Keap1-Nrf2 signaling activation by Bardoxolone-methyl ameliorates high glucose-induced oxidative injury in human umbilical vein endothelial cells

In cultured human umbilical vein endothelial cells (HUVECs) high glucose (HG) stimulation will lead to significant cell death. Bardoxolone-methyl (BARD) is a NF-E2 p45-related factor 2 (Nrf2) agonist. In this study we show that BARD, at only nM concentrations, activated Nrf2 signaling in HUVECs. BARD induced Keap1-Nrf2 disassociation, Nrf2 protein stabilization and nuclear translocation, increasing expression of antioxidant response element (ARE) genes. BARD pretreatment in HUVECs inhibited HG-induced reactive oxygen species production, oxidative injury and cell apoptosis. Nrf2 shRNA or knockout (using a CRISPR/Cas9 construct) reversed BARD-induced cytoprotection in HG-stimulated HUVECs. Conversely, forced activation of Nrf2 cascade by Keap1 shRNA mimicked BARD’s activity and protected HUVECs from HG. Importantly, BARD failed to offer further cytoprotection against HG in the Keap1-silened HUVECs. Taken together, Keap1-Nrf2 cascade activation by BARD protects HUVECs from HG-induced oxidative injury.


INTRODUCTION
Vascular endothelial cell injury is a primary medical issue in the pathogenesis and progression of cardiovascular complications in diabetes mellitus patients [1,2]. Sustained high glucose (HG) exposure to vascular endothelial cells will induce robust reactive oxygen species (ROS) production and oxidative injury. Subsequently, it will induce calcium overload, lipid peroxidation, as well as profound protein and DNA damage. These events together will induce cell death and apoptosis [3][4][5][6][7]. Human umbilical vein endothelial cells (HUVECs) were cultured with HG-containing medium, mimicking diabetes mellitus-induced pathogenesis in vascular endothelial cells [8][9][10][11][12][13].
NF-E2 p45-related factor 2 (Nrf2) is a mater transcriptional factor suppressing cellular oxidative injury and other stresses [14][15][16][17][18]. Once activated, the transcription factor will depart from its suppressor protein Keap1. The Nrf2 protein will be then stabilized and translocate to cell nuclei. At last it will bind to antioxidant response element (ARE), essential for the basal and inducible expression of many different genes. These genes encode detoxification enzymes, antioxidant proteins and many other cytoprotective proteins [14][15][16][17][18].
Studies have demonstrated that forced activation of Keap1-Nrf2 cascade can efficiently protect vascular endothelial cells from HG and other oxidative injury. Tang  AGING protected HUVECs from HG through activation of Nrf2 signaling cascade [8]. A study by Yang et al., showed that phloretin activated AMPK (AMP-activated protein kinase)-Nrf2 cascade to alleviate endothelial cell injury by palmitic acid [19]. Zeng et al., reported that a long non-coding RNA MALAT1 activated Nrf2 cascade to protect HUVECs from hydrogen peroxide [20]. Therefore, Nrf2 activation should be a valuable strategy to protect endothelial cells from oxidative injury.

BARD robustly activates Nrf2 signaling cascade in HUVECs
BARD can induce Nrf2 signaling cascade activation by releasing Nrf2 from Keap1 [21,22]. A coimmunoprecipitation (Co-IP) assay was carried out in cultured HUVECs. Results, in Figure 1A, demonstrated that the cytosol Keap1-Nrf2 association was disrupted with treatment of BARD (10-100 nM) for 3h. The input control results demonstrated that Nrf2 protein levels were elevated in BARD-treated HUVECs ( Figure 1B), where Keap1 levels were unchanged ( Figure 1B). By testing the nuclear fraction proteins, we found that the Nrf2 protein was enriched in the nuclei of BARD (10-100 nM)-treated HUVECs, with significant increase of ARE activity ( Figure 1D). Based on the results we propose that BARD treatment disrupted Nrf2-Keap1 binding, causing cytosol Nrf2 protein stabilization and nuclear translocation, thus increasing ARE activity in HUVECs.
Further results show that mRNA expression of Nrf2-ARE-dependent genes, including HO1, NQO1 and GCLC, was significantly increased following BARD (10-100 nM) treatment in HUVECs ( Figure 1E). Expression of Nrf2 mRNA was, however, unchanged ( Figure 1F). Protein levels of HO1, NQO1 and GCLC were augmented as well in BARD-treated HUVECs ( Figure 1G). Therefore, BARD efficiently (at nM concentrations) activated Nrf2 signaling cascade in HUVECs. Since 50 nM BARD induced robust Nrf2 cascade activation, this concentration was chosen for the following studies.
Importantly, in the Keap1-silenced HUVECs, treatment with BARD failed to further increase Nrf2 protein levels ( Figure 4A), ARE activity ( Figure 4B) or the expression of ARE-dependent genes (HO1, NQO1 and GCLC, Figure 4C, 4D). Therefore, BARD was ineffective on Nrf2 signaling in HUVECs with Keap1 silencing. Functional studies demonstrated that sh-Keap1-HUVECs were protected from HG, presenting AGING with significantly decreased viability (CCK-8 OD) reduction ( Figure 4E) and apoptosis activation ( Figure  4F), when compared to the control HUVECs with scramble non-sense shRNA lentivirus ("sh-C"). Significantly, in the sh-Keap1-HUVECs treatment with BARD failed to offer further cytoprotection against HG-induced cell death and apoptosis ( Figure  4E, 4F). These results further suggest that activation of Keap1-Nrf2 cascade mediated BARD-induced cytoprotection in HG-stimulated HUVECs.

