Protective Effect of Artemisia argyi and Its Flavonoid Constituents against Contrast-Induced Cytotoxicity by Iodixanol in LLC-PK1 Cells

Preventive effects and corresponding molecular mechanisms of mugwort (Artemisia argyi) extract and its flavonoid constituents on contrast-induced nephrotoxicity were explored in the present study. We treated cultured LLC-PK1 cells with iodixanol to induce contrast-induced nephrotoxicity, and found that A. argyi extracts ameliorated the reduction in cellular viability following iodixanol treatment. The anti-apoptotic effect of A. argyi extracts on contrast-induced nephrotoxicity was mediated by the inhibition of mitogen-activated protein kinase (MAPK) phosphorylation and the activation of caspases. The flavonoid compounds isolated from A. argyi improved the viability of iodixanol-treated cells against contrast-induced nephrotoxicity. Seven compounds (1, 2, 3, 15, 16, 18, and 19) from 19 flavonoids exerted a significant protective effect. Based on the in silico oral-bioavailability and drug-likeness assessment, which evaluate the drug potential of these compounds, compound 2 (artemetin) showed the highest oral bioavailability (49.55%) and drug-likeness (0.48) values. We further investigated the compound–target–disease network of compound 2, and proliferator-activated receptor gamma (PPAR-γ) emerged as a predicted key marker for the treatment of contrast-induced nephrotoxicity. Consequently, compound 2 was the preferred candidate, and its protective effect was mediated by inhibiting the contrast-induced inflammatory response through activation of PPAR-γ and inhibition of MAPK phosphorylation and activation of caspases.


Introduction
Artemisia is a large genus of herbal plants that are commonly used as functional foods and herbal medicine for the treatment of many diseases, including cancer and inflammation, in Korea, China, and Japan [1][2][3]. Artemisia extracts and their active compounds show antioxidant and anti-inflammatory activities against stress-related mucosal damage, HCl/EtOH-induced gastric damage, Helicobacter pylori-induced gastric damage, microglial neurotoxicity, and pancreatic and In our previous study, we reported that Artemisia extracts and the flavonoid constituent eupatilin could prevent cisplatin-induced nephrotoxicity in porcine renal proximal tubular LLC-PK1 cells, and explored the underlying molecular mechanisms of their protective actions [26]. A. argyi extracts, which contain eupatilin, may be effective for the treatment of cisplatin-induced nephrotoxicity. Eupatilin is also known to prevent acute ischemia-induced kidney injury in mice through the upregulation of the expression of Hsp70 protein, which has antioxidant and cytoprotective properties against the oxidative stress that is caused by renal tubular cell apoptosis in acute kidney injury [27,28].

Effect of Artemisia argyi Extract on the Expression of Apoptosis-Related Proteins in LLC-PK1 Cells Exposed to Iodixanol
Owing to these observed protective effects, we also investigated the effect of A. argyi extract and NAC on MAPKs and caspase-8, -9, and -3 protein expression, which are representative markers of renal damage [25,29], against iodixanol-induced nephrotoxicity in LLC-PK1 cells. Although the clear mechanisms of contrast-induced nephrotoxicity are not well identified, several pathways such as ROS-regulated signaling and apoptotic signaling have been implicated [17,25,[30][31][32]. Caspase-1, -3, -8, and -9, Bcl-2, and Bax are known to play a pivotal role in contrast-induced kidney injury as apoptotic signaling pathways [25,29]. In the present study, we observed that the activation of caspase-8 and -3 may contribute to apoptotic renal tubular injury from treatment with contrast agents. However, the activation of caspase-8 and -3 by contrast media was abrogated after co-treatment with the A. argyi extract or NAC ( Figure 3).
MAP kinases, apoptosis, and inflammatory mediators are known to play important roles in the cellular response to cytokines and external stress signals [33]. In our present study, we observed a marked increase in the phosphorylation of c-Jun-N-terminal kinase (JNK), extracellular-signal-regulated kinase (ERK), and p38 MAP kinases, which are known to be involved in the intracellular signaling associated with cell survival/proliferation in contrast-induced nephrotoxicity. The elevated phosphorylated levels of JNK, ERK, and MAP kinase returned to basal levels within 3 h of treatment with 50 µg/mL A. argyi extract. However, in comparison with treatment with 10 mM NAC, there was no change in the p38 MAP kinase levels. Therefore, although A. argyi and NAC exert similar antioxidant effects, based on the analysis of three representative MAPKs, their mechanisms of action against contrast-induced nephrotoxicity may differ ( Figure 3). signaling associated with cell survival/proliferation in contrast-induced nephrotoxicity. The elevated phosphorylated levels of JNK, ERK, and MAP kinase returned to basal levels within 3 h of treatment with 50 μg/mL A. argyi extract. However, in comparison with treatment with 10 mM NAC, there was no change in the p38 MAP kinase levels. Therefore, although A. argyi and NAC exert similar antioxidant effects, based on the analysis of three representative MAPKs, their mechanisms of action against contrast-induced nephrotoxicity may differ ( Figure 3).

