Journal of Chromatography B, 961 (2014) 91–96
Contents lists available at ScienceDirect
Journal of Chromatography B journal homepage: www.elsevier.com/locate/chromb
Highly sensitive, selective and rapid LC–MS method for simultaneous quantification of diadenosine polyphosphates in human plasma Anna Schulz a , Vera Jankowski b , Walter Zidek a , Joachim Jankowski b,∗ a b
Charité-Universitätsmedizin Berlin (CBF), Medizinische Klinik IV, Germany Universitätsklinikum RWTH Aachen, Institute of Molecular Cardiovascular Research, Aachen, Germany
a r t i c l e
i n f o
Article history: Received 16 February 2014 Accepted 11 May 2014 Available online 19 May 2014 Keywords: Dinucleoside polyphosphates Mediators Quantification Biomolecule Hypertension
a b s t r a c t Background: Diadenosine polyphosphates (Apn As) are endogenous mediators involved in large number of physiologic and pathophysiologic processes. The quantification of diadenosine polyphosphates in plasma and biological matrices is still challenging. Therefore, there is an urgent need for a simple and reliable quantification method suitable for clinical studies. The classical quantification of diadenosine polyphosphates is based on chromatographic separation and UV adsorption of the resulting fractions. These procedures are associated with low selectivity due to co-eluting plasma components. Therefore, we developed and validated a highly sensitive, selective and rapid LC–ESI–MS method for simultaneous quantification of Apn As (with n = 3–6) in human plasma within this study. The identities of the endogenous Apn As (with n = 3–6) were revealed by comparison of ESI-MS/MS fragment spectra of isolated endogenous compounds with those of authentic Apn As. Methods: Diadenosine polyphosphates were extracted from 100 l human plasma using weak anionexchange extraction cartridges. The separation of Apn As was achieved using capillary C18 columns. ESIHCT mass spectrometer (Bruker Daltonik, Germany) operated in negative ion mode was used for detection and quantification of Apn As. Results: A calibration curve was established for diadenosine polyphosphate free plasma in the concentration range 1.9–125 nM (r2 > 0.998) for all analytes. The intra- and inter-day accuracies were in the range of 91.4% and 110.9%. The intra- and inter-day precisions were determines as 0.1% and 11.4%, respectively. The mean plasma concentrations of Apn As were quantified as 31.9 ± 5.9 nM for Ap3 A, 40.4 ± 6.6 nM for Ap4 A, 10.7 ± 1.5 nM for Ap5 A and 10.0 ± 18.9 nM for Ap6 A. Discussion: The developed and validated ESI MS-based method for quantification of diadenosine polyphosphates in human plasma was successfully evaluated within the study. Conclusion Since the quantification is based on a volume of 100 l plasma, this method is highly applicable for clinical applications aiming at the validation of the impact of highly physiological and pathophysiological active diadenosine polyphosphates. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Diadenosine polyphosphates (Apn As) are endogenous mediators, isolated from e.g. blood platelets [1,2], adrenal medullary chromaffin granules [3,4], the central nervous system [4,5] and human plasma [6]. They are formed by two adenosine nucleosides interconnected by a phosphate chain, which varies from 3 up to 6 phosphates [7,8]. The schematic chemical structure of
∗ Corresponding author at: RWTH University of Aachen, Institute of Molecular Cardiovascular Research, Pauwelsstrasse 30, 52074 Aachen, Germany. Tel.: +49 241 80 80580; fax: +49 241 80 52716. E-mail address: [email protected] (J. Jankowski). http://dx.doi.org/10.1016/j.jchromb.2014.05.018 1570-0232/© 2014 Elsevier B.V. All rights reserved.
Apn As is shown in Fig. 1A. Apn As have been demonstrated to control numerous physiological and pathophysiological functions like neurotransmission, homeostasis as well as vascular regulation [9–11]. Furthermore, they are involved in cardio-renal processes and may have an impact on the progression of chronic kidney disease (CKD) [12,13]. Therefore, the quantification of the endogenous Apn As concentrations is of utmost importance since Apn As might be potential biomarker or target for specific therapy in the progression of chronic kidney disease. Until now, just a few methods have been described for the determination of Apn As in human plasma. These analytical methods are mostly based on liquid chromatography coupled with detection of the UV absorbance at 254 nm. The sensitivity is not exactly specified in those methods. Furthermore, for the quantification of Apn As
92
A. Schulz et al. / J. Chromatogr. B 961 (2014) 91–96
NH2
NH 2 N
N
N N
N
O
N O H
H
OH
H OH
H
A
2.2. High performance liquid chromatography conditions
B
An Agilent 1200 (Agilent, Böblingen, Germany) capillary LC system equipped with micro vacuum degasser (G1379B) and capillary pump (G1376A) along with an autosampler (G1377A) was used to inject 2 l aliquots of prepared samples. The separation was carried out on a Zorbax C18 XDB 0.5 mm × 35 mm column with 3.5 m particle size (Agilent, Böblingen, Germany). The temperature of the column was maintained at constant 50 ± 0.5 ◦ C. The eluent A was consisted of 0.1% N,N-DMHA in 1 mM ammonium formate prepared in LC–MS grade water. The pH of eluent A was adjusted using formic acid to 9.0. Eluent B was LC–MS grade acetonitrile. The injection volume was 2 l. The following gradient was used to separate Apn As: 0 min, 0% B, 0–1 min 0–12% B, 1–1.75 min 1–16% B, 1.75–2.5 min 16–22% B, 2.5–5.5 min 22–98% B, 5.5–6.5 min 98% B, 6.5–6.6 min 98–0% B, 6.6–12.5 min 0% B. The primary flow was 500 l/min and it was split to 100 l/min by using electronic flow control. The total chromatographic run time was 12.5 min.
N N
O
O P O O-
H
H
OH
H OH
H n=3-6
dilution of plasma with 100 mM ammonium formate weak anion exchange SPE Evaporation Reconstitution in 1% N,N-DMHA LC-ESI-MS
Fig. 1. (A) Molecular structure of diadenosine polyphosphates (Ap3–6 A) and (B) scheme of the preparation steps for the quantification of diadenosine polyphosphates.
by using those methods a high plasma volume is required. Since the CKD patients are suffering from anaemia the plasma volume should be minimized. In general, the quantification of endogenous Apn As levels and the validation of the analytical method are still challenging. The limitations are: (a) the concentration of diadenosine polyphosphates is in the nanomolar range, (b) short half-life times of diadenosine polyphosphates, and (c) the minimal differences in the chemical characteristics of diadenosine polyphosphates. In addition, since diadenosine polyphosphates are endogenous components of plasma, a diadenosine polyphosphates-free matrix is not directly available, which has an impact on the determination of the lower limit of quantification. Since chromatographic fractionation diadenosine polyphosphates are highly polar, the use of buffer systems and/or ion pair reagent to ensure a sufficient retention by chromatographic media are mandatory. To avoid the co-detection of eluting substances from the chromatographic column, a highly sensitive and selective method for detection of the eluting substances is essential. UV adsorption as used in former studies does not fulfil this criterion. In recent years, liquid chromatography tandem mass spectrometry has been characterized as a selective and highly robust method for quantification of metabolites and endogenous mediators [14]. Therefore, we developed, established and validated a liquid chromatography electrospray-ionization massspectrometric-based method for simultaneous quantification of Apn As in human plasma in this study.
