Targeting NF-κB by the Cell-Permeable NEMO-Binding Domain Peptide Improves Albuminuria and Renal Lesions in an Experimental Model of Type 2 Diabetic Nephropathy

Diabetic nephropathy (DN) is a multifactorial disease characterized by hyperglycemia and close interaction of hemodynamic, metabolic and inflammatory factors. Nuclear factor-κB (NF-κB) is a principal matchmaker linking hyperglycemia and inflammation. The present work investigates the cell-permeable peptide containing the inhibitor of kappa B kinase γ (IKKγ)/NF-κB essential modulator (NEMO)-binding domain (NBD) as therapeutic option to modulate inflammation in a preclinical model of type 2 diabetes (T2D) with DN. Black and tan, brachyuric obese/obese mice were randomized into 4 interventions groups: Active NBD peptide (10 and 6 µg/g body weight); Inactive mutant peptide (10 µg/g); and vehicle control. In vivo/ex vivo fluorescence imaging revealed efficient delivery of NBD peptide, systemic biodistribution and selective renal metabolization. In vivo administration of active NBD peptide improved albuminuria (>40% reduction on average) and kidney damage, decreased podocyte loss and basement membrane thickness, and modulated the expression of proinflammatory and oxidative stress markers. In vitro, NBD blocked IKK-mediated NF-κB induction and target gene expression in mesangial cells exposed to diabetic-like milieu. These results constitute the first nephroprotective effect of NBD peptide in a T2D mouse model that recapitulates the kidney lesions observed in DN patients. Targeting IKK-dependent NF-κB activation could be a therapeutic strategy to combat kidney inflammation in DN.


Introduction
Type 2 diabetes (T2D) is one of the biggest problems in global public health and considered a pandemic problem both in emerging nations and industrialized countries [1,2]. Although sodium-glucose transport protein 2 inhibitors and glucagon like-peptide receptors agonist allow better metabolic control in these patients, diabetic nephropathy (DN) remains the main cause of end-stage renal disease worldwide [3,4]. Novel anti-inflammatory therapeutic strategies could be

In Vivo/Ex Vivo Biodistribution of NBD Peptide in BTBR Ob/Ob Mice
Optical fluorescence imaging is a new tool for the preclinical study of pharmacological approach in small animals and human surgery intervention [35,36]. Near-infrared (NIR) fluorophore IRDye 800Cw allows high contrast, sensitivity and extinction coefficient for in vivo studies [36]. Figure 1A schematizes the generation of NBD peptide conjugated to IRDye 800Cw fluorophore and the purification by reverse phase high-performance liquid chromatography (HPLC) before image detection in Odyssey CLx ® system ( Figure 1B,C). In addition, HPLC analysis of mouse urine samples evidenced the renal metabolization of NBD peptide-fluorophore conjugate ( Figure 1D).

In Vivo/Ex Vivo Biodistribution of NBD Peptide in BTBR Ob/Ob Mice
Optical fluorescence imaging is a new tool for the preclinical study of pharmacological approach in small animals and human surgery intervention [35,36]. Near-infrared (NIR) fluorophore IRDye 800Cw allows high contrast, sensitivity and extinction coefficient for in vivo studies [36]. Figure 1A schematizes the generation of NBD peptide conjugated to IRDye 800Cw fluorophore and the purification by reverse phase high-performance liquid chromatography (HPLC) before image detection in Odyssey CLx ® system ( Figure 1B,C). In addition, HPLC analysis of mouse urine samples evidenced the renal metabolization of NBD peptide-fluorophore conjugate ( Figure 1D). The administration of fluorophore-conjugated NBD peptide showed rapid systemic distribution after intravascular, subcutaneous and intraperitoneal injection (starting 15 min post-injection; Figure 2A). Ex vivo image analysis at 4 h post-injection revealed peptide incorporation in all the organs evaluated, evidencing higher uptake intensity by liver and kidney ( Figure 2B). In addition, long-term pharmacokinetic behavior revealed that NBD peptide was   The administration of fluorophore-conjugated NBD peptide showed rapid systemic distribution after intravascular, subcutaneous and intraperitoneal injection (starting 15 min post-injection; Figure 2A). Ex vivo image analysis at 4 h post-injection revealed peptide incorporation in all the organs evaluated, evidencing higher uptake intensity by liver and kidney ( Figure 2B). In addition, long-term pharmacokinetic behavior revealed that NBD peptide was detected for up to 48 h following intraperitoneal administration, with high accumulation in bladder indicating predominant renal elimination ( Figure 2C). Based on the stability and kidney biodistribution, intraperitoneal route and three times per week dosing were chosen for intervention experiments.

