Dietary Inorganic Nitrate Protects Hepatic Ischemia-Reperfusion Injury Through NRF2-Mediated Antioxidative Stress

Objectives: Hepatic ischemia-reperfusion injury (HIRI) is of common occurrence during liver surgery and liver transplantation and may cause hepatic impairment, resulting in acute liver dysfunction. Nitrate plays an important physiological regulatory role in the human body. Whether dietary nitrate could prevent HIRI is, however, unknown. Methods: A HIRI mouse model was established in that the blood supply to the median lobe and left lateral lobe was blocked for 60 min through the portal vein and related structures using an atraumatic clip. Sodium nitrate (4 mM) was administrated in advance through drinking water to compare the influence of sodium nitrate and normal water on HIRI. Results: Liver necrosis and injury aggravated after HIRI. The group treated with sodium nitrate showed the lowest activities of plasma aminotransferase and lactate dehydrogenase and improved outcomes in histological investigation and TUNEL assay. Mechanistically, sodium nitrate intake increased plasma and liver nitric oxide levels, upregulated nuclear factor erythroid 2-related factor 2 (NRF2)–related molecules to reduce malondialdehyde level, and increased the activities of antioxidant enzymes to modulate hepatic oxidative stress. Conclusions: Dietary inorganic nitrate could prevent HIRI, possibly by activating the NRF2 pathway and modulating oxidative stress. Our study provides a novel therapeutic compound that could potentially prevent HIRI during liver transplantation or hepatic surgery.


INTRODUCTION
Hepatic ischemia-reperfusion injury (HIRI) is common during liver resection or transplantation, which remains a major cause of hepatic failure following hepatic surgery (Mendes-Braz et al., 2012). A series of pathological alterations are referred to liver ischemia-reperfusion (IR) injury, including initial sterile hypoxic or ischemic tissue injury, reperfusion-induced oxidative stress, inflammatory response, and microvascular dysfunction (Massip-Salcedo et al., 2007), and effective prevention or treatment methods are still lacking in clinics.
The mainstream research direction includes oxidative stress, immune response, and inflammatory response (Dar et al., 2019). Nitric oxide (NO) plays an important role in organ IR injuries (Peralta et al., 2013). Low tissue oxygen tension significantly decreased the oxygen-dependent NO synthesis of endothelial NO synthase during ischemia (Webb et al., 2004). Furthermore, overproduction of malondialdehyde (MDA) in the reperfusion phase further consumes endogenous NO (Weitzberg et al., 2010). Reduced bioavailability of NO can lead to endothelial and microvascular functional imbalance, bringing about the "noreflow phenomenon" after ischemic tissue reperfusion is initiated (Eltzschig and Eckle, 2011). Therefore, safe and effective methods to reduce oxidative stress, alleviate inflammation, and maintain NO bioavailability may become a new strategy for the prevention and treatment of hepatic IR injury.
Nitrates are quite abundant in our ordinary daily diets, especially in green leafy vegetables such as spinach, lettuce, or beetroot (Song et al., 2015). In oral microenvironment, commensal nitrate-reducing bacteria effectively degrade nitrate to nitrite, which is swallowed with 1 L saliva per day and continuously enters the circulation (Mensinga et al., 2003;Lundberg et al., 2011). Nitrate can be metabolized in vivo to form NO and other bioactive nitrogen oxides (Lundberg et al., 2008). In contrast to NO synthases, the nitrate-nitrite NO pathway is independent of oxygen and L-arginine. Therefore, the formation of NO from this source is unaffected during IR damage (Zweier et al., 1995). The protective response of nitrite on IR damage have been verified in multiple organs or tissues such as the brain (Jung et al., 2006), lungs (Sugimoto et al., 2012), and heart (Omar et al., 2016;Liu et al., 2020). However, the potential therapeutic value of inorganic nitrate in liver IR injury remains controversial.
In this study, we applied a mouse HIRI model to investigate the hepatic and systemic protection of dietary inorganic nitrate supplementation and explore the underlying mechanism of action.