2. Materials and methods 2.1. Chemicals and reagents Diadenosine polyphosphates (Ap3 A, Ap4 A, Ap5 A, Ap6 A), analytical grade ammonium formate, formic acid, ammonium hydroxide and N,N-dimethylhexylamine (DMHA) were purchased from Sigma Aldrich (Seelze, Germany). Liquid chromatography-mass spectrometry (LC–MS) grade water and acetonitrile were purchased from Fisher Scientific (Fair Lavn, United States). Oasis WAX cartridges were obtained from Waters (Eschborn, Germany).
2.3. Mass-spectrometric analysis The detection of Apn As was achieved in negative ion mode for all analytes by using a HCT mass spectrometer (Bruker Daltonic, Bremen, Germany), equipped with an electrospray ionization interface. The ionization parameters of nebulizer gas, dry gas and dry temperature were 25 psi, 9 l/min, and 300 ◦ C respectively. The ion spray voltage was set to 3000 V. The maximal accumulation time was set to 200 ms. Extracted ion chromatograms were used for quantification of Ap3 A: m/z 755.0 ± 0.2, Ap4 A: m/z 835.0 ± 0.2, Ap5 A: m/z 915.0 ± 0.2 and Ap6 A: m/z 994.9 ± 0.2. To evaluate the identity of the Apn As an aliquot of 2 l was injected and MS/MS ESI-fragment spectra were acquired. All data were acquired and processed using “Compass 1.3 Software” (Bruker Daltonik, Bremen, Germany). Calculations including calibration curve regressions, sample concentrations values and statistics were performed by using “GraphPad Prism 5.0” software (GraphPad Software, San Diego, USA). 2.4. Immobilization of phosphodiesterase-I Phosphodiesterase-I (10 units) was diluted in 1000 l of coupling buffer (0.5 M NaCl, 0.1 M NaHCO3 , pH 8.3) for immobilization of the enzyme. Cyanogen-bromide activated sepharose resin (500 l) was swollen in 10 ml of 1 mM cold HCl for 1 h. The resin was than washed 3 times with 10 ml of 1 mM HCl and 3 times with 10 ml of coupling buffer and the diluted phosphodiesterase I was immediately transferred to 500 l of the washed resin. The immobilization was carried out for 1 h at room temperature. After the immobilization the unbound substrate was washed out 3 times using 1.5 ml of blocking buffer (1 M NaCl, 0.05 M glycine, pH 3.5). The supernatant was removed and the unbound reactive group of the sepharose resin were blocked using blocking buffer incubated for 2 h at room temperature. After the blocking procedure the resin was washed alternately using 1.5 ml of washing buffer 1 (0.5 M NaCl, 0.1 M NaCH3 COO, pH 4) and 1.5 ml of washing buffer 2 (0.5 M NaCl, 0.1 M Tris, pH 8). The immobilized phosphodiesterase-I was stored in PBS containing 0.02% NaN3 at 4 ◦ C until use. Prior to use the immobilized phosphodiesterase-I was washed 5 times using 1 ml PBS. 2.5. Preparation of diadenosine polyphosphate free plasma To degrade endogenous diadenosine polyphosphates an aliquot of plasma (500 l) was incubated with immobilized phosphodiesterase-I (100 l resin) for 24 h at 37 ◦ C. An aliquot of
A. Schulz et al. / J. Chromatogr. B 961 (2014) 91–96
100 l was screened for presence of the endogenous diadenosine polyphosphate. 2.6. Preparation of calibration standards and quality control samples Standard stock solutions of Ap3 A, Ap4 A, Ap5 A and Ap6 A were prepared in LC–MS grade water in polypropylene tubes at a concentration of 10 M. The quality control samples (QC) were prepared in diadenosine polyphosphate free plasma with Ap3 A, Ap4 A, Ap5 A, Ap6 A at a concentration of 1.9 nM for lower limit of quantification (LLOQ), at 12.5 nM for low quality control (LQC), at 50 nM for middle quality control (MQC) and at 100 nM for high quality control (HQC). For quantification of concentration-mass signal intensity correlation, a serial dilution of calibration standards (CS) in the range of 1.9–125 nM for Ap3 A, Ap4 A, Ap5 A, Ap6 A was prepared.
93
centrifuged at 2500 × g for 5 min at 4 ◦ C in order to separate plasma from blood cells. The samples were divided in 100 l aliquots and stored at −20 ◦ C prior to use. Before sample preparation, frozen samples were thawed at room temperature and vortexed. The samples (100 l) were diluted with 900 l of 100 mM ammonium formate solution (pH 4.5). The extraction was performed using Oasis WAX cartridges (Waters, Germany). The cartridges were conditioned using 1 ml methanol, equilibrated using 1 ml 100 mM ammonium formate (pH 4.5) and the diluted plasma samples were transferred to the cartridge. After washing with 1 ml 100 mM ammonium formate, 1 ml LC–MS grade water and 1 ml 100% methanol the retained compounds were eluted using 1 ml of 5% ammonium hydroxide in 50% methanol/water. The eluted fraction was lyophilized and reconstituted in 10 l of 1% DMHA. An aliquot (2 l) was injected onto LC–MS system for analysis. 2.8. Method validation
2.7. Preparation of plasma samples Pheripheral blood (5 ml) of healthy subjects (n = 17) was obtained by catherization of the cubital vein and was collected in tubes containing K2 -EDTA (7.2 mg). The blood samples were
The validation of the method was performed using calibration standards (CS) and quality controls (QC). The quantification method was validated in terms of selectivity, matrix effects, calibration range, intra- and inter-day precision, intra- and inter-day accuracy,
Fig. 2. Representative extracted ion chromatograms of endogenous diadenosine polyphosphates (A1, B1, C1, D1) as well as the corresponding negative ion ESI-MS (A2, B2, C2, D2) and ESI-MS/MS spectra (A3, B3, C3, D3).
94
A. Schulz et al. / J. Chromatogr. B 961 (2014) 91–96
and recovery. Due to the presence of endogenous Apn As in human plasma the method was validated using plasma, which was previously incubated with immobilized phosphodiesterase I in order to degrade endogenous Apn As.
we used plasma incubated with immobilized phosphodiesterase-I in order to degrade endogenous Apn As.
