Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2009 May 19.
Published in final edited form as: Semin Nephrol. 2007 Nov;27(6):609–620. doi: 10.1016/j.semnephrol.2007.09.006

Metabolomics as an Extension of Proteomic Analysis: Study of acute kidney injury

Didier Portilla, Laura Schnackenberg, Richard D Beger
PMCID: PMC2684501  NIHMSID: NIHMS36079  PMID: 18061843

Abstract

While proteomics studies the global expression of proteins, metabolomics characterizes and quantifies their end products: the metabolites, produced by an organism under certain set of conditions. From this perspective it is apparent that proteomics and metabolomics are complementary and when joined allow a fuller appreciation of an organism’s phenotype. Our studies using 1H-nuclear magnetic resonance (NMR) spectroscopic analysis demonstrated the presence of glucose, aminoacids, and trichloroacetic acid cycle metabolites in the urine after 48 hr of cisplatin administration. These metabolic alterations precede changes in serum creatinine. Biochemical studies confirmed the presence of glucosuria, but also demonstrated the accumulation of nonesterified fatty acids, and triglycerides in serum, urine, and kidney tissue, in spite of increased levels of plasma insulin. These metabolic alterations were ameliorated by the use of fibrates. We propose that the injury-induced metabolic profile may be used as a biomarker of cisplatin-induced nephrotoxicity. These studies serve to illustrate that metabolomic studies add insight into pathophysiology not provided by proteomic analysis alone.

Introduction

Transcriptomics, proteomics, and metabolomics technologies comprise what is referred to as systems biology. The hope is that these “omics” technologies will identify translational biomarkers that are applicable to both preclinical investigations and in the clinical setting. This is especially true for markers of acute kidney injury where traditional serum markers such as blood urea nitrogen (BUN) and creatinine have proven to be insensitive and nonspecific (1). Metabolomics refers to the study of metabolite pool that exists within a cell, tissue, or biofluid under a particular set of conditions (2) while metabonomics has been defined as the “quantitative measurement of the dynamic metabolic reponse of living systems to pathophysiological stimuli or genetic modification” (3). Metabolomics, metabonomics, or global metabolic profiling use analytical technologies such as nuclear magnetic resonance (NMR) spectroscopy and mass spectrometry (MS) to identify and quantify as many metabolites as possible in cells, tissues, and biofluids. The quantified metabolites can then be evaluated statistically to determine biomarkers of health and disease status and to help elucidate mechanisms of drug metabolism and drug toxicity.

Typically, changes in metabolite levels represent the downstream changes in the genome and the proteome, but in reality they represent not only changes in gene function but also are influenced by factors related to environment and phenotype (2). Further, up or down regulation of a particular transcript does not necessarily result in an instantaneous change in enzymatic activity (4). This means that it is important to investigate the changes in gene transcripts, proteins, and metabolites over multiple time points in order to understand their interactions at a systems level. In this case, metabolic profiling studies have the advantage in terms of ease of sample collection of biofluids, preparation and analysis. Therefore, metabolic profiling derived data can be used in order to pinpoint time points of interest for further genomics and proteomics studies.

Metabolic profiling studies will be critical in forwarding the concept of personalized medicine since the metabolic profile encodes information about both the phenotype and the genotype (2). Phenotype is influenced by numerous environmental factors including health to disease status, gut microflora content, nutritional intake, and age among others. The bottleneck of metabolic profiling is trying to understand the link between metabolites and their phenotype or genotype information, which requires a strong bioinformatics component and the proper experimental design. This being said, an individual’s metabolic profile, therefore, reflects both genotype and phenotype information, which can be used by health care providers to provide more accurate health to disease assessment and individualized medical treatment.

Metabolomic study of cisplatin induced acute renal failure

Nephrotoxicity is a common side effect of cisplatin-based chemotherapeutic regimens (41–44). Our previous studies demonstrated reduced protein expression and enzyme activities of several kidney PPARα target genes in response to cisplatin nephrotoxicity, and also that the use of PPARα ligands such as fibrate compound Wy-14,643 protected renal function by preventing proximal tubule cell death in the animal models of ischemia-reperfusion and cisplatin-induced acute renal failure (ARF). We have examined the metabolic alterations that occur in ARF using high resolution 1H NMR spectroscopy coupled with pattern recognition (9). Experimental acute renal failure was induced in 8-to 10-week-old male mice (strain Sv129) using cisplatin administration. Animals were maintained on standard chow and, as indicated, a group of animals was fed with a special diet containing fibrate compound WY-14,643 (1%wt/wt) for 7 days prior to the induction of acute renal failure. Cisplatin was administered by a single intraperitoneal injection of 20 mg cisplatin/kg body weight. After the induction of renal failure, the animals were returned to their cages and allowed free access to food and water. Urine samples were collected for NMR analysis and biochemical measurements of metabolite formation at day 1, 2, 3 after cisplatin administration. Urine samples were analyzed on a Bruker Avance NMR spectrometer equipped with a triple resonance cryoprobe. A standard water suppression pulse program was employed for the presaturation of the water peak. In the studies where the PPARα ligand WY compound was used, urine samples obtained after 2 days of cisplatin injection were analyzed, since the urine metabolic profile at this point was quite different from the one obtained from control mice. Spectra were Fourier transformed, phased, baseline corrected, and intelligently binned using ACD/Labs 1D NMR Manager, version 9.0 (ACD/Labs, Toronto, Canada). Individual spectra were exported as jcamp files for further analysis by Chenomx Eclipse software. Chenomx Eclipse software (Chenomx, Edmonton, Canada) and databases were applied for identification of specific metabolites in the spectra.