Mice
This study was approved by the Institutional Animal Care and Ethics Committee at Beijing Friendship Hospital (IACUC ID: 18-2009) and performed in accordance with the Ethical Guidelines for Animal Studies.
Male C57BL/6 mice (8-week-old) weighing 20-30 g were obtained from Beijing Vital River Laboratory (Beijing, China) and housed under a specific pathogen-free, temperature-controlled, and humidity-controlled environment with free access to rodent chow and tap water at the animal facilities at Beijing Friendship Hospital (Beijing, China).

Mouse HIRI Model
A mouse HIRI model was successfully established as previously described (Jin et al., 2019). In brief, mice were anesthetized and heparinized to prevent blood clotting. A midline laparotomy incision was performed and an atraumatic clip was used to completely clamp the hepatic artery and portal vein, causing ischemia of the left lateral and median lobes of the liver. The mice were wrapped with a heating pad to maintain body temperature at 37°C. After 60 min, the atraumatic clip was removed to allow reperfusion. Then the mice were sutured. 6 h later, the liver and plasma were collected for measurement (Abe et al., 2009).

Plasma Transaminase Activities and Lactate Dehydrogenase (LDH) Enzyme Activities
Plasma alanine aminotransferase (ALT) and aspartate aminotransferase (AST) activities were determined as quantitative indices of liver injury using an alanine aminotransferase assay kit (C009-2, NJJC Bio Inc., Nanjing, China) and aspartate aminotransferase assay kit (C010-2, NJJC Bio Inc., Nanjing, China), respectively, according to the manufacturer's instructions. Plasma LDH activity was analyzed as per the method recommended for the lactate dehydrogenase assay kit (A020-2, NJJC Bio Inc., Nanjing, China).
Nitrate, Nitrite, and NO Levels in the Plasma and Liver Tissue 10,000 MW filters and diluted. The total nitric oxide and nitrate/ nitrite parameter assay kit (KGE001, R&D, MN, United States) was employed to determine the concentration of nitrate and nitrite, while a mouse Nitric Oxide (NO) ELISA Kit (EK18797, SAB, Maryland, United States) was used to measure the NO level.

Histological and Immunohistochemical Analysis
Liver samples were fixed in phosphate-buffered formalin and embedded in paraffin. Sections (5 µm) were stained with hematoxylin and eosin. The stained sections were semiquantitatively evaluated and examined in a blinded manner for each liver sample and scored from 0 to 4 to assess the condition of liver necrosis and steatosis, sinusoidal congestion, and hepatic cytoplasmic vacuolation-infiltrating polymorphonuclear leukocytes according to Suzuki Score (Suzuki et al., 1993).
For immunohistochemical analysis, neutrophils were evidenced in paraffin sections using anti-Ly6G antibody. Ly6G positive cells were calculated by Image J software (NIH, MD, United States).

Terminal Deoxynucleotidyl Transferase dUTP Nick End Labeling (TUNEL) Assay
Paraffinized samples were sectioned at 6 μm thickness, mounted on silane-coated glass slides. TUNEL staining was performed with a Click-iT ® Plus TUNEL Assay Kit (C10617, Thermo Fisher Scientific Inc., PA, United States) according to the manufacturer's instructions. Images were captured under a confocal microscope and the number of TUNEL-positive cells was calculated using Image J software (NIH, MD, United States).

Determination of Plasma Cytokine Levels
According to the manufacturer's instructions, the multiple cytokine levels of plasma were measured using the LEGENDplex ™ mouse inflammation panel (740150, BioLegend, CA, United States). Then, the prepared samples were tested using Aria II flow cytometer (BD Biosciences), and the data were analyzed by LEGEND plex software v8.0 (Biolegend, CA, United States).
Approximately 1×10 5 cells per well were cultured in a 24-well plate for 12 h. 500 μM H 2 O 2 was used to mimic the oxidative stress environment. Exogenous NO and hepatocyte NRF2 inhibition were provided by 25 μM SNP and 4 μM ML385, respectively.

Intracellular NO Levels
An intracellular NO Detected Kit (S0021S, Beyotime, Shanghai, China) was used to test the level of NO. According to the instructions, the intracellular NO level was measured and analyzed.

Flow Cytometry Analysis of ROS and Apoptosis
Cells were suspended and centrifuged. Part of the centrifuged cells were rinsed with PBS and incubated with ROS probe. Other cells were rinsed with Annexin V binding buffer and incubated with Annexin V. After the process finished, all samples were acquired on a FACS Aria II flow cytometer (BD Biosciences), and data were analyzed using FlowJo software (Tree Star, OR, United States).