2.9. Matrix effects
We make use of the polar nature of Apn As to extract the substances from plasma by using weak anion-exchange cartridges. For sufficient retention time and chromatographic separation of Apn A (with n = 3–6) on reversed phase column, usage of ion pair reagent was essential. N,N-dimethylhexylamine (0.1%) was found to be optimal for the chromatographic separation as well as massspectrometric detection of Apn As. The basic mobile phase (pH 9.0) improved the separation and facilitated the ionization of Apn As in the negative ion mode. We used 1 mM ammonium formate buffer to ensure the pH stability over several days without any effect on the ionization efficiency. The chromatographic separation of Apn As and the corresponding negative ESI-MS ion spectra of Apn As are shown in Fig. 2. The identity of the endogenous Apn As was revealed by
The qualitative analysis of matrix effects of endogenous plasma compounds from six plasma extracts was performed using the method defined by Bonfiglio et al. [15]. The post-column infusion system was used for these experiments. Two l of the plasma extract was injected onto column and the gradient programme was started while the analytes were being infused post column (concentration 1 M) at a flow rate 4 l/min. Additionally, we quantified the matrix effects of plasma extracts from six healthy volunteers by comparison of the peak area of Apn As at a concentration of 20 nM in plasma extract and in 1% DMHA.
3.1. Hplc and mass spectrometric operating conditions
2.10. Calibration range The calibration range of the assay was assessed using calibration standards (QS) at seven different concentrations within the range 1.9–125 nM for all Apn As (with n = 3–6). Calibration curves were calculated from concentrations of 1.9 nM, 3.9 nM, 7.8 nM, 15.6 nM, 31.2 nM, 62.5 nM, 125 nM prepared in Apn As free plasma. The calibration curve was calculated using peak area of the analyte vs. concentration.
A
ESI interface
syringe
mass detector
tee HPLC column
B 1.2×10 6
2.11. Precision and accuracy
8.0×10 5
2.12. Recovery rate
4.0×10 5 0 0
1
2
3
4
5
6
C 1.2 ×10 6
relative MS intensity
The intra-day (n = 3) and inter-day (n = 9 within 3 separate days) precisions and accuracies were determined by analysing spiked quality controls in plasma. The quality control sample at the lower limit of quantification (LLOQ) was spiked with Ap3 A, Ap4 A, Ap5 A and Ap6 A at the end concentration of 1.9 nM. Low quality control (LQC) was prepared at a concentration of 12.5 nM, middle quality control (MQC) at a concentration of 50 nM and high quality control (HQC) at a concentration of 100 nM. The quality controls represented the entire range of the calibration curve. Precision was calculated from the coefficient of variation (% CV) of replicates. Accuracy was calculated by comparison of the measured concentration of spiked analyte with expected concentrations (% bias). Coefficient of variation in the range of mean observed concentration did not exceed 15% at all concentrations. Accuracy was within 15%. At LLOQ accuracy was within 20% and precision was less than 20%.
8.0 ×10 5 4.0 ×10 5 0 0
1
2
3
4
5
6
D
3×10 0 6 2×10 0 6 1×10 0 6 0
The recovery rate of Ap3 A, Ap4 A, Ap5 A and Ap6 A were evaluated by comparison of peak area of each analyte with standards prepared in 1% DMHA at the concentrations of the quality controls representing 100% recovery. 2.13. Statistics All results are presented as means ± SEM. 3. Results and discussion We developed and validated a simple ESI–LC–MS method for determination of Apn As (n = 3–6) in human plasma in this study. The schematic overview of the established method is shown in Fig. 1B. Due to lack of Apn As free matrix for the validation of our method
0
1
2
3
4
5
6
E 3×10 0 6 2×10 0 6 1×10 0 6 0 0
1
2
3
4
5
6
retention time [min] Fig. 3. Investigation of ion suppression effects by the post column infusion method defined by Bonfiglio et al. [15]. (A) Scheme of the post-column infusion system; (B)–(E) overlay chromatograms showing the matrix effects due to endogenous plasma compounds for Ap3 A (B), Ap4A (C), Ap5A (D) and Ap6 A (D).
A. Schulz et al. / J. Chromatogr. B 961 (2014) 91–96
95
Table 1 Accuracy and precision of quality control samples. QC
Concentration (nM)
Intra-day (n = 3) Accuracy (%)a
Intra-day (n = 9) Precision (%)b
Accuracy (%)a
Precision (%)b
Ap3 A
LLOQ LQC MQC HQC
1.9 12.5 50 100
109.6 95.5 100.3 109.7
5.1 6.2 3.8 1.0
106.6 96.2 102.0 106.1
9.6 6.2 4.0 3.4
Ap4 A
LLOQ LQC MQC HQC
1.9 12.5 50 100
108.2 100.0 101.9 106.9
1.7 8.6 2.5 0.6
102.4 97.7 102.9 105.2
10.5 7.3 1.6 0.9
Ap5 A
LLOQ LQC MQC HQC
1.9 12.5 50 100
100.6 100.9 97.7 110.9
11.4 5.1 3.9 1.3
105.4 97.1 97.8 106.2
11.4 3.49 3.2 0.9
Ap6 A
LLOQ LQC MQC HQC
1.9 12.5 50 100
106.3 90.5 107.0 91.8
108.9 91.4 101.0 98.8
8.7 3.9 3.2 0.8
a b
6.4 1.62 2.42 0.1
Calculated as (mean determined amount/nominal amount × 100). Calculated as % CV (SD/mean) × 100.
comparison of ESI-MS/MS fragment spectra of isolated endogenous compounds with those of authentic Apn As (Fig. 2A3–D3). 3.2. Selectivity To ascertain the selectivity of the method Apn As free plasma (incubated with phosphodiesterase-I) samples and plasma samples containing endogenous Apn As were analyzed in detail. No significant interfering peaks from the Apn As free plasma were found at the retention times of each analyte. 3.3. Matrix effects The post column infusion method, defined by Bonfiglio et al. [15] was used to evaluate the matrix effects. The schematic post-column infusion system is shown in Fig. 3A. Significant signal suppression was not observed for Apn As (n = 3–6) at their specific retention time.
3.4. Calibration range The seven point calibration curves of Apn As (n = 3–6) showed a reliable reproducibility in the concentration range 1.9–125 nM. The calibration curve prepared by plotting peak-area vs. concentration was fitted to quadratic regression. The coefficients of determination r2 for validation were found to be greater than 0.998 (Fig. 4A–D).
A
2.1×10 6
2
R =0.998
1.4×10 6 7.0×10 5
0
25
50
75
100
R =0.998 1×10 0 6
125
0
2.0×10 6
25
50
75
100
concentration [nM]
50
75
100
125
125
D
4.5×10 6
peak area Ap 6 A
peak area Ap 5 A
R2=0.998
0
25
concentration [nM] C
4.0×10 6
2
2×10 0 6
concentration [nM] 6.0×10 6
B
3×10 0 6
peak area Ap4 A
peak area Ap 3 A
Only a slight drop in the mass peak intensity was detected at 0.3 and 1.8 min due to poorly retained plasma compounds. Fig. 3 shows matrix effects chromatogram overlaid by chromatogram of Apn As (n = 3–6) standards prepared in 1% DMHA to indicate the elution profile for the analyte over the infusion matrix effect baseline. The quantitative determination of the matrix effects revealed minor influence was quantified as −1.1 ± 2.6 for Ap3 A, 0.9 ± 2.4 for Ap4 A, −3.2 ± 1.7 for Ap5 A and 0.3 ± 2.7 for Ap6 A (% ± SEM each)
R2 =0.998 3.0×10 6 1.5×10 6
0
25
50
75
100
concentration [nM]
Fig. 4. Calibration curves of Ap3 A (A), Ap4 A (B), Ap5 A (C) and Ap6 A (D).
