Targeting HO-1 by Epigallocatechin-3-Gallate Reduces Contrast-Induced Renal Injury via Anti-Oxidative Stress and Anti-Inflammation Pathways

Both oxidative stress and inflammation are involved in the pathogenesis of contrast-induced nephropathy (CIN). Epigallocatechin-3-gallate (EGCG), a purified catechin from green tea, has antioxidant and anti-inflammatory effects. However, it is unknown whether or not EGCG is effective in treating CIN. Our present study found that intravenous administration of EGCG, either before or just after the establishment of CIN, had a protective effect, determined by normalization of serum creatinine and blood urea nitrogen levels, improvement in renal histopathological scoring and alleviation of apoptosis, accompanied by decreased oxidative stress and inflammation. Because EGCG is a potent inducer of the antioxidant heme oxygenase-1 (HO-1), we studied HO-1 signaling in CIN. HO-1 levels were increased in CIN; treatment with EGCG further increased HO-1 levels, accompanied by an increase in Nrf2, a regulator of antioxidant proteins. Interestingly, blockade of HO-1 with protoporphyrin IX zinc(II) (ZnPP) prevented the protective effect of EGCG on CIN. ZnPP also blocked the ability of EGCG to increase the activity of an antioxidant (superoxide dismutase), and decrease markers of oxidative stress (myeloperoxidase and malondialdehyde) and inflammation (myeloperoxidase and IL-1β), indicating that HO-1 is the upstream molecule that regulates the EGCG-mediated protection. To determine further the role of HO-1 on the EGCG-mediated inhibition of inflammation, we studied the effect of EGCG on the NLRP3 inflammasome, an upstream signaling of IL-1β. EGCG down-regulated NLRP3 expression, which was blocked by ZnPP, indicating that HO-1 links EGCG with NLRP3. Therefore, EGCG, via up-regulation of HO-1, protects against CIN by amelioration of oxidative stress and inflammation.


Introduction
Contrast-induced nephropathy (CIN) continues to be a common iatrogenic cause of acute kidney injury (AKI) after exposure to iodinated contrast medium (CM), e.g., during percutaneous coronary intervention (PCI), despite the tailored preventive strategies that include risk stratification for the individual patient, hydration, newer and safer CM and additional preventive methods (e.g., N-acetylcysteine [1], sodium bicarbonate [2], fenoldopam [3], statins [4,5], limb ischemic preconditioning [6], and preemptive hemodialysis [7]). Patients susceptible to CIN have comorbidities such as diabetes, chronic kidney disease, heart failure, and advanced age. The incidence of CIN in these patients can reach 20-30% [8,9], increasing the potential for the development of long-term loss of renal function [9,10]. To date, there is still lack of evidenceproved prevention or treatment of CIN.
Although the precise mechanisms that cause CIN are not fully understood, there seems to be a consensus that acute ischemia/hypoxia caused by CM or its toxicity per se, leads to acute tubular necrosis. Oxidative stress and inflammation have been implicated in the pathogenesis of CIN [8,9]. CM filtered by the glomerulus, actively taken up by renal tubular cells and retained within the cells and peritubular space, especially in patients with chronic kidney impairment, not only has a direct toxic action on tubular cells, increasing oxygen consumption, but also induces vasoconstriction of the vasa recta, decreasing oxygen delivery, inducing a state of hypoxia. CM triggers a series of reactions that lead to the release of free radicals, causing cellular damage and initiating the vicious cycle of oxidative stress and inflammation. Thus, a possible treatment strategy could involve the use of medications that target the regulators of both renal oxydative stress and inflammation.

Animal care
Male Sprague-Dawley (SD) rats, weighing 220-250g, were used for the experiments. The SD rats had free access to tap water and standard rat chow on a 12-h-light/dark cycle. The experimental protocol was approved by the Animal Care and Use Committee of Third Military Medical University (Permit Number: 2013-12). All surgery was performed under sodium pentobarbital anesthesia and all efforts were made to minimize suffering. histone H3 polyclonal antibody, and rabbit anti-nucleotide-binding oligomerization domain receptor (NOD-like receptor, NLR) subset 3 (NLRP3) polyclonal antibody were purchased from Proteintech Group Inc. (Chicago, IL), rabbit anti-GAPDH from Goodhere Technology (Hangzhou, China), and goat anti-rabbit IRDye 800 CW from LI-COR Biosciences (Lincoln, NE). Sodium citrate-EDTA antigen retrieval solution, immunostaining blocking solution and Cy3-labeled goat anti-rabbit IgG were purchased from Beyotime (Shanghai, China).

