Anal Bioanal Chem DOI 10.1007/s00216-015-8481-0

RESEARCH PAPER

Metabonomics revealed xanthine oxidase-induced oxidative stress and inflammation in the pathogenesis of diabetic nephropathy Jingping Liu & Chengshi Wang & Fang Liu & Yanrong Lu & Jingqiu Cheng

Received: 16 October 2014 / Revised: 6 January 2015 / Accepted: 10 January 2015 # Springer-Verlag Berlin Heidelberg 2015

Abstract Diabetic nephropathy (DN) is a serious complication of diabetes mellitus (DM), which is a major public health problem in the world. To reveal the metabolic changes associated with DN, we analyzed the serum, urine, and renal extracts obtained from control and streptozotocin (STZ)-induced DN rats by 1H NMR-based metabonomics and multivariate data analysis. A significant difference between control and DN rats was revealed in metabolic profiles, and we identified several important DN-related metabolites including increased levels of allantoin and uric acid (UA) in the DN rats, suggesting that disturbed purine metabolism may be involved in the DN. Combined with conventional histological and biological methods, we further demonstrated that xanthine oxidase (XO), a key enzyme for purine catabolism, was abnormally activated in the kidney of diabetic rats by hyperglycemia. The highly activated XO increased the level of intracellular ROS, which caused renal injury by direct oxidative damage to renal cells, and indirect inducing inflammatory responses via activating NF-κB signaling pathway. Our study highlighted that metabonomics is a promising tool to reveal the metabolic changes and the underlying mechanism involved in the pathogenesis of DN.

Electronic supplementary material The online version of this article (doi:10.1007/s00216-015-8481-0) contains supplementary material, which is available to authorized users. J. Liu : C. Wang : Y. Lu (*) : J. Cheng (*) Key Laboratory of Transplant Engineering and Immunology, Regenerative Medicine Research Center, West China Hospital, Sichuan University, No. 1 Keyuan 4th Road, Gaopeng Ave, Chengdu 610041, China e-mail: [email protected] e-mail: [email protected] F. Liu Division of Nephrology, West China Hospital, Sichuan University, No. 37 Guo Xue Xiang, Chengdu 610041, China

Keywords Diabetic nephropathy . Metabonomic . Xanthine oxidase . Oxidative stress . Inflammation . NF-κB

Introduction Diabetic nephropathy (DN) is one of the major complications of diabetes mellitus (DM) and is now the leading cause of endstage renal disease (ESRD) which has high morbidity and mortality, and affects about one third of subjects with DM [1–3]. Currently, the efficacy of early diagnosis and treatment of DN is relatively poor due to its complex pathogenesis [2]. Numerous factors have been implicated in the pathogenesis of DN, such as advanced glycation end products (ACEs), transforming growth factor (TGF-β), the polyol pathway, protein kinase C, reactive oxygen species (ROS), hemodynamic alterations, and so on [4–6]. Despite many factors having been reported, emerging evidence indicates that these factors are only a partial aspect of a much more complex picture. Therefore, the systemic changes related to diabetic renal injury still need to be investigated, which may provide additional information for the discovery of novel markers and therapeutic targets to DN. Metabonomics is a systemic biological approach which can detect the global changes of small molecular metabolites in biological samples rapidly, and can reveal metabolic responses of living systems to genetic, environmental, and pathological stimuli as well [7, 8]. There are several analytical methods used in metabonomic studies, including nuclear magnetic resonance spectroscopy (NMR), gas chromatography coupled mass spectrometry (GC-MS), and liquid chromatogr ap h y c o u p l e d m a s s s p e c t r o m e t r y ( L C - M S ) [ 9] . Metabonomics has been widely used in DN-related disease prediction, diagnosis, and treatment efficacy evaluation studies [10–12]. For instance, Hirayama et al. identified 19

