S

Sci

Ya

b

c

d

a

ARRAA

KACCHR

1

(fmoasorccbngt

0d

Journal of Pharmaceutical and Biomedical Analysis 66 (2012) 258– 263

Contents lists available at SciVerse ScienceDirect

Journal of Pharmaceutical and Biomedical Analysis

j ourna l ho me p a ge: www.elsev ier .com/ locate / jpba

hort communication

imultaneous determination of adenine nucleotides, creatine phosphate andreatine in rat liver by high performance liquid chromatography–electrosprayonization-tandem mass spectrometry

ang Jianga, Chengjun Suna, Xueqin Dingb, Ding Yuanc, Kefei Chenc, Bo Gaob, Yi Chend, Aimin Sunb,∗

West China School of Public Health, Sichuan University, No. 17, Section 3, South Renmin Road, Chengdu, Sichuan 610041, People’s Republic of ChinaAnalytical & Testing Center, Sichuan University, No. 29, Wangjiang Road, Chengdu, Sichuan 610064, People’s Republic of ChinaRegenerative Medicine Research Center, West China Hospital, Chengdu, Sichuan 610041, People’s Republic of ChinaLaboratory of Aging and Geriatrics Research, West China Hospital, Chengdu, Sichuan 610041, People’s Republic of China

r t i c l e i n f o

rticle history:eceived 1 December 2011eceived in revised form 13 March 2012ccepted 14 March 2012vailable online 24 March 2012

eywords:denine nucleotidesreatine

a b s t r a c t

A high performance liquid chromatography–electrospray ionization-tandem mass spectrometric method(HPLC–ESI-MS/MS) was developed for simultaneous determination of adenosine 5′-triphosphate (ATP),adenosine 5′-diphosphate (ADP), adenosine 5′-monophosphate (AMP), creatine phosphate (CP), and cre-atine in rat liver. After extraction with pre-cooled (4 ◦C) methanol/water (1:1, v/v), the analytes wereseparated on a porous graphitic carbon (Hypercarb) column (2.1 mm × 150 mm, 5 �m) using a pro-grammed gradient elution with a mobile phase consisting of 2 mmol/L ammonium acetate in water and2 mmol/L ammonium acetate in acetonitrile (pH = 10.0). The analytes were detected in a way of multiplereaction monitoring (MRM) under negative scan mode by a triple quadrupole mass spectrometer with

reatine phosphatePLC–ESI-MS/MSat liver

electrospray ionization (ESI). An external calibration method with linear ranges from 10 to 5000 ng/mLfor the five target compounds was used for quantification with a correlation coefficients ≥ 0.9973. Thelimits of detection and limits of quantification for all analytes were in ranges from 0.50 to 1.5 ng/mL and1.6 to 0.5 ng/mL, respectively. The average recoveries spiked in three levels were from 77.2% to 102% andprecisions expressed in RSDs were from 0.2% to 4.8%. The established method was successfully appliedto determination of ATP, ADP, AMP, CP and creatine in liver tissue.

. Introduction

Adenosine 5′-triphosphate (ATP), adenosine 5′-diphosphateADP) and creatine phosphate (CP) are important biomoleculesor energy storage and conversion in tissues. Molecules of 5′-

onophosphate (AMP) and creatine are involved in metabolismf these energy biomolecules. Liver is one of the most complexnd metabolically active organs and performs many functions,uch as biosyntheses and detoxification in which a large amountf energy molecules are needed. Concomitantly, liver cells areich in active mitochondria that can provide abundant energy forellular metabolism. In steatosis, energy metabolism of hepato-yte mitochondrial is impaired [1], implying alterations in energyiomolecules. Therefore, simultaneous quantification of adenine

ucleotides, creatine phosphate and creatine in liver tissue is ofreat importance for understanding the energy state and evaluatinghe function of liver cells.

∗ Corresponding author.E-mail address: sunaiminsc@sina.com (A. Sun).

