mass spectrometry

Environmental Toxicology and Pharmacology 46 (2016) 234–240

Contents lists available at ScienceDirect

Environmental Toxicology and Pharmacology journal homepage: www.elsevier.com/locate/etap

A metabolic profiling analysis of the nephrotoxicity of acyclovir in rats using ultra performance liquid chromatography/mass spectrometry Wenmin Xing a , Lili Gu b,1 , Xinyue Zhang c , Jiadong Xu b , Hong Lu b,∗ a b c

Zhejiang Provincial Key Lab of Geriatrics and Geriatrics Institute of Zhejiang Province, Zhejiang Hospital, Hangzhou 310013, China School of Pharmacology, Zhejiang Chinese Medical University, Hangzhou 310053, China Institute of Materia Medica, Zhejiang Academy of Medical Sciences, Hangzhou 310013, China

a r t i c l e

i n f o

Article history: Received 17 March 2016 Received in revised form 14 June 2016 Accepted 27 July 2016 Available online 28 July 2016 Keywords: Acyclovir Nephrotoxicity Dose-time-effect relationship Metabonomics

a b s t r a c t Acyclovir (ACV) exposure is a common cause of acute kidney injury (AKI). The toxicity mechanism of ACV has always been a matter of debate. The present study investigated into the time-effect relationship and dose-effect relationship of ACV-induced nephrotoxicity in rats using metabonomics. Twenty-four rats were randomly divided into four groups: a 0.9% NaCl solution group, and 100, 300, and 600 mg/kg ACV-treated groups; the ACV or vehicle solution was administered with a single intravenous injection. Urine was collected at different time periods (12 h before administration, and 0–6 h, 7–12 h, and 13–24 h after administration). Routine urinalysis was conducted by a urine automatic analyzer. Renal markers, including urine urea nitrogen, urine creatinine, and urinary N-acetyl-␤-d-glucosaminidase (NAG) activity, were determined using established protocols. Urinary metabolites were evaluated using ultra performance liquid chromatography/mass spectrometry (UPLC/MS). In the ACV-treated rats, increased levels of protein (PRO), occult blood (BLD), white blood cell (WBC), and NAG activity in urine were observed, while the urine creatinine and urea nitrogen levels showed a decrease compared with the control. Moreover, urine metabolites significantly changed after the treatment with ACV, and all the effects induced by ACV were dose–time dependent. Finally, 4 metabolites (guanine, 4-guanidinobutyric acid, creatinine, and urea) were identified, which can be used for further research on the mechanism of ACV-induced nephrotoxicity. © 2016 Published by Elsevier B.V.

1. Introduction Acyclovir [9-(2-hydroxyethoxymethyl) guanine] (ACV), which is an acyclic nucleoside, has been widely used for the treatment of viral infections, such as herpes simplex, herpes zoster, and hepatitis B. However, with the extensive clinical application of ACV, adverse drug reactions (ADRs), especially acute renal injury, have rapidly increased (Fleischer and Johnson, 2010; Obada et al., 2010). The Chinese State Food and Drug Administration has also issued a drug use warning for ACV (Monitoring, 2009). Therefore, evaluating its safety is imperative. It has been widely believed that ACV-induced nephrotoxicity is secondary to crystalluria (Lyon et al., 2002; Sawyer et al., 1988). For example, as reported by Giustina, treatment with low-dose

∗ Corresponding author at: Chinese Medical University, No.548 of Bin-wen Road, BinJiang District, ZheJiang, China. E-mail address: [email protected] (H. Lu). 1 First co-author. http://dx.doi.org/10.1016/j.etap.2016.07.018 1382-6689/© 2016 Published by Elsevier B.V.

intravenous ACV (5 mg/kg/d for 2 days) caused acute renal failure (Giustina et al., 1988). In addition, there have also been several reports of ACV-induced nephrotoxicity with biopsy evidence of tubular damage in the absence of crystal formation (Ahmad et al., 1994; Vachvanichsanong et al., 1995; Vomiero et al., 2002). These data may suggest that ACV induces direct insult to tubular cells. In our previous work, we found that administration of ACV at a human clinical equivalent dosage (150 mg/kg) for 9 days led to renal dysfunction and an increase in both the serum creatinine and blood urea nitrogen levels and fibrous proliferation in the glomerulus and renal tubules(Lu et al., 2014). However, no crystal formation was found. Therefore, metabolomics is ideally suited to address this issue through overview the whole changes of the metabolism related to ACV-induced nephrotoxicity. This research results may support the viewpoint that ACV induces direct insult to renal tubular cells. Metabolites considered to be relevant to kidney disease have been analyzed in a distinctly ‘low-tech’ manner by physicians since the Middle Ages (Weiss and Kim, 2012). Metabonomics is now defined as the quantitative measurement of the dynamic