J. Liu et al.

metabolites in serum from diabetic patients using metabolomics, which could distinguish diabetic patients without albuminuria from DN patients with albuminuria [11]. A recent metabolomic study reported that several combined urine and plasma metabolites could predict further development of macroalbuminuria in T2DM patients [12]. These studies showed that metabolomics was a potential tool for discrimination and prediction of DN, but the underlying pathways related to the reported metabolites and the molecular mechanism leading to renal damage remains uncertain. Zhao et al. used NMR-based metabonomics to analyze urine and kidney extract from STZ-induced diabetic rats, and identified disturbed lipid/ ketone body synthesis, TCA cycle and glycolysis in DN rats [13], suggesting that metabonomics might be a useful tool to reveal underlying pathways related to DN. In this study, we used NMR-based metabonomics to identify the significantly changed metabolites in serum, urine, and renal extract from control and DN rats, and combined with conventional clinical chemistry, histological and biological methods, we further elucidated the molecular mechanism of metabolic changes which contributed to the development of DN.

Materials and methods Animal experiment and sample collection Male Sprague-Dawley (SD) rats were purchased from Experimental Animal Center of Sichuan University (Chengdu, China). Animals were housed in individual cage with controlled temperature, humidity, and 12 h cycles of light and darkness, and fed with standard chow and tap water ad libitum. Animal experiment was performed according to the guidelines of Animal Care and Use Committee of Sichuan University. Rats were randomly divided into control and DN group (n = 10) and were made diabetic by intraperitoneally injection of STZ (65 mg/kg, Yuyang High-tech Developing Co. Ltd., Chengdu, China) dissolved in citrate buffer (10 mM, pH = 4.5) as previously reported [13]. The fasting blood glucose (FBG) level of rats were measured daily by glucometer (Accu-Chek, Roche) after STZ injection, and rats with FBG>16.7 mM for three consecutive tests were defined as diabetic. At 8 weeks after STZ injection, blood, urine, and kidney from rats were collected, respectively. Prior to sacrifice, each rat was kept in a metabolic cage, and 24-h urine sample was collected with addition of 0.1% (w/v) sodium azide. The supernatants of urine were collected by centrifuging and stored at −80 °C. The rats were anesthetized with an intraperitoneal injection of sodium pentobarbital (40 mg/kg), and the blood

samples were obtained by cardiac puncture, and allowed to clot at 4 °C. Sera were collected after centrifuging and stored at −80 °C. Kidneys were removed and weighed, and then the left kidney was fixed in 4% paraformaldehyde for histological examination and right kidney was stored in liquid nitrogen for further analysis. Biochemical measurement Clinical biochemistry analysis was performed on a Automatic Biochemistry Analyzer (Cobas Integra 400 plus, Roche) using commercial kits with the following parameters: blood glucose (GLU), cholesterol (TC), triglyceride (TG), blood urea nitrogen (BUN), uric acid (UA), and urinary albumin to creatinine ratio (UACR). 1