731-7085/$ – see front matter © 2012 Elsevier B.V. All rights reserved.oi:10.1016/j.jpba.2012.03.027

© 2012 Elsevier B.V. All rights reserved.

High-performance liquid chromatography (HPLC) has beenwidely used for determination of the aforementioned compoundsin biological samples [2,3]. Considering high polarity and non-volatility of the analytes, as well as background interference ofbiological matrix, several HPLC–MS methods have been developedfor analysis of the energy-related biomolecules [4–8]. However,these methods only focused on the determination of adeninenucleotides or creatine and in some cases [4,5] ion-pairing reagentswere added to mobile phase, which could result in contaminationof ion source and frequent cleaning was required. Besides, the sam-ple matrixes of these methods were mainly kidney tissues [4], cells[5,6] urine [7]. Until now, there seems no LC–MS/MS method inmultiple reaction monitoring (MRM) mode reported on simultane-ous determination of adenine nucleotides, CP and creatine in livertissue.

In this study, a simple and reliable HPLC–ESI-MS/MS methodwas developed for the first time for simultaneous determination

of ATP, ADP, AMP, CP and creatine. Using porous graphitic car-bon (PGC, Hypercarb) column, ion-pairing reagents were avoidedand reasonable retention time for the analytes was achieved. Toensure high specificity, quantification was performed in MRM

l and Biomedical Analysis 66 (2012) 258– 263 259

meppftsdh

2

2

fasgs

2

TUsai(Um

2

cSm(vgArpt

2

aswra(

2

astmw2

Table 1Multiple reaction monitoring MRM transitions for quantification of the five targetcompounds.

Analyte Precursor Product Collisionenergy (V)

Tube lens Retentiontime (min)

ATP 506 159a 35 113 7.42177b 24408b 22

ADP 426 134b 24 97 7.33159b 24328a 19

AMP 346 79a 36 95 7.23134b 35211b 15

CP 210 79a 35 76 1.7497b 10

121b 15

Creatine 130 59b 13 63 2.6288a 13

112b 14

a Quantitation ion.

Y. Jiang et al. / Journal of Pharmaceutica

ode. Moreover, methanol–water solution was used for samplextraction, and the procedure was rapid, simple as well as com-atible with ESI-MS/MS analysis compared with the commonerchloric acid-based extraction [4,6]. The method was success-ully applied to analysis of ATP, ADP, AMP, CP and creatine in liverissue of 8 normal rats. In addition, this assay is potentially useful intudying energy metabolism of fatty liver that is used as marginalonor liver, which is of great significance to further research onuman liver transplantation.

. Materials and methods

.1. Chemicals and reagents

Standards of ATP, ADP, AMP, CP and creatine were purchasedrom Sigma (St. Louis, MO, USA). HPLC-grade acetonitrile (ACN)nd methanol were purchased from Honeywell Burdick & Jack-on (Muskegon, MI, USA). Other chemicals used were of analyticalrade. Water used in the experiment was obtained from a Milli-Qystem (18.2 M� cm, Millipore, Bedford, MA, USA).

.2. Instrumentation

The HPLC–MS/MS system, Surveyor Plus HPLC System andSQ Quantum Ultra Mass Spectrometer (Thermo Fisher Scientific,SA), is equipped with an autosampler, ESI electrospray ionization

ource, and TSQ mass analyzer. LCQUAN quantitative software waspplied for data acquisition and processing. Other instruments usedn sample preparation included a high speed refrigerated centrifugeHC-2518 R, Zonkia, Hefei, China), a vacuum freeze dryer (Labconco,SA), a vortex shaker (QL-861, Kylin-Bell, Haimen, China), and a pHeter (PHS-3C, SPSIC, Shanghai, China).