DPPH-Radical-Scavenging Effects of Flavonoid Compounds Isolated from Artemisia argyi Extracts
We attempted to isolate various flavonoids, including eupatilin, from the A. argyi extract for the identification of active renoprotective compounds against contrast-induced nephrotoxicity. Nineteen previously reported flavonoids were isolated and identified by the comparison of their spectroscopic

DPPH-Radical-Scavenging Effects of Flavonoid Compounds Isolated from Artemisia argyi Extracts
We attempted to isolate various flavonoids, including eupatilin, from the A. argyi extract for the identification of active renoprotective compounds against contrast-induced nephrotoxicity. Nineteen previously reported flavonoids were isolated and identified by the comparison of their spectroscopic data with published literature data (see Supplementary Materials). For the flavonoids isolated from A. argyi, the free radical-scavenging activity against DPPH was determined spectrophotometrically, and we reported the IC 50 value, which is the sample concentration at which 50% of the DPPH radicals were scavenged (Table 1). Five compounds (6, 15, 16, 18, and 19) from nineteen flavonoids in the A. argyi extract showed free radical-scavenging activity. Among them, the effects of compound 6 (eupafolin) and compound 15 (5,7,3 ,4 -tetrahydroxyflavone) revealed significant donation of electrons to the stable free radical DPPH at a similar level to ascorbic acid (positive control). The excellent antioxidant, anti-inflammatory, and antiproliferative effects of these compounds have been previously reported [34][35][36].

Comparison of the Protective Effects of the Flavonoid Compounds Isolated from Artemisia argyi Extracts against Iodixanol-Induced Nephrotoxicity in LLC-PK1 Cells
To investigate the relationship between the structural characteristics of flavonoids and the renoprotective effect on contrast-induced nephrotoxicity in porcine renal proximal tubular LLC-PK1 cells, we treated the cells with various concentrations of flavonoids ( Figure 4A-G). Seven compounds (1, 2, 3, 15, 16, 18, and 19) from nineteen flavonoids exerted a protective effect. The reduction in cell viability by iodixanol was recovered by more than 80% by co-treatment with all seven compounds at 100 µM. These results suggest that the antioxidant activity may also partly contribute to the protective effect on contrast-induced nephrotoxicity.   (19). * p < 0.05 compared to the not-treated value.

DL and OB Evaluation of Flavonoid Compounds with Profound Protective Effects against Iodixanol-Induced Nephrotoxicity in LLC-PK1 Cells
We further investigated the drug potential of the four most efficacious compounds (2, 3, 18, and 19) to propose the most reliable candidates for the treatment of contrast-induced nephrotoxicity. The properties of oral-bioavailability (OB) and drug-likeness (DL), which are predicted values using in silico models, were evaluated, and all the four compounds showed feasible properties given the  (19). * p < 0.05 compared to the not-treated value.

