点击显示 收起
【关键词】 diabetic
Departments of Medicine, Pathology and Biochemistry, and Physiology, Case Western Reserve University, and Lerner Research Institute, Cleveland Clinic Foundation, Cleveland, Ohio
Department of Medicine and Pathology and Diabetes Research and Training Center, Albert Einstein College of Medicine, Bronx, New York
ABSTRACT
Chronic hyperglycemia causes structural alterations of proteins through the Maillard reaction. In diabetes, methylglyoxal (MGO)-induced hydroimidazolones are the predominant modification. In contrast to acute hyperglycemia, mitochondrial respiration is depressed in chronic diabetes. To determine whether MGO-derived protein modifications result in abnormalities in mitochondrial bioenergetics and superoxide formation, proteomics and functional studies were performed in renal cortical mitochondria isolated from rats with 2, 6, and 12 mo of streptozotocin-induced diabetes. MGO-modified proteins belonged to the following two pathways: 1) oxidative phosphorylation and 2) fatty acid -oxidation. Two of these proteins were identified as components of respiratory complex III, the major site of superoxide production in health and disease. Mitochondria from rats with diabetes exhibited a diminution of oxidative phosphorylation. A decrease in the respiratory complex III activity was significantly correlated with the quantity of MGO-derived hydroimidazolone present on mitochondrial proteins in both diabetic and control animals. In diabetes, isolated renal mitochondria produced significantly increased quantities of superoxide and showed evidence of oxidative damage. Administration of aminoguanidine improved mitochondrial respiration and complex III activity and decreased oxidative damage to mitochondrial proteins. Therefore, posttranslational modifications of mitochondrial proteins by MGO may represent pathogenic events leading to mitochondria-induced oxidative stress in the kidney in chronic diabetes.
methylglyoxal; nephropathy; complex III; proteomics
MITOCHONDRIA ARE THE MAJOR source of reactive oxygen species (ROS) in the cell because of continuous generation of superoxide, a by-product of electron transport (9). ROS generated by mitochondria are responsible for the activation of major, independent but interrelated, pathogenic mechanisms for diabetic complications as modeled in endothelial cells exposed to hyperglycemia in vitro (47).
Longstanding diabetes is associated with alterations in mitochondrial metabolism that result in both increased formation of ROS and failure of bioenergetics. In particular, diabetes causes dysfunction of mitochondria in those tissues highly dependent on aerobic metabolism, such as heart (16), brain (27), and skeletal muscle (29). The degree of mitochondrial failure has been correlated with the duration of diabetes. Complexes I (66), III, and IV of the electron respiratory chain (ERC) are the main mitochondrial targets of hyperglycemia-induced injury.
The development of cardiomyopathy in diabetes is accompanied by both changes in protein expression (67) and posttranslational modifications of mitochondrial proteins (66). However, little is known about the nature of mitochondrial dysfunction in the kidney in diabetes or the responsible structural modifications of mitochondrial proteins (26). In early stages of diabetic nephropathy, ultrastructural changes in renal proximal tubular mitochondria have been found to correlate with disturbances in the main functions of renal tubular cells (25) and typical features of diabetic nephropathy (46, 49). In addition, there is ample evidence that oxidative stress is involved in the pathogenesis of diabetic nephropathy (33).
One important explanation for structural modifications of proteins in diabetes is the nonenzymatic glycation of proteins to form advanced glycation end products (AGEs), the Maillard reaction (6, 44). A recent quantitative study showed higher levels of AGEs on intracellular proteins compared with plasma proteins and identified dicarbonyl-induced hydroimidazolones as the predominant biomarker in diabetes (64). Inhibition of advanced glycation with aminoguanidine has been shown to attenuate the development and progression of renal injury in humans (5) and experimental diabetes (48, 59), without influencing glycemic control.
Chronic hyperglycemia is thought to alter mitochondrial function through glycation of mitochondrial proteins (26). Levels of methylglyoxal (MGO), a highly reactive -dicarbonyl by-product of glycolysis, are increased in diabetes (3). MGO readily reacts with arginine, lysine, and sulfhydryl groups of proteins (38) and nucleic acids (50), inducing the formation of a variety of structurally identified AGEs. As the major precursor of AGE formation, both in the cell and in the plasma (64), MGO can influence multiple aspects of cellular biology in diabetes (62).
Previous in vitro work in this laboratory demonstrated that MGO has an inhibitory effect on kidney mitochondrial respiration and that MGO-induced modifications are targeted to specific mitochondrial proteins (55). These observations led to the goal of this work, to identify the main mitochondrial protein targets of MGO, the alterations of mitochondrial bioenergetics, and the effect of reversing MGO activity with aminoguanidine, in the diabetic kidney. The present study demonstrates that MGO-related posttranslational modifications of mitochondrial proteins involve the following two major mitochondrial pathways: 1) oxidative phosphorylation and 2) fatty acid -oxidation. These changes occur concomitantly with progressive decrements of mitochondrial bioenergetics, targeted to complex III, and with increases in the formation of superoxide by mitochondria. Furthermore, the administration of aminoguanidine improves mitochondrial respiration and diminishes oxidative damage to mitochondrial proteins in diabetes.
MATERIALS AND METHODS
Animals
Diabetes was induced by a single tail vein injection of 45 mg/kg streptozotocin (STZ; Sigma-Aldrich, St. Louis, MO) in 0.1 mol/l sodium citrate buffer, pH 4.5, to 8-wk-old male Lewis rats weighing 150 g at the start of the study (Charles River, Wilmington, MA). Control animals received a vehicle injection. All procedures were performed in compliance with the guidelines of the Case Western Reserve University Institutional Animal Care and Use Committee. Rats were maintained throughout the study in specific pathogen-free conditions using sterilized cages, feed, and bedding. Animals with plasma glucose >16 mmol/l were classified as diabetic. To maintain body weight and to limit hyperglycemia, 2 IU ultralente insulin (Humulin U; Eli Lilly, Indianapolis, IN) were administered to diabetic animals three times per week to prevent ketosis and catabolism. Glycemia was monitored by monthly determination of plasma glucose (Trinder assay kit; Sigma) and measurement of glycated Hb (boronate affinity chromatography; Sigma) at 2, 6, and 12 mo.
Protocol A: 2- and 12-mo diabetes model. There were a total of four groups (control and STZ diabetes animals at 2 and 12 mo). Two- and 12-mo time points were selected specifically to examine the alterations of mitochondrial function in the context of evolution of diabetes.
Protocol B. A 6-mo diabetes experiment was performed with control, experimental diabetic, and aminoguanidine-treated diabetic rats. Aminoguanidine was administered in the diet (2.5 g/kg diet) as reported previously for inhibition of diabetic retinopathy (13). A 6-mo time point was selected specifically to detect the alterations in mitochondrial bioenergetics induced by diabetes at a time point previously reported to attenuate albuminuria in diabetic rats (48, 59).