Metabolic changes in urine samples by 1H-NMR

Figure 1 shows representative spectra from the urine of the same mouse obtained prior to dosing, and for three days following dosing with 20 mg cisplatin/kg body weight (BW). Visual inspection of the spectra shows changes in various urine metabolites after the mouse had been given a dose of cisplatin including increases in glucose, lactate, pyruvate, and 2-oxoglutarate 48 hours post-dosing. Principal component analysis (PCA) of the spectral integrals showed distinct clusters for control urine and for urine collected on days 1, 2, and 3 after cisplatin injection, as shown in Figure 2. The loadings plot (data not shown) indicated that increases in proton resonances associated with 2-oxoglutarate and creatinine were the major metabolites separating day 1 samples from control, while increases in the glucose and lactate chemical shifts drove the clustering of the day 2 and day 3 urine samples away from control samples. Analysis of the spectra showed that the concentration of 2-oxoglutarate increased over the first 48 hours after dosing followed by a decrease from 48–72 hours. Figure 3 shows the 3D PCA plot obtained from analysis of day 2 urine samples from four groups of four mice each. These groups included control mice shown in blue, mice administered a single dose of cisplatin shown in red, and two groups, one maintained on a special diet of WY-14,643 alone, and the other one maintained on the fibrate diet for 7 days prior to cisplatin injection shown in yellow and green respectively. The PCA plot shows that the four groups separate into distinct clusters, with the control group on the special diet clustered fairly close to the control group on a normal diet. The cisplatin group clustered below the three groups along the PC3 axis.

Figure 1.

Figure 1

Open in a new tab

Representative 1D 1H NMR spectra of C) Control and following administration of cisplatin (D1) 24 hours (D2) 48 hours post-dosing (D3) 72 hours post-dosing.

Figure 2.

Figure 2

Open in a new tab

PC scores plot showing the changes from control over a 72 hour period following administration of cisplatin (●) Control animals (Δ) 24 hours after dosing (■) 48 hours after dosing (+) 72 hours after dosing.

Figure 3.

Figure 3

Open in a new tab

3D PCA scores plot of NMR data of day two urine indicating the effects of the PPARα ligand (WS) on cisplatin-induced ARF. (●) Control animals (■) Animals fed WY (Δ) Animals administered cisplatin following feeding with WY (+) Animals administered cisplatin.

Quantitative analysis of select metabolites is shown in Table 1 and indicated that cisplatin treatment for 2 days induced significant changes in the urine levels of glucose (200-fold increase), alanine (12-fold increase), lactate (4-fold increase), leucine (7-fold increase), methionine (4-fold increase), 2-oxoglutarate (2-fold increase), pyruvate (3.5-fold increase), and valine (8-fold increase). Pretreatment with the diet containing PPARα ligand WY provided protection from the metabolic changes induced by cisplatin by reducing urine levels of glucose, lactate, methionine, 2-oxoglutarate, and pyruvate. The data also indicate that the PPARα ligand increased in urine levels of amino acids proline and tyrosine when compared to control or saline treated mice. In addition, metabolite analysis showed that 1-methylnicotinamide was elevated in both groups maintained on the diet containing PPARα ligand WY-16,463. This metabolite has recently been shown to be a biomarker for peroxisome proliferation (45).

Table 1.

Select urine metabolite concentrations measured by 1H NMR spectroscopy and normalized to the concentration of creatinine in the urine.

Metabolite Control N=7 Cisplatin Day 2 N=7 Wy N=7 Wy + Cisplatin Day 2 N=7
Alanine 0.02±0.02 0.25±0.13*
p = 0.003		 0.07±0.06 0.19±0.16*
p = 0.037
Glucose 0.30±0.30 60.98±34.82*
p = 0.004		 0.66±0.61 18.08±23.98
Lactate 0.21±0.30 0.89±0.95 0.43±0.37 0.57±0.31
Leucine 0.03±0.02 0.22±0.10*
p = 0.004		 0.04±0.01 0.28±0.36
Methionine 0.03±0.01 0.13±0.07*
p = 0.009		 0.02±0.01 0.05±0.07
1-Methylnicotinamide 0.11±0.06 0.07±0.02 0.24±0.12*
p = 0.038		 0.22±0.08*
p = 0.016
2-Oxoglutarate 7.97±5.37 15.79±2.78*
p = 0.008		 15.60±3.76*
p = 0.010		 6.77±3.21
Proline 0.26±0.18 0.41±0.22 0.46±0.42 0.50±0.41
Pyruvate 0.12±0.07 0.44±0.21*
p = 0.008		 0.22±0.08*
p = 0.043		 0.27±0.18
Trimethylamine 0.90±0.89 0.28±0.26 0.07±0.06*
p = 0.048		 0.10±0.09
Tyrosine 0.17±0.08 0.15±0.06 0.50±0.56 0.41±0.39
Valine 0.02±0.01 0.16±0.09*
p = 0.006		 0.02±0.01 0.25±0.36
Open in a new tab

Values are reported as mg metabolite/mg creatinine

Cisplatin increases nonesterified fatty acids in serum, urine, and kidney tissue

Since neutral lipid accumulation detected by oil-red-O stain represents the accumulation of nonesterified fatty acids (NEFA), triglycerides, diglycerides, and cholesterol esters, we next examined specifically the effects of cisplatin on NEFA levels measured in serum, urine and kidney tissue. As shown in Figure 4A, cisplatin-treated mice exhibited a time-dependent increase in NEFA levels not only in serum and urine, but also in kidney tissue. At day 3 there was a 3-fold increase in serum levels, and also at day 3 after cisplatin injection there was a 7-fold increase in kidney tissue levels of NEFA. Our assay did not detect NEFA in the urine samples obtained from saline treated mice. However, in cisplatin treated mice there was a detectable level of NEFA in the urine obtained at day 3 after cisplatin injection (0.15 ± 0.01 mEq/L/mg protein).

