- Email: [email protected]
The separation and analysis of symmetric and asymmetric dimethylarginine and other hydrophilic isobaric compounds using aqueous normal phase chromatography
Journal of Chromatography A, 1441 (2016) 52–59
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
The separation and analysis of symmetric and asymmetric dimethylarginine and other hydrophilic isobaric compounds using aqueous normal phase chromatography Joseph J. Pesek ∗ , Maria T. Matyksa, Brent Modereger, Alejandra Hasbun, Vy T. Phan, Zahra Mehr, Mariano Guzman, Seiichiro Watanable Department of Chemistry, San Jose State University, San Jose, CA 95192, USA
a r t i c l e
i n f o
Article history: Received 26 January 2016 Received in revised form 19 February 2016 Accepted 22 February 2016 Available online 27 February 2016 Keywords: In-source fragmentation Silica hydride columns Clinical analyses
a b s t r a c t Two biologically important compounds with clinical relevance, asymmetric dimethylarginine and symmetric dimethylarginine, are analyzed using aqueous normal phase chromatography on silica hydride-based columns. Two different stationary phases were tested, a commercially available Diamond HydrideTM and a 2-acrylamido-2-methylpropane sulfonic acid experimental column. Two types of analytical protocols were investigated: analysis of the compounds when separation was achieved and analysis of the compounds with partial chromatographic separation. Urine samples from tuberculosis patients were tested for levels of asymmetric and symmetric dimethylarginine. The mass spectrometric technique of in-source fragmentation that can provide data similar to a tandem mass analyzer was evaluated as a means of identification and quantitation of the two compounds when complete separation is not achieved. This same protocol was also evaluated for two other isobaric compounds, glucose-1 and glucose-6 phohsphate, and leucine and isoleucine. © 2016 Elsevier B.V. All rights reserved.
1. Introduction Both asymmetric dimethylarginine (ADMA) and symmetric dimethylarginine (SDMA) plasma levels have been associated with various forms of renal and cardiac physiological conditions [1,2]. ADMA has been shown to be a potent inhibitor of the enzyme nitric oxide synthase (NOS) which regulates the conversion of arginine into the vasodilator nitric oxide [3–5]. ADMA levels in biological fluids have been shown to correlate closely with renal functions and could be used as a biomarker for certain diseases [6]. Also, it has been shown that above normal amounts of ADMA and SDMA in blood plasma can be used to predict the extent of detrimental effects from stroke [7]. Monitoring of ADMA has also been found to be of importance in acute critical conditions such as septic shock [8–10]. The extent of the clinical applications of ADMA and SDMA as well as other small polar physiological analytes makes the development of new approaches for their determination a topic of significant importance.
∗ Corresponding author. Fax: +1 408 924 4945. E-mail address: [email protected] (J.J. Pesek). http://dx.doi.org/10.1016/j.chroma.2016.02.071 0021-9673/© 2016 Elsevier B.V. All rights reserved.
HPLC based methods have been the most successful to date for the analysis of ADMA and SDMA [11]. A number of approaches have been proposed in developing analytical schemes for these two compounds. Because of their polarity, in many cases the compounds are derivatized in order to be retained in reversedphase and in some cases to also enhance detection. A very early method for the preparation of the two compounds utilized an ion-exchange column that required almost five hours for the separation to be completed [12]. Another early protocol described the analysis of underivatized ADMA/SDMA in plasma from renal failure patients that required considerable sample preparation and a 20 min analysis time using UV detection at 200 nm [13]. Typical examples of pre-column derivatization are methods for the determination of ADMA and SDMA in plasma samples that allow for reversed-phase retention and high sensitivity by fluorescence detection [14–17]. Another reversed-phase method with derivatization utilized MS/MS for detection and quantitation with the each compound having a daughter ion that was unique [18]. One other approach reported utilized hydrophilic liquid interaction chromatography (HILIC) to analyze underivatized ADMA and SDMA with MS/MS detection [19]. Some sensitivity that would have been gained at the high acetonitrile concentrations in the HILIC method may have been at least partially lost because the mobile phase con-
J.J. Pesek et al. / J. Chromatogr. A 1441 (2016) 52–59
53
Fig. 1. Overlayed 400 Mz proton magnetic resonance spectra of ADMA and SDMA standards used for chromatographic evaluation. (A) methyl region and (B) methylene region.
tained 0.025% trifluoroacetic acid, a known suppressor of MS signal intensity. Aqueous normal phase (ANP) chromatography utilizing silica hydride-based stationary phases is a versatile technique that can be applied to a broad range of analytical separations [20–22]. It versatility is based on the fact that all columns packed with silica hydride phases can operate in both the reversed-phase and normal-phase modes. There are a number of crucial differences between silica hydride and ordinary silica that lead to chromatographic advantages for the hydride material. These features include the dual retention capabilities cited above. In addition, the surface of silica hydride is slightly hydrophobic rather than highly polar due to silanols (as in HILIC phases) and thus often result in better peak shape for certain analytes, particularly bases. Other notable features of silica hydride include a very thin water shell
(∼0.5 monolayer) on the surface [23], rapid equilibration of the surface solvation layer after gradients and highly reproducible retention for polar compounds in the aqueous normal-phase mode. These properties have been used advantageously for the analysis of metabolites [24–28], nucleotides [29,30], pharmaceuticals [31–34], food and beverage components [35–39], lipophilic interferences [40] and chiral compounds [41]. Considering the proven capabilities of the silica hydride-based stationary phase for the analysis of hydrophilic compounds, the objective of this study was to determine if underivatized ADMA and SDMA could be better separated in a reasonable analysis time than in previous studies thus eliminating the need for extensive sample preparation or additional on-line procedures. In addition, the use of a newer and simpler mass spectrometric detection method, in-source fragmentation, for the detection of unique fragment ions
54
J.J. Pesek et al. / J. Chromatogr. A 1441 (2016) 52–59
on 4 m particle size silica hydride in a 150 mm × 2.1 mm column (MicroSolv Technology Corp.). Some preliminary screening experiments were also performed using a 50 mm × 2.1 mm AMPS column. ADMA was purchased from Sigma–Aldrich Co. LLC. (Sigma–Aldrich, St. Louis, MO, USA) and SDMA was purchased from Calbiochem (Billerica, MA, USA). Acetonitrile was obtained from GFS Chemicals (Columbus, OH, USA). Water was obtained from a Milli-Q water purification system (Millipore Corp., Billerica, MA, USA), and filtered with a 45 m Nylon filter (MicroSolv Technology Corp.). Acetonitrile, formic acid and glacial acetic acid were purchased from Sigma–Aldrich (LC-MS grade). Tuberculosis urine samples were obtained from Cornell-Weill Medical Center in New York, NY.