Cell-Permeable NBD Peptide Reduced Albuminuria and Morphologic KidneyLesions in BTBR Ob/Ob Mice
Preliminary data confirmed progressive obesity, hyperglycemia and kidney damage in BTBR ob/ob mice [33,37]. Although BTBR ob/ob is described as advanced DN model, albuminuria, podocytopenia, glomerular and tubular histopathologic changes were observed as early as 12 weeks
For this reason, we further explored the therapeutic potential of cell-permeable NBD peptide targeting NF-κB in the diabesity model. Therefore, diabetic mice at 6 weeks of age were treated intraperitoneally with active NBD cell permeable peptide at two different doses (6 and 10 µg/g body weight) for 7 weeks, using vehicle (acetonitrile ≤ 0.25%) and inactive mutant peptide (Mut 10 µg/g body weight) as control groups.
No significant differences among groups were observed in body weight ( Figure 4A) and glycemia ( Figure 4B) during the intervention period. At the end of the follow-up period, changes in kidney weight, serum creatinine, uric acid and lipids were not significant, except for total cholesterol in NBD 10µg group ( Figure 4C and Table 1). Remarkably, both doses of NBD peptide reduced albuminuria by 40-46% compared to vehicle control group ( Figure 4D).  The severity of histopathologic lesions in BTBR ob/ob mouse kidney was evaluated through semiquantitative scoring. In the vehicle group, mesangial expansion and glomerulomegaly were the most relevant glomerular changes. Arteriolar hyalinosis was also found in some mice. At tubulointerstitial level, mild focal inflammatory infiltrate and tubular flattening were observed. Noteworthy, administration of active NBD peptide significantly reduced glomerular and tubulointerstitial lesions in diabetic mice ( Figure 4E-H).

Cell-Permeable NBD Peptide Reduced Podocyte Damage and Basement Membrane Thickness in BTBR Ob/Ob Mouse Kidney
DN is one of the main causes of podocytopathy described in nephrology, owing its current incidence in the worldwide population [38,39]. The total podocyte count was assessed by Wilms' tumor protein-1 (WT-1) immunostaining in kidney sections. Only intraglomerular positive cells were quantified, discarding WT-1 + parietal cells. Compared with vehicle group, administration of cell-permeable NBD peptide significantly increased total podocyte content in a dose-dependent manner ( Figure 5A). Furthermore, ultrastructural studies by transmission electron microscopy evidenced that both doses of NBD peptide improved podocyte foot process, pedicellar effacement and also regularized the outline and thickness of glomerular basement membrane ( Figure 5B). Nevertheless, the thickening of tubular basement membrane, an indicator of diabetic tubulointerstitial disease, was only reduced in mice receiving a low dose of NBD peptide, with a mild irregular thickening observed in higher dose ( Figure 5C).

Cell-Permeable NBD Peptide Modulated the Proinflammatory and Oxidative Stress Markers in BTBR Ob/Ob Mice and Cultured Cells
Real-time PCR analyses in diabetic kidneys evidenced a significant decrease in the gene expression of STAT transcription factors (Stat1 and Stat3), inflammatory cytokines (Tnfα, Il-1β and Il-12) and chemokines (Ccl2, Ccl5 and C-X-C motif chemokine ligand 10; Cxcl10) in NBD-treated mice compared to vehicle group ( Figure 6A). Cell-permeable NBD peptide also modified redox gene expression in the kidney by reducing pro-oxidant enzyme NADPH oxidase 4 (Nox4) and increasing Nrf2-dependent antioxidant genes (Nrf2/Nfe2l2 and superoxide dismutase 1; Sod1) ( Figure 6B).