Figure 4.

Figure 4

Figure 4

Open in a new tab

Figure 4A. Effect of cisplatin on NEFA levels in serum, urine and kidney tissue. Mice were administered saline (control) or cisplatin (20 mg/kg body weight) by a single intraperitoneal injection. NEFA levels were measured at day1, day2, day3 after cisplatin injection in serum, urine, and kidney tissue homogenates as described in Methods. Bars correspond to means ± SE of at least 6 independent experiments under each condition. *p < 0.05, compared with control by unpaired student’s t-test.

Figure 4B Effect of cisplatin and fibrate (WY) on NEFA levels. Wild-type mice were fed with either a regular or WY-containing diet and then were given saline (control and WY groups) or cisplatin (cisplatin and cisplatin + WY groups). NEFA levels were measured at day 3 after cisplatin injection in serum, urine, and kidney tissue homogenates as described in Methods Results are expressed as the means ± SE. †P < 0.05 compared to cisplatin, *P < 0.05, compared with control by unpaired Student’s t-test.

PPARα ligand prevents cisplatin induced accumulation of nonesterified fatty acids

Our most recent studies have shown that the use of PPARα ligands prevents the development of cisplatin-induced proximal tubule cell death by preventing the inhibition of fatty acid oxidation (46). Therefore, we examined the effect of PPARα ligand on cisplatin-induced increased NEFA levels in serum, urine, and kidney tissue. At day 3 after cisplatin injection, there was a 3-fold increase in the serum levels of NEFA. In contrast, the group of mice that received the diet containing PPARα ligand and cisplatin exhibited comparable levels of serum NEFA to the mice treated with saline alone, as shown in Figure 4B. We also measured NEFA levels in urine samples of mice treated with PPARα ligand and cisplatin, and again similarly to saline treated mice, we were not able to detect measurable amounts of NEFA in the urine samples from these mice. We next examined the effect of PPARα ligand on cisplatin-mediated accumulation of NEFA in kidney tissue. In the group of mice that received the diet containing PPARα ligand and cisplatin, there was a 65% reduction in the levels of NEFA in kidney tissue when compared to cisplatin treated mice. These results are shown in Figure 4B.

Cisplatin increases triglyceride levels in serum and kidney tissue

Since triglycerides (TG) represent the major component of neutral lipids in kidney tissue, we measured TG levels in serum, urine and kidney tissue homogenates of mice treated with saline (control) and mice treated with cisplatin. As shown in Figure 7A we were able to detect TG only in the serum and kidney tissue homogenates. Our assay was not able to detect measurable amounts of TG in urine samples obtained from control or cisplatin treated mice. As shown in Figure 5A, cisplatin-treated mice exhibited not only a time-dependent increase in TG levels in serum, but also in kidney tissue. At day 3 there was a 3-fold increased in serum TG levels, and also at day 3 after cisplatin injection there was a 6-fold increase in TG levels in kidney tissue homogenates.

Figure 5.

Figure 5

Figure 5

Open in a new tab

Figure 5A. Effect of cisplatin on TG levels in serum, urine and kidney tissue. Mice were administered saline (control) or cisplatin (20 mg/kg body weight) by a single intraperitoneal injection. TG levels were measured in serum, urine and kidney tissues at day1, day2, day3 after cisplatin injection as described in Methods. Bars correspond to means ± SE of at least 6 independent experiments under each condition. **p < 0.005 compared with control by unpaired student’s t-test.

Figure 5 B. Effect of cisplatin and fibrate (WY) on TG levels. Wild-type mice were fed with either a regular diet or WY-containing diet and then were given saline (control and WY groups) or cisplatin (cisplatin and cisplatin+WY groups). TG levels were measured in serum, and kidney tissue homogenates at day 3 after cisplatin injection as described in Methods. Results are expressed as the means ± SE. †P < 0.05 compared to cisplatin, *P < 0.05, **P < 0.005 compared with control by unpaired Student’s t-test.

PPARα ligand prevents cisplatin induced accumulation of TGs

We next examined the effect of PPARα ligand on cisplatin-mediated increased levels of TG in serum, and kidney tissue. At day 3 after cisplatin injection, there was a 3-fold increase in the serum levels of TG. In contrast, the group of mice that received the diet containing PPARα ligand and cisplatin, exhibited a 70 % reduction in serum TG levels, when compared to cisplatin treated mice, as shown in Figure 5B. We also examined the effect of PPARα ligand on cisplatin-mediated accumulation of TG in kidney tissue. In the group of mice that received the diet containing PPARα ligand and cisplatin there was a 55% reduction in TG levels in kidney tissue when compared to cisplatin treated mice. These results are shown in Figure 5B.