Real-Time Polymerase Chain Reaction (RT-PCR)
Total RNA was extracted from the liver tissue using an Eastep Super Total RNA Extraction Kit (LS1040, Shanghai Promega, Shanghai, China) in accordance with the manufacturer's protocol and reverse-transcribed to cDNA using a PrimeScript RT Reagent Kit (RR037A, TaKaRa, Tokyo, Japan). Quantitative RT-PCR analysis was performed using an ABI 7500 Sequence Detection System (Applied Biosystems, CA, United States). The PCR mixture comprised 10 μl SYBR Green Master Mix, 0.5 μM forward and reverse primers, and 1 μl cDNA sample. After normalization of target gene expression, the data was quantified by the 2 −ΔΔCt method. The genes and primer sequences are listed in Supplementary Table S1.

Statistical Analysis
Sample collection and data analysis were performed by different members according to the double-blind principle. The data were analyzed with GraphPad Prism 8.0 software (GraphPad Software, CA, United States). Results are presented as mean ± standard error of mean (SEM). Student's t-test was used to compare difference between two groups with normal distribution or Kruskal-Wallis test for others. Among multiple groups, oneway ANOVA with post-hoc test for the normal distribution and Kruskal-Wallis test for others. *p < 0.05, **p < 0.01, and ***p < 0.001 were considered significant, and NS was not significant. Statistical analysis was performed using SPSS 20.0.

Dietary Nitrate Attenuated IR-Induced Liver Injury
The HIRI model was established as shown in Figures 1A,B. Mice were administrated with 4 mM dietary nitrate through drinking water 5 days before the operation. The water intake, food intake, and body weight of each mouse were detected (Supplementary Figure S1), while no difference was found among the groups. The total intake dosage of sodium nitrate was 78.59 ± 6.93 μg/day per kilogram of weight. As shown in Figures 1C-E, plasma ALT, AST, and LDH activities were significantly higher in HIRI group than in the control group and sham group. Mice exposed to 4 mM nitrate for 5 days prior to HIRI showed a significant decrease in the activities of ALT, AST, and LDH as compared with unpretreated mice.
To investigate other IR-related factors, we also detected the damage-associated molecular pattern molecules (DAMPs) and revealed that the mRNA level of Hmgb1 and S100b was significantly increased in HIRI group and remarkably decreased in nitrate intake group ( Figure 1F).
Consistent with the changes in plasma markers, morphological alterations such as tubular necrosis and infiltration of inflammatory cells were obvious in the HIRI liver ( Figure 1G), while nitrate pretreatment had the ability to attenuate IR-induced liver damage. The histopathological score of the liver also revealed the nitrate-mediated significant amelioration of IR-related liver injury ( Figure 1H).
TUNEL assay (Figures 1I,J) showed that apoptosis level increased after IR but significantly decreased by nitrate pretreatment. RT-PCR of liver tissues ( Figure 1K) revealed the downregulation in the expression of the anti-apoptosis genes Bcl-2 and Bcl-xL in HIRI group, and an upregulation in 4 mM nitratetreated group. The protein expression ( Figures 1L,M) also showed the similar consequences that the lowest expression level in HIRI mice and a reversion could be seen in nitrate intake mice. The expression of anti-apoptosis gene and protein was consistent with the result of TUNEL assay.

Dietary Nitrate Reduced IR-Associated Inflammatory Responses
Tumor necrosis factor (TNF)-α and interleukin (IL)-1β are considered as proinflammatory factors, while IL-10 is an antiinflammatory cytokine. As shown in Figures 2A-G, the relative mRNA levels of Tnf-α and Il-1β increased and that of Il-10 decreased in HIRI group as compared to those in control groups. In addition, the levels of Il-1α, Il-17A, Il-27, and interferon (Ifn)-c also increased in HIRI group. However, nitrate administration downregulated the expression of Tnf-α, Il-1β, Il-1α, Il-17A, Il-27, and Ifn-c in the liver tissue. Il-10 expression was the lowest in HIRI group but was significantly restored in 4 mM nitrate pretreatment group.
We also evaluated the changes in the plasma levels of several pro-inflammatory factors, including TNF-α, IL-1β, IL-10, IL-1α, IL-17A, IL-27, and IFN-c. As shown in Figures 2H-N, the plasma levels of above cytokines were altered, consistent with the results of RT-PCR. The levels of TNF-α, IL-1β, IL-1α, IL-17A, IL-27, and IFN-c were higher in HIRI group while IL-10 was lower than in controls. However, nitrate administration in HIRI + Nit group decreased TNF-α, IL-1β, IL-1α, IL-17A, IL-27, and IFN-c levels in the plasma while IL-10 increased as compared to unpretreated HIRI group.
In IR model, the augment of neutrophils could be seen remarkably. To investigate the immune and inflammatory condition, the infiltration of neutrophils in the liver was tested by immunohistochemistry. Nitrate intake markedly decreased Ly6G positive cells in liver with HIRI ( Figures 2O,P).