Materials
Except as noted, all chemical reagents were purchased from Sigma-Aldrich. Lucigenin (bis-N-ethylacridinium nitrate) was purchased from Molecular Probes (Eugene, OR). The MGO-derived imidazole AGE antibody (1H7G5) was described previously (47). Other reagents were of the highest purity available.
Methods
Measurement of mitochondrial respiration.
Renal mitochondrial isolation. Mitochondria were isolated from the kidneys of Lewis rats as described previously (55). The tissue was suspended in cold isolation buffer (0.25 M sucrose, 2 mM EDTA, and 25 mM Tris?HCl, pH 7.4), homogenized, and subjected to differential centrifugation. The mitochondria were pelleted at 5,000 g for 10 min. The pellet was resuspended to a final protein concentration of 25 mg/ml in isolation buffer. With the use of this technique, mitochondria could be stored at 4°C for up to 5 h without a change in state 3 or state 4 respiratory rates. Proteins were determined using a bicinchoninic acid method (BCA Protein Assay Kit; Pierce, Rockford, IL).
Mitochondrial respiration. Respiration assays were performed using a polarographic Clark electrode (Instech Laboratories, Philadelphia, PA). Reactions were conducted at 25°C in a 750-μl chamber containing 0.5 mg/ml mitochondrial proteins in a respiratory buffer of 125 mM KCl and 5 mM KH2PO4 at pH 7.25. State 2 respiration was initiated by adding NADH-linked substrates (10 mM glutamate and 5 mM malate). State 3 respiration was determined by adding 300 nmol ADP. State 4 respiration was defined as oxygen consumed in the presence of the adequate substrates after ADP exhaustion.
2,4-dinitrophenol-dependent respiration. Renal mitochondrial ADP-dependent respiration might be altered by diabetes in one of two ways: through an effect on adenine nucleotide translocase and/or ATPase or by suppressing TCA and ERC. To differentiate these possibilities, renal mitochondrial respiration was measured in the presence of 60 μM 2,4-dinitrophenylhydrazine (DNP). DNP increases O2 consumption independent of ADP transport or ATP synthesis, but its effect is dependent on the supply of reducing equivalents and ERC components.
Nadh oxidase. To determine the ability of the ERC to utilize exogenously added NADH, NADH oxidase was assayed using a protocol developed by Humphries et al. (20, 21). Mitochondria were diluted to a protein concentration of 0.2 mg/ml with a hypotonic buffer (10 mM MOPS and 0.5 mM EDTA, pH 7.4) and then sonicated for 30 s to provide access of NADH and cytochrome c to the ERC. O2 consumption was monitored after the addition of 9 μM cytochrome c and 0.4 μM NADH.
Erc complexes. To further define the activity of NADH-linked ERC complexes, renal mitochondria were exposed to various experimental conditions (20, 21).
For analysis of complex I, mitochondria (0.5 mg/ml) were diluted to 50 μg/ml in a hypotonic buffer (25 mM KH2PO4, 0.5 mM EDTA) and then sonicated. Upon addition of 5 μM antimycin A, 2 mM KCN, 60 μM ubiquinone-1, and 100 μM NADH, consumption of NADH was monitored by spectrophotometer at 340 nm (molar absorptivity, = 6,200 M1?cm1).
To measure complex III activity, mitochondria (0.5 mg/ml) were diluted to 2.5 μg/ml in hypotonic buffer and then sonicated. Upon addition of 40 μM reduced decylubiquinone, 50 μM cytochrome c, and 2 mM KCN, the reduction of cytochrome c was measured at 550 nm ( = 18,500 M1?cm1). Reaction was started by the addition of cytochrome c.
Complex IV activity was assayed using an oxygen electrode to record the rate of oxygen consumption upon the addition of 5 mM ascorbate, 250 μM N,N,N',N'-tetramethyl-p-phenylenediamine, and 10 μM cytochrome c to 100 μg mitochondrial protein, after sonication.
Detection of superoxide anion in isolated mitochondria. Lucigenin was used to measure superoxide anion production by the mitochondrial ERC (12). Reaction blanks (cuvettes containing all components except mitochondria) were measured first. Renal mitochondria (0.5 mg/ml) were sonicated in respiratory buffer and then diluted to 50 μg/ml in 1 ml hypotonic buffer (10 mM MOPS, 0.5 mM EDTA, pH 7.6) complemented with 20 μM lucigenin (Molecular Probes). Mitochondrial respiration was started by the addition of 70 μM NADH as a substrate initiating ERC activity (52). The luminometer (Lumat LB 9501; Berthold, Bundoora, Australia) was set to count every fifth second at 25°C. Rates of superoxide production were expressed as counts per second. Tiron (4,5-dihydroxy-1,3-benzene-disulfonic acid; 100 mM) was added as a scavenger of superoxide (12, 35).
Western blot detection of posttranslational modifications of mitochondrial proteins by oxidative pathway. One hallmark of the oxidation status of proteins is the immunodetection of carbonyl groups. An OxyBlot Protein Oxidation Kit (Intergen) was used to test for the presence of these moieties. The carbonyl groups in the mitochondrial proteins (7 μg/μl) were derivatized to 2,4-dinitrophenylhydrazone (DNPH) by reaction with DNP. For negative controls, control solution was added instead of the DNPH solution. The DNP-derivatized protein samples were separated by PAGE under antioxidant conditions (412% NuPAGE Bis-Tris; Invitrogen) followed by blotting on the nitrocellulose membrane. The membrane was probed with a primary antibody (dilution 1:150) specific to the DNP moiety of the proteins. Binding of this antibody was detected by incubating the membrane with a 1:600 dilution of an anti-rabbit IgG conjugated to horseradish peroxidase. After incubation with a chemiluminescent substrate, primary antibody bound to DNP moiety was observed by autoradiography. To ensure equal loading among lanes, the membranes were stained with Ponceau S (Sigma) for 3 min until the bands appeared and then washed three times in PBS-Tween buffer. The intensities of the bands were visually confirmed to be similar across the lanes.
Western blot detection of specific posttranslational modifications of mitochondrial proteins by the Maillard reaction. Mitochondrial proteins were loaded at 14 μg/lane on duplicate 412% NuPAGE gels (Invitrogen, Carlsbad, CA). One gel of the duplicate was stained with Coomassie blue to assess the size distribution of mitochondrial proteins and to ensure equal loading among lanes. The second gel was electroblotted on a nitrocellulose membrane and probed with a monoclonal antibody to MGO-derived imidazole AGE (1H7G5) at 1:6,000 (47). The specific sites of antibody binding were detected with a 1:40,000 dilution of anti-mouse IgG conjugated to horseradish peroxidase. After the chemiluminescent substrate was added (Pierce), MGO-modified proteins were observed by autoradiography. Densitometric evaluation of the blots was carried out using the VersaDoc imaging system and Quantity One software (Bio-Rad).