.3. Chromatographic conditions

The chromatographic separation was performed on a Hypercarbolumn (2.1 mm × 150 mm, 5 �m) from Thermo (Thermo Fishercientific, USA). The standards and samples were separated using aobile phase consisting of 2 mmol/L ammonium acetate in water

solvent A) and 2 mmol/L ammonium acetate in acetonitrile (sol-ent B), which were both adjusted with ammonia to pH 10.0. Theradient elution program was: 0–2 min, 98% A; 2–4 min, 98–85%; 4–10 min, 85% A; 10–13 min, 50% A; and then back to 98% A toeconditioning the column for 7 min. The flow rate of the mobilehase was 200 �L/min, and the column temperature was main-ained at 35 ◦C.

.4. Mass spectrometric parameters

ESI ionization was performed in the negative ion mode with spray voltage of 3000 V. Sheath gas pressure and aux gas pres-ure were 40 arb and 10 arb, respectively. Capillary temperatureas set at 350 ◦C. Quantification was performed in the multiple

eaction monitoring (MRM) mode. The optimized parameters, suchs quantification and confirmation ion transitions, collision energyCE), tube lens for each compound, were summarized in Table 1.

.5. Preparation of standard solutions

Proper amounts of ATP, ADP, AMP, CP and creatine wereccurately weighed and dissolved in water to prepare stocktandard solutions at 1.0 mg/mL. Mixed working standard solu-

ion was prepared by dilution of the stock standards solution in

ethanol–water (1:1, v/v). Then mixed working standard solutionas serially diluted to prepare the concentration series of 10, 50,

50, 500, 2000 and 5000 ng/mL. Stock standard solutions were

b Confirmation ion.

stable for one week stored at −20 ◦C, and mixed working standardsolution was freshly prepared prior to use.

2.6. Sample preparation

The freeze-dried (−70 ◦C for 36 h) samples of rat liver werequickly ground and blended in a mortar. An aliquote of 10 mg ofpowdered liver tissue was placed into a 1.5 mL scaled Eppendorftube containing 500 �L pre-cooled (4 ◦C) methanol to precipitateproteins by vortex shaking for 1 min. And then 500 �L pre-cooled(4 ◦C) water was immediately added to extract the analytes by vor-tex shaking for 3 min. After centrifugation of the extract at −10 ◦Cfor 20 min at 16,000 × g, the supernatant was stored at −20 ◦C priorto the LC–MS/MS analysis.

3. Results and discussion

3.1. Optimization of chromatographic conditions

3.1.1. Selection of column for separationSeveral different chromatographic columns were tested, includ-

ing Hypersil Gold C18 column (150 mm × 2.1 mm, 5 �m) fromThermo, Ultra C8 column (15 mm × 2.1 mm, 5 �m) from Restek,Spursil C18-EP (150 mm × 2.1 mm, 3 �m) from Dikma, Hyper-carb column (100 mm × 4.6 mm, 5 �m) and Hypercarb column(150 mm × 2.1 mm, 5 �m) from Thermo. Due to their high polar-ity, all the target compounds were weakly retained on thesecolumns except Hypercarb column. Wang et al. also found thatvarious reversed-phase columns could not provide reasonableretention time for adenine nucleotides [6]. The retention mecha-nism of porous graphitic carbon (PGC, Hypercarb) is different fromconventional silica-bonded stationary phase, which can provideremarkable retention and selectivity for polar compounds. In addi-tion, Hypercarb column is stable throughout the pH range of 0–14.Moreover, HPLC with PGC column would be more compatible forMS detection because the ion source contamination from buffersalts or ion-pairing agents could be avoided. Therefore, Hypercarb

column was selected in this study. Fig. 1 shows the LC–MS/MS chro-matogram of a standard mixture of the five target compounds inMRM mode.

260 Y. Jiang et al. / Journal of Pharmaceutical and B

Fp

3

smLgstncft1cs2rrwtfsitiwHhf

3

se

ig. 1. The LC–MS/MS chromatograms of a standard mixture of the five target com-ounds (2000 ng/mL) in MRM mode.