2.2. Instrumentation NMR spectra were acquired on a Varian (Palo Alto, CA, USA) INOVA 400 spectrometer. One HPLC system used in this study was an Agilent (Little Falls, DE, USA) 1100 Series LC system, including degasser, binary pump, temperature-controlled autosampler and temperature-controlled column compartment. The mass spectrometer system was an Agilent (Santa Clara, CA, USA) Model 6210 MSD TOF with a dual sprayer electrospray ionization (ESI) source. Additional LCMS studies were done using a Flexar (Perkin-Elmer, Waltham, MA, USA) UHPLC system, and a Flexar SQ 300 MS.
2.3. Methods
Fig. 2. Chromatographic elution of ADMA and SDMA under ANP conditions. (A) DH column: mobile phase; A, DI water + 0.1% acetic acid and B, acetonitrile + 0.1% acetic acid. Gradient; 0–1 min 80% B, 1–3 min to 70% B, 3–4 min 70% B, 4–6 min to 60% B, 6–7 min 60% B, 7–9 min to 50% B, 9–10 min 50% B, 10–12 min to 40% B, 12–13 min 40% B, 13–15 min to 30% B, 15–16 min 30% B, 16–18 to 10% B, 18–23 min 10%B. (B) AMPS column: mobile phase; A, DI water + 0.1% formic acid and B, acetonitrile + 0.1% formic acid. Isocratic at 70% B. (C) AMPS column: mobile phase; A, DI water + 0.2% acetic acid and B, acetonitrile + 0.2% acetic acid. Gradient; 0–1 min 95% B, 1–3 min to 70% B, 3–4 min 70% B, 4–5 min to 50% B, 5–6 min 50% B, 6–7 min to 20% B, 7–14 min 20% B. Detection at [M + H]+ 203.1503. Flow rate = 0.4 mL/min.
was explored. The objective was to determine if this approach has general applicability to the analysis of isobaric compounds that are partially separated chromatographically. These techniques were then applied to the analysis of real clinical samples. This format could be used to develop validated methods for these and other hydrophilic analytes including isobaric compounds.
Stock solutions of ADMA and SDMA were prepared at a concentration of 1 mg/mL in DI water. Sample solutions were prepared by diluting stock solutions in 50:50 acetonitrile/water. ADMA and SDMA samples were prepared by diluting the stock solution 1:19. Urine samples of healthy individuals were prepared by the Dept. of Nutrition, Food Science and Packaging at San Jose State University. Urine samples of tuberculosis patients and controls were prepared by Division of Infectious Diseases, Department of Medicine, Weill Cornell Medical College, New York NY. The mixed sample was prepared by combining equal volumes of stock solution, and then further diluting 1:10 with 50:50 DI water/ACN. The mobile phase organic solvent was composed of 0.1%, 0.2%, or 0.5% acetic acid in acetonitrile. Water containing 0.1%, 0.2%, or 0.5% acetic acid made up the difference, pairing solvents of equal acetic acid composition. The column flow rate was 0.4 mL/min. The column temperature was kept at 25 ◦ C. Various gradients were used in testing the columns and those with significance are reported in the text or figure captions. The gradient flow rate was 0.4 mL/min. MS detection was operated in the positive ion mode, monitoring at 203.1503 m/z (for exact mass detection) or 203.0000–203.2000 m/z. Fragmentation studies were monitored at the parent ion 203.1503 m/z as well as 46.0000 m/z (ADMA), and 172.0000 m/z (SDMA).
3. 3. Results and Discussion 2. Experimental 3.1. NMR sample authentication 2.1. Materials Two silica hydride stationary phases were used in this study. One phase was a Diamond HydrideTM material (4 m particle size, 150 mm × 2.1 mm in stainless steel tubing and 100 mm × 2.1 mm in PEEK tubing) obtained from Microsolv Technology (Eatontown, NJ, USA). The second phase was an experimental material with functionalized 2-acrylamido-2-methylpropane sulfonic acid (AMPS® )
Since the structure of ADMA and SDMA are very similar and previous HPLC studies of the underivatized compounds have shown little or no separation, it was necessary to verify the integrity of the reference materials. The subtle differences in the structure of the compounds are potentially distinguishable by NMR spectroscopy. Fig. 1A shows the 400 Mz proton spectrumof the methyl region for the two compounds.