Cisplatin-induced hyperglycemia and glucosuria precede cisplatin induced ARF

1H NMR analysis of urine samples revealed that glucose was present within the first 24–48 hrs post-dosing with cisplatin. In order to corroborate these findings, we measured glucose levels in serum, urine, and kidney tissue by an enzymatic colorimetric assay. As shown in Figure 6A, cisplatin induced a time-dependent increase in glucose levels in serum, urine, and kidney tissue. At day 3 after cisplatin injection there was a 3.5-fold increase in serum glucose levels when compared to saline treated mice. Cisplatin treated mice also exhibited a remarkable 75-fold increase in urine glucose levels, and a 2.5 fold increase in kidney tissue levels of glucose, when compared to saline treated mice.

Figure 6.

Figure 6

Figure 6

Open in a new tab

Figure 6A. Effect of cisplatin on glucose levels in serum, urine, and kidney tissue. Mice were administered saline (control) or cisplatin (20 mg/kg body weight) by a single intraperitoneal injection. Glucose levels were measured at day1, day2, day3 after cisplatin injection in serum, urine, and kidney tissue homogenates as described in Methods. Bars correspond to means ± SE of at least 6 independent experiments under each condition. *P < 0.05, **P < 0.005 compared with control by unpaired student’s t-test.

Figure 6B. Effect of cisplatin and fibrate(WY) on glucose levels. Wild-type mice were fed with either a regular or WY-containing diet and then were given saline (control and WY groups) or cisplatin (cisplatin and cisplatin+WY groups). Glucose levels were measured at day 3 after cisplatin injection in serum, urine, and kidney tissue homogenates as described in Methods Results are expressed as the means ± SE. †P < 0.05compared to cisplatin, *P < 0.05, **P < 0.005 compared with control by unpaired Student’s t-test.

PPARα ligand prevents cisplatin-induced hyperglycemia and glucosuria

Our most recent studies have shown that the use of PPARα ligands prevent the development of proximal tubule cell death during ARF by preventing not only fatty acid oxidation, but also by preventing the inhibition of PDC activity in the mitochondria (46). In addition to its stimulation of fatty acid oxidation and anti-apoptotic effects, recent reports suggest that PPARα may also play a role in the regulation of metabolic abnormalities associated with the metabolic syndrome (47). Therefore, we examined the effect of PPARα ligand on cisplatin-induced increased levels of glucose in serum, urine, and kidney tissue. As shown in Figure 6B, serum glucose levels of wild-type mice receiving a regular diet were significantly increased by cisplatin treatment. At day 3, there was a 3.5-fold increase in serum glucose levels. By contrast, cisplatin-treated mice receiving a PPARα ligand in their diet, exhibited comparable levels of serum glucose to the mice treated with saline alone. Similar to our NMR results, our biochemical analysis confirmed that cisplatin treated mice exhibited remarkable increases in urine glucose levels (75-fold) when compared to saline treated mice. In the group of mice that received the diet containing PPARα ligand and cisplatin, there was a 78% reduction in urine glucose levels when compared to cisplatin treated mice. We next examined the effect of PPARα ligand on cisplatin-mediated accumulation of glucose in kidney tissue. In the group of mice that received the diet containing PPARα ligand and cisplatin, there was a 70% reduction in glucose levels in kidney tissue homogenates, when compared to cisplatin treated mice. These results are shown in Figure 8B.

Effect of Cisplatin and Fibrate (WY) on serum levels of Insulin

In the next series of experiments, we measured the effects of cisplatin on serum insulin levels. Three days after one single injection of cisplatin, there were significant increases in serum insulin levels from control to cisplatin-treated mice of 0.506±0.11 to 0.926±0.14 μg/L ( N=6, p<.05). By contrast, cisplatin-treated mice receiving the diet containing PPARα ligand exhibited comparable levels of insulin to the mice treated with saline alone as shown in Table 2.

Table 2.

Effect of Cisplatin and Fibrate (WY) on serum levels of Glucose, Insulin, NEFA and TGs.

Serum Control 3 Day cisplatin WY WY+Cisplatin
Glucose(mg/dL) 145.87± 17.00 507.33+ 85.54** 91.2+12.3 142.75± 26.78†
Insulin(ug/l) 0.51±0.11 0.93±0.14 0.37±0.28 0.33±0.11
NEFA(mEq/L) 0.11±0.02 0.36±0.05* 0.06±0.02 0.11±0.04.†
TG(mg/dL) 25.33± 1.45 79.00± 9.46** 22.33± 0.33 42.33± 5.81†
Open in a new tab

Results are expressed as the means ± SE.

†

P < 0.05 compared to cisplatin,

*

P < 0.05,

**

P < 0.005 compared with control by unpaired Student’s t-test.