.1.2. Selection of mobile phase for separationThe composition of the mobile phase was optimized through

everal steps on the Hypercarb column. Firstly, the effect ofethanol and acetonitrile as organic mobile phase using the same

C gradient and ammonium acetate concentration was investi-ated. There was no elution for the analytes in 30 min whenubstituting methanol for acetonitrile. Xing et al. also reportedhat acetonitrile was much more effective in displacing adenineucleotides from the PGC column and exhibited a stronger elutionapability than methanol [8]. Therefore, acetonitrile was selectedor further experiments in this study. Secondly, based on the litera-ures [6,8], different concentrations of ammonium acetate (2, 5 and0 mmol/L) were tested. It was found that the increase of salt con-entration resulted in suppression of MS signal intensity, because ofalt precipitation in ion source and ion transportation tube. Hence,

mmol/L ammonium acetate was added in the mobile phase, andeasonable retention and separation as well as good sensitivity andeproducibility were obtained. Thirdly, the pH of the mobile phaseas also examined using acetic acid and ammonia for adjusting

he pH of the mobile phase. In the preliminary experiments, it wasound that the ESI-MS/MS intensity of adenine nucleotides con-iderably decreased when the pH was lower than 8. Hence, thenfluence of the mobile phase pH on the detection sensitivity ofhe analytes was investigated over the pH range of 8–11. With thencrease of pH, the signal intensity of ESI-MS/MS for ATP and ADP

as greatly enhanced, while the intensity of AMP and CP was not.owever, there was significant decrease for creatine. By compre-ensive consideration, the mobile phase’s pH was adjusted to 10.0

or the analysis of the target compounds.

.2. Optimization of mass spectrometric conditions

Due to high polarity of the analytes, the ESI ion source waselected. The mass spectrometric conditions were optimized forach compound by continuously infusing corresponding standard

iomedical Analysis 66 (2012) 258– 263

solutions at a flow rate of 10 �L/min using a syringe infusion pump.During the method development, MS parameters in both positiveand negative ionization modes were tuned for ATP, ADP, AMP, CPand creatine. However, it was found that the response was muchhigher in the negative ionization mode for ATP, ADP, AMP and CPthan that in the positive mode due to their acidic nature. There-fore, precursor ions and product ions were detected in ESI negativemode and quantification was performed in MRM mode. Fig. 2 showschemical structures and full scan product ion of precursor ions ofATP, ADP, AMP, CP and creatine. Three product ions were moni-tored, one for quantification and the other two for confirmation.The optimized MS parameters are described in Section 2.4.

3.3. Sample extraction

In the previous literatures, perchloric acid [4,6], acetoni-trile/water [9] and methanol [5] were used for extraction.Considering good water solubility of the analytes, we evaluateddifferent solvents for the extraction, including 0.42 mol/L per-chloric acid, acetonitrile:water (1:1), methanol:water (7:3) andmethanol:water (1:1). For the perchloric acid extract, we found thatthe pH of the extract should be carefully monitored at 7 after KOHsolution was used to neutralize perchloric acid, otherwise redun-dant acid or base would be not compatible with MS system andfurther interfere with the detection. Cordell et al. [5] reported thatthe perchloric acid extract was not compatible with LC–MS becauseof the presence of a precipitate when evaporating the neutralizedsamples. Hence, we tested several organic solvents that can pre-cipitate proteins. The recoveries for the acetonitrile:water (1:1)extraction varied from 10% to 33%. Methanol:water (7:3) systemdemonstrated a lower than 20% recovery for ATP and a lower than50% recovery for CP. With the proportion of water increase to 50%,the recoveries for the five target compounds varied from 68.7% to87.0%. Therefore, methanol:water (1:1, v/v) was used for extraction,which was easy to prepare and more MS-compatible, comparedwith the perchloric acid-based extraction.

3.4. Method validation

3.4.1. Linearity, limit of detection and limit of quantificationTable 2 shows the equations of the external standard curves,

correlation coefficients, linear range, limits of detection (LODs), andlimits of quantification (LOQs) for the five target compounds. Goodlinearities (r ≥ 0.9973) were obtained for all the five analytes. Here,the LODs and LOQs were defined as the concentration with a signal-to-noise ratio of 3 and 10, respectively.