J.J. Pesek et al. / J. Chromatogr. A 1441 (2016) 52–59
There is a clear difference in this part of the spectrum with the peak at 2.812 ppm due to the two methyl groups on SDMA which are either slightly different from each other or undergoing a kinetic effect as a result of the migration of the double bond between the two nitrogens making it broader than the two equivalent methyl groups on ADMA that are at 3.007 ppm. The closer proximity of the double bond to the methyl groups in SDMA accounts for lower chemical shift value with respect to ADMA. Differences are also noted in the methylene portion of the spectrum (Fig. 1B) for the two CH2 groups bonded to the middle methylene group on both molecules. Thus it is clear from the NMR spectra that the two standards to be used for chromatographic testing are different from each other and can be assigned to ADMA and SDMA as shown in the structures above. 3.2. Column evaluation While all columns based on silica hydride have the potential to retain polar compounds, the Diamond HydrideTM (DH) has proved to be the most versatile [20–22]. Thus a variety of gradients using water/acetonitrile with formic acid and acetic acid concentrations ranging from 0.1 to 0.5% were tested in order to determine if was possible to separate underivatized ADMA and SDMA. Only marginal
55
separation was achieved under some conditions. The best results achieved are shown in Fig. 2A. While this separation is somewhat better than obtained for the underivatized compounds using HILIC [19], it is still not close to the baseline resolution that would be ideal. Therefore, a new experimental column also based on a silica hydride particle containing the bonded organic moiety acrylamido2-methylpropane sulfonic acid (AMPS) was selected for further evaluation. The advantage of this particular stationary phase would be that at pH values as low as 2, the practical lower limit for most silica-based stationary phases, the sulfonic acid group would be unprotonated resulting in a negative charge. Thus ionic or ion-exchange interactions should be possible with the analytes providing another mechanism that could improve their separation. Fig. 2B shows the result of an isocratic run on the AMPS column. Under these conditions the two compounds are essentially baseline separated, but the analysis time is relatively long at 28 min. The analysis can be shortened by use of a gradient, but at the loss of some resolution as shown in Fig. 2C. While quantitative analysis would be possible under these conditions, the degree of accuracy and precision could be degraded with respect to the isocratic conditions. One possible future solution would be to fabricate the AMPS stationary phase on smaller particles. The DH stationary phase is currently available on 2 m particles. In order to test this hypothesis, a comparison of efficiencies for ADMA was made on DH columns with 4 m and 2 m particle sizes. The peak width for ADMA on the 2 m column was approximately one-half of the width on the 4 m column under the same experimental conditions (sample volume and concentration were identical). Thus if the analysis shown in Fig. 2C was done with a 2 m AMPS column the two peaks would be essentially baseline separated. 3.3. Analysis using MS fragmentation The use of MS/MS for the analysis of ADMA and SDMA has been the most convenient approach since each compound has a distinct fragment ion thus making their separation less important [18,19]. However, typical tandem mass spectrometry is an expensive approach for this and similar analyses involving isobaric compounds like ADMA and SDMA. An often overlooked alternative to MS/MS is to use in-source collision induced fragmentation of the parent ion that can result in daughter ions similar to those produced in tandem MS [42]. Since ADMA and SDMA have been studied by tandem mass spectrometry, the identity of the most likely fragment ions for each compound is readily available. The unique fragment ions are m/z 46 for ADMA and m/z 172 for SDMA. To produce more fragmentation in the ESI source for a single quad mass analyzer, the capillary exit voltage was increased from +80 V to +140 V while the skimmer voltage was kept constant at +25 V. Fig. 3 shows the extracted ion chromatogram (EIC) for each of the unique fragment ions under the best chromatographic conditions on the AMPS column. Two different samples were run having 1:3 (Fig. 3A) and 3:1 (Fig. 3B) concentration ratios for SDMA/ADMA. In this case the data were obtained using the same gradient as in Fig. 2C but with 0.1% acetic acid in the mobile phase and on a different LCMS system with a single quad mass analyzer (see Section 2.2). It should be noted that this approach cannot replace all analyses requiring MS/MS and has limitations where complex matrices exist as well as lower reliability as the m/z value decreases. 3.4. Quantitation of ADMA/SDMA without separation
Fig. 3. Extracted ion chromatograms of ADMA and SDMA at two different concentration ratios using in-source fragmentation. Conditions same as Fig. 2C but with 0.1% acetic acid in the mobile phase. Fragment ions in the positive mode are m/z 46 for ADMA and m/z 172 for SDMA.
Under some circumstances, useful clinical or physiological information can be obtained from the total combined concentration of ADMA and SDMA. For this type of analysis, a simple calibration
56
J.J. Pesek et al. / J. Chromatogr. A 1441 (2016) 52–59
Fig. 4. Overlaid extracted ion chromatograms of normal subjects and TB patients for agrinine (range m/z 175.00–175.20) and ADMA/SDMA (m/z 203.00–203.20) in urine samples on the DH column. Mobile phase: A; 50:50 DI water/2-propanol + 0.1% formic acid and B; acetonitrile + 0.1% formic acid. Gradient: 0.0 min 90% B, 0–5 min to 30% B, 5–8 min 30% B. Flow rate 0.4 mL/min.
Fig. 5. Analysis of ADMA and SDMA in urine samples utilizing the parent ion (m/z 203) and the in-source fragment ions of m/z 46 for ADMA and m/z 172 for SDMA. Conditions same as Fig. 3.
curve was created for ADMA over the expected concentration range for urine samples on the DH column. The least squares correlation coefficient (R2 ) for this calibration curve is 0.99. The result for a combined ADMA/SDMA analysis in the urine of three healthy individuals was in the range of 0.91 to 0.96 g/mL thus establishing
a level to be used for diagnostic purposes using the same sample preparation procedure. This protocol was tested with a set of urine samples obtained from patients diagnosed with tuberculosis and another set from a control group. Fig. 4 shows typical results obtained when urine samples of healthy individuals are compared to those from TB
J.J. Pesek et al. / J. Chromatogr. A 1441 (2016) 52–59
Fig. 6. In source fragmentation extracted ion chromatograms for glucose-1- and glucose-6-phosphate. A: m/z 241 and B m/z 96.7. Mobile phase: A; DI water + 10 mM ammonium acetate and B; 97:3 acetonitrile/water + 10 mM ammonium acetate. Gradient: 0.0 min 90% B, 0–6 min to 20% B, 6–8 min 20%B. Column: DH packed in PEEK tubing. Flow rate 0.4 mL/min.
patients. There is a dramatic increase in the total ADMA/SDMA concentration for the TB urine samples. The average value in the control group was 0.93 g/mL, consistent with the results reported above, while that in the TB group was 12.8 g/mL. The same figure also shows a comparison for another potential compound to be used for better understanding the disease process, arginine. The green traces represent the range of results obtained for the control sample while the blue trace is a typical result from the TB group. Based on this limited data, it would appear that arginine blood urine levels for TB patients are only about 20% of those found in healthy individuals. The aqueous component of the mobile phase contained 2-propanol as means of preventing column contamination from the urine samples. A sample from another TB patient and a control were investigated using the in-source parent ion fragmentation protocol. The results for these analyses are shown in Fig. 5. For this particular sample the combined ADMA/SDMA concentration is about two times the concentration for the TB sample (Fig. 5A) in comparison to the control (Fig. 5B). There is no significant difference between the relative concentrations of ADMA and SDMA when comparing the TB patient samples to the control group. This result is found for the parent ion m/z 203 as well as analysis of the fragment ions (m/z 46 and 172). In any case, further investigations with a larger sample set are needed to verify and/or refine the results reported here. 3.5. Analysis of other hydrophilic metabolites with in-source fragmentation As a further verification of the utility of in-source fragmentation, two other isobaric compound pairs were tested. Fig. 6 shows the
57
Fig. 7. In source fragment ion chromatograms for leucine and isoleucine. A: leucine with fragment ion m/z 30; and B isolueucine with fragment ions m/z 30 and m/z 68. Mobile phase: A; DI water + 0.1% acetic acid and B; acetonitrile + 0.1% acetic acid. Gradient same as Fig. 2C. AMPS column. Flow rate 0.4 mL/min.