In summary, we find that cisplatin-induced ARF produces an endogenous metabolic profile revealed by 1H NMR analysis of urine. As indicated in Table 1, the most marked changes induced by cisplatin, and revealed by NMR spectroscopy occurred within the first 48 hours, and were characterized by increased concentrations of glucose, lactate, amino acids such as alanine, valine, leucine, methionine, and the presence of TCA cycle metabolites such as pyruvate and lactate in urine. Our results confirm previous studies where quantitative changes by NMR spectroscopic metabolite patterns provided information on the location and severity of toxic lesions. Those studies detected the presence of glucosuria and aminoaciduria after exposure to known proximal tubule nephrotoxins such as gentamicin, mercuric chloride, and D-serine (7, 8, 48). Glucosuria and enhanced excretion of amino acids are both strong indicators of proximal tubule damage in general, likely caused by impaired tubular reabsorption in the proximal tubule via sodium-dependent glucose transport. In the case of gentamicin, recent in vivo and in vitro studies performed in LLCPK1 cells, as well as in mouse kidney tissue, have shown that aminoglycoside antibiotics reduce glucose reabsorption in kidney tissue by reducing mRNA, protein expression, and function of the sodium-dependent glucose transporter (SGLT1), which is located in the apical membrane of the proximal tubule (49). Our most recent unpublished studies, using alpha-D-14C-glucopyranoside (AMG) to examine the effects of cisplatin on glucose uptake in LLCPK1 cells also lend support to the hypothesis that the inhibition of sodium dependent glucose transport represents an early intracellular event which precedes proximal tubule cell death when LLCPK1 cells are exposed to cisplatin. Therefore, it is quite possible that a common mechanism of proximal tubule nephrotoxicicy for both gentamicin and cisplatin relates to reduced expression and function of sodium-dependent glucose transporters. This mechanism of proximal tubule nephrotoxicity explains our findings by NMR analysis of the early presence of glucosuria and aminoaciduria in cisplatin-treated mice.

PPARα ligand prevents cisplatin-induced hyperglycemia and glucose accumulation in kidney tissue

Previous studies done in rats have shown that cisplatin-induced hyperglycemia is secondary to the presence of marked glucose intolerance, in association with an impaired insulin response, and abnormal glucagon response to a glucose stimulus (50, 51). Our studies corroborate those previous findings in mice treated with cisplatin, including the presence of severe hyperglycemia, as well as the presence of hyperinsulinemia, as shown in Table 2. In addition, our studies also confirmed the presence of an inappropriate insulin response when compared to the level of hyperglycemia induced by cisplatin in these mice. On the other hand, fibrate treatment significantly ameliorated cisplatin-induced hyperglycemia and hyperinsulinemia, and also prevented cisplatin-mediated accumulation of glucose in kidney tissue. These results suggest that fibrate treatment leads to an improved insulin action in kidney tissue. This beneficial effect of fibrates or PPARα ligands on hyperglycemia and insulin levels has been previously observed in animal models of diabetes, and also in diabetes-prone animals (52–54). Using Otsuka Long Evans Tokushima Fatty (OLEF) rats, Koh et al (55) have shown that fibrate treatment prevented the development of diabetes by several mechanisms which included: 1) reduced basal plasma insulin concentrations, 2) reduced adiposity, 3) improved peripheral insulin action, and 4) by exerting beneficial effects on pancreatic β cell function.

Cisplatin-mediated renal lipid accumulation is ameliorated by PPARα ligand

A previous study had reported that cisplatin nephrotoxicity in rats was accompanied by significant elevations in serum total cholesterol and TG concentrations, an effect that was not accompanied by injury to the liver (56). Our results extend those previous observations showing also accumulation of TG and NEFA in kidney tissue of cisplatin treated mice. The mechanisms by which cisplatin increases both NEFA and TG accumulation in kidney tissue likely reflect in part cisplatin-mediated inhibition of fatty acid oxidation. We have shown that cisplatin causes a significant reduction in mRNA levels and enzyme activity of mitochondrial medium chain acyl-CoA dehydrogenase (MCAD), and that the use of PPARα ligand WY prevented cisplatin-induced reduction of mRNA levels and enzyme activity of MCAD and ameliorated acute renal failure (46). By contrast, cisplatin treated PPARα null mice fail to demonstrate a protective effect of fibrate treatment and do not reverse cisplatin-mediated inhibition of MCAD (46). In our most recent study done in renal epithelial cells in culture, we found that cisplatin directly inhibited PPARα activity in these cells, and this event was accompanied by increased accumulation of NEFA (57). Pretreatment with fibrate prevented the inhibition of PPARα activity, and the accumulation of NEFA, and also prevented cisplatin-induced proximal tubule cell death, supporting our in vivo observations about the protective role of PPARα ligands in reducing accumulation of NEFA and preventing cisplatin-mediated nephrotoxicity.

In addition to the protective effects of fibrates in preventing cisplatin-mediated inhibition of fatty acid oxidation in kidney tissue, our study cannot rule out the presence of protective systemic effects of fibrates on kidney tissue lipotoxicity. For example, the most pronounced systemic effect of fibrates is to reduce plasma triglyceride-rich lipoproteins (TRLs). The hypotriglyceridemic action of fibrates involves combined effects on lipoprotein lipase (LPL) and apo-CIII expression (55), resulting in increased lipolysis. The induction of LPL expression occurs at the transcriptional level and is mediated by PPARα. In contrast to LPL, transcription of the apoCIII gene is inhibited by fibrates, resulting in decreased production of apoCIII in the liver and reduced serum TG levels (58). The repression of apoCIII gene expression by fibrates is mediated by PPARα (59). Previous studies also have shown that in addition to increased lipolysis, fibrates also increase the hepatic uptake of free fatty acids by inducing the expression of specific fatty acid transport proteins and by increasing the formation of acyl CoA esters by acyl CoA synthetase (60). Therefore, fibrates could potentially change free fatty acid metabolism from triglyceride synthesis to catabolism. Additional cellular mechanisms by which cisplatin increases the accumulation of TG in the proximal tubule are likely to be involved. In a recent study by Peters et al. (61), the authors examined the potential mechanisms by which proximal tubule injury leads to TG accumulation. Significant differences in the expression and function of TG synthetic and degradative pathways were found, depending on the in vivo or in vitro model of tubular cell damage used. A significantly increased expression in acyl coenzyme A:diacylglycerol acyltransferase (DGAT) expression was seen in the antimycin and endotoxin models, while reduced expression of TG lipase was seen in the glycerol model of ARF. Although we have not examined in detail the cellular mechanisms by which cisplatin induces TG accumulation in the proximal tubule, our recent gene array analysis of kidney tissue of mice treated with cisplatin show that mRNA levels of fatty acid synthetase (FAS) gene were significantly reduced by cisplatin. This observation suggests that reduced TG catabolism by cisplatin and increased accumulation of free fatty acids represent important mechanism(s) of TG accumulation in the proximal tubule.