Dietary Nitrate Increased Nitrate, Nitrite and NO in the Plasma and Liver and Reduced IR-Associated Oxidative Stress
To explore the functional role of dietary nitrate in HIRI, we measured the concentrations of nitrate, nitrite, and NO in the plasma and liver tissues. As shown in Figures 3A,B, 5 days of 4 mM nitrate intake could significantly increase nitrate, nitrite, and NO levels both in plasma and liver of normal mice. Without nitrate administration, the nitrate levels in plasma were lower in HIRI group. The mice pretreated with 4 mM nitrate showed statistically higher levels of plasma nitrate than mice from HIRI group ( Figure 3C). The plasma nitrite levels in the liver also revealed the similar tendency although without significant difference ( Figure 3D). As a key functional molecule indicative of nitrate activity, we observed that the NO levels in the plasma of HIRI group were significantly lower than those in the liver of the control group and the sham group ( Figure 3E). However, the mice treated with nitrate showed statistically higher levels of liver NO than the mice from HIRI group. Similar changes in liver tissue could be seen in Figures 3F-H. Dietary nitrate could reverse the reduction of nitrate, nitrite, and NO in liver tissue.
The level of MDA in the liver ( Figure 3I), an index of lipid peroxidation, was significantly higher in HIRI group than that in the control and sham group, but 4 mM nitrate administration significantly alleviated the levels of MDA.
As shown in Figures 3J-L, the activities of antioxidant enzyme CAT, GSH-PX, and SOD were significantly lower in HIRI group than those in the control and sham group, possibly owing to the depletion of the antioxidant pool that was consumed to remove excess of ROS produced during HIRI. The antioxidant enzyme activities were restored in mice administered with 4 mM nitrate.

Dietary Nitrate Activated the NRF2 Pathway in HIRI
Previous reports have shown that NRF2 activation is associated with oxidative stress and exerts a strong protective effect against hepatotoxicity through increased autophagy (Bellezza et al., 2018). To explore the potential protective mechanisms of nitrate on liver, we examined the NRF2 pathway. Our findings ( Figure 4A) indicated that nitrate pretreatment efficiently increased the mRNA levels of Nrf2 and Nrf2-related genes such as quinone oxidoreductase 1 (Nqo1), heme oxygenase 1 (Ho1), thioredoxin reductase (TrxR), CAT, and Glutathione Peroxidase 1 (GPX1). Further, nitrate pretreatment downregulated the expression of NRF2 repressor protein Kelch-like ECH-associated protein 1 (Keap1). Also, relative protein expression revealed that NRF2, NQO1, and HO1 in total protein ( Figures 4B,C) reduced in HIRI group than those in control and sham group, while dietary nitrate could increase the expression significantly. Contrast to NRF2, the protein expression of KEAP1 was increased in HIRI group and decreased in HIRI + Nit group. The results of total protein expression were consistent with the mRNA levels. When NRF2 activated, parts of it should transport into the nucleus to exert function, so we also detected the expression of cytosolic and nuclear NRF2. We found that the expression had no significance in cytosolic NRF2 ( Figures 4D,E), however, remarkably alteration of NRF2 was revealed in nucleus ( Figures 4F,G). A significant increase could be observed in mice treated with nitrate.