Immunoprecipitation of mitochondrial proteins. Mitochondrial pellet (50 μl) was resuspended in 900 μl lysis buffer (20 mM Tris, pH 8.0, 1 mM EDTA, 200 mM NaCl, and 1% Triton X-100), sonicated, and ultracentrifuged at 150,000 g for 30 min. The collected supernatant with solubilized proteins was subjected to an immunoprecipitation procedure using Seize X Protein A Kit (Pierce), with the modification that we use protein G plus agarose beads (Santa Cruz Biotechnology, Santa Cruz, CA) because of their higher binding affinity for mouse antibodies. Briefly, 100 μg solubilized proteins were incubated overnight with 5 μg purified MGO-derived AGE imidazole antibody linked to the protein G support. Immunoprecipitated antigen was eluted and then subjected to one-dimensional PAGE as described.
2-Dimensional electrophoresis of mitochondrial proteins. Mitochondrial protein samples were prepared for 2-dimensional (2D) electrophoresis using a protocol described by Lopez et al. (39) and Molloy (43) with the following modifications. Briefly, 350 μl mitochondrial pellet (22 μg mitochondrial protein/μl) were resuspended in 350 μl rehydration buffer containing 7 M urea, 2 M thiourea, 1% CHAPS, 1% Triton X-100, sonicated, and then ultracentrifuged at 150,000 g for 30 min to remove particulate matter. The samples were diluted with rehydration buffer to a protein concentration of 1 μg/μl. Dithiothreitol (64 mM) and Bio-Lyte 3/10 ampholyte (1%) were added to the collected supernatant. The supernatant with solubilized mitochondrial proteins was used to rehydrate 7-cm ReadyStrip IPG strips (Bio-Rad). Strips were rehydrated passively with 180 μg mitochondrial proteins. Isoelectric focusing was performed at 20,000 Vh in a PROTEAN IEF cell (Bio-Rad). Strips were equilibrated in 1) NuPAGE LDS sample buffer, 10% NuPAGE sample reducing agent and then 2) 1x NuPAGE LDS sample buffer, 125 mM iodoacetamide (Invitrogen). The proteins resolved on IPG strips were applied to second-dimension gels (412% NuPAGE Novex Bis-Tris ZOOM Gel; Invitrogen). One gel was used for Coomassie stain and another one was electroblotted, probed with antibody (1H7G5), and visualized as above. To adequately separate the 1,000 proteins theorized to be present in mitochondria, the same sample was applied to three narrow-range overlapping gradient strips [isoelectric point (pI) 36, 58, 710] to improve protein separation by expanding a small pI range across the entire width of the gel (7 cm).
Identification of MGO-modified proteins. The MGO-modified proteins were identified by overlapping the appropriate Coomassie-stained one-dimensional and 2D gel with specific membranes probed with anti-MGO-derived imidazole antibody. Proteomic identification of MGO-modified proteins was performed as previously described (28). Spots were excised, destained, washed, and dehydrated before reduction with dithiothreitol and alkylation with iodoacetamide. The protease trypsin was driven into the gel pieces by rehydrating them in 30 μl of 20 ng/μl trypsin in 50 mM ammonium bicarbonate on ice for 10 min. Any excess trypsin solution was removed, and 20 μl of 50 mM NH4HCO3 were added. The sample was digested overnight at room temperature. The resulting peptides were extracted in two aliquots of 30 μl of 50% acetonitrile/5% formic acid, combined, and evaporated to <20 μl for LC-MS analysis.
Analyses were performed on a Finnigan LCQ-Deca ion trap mass spectrometer system with a Protana microelectrospray ion source interfaced to a self-packed 10 cm x 75 μm ID Phenomenex Jupiter C18 reversed-phase capillary chromatography column. The microelectrospray ion source was operated at 2.5 kV. The digest was analyzed using the data-dependent multitask capability of the instrument, acquiring full-scan mass spectra to determine peptide molecular weight and product ion spectra to determine amino acid sequence in successive instrument scans. This mode of analysis produces 1,000 collisionally induced dissociation (CID) spectra of ions ranging in abundance over several orders of magnitude. Not all CID spectra are derived from peptides.
The data were analyzed by using all CID spectra collected in the experiment to search the nonredundant protein sequence database of the National Center for Biotechnology Information with the search program TurboSequest. All matching spectra were verified by manual interpretation. The interpretation process was also aided by additional searches using the programs Mascot and Fasta.
Statistical methods. Between-group differences were determined by ANOVA, or t-test, with significance calculated by the method of Tukey, using JMP 5.0.1 (SAS, Cary, NC). Bivariate linear regression analysis was performed for continuous variables, calculating for the fitted line. Differences were considered significant at P < 0.05.
RESULTS
Glycemic Control
At every time point, plasma glucose (data not shown) and glycated Hb levels were significantly higher in rats with STZ diabetes compared with controls. At 2 mo, glycated Hb was 8.9 ± 1.1% in diabetes compared with 2.1 ± 0.9% in controls; at 6 mo, diabetes 9.9 ± 1.0%, controls 3.8 ± 0.8%; and at 12 mo diabetes 11.4 ± 1.0%, controls 5.8 ± 0.5. Treatment with aminoguanidine did not alter plasma glucose or glycated Hb levels (8.9 ± 2.3%).
Twelve Months of Diabetes Cause Diminution of Renal Mitochondrial Respiration
ADP-dependent NADH-linked respiration. The effect of the diabetic state on NADH-linked state 3 (ADP-dependent) respiratory rates was evaluated by using intact cortical renal mitochondria given glutamate and malate as substrates. Isolated mitochondria respond with a characteristic increase in the rate of oxygen consumption after the addition of ADP (state 3) and a leveling off of consumption with ADP exhaustion (state 4). Mitochondria isolated from 2 mo control rats had a state 3 respiratory rate of 210.33 ± 31.74 natom O?mg mitochondrial protein1?min1, a state 4 respiratory rate of 39.33 ± 4.58 natom O?mg mitochondrial protein1?min 1, and a respiratory quotient (RQ, state 3/state 4) of 5.36 ± 0.46 with glutamate and malate as substrates. Data ranges were similar to previously published observations (54, 55, 58).