125
96
A. Schulz et al. / J. Chromatogr. B 961 (2014) 91–96
The lower limit of quantification (LLOQ) was calculated at the lowest amount of the calibration curve with a precision not exceeding 20% and accuracy within 20% of the actual value. The LLOQ for all Apn As (n = 3–6) was found to be 1.9 nM. 3.5. Precision and accuracy The intra- (n = 3) and inter-day (n = 9) precisions and accuracies were determined for each Apn As (with n = 3–6) at a concentration of 1.9 nM (LLOQ), 12.5 nM (LQC), 50 nM (MQC) and 100 nM (HQC). The inter-day precision and accuracy were assessed within three separate days. The inter- and intra-day precisions expressed as percent coefficient of variation (% CV) ranged from 1.0% to 9.6% for Ap3 A, from 0.6% to 10.5% for Ap4 A, from 0.9% to 11.4% for Ap5 A, from 0.1% to 8.7% for Ap6 A. The inter- and intra-day accuracies defined as percent of nominal values were between 95.5% and 109.7% for Ap3 A, 97.7% and 108.2% for Ap4 A, 97.1% and 110.9% for Ap5 A, 90.5% and 108.9% for Ap6 A. The analysis of the accuracy and precision is given in Table 1. Those results indicate an acceptable precisions and accuracies of this new method for quantification of Apn As (n = 3–6) in human plasma. 3.6. Recovery rates The mean recovery rate was found to be 74.3 ± 2.0% for Ap3 A, 60.1 ± 1.2% for Ap4 A, 29.5 ± 1.1% for Ap5 A and 34.3 ± 2.2% for Ap6 A. The recovery rates seem to be less dependent on the number of phosphate groups and pKa value resulting in differently strong elution from the weak anion exchange sorbent. The Apn As were extracted from plasma by using a mixed-mode weak anion exchange and reversed-phase sorbent, which is used for the isolation of strong acid from different matrices. It combines hydrophobic and ionic retention of analytes. This extraction is based on the assumption that strong acid are always ionized and adsorb on the ionized sorbent. Trough the pH change, the sorbent become neutral and the analytes are eluted. Since the different Apn As are characterized by different pKa values there may be still unionized part of Apn As especially for Ap3 A (pKa = 0.9) and Ap4 A (pKa = 0.6) and this part is retained using hydrophobic interactions resulting in a higher recovery rate than for Ap5 A (pKa = 0.4) and Ap6 A (pKa = 0.2). However, each recovery rate is sufficient to quantify the analyte in the determined calibration range. 3.7. Quantification of diadenosine polyphosphates in human plasma samples Plasma samples from seventeen healthy volunteers were analyzed for diadenosine polyphosphates concentrations. The mean
plasma concentrations of Apn As were found to be 31.9 ± 5.9 nM for Ap3 A, 40.4 ± 6.6 nM for Ap4 A, 10.7 ± 1.5 nM for Ap5 A and 10.0 ± 18.9 nM for Ap6 A. These concentrations were comparable with these recently published by Wright et al. [16]. In conclusion, the method described in this study is simple, reliable, highly sensitive and selective for simultaneous quantification of Apn As (with n = 3–6) in human plasma. The method involves a simple sample preparation using weak anion exchange extraction cartridges followed by chromatographic separation within 12.5 min and mass spectrometric detection in negative ion mode. Comparison of the MS/MS spectra of the endogenous Apn As with those of synthetic Apn As confirmed the identity, which cannot be ensured by using UV absorbance as a detection method. Additionally, high sensitivity is achieved using only 100 l plasma. This new method is suitable for determination of diadenosine polyphosphates in plasma of patients in large clinical studies. Acknowledgements This study was supported by a grant of the German Research Foundation (DFG, Ja-972/11–1/3), by grant FP7-HEALTH “SysKid” (grant agreement 241544) and to “Mascara” (grant agreement 278249) from the European Union. References [1] H. Flodgaard, H. Klenow, Biochem. J. 208 (1982) 737. [2] J. Luthje, A. Ogilvie, Biochem. Biophys. Res. Commun. 115 (1983) 253. [3] E. Castro, M. Torres, M.T. Miras-Portugal, M.P. Gonzalez, Br. J. Pharmacol. 100 (1990) 360. [4] A. Klishin, N. Lozovaya, J. Pintor, M.T. Miras-Portugal, O. Krishtal, Neuroscience 58 (1994) 235. [5] E.B. Pivorun, A. Nordone, J. Neurosci. Res. 44 (1996) 478. [6] J. Jankowski, V. Jankowski, U. Laufer, M. van der Giet, L. Henning, M. Tepel, W. Zidek, H. Schluter, Arterioscler. Thromb. Vasc. Biol. 23 (2003) 1231. [7] J. Jankowski, P. Grosse-Huttmann, W. Zidek, H. Schluter, Rapid Commun. Mass Spectrom. 17 (2003) 1189. [8] V. Jankowski, A. Schulz, A. Kretschmer, H. Mischak, M. van der Giet, D. Janke, M. Schuchardt, R. Herwig, W. Zidek, J. Jankowski, J. Mol. Med. 91 (2013) 1095. [9] M.D. Baxi, J.K. Vishwanatha, J. Pharmacol. Toxicol. Methods 33 (1995) 121. [10] V. Jankowski, M. van der Giet, H. Mischak, M. Morgan, W. Zidek, J. Jankowski, Br. J. Pharmacol. 157 (2009) 1142. [11] T. Matsumoto, R.C. Tostes, R.C. Webb, Pharmacol. Res. 65 (2012) 81. [12] J. Jankowski, J. Hagemann, M.S. Yoon, M. van der Giet, N. Stephan, W. Zidek, H. Schluter, M. Tepel, Kidney Int. 59 (2001) 1134. [13] H. Chang, I.B. Yanachkov, A.D. Michelson, Y. Li, M.R. Barnard, G.E. Wright, A.L. Frelinger 3rd, Thromb. Res. 125 (2010) 159. [14] E. Gelpi, J. Chromatogr. A 703 (1995) 59. [15] R. Bonfiglio, R.C. King, T.V. Olah, K. Merkle, Rapid Commun. Mass Spectrom. 13 (1999) 1175. [16] M. Wright, A.D. Miller, Anal. Biochem. 432 (2013) 103.