In summary, our analysis by NMR spectroscopy demonstrates that exposure to cisplatin results in a marked change in the urinary metabolic profile that precedes changes in known biomarkers of nephrotoxicity, such as BUN and serum creatinine. Future analysis by UPLC/MS-based techniques should allow us to further differentiate the presence of spurious markers that merely reflect the administration of cisplatin, from genuine biomarkers of toxicity. In addition, our results further support the protective role of PPARα ligands on renal function by preventing systemic and renal alterations in glucose and lipid metabolism caused by cisplatin.

Systems Toxicology: Integrated Metabonomics and Proteomics in Liver tissue

While our studies focused solely on metabolomic analysis, others have sought to integrate metabolomic and proteomic analyses. Metabonomics, proteomics, and gene expression micro array platforms were employed in a systems biology study of acute hepatotoxcity of valproic acid (62). Pregnant CD-1 mice were injected subcutaneously with 600 mg/kg valproic acid or vehicle control. Urine, serum, and liver tissue were collected at 6, 12, and 24 hours after dosing. 1D proton NMR experiments were applied to the urine and aqueous extracts of terminal plasma and tissue samples using a Bruker 600 MHz NMR. Principle component analysis of the binned NMR spectral data of urine samples showed the VPA dosed groups were clustered away from the controls due to altered glucose concentrations in urine samples at 12 and 24 hours. NMR metabonomics procedures applied on aqueous liver tissue extracts showed altered glucose levels at 12 hours after VPA administration. Proteomics was applied to crude liver mitochondrial enriched fractions that were prepared from 24 hour control and VPA treated mouse liver homogenates. Approximately 30 μg from each pooled sample was dissolved in loading buffer and separated by SDS-PAGE gels. The SDS gel was into 38 bands were transferred to a 96-well plate. Gel bands were reduced with dithiothreitol, alkylated with iodoacetamide and digested with trypsin as previously described (63,64). The resultant 50 μl peptide pools were analyzed using nano LC/MSMS on a LCQ Deca XP Plus ion trap mass spectrometer. Samples were eluted with a 50 min gradient and MS/MS was performed on the top four ions in each MS scan using the data-dependent acquisition mode. Product ion data were searched against the European Bioinformatics Institute’s (EBI) mouse international protein index (Mouse IPI) protein database. Proteomics studies identified two proteins, glycogen phosphorylase and amylo-1,6-glucosidase, which were increased in dosed animals relative to control. Both of these proteins are involved in converting glycogen to glucose. The combined metabolomic and proteomic studies indicated a perturbation in the glycogenolysis pathway following administration of valproic acid.

Conclusion

The application of metabolomics to evaluate and monitor the presence of acute kidney disease is still under development. From the work presented in this review, it is apparent that a number of known and unknown metabolites that can be easily measured in urine, serum, and kidney tissue could provide a reliable indication of organ function. Validation of these metabolic biomarkers with respect to preclinical and clinical use, sensitivity, and specificity could provide additional tools in the detection of the onset and severity of kidney injury. Moreover, a significant opportunity exists to integrate metabolomic and proteomic analyses in the study of renal pathophysiology. The literature indicates that the two approaches have yet to be applied in a single study of renal function or pathology. The application of such a systems biology approach is likely to provide new insights into renal disease.