Establishment of rat CIN model
The rat CIN model was established as previously reported, with minor modifications [24,25]. The rats were deprived of water 24h before the acute insult. After the rats were anesthetized with an intraperitoneal injection of sodium pentobarbital (50mg/kg body wt), and placed on a heating table to maintain body temperature at 37°C, the left external jugular vein was cannulated with PE-10 tubing. In the CM group, indomethacin in ethanol (10mg/kg body wt), L-NAME in normal saline (10mg/kg body wt), and iopromide (1.8g(I)/kg body wt), were sequentially injected at 15min intervals. The vehicle group received the same amount of solvents.
EGCG (5, 10, 20mg/kg body wt) in normal saline was given intravenously at the indicated time-points. ZnPP, a HO-1 inhibitor, in normal saline (30mg/kg body wt), was given intraperitoneally 7h before EGCG and other treatments [26,27]. Then, the rats were allowed to recover at the indicated times (24-72h) and continued to have free access to water and rat chow. The rats were kept in metabolic cages for 24h urinary collections.
At the end of experiments, the rats were re-anesthetized with sodium pentobarbital, 100mg/ kg body wt. After laparotomy, all available blood was withdrawn from the abdominal aorta. Then, the kidneys were removed and rinsed twice in ice-cold phosphate-buffered saline (PBS). One longitudinal half of the left kidney was fixed in 4% (w/v) paraformaldehyde in PBS for histological assessment. The remaining half was stored at -80°C until use.

Biochemical assays
Blood sera were separated by centrifugation at 1,200g for 10min and stored at -80°C until use. Serum creatinine (Cr), blood urea nitrogen (BUN), and urinary Cr were measured by an automatic biochemistry analyzer, Analyzer Medical Systems (SaBa-18, Rome, Italy), using commercial kits (ZhongSheng BeiKong Bio-Technology and Science Inc., Beijing, China). The creatinine clearance (CrCl) was calculated according to the formula: CrCl = UV/P: U represents the urinary Cr concentration (μmol/L); V is the total urine volume collected for 24hrs (ml/min); and P is serum Cr concentration (μmol/L) [28].

Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL)
Apoptosis in paraffin-embedded kidney sections was detected using an In Situ Cell Death Detection kit (POD; Roche Applied Bio Sciences, Basel, Switzerland), as previously described [29]. Nuclei were identified with DAPI staining. Images under ×200 magnification field were obtained using a fluorescence microscope (Eclipse Ti-U, Nikon Corporation, Tokyo, Japan) at an excitation wavelength of 405nm for DAPI and 488nm for TUNEL. The number of TUNELpositive nuclei was quantified in 10 fields in renal medulla per section from three different sections with the examiner blinded from the experimental protocol. Data were expressed as the average number of TUNEL-positive nuclei per ×200 magnification field.

Assays of oxidative stress and inflammatory markers
Renal tissues were homogenized in ice-cold sucrose buffer (pH 7.4) according to the instructions of the assay kit. A marker of lipid peroxidation, malondialdehyde (MDA), and a marker of oxidative stress and inflammation, myeloperoxidase (MPO) were analyzed according to the protocols of commercial assay kits from Jiancheng Bioengineering Institute (Nanjing, China). The activity of the antioxidant, superoxide dismutase (SOD) was measured using a SOD assay kit from Dojindo Laboratories (Kumamoto, Japan). The level of the pro-inflammatory cytokine IL-1β was measured using an IL-1β enzyme-linked immunosorbent assay kit from R&D Systems (Minneapolis, MN). The values were normalized by tissue protein concentration.

Immunofluorescence microscopy
The kidney sections were deparaffinized and rehydrated. Antigen retrieval was performed by microwave heating in sodium citrate-EDTA antigen retrieval solution. After natural cooling and PBS rinses, the tissue sections were mixed with immunostaining blocking solution for 1h at room temperature to prevent nonspecific antibody binding. Then, the sections were incubated with anti-NLRP3 antibody (1:25) or anti-HO-1 antibody (1:50) at 4°C overnight. After washing with PBS for 5 min 3 times, the sections were incubated with secondary antibody, Cy3-labeled goat anti-rabbit IgG (1:200), at room temperature for 1h. Finally, after washing with PBS (5 min, 3 times), the sections were stained with DAPI before being imaged under a fluorescence microscope (Eclipse Ti-U, Nikon Corporation, Tokyo, Japan) at an excitation wavelength of 405nm for nuclei and 543nm for NLRP3 and HO-1.