3.4.2. Accuracy and precisionAccuracy of the method was evaluated by the recoveries of

spiked liver tissue samples and precision was evaluated using rel-ative standard deviations (RSDs). Three levels of mixture standardsolutions were added to each sample and then analyzed using theestablished method. Each level was performed in six replicates. Theresults are shown in Table 3. In the three spiked levels, the aver-aged recoveries of this method were from 77.2% to 102%. The RSDsof method ranged from 0.2% to 4.8%.

3.5. Application of the method

This method has been successfully applied to determination ofATP, ADP, AMP, CP and creatine in liver tissue from eight normalrats. The collection of rat liver tissues was conducted in accordance

with the experimental protocols approved by the ExperimentalAnimal Center of Sichuan University. The results are summarizedin Table 4 and a typical LC–MS/MS chromatogram from rat liveris displayed in Fig. 3. In the previous studies, HPLC method was

Y. Jiang et al. / Journal of Pharmaceutical and Biomedical Analysis 66 (2012) 258– 263 261

Fig. 2. Chemical structures and full scan product ion of precursor ions of ATP (a), ADP (b), AMP (c), CP (d) and creatine (e).

Table 2Linear regression, LODs and LOQs for the five target compounds.

Analyte Calibration curve r Linear range (ng/mL) LOD (ng/mL) LOQ (ng/mL)

ATP y = 259.4x − 2343 0.9995 10–5000 0.5 1.6ADP y = 338.5x − 1694 0.9992 10–5000 0.5 1.6AMP y = 369.0x − 516.9 0.9999 10–5000 0.5 1.6CP y = 214.9x − 2058 0.9982 10–5000 1.0 3.3Creatine y = 712.6x − 495.4 0.9973 10–5000 1.5 5.0

262 Y. Jiang et al. / Journal of Pharmaceutical and Biomedical Analysis 66 (2012) 258– 263

Table 3Accuracy and precision of the method for the five target compounds (n = 6).

Analyte Background (ng/mL) Spiked (ng/mL) Mean found (ng/mL) Recovery (%) RSD (%)

ATP 706.5 250 240.2 96.1 2.6500 488.3 97.7 2.21000 1011 101 2.4

ADP 1132 250 242.9 97.1 4.8500 488.9 97.8 3.31000 1006 101 2.4

AMP 445.8 250 244.1 97.6 3.9500 501.3 100 2.41000 984.2 98.4 2.7

CP 69.3 200 156.2 78.1 1.4500 386.2 77.2 0.21000 810.7 81.1 0.3

Creatine 188.5 250 255.5 102 2.1500 495.9 99.2 1.31000 983.7 98.4 1.4

Table 4The analytical results of the five target compounds in liver tissues by using the proposed method (mg/kg).

Sample no. ATP ADP AMP CP Creatine

1 5606 2707 492.4 24.28 188.82 6470 2092 374.4 40.96 197.73 3026 1969 521.8 29.63 229.84 1414 1049 301.4 9.032 127.45 6105 2830 484.2 9.628 176.66 4305 2551 471.7 9.771 141.47 4548 3469 419.5 16.87 202.98 5172 4250 507.2 18.06 202.5Mean ± SD 4581 ± 1683 2615 ± 971.3

Fig. 3. Chromatograms of the five target compounds in rat liver.

446.6 ± 76.06 19.78 ± 11.31 183.4 ± 33.97

widely used in determination of adenine nucleotides in liver tis-sue. Mikami et al. reported the change levels of ATP, ADP and AMPin rat liver induced by intense exercise [10]. And the contents ofATP, ADP and AMP in rat liver tissues during different day timeswere also reported by Abadi [11]. However, there is no report aboutsimultaneous determination of ATP, ADP, AMP, CP and creatine inliver tissue by LC–MS/MS.