DL and OB Evaluation of Flavonoid Compounds with Profound Protective Effects against Iodixanol-Induced Nephrotoxicity in LLC-PK1 Cells
We further investigated the drug potential of the four most efficacious compounds (2, 3, 18, and 19) to propose the most reliable candidates for the treatment of contrast-induced nephrotoxicity. The properties of oral-bioavailability (OB) and drug-likeness (DL), which are predicted values using in silico models, were evaluated, and all the four compounds showed feasible properties given the threshold values of OB ≥ 30% and DL ≥ 0.18 (Table 2). In particular, compound 2 had the highest OB (49.55%) and DL values (0.48), which made it the preferred candidate. Furthermore, we conducted in silico network pharmacological analysis to investigate the possible mechanism of compound 2 on nephrotoxicity. The compound-target-disease network of compound 2 was constructed using TCMSP (see Materials and Methods). We found that 26 targets were linked with compound 2 in the constructed network. Among them, we focused on peroxisome proliferator-activated receptor gamma (PPAR-γ), which was related to the disease node "inflammation" ( Figure 5). The nuclear receptor PPAR-γ regulates transcription factors that involved in lipid metabolism, fatty acid metabolism, glucose homeostasis, cell proliferation, inflammation, and related metabolic disorders [37][38][39]. In the kidney, various renal cell types including the proximal tubules and medullary collecting duct cells have endogenous PPAR-γ expression and activity. Previous studies have shown that the suppression of PPAR-γ involved in p53 and Bax interaction in renal tubular cell apoptosis [40] resulted in renal injury associated with ischemia/reperfusion in rats. In addition, the later suppression of PPAR-γ may result in hyperuricemia patients with chronic renal injury [41]. Therefore, expression of PPAR-γ in the kidney may play an important role in maintaining normal renal function [42].
tubules and medullary collecting duct cells have endogenous PPAR-γ expression and activity. Previous studies have shown that the suppression of PPAR-γ involved in p53 and Bax interaction in renal tubular cell apoptosis [40] resulted in renal injury associated with ischemia/reperfusion in rats. In addition, the later suppression of PPAR-γ may result in hyperuricemia patients with chronic renal injury [41]. Therefore, expression of PPAR-γ in the kidney may play an important role in maintaining normal renal function [42].  Consequently, we focused on the identification of the protective mechanism of compound 2 on contrast-induced LLC-PK1 cell damage including PPAR-γ as a predicted key marker of the compound-target-disease network of compound 2. As shown in Figure 6A, exposure to contrast media induced morphological changes in LLC-PK1 cells, which appeared blebbed and shrunk, as detected by phase-contrast inverted microscopy. These abnormal morphological changes were recovered in cells exposed to 50 and 100 µM compound 2 ( Figure 6A). In addition, the nuclear condensation induced by iodixanol, as detected spectrofluorometrically, was significantly ameliorated after treatments of cells with compound 2 at concentrations of 50 and 100 µM ( Figure 6A). In addition, we observed that the suppression of PPAR-γ may contribute to apoptotic renal tubular injury from treatment with contrast agents. However, this suppression of PPAR-γ by contrast media returned to basal levels upon treatment of cells with compound 2 ( Figure 6B). recovered in cells exposed to 50 and 100 μM compound 2 ( Figure 6A). In addition, the nuclear condensation induced by iodixanol, as detected spectrofluorometrically, was significantly ameliorated after treatments of cells with compound 2 at concentrations of 50 and 100 μM ( Figure  6A). In addition, we observed that the suppression of PPAR-γ may contribute to apoptotic renal tubular injury from treatment with contrast agents. However, this suppression of PPAR-γ by contrast media returned to basal levels upon treatment of cells with compound 2 ( Figure 6B). Based on the possible mechanisms of PPAR-γ in the iodixanol-induced cytotoxicity of LLC-PK1 cells, we assessed the protective effect of PPAR-γ agonist rosiglitazone ( Figure 6C) on iodixanol-induced cytotoxicity. After treatment with 25 mg/mL iodixanol, the LLC-PK1 cell viability was reduced to 58.7% ( Figure 6C). However, this reduced cell viability by iodixanol was recovered to 83.4% by co-treatment with 1 µM rosiglitazone. In addition, we assessed the effect of PPAR-γ selective antagonist GW9662 on the protection effect of artemetin on iodixanol-induced cytotoxicity. GW9662 and artemetin used at concentrations that had no considerable effect on cell viability. The reduction in cell viability by iodixanol was recovered by 89.0% by co-treatment with 100 µM artemetin, and this protective effect was reduced by 71.3% after co-treatment with 1 µM GW9662 ( Figure 6D). Taken together, our data suggest that artemetin exerts its protection effect through PPAR-γ in LLC-PK1 cells.