Statistical analysis
All data were analyzed by SPSS 13.0 (Chicago, IL) and presented as mean ± SEM. Data were compared by one-way ANOVA with Bonferroni post-hoc test for multiple comparisons, after checking for normality (Kolmogorov-Smirnov) and homogeneity (Levene). The H&E scores among groups were compared by the nonparametric Kruskal-Wallis test. P<0.05 was considered statistically significant.

The renal protective effect of EGCG on CIN
In the present model, the levels of serum Cr and BUN were found to peak at 24h, decreasing at 48h, and almost receding back to normal at 72h. Thus, the time-point at 24h was used to evaluate renal function in this contrast-induced AKI model (Fig 1A and 1B). Intravenous pretreatment with 5 to 20mg/kg body wt of EGCG was able to reduce the extent of CIN, as assessed by renal function markers, i.e., serum Cr, BUN, and CrCl at 24h after the injury (Fig 1C-1E). The maximal effect of EGCG was found at the dosage of 10mg/kg body wt, which was chosen for subsequent experiments.
We also determined if the time-window of intervention, before or just after the contrast exposure, made any difference in the severity of CIN. We found that regardless of the time of pre-or post-EGCG treatment, a dosage of 10mg/kg body wt, significantly ameliorated the CIN-associated increase in serum Cr and BUN, as compared with vehicle treatment. There were no differences between these two intervention time-points (Fig 2A and 2B), suggesting that EGCG treatment could be effective in both pre-and post-injection of CM. The protective effects of EGCG on medullary damage and apoptosis were further confirmed by histological examination. In the CM group, CIN induced tubular vacuolar degeneration/necrosis, protein (hyaline) and cellular casts that were associated with erythrocytes and infiltration of polymorphonuclear cells in the interstitium ( Fig  2C). The major damages were in the renal outer medulla (location of medullary thick ascending limb, mTAL), rather than in the inner medulla, cortex or conjuction of cortex and medulla (location of the pars recta segment (S3) of the proximal tubule) (S1 Fig), consistent with the reports of Agmon Y [24] and Bird JE [32]. The above-mentioned abnormal changes were significantly restored after treatment with EGCG (10mg/kg body wt). The histopathologic scores of medullary damage were lower in the EGCG-treated groups than the untreated CM group (Fig 2C and 2E). The number of TUNEL-positive nuclei, reflecting the apoptosis, was substantially increased after contrast-injudced AKI, and reduced by EGCG either pre-or post-treatment (Fig 2D and 2F).

Role of HO-1 in the antioxidant and renal protective effect of EGCG on CIN
Oxidative stress is an initiator and major component in pathogenesis of CIN [8,9]. To evaluate the effect of EGCG on renal oxidative stress, we measured the renal expression of the oxidative  stress marker MDA. CIN was associated with oxidative stress because renal MDA content was increased; the increase was almost normalized by EGCG pretreatment (Fig 3A).
Though EGCG has no effect on baseline reactive oxygen species [15,16], it has been reported to increase the levels of several antioxidant enzymes, including SOD, catalase, glutathione peroxidase (GPx), and heme oxygenase-1 (HO-1) in pathological condictions [17,18,33,34]. We wondered which one is the key enzyme involved in the antioxidant effect of EGCG. We found that the CIN-induced decrease in the activity of SOD was almost normalized by EGCG pretreatment ( Fig 3B). However, HO-1, as a stress inducible antioxidant enzyme, has been reported as a powerful cytoprotective protein in several disease states, including that from renal injury [17,18,31]. Therefore, we measured HO-1 levels and found that HO-1 was increased in CIN compared with the vehicle group. While EGCG was reported to have no effect on baseline HO-1 in kidney [35], we found that EGCG treatment increased HO-1 expression to an even higher level in CIN (Fig 3C) both in the cortex and medulla (S2 Fig).
To confirm the key role of HO-1 in the protective effect of EGCG, we studied the effect of EGCG on CIN when HO-1 was inhibited by ZnPP (30mg/kg body wt). ZnPP acts as a competitive inhibitor of HO-1 [26,27]. Blocking HO-1 activity by ZnPP almost completely abrogated the renal protective effect of EGCG on CIN; it reversed the beneficial effect of EGCG on serum Cr, BUN (Fig 3D and 3E), and kidney MDA, SOD (Fig 3A and 3B), indicating that these renal protective and antioxidant effects of EGCG may via HO-1. To reconfirm the role of HO-1 in EGCG's protective action, another inhibitor SnPP was used to block the activity of HO-1 [36,37]. SnPP (10mg/kg body wt i.p. 2h before EGCG pretreatment) also significantly offsetted the reno-protective effect of EGCG, assessed by serum Cr and BUN (S3 Fig). Although the exact mechanism by which EGCG upregulates HO-1 expression is not clear, Nrf2, a transcription factor of antioxidant genes with antioxidant response element (ARE), has been found as the regulator of HO-1. Nrf2/HO-1 signaling has also been implicated in the protective effect of EGCG on renal injury [17,18]. Indeed, we found that EGCG increased the protein expression of Nrf2 in CIN (Fig 3F), indicating EGCG may increase the expression of HO-1 at the level of transcription.