Two months of diabetes did not alter renal cortical mitochondrial respiration (Fig. 1). In contrast, 12 mo of diabetes caused a significant inhibition of state 3 respiratory activity in the presence of NADH-linked substrates (P = 0.0003) compared with age-matched controls (Fig. 1A). Mitochondria isolated from 12 mo diabetic animals exhibited a significant increase in respiratory state 4 (P < 0.0001; Fig. 1B). RQ, the ratio of state 3 to state 4, reflects the respiratory function of the inner mitochondrial membrane and provides a measure of mitochondrial integrity. RQ declines with aging and 12 mo diabetes (P < 0.0001; Fig. 1C).
DNP-dependent NADH-linked respiration. Uncoupling agents collapse the proton gradient, thereby promoting maximum rates of mitochondrial respiration dependent on TCA and ERC activities, and independent of ATP synthase and ADP translocase. Renal mitochondria obtained from animals with 12 mo of chronic hyperglycemia demonstrated a significant inhibition of uncoupled respiration compared with controls (P = 0.002; Fig. 2A). This inhibition of uncoupled respiration was of the same magnitude as the inhibition of ADP-dependent respiration (20%), supporting the idea that long-term diabetes exerts its effect on the TCA or ERC rather than on the ATP synthase and ADP translocase.
NADH oxidase assay. The NADH oxidase assay (Fig. 2B) tests the ability of the ERC to utilize exogenously added NADH. Twelve months of diabetes was associated with significant inhibitions of ERC activity by 19% compared with control (P = 0.001). ERC activity was not affected by 2 mo of diabetes.
To further define the activity of NADH-linked ERC complexes, 12 mo control or diabetic renal mitochondria were exposed to various experimental conditions (20, 21). The activities of complexes I and IV of the ERC were not affected by chronic hyperglycemia. Studies of complex III revealed a 20% decrease in the activity in diabetic mitochondria isolated from 12 mo diabetic rats (P = 0.0006; Table 1).
View this table:
Diabetic Mitochondria are a Source of Oxidative Stress
We measured superoxide production by renal mitochondria isolated from normal and diabetic kidney. Superoxide production was quantitated for a period of 75 s in each individual mitochondrial sample oxidizing NADH. As expected, control mitochondria respond with an increase in the rate of superoxide production upon the addition of NADH and the initiation of the ERC activity, consistent with the accepted concept that 14% of oxygen consumed is released as free radicals in normal conditions. Figure 3A shows a representative experiment contrasting superoxide production by control and diabetic mitochondria. Mitochondria from diabetic animals produced superoxide at increased rates. Tiron scavenged superoxide produced by both control and diabetic mitochondria. Data for individual experiments used to generate curves similar to those shown in Fig. 3A were analyzed at 55 s. Mitochondria from animals with 12 mo of diabetes showed a 19.8% increase in superoxide production compared with their matched controls (P = 0.0003, data not shown).
Diabetic Mitochondria are a Target of Oxidative Stress
The formation of carbonyl groups is associated with protein oxidation. Using an OxyBlot immunodetection method, mitochondrial proteins over a broad range of molecular masses are shown to be targets of progressive oxidative modifications with aging and duration of diabetes. Four bands, consistently present in all samples(>62, 62, 49, and 38 kDa; Fig. 3B), were quantitated using densitometry. Twelve months of diabetes resulted in an increase in mitochondrial protein oxidation (P < 0.05 for >62-kDa band, and P < 0.001 for 62-kDa band; Fig. 3C).
Diabetes Causes Carbonyl-Induced Modifications of Mitochondrial Proteins
Western blot analysis using antibody against MGO-imidazole revealed the presence of three protein bands at 14, 28, and 38 kDa. However, the intensity of staining for the 38- and 14-kDa proteins was not significantly modified in diabetes compared with control (Fig. 4A). In contrast, the signal intensity of the 28-kDa band was markedly increased in all mitochondria isolated from animals with diabetes (Fig. 4A). The proteins at 14 and 38 kDa revealed positive staining in the absence of the antibody (data not shown), the result of either endogenous peroxidase activity or the presence of trace levels of contaminating metals in buffers used for routine Western blotting procedures (7).
To circumvent these issues, immunoprecipitation of those proteins having MGO-derived imidazole epitopes was performed. This approach identified only those proteins of 28 kDa molecular mass. As demonstrated by immunoprecipitation (Fig. 4B), the intensity of MGO-induced modifications increased with the duration of diabetes.
Signal intensity of the 28-kDa band obtained from each 12-mo animal was measured using densitometry. There was a significant negative correlation between the signal intensity and complex III activity in both control ( = 0.77, P < 0.03) and diabetic animals ( = 0.70, P < 0.05) animals (Fig. 4C).
Aminoguanidine Improves the Mitochondrial Respiration and Complex III Activity and Diminishes the Oxidative Modifications of Mitochondrial Proteins
Renal mitochondria showed a significant inhibition of state 3 respiratory activity in the presence of NADH-linked substrates at 6 mo diabetes compared with their age-matched controls (P < 0.01; Fig. 5A). Figure 5B shows that mitochondria isolated from rats with 6 mo diabetes exhibited a significant increase in respiratory state 4 (P = 0.005) and a decline in RQ (P < 0.01; Fig. 5C). Treatment with aminoguanidine resulted in an improvement of state 3 respiration, as shown in Fig. 5A, although aminoguanidine did not normalize respiratory state 4 (Fig. 5B) and had no influence on RQ (Fig. 5C). Importantly, aminoguanidine treatment normalized the diabetes-induced decline in complex III activity (Fig. 5D).
Western blot analysis revealed the presence of three MGO-imidazole containing proteins at 14, 28, and 38 kDa. Similar to the 12-mo time point, the 38- and 14-kDa proteins were not significantly modified in 6 mo diabetes compared with control (Fig. 6A). In contrast, the signal intensity of the 28-kDa band was significantly increased in mitochondria isolated from 6 mo diabetic animals. As expected, aminoguanidine treatment was associated with a significant decrease in MGO-induced modifications of proteins running at 28 kDa (P = 0.0004; Fig. 6B).
Mitochondrial proteins are a target of oxidative modifications as early as 6 mo of diabetes, as demonstrated by Fig. 7, A and B. Diabetes resulted in an increase in mitochondrial protein oxidation apparent in the density of the >62-, 62-, and 49-kDa bands (P = 0.005, P = 0.01, and P = 0.003, respectively). Aminoguanidine treatment reversed this pattern, decreasing oxidation associated with diabetes toward the normal level (Fig. 7, A and B).
Identification of MGO-Modified Proteins
Figure 8A shows a representative 58 pI gel from a series of 2D gel electrophoresis analyses of proteins in mitochondria isolated from the cortical kidney of 12 mo diabetic animals. Repetitive analysis of these samples gave similar and reproducible patterns of protein expression for each type of sample.