H NMR measurement

All samples were thawed and centrifuged to remove any precipitates prior to NMR analysis. To obtain a deuterium lock signal, 200 μl serum was mixed with 300 μl D2O (Cambridge Isotope Laboratories, Inc. USA), and the mixture was pipetted into a 5-mm NMR tube (Wilmad, USA). To minimize chemical shift variation due to different pH, 400 μl urine was mixed with 100 μl PBS (0.2 M, pH=7.0) and 50 μl D2O containing 1% (w/v) 3-trimethylsilyl propionic acid-d4 sodium salt (TSP, Cambridge Isotope Laboratories, Inc. USA), and was pipetted to NMR tube. Renal tissues were homogenized and extracted using a chloroform/methanol/H2O method as previously reported [14]. The lyophilized extracts were resuspended in 500 μl PBS with addition of 50 μl TSP (1%, in D2O), and 550 μl supernatant was transferred to NMR tube after centrifugation. 1 H NMR spectra were measured on a Bruker Avance II 600-MHz spectrometer (Bruker Biospin, Germany) at 298 K with a 5-mm PATXI probe. Serum NMR spectra were acquired using a water-suppressed Carr-PurcellMeiboom-Gill (CPMG) pulse sequence (RD-90°-(τ180°-τ)n-acq) with the following parameter: relaxation delay=5 s, τ=400 μs, n=100, acquisition time=2.65 s. The 90° pulse length was adjusted to about 10 μs and 64 scans were collected into 64 K data points with a spectral width of 20 ppm. The induction decays (FIDs) were weighted by an exponential function with a 0.3 Hz line broadening and zero-filled to 128 k data points prior to Fourier transformation (FT). The chemical shifts were referenced to lactate at 1.33 ppm. The NMR spectra of urine and renal extracts was acquired by a NOESY-presaturation (NOESYPR1D) pulse sequence (RD-90°-t1-90°-tm-90°-acq) with the following parameter: relaxation delay=5 s, mixing time=100 ms, acquisition time=2.48 s. The 90° pulse length was adjusted to about 10 μs and 64 scans were collected into 32 K

1

H NMR-based metabonomic study of rat DN

data points with a spectral width of 11 ppm. The FIDs were weighted by an exponential function with a 0.3 Hz line broadening and zero-filled to 32 k data points prior to FT. The chemical shifts were referenced to TSP at 0 ppm.

Real-Time PCR Detection System (Bio-Rad, USA) with SYBR green supermix (SsoFast EvaGreen, Bio-Rad). The relative changes of mRNA level were calculated by delta-delta Ct method with β-actin as internal reference gene.

Multivariate data analysis ELISA measurement All NMR spectra were phased and baseline corrected, and were then segmented into equal widths (0.003 ppm) corresponding to the 0.5–9.5 ppm region using the MestReC software (version 4.9.9.9, Mestrelab Research, Spain). The regions including water (4.5–5.2 ppm), urea (5.6–6.2 ppm), and glucose (3.4–3.9 ppm) were removed prior to normalization. The remaining segments were normalized to the sum intensity of each spectrum excluding the water, urea, and glucose regions. The normalized integrals were then subjected to multivariate pattern recognition analysis. The distribution of all data including the clustering and the outlier was first observed by principal component analysis (PCA) using SIMCAP+ (version 11.5, Umetrics, Sweden). Statistical total correlation spectroscopy (STOCSY) analysis was performed using MATLAB R2009a software (Mathworks, Natick, MA). Orthogonal projection to latent structure with discriminant analysis (OPLS-DA) with unit-variance (UV) scaling was used to identify distinct metabolites between the control and the experimental group. The quality of OPLS-DA model was evaluated by the parameters of R2X, R2Y, and Q2. Each model was validated by a seven-fold cross validation method with CV-ANOVA and a permutation test (permutation number= 200) to avoid over-fitting using SIMCA-P+. The OPLS-DA coefficient plot was generated with an in-house-developed MATLAB script and was color coded with absolute value of correlation coefficients (|r|), which indicated significant metabolites that contributed to group differentiation. Quantitation of metabolites in the 1H NMR spectra of urine and kidney was done by Chenomx NMR Suite as previously reported [15]. Chenomx software allows fitting spectral lines by using standard metabolite library for 1H NMR spectra, and the known reference signal (TSP) was chosen as internal standard to automatically calculate the concentration of metabolites by comparing peak areas with appropriate proton ratio. Quantitative real-time PCR Total RNA was isolated from renal tissues and cultured cells by Trizol Reagent (GIBCO, Life Technologies) according to the manufacturer’s instruction. The RNA was quantified by NanoDrop 2000 microspectrophotometer (Thermo Fisher Scientific, Inc.). cDNA was synthesized by an iScript cDNA Synthesis Kit (Bio-Rad, CA, USA). The primers were synthesized in Invitrogen Inc., and their sequences were listed in Table S2. PCR reactions were carried out on an iCycler iQ

To assess the degree of oxidative stress in kidney of rats, oxidative markers including 8-OHdG and 8-isoProstaglandin F2α in urine and renal tissue homogenate was measured by commercial rat ELISA kits (Cusabio Biotech Co. Ltd., China). All assays were performed following the manufacturer’s instructions.