Acknowledgments

This work was supported by a VA Merit Award Grant to Dr Didier Portilla. We thank Jinchun Sun for her work on some of the Metabolomics results presented in this chapter. Jinchun Sun was supported in part by appointments to ORAU Research Program at the National Center for Toxicological Research administered by the Oak Ridge Associated Universities through an interagency agreement. ACD/Labs 1D NMR manager is part of “beta test” collaboration between the NCTR and ACD/Labs. The views presented in this article do not necessarily reflect those of the U. S. Food and Drug Administration.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  • 1.Vaidya VS, Bonventre JV. Mechanistic biomarkers for cytotoxic acute kidney injury. Expert Opin Drug Metab Toxicol. 2006;2:697–713. doi: 10.1517/17425255.2.5.697. [DOI] [PubMed] [Google Scholar]
  • 2.Fiehn O. Metabolomics--The link between genotypes and phenotypes. Plant Mol Biol. 2002;48:155–171. [PubMed] [Google Scholar]
  • 3.Nicholson JK, Lindon JC, Holmes E. ‘Metabonomics’: Understanding the metabolic responses of living systems to pathophysiological stimuli via multivariate statistical analysis of biological NMR spectroscopic data. Xenobiotica. 1999;29:1181–1189. doi: 10.1080/004982599238047. [DOI] [PubMed] [Google Scholar]
  • 4.Raamsdonk LM, Teusink B, Broadhurst D, Zhang N, Hayes A, Walsh MC, Berden JA, Brindle KM, Kell DB, Rowland JJ, et al. A functional genomics strategy that uses metabolome data to reveal the phenotype of silent mutations. Nature Biotechnol. 2001;19:45–50. doi: 10.1038/83496. [DOI] [PubMed] [Google Scholar]
  • 7.Williams RE, Jacobsen M, Lock EA. 1H NMR pattern recognition and 31P NMR studies with D-serine in rat urine and kidney, time- and dose-related metabolic effects. Chem Res Toxicol. 2003;16:1207–1216. doi: 10.1021/tx030019q. [DOI] [PubMed] [Google Scholar]
  • 8.Lenz EM, Bright J, Knight R, Westwood FR, Davies D, Major H, Wilson ID. Metabonomics with 1H-NMR spectroscopy and liquid chromatography-mass spectrometry applied to the investigation of metabolic changes caused by gentamicin-induced nephrotoxicity in the rat. Biomarkers. 2005;10:173–187. doi: 10.1080/13547500500094034. [DOI] [PubMed] [Google Scholar]
  • 9.Portilla D, Li S, Nagothu K, Megyesi J, Schnackenberg L, Safirstein RL, Beger RD. Metabolomic study of cisplatin-induced nephrotoxicity. Kidney Int. 2006;69:2194–2204. doi: 10.1038/sj.ki.5000433. [DOI] [PubMed] [Google Scholar]
  • 41.Naziroglu M, Karaoglu A, Aksoy AO. Selenium and high dose vitamin E administration protects cisplatin-induced oxidative damage to renal, liver and lens tissues in rats. Toxicology. 2004;195:221–230. doi: 10.1016/j.tox.2003.10.012. [DOI] [PubMed] [Google Scholar]
  • 42.Price PM, Safirstein RL, Megyesi J. Protection of renal cells from cisplatin toxicity by cell cycle inhibitors. Am J Physiol Renal Physiol. 2004;286:F378–F384. doi: 10.1152/ajprenal.00192.2003. [DOI] [PubMed] [Google Scholar]
  • 43.Atasoyu EM, Yildiz S, Bilgi O, Cermik H, Evrenkaya R, Aktas S, Gultepe M, Kandemir EG. Investigation of the role of hyperbaric oxygen therapy in cisplatin-induced nephrotoxicity in rats. Arch Toxicol. 2005;79:289–293. doi: 10.1007/s00204-004-0627-3. [DOI] [PubMed] [Google Scholar]
  • 44.Santos NA, Catao CS, Martins NM, Curti C, Bianchi ML, Santos AC. Cisplatin-induced nephrotoxicity is associated with oxidative stress, redox state unbalance, impairment of energetic metabolism and apoptosis in rat kidney mitochondria. Arch Toxicol. 2007 doi: 10.1007/s00204-006-0173-2. in press. [DOI] [PubMed] [Google Scholar]
  • 45.Delaney J, Hodson MP, Thakkar H, Connor SC, Sweatman BC, Kenny SP, McGill PJ, Holder JC, Hutton KA, Haselden JN, et al. Tryptophan-NAD+ pathway metabolites as putative biomarkers and predictors of peroxisome proliferation. Arch Toxicol. 2005;79:208–223. doi: 10.1007/s00204-004-0625-5. [DOI] [PubMed] [Google Scholar]
  • 46.Li S, Wu P, Yarlagadda P, Vadjunec NM, Proia AD, Harris RA, Portilla D. PPAR alpha ligand protects during cisplatin-induced acute renal failure by preventing inhibition of renal FAO and PDC activity. Am J Physiol Renal Physiol. 2004;286:F572–580. doi: 10.1152/ajprenal.00190.2003. [DOI] [PubMed] [Google Scholar]
  • 47.Chinetti-Gbaguidi G, Fruchart JC, Staels B. Role of the PPAR family of nuclear receptors in the regulation of metabolic and cardiovascular homeostasis: new approaches to therapy. Curr Opin Pharmacol. 2005;5:177–183. doi: 10.1016/j.coph.2004.11.004. [DOI] [PubMed] [Google Scholar]
  • 48.Lenz EM, Bright J, Knight R, Wilson ID, Major H. A metabonomic investigation of the biochemical effects of mercuric chloride in the rat using 1H NMR and HPLC-TOF/MS: Time dependent changes in the urinary profile of endogenous metabolites as a result of nephrotoxicity. Analyst. 2004;129:535–541. doi: 10.1039/b400159c. [DOI] [PubMed] [Google Scholar]