Subsequent LC-MS analysis permitted unambiguous identification of the proteins. The spots in Fig. 8A that are indicated by the arrows are proteins containing MGO-derived imidazole epitopes, as was found after Western blot analysis (Fig. 8B).
Digestion and analysis of these spots (Fig. 8A) identified the proteins with corresponding NCBI accession numbers, molecular masses, and pI values shown in Table 2. Commercial antibodies against these proteins are not available yet, so their identity was not confirmed by immunoprecipitation. However, these observations were established after repeated separation of proteins using both one-dimensional and 2D gel electrophoresis.
View this table:
DISCUSSION
The present study describes, for the first time, a direct relationship between the formation of intracellular AGEs on renal mitochondrial proteins, the decline in mitochondrial function, and the formation of ROS.
Nonenzymatic modifications of proteins by glucose to form AGEs has been linked to the development of diabetic nephropathy (5, 17, 24, 57, 69). The kidney may be uniquely vulnerable to this process because it is the target both of AGEs that form within renal cells and AGEs excreted in the urine. In addition, the main marker of Maillard-induced alterations in diabetes is the presence of MGO-induced hydroimidazolone on intracellular proteins (64).
In vitro studies in this laboratory have demonstrated that mitochondrial proteins are subcellular targets of dicarbonyls, particularly the highly reactive compound MGO. In renal mitochondrial suspension, MGO attaches to mitochondrial proteins, forms MGO-derived imidazole AGEs, and nearly instantaneously inhibits mitochondrial oxygen consumption by the ERC (55). There is evidence that MGO plays a role in protein structural changes and diabetic complications. MGO is markedly elevated in plasma from patients with diabetes (31, 40). Circulating MGO levels are correlated with the level of glycemia (3, 41) and associated with diabetic complications (45), including diabetic nephropathy (3).
The strategy to focus on MGO-derived imidazole modifications of mitochondrial proteins, through the use of the specific anti-imidazole AGE antibody, yielded six proteins in mitochondria from rats with 12 mo of diabetes. These proteins are known to be involved in two specific mitochondrial metabolic pathways associated with energy production: 1) oxidative phosphorylation and 2) fatty acid -oxidation.
Four of the MGO-modified proteins are components of the ERC. Two of the modifications were not associated with functional changes. Thus the G chain of F1-ATPase was shown to possess MGO-derived imidazole motifs, although there were no changes in the phosphorylation apparatus. In addition, the MGO-induced hydroimidazolone epitope present on the 30-kDa subunit of complex I was not associated with a change in the maximal activity of complex I.
In contrast, both proteomic and functional studies show that complex III is the key target of MGO-induced modifications. Complex III spans the central part of the respiratory chain, catalyzing electron transfer from ubiquinol to oxidized cytochrome c. This study shows that 2 out of 11 complex III components are modified by MGO in diabetes (core protein 1 and cytochrome c1). Based on the crystal structure of complex III, both proteins are exposed to the mitochondrial intermembrane space, which is conceptually more accessible to MGO from the cytoplasm. More than that, MGO may be formed within the intermembrane space in close proximity to these proteins by spontaneous dephosphorylation and decomposition of dihydroxyacetone phosphate, a compound of the glycerol 3-phosphate shuttle (51). Owing to this design, MGO may readily attach to the arginine residue of these two proteins. The role of core protein 1 in mitochondrial respiration is uncertain, but it may play a role in complex III assembly (4, 23) and in vectorial proton translocation activity (11, 53). Cytochrome c1 channels the electrons from the iron-sulfur protein (FeS) to cytochrome c. The tight interaction between the two type 1 hemes located on these two cytochromes suggests a direct heme-to-heme transfer, and it is mainly stabilized by electrostatic forces between amino acid residues (22, 65). A potential target is the arginine residue in the 166 position (Arg166 in yeast corresponds to Arg102 in mammals). This arginine seems to be essential for cytochrome c1 function, since it is highly conserved in phylogenesis and lies in the loop of the cytochrome c binding site of complex III, and direct mutagenesis of this site results in loss of function (1). Therefore, we hypothesize that, in diabetes, acquired modification of essential arginine residues of cytochrome c1 by MGO present in the intermembrane space may interfere with catalytic activity of these proteins. Imidazolone-modified cytochrome c1 may not be able to transfer electrons to cytochrome c, resulting in more reduced proximal components of complex III and increased pressure at the quinol binding site of complex III close to the outer face of the membrane (Qo site) where the electrons may be donated to molecular oxygen and form superoxide (34, 36).
This study has relied heavily on the use of an immunochemical method as a common procedure for the analysis of an MGO-induced protein glycation product imidazole, using a well-characterized monoclonal antibody against a specific AGE structure. However, this immunochemical method is restricted to detection and semiquantitation of a single marker compound of glycation. Yet, MGO-induced protein glycation products in vivo are heterogeneous. Therefore, the choice of this antibody is based on robust literature implicating MGO-induced AGE-imidazole as the major quantitative and qualitative modification in the glomeruli in patients with diabetes (64). However, other MGO-induced AGE products may be formed in high yields in long-term diabetes on mitochondrial proteins. More then that, because of steric constrains, not all AGE epitopes on a protein may be available for the interaction with the antibody. AGE immunoassays used in this study must be considered to yield semiquantitative results only.
The decline in renal mitochondrial respiration in diabetes, in response to complex I-related substrates, localizes the defect to complexes I, III, and IV. In addition, a decline of ERC activity (as demonstrated by the decreased utilization of exogenous NADH in diabetes) parallels the global diminution of mitochondrial respiration. In vitro, the ERC is much more sensitive to MGO modification than the TCA (55). Thus the presence of MGO-induced imidazole epitopes on critical protein components of complex III may explain specific inhibition of complex III in diabetes. Furthermore, inhibition of MGO-induced modifications with aminoguanidine restores mitochondrial ERC function. An important negative finding is the absence of MGO modifications of protein components of TCA (Table 2). In the cell, MGO formation is mainly external to the mitochondria (62). Because components of the ERC are localized to the mitochondrial inner membrane, they are likely to have increased vulnerability to modification by MGO in contrast to the TCA situated in the mitochondrial matrix.
The magnitude of the inhibition of complex III demonstrated in this study is sufficient to account for the diminution of whole mitochondrial respiration seen in the kidney in chronic diabetes. Metabolic expression of an existing defect in oxidative phosphorylation is dependent on a biochemical threshold that varies across very large ranges according to the tissue (56, 61). Normal kidney mitochondria have an excess of available complex III and demonstrate a high biochemical threshold after experimental inhibition (56). In contrast in long-term diabetes, the diminution of complex III activity parallels the magnitude of the decline in oxidative phosphorylation. We speculate that the threshold curve might be modified in chronic diabetes by both posttranslational modifications and a decrease in the quantity of components of complex III, as previously shown in another highly oxidative tissue, the heart (67).