Contents lists available at ScienceDirect
Journal of Chromatography B journal homepage: www.elsevier.com/locate/chromb
Highly sensitive, selective and rapid LC–MS method for simultaneous quantification of diadenosine polyphosphates in human plasma Anna Schulz a , Vera Jankowski b , Walter Zidek a , Joachim Jankowski b,∗ a b
Charité-Universitätsmedizin Berlin (CBF), Medizinische Klinik IV, Germany Universitätsklinikum RWTH Aachen, Institute of Molecular Cardiovascular Research, Aachen, Germany
a r t i c l e
i n f o
Article history: Received 16 February 2014 Accepted 11 May 2014 Available online 19 May 2014 Keywords: Dinucleoside polyphosphates Mediators Quantification Biomolecule Hypertension
a b s t r a c t Background: Diadenosine polyphosphates (Apn As) are endogenous mediators involved in large number of physiologic and pathophysiologic processes. The quantification of diadenosine polyphosphates in plasma and biological matrices is still challenging. Therefore, there is an urgent need for a simple and reliable quantification method suitable for clinical studies. The classical quantification of diadenosine polyphosphates is based on chromatographic separation and UV adsorption of the resulting fractions. These procedures are associated with low selectivity due to co-eluting plasma components. Therefore, we developed and validated a highly sensitive, selective and rapid LC–ESI–MS method for simultaneous quantification of Apn As (with n = 3–6) in human plasma within this study. The identities of the endogenous Apn As (with n = 3–6) were revealed by comparison of ESI-MS/MS fragment spectra of isolated endogenous compounds with those of authentic Apn As. Methods: Diadenosine polyphosphates were extracted from 100 l human plasma using weak anionexchange extraction cartridges. The separation of Apn As was achieved using capillary C18 columns. ESIHCT mass spectrometer (Bruker Daltonik, Germany) operated in negative ion mode was used for detection and quantification of Apn As. Results: A calibration curve was established for diadenosine polyphosphate free plasma in the concentration range 1.9–125 nM (r2 > 0.998) for all analytes. The intra- and inter-day accuracies were in the range of 91.4% and 110.9%. The intra- and inter-day precisions were determines as 0.1% and 11.4%, respectively. The mean plasma concentrations of Apn As were quantified as 31.9 ± 5.9 nM for Ap3 A, 40.4 ± 6.6 nM for Ap4 A, 10.7 ± 1.5 nM for Ap5 A and 10.0 ± 18.9 nM for Ap6 A. Discussion: The developed and validated ESI MS-based method for quantification of diadenosine polyphosphates in human plasma was successfully evaluated within the study. Conclusion Since the quantification is based on a volume of 100 l plasma, this method is highly applicable for clinical applications aiming at the validation of the impact of highly physiological and pathophysiological active diadenosine polyphosphates. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Diadenosine polyphosphates (Apn As) are endogenous mediators, isolated from e.g. blood platelets [1,2], adrenal medullary chromaffin granules [3,4], the central nervous system [4,5] and human plasma [6]. They are formed by two adenosine nucleosides interconnected by a phosphate chain, which varies from 3 up to 6 phosphates [7,8]. The schematic chemical structure of
∗ Corresponding author at: RWTH University of Aachen, Institute of Molecular Cardiovascular Research, Pauwelsstrasse 30, 52074 Aachen, Germany. Tel.: +49 241 80 80580; fax: +49 241 80 52716. E-mail address: [email protected] (J. Jankowski). http://dx.doi.org/10.1016/j.jchromb.2014.05.018 1570-0232/© 2014 Elsevier B.V. All rights reserved.
Apn As is shown in Fig. 1A. Apn As have been demonstrated to control numerous physiological and pathophysiological functions like neurotransmission, homeostasis as well as vascular regulation [9–11]. Furthermore, they are involved in cardio-renal processes and may have an impact on the progression of chronic kidney disease (CKD) [12,13]. Therefore, the quantification of the endogenous Apn As concentrations is of utmost importance since Apn As might be potential biomarker or target for specific therapy in the progression of chronic kidney disease. Until now, just a few methods have been described for the determination of Apn As in human plasma. These analytical methods are mostly based on liquid chromatography coupled with detection of the UV absorbance at 254 nm. The sensitivity is not exactly specified in those methods. Furthermore, for the quantification of Apn As
92
A. Schulz et al. / J. Chromatogr. B 961 (2014) 91–96
NH2
NH 2 N
N
N N
N
O
N O H
H
OH
H OH
H
A
2.2. High performance liquid chromatography conditions
B
An Agilent 1200 (Agilent, Böblingen, Germany) capillary LC system equipped with micro vacuum degasser (G1379B) and capillary pump (G1376A) along with an autosampler (G1377A) was used to inject 2 l aliquots of prepared samples. The separation was carried out on a Zorbax C18 XDB 0.5 mm × 35 mm column with 3.5 m particle size (Agilent, Böblingen, Germany). The temperature of the column was maintained at constant 50 ± 0.5 ◦ C. The eluent A was consisted of 0.1% N,N-DMHA in 1 mM ammonium formate prepared in LC–MS grade water. The pH of eluent A was adjusted using formic acid to 9.0. Eluent B was LC–MS grade acetonitrile. The injection volume was 2 l. The following gradient was used to separate Apn As: 0 min, 0% B, 0–1 min 0–12% B, 1–1.75 min 1–16% B, 1.75–2.5 min 16–22% B, 2.5–5.5 min 22–98% B, 5.5–6.5 min 98% B, 6.5–6.6 min 98–0% B, 6.6–12.5 min 0% B. The primary flow was 500 l/min and it was split to 100 l/min by using electronic flow control. The total chromatographic run time was 12.5 min.
N N
O
O P O O-
H
H
OH
H OH
H n=3-6
dilution of plasma with 100 mM ammonium formate weak anion exchange SPE Evaporation Reconstitution in 1% N,N-DMHA LC-ESI-MS
Fig. 1. (A) Molecular structure of diadenosine polyphosphates (Ap3–6 A) and (B) scheme of the preparation steps for the quantification of diadenosine polyphosphates.
by using those methods a high plasma volume is required. Since the CKD patients are suffering from anaemia the plasma volume should be minimized. In general, the quantification of endogenous Apn As levels and the validation of the analytical method are still challenging. The limitations are: (a) the concentration of diadenosine polyphosphates is in the nanomolar range, (b) short half-life times of diadenosine polyphosphates, and (c) the minimal differences in the chemical characteristics of diadenosine polyphosphates. In addition, since diadenosine polyphosphates are endogenous components of plasma, a diadenosine polyphosphates-free matrix is not directly available, which has an impact on the determination of the lower limit of quantification. Since chromatographic fractionation diadenosine polyphosphates are highly polar, the use of buffer systems and/or ion pair reagent to ensure a sufficient retention by chromatographic media are mandatory. To avoid the co-detection of eluting substances from the chromatographic column, a highly sensitive and selective method for detection of the eluting substances is essential. UV adsorption as used in former studies does not fulfil this criterion. In recent years, liquid chromatography tandem mass spectrometry has been characterized as a selective and highly robust method for quantification of metabolites and endogenous mediators [14]. Therefore, we developed, established and validated a liquid chromatography electrospray-ionization massspectrometric-based method for simultaneous quantification of Apn As in human plasma in this study.