result of using in-source fragmentation for the chromatographic analysis of glucose-1-phosphate and glucose-6-phosphate. The stationary phase is the same DH material used in the ADMA/SDMA studies but packed in PEEK tubing to control the peak distortions that often occur with phosphate compounds when eluted through stainless steel columns and frits. Under the conditions used in this experiment, there is minimal separation of the two isobaric compounds. However, each of the two compounds has a unique fragmentation ion that can be potentially used for both identification and quantitation. Fig. 6A shows the extracted fragment ion chromatogram for m/z 241 which is unique for glucose-1phosphate (black trace) as can be seen by the lack of any signal when glucose-6-phosphate (red trace) is injected with detection at the same mass-to-charge ratio. Fig. 6B shows an identical experiment with detection at m/z 96.7. In this case a peak for glucose-6phosphate (red trace) is readily detected and there is no signal for glucose-1- phosphate (black trace). Thus each compound can be readily identified and quantified through their unique fragment ions even though there is little or no chromatographic separation and without the use of an MS/MS instrument. Another possible utilization of in-source fragmentation occurs when only one of the non- or partially separated pairs of isobaric compounds has a unique fragment. This situation is illustrated by the example of leucine and isoleucine shown in Fig. 7. One distinguishable fragment at m/z 30 is obtained for leucine (Fig. 7A). However, isoleucine has two readily identified fragments by insource fragmentation, m/z 68 and m/z 30 (Fig. 7B). Thus a calibration curve for isoleucine can be created based on the unique fragment at m/z 68. A calibration curve for both leucine and isoleucine at m/z 30 can be prepared separately using standard solutions of the two
58
J.J. Pesek et al. / J. Chromatogr. A 1441 (2016) 52–59
compounds. For a mixture, the concentration of iosleucine is determined from the m/z 68 peak. Once this is determined, the signal corresponding to isoleucine can be subtracted from the total signal at m/z 30, with the remainder due to leucine. From this difference, then the concentration of leucine can be determined.
[13]
[14]
4. Conclusions
[15]
Aqueous normal phase chromatography was shown to be an excellent alternative for the analysis of the two clinically important compounds ADMA and SDMA. No derivatization or mobile phase additives that are incompatible with mass spectrometery are needed. Partial separation of the two compounds was achieved with the DH stationary phase and baseline resolution was obtained on the experimental AMPS column. The combined ADMA/SDMA levels as well as those for arginine in a limited number of TB patients were measured and it was determined that the former were considerably higher and the latter considerably lower than those of normal subjects. It was also demonstrated that ADMA and SDMA could be analyzed individually without complete separation through the use of “in-source fragmentation” where each of the compounds has a unique daughter ion. This protocol could potentially be used for a number of analyses thus avoiding the need for more expensive MS/MS instrumentation. Finally, the in-source approach was shown to be viable for two other pairs of isobaric compounds, glucose-1- and glucose-6-phosphate, and leucine and isoleucine.
[16]
Acknowledgments The authors would like to thank Microsolv Technology Corp. for donation of the HPLC columns used in this study and Dr. Kyu Rhee of Weill Cornell Medical Center for the urine samples.
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
References [26] [1] D. Fliser, Asymmetric dimethylarginine (ADMA): the silent transition froman ‘uraemic toxin’ to a global cardiovascular risk molecule, Eur. J. Clin. Invest. 35 (2005) 71–79. [2] R.H. Böger, Asymmetric dimethylarginine (ADMA) and cardiovascular disease: insights from prospective clinical trials, Vasc. Med. 10 (2005) 19–25. [3] C. Fleck, A. Janz, F. Schweitzer, E. Karge, M. Schwertfeger, G. Stein, Serum concentrations of asymmetric (ADMA) and symmetric (SDMA) dimethyl arginine in renal failure patients, Kidney Int. Suppl. 78 (2001) S14–S18 (1974–2011). [4] A. Sasaki, T. Mabuchi, K. Serizawa, I. Takasaki, T. Andoh, K. Shiraki, S. Ito, Y. Kuraishi, Different roles of nitric oxide synthase-1 and -2 between herpetic and postherpetic allodynia in mice, Neuroscience 150 (2007) 459–466. [5] S. Moncada, E.A. Higgs, The discovery of nitric oxide and its role in vascular biology, Br. J. Pharmacol. 147 (2006) S193–201. [6] J.T. Kielstein, R. Boger, S.M. Bode-Böger, J. Schaffer, M. Barbey, K.M. Koch, J.C. Frolich, Asymmetric dimethylargine plasma concentrations differ in patients with end-stage renal disease: relationship to treatment methods and atherosclerotic disease, J. Am. Soc. Nephrol. 10 (1999) 594–600. [7] H. Worthmann, S. Chen, J. Martens-Lobenhoffer, N. Li, M. Deb, A.B. Tryc, A. Goldbecker, Q. Dong, J.T. Kielstein, J. Thomas, S.M. Bode-Böger, K. Weissenborn, High plasma dimethylarginine levels are associated with adverse clinical outcome after stroke, J. Atheroscler. Thromb. 18 (2011) 753–761. [8] R.J. Nijveldt, T. Teerlink, P.A. van Leeuwen, The asymmetric dimethylarginine (ADMA)-multiple organ failure hypothesis, Clin. Nutr. 22 (2003) 99–104. [9] R.J. Nijveldt, T. Teerlink, R. Van Der Hoven, M.P. Siroen, D.J. Kuik, A. Rauwerda, P.A. van Leeuwen, Asymmetrical dimethylarginine (ADMA) in critically ill patients: high plasma ADMA concentration is an independent risk factor of ICU mortality, Clin. Nutr. 22 (2003) 23–30. [10] M.J. O’Dwyer, F. Dempsey, V. Crowley, D.P. Kelleher, R. McManus, R. Ryan, Septic shock is correlated with asymmetrical dimethyl arginine levels, which may be influenced by a polymorphism in the dimethylarginine dimethylaminohydrolase II gener: a prospespective observational study, Crit. Care 10 (2006) R139. [11] T. Teerlink, HPLC analysis of ADMA and other methylated l-arginine analogs in biological fluids, J. Chromatogr. B 851 (2007) 21–29. [12] Y. Kakimoto, S. Akazawa, Isolation and identification of NG ,NG - and NG ,N G -dimethylarginine, N -mono, di, and trimethyllysine, and