NO Pretreatment Attenuated Oxidative Stress Induced by H 2 O 2 and Upregulated NRF2 in vitro
To validate the mechanism of NRF2-pathway, AML12 cells were treated with 500 μM H 2 O 2 to mimic the oxidative condition in vitro. SNP, a common NO donor, provided exogenous NO in the culture medium. A NRF2 inhibitor ML385 (Singh et al., 2016) was also used in this study to block NRF2-pathway in AML12 cells. As shown in Figure 5A, AML12 cells were pretreated with SNP or SNP + ML385 for 12h, and followed Frontiers in Pharmacology | www.frontiersin.org June 2021 | Volume 12 | Article 634115 by 12h-H 2 O 2 stimulation. AML12 cells treated with 4 μM ML385 alone for 24 h had no changes of morphology ( Figure 5B), ROS, and Annexin V levels ( Figure 5C). Also, AML12 cells treated with 25 μM SNP alone for 24 h had no alterations of morphology ( Figure 5D), ROS, and Annexin V levels ( Figure 5E). AML12 cells showed ( Figure 5F) shrinkage, uncleared outline and increased intercellular particles after H 2 O 2 loading. SNP pretreatment could maintain the normal morphology of AML12 cells, however, the administration with ML385 could reverse the protection of SNP on AML 12 cells. As shown in Figure 5G, although the intracellular concentrations of NO were increased in both SNP-treated cells, the protection of AML12 cells toward oxidative stress induced by H 2 O 2 could only be seen in SNP pretreated cells but not in SNP and ML385 combination treated cells (Figures 5H-J). Furthermore, the relative mRNA levels of Nrf-2, Nqo1, and associated genes were lower and Keap1 was higher in H 2 O 2 stimulated cells, SNP pretreatment restored the changes accordingly. However, blocking the NRF2 pathway in AML12 cells by ML385 could partially inhibit the alteration induced by SNP ( Figure 5K). Taken together, NO pretreatment attenuated oxidative stress induced by H 2 O 2 mainly through NRF2 pathway.