Effects of Artemetin on Iodixanol-Induced Apoptosis in LLC-PK1 Cells
We further explored whether compound 2 could decrease apoptosis in LLC-PK1 cells exposed to contrast by using the annexin V Alexa Fluor 488 and propidium iodide staining ( Figure 7A). Apoptotic cell death, observed via staining with annexin V, increased from 4.6 ± 0.5% to 53.0 ± 2.0% after 25 mg/mL contrast treatment, whereas a decrease by 14.0 ± 1.0% and 11.0 ± 1.7% was observed after treatment with 50 and 100 µM of compound 2, respectively, as shown in the quantified graph for the percentage of apoptotic cells ( Figure 7B). Based on the results of protein analysis associated with inflammation and apoptosis via Western blot, the increased phosphorylation of JNK and ERK MAPKs by iodixanol returned to basal levels with treatment of compound 2. Furthermore, the activation of caspase-8 and -3 by iodixanol was abrogated after treatment with the compound 2 ( Figure 7C). Therefore, the protective effect of compound 2 on iodixanol-induced apoptosis in LLC-PK1 cells was mediated by the inhibition of MAPK phosphorylation and caspase activation. selective antagonist GW9662 on the protection effect of artemetin on iodixanol-induced cytotoxicity. GW9662 and artemetin used at concentrations that had no considerable effect on cell viability. The reduction in cell viability by iodixanol was recovered by 89.0% by co-treatment with 100 μM artemetin, and this protective effect was reduced by 71.3% after co-treatment with 1 μM GW9662 ( Figure 6D). Taken together, our data suggest that artemetin exerts its protection effect through PPAR-γ in LLC-PK1 cells.

Effects of Artemetin on Iodixanol-Induced Apoptosis in LLC-PK1 Cells
We further explored whether compound 2 could decrease apoptosis in LLC-PK1 cells exposed to contrast by using the annexin V Alexa Fluor 488 and propidium iodide staining ( Figure 7A). Apoptotic cell death, observed via staining with annexin V, increased from 4.6 ± 0.5% to 53.0 ± 2.0% after 25 mg/mL contrast treatment, whereas a decrease by 14.0 ± 1.0% and 11.0 ± 1.7% was observed after treatment with 50 and 100 μM of compound 2, respectively, as shown in the quantified graph for the percentage of apoptotic cells ( Figure 7B). Based on the results of protein analysis associated with inflammation and apoptosis via Western blot, the increased phosphorylation of JNK and ERK MAPKs by iodixanol returned to basal levels with treatment of compound 2. Furthermore, the activation of caspase-8 and -3 by iodixanol was abrogated after treatment with the compound 2 ( Figure 7C). Therefore, the protective effect of compound 2 on iodixanol-induced apoptosis in LLC-PK1 cells was mediated by the inhibition of MAPK phosphorylation and caspase activation.

DPPH Radical-Scavenging Assay
The DPPH radical-scavenging assay analyzes antioxidant behavior based on an electron-transfer reaction. The change in color from purple to yellow is observed as DPPH reacts with any antioxidant compounds present. The reaction mixture consisted of a solution of antioxidant compounds (100 µL) and an equal volume of DPPH solution (0.1 mM) in ethanol, and was incubated for 30 min at room temperature in the dark. Ascorbic acid was used as the positive control. DPPH reaction was monitored by measuring the absorbance (Abs.) at 517 nm by using a microplate reader (PowerWave XS; Bio-Tek Instruments, Winooski, VT, USA). The radical-scavenging activity is presented as an IC 50 value, which is the concentration of the antioxidant compounds required to inhibit 50% of the free radicals, and was calculated using the following equation: Scavenging activity (%) = [(Control Abs. − Sample Abs.)/Control Abs.] × 100.

