plays a pivotal role in the IS-mediated inhibition of valsartan-conferred beneficial effects on endothelial function in vitro and neovascularization in subtotal nephrectomy mice. We proposed a novel causative role for IS in cardiovascular complications in CKD patients.

Keywords: Chronic kidney disease; Indoxyl sulfate; Neovascularization; Valsartan.

Fig. 1

IS impairs NO generation and attenuates valsartan-mediated NO production and tube formation in HAECs. (A) HAECs were treated with IS (0, 15, 30, 60, 120, 240 μg/mL) for 18 h. (B) HAECs were treated with 240 μg/mL IS for the indicated times (0, 3, 6, 9, 12, 18 h). The level of nitrite in the culture medium was measured by the Griess assay. (C) HAECs were pretreated with IS (15, 30, 60, 120, 240 μg/mL) for 1 h and then with valsartan (10 μM) for 18 h. (D) HAEC proliferation or (E) migration was determined by BrdU incorporation assay or wound healing migration assay, respectively. (F) HAECs were cultured in precoated ECL Cell Attachment Matrix with the indicated treatment agents. Tube formation was visualized; the bar graphs indicate the fold of branch points in 5 randomly selected microscopy views. (G–J) HAECs were pretreated with indicated concentrations of IS for 1 h in the presence of human serum albumin (4  g/dL) and then with valsartan (10 μM) for 18 h. (G) Nitrite level in the culture medium, (H) proliferation, (I) migration, and (J) tube formation were determined. Values are presented as the mean ± SEM of 5 separate experiments. *P < 0.05 vs. vehicle-treated cells, ^#P < 0.05 vs. valsartan-treated cells.

Fig. 2

Effects of IS on eNOS phosphorylation at various sites in HAECs. HAECs were treated with IS (240 μg/mL) for the indicated times (0, 0.5, 1, 2, 4, 8, 12, 18 h). (A–E) Cellular lysates were subjected to Western blot analysis to evaluate the phosphorylation of eNOS at Thr495, Ser615, Ser633, Ser1177 or eNOS. Data are expressed as the mean ± SEM. *P < 0.05 vs. vehicle-treated cells.

Fig. 3

Effect of IS on PKC phosphorylation, valsartan-mediated NO production and tube formation in HAECs. (A) HAECs were treated with IS (240 μg/mL) for the indicated times (0, 3, 6, 9, 12, 18 h). (B) Cells were preincubated with the PKC inhibitor calphostin C (1 μM) for 2 h and then with IS (240 μg/mL) for 18 h. Cellular lysates were subjected to Western blot analysis to evaluate the phosphorylation of PKCα^Ser657 and eNOS^Thr495. HAECs were incubated with the indicated treatment agent and (C) the level of nitrite in the culture medium, (D) proliferation, (E) migration and (F) tube formation were assessed. (G) The representative images for cell migration. (H) The representative images for tube formation. Data are expressed as the mean ± SEM. *P < 0.05 vs. vehicle-treated cells. ^#P < 0.05 vs. valsartan-treated cells, ^&P < 0.05 vs. IS + valsartan-treated cells.

Fig. 4

The NADPH oxidase-ROS pathway plays a crucial role in the harmful effect of IS on valsartan-mediated NO production and tube formation in HAECs. (A–C) HAECs were treated with IS (240 μg/mL) for the indicated times. The membrane-permeable probe hydroethidine and 2′,7′-dichlorofluorescin diacetate were used to examine intracellular ROS levels. (D) HAECs were pretreated with or without the NAD(P)H oxidase inhibitor apocynin (150 μM) for 2 h and then with IS (240 μg/mL) for 60 min. Intracellular ROS levels and NADPH oxidase activities were analyzed by the intensities of the red fluorescent ethidium (ETH) product and the green fluorescent product dichlorofluorescin (DCF) and the NADP⁺/NADPH assay kit, respectively. (E) Cells were preincubated with apocynin (150 μM) for 2 h and then with IS (240 μg/mL) for 9 h or 18 h. Cellular lysates were subjected to Western blot analysis to evaluate the phosphorylation of PKCα and eNOS^Thr495. Cells were incubated with the indicated treatment agent and (F) the levels of nitrite in the culture medium and (G) tube formation was assessed. Data are expressed as the mean ± SEM. *P < 0.05 vs. vehicle-treated cells. ^#P < 0.05 vs. IS-treated cells, ^&P < 0.05 vs. IS + valsartan-treated cells.

Fig. 5

IS impairs valsartan-induced angiogenesis in vivo. Eight-week-old male wild-type mice were subcutaneously injected with Matrigel plugs containing the indicated treatment agents. At 7 days post administration, the plugs were removed and photographed, and the hemoglobin content was analyzed. Data are presented as the mean ± SD from 10 mice. *P < 0.05 vs. vehicle-treated mice. ^#P < 0.05 vs. valsartan-treated mice, ^&P < 0.05 vs. valsartan + IS-treated mice.

Fig. 6

Blood perfusion in the ischemic limbs of mice with subtotal nephrectomy after hindlimb ischemic surgery. (A) Protocol illustration of the effect of IS on neovascularization in SNx mice. Uremia was induced in 8-week-old male C57BL/6 mice using two-step subtotal nephrectomy (SNx). One week after subtotal nephrectomy, six SNx mouse groups receiving vehicle or medications in the following six weeks were created as follows: mice were administered with vehicle, mice were administered with valsartan (60 mg/kg/day in drinking water), mice were administered with IS intraperitoneally (IP, 100 mg/kg thrice a week), mice were administered with IS + valsartan, mice were administered with IS + valsartan + IP calphostin C (10 μg/kg thrice a week), mice were administered with IS + valsartan + IP apocynin (50 mg/kg thrice a week). (B) Blood perfusion in the ischemic hindlimb was measured before, immediately after, and at 1 and 2 weeks after hindlimb ischemic (HI) surgery by laser Doppler. (C) Quantification and (D) representative images of blood perfusion in the ischemic hindlimb was measured at 2 weeks after HI surgery by laser Doppler. Data are presented as the mean ± SEM from 8 mice. *P < 0.05 vs. vehicle-treated mice. ^#P < 0.05 vs. valsartan-treated mice, ^&P < 0.05 vs. IS + valsartan-treated mice.

Fig. 7

IS impairs valsartan-induced neovascularization in the ischemic limbs of SNx mice after HI surgery. (A) Circulating Sca-1⁺/Flk-1⁺ cells were analyzed by flow cytometry at baseline and at 2 days after HI surgery. (B) Representative images and quantification of CD31 positive capillaries in the ischemic muscles. Scale bar = 50 μm. (C) Verification and quantification of phosphor-PKCα^Ser657 (p-PKCα^Ser657), phosphor-eNOS^Thr497 (p-eNOS^Thr497), PKCα and eNOS in the ischemic hindlimbs by Western blot analysis. (D) Schematic illustration of the proposed mechanism by which IS inhibits valsartan-induced neovascularization via the NADPH oxidase (NOX) and PKC signaling pathways in SNx mice. The results are presented as the means ± SEM. *P < 0.05 vs. vehicle-treated mice. ^#P < 0.05 vs. valsartan-treated mice, ^&P < 0.05 vs. IS + valsartan-treated mice.