W. Xing et al. / Environmental Toxicology and Pharmacology 46 (2016) 234–240

multi-parametric metabolic responses of living systems to pathophysiolo- gical stimuli or genetic modifications, which is widely used to characterize the biochemical patterns of the endogenous metabolites in cells, body fluids or tissues (Zhao, 2013). It found a series of molecules such as amino acids, lipids or nucleic acids, which provides an overview of metabolism related to disease or drug exposure. Recently, metabolomics has evolved into a valuable tool in nephrotoxicity, and play an important role in providing novel and specific biomarkers for renal injury (Zhao and Lint, 2014). Therefore, mang urinary and kidney metabolomics was performed in gentamicin, cisplatin, or tobramycin-induced nephrotoxicity (Boudonck et al., 2009). For example, in gentamicin-induced nephrotoxicity, amino acid and sugar excretion was increased prior to histopathologically evident kidney injury (Portilla et al., 2006; van de Poll et al., 2004). In the present study, we studied the changes of renal damage indexes and urinary metabolic profiles in rats induced by ACV administration (in different concentrations and for different durations) used metabolomics technology to elucidate the time-effect relationship and dose-effect relationship of biomarkers to ACVinduced nephrotoxicity. These results may provide information for further study on the mechanism of ACV-induced nephrotoxicity. 2. Materials and methods 2.1. Animals and treatment Twenty-four male SD rats (180–220 g) bred in the animal house of Zhejiang Experimental Animal Center were used for the experiment. All animals were housed in standard animal enclosures with a regulated temperature (22–24 ◦ C), relative humidity (60–80%), and 12 h light/dark cycle. Food and tap water were provided ad libitum. All experiments were carried out according to the guidelines of China for the care and use of laboratory animals. Each rat was randomly assigned to 1 of 4 experimental groups (each n = 6): high-dose ACV (600 mg/kg, slightly higher than the clinical highest dosage for humans), medium-dose ACV (300 mg/kg), low-dose ACV (100 mg/kg, the human clinical equivalent dosage), and a control group. ACV was administered intravenously to the rats once. ACV (Wuhan Humanwell Pharmaceutical Co., Ltd., China) was diluted in a 0.9% sodium chloride injection solution (Huadong Pharmaceutical Co., Ltd., Hangzhou, China). The control group received an equal volume of 0.9% sodium chloride injection solution. 2.2. Collection and preparation of urine samples The urine samples were collected at different time points (12 h pre-dose, and 0–6 h, 7–12 h, and 13–24 h after administration). Some samples were used for analyzing renal function, while the other samples were stored at −80 ◦ C for the purpose of UPLC/MS analysis. Prior to the UPLC/MS analysis, 200 ␮l of urine samples was thawed and then added to 500 ␮l of double-distilled water. Urine samples were then centrifuged at 12 000 rpm for 8 min at 4 ◦ C. Supernatant was transferred to a new tube and then filtered through 0.22 ␮m nylon filters for UPLC/MS analysis. 2.3. UPLC/MS analysis of urine samples UPLC/MS analysis was performed using an Agilent 1260 infinity ultra performance liquid chromatograph coupled to an Agilent 6520 Q-TOF/MS spectrometer with a dual ESI source (Agilent Technologies, Santa Clara, CA, USA). The system used an Agilent Eclipse Plus C18 column (2.1 mm × 50 mm, 1.8 ␮m). The column was maintained at 35 ◦ C, and a 5-␮l aliquot of the sample was introduced to the column. The UPLC mobile phase consisted of 0.1% formic acid (Tedia, Fairfield, OH, USA) in water purified by a Milli-Q water