DISCUSSION
Under the resting conditions, Nrf2 binds to its suppressor protein Keap1. It is distributed in an inactive state and eventually degraded in the cytosol. Following activation, the Nrf2 protein will disassociate with Keap1, causing Nrf2 cytoplasmic entry into the nuclei and binding to antioxidant genes. This will promote the transcription and activate cellular antioxidant functions [17,18,32]. The emerging studies have reported that BARD could activate Nrf2 signaling and ameliorate oxidative stress through disrupting Keap1-Nrf2 association and induction of antioxidant genes [22,33,34].
In the present study we show that BARD induced robust Nrf2 signaling activation in HUVECs. Following BARD treatment Nrf2 departed from KEAP1 in cytosol, causing Nrf2 protein stabilization. The active Nrf2 protein then translocated to cell nuclei in HUVECs, promoting transcription of ARE genes, including HO1, NQO1 and GCLC. Importantly, HGinduced ROS production, oxidative injury and cell death/apoptosis were largely attenuated by BARD pretreatment in HUVECs. Thus, BARD activated Nrf2 cascade and ameliorated HG-induced oxidative injury in HUVECs.

Figure 2. BARD inhibits high glucose-induced oxidative injury in HUVECs.
HUVECs were pretreated with Bardoxolone Methyl (BARD, at 50 nM) for 1h, followed by HG stimulation and cultured for applied time periods, the cellular superoxide contents (A), the GSH/GSSH ratio (B) and mitochondrial depolarization (JC-1 green intensity, C) were tested; Cell viability and death were tested by CCK-8 (D) and medium LDH release (E) assays, respectively, with cell apoptosis analyzed by caspase-3 activity (F), nuclear TUNEL staining (G) and Annexin V-FACS (H) assays. For TUNEL staining assays, at least 500 nuclei in five random views (1×200 magnification) for each condition were included to calculate the TUNEL/DAPI ratio (same for all Figures). Error bars stand for mean ± standard deviation (SD, n=5). "Ctrl" stands for cells-cultured in the normal glucose medium (same for all Figures). ** p<0.01 vs. "Ctrl" treatment. ## p<0.01. vs. HG only treatment (no BARD pretreatment). Each experiment was repeated five times to insure the consistency of experimental results.