Numerous reports have demonstrated that oxidative stress induced by diabetes plays a major role in the development of diabetic nephropathy (30). However, there is no consensus about the origin of mitochondrial ROS in diabetes. Our results suggest that the selective diabetes-related alterations in complex III may enhance the production of ROS. Electron transfer among the components of complex III occurs via a bifurcation of electron flow mediated by motion of FeS between cytochrome c1 and cytochrome b (36). After abstraction of an electron by FeS, the remaining ubisemiquinone donates the second electron to cytochrome b. The electron from FeS sequentially reduces cytochrome c1, then cytochrome c, via a docking interaction facilitated by small subunits of complex III. Thus impaired electron transfer in the FeS-cytochrome c1-cytochrome c loop of complex III, resulting from loss of catalytic activity of cytochrome c1, would lead to an increased relative reduction of cytochrome b and ubisemiquinone with generation of ROS via direct electron leak to molecular oxygen (68). The inhibition of complex III has been shown to be responsible for the creation of 7080% of superoxide in the ubiquinone cycle (8) or for triggering of a reverse electron transfer to complex I (37). This study clearly demonstrates that treatment with an inhibitor of AGE-induced posttranslational modifications of complex III components improves mitochondrial respiration and complex III activity and diminishes the oxidative damage to mitochondrial proteins. Aminoguanidine is the prototype of an efficient dicarbonyl scavenging agent that prevents the formation of AGEs from dicarbonyl precursors (63). In pharmacological doses, aminoguanidine has been shown to delay dramatically the development of chronic complications of diabetes (13, 48, 59). Aminoguanidine is a potent inhibitor of the inducible form of nitric oxide synthase with an IC50 of 31 μM (2) and can act as an antioxidant in vivo by decreasing oxidant-induced ROS production and lipid peroxidation products (18). However, the kinetics of scavenging lipid peroxidation products are 1,000-fold slower than for dicarbonyls (63). In addition, hydroxyl radicals are scavenged only at very high concentrations (110 mM; see Ref. 70). Therefore, the achievable concentrations of aminoguanidine in kidney (740 nmol/g in the whole kidney) are too low to inhibit nitrosylation of mitochondrial proteins (2) or to exert direct antioxidant effects (48). In contrast, aminoguanidine was shown to have a prooxidant effect (63). In summary, aminoguanidines extend beyond the scavenging of dicarbonyls. Because of nonspecific effects, the actions of this agent on mitochondrial ROS production must be interpreted with great caution in long-term models of diabetes. A stronger study would be to demonstrate the effects of MGO in an experimental model in which complete ablation of MGO production is possible, such as glyoxalase I transgenic animals. Experiments in these models are presently in the planning phase in our laboratory.
Existing literature suggests that the MGO-imidazolone modification identified on enzymes of fatty acid -oxidation are not likely to alter function. The active site architecture of mitochondrial enoyl-CoA hydratase (crotonase) reveals the presence of one glutamic acid as the main catalytic structure (14) rather than a more MGO-susceptible amino acid (such as arginine). Crotonase catalyzes the hydration of 2-trans-enoyl-CoA intermediates of fatty acid oxidation (15) and is readily upregulated in the presence of high levels of substrates as described in the diabetic heart (67). Electron transfer flavoprotein is a necessary electron acceptor for many mitochondrial dehydrogenases, including those involved in fatty acid degradation. The identification of MGO-induced modifications in the -oxidation pathway is likely to reflect a generalized upregulation of these enzymes in response to high levels of substrates.
The findings reported in this paper extend knowledge of the central role played by mitochondria in diabetes from acute hyperglycemia to a model of chronic disease. In cell culture, acute hyperglycemic conditions cause increased mitochondrial superoxide formation as a result of increased substrate, increased availability of reducing equivalents provided by the TCA to ERC, and mitochondrial membrane hyperpolarization (47). There is a threshold above which even a small increase in mitochondrial potential gives rise to a significant increase in hydrogen peroxide production (32). ROS production may derive from MGO-induced posttranslational modification of mitochondrial enzymes. Gredilla et al. (19) demonstrated that complexes I and III are the main sites of superoxide production by intact renal mitochondria. The direct correlation between the intensity of MGO-induced modifications and the inhibition of complex III in the kidney in chronic diabetes is consistent with prior work localizing ROS production to complex III in intact (10, 34, 60), aged (42), and ischemic-reperfused (36) heart mitochondria and in diabetic liver (34).
In conclusion, in chronic diabetes, mitochondrial proteins are prone to posttranslational modifications resulting from glycation and oxidation. Carbonyl-induced modifications are targeted to key protein components of major mitochondrial cycles, such as oxidative phosphorylation and -oxidation. In these studies, there is a statistically significant negative correlation between the extent of MGO-induced modifications and mitochondrial bioenergetics in the kidney of rats with long-term diabetes. In addition, functionally damaged mitochondria give rise to increased quantities of superoxide and oxidative damage to mitochondrial proteins. At 6 mo, an inhibitor of advanced glycation restores functional and structural alterations of renal mitochondria. Therefore, this study provides early evidence in support of the hypothesis that glycation of mitochondrial proteins may be the primary cause of mitochondrial dysfunction and oxidative stress in chronic diabetes.
GRANTS
This work was supported by supported by a fellowship grant from the Juvenile Diabetes Foundation International (M. G. Rosca), a grant from the Leonard B. Rosenberg Renal Research Foundation of the Center for Dialysis Care, Cleveland, Ohio, National Institutes of Health Grants ES-11461 (M. F. Weiss), DK-57733 (T. S. Kern, V. M. Monnier, and M. F. Weiss), and DK-45619 (M. F. Weiss), and a mentorship grant from the American Diabetes Association (V. M. Monnier).
ACKNOWLEDGMENTS
We thank Penny Erhard for technical expertise.
FOOTNOTES
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
REFERENCES
Ahmad Z and Sherman F. Role of Arg-166 in yeast cytochrome c1. J Biol Chem 276: 1845018456, 2001.
Alderton W, Cooper C, and Knowles R. Nitric oxide synthases: structure, function and inhibition. Biochem J 357: 593615, 2001.
Beisswenger P, Howell S, Nelson R, Mauer M, and Szwergold B. Alpha-oxoaldehyde metabolism and diabetic complications. Biochem Soc Trans 31: 13581363, 2003.
Berry E, Guergova-Kuras M, Huang L, and Crofts A. Structure and function of cytochrome bc complex. Annu Rev Biochem 69: 1005- 1075, 2000.