235

purification system (Millipore, Bedford, MA, USA) (solution A) or acetonitrile (Tedia, Fairfield, OH, USA) (solution B). The linear gradient increased from 2% to 10% B in 2 min, and increased to 45% B in another 9.5 min. Then solution B was continuously increased from 45% to 60% in 1.5 min, and increased to 90% B in 4 min. Finally, the mobile phase was kept at 90% B for 2 min. Mass spectra were obtained on full-scan operation in positive ion mode with an m/z range of 50–1000. The capillary voltage was set at 4000 V. The skimmer voltage was set at 60 V and the fragmentor voltage at 150 V. Nitrogen flow was 10 l/min, and the gas temperature was 350 ◦ C. 2.4. Measurement of renal function Blood was collected 24 h after administration from the abdominal aorta; the kidneys were extracted and weighed immediately after urine collection. The renal index was calculated as kidney weight/body weight × 100 (%). Blood samples were then centrifuged at 3000 rpm for 10 min at 4 ◦ C, and urine samples were centrifuged at 5000 rpm for 6 min at 4 ◦ C. The concentration of creatinine and urea nitrogen in urine and blood and the activity of NAG in urine were measured using commercial kits (Jiancheng Bioengineering Institute, Nanjing, China) according to the manufacturer’s instructions. Qualitative examination of urine was performed by a urine analyzer (AU-4290, Arkray, Japan), including urine specific gravity (SG), pH, WBC, nitrite (NIT), PRO, glucose (GLU), ketones (KET), urobilinogen (URO), bilirubin (BIL), and BLD. 2.5. Data analysis Data were expressed as mean ± SD. Statistical analysis was performed using a t-test with SPSS data analysis software version 13.0. P-values ≤0.05 were considered significant. For UPLC/MS, XCMS software (Fraga et al., 2010) was used for raw peaks extraction, data baseline filtering, and peak alignment, and a standard normalization method was used in this data analysis. The resulting three-dimensional data involving the peak number, sample name, and normalized peak area were fed into SIMCA-P (version 11.0, Sweden) for principal component analysis (PCA) and partial least squares discriminant analysis (PLSDA). For the identification of potential markers, once potential m/z was obtained, the biomarkers were first derived by searching comparison with free online databases, such as ChemSpider (http://www.chemspider.com), the Human Metabolome Database (HMDB) (http://www.hmdb.ca), Metlin (http://metlin.scripps.edu/ ) and Pubchem (http://pubchem.ncbi.nlm.nih.gov/). Finally, the biomarkers were further confirmed by retention times and MS/MS spectra. 3. Results 3.1. Renal biomarker changes in the ACV-treated rats As shown in Fig. 1, ACV significantly decreased urine creatinine concentration as well as urine urea nitrogen concentration compared with the control (P < 0.01, P < 0.05). ACV-treated rats exhibited a significant increase in NAG activity compared with the control rats (Fig. 2A) (P < 0.01, P < 0.05). The most obvious changes were observed 0–6 h after administration. Moreover, all the effects induced by ACV were dose dependent. In addition, obvious changes in serum creatinine levels and blood urea nitrogen levels were also observed 24 h after ACV administration (Fig. 2C, D) (P < 0.01, P < 0.05). The renal index significantly increased in the ACV-exposed groups compared with the control (Fig. 2B, P < 0.01). These data suggested that ACV treatment induced renal damage in SD rats.

236

W. Xing et al. / Environmental Toxicology and Pharmacology 46 (2016) 234–240

Fig. 1. Effects of ACV on urine creatinine (A) and urine urea nitrogen (B) of rats (x ± s, n = 6). * compared with control (P < 0.05), ** compared with control (P < 0.01).

Fig. 2. Effects of ACV on urinary NAG activity, serum creatinine, and blood urea nitrogen, kidney coefficient of rats (x ± s, n = 6). * compared with control (P < 0.05), ** compared with control (P < 0.01). B: change of kidney coefficient of rats 24 h after administration of ACV. C and D: changes of blood urea nitrogen and creatinine of rats 24 h after administration of ACV.

3.2. Urinalysis SG, pH, NIT, GLU, KET, URO, and BIL were unaltered in urine after administration of ACV compared with the control group. However, PRO, BLD, and WBC were detected in several rats after administration of ACV (Tables 1–3 ). The positive results for PRO, BLD, and WBC appeared rapidly 0–6 h after the administration of ACV. All the effects were in a dose-dependent manner. 3.3. Urinary metabolomic analysis The typical base peak intensity chromatograms of urinary samples was presented from the control and ACV nephrotoxicity rats

(Supplementary materials 1), which indicated that the method could be used for the analysis of samples. When all the samples were grouped according to different doses, the PCA score plot (Fig. 3) and PLS-DA score plot (Fig. 4) of LC/MS spectral data showed a clear dose-dependent separation of controls and ACV-treated animals. The high-dose group was farthest away from the control group; the low-dose group was closest to the control group; and the medium-dose group was located between the high-dose group and the low-dose group. When all the samples were grouped according to different time periods, the PCA score plot (Fig. 5) and PLS-DA score plot (Fig. 6) could be readily divided into four clusters. The group 0–6 h after the administration of ACV was the farthest away from the control group, which indicated that the

Table 1 Effects of ACV on PRO of Rats. group

control group 100 mg/kg 300 mg/kg 600 mg/kg

urine protein (PRO) 12 h pre-dose

0–6 h

7–12 h

13–24 h

− − +(1) +(1)