Histological examination The fixed renal tissues were embedded in paraffin, and were then cut into 5-μm-thick sections. Renal sections were deparaffinized in xylene and rehydrated in graded ethanols, and stained with hematoxylin and eosin (HE) and periodic acid-schiff (PAS). Renal injury was assessed by morphometric analysis, and the glomerular and tubular injury index were calculated as previously reported [16]. For immunohistochemical (IHC) staining, sections were blocked with 1% BSA, and incubated with diluted primary antibodies including rabbit anti-xanthine oxidase (Santa Cruz, USA), rabbit antiIL-1β (Bioworld, USA) and rabbit anti-CD68 (Santa Cruz, USA), then incubated with HRP-conjugated secondary antibody (DAKO, USA), and finally stained with DAB substrate and hematoxylin. The stained sections were observed and recorded with a microscope (Olympus, Japan). Quantitative analysis of positive staining areas (%) in micrographs was done using Image J software.

Cell culture and treatment Human renal proximal tubule epithelial cell line (HK-2) were cultured in Dulbecco’s modified Eagle’s medium (DMEM, Invitrogen) supplemented with 10% fetal bovine serum (Gibco, USA), 50 U/ml penicillin and 50 μg/ml streptomycin in a humidified atmosphere at 37 °C with 5% CO2. Prior to treatment, cells were incubated overnight in serum-free medium supplemented with xanthine (100 μM, Sigma) as substrate. HK-2 cells were then stimulated with normal glucose (NG, 5 mM), high glucose (HG, 30 mM), HG+allopurinol (ALP, XO inhibitor, Sigma), respectively. For ALP treatment, cells were pre-incubated with ALP dissolved in DMSO for 2 h. After exposed to the indicated conditions for 3 days, cultured cells were subjected to further analysis.

J. Liu et al.

Measurement of intracellular ROS Commercial 2′,7′-dichlorofluorescin diacetate (DCFH-DA, Beyotime Institute of Biotechnology, China) kit was used to measure intracellular ROS according to the manufacturer’s instructions. Briefly, HK-2 cells were first cultured under indicated conditions for 24 h, then incubated with 10 μM DCFH-DA solution at 37 °C for 20 min and washed with PBS, and observed by fluorescent microscope (IX71, Olympus, Japan). For flow-cytometric analysis, the treated cells were digested to cell suspension and then analyzed on a flow cytometer (BD FC500, USA). Immunocytochemistry The HK-2 cells were seeded on coverslips loaded into a sixwell plate. After being exposed to the indicated conditions, cells were fixed with 4% paraformaldehyde for 15 min at room temperature, and permeabilized with 0.3% Triton X100 for 10 min. After blocking with 1% BSA for 30 min, cells were incubated with diluted primary antibodies including rabbit anti-XO (Santa Cruz, USA), mouse anti-p-IκB-α (Abcam, USA), mouse anti-NF-κB p65 (Santa Cruz, USA), and rabbit anti-IL-1β (Bioworld, USA) overnight at 4 °C, and followed by incubating with FITC-conjugated goat anti-mouse IgG or TRITC-conjugated goat anti-rabbit IgG (Millipore, USA) at 37 °C for 1 h. Cells were stained with DAPI (Sigma, USA) and washed with PBS, and then covered with mounting medium. Images of stained cells were acquired by a fluorescent microscope (Olympus IX71, Japan). Quantitative analysis of positive areas in micrographs was done using Image J software. Statistical analysis Quantitative data were presented as mean±SD and analyzed by using SPSS software (version 11.5, SPSS Inc., IL, USA) with Student’s t test, and p