Role of HO-1 in the anti-inflammatory effect of EGCG on CIN
In addition to oxidative stress, inflammation also plays an important role in the CM-induced renal damage [8,9]. Our present study found that signs of inflammation in CIN were increased, as indicated by increased MPO (marker of both oxidative stress and inflammation) and IL-1β (marker of inflammation) levels (Fig 4A and 4B). EGCG treatment reduced the increased MPO and IL-1β levels in CIN (Fig 4A and 4B). To determine the effect of EGCG on inflammatory signaling, we measured the expression of NLRP3 inflammasome, which is upstream of IL-1β. We found that NLRP3 expression that was increased in CIN was reduced by EGCG treatment (Fig 4C and 4D), indicating that EGCG, via NLRP3, regulated the IL-1β levels.
As indicated in the above-mentioned results, amelioration of oxidative stress and inflammation is involved in the protective effect of EGCG. Although the lower level of inflammation induced by EGCG may be subsequent to its antioxidant action, we wondered if there is a molecular connection between actions of EGCG on reactive oxygen species production and inflammation. To test the hypothesis that HO-1 plays a pivotal role in the EGCG-mediated negative regulation of inflammation, we studied the effect of ZnPP, a HO-1 inhibitor. We EGCG (10mg/kg body wt) was intravenously infused 15min before the establishment of CIN. The HO-1 inhibitor ZnPP (30mg/kg body wt) was injected intraperitoneally 7h before EGCG pretreatment. The rats were sacrificed at 24h after the establishment of CIN. Renal MDA level (A) and SOD activity (B) were measured. Renal HO-1 protein detected by immunoblotting was found that ZnPP prevented the ability of EGCG to reduce the increased activity and content of the inflammatory markers, MPO and IL-1β in CIN (Fig 4A and 4B). Moreover, the ability of ZnPP to decrease NLRP3 protein expression was evident by immunofluorescence microscopy and immunoblotting (Fig 4C and 4D). Taken together, these results indicate that HO-1 is the key target of EGCG in reducing both the oxidative stress and inflammation in CIN.