+(4), ++++(2) +(2), +++(1), ++++(3) +++(2), ++++(4) +++(3), ++++(3)

+(2) +(2), ++++(4) +(2), ++++(4) +(2), +++(1), ++++(2)

+(3), ++++(2) +(5), ++++(1) +(4), ++++(1) +(5)

The values of the brackets mean the number of rats. "−" indicates a negative degree; "+,” “++,” “+++,” and “++++" indicate a positive degree. Urinary protein level: −: urine protein < 0.1 g/L; +: 0.1 g/L ≤ urine protein < 1.0 g/L; ++: 1.0 g/L ≤ urine protein < 2.0 g/L; +++: 2.0 g/L ≤ urine protein < 4.0 g/L; ++++: urine protein > 4.0 g/L.

W. Xing et al. / Environmental Toxicology and Pharmacology 46 (2016) 234–240

237

Table 2 Effects of ACV on BLD of Rats. group

control group 100 mg/kg 300 mg/kg 600 mg/kg

urine occult blood (BLD) 12 h pre-dose

0–6 h

7–12 h

13–24 h

− − − −

− +(1), ++(1) +(3) +(3), +++(3)

− − +(1) ++(1), ++++(1)

+(1) − +(1) +(2), ++(1)

The values of the brackets mean the number of rats. "−" indicates a negative degree; "+,” “++,” “+++,” and “++++" indicate a positive degree. Urine occult blood level: −: occult blood < 10 cells/␮L; +: 10 cells/␮L ≤ occult blood < 20 cells/␮L; ++: 20 cells/␮L ≤ occult blood < 30 cells/␮L; +++: 30 cells/␮L ≤ occult blood < 40 cells/␮L; ++++: occult blood > 40 cells/␮L. Fig. 6. PLS-DA score plot of LC/MS spectral data of urine samples obtained 0–6 h after administration (red), 7–12 h after administration (green), 13–24 h after administration (blue) and before administration (black). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 3. PCA score plot of LC/MS spectral data of urine samples obtained from the ACV 600 mg/kg group (red), 300 mg/kg group (green), 100 mg/kg group (blue), and control group (black). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) Fig. 7. S-plot showing the time-dependent contribution to the clustering of different groups. The p[1] axis represents the magnitude of the spectral bins. Among them, bins with high p(corr)[1] <−0.2 for increased and >0.2 for decreased.

Fig. 4. PLS-DA score plot of LC/MS spectral data of urine samples obtained from the ACV 600 mg/kg group (red), 300 mg/kg group (green), 100 mg/kg group (blue), and control group (black). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 8. S-plot showing the dose-dependent contribution to the clustering of different groups. The p[1] axis represents the magnitude of the spectral bins. Among them, bins with high p(corr)[1] <−0.2 for increased and >0.2 for decreased.

urine metabolic pattern changed mostly 0–6 h after the administration of ACV. Potential discriminating metabolites were chosen by the S-plot (Figs. 7 and 8), which combined the modeled covariance and variables with high value in magnitude and reliability (Chen et al., 2009). 3.4. Identification of potential metabolites in ACV-induced nephrotoxicity

Fig. 5. PCA score plot of LC/MS spectral data of urine samples obtained 0–6 h after administration (red), 7–12 h after administration (green), 13–24 h after administration (blue), and before administration (black). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Metabolites both existed in dose-dependent analysis (VIP > = 1 and p value<0.05) and time series-dependent analysis (VIP> = 1 and p value < 0.05) were included for final potential markers detection. Briefly, the PLS-DA was used to isolate the variables responsible for differences among the various groups, with the variable importance in projection (VIP) > 1. The total number of potential

238

W. Xing et al. / Environmental Toxicology and Pharmacology 46 (2016) 234–240

Table 3 Effects of ACV on WBC of Rats. group

white blood cell (WBC)

control group 100 mg/kg 300 mg/kg 600 mg/kg

12 h pre-dose

0–6 h

7–12 h

13–24 h

+(2) − +(1) +(2)

− ++(3), ++++(3) +(1), ++(1), ++++(4) ++(3), +++(1), ++++(2)

− +(1), +++(1) ++(3), +++(1), ++++(2) +(2), +++(2)++++(1)

− − +(1) +(3), ++(1)

The values of the brackets mean the number of rats. "−" indicates a negative degree; "+,” “++,” “+++,” and “++++" indicate a positive degree. White blood cell level: −: white blood cell < 70 cells/␮L; +: 70 cells/␮L ≤ white blood cells < 100 cells/␮L; ++: 100 cells/␮L ≤ white blood cells < 150 cells/␮L; +++: 150 cells/␮L ≤ white blood cells < 200 cells/␮L; ++++: white blood cells > 200 cells/␮L.