Cell-Permeable NBD Peptide Modulated the Proinflammatory and Oxidative Stress Markers in BTBR Ob/Ob Mice and Cultured Cells
Real-time PCR analyses in diabetic kidneys evidenced a significant decrease in the gene expression of STAT transcription factors (Stat1 and Stat3), inflammatory cytokines (Tnfα, Il-1β and Il-12) and chemokines (Ccl2, Ccl5 and C-X-C motif chemokine ligand 10; Cxcl10) in NBD-treated mice compared to vehicle group ( Figure 6A). Cell-permeable NBD peptide also modified redox gene expression in the kidney by reducing pro-oxidant enzyme NADPH oxidase 4 (Nox4) and increasing Nrf2-dependent antioxidant genes (Nrf2/Nfe2l2 and superoxide dismutase 1; Sod1) ( Figure 6B).

Cell-Permeable NBD Peptide Modulated the Proinflammatory and Oxidative Stress Markers in BTBR Ob/Ob Mice and Cultured Cells
Real-time PCR analyses in diabetic kidneys evidenced a significant decrease in the gene expression of STAT transcription factors (Stat1 and Stat3), inflammatory cytokines (Tnfα, Il-1β and Il-12) and chemokines (Ccl2, Ccl5 and C-X-C motif chemokine ligand 10; Cxcl10) in NBD-treated mice compared to vehicle group ( Figure 6A). Cell-permeable NBD peptide also modified redox gene expression in the kidney by reducing pro-oxidant enzyme NADPH oxidase 4 (Nox4) and increasing Nrf2-dependent antioxidant genes (Nrf2/Nfe2l2 and superoxide dismutase 1; Sod1) ( Figure 6B).  In order to confirm the in vivo findings, we performed in vitro studies using mesangial cells and macrophages under hyperglycemic and/or inflammatory conditions. In mesangial cells, NBD peptide disrupted the interaction of NEMO and IKKα/β in the kinase complex ( Figure 7A) and prevented the nuclear translocation of p65 subunit ( Figure 7B) induced by high-glucose in combination with TNFα. Real-time PCR revealed a dose-dependent inhibition of NF-κB-dependent genes (Ccl2, Ccl5 and Cxcl10) by NBD peptide (Figure 7C), while the inactive Mut peptide had no significant effect. Furthermore, pretreatment with NBD reduced the NOX activity in macrophages ( Figure 7D). These data suggest that the functional and structural renal improvement observed in NBD-treated mice may be due, at least in part, to the modulation of the local inflammation and oxidative stress.