[27]
[28]
[29]
[30]
[31]
[32]
[33] [34]
[35]
[36]
[37]
glucosylgalactosyl- and galactosyl-␦-hydroxylysine from human urine, J. Biol. Chem. 245 (1970) 5751–5758. R.J. McAllister, M.H. Rambausek, P. Vallance, D. Williams, K.-H. Hoffman, E. Ritz, Concentration of dimethyl-l-arginine in the plasma of patients with end-stage renal failure, Nephrol. Dial. Transplant 11 (1996) 2449–2452. M. Ivanova, C. Artusi, G.M. Boffa, M. Zaninotto, M. Plebani, HPLC determination of plasma dimethylarginines: method validation and preliminary clinical application, Clin. Chim. Acta 411 (2010) 1632–1636. C.E. Jones, C.J. Darcy, T. Woodberry, N.M. Anstey, Y.R. McNeil, HPLC analysis of asymmetric dimethylarginine in small plasma volumes using a Gemini-NC column at high pH, J. Chromatogr. B 878 (2010) 8–12. M. Davids, E. Swieringa, F. Palm, D.E.C. Smith, Y.M. Smulders, P.G. Scheffer, H.J. Blom, T. Teerlink, Simultaneous determination of asymmetric and symmetric dimethylarginine, l-monomethylarginine, l-arginine, and l-homoarginine in biological samples using stable isotope dilution liquid chromatography tandem mass spectrometry, J. Chromatogr. B 900 (2012) 38–47. S. Blackwell, D. St. J. O’Reilly, D.K. Talwar, HPLC analysis of asymmetric dimethylarginine (ADMA) and related arginine metabolites in human plasma using a novel non-endogenous internal standard, Clin. Chim. Acta 401 (2009) 14–19. I.M. Di Gangi, L. Chiandetti, A. Gucciardi, V. Moret, M. Naturale, G. Giordano, Simultaneous quantitative determination of NG,NG-dimethyl-l-arginine or asymmetric dimethylarginine and related pathway’s metabolites in biological fluids by ultrahigh-performance liquid chromatography/electrospray ionization-tandem mass spectrometry, Anal Chim. Acta 677 (2010) 140–148. J. Martens-Lobenhoffer, S.M. Bode-Böger, Quantification of l-arginine, asymmetric dimethylarginine and symmetric dimethylarginine in human plasma: a step improvement in precision by stable isotope dilution mass spectrometry, J. Chromatogr. B 904 (2012) 140–143. J.J. Pesek, M.T. Matyska, Hydrophilic Interaction Chromatography (HILIC) and Advanced Applications, in: P.G. Wang, W. He (Eds.), CRC Press, Boca Raton, 2011, pp. 1–26. J.J. Pesek, M.T. Matyska, R.I. Boysen, Y. Yang, M.T.W. Hearn, Aqueous normal phase chromatography using silica hydride-based stationary phases, Trends Anal. Chem. 42 (2013) 64–73. J.J. Pesek, R.I. Boysen, M.T.W. Hearn, M.T. Matyska, Hydride-based stationary phases: A rapidly evolving technology for the development of new bio-analytical methods, Anal. Methods 6 (2014) 4496–4503. J. Soukup, P. Janas, P. Jandera, Gradient elution in aqueous normal-phase liquid chromatography on hydrosilated silica-based stationary phases, J. Chromatogr. A 1286 (2013) 111–118. S. Jenkins, S.M. Fischer, L. Chen, T.R. Sana, Global LC/MS metabolomics profiling of calcium stressed and immunosuppressant drug treated Saccharomyces cerevisiae, Metabolites 3 (2013) 1102–1117. H. Eoh, K.Y. Rhee, Multifunctional essentiality of succinate metabolism in adaptation to hypoxia in Mycobacterium tuberculosis, PNAS 110 (2013) 6554–6559. N. Putluria, A. Shojaieg, V.T. Vasub, et al., Metabolomic profiling reveals potential markers and bioprocesses altered in bladder cancer progression, Cancer Res. 71 (2011) 7376–7386. K.R. Chalcraft, B.E. McCarry, Tandem LC columns for the simultaneous retention of polar and nonpolar molecules in comprehensive metabolomics analysis, J. Sep. Sci. 36 (2013) 3478–3485. J.J. Pesek, M.T. Matyska, J.A. Loo, S.M. Fischer, T.R. Sana, Analysis of hydrophilic metabolites in physiological fluids by HPLC-MS using a silica hydride-based stationary phase, J. Sep. Sci. 32 (2009) 2200–2208. J.J. Pesek, M.T. Matyska, M.T.W. Hearn, R.I. Boysen, Aqueous normal-phase retention of nucleotides on silica hydride columns, J. Chromatogr. A 1216 (2009) 1140–1146. J.J. Pesek, M.T. Matyska, J. Duley, M. Zamzami, S.M. Fischer, Aqueous normal phase (ANP) retention of nucleotides on silica hydride-based columns. method development strategies for analytes relevant in clinical analysis, J. Sep. Sci. 33 (2010) 930–938. J. Ge, F. Liu, E.H. Holmes, G.K. Ostrander, Q.X. Li, Aqueous normal phase liquid chromatography coupled with tandem time-of-flight quadrupole mass spectrometry for determination of zanamivir in human serum, J. Chromatogr. B 906 (2012) 58–62. J.J. Pesek, M.T. Matyska, A. Dang, Analysis of cycloserine and related compounds using aqueous normal phase chromatography/mass spectrometry, J. Pharm. Biomed. Anal. 64-65 (2012) 72–76. J.J. Pesek, M.T. Matyska, A. Kim, Evaluation of silica hydride-based stationary phases for the analysis of drugs of abuse, J. Sep. Sci. 36 (2013) 2760–2766. J.J. Pesek, M.T. Matyska, A. Dang, Analysis of ethyl glucuronide and ethyl sulfate using aqueous normal phase chromatography/mass spectrometry, J. Sep. Sci. 38 (2015) 1515–1520. J.E. Young, M.T. Matyska, J.J. Pesek, Liquid chromatography/mass spectrometry compatible approaches for the quantitation of folic acid in fortified juices and cereals using aqueous normal phase conditions, J. Chromatogr. A 1218 (2011) 2121–2126. A. Dang, M.T. Matyska, J.J. Pesek, The use of aqueous normal phase chromatography as an analytical tool for food analysis. Determination of histamine as a model system, Food Chem. 141 (2013) 4226–4230. R. Le, J.E. Young, J.J. Pesek, M.T. Matyska, Separation of 1,3-dimethylamylamine and other polar compounds in dietary supplement formulation using aqueous normal phase chromatography with mass spectrometry, J. Sep. Sci. 36 (2013) 2578–2583.