Table 4 Identified potential biomarkers in the Database. Names

m/z

RT(s)

VIP(Time)a

VIP(Dose)b

M152T36 M226T35 M227T36 1 M231T35 M451T35 M588T1079 M61T34 1 M114T32 1 M146T34 M249T34 M264T34 M151T30 1 M588T940 M337T337 M211T502 M227T36 2 M248T35 M473T35 M489T34 M171T502 M91T27

152.0562 226.2342 227.0952 230.99 451.1779 588.409 61.04058 114.0663 146.0917 249.078 264.0494 151.1439 588.4088 337.1421 211.0937 227.1245 248.0751 473.1608 489.1358 171.1009 90.977

35.761 35.301 35.799 35.249 35.0685 1079.32 34.495 32.205 34.471 34.0705 34.2365 30.2645 940.067 337.341 501.623 36.0695 34.609 34.539 34.285 501.623 26.7905

2.59808 1.46878 2.55932 2.42722 2.79026 2.87639 5.38396 4.25884 2.73331 1.29855 4.98677 2.78616 1.977 2.99926 1.56598 2.76465 4.39747 4.35458 2.1515 1.52742 1.69944

3.36006 1.7403 3.00266 2.19072 3.13457 3.66693 1.82625 2.93956 2.72307 1.66881 4.94409 3.24094 1.66327 2.80304 2.45806 3.32984 5.8917 5.23689 3.62454 2.34289 1.76968

a b c d e

Isotopesc

adductd

pcgroupe

[M + K]+ 113.09

3 3 3 3 3 4 6 6 6 6 6 12 15 21 24 179 189 192 193 235 301

[M + Na]+ 204.105 [M + K]+ 192.027 [M + H]+ 450.173 [107][M]+ [M + H]+ 60.0324 [2][M]+ [3][M]+ [24][M + 1]+ [32][M]+ [4][M]+ [46][M]+ [15][M]+ [24][M]+ [85][M]+ [7][M]+

[M + H]+ 263.043

[M + H]+ 336.138

The VIP value of different periods after ACV administration. The VIP value of different ACV dose. CAMERA annotated isotope ion. CAMERA annotated adduct ion. CAMERA annotated group.

metabolites listed in Table 4 was based on the value of VIP > 1 and P < 0.05 and the CAMERA database. These metabolites could be suggested as potential metabolites in ACV-induced nephrotoxicity. We then further identified the potential metabolites by a secondary mass spectrum (Waters Orbitrap LTQ). Biomarkers were interpreted with available secondary mass spectrum databases, such as Metlin, Massbank, and HMDB. Finally, 4 metabolites, listed in Table 5, were identified: guanine, 4-guanidinobutyric acid, creatinine, and urea. The identified biomarkers’ MS and MS/MS (guanine, 4-guanidinobutyric acid, creatinine, and urea) were shown in Supplementary materials 2. 4. Discussion Creatinine and urea nitrogen are commonly used to assess the glomerular filtration rate, as well as the concentrating and diluting Table 5 Identified significant metabolietes by secondary mass spectrum. Names

m/z

VIP(s)

VIP(Dose)

Metabolites

M61T34 1 M114T32 1 M146T34 M152T36

61.04058361 114.0662568 146.0916888 152.0562391

5.38396 4.25884 2.73331 2.59808

1.82625 2.93956 2.72307 3.36006

Urea Creatinine 4-guanidinobutyric acid Guanine

a:The VIP value of different periods after ACV administration; b: the VIP value of different ACV dose c: CAMERA annotated isotope ion; d: CAMERA annotated adduct ion; e: CAMERA annotated group.

capacity of the tubular functions of the kidney. The change in these markers may indicate the development and extent of renal tubular damage (Fuchs and Hewitt, 2011). NAG is a lysosomal enzyme present in proximal tubular cells, and an increase in the urinary activity of this enzyme is a sensitive and reasonably specific measure of renal tubular damage. It was precluded and filtrated by the glomerulus because of its relatively large molecular weight, and therefore the increase of NAG excretion is a consequence of renal tubular cell breakdown (Westhuyzen et al., 2003). In this study, administration of ACV obviously increased serum creatinine and blood urea nitrogen levels compared with the control. ACVtreated rats exhibited a significant decrease in urine creatinine and urine urea nitrogen levels compared with the control. NAG activity was markedly increased by ACV compared to the control. These data suggest that ACV treatment induced renal damage in SD rats. In addition, ACV-treated rats showed that the kidney coefficient was significantly increased compared with the control. This phenomenon was probably explained by the edema that was caused by ACV-induced tubular injury. Furthermore, PRO, BLD, and WBC were observed in several rats after administration of ACV, and the effect was the mostly obvious 0–6 h after administration. Similarly, the maximum changes of urine creatinine and urine urea nitrogen levels as well as NAG activity, were observed 0–6 h after the administration of ACV. Moreover, all of the above data were dose dependent. Metabonomics can be used not only to distinguish different states of intoxication but also to identify early toxicity