Cell Culture
LLC-PK1 (pig kidney epithelium, CL-101) cells were purchased from American Type Culture Collection (Rockville, MD, USA), and were cultured in DMEM (Cellgro, Manassas, VA, USA) supplemented with 10% FBS, 1% penicillin/streptomycin (Invitrogen Co., Grand Island, NY, USA), and 4 mM L-glutamine in an atmosphere of 5% CO 2 at 37 • C. The cells were routinely passaged before they reached 80% confluence, and the culture medium was replaced with fresh medium every 2 days.

Renoprotective Effect against Iodixanol-Induced Damage in Kidney Cells
The experiment was performed according to the method reported previously in our study about renoprotective effects against iodixanol-induced damage in kidney cells [62]. Briefly, the cells were seeded onto 96-well culture plates at 1 × 10 4 cells/well and allowed to adhere for 24 h. Cells were treated with the vehicle control (0.5% DMSO), positive control (10 mM NAC), indicated concentrations of A. argyi extract, and its flavonoid constituents. After incubation for 2 h, 25 mg/mL iodixanol was added to each well and incubated for a further 3 h. Cell viability was determined using the Ez-Cytox cell viability detection kit in accordance with the manufacturer's instructions, and evaluated by measuring the absorbance at 450 nm using a microplate reader (PowerWave XS; Bio-Tek Instruments, Winooski, VT, USA).

Effect of PPAR-γ Ligand (Rosiglitazone) and PPAR-γ Antagonist (GW9662) against Iodixanol-Induced Damage in Kidney Cells
To identify the effect of rosiglitazone on iodixanol-induced cytotoxicity, the cells were seeded onto 96-well culture plates at 1 × 10 4 cells/well and allowed to adhere for 24 h. Cells were treated with the vehicle control (0.5% DMSO), iodixanol (25 mg/mL) or rosiglitazone (1 µM) alone and also treated with combinations of rosiglitazone (1 µM) and iodixanol (25 mg/mL). To verify the effect of GW9662 in the presence of the artemetin, the cells were seeded onto 96-well culture plates at 1 × 10 4 cells/well and allowed to adhere for 24 h. Cells were treated with the vehicle control (0.5% DMSO), iodixanol (25 mg/mL), GW9662 (1 µM) or artemetin (100 µM) alone and also treated with combinations of iodixanol (25 mg/mL) and artemetin (100 µM) or combinations of iodixanol (25 mg/mL), artemetin (100 µM) and GW9662 (1 µM). Cell viability was determined using the Ez-Cytox cell viability detection kit in accordance with the manufacturer's instructions, and evaluated by measuring the absorbance at 450 nm using a microplate reader.

Nuclear Staining with Hoechst 33342
Cells were seeded onto 6-well culture plates at 4 × 10 5 cells/well. allowed to adhere for 24 h, and were treated with the vehicle control (0.5% DMSO) and compound 2. After incubation for 2 h, 25 mg/mL iodixanol was added to each well. After incubation for 3 h, Hoechst 33342 solution was added to each well and incubated for a further 10 min. Stained cells were visualized by fluorescence microscopy.

Image-Based Cytometric Assay
Cells were seeded onto 6-well culture plates at 4 × 10 5 cells/well and allowed to adhere for 24 h. Cells were treated with the vehicle control (0.5% DMSO) and compound 2. After incubation for 2 h, annexin V Alexa Fluor 488 and propidium iodide were added to each well and incubated for a further 30 min in darkness. Apoptotic cells (stained green with annexin V Alexa Fluor 488), dead cells (stained red with propidium iodide and green with annexin V Alexa Fluor 488), and live cells (unstained) were visualized and counted by using a Tali image-based cytometer (Invitrogen, CA, USA) [63].