J.J. Pesek et al. / J. Chromatogr. A 1441 (2016) 52–59 [38] Y. Nolvachai, C. Kulsing, R.I. Boysen, M.T. Matyska, J.J. Pesek, P.J. Marriott, M.T.W. Hearn, Comparison of the performance of different silica hydride particles for the solid-phase extraction of non-volatile analytes from dark chocolate with analysis by gas chromatography-quadrupole mass spectrometry, Food Chem. 174 (2015) 434–439. [39] J.E. Young, J.J. Pesek, M.T. Matyska, LC–MS compatible approaches for quantitation of limonin in citrus juice, LCGC 33 (2015) 192–199. [40] R. MacNeill, R. Stromeyer, B. Urbanowicz, V. Acharya, M. Moussaille, J.J. Pesek, Silica hydride-based chromatography of LC–MS response altering compounds native to human plasma, Bioanalysis 4 (2012) 2877–2886.
59
[41] S. Rocchi, A. Rocco, J.J. Pesek, M.T. Matyska, D. Capitani, S. Fanali, Enantiomers separation by nano-liquid chromatography: Use of a novel-sub-2 m vancomycin silica hydride stationary phase, J. Chromatogr. A 1381 (2015) 149–159. [42] L. Abranko, J.F. Garcia-Reyes, A. Molina-Diaz, In-source fragmentation and accurate mass analysis of multiclass flavonoid conjugates by electrospray ionization time-of-flight mass spectrometry, J. Mass Spec. 46 (2011) 478–488.
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
The separation and analysis of symmetric and asymmetric dimethylarginine and other hydrophilic isobaric compounds using aqueous normal phase chromatography Joseph J. Pesek ∗ , Maria T. Matyksa, Brent Modereger, Alejandra Hasbun, Vy T. Phan, Zahra Mehr, Mariano Guzman, Seiichiro Watanable Department of Chemistry, San Jose State University, San Jose, CA 95192, USA
a r t i c l e
i n f o
Article history: Received 26 January 2016 Received in revised form 19 February 2016 Accepted 22 February 2016 Available online 27 February 2016 Keywords: In-source fragmentation Silica hydride columns Clinical analyses
a b s t r a c t Two biologically important compounds with clinical relevance, asymmetric dimethylarginine and symmetric dimethylarginine, are analyzed using aqueous normal phase chromatography on silica hydride-based columns. Two different stationary phases were tested, a commercially available Diamond HydrideTM and a 2-acrylamido-2-methylpropane sulfonic acid experimental column. Two types of analytical protocols were investigated: analysis of the compounds when separation was achieved and analysis of the compounds with partial chromatographic separation. Urine samples from tuberculosis patients were tested for levels of asymmetric and symmetric dimethylarginine. The mass spectrometric technique of in-source fragmentation that can provide data similar to a tandem mass analyzer was evaluated as a means of identification and quantitation of the two compounds when complete separation is not achieved. This same protocol was also evaluated for two other isobaric compounds, glucose-1 and glucose-6 phohsphate, and leucine and isoleucine. © 2016 Elsevier B.V. All rights reserved.
1. Introduction Both asymmetric dimethylarginine (ADMA) and symmetric dimethylarginine (SDMA) plasma levels have been associated with various forms of renal and cardiac physiological conditions [1,2]. ADMA has been shown to be a potent inhibitor of the enzyme nitric oxide synthase (NOS) which regulates the conversion of arginine into the vasodilator nitric oxide [3–5]. ADMA levels in biological fluids have been shown to correlate closely with renal functions and could be used as a biomarker for certain diseases [6]. Also, it has been shown that above normal amounts of ADMA and SDMA in blood plasma can be used to predict the extent of detrimental effects from stroke [7]. Monitoring of ADMA has also been found to be of importance in acute critical conditions such as septic shock [8–10]. The extent of the clinical applications of ADMA and SDMA as well as other small polar physiological analytes makes the development of new approaches for their determination a topic of significant importance.
∗ Corresponding author. Fax: +1 408 924 4945. E-mail address: [email protected] (J.J. Pesek). http://dx.doi.org/10.1016/j.chroma.2016.02.071 0021-9673/© 2016 Elsevier B.V. All rights reserved.