Bolton W, Cattran D, Williams M, Adler S, Appel G, Cartwright K, Foiles P, Freedman B, Raskin P, Ratner R, Spinowitz B, Whittier F, Wuerth J, and Action Investigator Group I. Randomized trial of an inhibitor of formation of advanced glycation end products in diabetic nephropathy. Am J Nephrol 24: 3240, 2004.
Brownlee M. Glycation and diabetic complications. Diabetes 43: 836841, 1994.
Buettner G. Use of ascorbate as test for the catalytic metals in simple buffers. Methods Enzymol 186: 125127, 1990.
Casteilla L, Rigoulet M, and Penicaud L. Mitochondrial ROS metabolism: modulation by uncoupling proteins. IUBMB Life 52: 181188, 2001.
Chance B, Sies H, and Boveris A. Hydroperoxide metabolism in mammalian organs. Physiol Rev 59: 527605, 1979.
Chen Q, Vazquez E, Moghaddas S, and Hoppel C. Production of reactive oxygen species by mitochondria. Central role of complex III. J Biol Chem 278: 3602736031, 2003.
Cocco T, Paola MD, Papa S, and Lorusso M. Localization of acidic residues involved in the proton pumping activity of the bovine heart mitochondrial bc1 complex. FEBS Lett 456: 3740, 1999.
Du Y, Miller C, and Kern T. Hyperglycemia increases mitochondrial superoxide in retina and retinal cells. Free Radic Biol Med 35: 14911499, 2003.
Du Y, Smith M, Miller C, and Kern T. Diabetes-induced nitrative stress in the retina, and correction by aminoguanidine. J Neurochem 80: 771779, 2002.
Engel C, Mathieu M, Zeelen J, Hiltunen L, and Wierenga R. Crystal structure of enoyl- coenzyme A (CoA) hydratase at 2.5 A resolution: a spiral fold defines the CoA- binding pocket. EMBO J 15: 51355145, 1996.
Fersht A. Enzyme Structure and Mechanisms. New York:Freeman, 1985.
Flarsheim C, Grupp I, and Matlib M. Mitochondrial dysfunction accompanies diastolic dysfunction in diabetic rat heart. Am J Physiol Heart Circ Physiol 271: H192H202, 1996.
Forbes J, Thallas V, Thomas M, Founds H, Burns W, Jerums G, and Cooper M. The breakdown of preexisting advanced glycation end products is associated with reduced renal fibrosis in experimental diabetes. FASEB J 17: 17621764, 2003.
Giardino I, Fard A, Hatchell D, and Brownlee M. Aminoguanidine inhibits reactive oxygen species formation, lipid peroxidation, and oxidant-induced apoptosis. Diabetes 47: 11141120, 1998.
Gredilla R, Phaneuf S, Selman C, Kendaiah S, Leeuwenburgh C, and Barja G. Short-term caloric restriction and sites of oxygen radical generation in kidney and skeletal muscle mitochondria. Ann NY Acad Sci 1019: 333342, 2004.
Humphries K and Szweda L. Selective inactivation of alpha-ketoglutarate dehydrogenase and pyruvate dehydrogenase: reaction of lipoic acid with 4-hydroxy-2-nonenal. Biochemistry 37: 1583515841, 1998.
Humphries M, Yoo Y, and Szweda L. Inhibition of NADH-linked mitochondrial respiration by 4-hydroxy-2-nonenal. Biochemistry 37: 552557, 1998.
Hunte C, Solmaz S, and Lange C. Electron transfer between yeast cytochrome bc1 complex and cytochrome c: a structural analysis (Abstract). Biochem Biophys Acta 1555: 21028, 2002.
Islam M, Tanaka M, Suzuki H, Torii K, Torii N, Hattori N, and Ozawa T. A complete cDNA sequence for core I protein subunit of human ubiquinol-cytochrome c reductase. Biochem Mol Biol Int 32: 797805, 1994.
Jerums G, Panagiotopoulos S, Forbes J, Osicka T, and Cooper M. Evolving concepts in advanced glycation, diabetic nephropathy, and diabetic vascular disease. Arch Biochem Biophys 419: 5562, 2003.
Kaneda K, Iwao J, Sakata N, and Takebayashi S. Correlation between mitochondrial enlargement in renal proximal tubules and microalbuminuria in rats with early streptozotocin- induced diabetes. Acta Pathol Jpn 42: 855860, 1992.
Kang Y, Edwards L, and Thornalley P. Effect of methylglyoxal on human leukemia 60 cell growth: modification of DNA, growth arrest and induction of apoptosis. Leuk Res 20: 397405, 1996.
Kaur G and Bhardwaj S. The impact of diabetes on CNS. Role of bioenergetic defects. Mol Chem Neuropathol 35: 119131, 1998.
Keightley J, Shang L, and Kinter M. Proteomic analysis of oxidative stress-resistant cells: a specific role for aldose reductase overexpression in cytoprotection. Mol Cell Proteomics 3: 167175, 2004.
Kelley D, Jing H, Menshikova E, and Ritov V. Dysfunction of mitochondria in human skeletal muscle in type 2 diabetes. Diabetes 51: 29442949, 2002.
Kitada M, Koya D, Sugimoto T, Isono M, Araki S, Kashiwagi A, and Haneda M. Translocation of glomerular p47phox and p67phox by protein kinase C-beta activation is required for oxiadtive stress in diabetic nephropathy. Diabetes 52: 26032614, 2003.
Knecht K, Feather M, and Baynes J. Detection of 3-deoxyfructose and 3-deoxyglucosone in human urine and plasma: evidence for intermediate stage of the Maillard reaction in vivo. Arch Biochem Biophys 294: 130137, 1992.
Korshunov S, Skulachev V, and Starkov A. High protonic potential actuates a mechanism of production of reactive oxygen species in mitochondria. FEBS Lett 416: 1518, 1997.
Koya D, Hayashi K, Kitada M, Kashiwagi A, Kikkawa R, and Haneda M. Effects of antioxidants in diabetes-induced oxidative stress in the glomeruli of diabetic rats. J Am Soc Nephrol 14: S250S253, 2003.
Kristal B, Jackson C, Chung HY, Matsuda M, Nguyen H, and Yu B. Defects at center P underlie diabetes-associated mitochondrial dysfunction. Free Radic Biol Med 22: 823833, 1997.
Ledenev A, Konstantinov A, Popova E, and Ruuge E. A simple assay of the superoxide generation rate with Tiron as an EPR-visible radical scavenger. Biochem Int 13: 391396, 1986.
Lesnefsky E and Hoppel C. Ischemia-reperfusion injury in the aged heart; role of mitochondria. Arch Biochem Biophys 420: 287297, 2003.