2. Materials and methods 2.1. Chemicals and reagents Diadenosine polyphosphates (Ap3 A, Ap4 A, Ap5 A, Ap6 A), analytical grade ammonium formate, formic acid, ammonium hydroxide and N,N-dimethylhexylamine (DMHA) were purchased from Sigma Aldrich (Seelze, Germany). Liquid chromatography-mass spectrometry (LC–MS) grade water and acetonitrile were purchased from Fisher Scientific (Fair Lavn, United States). Oasis WAX cartridges were obtained from Waters (Eschborn, Germany).
2.3. Mass-spectrometric analysis The detection of Apn As was achieved in negative ion mode for all analytes by using a HCT mass spectrometer (Bruker Daltonic, Bremen, Germany), equipped with an electrospray ionization interface. The ionization parameters of nebulizer gas, dry gas and dry temperature were 25 psi, 9 l/min, and 300 ◦ C respectively. The ion spray voltage was set to 3000 V. The maximal accumulation time was set to 200 ms. Extracted ion chromatograms were used for quantification of Ap3 A: m/z 755.0 ± 0.2, Ap4 A: m/z 835.0 ± 0.2, Ap5 A: m/z 915.0 ± 0.2 and Ap6 A: m/z 994.9 ± 0.2. To evaluate the identity of the Apn As an aliquot of 2 l was injected and MS/MS ESI-fragment spectra were acquired. All data were acquired and processed using “Compass 1.3 Software” (Bruker Daltonik, Bremen, Germany). Calculations including calibration curve regressions, sample concentrations values and statistics were performed by using “GraphPad Prism 5.0” software (GraphPad Software, San Diego, USA). 2.4. Immobilization of phosphodiesterase-I Phosphodiesterase-I (10 units) was diluted in 1000 l of coupling buffer (0.5 M NaCl, 0.1 M NaHCO3 , pH 8.3) for immobilization of the enzyme. Cyanogen-bromide activated sepharose resin (500 l) was swollen in 10 ml of 1 mM cold HCl for 1 h. The resin was than washed 3 times with 10 ml of 1 mM HCl and 3 times with 10 ml of coupling buffer and the diluted phosphodiesterase I was immediately transferred to 500 l of the washed resin. The immobilization was carried out for 1 h at room temperature. After the immobilization the unbound substrate was washed out 3 times using 1.5 ml of blocking buffer (1 M NaCl, 0.05 M glycine, pH 3.5). The supernatant was removed and the unbound reactive group of the sepharose resin were blocked using blocking buffer incubated for 2 h at room temperature. After the blocking procedure the resin was washed alternately using 1.5 ml of washing buffer 1 (0.5 M NaCl, 0.1 M NaCH3 COO, pH 4) and 1.5 ml of washing buffer 2 (0.5 M NaCl, 0.1 M Tris, pH 8). The immobilized phosphodiesterase-I was stored in PBS containing 0.02% NaN3 at 4 ◦ C until use. Prior to use the immobilized phosphodiesterase-I was washed 5 times using 1 ml PBS. 2.5. Preparation of diadenosine polyphosphate free plasma To degrade endogenous diadenosine polyphosphates an aliquot of plasma (500 l) was incubated with immobilized phosphodiesterase-I (100 l resin) for 24 h at 37 ◦ C. An aliquot of
A. Schulz et al. / J. Chromatogr. B 961 (2014) 91–96
100 l was screened for presence of the endogenous diadenosine polyphosphate. 2.6. Preparation of calibration standards and quality control samples Standard stock solutions of Ap3 A, Ap4 A, Ap5 A and Ap6 A were prepared in LC–MS grade water in polypropylene tubes at a concentration of 10 M. The quality control samples (QC) were prepared in diadenosine polyphosphate free plasma with Ap3 A, Ap4 A, Ap5 A, Ap6 A at a concentration of 1.9 nM for lower limit of quantification (LLOQ), at 12.5 nM for low quality control (LQC), at 50 nM for middle quality control (MQC) and at 100 nM for high quality control (HQC). For quantification of concentration-mass signal intensity correlation, a serial dilution of calibration standards (CS) in the range of 1.9–125 nM for Ap3 A, Ap4 A, Ap5 A, Ap6 A was prepared.
93
centrifuged at 2500 × g for 5 min at 4 ◦ C in order to separate plasma from blood cells. The samples were divided in 100 l aliquots and stored at −20 ◦ C prior to use. Before sample preparation, frozen samples were thawed at room temperature and vortexed. The samples (100 l) were diluted with 900 l of 100 mM ammonium formate solution (pH 4.5). The extraction was performed using Oasis WAX cartridges (Waters, Germany). The cartridges were conditioned using 1 ml methanol, equilibrated using 1 ml 100 mM ammonium formate (pH 4.5) and the diluted plasma samples were transferred to the cartridge. After washing with 1 ml 100 mM ammonium formate, 1 ml LC–MS grade water and 1 ml 100% methanol the retained compounds were eluted using 1 ml of 5% ammonium hydroxide in 50% methanol/water. The eluted fraction was lyophilized and reconstituted in 10 l of 1% DMHA. An aliquot (2 l) was injected onto LC–MS system for analysis. 2.8. Method validation
2.7. Preparation of plasma samples Pheripheral blood (5 ml) of healthy subjects (n = 17) was obtained by catherization of the cubital vein and was collected in tubes containing K2 -EDTA (7.2 mg). The blood samples were
The validation of the method was performed using calibration standards (CS) and quality controls (QC). The quantification method was validated in terms of selectivity, matrix effects, calibration range, intra- and inter-day precision, intra- and inter-day accuracy,
Fig. 2. Representative extracted ion chromatograms of endogenous diadenosine polyphosphates (A1, B1, C1, D1) as well as the corresponding negative ion ESI-MS (A2, B2, C2, D2) and ESI-MS/MS spectra (A3, B3, C3, D3).
94
A. Schulz et al. / J. Chromatogr. B 961 (2014) 91–96
and recovery. Due to the presence of endogenous Apn As in human plasma the method was validated using plasma, which was previously incubated with immobilized phosphodiesterase I in order to degrade endogenous Apn As.
we used plasma incubated with immobilized phosphodiesterase-I in order to degrade endogenous Apn As.