Discussion
The symptomatic silence of DN only allows clinical recognition through the presence of microalbuminuria or estimated glomerular filtration rate (eGFR), main markers of kidney damage progression [40]. New research shows that certain patients with preserved eGFR do not develop proteinuria, which makes the diagnosis much more complex, and therefore proteinuria is being considered a marker of disease progression rather than a diagnostic marker [41][42][43]. For this reason, the early biomarkers validation in the DN progression is one of the most relevant challenges for the coming years [44]. At this point, the incorporation of preclinical models that represent reliably the characteristic findings of DN is a necessary pillar in the search of novel biomarkers and therapeutic targets of the disease [45]. In these sense, the BTBR ob/ob mouse model used in this study meets most of the criteria established by the Animal models diabetic complications consortium, being a valuable preclinical tool for the development of new therapeutic strategies to quickly translate the results into clinical practice [33]. This suitable preclinical model of DN recapitulating the human disease helped us to demonstrate the renoprotective potential of NF-κB-targeting peptide in T2D. In order to confirm the in vivo findings, we performed in vitro studies using mesangial cells and macrophages under hyperglycemic and/or inflammatory conditions. In mesangial cells, NBD peptide disrupted the interaction of NEMO and IKKα/β in the kinase complex ( Figure 7A) and prevented the nuclear translocation of p65 subunit ( Figure 7B) induced by high-glucose in combination with TNFα. Real-time PCR revealed a dose-dependent inhibition of NF-κB-dependent genes (Ccl2, Ccl5 and Cxcl10) by NBD peptide (Figure 7C), while the inactive Mut peptide had no significant effect. Furthermore, pretreatment with NBD reduced the NOX activity in macrophages ( Figure 7D). These data suggest that the functional and structural renal improvement observed in NBD-treated mice may be due, at least in part, to the modulation of the local inflammation and oxidative stress. Values normalized in each sample by endogenous control gene 18s and expressed as n-fold of the average value from diabetic vehicle control. Data shown as scatter dot plots and mean ± SD of each group (n = 6-8 mice/group). *-p <0.05; **-p <0.01; ***-p <0.001 vs. diabetic vehicle control.
In order to confirm the in vivo findings, we performed in vitro studies using mesangial cells and macrophages under hyperglycemic and/or inflammatory conditions. In mesangial cells, NBD peptide disrupted the interaction of NEMO and IKKα/β in the kinase complex ( Figure 7A) and prevented the nuclear translocation of p65 subunit ( Figure 7B) induced by high-glucose in combination with TNFα. Real-time PCR revealed a dose-dependent inhibition of NF-κB-dependent genes (Ccl2, Ccl5 and Cxcl10) by NBD peptide (Figure 7C), while the inactive Mut peptide had no significant effect. Furthermore, pretreatment with NBD reduced the NOX activity in macrophages ( Figure 7D). These data suggest that the functional and structural renal improvement observed in NBD-treated mice may be due, at least in part, to the modulation of the local inflammation and oxidative stress.

Discussion
The symptomatic silence of DN only allows clinical recognition through the presence of microalbuminuria or estimated glomerular filtration rate (eGFR), main markers of kidney damage progression [40]. New research shows that certain patients with preserved eGFR do not develop proteinuria, which makes the diagnosis much more complex, and therefore proteinuria is being considered a marker of disease progression rather than a diagnostic marker [41][42][43]. For this reason, the early biomarkers validation in the DN progression is one of the most relevant challenges for the coming years [44]. At this point, the incorporation of preclinical models that represent reliably the characteristic findings of DN is a necessary pillar in the search of novel biomarkers and therapeutic targets of the disease [45]. In these sense, the BTBR ob/ob mouse model used in this study meets most of the criteria established by the Animal models diabetic complications consortium, being a valuable preclinical tool for the development of new therapeutic strategies to quickly translate the results into clinical practice [33]. This suitable preclinical model of DN recapitulating the human disease helped us to demonstrate the renoprotective potential of NF-κB-targeting peptide in T2D. The early activation of cell signaling pathways related to inflammation and oxidative stress is directly correlated with hyperglycemia and the increase in adipose tissue dependent-body weight, and is actually described as meta-inflammation [46,47]. In this insulin-resistant and hyperglycemic context, NF-κB pathway is strongly associated with proliferative and inflammatory response in diabetic kidney, as well as cell migration, differentiation and apoptosis processes that involve both renal infiltrating and resident (mesangial, endothelial, podocyte and tubular) cells [22,48].
Previous studies in human biopsies and experimental models of DN have reported high activation levels of NF-κB family members and inflammatory target genes such as cytokines, chemokines, adhesion molecules, advanced glycation end-products and transcription factors, among others [12,18,20,29]. This is in line with an increased activity of NF-κB found in our study, both in diabetic BTBR ob/ob kidneys and in cultured cells under hyperglycemic/inflammatory conditions. We also found a positive correlation between NF-κB levels and the renal content of activated STAT3 and Nrf2 pathways and inflammatory cells. This finding prompted us to explore the inhibition of NF-κB pathway through a cell-permeable peptide derived from the IKKα/β domain as a feasible strategy to delay DN evolution in T2D.
Previous studies in T1D model by streptozotocin injection in apolipoprotein E-deficient mice have proved that NBD peptide administration ameliorates diabetic kidney damage and atherosclerosis through the modulation of systemic and local inflammation [29]. Although our