4. Conclusion

In this paper, a HPLC–ESI-MS/MS method was developed forsimultaneous determination of ATP, ADP, AMP, CP and creatine inliver tissue. The use of porous graphitic carbon (PGC, Hypercarb)column enhanced retention of adenine nucleotides and meanwhileoffered MS-compatible chromatographic conditions because noion-pair agents were needed. In addition, the sample pre-treatmentprocedure was rapid, simple and well compatible with ESI-MS/MSanalysis. This method has been successfully applied to simultane-ous determination of ATP, ADP, AMP, CP and creatine in liver tissuefrom 8 normal rats. Moreover, the method can be expended forstudying energy metabolism of fatty liver that is used as marginaldonor liver, which is of great significance to further research onhuman liver transplantation.

Acknowledgement

The authors are grateful to the financial support from applica-tion infrastructure fund (2010JY0007) of Department of Science andTechnology of Sichuan Province, China.

l and B

R

[nucleotide and purine nucleotide synthesis via de novo pathway in the rat

Y. Jiang et al. / Journal of Pharmaceutica

eferences

[1] G. Vial, H. Dubouchaud, K. Couturier, C. Cottet-Rousselle, N. Taleux, A. Athias,A. Galinier, L. Casteilla, X.M. Leverve, Effects of a high-fat diet on energymetabolism and ROS production in rat liver, J. Hepatol. 54 (2011) 48–356.

[2] M.D. Scott, L.J. Baudendistel, T.E. Dahms, Rapid separation of creatine,phosphocreatine and adenosine metabolites by ion-pair reversed-phasehigh-performance liquid chromatography in plasma and cardiac tissue, J. Chro-matogr. 576 (1992) 149–154.

[3] M.G. Volonté, G. Yuln, P. Quiroga, A.E. Consolini, Development of an HPLCmethod for determination of metabolic compounds in myocardial tissue, J.Pharm. Biomed. Anal. 35 (2004) 647–653.

[4] J. Klawitter, V. Schmitz, J. Miljus, D. Leibfritz, U. Christians, Developmentand validation of an assay for the quantification of 11 nucleotides using

LC/LC–electrospray ionization-MS, Anal. Biochem. 365 (2007) 230–239.

[5] R.L. Cordell, S.J. Hill, C.A. Ortori, D.A. Barrett, Quantitative profiling ofnucleotides and related phosphate-containing metabolites in culturedmammalian cells by liquid chromatography tandem electrospray mass spec-trometry, J. Chromatogr. B 817 (2008) 115–124.

[

iomedical Analysis 66 (2012) 258– 263 263

[6] J. Wang, T. Lin, J.P. Lai, Z.W. Cai, M.S. Yang, Analysis of adenosine phosphatesin HepG-2 cell by a HPLC–ESI-MS system with porous graphitic carbon as sta-tionary phase, J. Chromatogr. B 877 (2009) 2019–2024.

[7] S. Young, E. Struys, T. Wood, Quantification of creatine and guanidinoacetateusing GC–MS and LC–MS/MS for the detection of cerebral creatine deficiencysyndromes, Curr. Protoc. Hum. Genet. (2007) 17.3.1–17.3.18.

[8] J. Xing, A. Apedo, A. Tymiak, N. Zhao, Liquid chromatographic analysis of nucle-osides and their mono-, di- and triphosphates using porous graphitic carbonstationary phase coupled with electrospray mass spectrometry, Rapid Com-mun. Mass Spectrom. 18 (2004) 1599–1606.

[9] J.L. Au, M.H. Su, M.G. Wientjes, Extraction of intracellular nucleosides andnucleotides with acetonitrile, Clin. Chem. 35 (1989) 48–51.

10] T. Mikami, J. Kitagawa, Intense exercise induces the degradation of adenine

liver, Appl. Physiol. 96 (2006) 543–550.11] R.H.H.B. Abadi, Analysis of free nucleotide pools of mouse liver tissue by

high-pressure liquid chromatography (HPLC), Biochem. Biophys. 40 (2003)209–212.