2.9. Matrix effects
We make use of the polar nature of Apn As to extract the substances from plasma by using weak anion-exchange cartridges. For sufficient retention time and chromatographic separation of Apn A (with n = 3–6) on reversed phase column, usage of ion pair reagent was essential. N,N-dimethylhexylamine (0.1%) was found to be optimal for the chromatographic separation as well as massspectrometric detection of Apn As. The basic mobile phase (pH 9.0) improved the separation and facilitated the ionization of Apn As in the negative ion mode. We used 1 mM ammonium formate buffer to ensure the pH stability over several days without any effect on the ionization efficiency. The chromatographic separation of Apn As and the corresponding negative ESI-MS ion spectra of Apn As are shown in Fig. 2. The identity of the endogenous Apn As was revealed by
The qualitative analysis of matrix effects of endogenous plasma compounds from six plasma extracts was performed using the method defined by Bonfiglio et al. [15]. The post-column infusion system was used for these experiments. Two l of the plasma extract was injected onto column and the gradient programme was started while the analytes were being infused post column (concentration 1 M) at a flow rate 4 l/min. Additionally, we quantified the matrix effects of plasma extracts from six healthy volunteers by comparison of the peak area of Apn As at a concentration of 20 nM in plasma extract and in 1% DMHA.
3.1. Hplc and mass spectrometric operating conditions
2.10. Calibration range The calibration range of the assay was assessed using calibration standards (QS) at seven different concentrations within the range 1.9–125 nM for all Apn As (with n = 3–6). Calibration curves were calculated from concentrations of 1.9 nM, 3.9 nM, 7.8 nM, 15.6 nM, 31.2 nM, 62.5 nM, 125 nM prepared in Apn As free plasma. The calibration curve was calculated using peak area of the analyte vs. concentration.
A
ESI interface
syringe
mass detector
tee HPLC column
B 1.2×10 6
2.11. Precision and accuracy
8.0×10 5
2.12. Recovery rate
4.0×10 5 0 0
1
2
3
4
5
6
C 1.2 ×10 6
relative MS intensity
The intra-day (n = 3) and inter-day (n = 9 within 3 separate days) precisions and accuracies were determined by analysing spiked quality controls in plasma. The quality control sample at the lower limit of quantification (LLOQ) was spiked with Ap3 A, Ap4 A, Ap5 A and Ap6 A at the end concentration of 1.9 nM. Low quality control (LQC) was prepared at a concentration of 12.5 nM, middle quality control (MQC) at a concentration of 50 nM and high quality control (HQC) at a concentration of 100 nM. The quality controls represented the entire range of the calibration curve. Precision was calculated from the coefficient of variation (% CV) of replicates. Accuracy was calculated by comparison of the measured concentration of spiked analyte with expected concentrations (% bias). Coefficient of variation in the range of mean observed concentration did not exceed 15% at all concentrations. Accuracy was within 15%. At LLOQ accuracy was within 20% and precision was less than 20%.
8.0 ×10 5 4.0 ×10 5 0 0
1
2
3
4
5
6
D
3×10 0 6 2×10 0 6 1×10 0 6 0
The recovery rate of Ap3 A, Ap4 A, Ap5 A and Ap6 A were evaluated by comparison of peak area of each analyte with standards prepared in 1% DMHA at the concentrations of the quality controls representing 100% recovery. 2.13. Statistics All results are presented as means ± SEM. 3. Results and discussion We developed and validated a simple ESI–LC–MS method for determination of Apn As (n = 3–6) in human plasma in this study. The schematic overview of the established method is shown in Fig. 1B. Due to lack of Apn As free matrix for the validation of our method
0
1
2
3
4
5
6
E 3×10 0 6 2×10 0 6 1×10 0 6 0 0
1
2
3
4
5
6
retention time [min] Fig. 3. Investigation of ion suppression effects by the post column infusion method defined by Bonfiglio et al. [15]. (A) Scheme of the post-column infusion system; (B)–(E) overlay chromatograms showing the matrix effects due to endogenous plasma compounds for Ap3 A (B), Ap4A (C), Ap5A (D) and Ap6 A (D).
A. Schulz et al. / J. Chromatogr. B 961 (2014) 91–96
95
Table 1 Accuracy and precision of quality control samples. QC
Concentration (nM)
Intra-day (n = 3) Accuracy (%)a
Intra-day (n = 9) Precision (%)b
Accuracy (%)a
Precision (%)b
Ap3 A
LLOQ LQC MQC HQC
1.9 12.5 50 100
109.6 95.5 100.3 109.7
5.1 6.2 3.8 1.0
106.6 96.2 102.0 106.1
9.6 6.2 4.0 3.4
Ap4 A
LLOQ LQC MQC HQC
1.9 12.5 50 100
108.2 100.0 101.9 106.9
1.7 8.6 2.5 0.6
102.4 97.7 102.9 105.2
10.5 7.3 1.6 0.9
Ap5 A
LLOQ LQC MQC HQC
1.9 12.5 50 100
100.6 100.9 97.7 110.9
11.4 5.1 3.9 1.3
105.4 97.1 97.8 106.2
11.4 3.49 3.2 0.9
Ap6 A
LLOQ LQC MQC HQC
1.9 12.5 50 100
106.3 90.5 107.0 91.8
108.9 91.4 101.0 98.8
8.7 3.9 3.2 0.8
a b
6.4 1.62 2.42 0.1
Calculated as (mean determined amount/nominal amount × 100). Calculated as % CV (SD/mean) × 100.
comparison of ESI-MS/MS fragment spectra of isolated endogenous compounds with those of authentic Apn As (Fig. 2A3–D3). 3.2. Selectivity To ascertain the selectivity of the method Apn As free plasma (incubated with phosphodiesterase-I) samples and plasma samples containing endogenous Apn As were analyzed in detail. No significant interfering peaks from the Apn As free plasma were found at the retention times of each analyte. 3.3. Matrix effects The post column infusion method, defined by Bonfiglio et al. [15] was used to evaluate the matrix effects. The schematic post-column infusion system is shown in Fig. 3A. Significant signal suppression was not observed for Apn As (n = 3–6) at their specific retention time.
3.4. Calibration range The seven point calibration curves of Apn As (n = 3–6) showed a reliable reproducibility in the concentration range 1.9–125 nM. The calibration curve prepared by plotting peak-area vs. concentration was fitted to quadratic regression. The coefficients of determination r2 for validation were found to be greater than 0.998 (Fig. 4A–D).
A
2.1×10 6
2
R =0.998
1.4×10 6 7.0×10 5
0
25
50
75
100
R =0.998 1×10 0 6
125
0
2.0×10 6
25
50
75
100
concentration [nM]
50
75
100
125
125
D
4.5×10 6
peak area Ap 6 A
peak area Ap 5 A
R2=0.998
0
25
concentration [nM] C
4.0×10 6
2
2×10 0 6
concentration [nM] 6.0×10 6
B
3×10 0 6
peak area Ap4 A
peak area Ap 3 A
Only a slight drop in the mass peak intensity was detected at 0.3 and 1.8 min due to poorly retained plasma compounds. Fig. 3 shows matrix effects chromatogram overlaid by chromatogram of Apn As (n = 3–6) standards prepared in 1% DMHA to indicate the elution profile for the analyte over the infusion matrix effect baseline. The quantitative determination of the matrix effects revealed minor influence was quantified as −1.1 ± 2.6 for Ap3 A, 0.9 ± 2.4 for Ap4 A, −3.2 ± 1.7 for Ap5 A and 0.3 ± 2.7 for Ap6 A (% ± SEM each)
R2 =0.998 3.0×10 6 1.5×10 6
0
25
50
75
100
concentration [nM]
Fig. 4. Calibration curves of Ap3 A (A), Ap4 A (B), Ap5 A (C) and Ap6 A (D).