  • 49.Fleck C, Schwertfeger M, Taylor PM. Regulation of renal amino acid (AA) transport by hormones, drugs and xenobiotics - a review. Amino Acids. 2003;24:347–374. doi: 10.1007/s00726-002-0316-6. [DOI] [PubMed] [Google Scholar]
  • 50.Fleck C, Kretzschel I, Sperschneider T, Appenroth D. Renal amino acid transport in immature and adult rats during chromate and cisplatinum-induced nephrotoxicity. Amino Acids. 2001;20:201–215. doi: 10.1007/s007260170060. [DOI] [PubMed] [Google Scholar]
  • 51.Takamoto K, Kawada M, Usui T, Ishizuka M, Ikeda D. Aminoglycoside antibiotics reduce glucose reabsorption in kidney through down-regulation of SGLT1. Biochem Biophys Res Commun. 2003;308:866–871. doi: 10.1016/s0006-291x(03)01502-x. [DOI] [PubMed] [Google Scholar]
  • 52.Goldstein RS, Mayor GH, Rosenbaum RW, Hook JB, Santiago JV, Bond JT. Glucose intolerance following cis-platinum treatment in rats. Toxicology. 1982;24:273–280. doi: 10.1016/0300-483x(82)90009-9. [DOI] [PubMed] [Google Scholar]
  • 53.Goldstein RS, Mayor GH, Gingerich RL, Hook JB, Rosenbaum RW, Bond JT. The effects of cisplatin and other divalent platinum compounds on glucose metabolism and pancreatic endocrine function. Toxicol Appl Pharmacol. 1983;69:432–441. doi: 10.1016/0041-008x(83)90266-1. [DOI] [PubMed] [Google Scholar]
  • 54.Jia D, Otsuki M. Bezafibrate, a peroxisome proliferator-activated receptor (PPAR)-alpha activator, prevents pancreatic degeneration in obese and diabetic rats. Pancreas. 2003;26:286–291. doi: 10.1097/00006676-200304000-00013. [DOI] [PubMed] [Google Scholar]
  • 55.Koh EH, Kim MS, Park JY, Kim HS, Youn JY, Park HS, Youn JH, Lee KU. Peroxisome proliferator-activated receptor (PPAR)-alpha activation prevents diabetes in OLETF rats: comparison with PPAR-gamma activation. Diabetes. 2003;52:2331–2337. doi: 10.2337/diabetes.52.9.2331. [DOI] [PubMed] [Google Scholar]
  • 56.Bihan H, Rouault C, Reach G, Poitout V, Staels B, Guerre-Millo M. Pancreatic islet response to hyperglycemia is dependent on peroxisome proliferator-activated receptor alpha (PPARalpha) FEBS Lett. 2005;579:2284–2288. doi: 10.1016/j.febslet.2005.03.020. [DOI] [PubMed] [Google Scholar]
  • 57.Nagothu KK, Bhatt R, Kaushal GP, Portilla D. Fibrate prevents cisplatin-induced proximal tubule cell death. Kidney Int. 2005;68:2680–2693. doi: 10.1111/j.1523-1755.2005.00739.x. [DOI] [PubMed] [Google Scholar]
  • 58.Abdel-Gayoum AA, El-Jenjan KB, Ghwarsha KA. Hyperlipidaemia in cisplatin-induced nephrotic rats. Hum Exp Toxicol. 1999;18:454–459. doi: 10.1191/096032799678840255. [DOI] [PubMed] [Google Scholar]
  • 59.Staels B, Vu-Dac N, Kosykh VA, Saladin R, Fruchart JC, Dallongeville J, Auwerx J. Fibrates downregulate apolipoprotein C-III expression independent of induction of peroxisomal acyl coenzyme A oxidase. A potential mechanism for the hypolipidemic action of fibrates. J Clin Invest. 1995;95:705–712. doi: 10.1172/JCI117717. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 60.van Dijk KW, Rensen PC, Voshol PJ, Havekes LM. The role and mode of action of apolipoproteins CIII and AV: synergistic actors in triglyceride metabolism? Curr Opin Lipidol. 2004;15:239–246. doi: 10.1097/00041433-200406000-00002. [DOI] [PubMed] [Google Scholar]
  • 61.Peters JM, Hennuyer N, Staels B, Fruchart JC, Fievet C, Gonzalez FJ, Auwerx J. Alterations in lipoprotein metabolism in peroxisome proliferator-activated receptor alpha-deficient mice. J Biol Chem. 1997;272:27307–27312. doi: 10.1074/jbc.272.43.27307. [DOI] [PubMed] [Google Scholar]
  • 62.Schnackenberg LK, Jones RC, Thyparambil S, Taylor JT, Han T, Tong W, Hansen DK, Fuscoe JC, Edmondson RD, Beger RD, Dragan YP. An Integrated Study of Acute Effects of Valproic Acid in the Liver Using Metabonomics, Proteomics, and Transcriptomics Platforms . OMICS. 2006;10:1–14. doi: 10.1089/omi.2006.10.1. [DOI] [PubMed] [Google Scholar]
  • 63.Edmonson RD, Vondriska TM, Biederman KJ, Zhang J, Jones RC, Zheng Y, Allen DL, Xiu JX, Cardwell EM, Pisano MR, Ping P. Protein kinase C epsilon signaling complexes include metabolism- and transcription/translation-related proteins: complimentary separation techniques with LC/MS/MS. Mol Cell Proteomics. 2002;1:421 – 433. doi: 10.1074/mcp.m100036-mcp200. [DOI] [PubMed] [Google Scholar]
  • 64.Gambus A, Jones RC, Sanchez-Diaz A, Kanemaki M, van Deursen F, Edmonson RD, Labib K. GINS-dependent replisome progression complex controls the advance of eukaryotic DNA replication forks. Nature Cell Biology. 2006;8:358–366. doi: 10.1038/ncb1382. [DOI] [PubMed] [Google Scholar]

RESOURCES