DISCUSSION
Nitrate is thought to be harmful owing to the potential production of carcinogenic nitrosamines (Bedale et al., 2016).  However, there is no clear evidence regarding dietary nitratemediated increase in the occurrence of cancer (Bryan et al., 2012); instead, studies have highlighted the potential benefits of nitrate. For years, we have performed several studies on the management of oral and systemic diseases using nitrate, and demonstrated that nitrate can effectively prevent Sjogren's syndrome (Xia et al., 2015), promote the effect of chemoradiotherapy (Chang et al., 2019), prevent gastrointestinal stress (Jin et al., 2013), facilitate weight loss (Ma et al., 2020), and alleviate aging (Wang et al., 2018). Systemic nitrates circulate in the blood, saliva, and tissues and are absorbed after a nitrate-rich diet with a peak plasma level of 15-30 min with a half-life period of about 5-8 h (Witter et al., 1979;Lundberg et al., 2011). Due to the bioavailability of nitrate in the stomach and small intestine, it is almost completely absorbed, with about 75% excreted in the urine and the rest reabsorbed by the kidneys, biliary tract, and salivary glands (Kahn et al., 1975;Fritsch et al., 1985). Normally, up to 25% recycled nitrate can be found in salivary glands, and nitrate concentrations in salivary glands are 10 times higher than in plasma (Spiegelhalder et al., 1976) . The role of nitrate in the human body mainly depends on the nitrate-nitrite-NO axis, and is mediated through NO (Witter et al., 1979). In this process, nitrate is converted to nitrite by oral and gastrointestinal bacteria. Nitrite is extremely unstable and gets converted to NO through enzymatic reactions (Bryan and Ivy, 2015).
The results of this study showed that oral intake of inorganic nitrate could effectively prevent HIRI. Previous studies have revealed the preventive effects of intravenous nitrite on HIRI; however, nitrite is unstable and inconvenient for common use (Shiva et al., 2007). Further, intravenously administrated nitrate may not perform the necessary functions (Duranski et al., 2005). The differences in results may be attributed to the different intake method adopted in our study. Oral and gastrointestinal bacteria are of the utmost importance for nitrate transfer to NO (Gaston et al., 1994). However, the access of intravenously injected nitrate to bacteria was limited, which may prevent the conversion of nitrate to nitrite and then into NO.
There were lots of indices about oxidative stress, including ROS, antioxidant enzymes, and MDA. ROS were highly reactive and attack biomolecules including proteins, DNA, and lipids such as polyunsaturated fatty acids. This phenomenon was generally known as "oxidative stress". The polyunsaturated fatty acid arachidonic acid could be peroxidized to finally form MDA. This particular reaction of ROS with lipids was generally known as "lipid peroxidation" (Tsikas, 2017). Previous study showed that a time-dependent increase in lipid peroxidation products was observed during ischemia in a model of HIRI which indicated MDA was a key factor of oxidative stress process in HIRI (Fukai et al., 2005). Our study revealed that dietary inorganic nitrate could regulate the level of oxidative stress, which is consistent with the results of previous studies. A previous study evaluated the role of dietary nitrate in unilateral renal IR injury and found that nitrate could regulate the level of oxidative stress (Yang et al., 2017). However, the mechanism of action of nitrate on oxidative stress was not clarified. Similar conclusions have been derived by intravenous nitrite administration in heart IR injury (Shiva et al., 2007).
NRF2 is a member of the cap 'n' collar family of basic region leucine zipper transcription factors that was first identified, cloned, and characterized in 1994 (Moi et al., 1994). Growing evidence implies that the activation of NRF2 signaling and the upregulation of downstream antioxidant enzymes are crucial to suppress oxidative stress and maintain cellular homeostasis (Bellezza et al., 2018). Many drugs exert anti-cancer, antiapoptosis, anti-oxidation, and anti-inflammatory effects by regulating the expression and activity of NRF2 (Copple et al., 2010;Sun et al., 2015). Also, a research about kidney IR showed that NRF2-knockout mice had a worse performance compared to wild type mice (Liu et al., 2009). In this study, we focused on NRF2 to see whether it could modulate hepatic injury caused by IR.
On NRF2 pathway, several key genes take vital roles. KEAP1, a NRF2 repressor and the main intracellular regulator of NRF2, has five domains, each of which is important for inhibiting NRF2 activity (Bellezza et al., 2018). NQO1 is a cytosolic flavin that catalyzes the twoelectron reduction and detoxification of quinones and other exogenous and endogenous chemicals in the redox cycle (Vasiliou et al., 2006). HO1 is a cytoprotective endogenous enzyme that exhibits both anti-inflammatory and antioxidative effects by catalyzing the first and rate-limiting step (Klaassen and Reisman, 2010). We investigated the expression of KEAP1, NQO1, and HO1 and found that total protein expression of NRF2 and NQO1 were increased under dietary nitrate. Importantly, we found nuclear NRF2 also upregulated compared to the HIRI group which means NRF2 had nuclear translocation to further influence the downstream genes.
In order to verify the in vivo results, we introduced in vitro study. With H 2 O 2 stimulation, the levels of ROS and Annexin V of AML12 cells increased significantly, while SNP, which could release NO, could decrease the oxidative stress and apoptosis caused by H 2 O 2 . Also, during SNP loading, NRF2 pathway was activated. In order to ensure the function of NRF2, we employed a NRF2 inhibitor-ML385, under a concentration without affecting the normal cells, could inhibit the expression of NRF2 associated genes. And more importantly, NRF2 inhibition could reverse the protective effects of SNP on H 2 O 2 treated AML12 cells. By experimental verification, we hypothesized that NO, which was derived from nitrate, could be transported through the cell membrane, and activate the NRF2 pathway. More NRF2 protein is transported into the nucleus to mediate the transcription of related genes. Finally, the proteins that were transcribed by these genes play pivotal roles in modulating hepatic oxidative stress ( Figure 5L).
In summary, here we found that dietary nitrate could effectively prevent HIRI in mice. We studied the underlying mechanism of action and found that nitrate converted to NO regulates oxidative stress by activating the NRF2 pathway. These data support the concept and the feasibility of dietary inorganic nitrate for the clinical prevention of HIRI.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

ETHICS STATEMENT
The animal study was reviewed and approved by Beijing Friendship Hospital.

AUTHOR CONTRIBUTIONS
All listed authors participated meaningfully in the study, and they have seen and approved the submission of this manuscript. SL participated in performing the research, analyzing the data, and initiating the original draft of the article. HJ, GS, CZ, JW, and HX participated in performing the research and collecting the data. SW and DZ established the hypotheses, supervised the studies, analyzed the data, and co-wrote the manuscript.