AGING
Our results implied that activation of Keap1-Nrf2 signaling is required for BARD-induced cytoprotection against HG in HUVECs. Nrf2 silencing or KO completely blocked BARD-induced increases of ARE activity and expression of ARE genes (HO1, NQO1 and GCLC). Furthermore, BARD-induced cytoprotection against HG was completely reversed with Nrf2 silencing or KO in HUVECs. Further studies found that forced Nrf2 cascade activation by Keap1 shRNA mimicked BARD's activity and protected HUVECs from HG-induced cell death/apoptosis. Importantly, BARD failed to offer further cytoprotection against HG in the Keap1-silened HUVECs. Therefore, Keap1 depletion not only mimicked, but also nullified, BARDinduced cytoprotection against HG in HUVECs. These results confirm that in HUVECs BARD-induced cytoprotection against HG requires Keap1-Nrf2 cascade activation.
As compared to other known Nrf2 activators, one significant and obvious advantage of this compound is that it directly acts on Keap1, causing fast and sustained Keap1-Nrf2 disassociation and profound Nrf2 cascade activation [22,23,34]. Therefore, BARD is an extremely efficient and promising Nrf2 activator.
Phase II clinical trials in chronic kidney diseases (CDKs) patients have demonstrated a long-term increment in glomerular filtration by BARD administration [35]. However, a phase III clinical trial evaluating patients with type 2 diabetes mellitus and stage 4 CDKs revealed that BARD failed to reduce the Nrf2 shRNA lentiviral particles ("sh-Nrf2") or the lenti-CRISPR-GFP-Nrf2 knockout (KO) construct ("ko-Nrf2"), as well as the parental control cells ("Pare"), were treated with Bardoxolone Methyl (BARD, at 50 nM) for applied time periods, expression of listed genes was tested by qPCR and Western blotting analyses (A, C, D); The relative ARE activity was examined as well (B); Alternatively, cells were pretreated with BARD (50 nM) for 1h, followed by HG stimulation and cultured for 48h, cell viability (CCK-8 assay, E) and apoptosis (nuclear TUNEL staining assay, F) were tested. Expression of the listed proteins was quantified, normalizing to the indicated loading control protein (A, D). Error bars stand for mean ± standard deviation (SD, n=5). ## p<0.01. vs. "Pare" cells. Each experiment was repeated five times to insure the consistency of experimental results. AGING risk of end-stage renal disease (ESRD) or death from cardiovascular causes [36]. The trial was terminated due HUVECs with Keap1 shRNA lentiviral particles ("sh-Keap1") were treated with or without Bardoxolone Methyl (BARD, at 50 nM) for applied time periods, control cells were transduced with the scramble control shRNA ("sh-C"), expression of listed genes was tested by qPCR and Western blotting analyses (A, C, D), with the relative ARE activity examined as well (B); Alternatively, cells were pretreated with BARD (50 nM) for 1h, followed by HG stimulation and cultured for 48h, cell viability (CCK-8 assay, E) and apoptosis (nuclear TUNEL staining assay, F) were tested. Expression of the listed proteins was quantified, normalizing to the indicated loading control protein (A and D). Error bars stand for mean ± standard deviation (SD, n=5). ## p<0.01. vs. "sh-C" cells. Each experiment was repeated five times to insure the consistency of experimental results.
to a high rate of heart-related adverse events in patients, including nonfatal myocardial infarction, nonfatal stroke, heart failure, or death from cardiovascular causes [36]. Interestingly, a new phase II clinical trial is recruiting rare CKD patients to better define the safety and efficacy profiles of BARD [37]. Furthermore, a phase II clinical trial is ongoing to evaluate BARD in pulmonary arterial hypertension patients [37].
It should be noted that the current results from in vitro studies shall not be directly translated to humans. A single concentration of HG stimulation in cultured HUVECs is also not equal to vascular injury in the diabetes mellitus patients. The efficacy and safety of BARD will need further in vivo characterizations.

CONCLUSIONS
Taken together, we conclude that Keap1-Nrf2 cascade activation by BARD protects HUVECs from HGinduced oxidative injury.

Materials and reagents
BARD, CCK-8 dye and puromycin were provided by Sigma-Aldrich (St. Louis, MO). Antibodies utilized in this study were provided by Santa Cruz Biotechnology (Santa Cruz, CA) and Abcam (Shanghai, China). The cell culture reagents, fetal bovine serum (FBS) and others, were obtained from Hyclone Co. (Logan, UT). TRIzol along with other PCR agents and Lipofectamine 2000 were provided by Thermo-Fisher Invitrogen (Shanghai, China). All the primers and sequences were synthesized by Shanghai Genechem Co. (Shanghai, China).

Culture of HUVECs
HUVECs were provided by Dr. Jiang [38] at Nanjing Medical University, cultured in the described medium [8,39]. For high glucose (HG) stimulation, HUVECs were cultured in the basal medium plus 50 mM glucose. Cells were further cultured for indicated time periods, before further biological analyses were performed. The protocols of this study were approved by the Ethics Committee of Nanjing Medical University.

Quantitative real-time reverse transcription PCR assay (qPCR)
Following the applied treatment, the total cellular RNA of HUVECs was extracted by TRIzol, that was reversely transcribed using a ReverTra Ace qPCR RT kit (Toyobo). qPCR was performed under an ABI Prism 7900 HT System (Applied Biosystems, Foster City, AGING CA). mRNA primers for Nrf2, HO1, NQO1, and GCLC were listed in Table 1. The product melting temperature was always calculated [40]. Quantization of the listed mRNAs was carried out through a 2 −∆∆Ct method, with GAPDH tested as the reference gene.