Discussion
The symptomatic silence of DN only allows clinical recognition through the presence of microalbuminuria or estimated glomerular filtration rate (eGFR), main markers of kidney damage progression [40]. New research shows that certain patients with preserved eGFR do not develop proteinuria, which makes the diagnosis much more complex, and therefore proteinuria is being considered a marker of disease progression rather than a diagnostic marker [41][42][43]. For this reason, the early biomarkers validation in the DN progression is one of the most relevant challenges for the coming years [44]. At this point, the incorporation of preclinical models that represent reliably the characteristic findings of DN is a necessary pillar in the search of novel biomarkers and therapeutic targets of the disease [45]. In these sense, the BTBR ob/ob mouse model used in this study meets most of the criteria established by the Animal models diabetic complications consortium, being a valuable preclinical tool for the development of new therapeutic strategies to quickly translate the results into clinical practice [33]. This suitable preclinical model of DN recapitulating the human disease helped us to demonstrate the renoprotective potential of NF-κB-targeting peptide in T2D.
The early activation of cell signaling pathways related to inflammation and oxidative stress is directly correlated with hyperglycemia and the increase in adipose tissue dependent-body weight, and is actually described as meta-inflammation [46,47]. In this insulin-resistant and hyperglycemic context, NF-κB pathway is strongly associated with proliferative and inflammatory response in diabetic kidney, as well as cell migration, differentiation and apoptosis processes that involve both renal infiltrating and resident (mesangial, endothelial, podocyte and tubular) cells [22,48].
Previous studies in human biopsies and experimental models of DN have reported high activation levels of NF-κB family members and inflammatory target genes such as cytokines, chemokines, adhesion molecules, advanced glycation end-products and transcription factors, among others [12,18,20,29]. This is in line with an increased activity of NF-κB found in our study, both in diabetic BTBR ob/ob kidneys and in cultured cells under hyperglycemic/inflammatory conditions. We also found a positive correlation between NF-κB levels and the renal content of activated STAT3 and Nrf2 pathways and inflammatory cells. This finding prompted us to explore the inhibition of NF-κB pathway through a cell-permeable peptide derived from the IKKα/β domain as a feasible strategy to delay DN evolution in T2D.
Previous studies in T1D model by streptozotocin injection in apolipoprotein E-deficient mice have proved that NBD peptide administration ameliorates diabetic kidney damage and atherosclerosis through the modulation of systemic and local inflammation [29]. Although our results are in agreement with these findings, the clinical significance of our study is the superior protective effect of NBD peptide in a T2D mouse model that recapitulates the human pathology. Indeed, systemic administration of NBD peptide to BTBR ob/ob mice caused a significant reduction of albuminuria (>40% reduction on average), ameliorated the histopathologic glomerular and tubulointerstitial damage and also attenuated podocytopenia, pedicellar effacement and basement membrane thickening. In addition.in vivo/ex vivo pharmacokinetics and biodistribution revealed selective renal tissue uptake at both short-and long-intervention periods. Finally, mechanistic in vitro studies confirmed that NBD peptide is a potent inhibitor of IKK-dependent NF-κB activation and target gene expression in mesangial cells and macrophages under diabetic conditions. Besides its anti-inflammatory action, NBD peptide also demonstrated antioxidant properties as evidenced by the induction of Nrf2-dependent pathway and antioxidant enzymes, as well as the suppression of pro-oxidant activity and ROS generation. Similar findings were reported in LPS-induced acute lung injury model, where NBD administration downregulated oxidative stress markers and pro-oxidant enzymes (Nox1/2/4) and improved Sod1 activity and total antioxidant capacity [49]. In a preclinical model of injury from intracerebral hemorrhage, NBD peptide relieved microglial inflammation and oxidative stress [30]. Therefore, the favorable effect of the NBD peptide is not only due to the inhibition of proinflammatory NF-κB signaling, but also through the restoration of redox balance. Our findings in the context of diabetes-induced kidney disease are consistent with other studies emphasizing the powerful therapeutic effects of different NBD-based strategies in preclinical models of Duchenne muscular dystrophy [28], Parkinson's disease [50], arthritis [27], breast cancer [31] and ischemic acute kidney injury [51]. A recent phase I trial of NBD peptide administration in dogs with large B-cell lymphoma demonstrated safety and efficacy [32], offering hope for translation to human disease.
In conclusion, our results constitute the first nephroprotective effect of cell-permeable NBD peptide in a T2D model that recapitulates the kidney lesions observed in patients with DN, with marked reduction of albuminuria and morphologic renal lesions. Targeting selective inhibition of canonical NF-κB pathway IKK-dependent could be a therapeutic strategy to combat kidney inflammation in DN patients.