W. Xing et al. / Environmental Toxicology and Pharmacology 46 (2016) 234–240

biomarkers, which are specific indicators of damage to a particular organ (Boudonck et al., 2009; Klawitter et al., 2010; Wei et al., 2009). Various analytical techniques were used in metabonomics, including nuclear magnetic resonance(NMR) and mass spectrometry (MS) (Zhang et al., 2012). Recently, UPLC/MS technique is widely used in metabonomics because of its high sensitivity in detecting metabolites, especially in large-scale untargeted metabonomics(Zhao, 2013). In recent years, UPLC/MS metabolomics was conducted to characterize the metabolic profile of plasma, urine, or kidney tissue extracts of Morning Glory Seed, Aristolochic acid (AA) and R. alismatis-induced nephrotoxicity (Zhao and Lint, 2014; Zhao et al., 2015). Most resulted metabolites were related to free fatty acids, energy, amino acid metabolism and the TCA cycle and so on. Moreover, UPLC-based metabolomics technique is now increasingly considered as a novel diagnostic approach in kidney diseases, especially in chronic kidney disease (CKD) (Zhao, 2013). For example, the metabolic profiling of kidney, serum and urine of adenine-induced CKD animal was investigated by UPLCbased metabolomics(Zhao et al., 2014). The result showed that the most important metabolites were polyunsaturated fatty acids, pcresyl sulfate and indoxyl sulfate. These results have given our new insights into the development of renal function. In our study, to investigate the time course of metabolite changes of ACV nephrotoxicity, the samples at each time point were analyzed by PCA score plot. Based on UPLC/MS data, the PCA score plot and PLS-DA score plot showed a clear separation between ACV-treated and control groups, which indicated that the urine metabolic pattern significantly changed after the administration of ACV. In addition, the group 0–6 h after administration and the high-dose group were farthest away from the groups of before administration and the control. These data suggested that the maximum effects were revealed 0–6 h after ACV administration, and the effect was also dose dependent. These changes indicated that ACV nephrotoxicity rats presented remarkable altered urinary metabolites, and hence altered urinary profiling could reflect the changes in urinary metabolites induced by ACV. ACV-induced nephrotoxicity is a rapidly progressing tubuleinterstitial toxicity. To the best of our knowledge, urinary metabolites and the difference in metabolic pattern have not been reported in ACV-induced nephrotoxicity. 4-guanidinobutyric acid, creatinine, and urea notably decreased in the urine of ACV-treated rats compared with the control. The decreases of these metabolites were dose dependent and most pronounced 0–6 h after ACV administration and then gradually returned to normal over time. The results for creatinine and urea were in agreement with those of the biochemistry measurement. Purines are closely related structurally, they are metabolized in so many different ways. Some of the metabolites, particularly of guanine, may cause a potential nephrotoxicity. Guanine is a kind of nucleoside, which is vital to nucleic acid synthesis. The increasing level of guanine in urine was observed in this study, and this probably reflects altered activity of renal transporters, altered synthesis or breakdown of nucleotides, or altered filtering by the kidneys (Boudonck et al., 2009). Moreover, guanine is directly converted to uric acid and allantoin without phosphorylation (Ishii and Green, 1973). Therefore, guanine caused a significant rise in the allantoin level and serum uric acid(Yokozawa et al., 1982), which could induced apparente renal injury. Additionally, previous studies have suggested guanosine stimulates the release of adenosine in cultured astrocyte, and both are released under excitotoxic condition (Ciccarelli et al., 2001). Interestingly, N-met- hyl-d-aspartate(NMDA)-induced excitotoxicity stimulate these production and release of both inosine and adenosine (Zamzow et al., 2009). These evidences may give new insight into the up regulation of guanine in ACV-induced nephrotoxicity.