Discussion
CIN is the third leading cause of AKI, accounting for 10-13% of cases in hospitalized patients [38]. The renal tubular damage in CIN is caused by a decrease in renal blood flow and direct cytotoxic effect. It is hard to establish a contrast-induced AKI in a normal rat without predispositions. Homeostasis of medullary oxygenation may depend on prostanoids and nitric oxide. Impaired endothelium-derived vasorelaxation in diabetes mellitus, hypertension, atherosclerosis and heart failure, results in regional hypoxia. Indomethacin, a cyclooxygenase inhibitor, and L-NAME, a nitric oxide synthase inhibitor in this model were used as predispositions to decrease the production of prostanoids and nitric oxide respectively for later contrast-induced AKI, which was thoroughly studied by Agmon Y [24]. This method has been widely used in experimental studies of CIN [26,30,39]. As low-osmotic CM, rather than a high osmotic CM, is more commonly used in clinic, therefore, iopromide, a low osmotic CM, was chosen to establish the model of CIN.
Previous studies have shown a beneficial effect of green tea on dextran sulfate sodium-or cisplatin-induced nephropathy [40,41]. However, there are no reports about the effect of EGCG in CIN. In the present study, we demonstrated that the intravenous administration of EGCG, either before or just after the establishment of CIN, had a protective effect, assessed by measurements of serum Cr and BUN, H&E histopathological scoring and apoptosis.
Oxidative stress and inflammation are two major factors involved in the pathogenesis of CIN. Oxidative stress can initiate the tubular injury and induce inflammation that in turn causes oxidative stress, resulting in a vicious cycle, augmenting the sterile damage. Sterile inflammation is a reaction of the immune system in response to tissue injury that is essential for clearance of cell debris and tissue repair. However, uncontrolled excessive and/or prolonged activation of inflammation causes tissue damage, and contributes to the pathogenesis of AKI that eventually leads into chronic kidney disease [42][43][44]. In our study, EGCG significantly reduced both oxidative stress and inflammation in the kidney, attested by the reduction in MDA level, increase in SOD activity, and reduction in MPO activity and IL-1β level.
Interstitial congestion and hemoglobin oxidation are typical of acute tubular necrosis caused by CM. Heme (iron(II)-protoporphyrin IX), released from hemoglobin following hemolysis, possesses pro-inflammatory and pro-oxidative properties. It is a hallmark of extensive tissue damage, playing a central role in the pathogenesis of malaria, sepsis, sickle cell disease [33,45,46], and AKI in the elderly [47]. Endogenous molecules from damaged cells are essential in the auto-inflammatory response [33,43,44]. Innate immune receptors, which are pattern recognition receptors (PRRs), including the widely distributed transmembrane Toll-like receptors (TLRs), provide a sensing network for endogenous ligands [48], like heme and its derivatives [33,45,[49][50][51][52]. Thus, heme may be an important risk factor in escalating the oxidation and inflammation of CIN. Interestingly, catabolism of heme by HO converts the cytotoxic heme into cytoprotective catabolites, including iron, biliverdin, and carbon monoxide, that have  antioxidant and anti-inflammatory properties. HO-1, an inducible isozyme, is activated by heme, oxidants, cytokines, glycated albumin, and other stressors, and may be part of the protective response in many diseases [12,32], including acute lung injury [53], lipopolysaccharideinduced acute liver failure [54], neurodegenerative disorders [55], and AKI [16][17][18]. In the present study, we found that the protective effect of EGCG on CIN is mediated by HO-1, because blockade of HO-1 activity abolished the protective effect of EGCG.
Although inhibition of PI3k/Akt and MAPK signaling pathways by EGCG has been reported involved in the anti-inflammatory mechanism in ameliorating crescentic glomerulonephritis [56], the NLRs, which are cystolic PRRs, and NLR-activated inflammasome, typically NLRP3, aroused our interest. NLRP3 inflammasone is a central component of innate immunity and the sterile inflammatory response that acts as a guardian, linking damage sensing to the initiation and amplification of the inflammatory response [48,57]. When triggered by bacterial toxins, and environmental or intracellular danger signals, NLRP3 oligomerizes and recruits ASC (apoptosis speck-like protein containing a caspase-recruitment domain). ASC interacts with pro-caspase-1 and induces auto-cleavage to form a platform, i.e., inflammasome, that stimulates maturation and secretion of IL-1β and IL-18, leads to pyroptosis, a type of programmed cell death [45]. The importance of NLRP3 inflammasome has been implied in many diseases of the heart [57], lung [58,59], liver [54], and kidney [50,[60][61][62]. EGCG has been reported to inhibit NLRP3 inflammasome, and subsequently IL-1β expression in human umbilical vein endothelial cells exposed to palmitate [63], and lupus nephritis [22]. Heme, the substrate of HO-1, can also activate the NLRP3 inflammasome [31,45]. Therefore we studied the effect of EGCG on the NLRP3 inflammasome, which promotes the maturation of IL-1β. EGCG down-regulated NLRP3 expression, which was blocked by ZnPP, indicating that HO-1 links EGCG with NLRP3, and thereby, decreasing the expression of IL-1β.
In conclusion, the peri-operative intravenous administration of EGCG could protect against CIN. HO-1 is key to the EGCG-mediated protection because blockade of HO-1 abolished the down-regulation of reactive oxygen species production and inflammation caused by EGCG (Fig 5).