HPLC based methods have been the most successful to date for the analysis of ADMA and SDMA [11]. A number of approaches have been proposed in developing analytical schemes for these two compounds. Because of their polarity, in many cases the compounds are derivatized in order to be retained in reversedphase and in some cases to also enhance detection. A very early method for the preparation of the two compounds utilized an ion-exchange column that required almost five hours for the separation to be completed [12]. Another early protocol described the analysis of underivatized ADMA/SDMA in plasma from renal failure patients that required considerable sample preparation and a 20 min analysis time using UV detection at 200 nm [13]. Typical examples of pre-column derivatization are methods for the determination of ADMA and SDMA in plasma samples that allow for reversed-phase retention and high sensitivity by fluorescence detection [14–17]. Another reversed-phase method with derivatization utilized MS/MS for detection and quantitation with the each compound having a daughter ion that was unique [18]. One other approach reported utilized hydrophilic liquid interaction chromatography (HILIC) to analyze underivatized ADMA and SDMA with MS/MS detection [19]. Some sensitivity that would have been gained at the high acetonitrile concentrations in the HILIC method may have been at least partially lost because the mobile phase con-
J.J. Pesek et al. / J. Chromatogr. A 1441 (2016) 52–59
53
Fig. 1. Overlayed 400 Mz proton magnetic resonance spectra of ADMA and SDMA standards used for chromatographic evaluation. (A) methyl region and (B) methylene region.
tained 0.025% trifluoroacetic acid, a known suppressor of MS signal intensity. Aqueous normal phase (ANP) chromatography utilizing silica hydride-based stationary phases is a versatile technique that can be applied to a broad range of analytical separations [20–22]. It versatility is based on the fact that all columns packed with silica hydride phases can operate in both the reversed-phase and normal-phase modes. There are a number of crucial differences between silica hydride and ordinary silica that lead to chromatographic advantages for the hydride material. These features include the dual retention capabilities cited above. In addition, the surface of silica hydride is slightly hydrophobic rather than highly polar due to silanols (as in HILIC phases) and thus often result in better peak shape for certain analytes, particularly bases. Other notable features of silica hydride include a very thin water shell
(∼0.5 monolayer) on the surface [23], rapid equilibration of the surface solvation layer after gradients and highly reproducible retention for polar compounds in the aqueous normal-phase mode. These properties have been used advantageously for the analysis of metabolites [24–28], nucleotides [29,30], pharmaceuticals [31–34], food and beverage components [35–39], lipophilic interferences [40] and chiral compounds [41]. Considering the proven capabilities of the silica hydride-based stationary phase for the analysis of hydrophilic compounds, the objective of this study was to determine if underivatized ADMA and SDMA could be better separated in a reasonable analysis time than in previous studies thus eliminating the need for extensive sample preparation or additional on-line procedures. In addition, the use of a newer and simpler mass spectrometric detection method, in-source fragmentation, for the detection of unique fragment ions
54
J.J. Pesek et al. / J. Chromatogr. A 1441 (2016) 52–59
on 4 m particle size silica hydride in a 150 mm × 2.1 mm column (MicroSolv Technology Corp.). Some preliminary screening experiments were also performed using a 50 mm × 2.1 mm AMPS column. ADMA was purchased from Sigma–Aldrich Co. LLC. (Sigma–Aldrich, St. Louis, MO, USA) and SDMA was purchased from Calbiochem (Billerica, MA, USA). Acetonitrile was obtained from GFS Chemicals (Columbus, OH, USA). Water was obtained from a Milli-Q water purification system (Millipore Corp., Billerica, MA, USA), and filtered with a 45 m Nylon filter (MicroSolv Technology Corp.). Acetonitrile, formic acid and glacial acetic acid were purchased from Sigma–Aldrich (LC-MS grade). Tuberculosis urine samples were obtained from Cornell-Weill Medical Center in New York, NY.
2.2. Instrumentation NMR spectra were acquired on a Varian (Palo Alto, CA, USA) INOVA 400 spectrometer. One HPLC system used in this study was an Agilent (Little Falls, DE, USA) 1100 Series LC system, including degasser, binary pump, temperature-controlled autosampler and temperature-controlled column compartment. The mass spectrometer system was an Agilent (Santa Clara, CA, USA) Model 6210 MSD TOF with a dual sprayer electrospray ionization (ESI) source. Additional LCMS studies were done using a Flexar (Perkin-Elmer, Waltham, MA, USA) UHPLC system, and a Flexar SQ 300 MS.
2.3. Methods
Fig. 2. Chromatographic elution of ADMA and SDMA under ANP conditions. (A) DH column: mobile phase; A, DI water + 0.1% acetic acid and B, acetonitrile + 0.1% acetic acid. Gradient; 0–1 min 80% B, 1–3 min to 70% B, 3–4 min 70% B, 4–6 min to 60% B, 6–7 min 60% B, 7–9 min to 50% B, 9–10 min 50% B, 10–12 min to 40% B, 12–13 min 40% B, 13–15 min to 30% B, 15–16 min 30% B, 16–18 to 10% B, 18–23 min 10%B. (B) AMPS column: mobile phase; A, DI water + 0.1% formic acid and B, acetonitrile + 0.1% formic acid. Isocratic at 70% B. (C) AMPS column: mobile phase; A, DI water + 0.2% acetic acid and B, acetonitrile + 0.2% acetic acid. Gradient; 0–1 min 95% B, 1–3 min to 70% B, 3–4 min 70% B, 4–5 min to 50% B, 5–6 min 50% B, 6–7 min to 20% B, 7–14 min 20% B. Detection at [M + H]+ 203.1503. Flow rate = 0.4 mL/min.
was explored. The objective was to determine if this approach has general applicability to the analysis of isobaric compounds that are partially separated chromatographically. These techniques were then applied to the analysis of real clinical samples. This format could be used to develop validated methods for these and other hydrophilic analytes including isobaric compounds.
Stock solutions of ADMA and SDMA were prepared at a concentration of 1 mg/mL in DI water. Sample solutions were prepared by diluting stock solutions in 50:50 acetonitrile/water. ADMA and SDMA samples were prepared by diluting the stock solution 1:19. Urine samples of healthy individuals were prepared by the Dept. of Nutrition, Food Science and Packaging at San Jose State University. Urine samples of tuberculosis patients and controls were prepared by Division of Infectious Diseases, Department of Medicine, Weill Cornell Medical College, New York NY. The mixed sample was prepared by combining equal volumes of stock solution, and then further diluting 1:10 with 50:50 DI water/ACN. The mobile phase organic solvent was composed of 0.1%, 0.2%, or 0.5% acetic acid in acetonitrile. Water containing 0.1%, 0.2%, or 0.5% acetic acid made up the difference, pairing solvents of equal acetic acid composition. The column flow rate was 0.4 mL/min. The column temperature was kept at 25 ◦ C. Various gradients were used in testing the columns and those with significance are reported in the text or figure captions. The gradient flow rate was 0.4 mL/min. MS detection was operated in the positive ion mode, monitoring at 203.1503 m/z (for exact mass detection) or 203.0000–203.2000 m/z. Fragmentation studies were monitored at the parent ion 203.1503 m/z as well as 46.0000 m/z (ADMA), and 172.0000 m/z (SDMA).