125
96
A. Schulz et al. / J. Chromatogr. B 961 (2014) 91–96
The lower limit of quantification (LLOQ) was calculated at the lowest amount of the calibration curve with a precision not exceeding 20% and accuracy within 20% of the actual value. The LLOQ for all Apn As (n = 3–6) was found to be 1.9 nM. 3.5. Precision and accuracy The intra- (n = 3) and inter-day (n = 9) precisions and accuracies were determined for each Apn As (with n = 3–6) at a concentration of 1.9 nM (LLOQ), 12.5 nM (LQC), 50 nM (MQC) and 100 nM (HQC). The inter-day precision and accuracy were assessed within three separate days. The inter- and intra-day precisions expressed as percent coefficient of variation (% CV) ranged from 1.0% to 9.6% for Ap3 A, from 0.6% to 10.5% for Ap4 A, from 0.9% to 11.4% for Ap5 A, from 0.1% to 8.7% for Ap6 A. The inter- and intra-day accuracies defined as percent of nominal values were between 95.5% and 109.7% for Ap3 A, 97.7% and 108.2% for Ap4 A, 97.1% and 110.9% for Ap5 A, 90.5% and 108.9% for Ap6 A. The analysis of the accuracy and precision is given in Table 1. Those results indicate an acceptable precisions and accuracies of this new method for quantification of Apn As (n = 3–6) in human plasma. 3.6. Recovery rates The mean recovery rate was found to be 74.3 ± 2.0% for Ap3 A, 60.1 ± 1.2% for Ap4 A, 29.5 ± 1.1% for Ap5 A and 34.3 ± 2.2% for Ap6 A. The recovery rates seem to be less dependent on the number of phosphate groups and pKa value resulting in differently strong elution from the weak anion exchange sorbent. The Apn As were extracted from plasma by using a mixed-mode weak anion exchange and reversed-phase sorbent, which is used for the isolation of strong acid from different matrices. It combines hydrophobic and ionic retention of analytes. This extraction is based on the assumption that strong acid are always ionized and adsorb on the ionized sorbent. Trough the pH change, the sorbent become neutral and the analytes are eluted. Since the different Apn As are characterized by different pKa values there may be still unionized part of Apn As especially for Ap3 A (pKa = 0.9) and Ap4 A (pKa = 0.6) and this part is retained using hydrophobic interactions resulting in a higher recovery rate than for Ap5 A (pKa = 0.4) and Ap6 A (pKa = 0.2). However, each recovery rate is sufficient to quantify the analyte in the determined calibration range. 3.7. Quantification of diadenosine polyphosphates in human plasma samples Plasma samples from seventeen healthy volunteers were analyzed for diadenosine polyphosphates concentrations. The mean
plasma concentrations of Apn As were found to be 31.9 ± 5.9 nM for Ap3 A, 40.4 ± 6.6 nM for Ap4 A, 10.7 ± 1.5 nM for Ap5 A and 10.0 ± 18.9 nM for Ap6 A. These concentrations were comparable with these recently published by Wright et al. [16]. In conclusion, the method described in this study is simple, reliable, highly sensitive and selective for simultaneous quantification of Apn As (with n = 3–6) in human plasma. The method involves a simple sample preparation using weak anion exchange extraction cartridges followed by chromatographic separation within 12.5 min and mass spectrometric detection in negative ion mode. Comparison of the MS/MS spectra of the endogenous Apn As with those of synthetic Apn As confirmed the identity, which cannot be ensured by using UV absorbance as a detection method. Additionally, high sensitivity is achieved using only 100 l plasma. This new method is suitable for determination of diadenosine polyphosphates in plasma of patients in large clinical studies. Acknowledgements This study was supported by a grant of the German Research Foundation (DFG, Ja-972/11–1/3), by grant FP7-HEALTH “SysKid” (grant agreement 241544) and to “Mascara” (grant agreement 278249) from the European Union. References [1] H. Flodgaard, H. Klenow, Biochem. J. 208 (1982) 737. [2] J. Luthje, A. Ogilvie, Biochem. Biophys. Res. Commun. 115 (1983) 253. [3] E. Castro, M. Torres, M.T. Miras-Portugal, M.P. Gonzalez, Br. J. Pharmacol. 100 (1990) 360. [4] A. Klishin, N. Lozovaya, J. Pintor, M.T. Miras-Portugal, O. Krishtal, Neuroscience 58 (1994) 235. [5] E.B. Pivorun, A. Nordone, J. Neurosci. Res. 44 (1996) 478. [6] J. Jankowski, V. Jankowski, U. Laufer, M. van der Giet, L. Henning, M. Tepel, W. Zidek, H. Schluter, Arterioscler. Thromb. Vasc. Biol. 23 (2003) 1231. [7] J. Jankowski, P. Grosse-Huttmann, W. Zidek, H. Schluter, Rapid Commun. Mass Spectrom. 17 (2003) 1189. [8] V. Jankowski, A. Schulz, A. Kretschmer, H. Mischak, M. van der Giet, D. Janke, M. Schuchardt, R. Herwig, W. Zidek, J. Jankowski, J. Mol. Med. 91 (2013) 1095. [9] M.D. Baxi, J.K. Vishwanatha, J. Pharmacol. Toxicol. Methods 33 (1995) 121. [10] V. Jankowski, M. van der Giet, H. Mischak, M. Morgan, W. Zidek, J. Jankowski, Br. J. Pharmacol. 157 (2009) 1142. [11] T. Matsumoto, R.C. Tostes, R.C. Webb, Pharmacol. Res. 65 (2012) 81. [12] J. Jankowski, J. Hagemann, M.S. Yoon, M. van der Giet, N. Stephan, W. Zidek, H. Schluter, M. Tepel, Kidney Int. 59 (2001) 1134. [13] H. Chang, I.B. Yanachkov, A.D. Michelson, Y. Li, M.R. Barnard, G.E. Wright, A.L. Frelinger 3rd, Thromb. Res. 125 (2010) 159. [14] E. Gelpi, J. Chromatogr. A 703 (1995) 59. [15] R. Bonfiglio, R.C. King, T.V. Olah, K. Merkle, Rapid Commun. Mass Spectrom. 13 (1999) 1175. [16] M. Wright, A.D. Miller, Anal. Biochem. 432 (2013) 103.