Synthesis of Cell-Permeable Peptides
The active NBD peptide containing a cationic cell-penetrating sequence (octalysine) fused to the IKKβ NBD region (TALDWSWLQTE) through a diglycine spacer was synthesized as described [28,29]. A biologically inactive (or less active) mutant peptide with two tryptophan-to-arginine substitutions was used. All peptides were synthesized by ProteoGenix (Schiltigheim, France). Lyophilized peptides were dissolved in acetonitrile and stock solutions were diluted in physiological saline solution before use.
For this study, male BTBR ob/ob diabetic mice were used. Breeding pairs BTBR heterozygotes (BTBR ob +/− ) were purchased from Jackson Laboratories (Bar Harbor, ME) and housed at a density of four animals per cage in a temperature-controlled room (20-22 • C) with 12-h light-dark cycles. Ad libitum water and standard food were available. The experimental protocol was approved (25 February 2016) by the Ethics Committee for Animal Experiments of the University Austral of Chile (Permit N • 245-2016) according to National Institutes of Health guidelines.
Six-week-old male BTBR ob/ob mice were randomized into the following groups: (i) Active NBD peptide (NBD groups) at doses of 6 µg/g body weight (NBD 6 µg group; n = 8) and 10 µg/g body weight (NBD 10 µg group; n = 7); (ii) Inactive mutant peptide at dose of 10 µg/g (Mut group; n = 8); (iii) Vehicle at final acetonitrile concentration ≤0.25% in physiological saline solution (Veh group; n = 6). All groups received three intraperitoneal injections per week for seven weeks of intervention. At the end of the intervention period, all mice were analyzed and euthanized for respective analyses. In some experiments, additional for treatment groups and according to the principle of animal experimental research (the 3Rs concept), age-matched BTBR ob/ob and wild type mice kidney samples from 12-week-old (n = 5-6 mice/group) were included as a reference and immunohistochemistry analyses.
Blood glucose (Accu-Chek glucometer, Roche Diagnostics, Rotkreuz, Switzerland) and body weight levels were monitored once weekly during the follow-up period. Serum cholesterol, triglyceride and uric acid levels were measured by automated procedures. Total cholesterol, triglycerides, uric acid and serum/urine creatinine levels were measured by Jaffé reaction (LiquiColor, Wiesbaden, Germany). Albuminuria was analyzed by ELISA (Mouse albumin ALPCO, Salem, NH, USA) and corrected for urinary creatinine values to obtain urinary albumin/creatinine ratio (UACR).

In Vivo and Ex Vivo Optical Fluorescence Imaging
The pharmacokinetic behavior of NBD peptide was performed through non-invasive in vivo/ex vivo study with NIR fluorophore IRDye 800CW NHS ester reactive (LI-COR Biosciences, Lincoln, NE, USA). IRDye 800CW Ester Reactive binds primary amines of amino acid sequence such as lysine and due to its selective range of visualization (λexc = 680 nm y λem = 820 nm) with minimal intrinsic interference (autofluorescence). Purification of NBD peptide and IRDye 800CW-conjugated NBD peptide was performed through reverse phase HPLC UltiMate 3000RS system, coupled to a TSQ Vantage mass spectrometer (Thermo Fisher Scientific, Waltham, MA, USA). Prior to animal injection (intraperitoneal, ventral subcutaneous and intravascular), IRDye 800CW-conjugated NBD peptide was evaporated to dryness in SpeedVac ® for 2 h, quantified by Pierce BCA Protein Assay Kit (Thermo Fisher ScientificWaltham, MA, USA) and dissolved in PBS 1× at a final concentration of 6 µg/g body weight.
BTBR ob/ob mice (12 weeks-old, n = 3) were anesthetized with oxygen/isoflurane mixture administered by SmartFlow ® vaporizer unit in induction chamber and then incorporated in a built-in heating base at 33 • C in Odyssey CLx infrared imaging system (LI-COR Biosciences, Lincoln, NE, USA) for images detection. After 4 h post-injection, BTBR ob/ob mice were euthanized; heart, liver, inguinal visceral fat tissue, eyeball and kidney were collected for ex vivo imaging kidney and heart were cut along the coronal plane.