239

Many guanidino compounds are known to exist in human body fluids and animal organs(Marescau et al., 1985). Some of them play important biological roles in the ureagenesis and muscular constraction. However, the toxicity of guanidino compounds also has been noticed in the biological and clinical fields(Mori, 1987). Many guanidino compounds such as 4-guanidinobutyric acid, l-homoarginine induce convulsions in experimental animals (Mori, 1987). 4-guanidinobutyric acid is a kind of guanidino compound, a structural metabolite of l-arginine, which has been found in the body fluids and brain of renal-failure patients and animals (Ahmad et al., 1994; Levillain et al., 2001). In addition, the accumulation of guanidinosuccinic acid and methylguanidine has also been reported in patients with renal failure (Giovannetti et al., 1973). 4-guanidinobutyric acid, d-Canavanine and guanidinoacetic acid as well as Arg acted as an amidine donor for 2-guanidinoethanol (GEt) formation in rat kidney. From our observations, 4-guanidinobutyric acid metabolism is considered to be strongly affected by ACV. However, the activity of 4-guanidinobutyric acid, 3-guanidinopropionic acid and l-homoarginine as donors may have not been clear(Watanabe et al., 1988), which needed to be further studied. Based on the phenomenon showed by the PCA score plot and PLS-DA score plot coupled with the biochemistry results, some conclusions can be made: the presence of substantial kidney damage after administration of ACV was confirmed; 100 mg/kg (the human clinical equivalent dosage) had already induced nephrotoxicity in rats; and the effect was dose dependent. Besides, we observed a time-toxic effect relationship of ACV wherein the toxic effect was most pronounced 6 h after ACV administration and then gradually returned to normal with time, which indicating the toxic damage can be reversed. Conflict of interest All the authors declare that there are no conflicts of interest. Funding This work was supported by the Zhejiang Medical Technology Program [2012RCA 015]. References Ahmad, T., Simmonds, M., McIver, A.G., McGraw, M.E., 1994. Reversible renal failure in renal transplant patients receiving oral acyclovir prophylaxis. Pediatr. Nephrol. 8, 489–491. Boudonck, K.J., Mitchell, M.W., Nemet, L., Keresztes, L., Nyska, A., Shinar, D., Rosenstock, M., 2009. Discovery of metabolomics biomarkers for early detection of nephrotoxicity. Toxicol. Pathol. 37, 280–292. Chen, Y., Zhang, R., Song, Y., He, J., Sun, J., Bai, J., An, Z., Dong, L., Zhan, Q., Abliz, Z., 2009. RRLC–MS/MS-based metabonomics combined with in-depth analysis of metabolic correlation network: finding potential biomarkers for breast cancer. Analyst 134, 2003–2011. Ciccarelli, R., Ballerini, P., Sabatino, G., Rathbone, M.P., D’Onofrio, M., Caciagli, F., Di Iorio, P., 2001. Involvement of astrocytes in purine-mediated reparative processes in the brain. Int. J. Dev. Neurosci. 19, 395–414. Fleischer, R., Johnson, M., 2010. Acyclovir nephrotoxicity: a case report highlighting the importance of prevention, detection, and treatment of acyclovir-induced nephropathy. Case Rep. Med. 2010. Fraga, C.G., Clowers, B.H., Moore, R.J., Zink, E.M., 2010. Signature-discovery approach for sample matching of a nerve-agent precursor using liquid chromatography–mass spectrometry, XCMS, and chemometrics. Anal. Chem. 82, 4165–4173. Fuchs, T.C., Hewitt, P., 2011. Biomarkers for drug-induced renal damage and nephrotoxicity-an overview for applied toxicology. AAPS J. 13, 615–631. Giovannetti, S., Balestri, P.L., Barsotti, G., 1973. Methylguanidine in uremia. Arch. Intern. Med. 131, 709–713. Giustina, A., Romanelli, G., Cimino, A., Brunori, G., 1988. Low-dose acyclovir and acute renal failure. Ann. Intern. Med. 108, 312. Ishii, K., Green, H., 1973. Lethality of adenosine for cultured mammalian cells by interference with pyrimidine biosynthesis. J. Cell Sci. 13, 429–439. Klawitter, J., Haschke, M., Kahle, C., Dingmann, C., Klawitter, J., Leibfritz, D., Christians, U., 2010. Toxicodynamic effects of ciclosporin are reflected by