ARE reporter assay
HUVECs were initially seeded onto a six-well plate (at 2 × 10 5 cells/well), and transfected with an AREinducible firefly luciferase vector (provided by Dr. Jiang at Nanjing Medical University [41]). The transfected cells were treated with BARD and cell lysates subjected to tests of the luciferase activity using a luminescence machine.

Western blotting
HUVECs with applied treatments were incubated with the described lysis buffer [42]. For each treatment, 40 μg of protein lysates were separated in a denaturing 10% SDS-PAGE gel, thereafter transferred onto a PVDF blot (Millipore, Shanghai, China). After blocking, the blot was incubated with applied primary antibody and corresponding secondary antibody at appropriate concentrations. An enhanced chemiluminescence (ECL) kit (Amersham) was utilized to test antibody-antigen binding based on the molecular weight of the targeted protein. An ImageJ software (National Institutes of Health, Bethesda, MD) was utilized for data quantification [43,44]. The separation of nuclear fraction lysates through a kit from Sigma was described previously [45].

Co-Immunoprecipitation (Co-IP)
The detailed protocols for Co-IP were described in elsewhere [44]. Briefly, following the applied BARD treatment, 1 mg of total cell lysates from HUVECs were pre-cleared by adding the protein A/G Sepharose ("Beads"). Afterwards, an anti-Keap1 antibody (15 μL for each treatment, Sana Cruz Biotech, Sana Cruz) was added to the pre-cleared lysates overnight. Thereafter, the protein A/G Sepharose were added back to the lysates for 3h. The Keap1-immunoprecipitated proteins were subjected to Western blotting analyses.

Cell viability
At the density of 30, 000 cells/cm 2 HUVECs were initially seeded into 96-well tissue culture plates. Following the indicated treatments, cell viability was measured by using the Cell Counting Kit-8 (CCK-8, Dojindo Laboratories, Kumamoto, Japan). The optical density (OD) values of CCK-8 were tested at 550 nm.

Caspase-3 activity
At 30, 000 cells/cm 2 HUVECs were initially seeded into six-well tissue culture plates. Following the applied HG stimulation, 25 μg of achieved total cell lysates were incubated with DEVD-AFC (the caspase-3 substrate, Invitrogen Thermo-Fisher, Shanghai, China). The relative caspase-3 activity, reflected by the AFC absorbance, was examined through a Fluoroskan Ascent FL equipment machine at 355 nm excitation and 525 nm emission [8].

Annexin V FACS
HUVECs with the applied treatments were co-incubated with 5 μg/mL of Annexin V and 5 μg/mL of propidium iodide (PI) for 20 min. A fluorescent-activated cell sorting (FACS) equipment (BD bioscience, Shanghai, China) was utilized to sort the apoptotic cells with positive Annexin V staining. Annexin V ratios were recorded.

Mitochondrial depolarization
After the applied treatment, mitochondrial depolarization ("∆Ψ") in HUVECs was measured by a mito-dye JC-1, which shall form green monomers with mitochondrial depolarization in stressed cells [46]. The protocol of JC-1 assay has been described elsewhere [47]. JC-1 green fluorescence absorbance was recorded at the test wavelength of 550 nm.

Superoxide assay
At 30, 000 cells/cm 2 HUVECs were initially seeded into six-well tissue culture plates. Following the applied HG stimulation, superoxide contents were tested through a superoxide colorimetric assay kit (BioVision, San Francisco, CA) following the attached protocols. At the wavelength of 450 nm the absorbance of superoxide was tested.

Glutathione contents
At 30, 000 cells/cm 2 HUVECs were initially seeded into six-well tissue culture plates. Following the applied HG AGING Table 1. Primers utilized in this study.

Nrf2 knockout
The monoclonal stable HUVECs with the lenti-CRISPR-GFP-Nrf2 knockout (KO) construct, or the Nrf2 KO HUVECs, as well as control cells with an empty vector, were provided by Dr. Xu [8]. Nrf2 KO was always verified by Western blotting and qPCR analyses.

Statistics analyses
Data were expressed as mean ± standard deviation (SD). Statistical analyses were performed by repeatedmeasures analysis of variance (RMANOVA) with Dunnett's post hoc test using a SPSS software (version 21.0, SPSS Inc., Chicago, IL). A two-tailed unpaired T test (Excel 2013, Microsoft) was carried out to examine significance between two treatment groups. P < 0.05 was considered statistically significant.

AUTHOR CONTRIBUTIONS
All authors carried out the experiments, participated in the design of the study and performed the statistical analysis, participated in its design and coordination and helped to draft the manuscript.

CONFLICTS OF INTEREST
The listed authors have no conflicts of interest.

FUNDING
This work is supported by the National Natural Science Foundation.