Histological Analysis and Immunohistochemistry
For euthanasia, mice were anesthetized with 2% 2,2,2-tribromoethanol (Sigma-Aldrich, Burlington, MA, USA) dissolved in 2-methyl-2-butanol (Sigma-Aldrich, Burlington, MA, USA). Blood samples were taken for serum collection and both kidneys were removed, decapsulated and hemisected along their perihilar sagittal plane. One-half of each kidney (right and left) was fixed in 4% formaldehyde. A small portion of the renal cortex was embedded in 2% glutaraldehyde for transmission electron microscopy. The remaining portion was stored immediately in liquid N 2 and processed for RNA extraction.
The kidneys were fixed in 4% formaldehyde, embedded in paraffin and cut in 3-4 µm tissue sections for histology (Periodic acid Schiff (PAS) and Masson's trichrome staining) and immunohistochemistry studies. Renal damage was examined in a blinded manner and scored using a semiquantitative histopathologic scale (0 to 4) to evaluate glomerular and tubulointerstitial lesions as previously described [52]. For analysis by transmission electron microscopy, the kidney tissue was fixed in 2% glutaraldehyde (Merck, Darmstad, Germany), post-fixed with 1% osmium tetroxide (Ted Pella, Inc., Redding, CA, USA) and observed under a Philips Tecnai 12 electron microscope (Philips, Eindhoven, The Netherlands) operated at 80 kV.

NOXActivity Assay
NOX-dependent superoxide anion generation in macrophages was measured by the lucigenin chemiluminescence assay as described [53]. Briefly, cells were homogenized in 50-mM phosphate buffer containing 0.01-mM EDTA, 0.32-mM sucrose and 0.1% protease inhibitor cocktail. Lysates were transferred to Röhren tubes containing 5-µM lucigenin and 100-µM NADPH (Sigma-Aldrich, Burlington, MA, USA) and chemiluminescence was measured with a luminometer (Berthold Technologies, Bad Wildbad, Germany) by counting the photon emission at 5-s intervals over 5 min. Values were expressed as relative light units (RLU) per mg of protein.

mRNA Expression
Total RNA from kidney tissue and cells was isolated with TRIzol reagent (Ambion Inc, New Haven, CT, USA). cDNA was synthesized by ImProm-II TM Reverse Transcription System (Promega, Madison, Wisconsin, USA) using 2 µg of total RNA primed with random hexamer primers. The analysis of quantitative gene expression was performed on a Rotor-Gene Q (Qiagen, Venlo, The Netherlands) using primers designed by Integrated DNA Technologies (IDT, Coralville, IA, USA) and KAPA SYBR FAST Universal Kit (Kapa Biosystems, Wilmington, MA, USA). The expression of target genes was analyzed in duplicate and normalized by 18S rRNA housekeeping gene. The primer sequences used in this study are detailed in Table 2.

Statistical Analysis
The data were expressed as scatter dot plots of total number of experiments. Statistical analyses were performed using the nonparametric Mann-Whitney U test for comparison between two groups, one-way ANOVA with Dunnett's posthoc test for multiple comparisons and Pearson coefficient test for correlation studies, considering differences statistically significant those with values of p < 0.05. All graphs and statistical tests were performed using GraphPad Prism 6 software.