Liu Y, Fiskum G, and Schubert D. Generation of reactive oxygen species by the mitochondrial electron transport chain. J Neurochem 80: 780787, 2002.
Lo T, Westwood M, McLellan A, Selwood T, and Thornalley P. Binding and modification of proteins by methylglyoxal under physiological conditions. A kinetic and mechanistic study with N alpha-acetylarginine, N alpha-acetylcysteine, N alpha-acetyl-lysine, and bovine serum albumin. J Biol Chem 269: 3229932305, 1994.
Lopez M, Kristal B, Chernokalskaya E, Lazarev A, Shestopalov A, Bogdanova A, and Robinson M. High-throughput profiling of the mitochondrial proteome using affinity fractionation and automation. Electrophoresis 21: 34273440, 2000.
McLellan A, Phillips S, and Thornalley P. The assay of methylglyoxal in biological systems by derivatization with 1,2-diamino-4,5-dimethoxybenzene. Anal Biochem 206: 1723, 1992.
McLellan A, Thornalley P, Benn J, and Sonksen P. Glyoxalase system in clinical diabetes mellitus and correlation with diabetic complications. Clin Sci (Colch) 87: 2129, 1994.
Moghaddas S, Hoppel C, and Lesnefsky E. Aging defect at the Qo site of complex III augments oxyradical production in rat heart interfibrillar mitochondria. Arch Biochem Biophys 414: 5966, 2003.
Molloy M. Two-dimensional electrophoresis of membrane proteins using immobilized pH gradients. Anal Biochem 280: 110, 2000.
Monnier V, Sell D, Miyata S, Nagaraj R, Odetti P, and Lapolla A. Advanced Maillard reaction products as markers for tissue damage in diabetes and uraemia: relevance to diabetic nephropathy. Acta Diabetol 29: 130135, 1992.
Monnier V, Vishwanath V, Frank K, Elmets C, Sauthot P, and Kohn R. Relation between complications of type I diabetes mellitus and collagen-linked fluorescence. N Engl J Med 314: 403408, 1986.
Nishi S, Ueno M, Hisaki S, Iimo N, Iguchi S, Oyama Y, Imai N, Arakawa M, and Gejyo F. Ultrastructural characteristics of diabetic nephropathy. Med Electron Microsc 33: 6573, 2002.
Nishikawa T, Edelstein D, Xue LD, Yamagishi S, Matsumura T, Kaneda Y, Yorek A, Beebe D, Oates J, Hammes H, Giardino I, and Brownlee M. Normalizing mitochondrial superoxide production blocks three pathways of hyperglycemic damage. Nature 404: 787790, 2000.
Nyengaard J, Chang K, Berhorst S, Reiser K, Williamson J, and Tilton R. Discordant effects of guanidines on renal structure and function and on regional vascular dysfunction and collagen changes in diabetic rats. Diabetes 46: 94106, 1997.
Obineche E, Mensah-Brown E, Chadranath S, Ahmed I, Naseer O, and Adem A. Morphological changes in the rat kidney following long-term diabetes. Arch Physiol Biochem 109: 241245, 2001.
Papoulis A, Al-Abed Y, and Bucala R. Identification of N2-(1-carboxyethyl)guanine (CEG) as a guanine advanced glycation endproduct. Biochemistry 34: 648652, 1995.
Phillips SA and Thornalley PJ. The formation of methylglyoxal from triose phosphates. Investigation using a specific assay for methylglyoxal. Eur J Biochem 212: 101105, 1993.
Pitkanen S and Robinson B. Mitochondrial complex I deficiency leads to increased production of superoxide radicals and induction of superoxide dismutase. J Clin Invest 98: 345351, 1996.
Rago Jd Sohm F, Boccia C, Dujardin G, Trumpower B, and Slonimski P. A point mutation in the mitochondrial cytochrome b gene obviates the requirement for the nuclear encoded core protein 2 subunit in the cytochrome bc1 complex in Saccharomyces cerevisiae. J Biol Chem 272: 46994704, 1997.
Rogers K, Friend W, and Higgins E. Metabolic and mitochondrial disturbances in streptozotocin-treated Sprague-Dawley and Sherman rats (42323). Proc Soc Exp Biol Med 182: 167175, 1986.
Rosca M, Monnier V, Szweda L, and Weiss M. Alterations in renal mitochondrial respiration in response to the reactive oxoaldehyde methylglyoxal. Am J Physiol Renal Physiol 283: F52F59, 2002.
Rossignol R, Malgat M, Mazat J, and Letellier T. Threshold effect and tissue specificity. J Biol Chem 247: 3342633432, 1999.
Sakurai S, Yonekura H, Yamamoto Y, Watanabe T, Tanaka N, Li H, Rahman A, Myint K, Kim C, and Yamamoto H. The AGE-RAGE system and diabetic nephropathy. J Am Soc Nephrol 14: S259S263, 2003.
Simon N, Jolliet P, Morin C, Zini R, Urien S, and Tillement J. Glucocorticoids decrease cytochrome c oxidase activity of isolated rat kidney mitochondria. FEBS Lett 435: 2528, 1998.
Soulis-Liparota T, Cooper M, Papazoglou D, Clark B, and Jerums G. Retardation by aminoguanidine of development of albumiunuria, mesangial expansion, and tissue fluorescence in streptozotocin-induced diabetic rat. Diabetes 40: 13281334, 1991.
St-Pierre J, Buckingham J, Roebuck S, and Brand M. Topology of superoxide production from different sites in the mitochondrial electron chain. J Biol Chem 277: 4478444790, 2002.
Taylor R, Birch-Machin M, Bartlett K, Lowerson S, and Turnbull D. The control of mitochondrial oxidation by complex III in rat muscle and liver mitochondria. Implications for our understanding of mitochondrial cytopathies in man. J Biol Chem 269: 35233528, 1994.
Thornalley P. Advanced glycation and the development of diabetic complications. Unifying the involvement of glucose, methylglyoxal and oxidative stress. Endocrinol Metab 3: 149166, 1996.
Thornalley P. Use of aminoguanidine (Pimagedine) to prevent the formation of advanced glycation endproducts. Arch Biochem Biophys 419: 3140, 2003.
Thornalley P, Battah S, Ahmed N, Karachalias N, Agalou S, Babaei-Jadidi R, and Dawnay A. Quantitative screening of advanced glycation endproducts in cellular and extracellular proteins by tandem mass spectrometry. Biochem J 375: 581592, 2003.
Trumpower B. A concerted, alternating sites mechanism of ubiquinol oxidation by the dimeric cytochrome bc1 complex. Biochem Biophys Acta 1555: 166173, 2002.
Turko I, Li L, Kulwant S%