240

W. Xing et al. / Environmental Toxicology and Pharmacology 46 (2016) 234–240

metabolite profiles in the urine of healthy individuals after a single dose. Br. J. Clin. Pharmacol. 70, 241–251. Levillain, O., Marescau, B., Possemiers, I., Al Banchaabouchi, M., De Deyn, P.P., 2001. Influence of 72% injury in one kidney on several organs involved in guanidino compound metabolism: a time course study. Pflugers Arch. 442, 558–569. Lu, H., Han, Y.J., Xu, J.D., Xing, W.M., Chen, J., 2014. Proteomic characterization of acyclovir-induced nephrotoxicity in a mouse model. PLoS One 9, e103185. Lyon, A.W., Mansoor, A., Trotter, M.J., 2002. Urinary gems: acyclovir crystalluria. Arch. Pathol. Lab. Med. 126, 753–754. Marescau, B., Qureshi, I.A., De Deyn, P., Letarte, J., Ryba, R., Lowenthal, A., 1985. Guanidino compounds in plasma, urine and cerebrospinal fluid of hyperargininemic patients during therapy. Clin. Chim. Acta 146, 21–27. Mori, A., 1987. Biochemistry and neurotoxicology of guanidino compounds. History and recent advances. Pavlovian J. Biol. Sci. 22, 85–94. Obada, E.N., Level, G., Mathieu, P., Parent, X., Gilson, B., Bindi, P., 2010. [Acute renal failure following a treatment with acyclovir]. Nephrol. Ther. 6, 125–127. Portilla, D., Li, S., Nagothu, K.K., Megyesi, J., Kaissling, B., Schnackenberg, L., Safirstein, R.L., Beger, R.D., 2006. Metabolomic study of cisplatin-induced nephrotoxicity. Kidney Int. 69, 2194–2204. Sawyer, M.H., Webb, D.E., Balow, J.E., Straus, S.E., 1988. Acyclovir-induced renal failure: clinical course and histology. Am. J. Med. 84, 1067–1071. Vachvanichsanong, P., Patamasucon, P., Malagon, M., Moore, E.S., 1995. Acute renal failure in a child associated with acyclovir. Pediatr. Nephrol. 9, 346–347. Vomiero, G., Carpenter, B., Robb, I., Filler, G., 2002. Combination of ceftriaxone and acyclovir − an underestimated nephrotoxic potential? Pediatr. Nephrol. 17, 633–637. Watanabe, Y., Yokoi, I., Watanabe, S., Sugi, H., Mori, A., 1988. Formation of 2-guanidinoethanol by a transamidination reaction from arginine and ethanolamine by the rat kidney and pancreas. Life Sci. 43, 295–302.

Wei, L., Liao, P., Wu, H., Li, X., Pei, F., Li, W., Wu, Y., 2009. Metabolic profiling studies on the toxicological effects of realgar in rats by (1)H NMR spectroscopy. Toxicol. Appl. Pharmacol. 234, 314–325. Weiss, R.H., Kim, K., 2012. Metabolomics in the study of kidney diseases. Nature reviews. Nephrology 8, 22–33. Westhuyzen, J., Endre, Z.H., Reece, G., Reith, D.M., Saltissi, D., Morgan, T.J., 2003. Measurement of tubular enzymuria facilitates early detection of acute renal impairment in the intensive care unit. Nephrol. Dial. Transplant. 18, 543–551. Yokozawa, T., Oura, H., Okada, T., 1982. Metabolic effects of dietary purine in rats. J. Nutr. Sci. Vitaminol. 28, 519–526. Zamzow, C.R., Bose, R., Parkinson, F.E., 2009. N-methyl-D-aspartate-evoked adenosine and inosine release from neurons requires extracellular calcium. Can. J. Physiol. Pharmacol. 87, 850–858. Zhang, A., Sun, H., Wang, P., Han, Y., Wang, X., 2012. Modern analytical techniques in metabolomics analysis. Analyst 137, 293–300. Zhao, Y.Y., Lint, R.C., 2014. Metabolomics in nephrotoxicity. Adv. Clin. Chem. 65, 69–89. Zhao, Y.Y., Cheng, X.L., Vaziri, N.D., Liu, S., Lin, R.C., 2014. UPLC-based metabonomic applications for discovering biomarkers of diseases in clinical chemistry. Clin. Biochem. 47, 16–26. Zhao, Y.Y., Tang, D.D., Chen, H., Mao, J.R., Bai, X., Cheng, X.H., Xiao, X.Y., 2015. Urinary metabolomics and biomarkers of aristolochic acid nephrotoxicity by UPLC–QTOF/HDMS. Bioanalysis 7, 685–700. Zhao, Y.Y., 2013. Metabolomics in chronic kidney disease. Clin. Chim. Acta 422, 59–69. van de Poll, M.C., Soeters, P.B., Deutz, N.E., Fearon, K.C., Dejong, C.H., 2004. Renal metabolism of amino acids: its role in interorgan amino acid exchange. The Am. J. Clin. Nutr. 79, 185–197.