Western Blotting Analysis
LLC-PK1 cells were seeded onto 6-well plates at 4 × 10 5 cells/well and treated with the vehicle control (0.5% DMSO), A. argyi extract (50 µg/mL), compound 2, and NAC (10 mM) as the positive control. After incubation for 2 h, 25 mg/mL iodixanol was added to each well and incubated for 3 h. The cells were lysed with RIPA buffer and supplemented with 1 mM phenylmethylsulfonyl fluoride (PMSF) immediately before use. Protein concentration was determined using the Pierce™ BCA Protein Assay Kit according to the manufacturer's instructions, and bovine serum albumin (BSA) was used as the standard protein. Equal amounts (20 µg/lane) of protein sample were separated via electrophoresis in 10% sodium dodecyl sulfate-polyacrylamide gel and transferred onto PVDF transfer membranes [64]. Proteins were analyzed with epitope-specific primary antibodies to JNK, phospho-JNK, p44/42 MAP Kinase (ERK), phospho-p44/42 (pERK), p38 MAP Kinase, phospho-p38, cleaved caspase-8, cleaved caspase-3, PPAR-γ, glyceraldehyde 3-phosphate dehydrogenase (GAPDH), and horseradish peroxidase (HRP) conjugated anti-rabbit secondary antibodies. Bound antibodies were detected using ECL Advance Western Blotting Detection Reagents and visualized on a FUSION Solo Chemiluminescence System (PEQLAB Biotechnologie GmbH, Erlangen, Germany).

Oral-Bioavailability (OB) and Drug-Likeness (DL) Evaluation
OB and DL values were obtained from the Traditional Chinese Medicine Systems Pharmacology Database (TCMSP, http://tcmspnw.com). OB is one of the most important pharmacokinetic properties for drug discovery. In TCMSP, OB values are calculated using in silico screening model OBioavail 1.1 [65]. The model was constructed using 805 structurally diverse drugs and drug-like molecules, and has been proven as an effective drug-screening model in many previous studies.
The property of drug-likeness is a qualitative concept used in drug design. The DLs determined in TCMSP are estimated values using molecular descriptors based on Lipinski's rule of five [66] and Tanimoto coefficients [67], which measure the structural similarities between herbal ingredients and drugs in the Drugbank database (http://www.drugbank.ca/). In this study, the threshold values for evaluation were OB ≥ 30% and DL ≥ 0.18, which are the default suggested thresholds for TCMSP.

Network Analysis
Compound-target-disease network is a tripartite network with three types of nodes: compounds, targets, and diseases. The edges between compounds and targets are defined as compound-target interactions (1 or 0). To construct a network, compound-target interaction information was extracted from TCMSP for all pairs of candidate compounds and target proteins in the database. This included experimentally validated interactions, but most of the interactions were predicted interactions, based on the machine learning methods (support vector machine and random forest) with validated drug-target interaction datasets. The performance of these predictive methods for compound-target interactions have been proven to be reliable [46]. The information to define edges between targets and diseases was extracted from PharmGKB (http://www.pharmgkb.org) and Therapeutic Targets Database (http://bidd.nus.edu.sg/BIDD-Databases/TTD/TTD.asp).

Statistical Analysis
Statistical significance was determined by analysis of variance (ANOVA) followed by a multiple comparison test with a Bonferroni adjustment. A value of p < 0.05 was considered statistically significant. The analysis was performed using SPSS ver. 19.0 (SPSS Inc., Chicago, IL, USA).

Conclusions
A. argyi extracts mitigated the reduced viability of iodixanol-treated LLC-PK1 kidney cells. The anti-apoptotic effects of A. argyi extracts on contrast-induced nephrotoxicity was mediated by the suppression of MAPKs and activation of caspases. In addition, flavonoid compounds isolated from the A. argyi extract improved the reduced viability of the cells treated with iodixanol. Compound 2 isolated from the A. argyi extract exhibited anti-apoptotic effects mediated by the suppression of MAPKs and activation of caspases. Moreover, PPAR-γ levels following suppression by contrast media returned to basal levels after treatment with compound 2. Compound 2 (artemetin) was the preferred candidate and its protective effect was mediated by inhibition of contrast-induced inflammatory response by activating PPAR-γ and inhibiting MAPK phosphorylation and the activation of caspases. Further studies on the in vivo effects to confirm the use of A. argyi extracts and their flavonoid constituents are necessary to prove their beneficial effects in reducing the adverse effects of contrast agents in the kidney.