3. 3. Results and Discussion 2. Experimental 3.1. NMR sample authentication 2.1. Materials Two silica hydride stationary phases were used in this study. One phase was a Diamond HydrideTM material (4 m particle size, 150 mm × 2.1 mm in stainless steel tubing and 100 mm × 2.1 mm in PEEK tubing) obtained from Microsolv Technology (Eatontown, NJ, USA). The second phase was an experimental material with functionalized 2-acrylamido-2-methylpropane sulfonic acid (AMPS® )
Since the structure of ADMA and SDMA are very similar and previous HPLC studies of the underivatized compounds have shown little or no separation, it was necessary to verify the integrity of the reference materials. The subtle differences in the structure of the compounds are potentially distinguishable by NMR spectroscopy. Fig. 1A shows the 400 Mz proton spectrumof the methyl region for the two compounds.
J.J. Pesek et al. / J. Chromatogr. A 1441 (2016) 52–59
There is a clear difference in this part of the spectrum with the peak at 2.812 ppm due to the two methyl groups on SDMA which are either slightly different from each other or undergoing a kinetic effect as a result of the migration of the double bond between the two nitrogens making it broader than the two equivalent methyl groups on ADMA that are at 3.007 ppm. The closer proximity of the double bond to the methyl groups in SDMA accounts for lower chemical shift value with respect to ADMA. Differences are also noted in the methylene portion of the spectrum (Fig. 1B) for the two CH2 groups bonded to the middle methylene group on both molecules. Thus it is clear from the NMR spectra that the two standards to be used for chromatographic testing are different from each other and can be assigned to ADMA and SDMA as shown in the structures above. 3.2. Column evaluation While all columns based on silica hydride have the potential to retain polar compounds, the Diamond HydrideTM (DH) has proved to be the most versatile [20–22]. Thus a variety of gradients using water/acetonitrile with formic acid and acetic acid concentrations ranging from 0.1 to 0.5% were tested in order to determine if was possible to separate underivatized ADMA and SDMA. Only marginal
55
separation was achieved under some conditions. The best results achieved are shown in Fig. 2A. While this separation is somewhat better than obtained for the underivatized compounds using HILIC [19], it is still not close to the baseline resolution that would be ideal. Therefore, a new experimental column also based on a silica hydride particle containing the bonded organic moiety acrylamido2-methylpropane sulfonic acid (AMPS) was selected for further evaluation. The advantage of this particular stationary phase would be that at pH values as low as 2, the practical lower limit for most silica-based stationary phases, the sulfonic acid group would be unprotonated resulting in a negative charge. Thus ionic or ion-exchange interactions should be possible with the analytes providing another mechanism that could improve their separation. Fig. 2B shows the result of an isocratic run on the AMPS column. Under these conditions the two compounds are essentially baseline separated, but the analysis time is relatively long at 28 min. The analysis can be shortened by use of a gradient, but at the loss of some resolution as shown in Fig. 2C. While quantitative analysis would be possible under these conditions, the degree of accuracy and precision could be degraded with respect to the isocratic conditions. One possible future solution would be to fabricate the AMPS stationary phase on smaller particles. The DH stationary phase is currently available on 2 m particles. In order to test this hypothesis, a comparison of efficiencies for ADMA was made on DH columns with 4 m and 2 m particle sizes. The peak width for ADMA on the 2 m column was approximately one-half of the width on the 4 m column under the same experimental conditions (sample volume and concentration were identical). Thus if the analysis shown in Fig. 2C was done with a 2 m AMPS column the two peaks would be essentially baseline separated. 3.3. Analysis using MS fragmentation The use of MS/MS for the analysis of ADMA and SDMA has been the most convenient approach since each compound has a distinct fragment ion thus making their separation less important [18,19]. However, typical tandem mass spectrometry is an expensive approach for this and similar analyses involving isobaric compounds like ADMA and SDMA. An often overlooked alternative to MS/MS is to use in-source collision induced fragmentation of the parent ion that can result in daughter ions similar to those produced in tandem MS [42]. Since ADMA and SDMA have been studied by tandem mass spectrometry, the identity of the most likely fragment ions for each compound is readily available. The unique fragment ions are m/z 46 for ADMA and m/z 172 for SDMA. To produce more fragmentation in the ESI source for a single quad mass analyzer, the capillary exit voltage was increased from +80 V to +140 V while the skimmer voltage was kept constant at +25 V. Fig. 3 shows the extracted ion chromatogram (EIC) for each of the unique fragment ions under the best chromatographic conditions on the AMPS column. Two different samples were run having 1:3 (Fig. 3A) and 3:1 (Fig. 3B) concentration ratios for SDMA/ADMA. In this case the data were obtained using the same gradient as in Fig. 2C but with 0.1% acetic acid in the mobile phase and on a different LCMS system with a single quad mass analyzer (see Section 2.2). It should be noted that this approach cannot replace all analyses requiring MS/MS and has limitations where complex matrices exist as well as lower reliability as the m/z value decreases. 3.4. Quantitation of ADMA/SDMA without separation
Fig. 3. Extracted ion chromatograms of ADMA and SDMA at two different concentration ratios using in-source fragmentation. Conditions same as Fig. 2C but with 0.1% acetic acid in the mobile phase. Fragment ions in the positive mode are m/z 46 for ADMA and m/z 172 for SDMA.
Under some circumstances, useful clinical or physiological information can be obtained from the total combined concentration of ADMA and SDMA. For this type of analysis, a simple calibration
56
J.J. Pesek et al. / J. Chromatogr. A 1441 (2016) 52–59
Fig. 4. Overlaid extracted ion chromatograms of normal subjects and TB patients for agrinine (range m/z 175.00–175.20) and ADMA/SDMA (m/z 203.00–203.20) in urine samples on the DH column. Mobile phase: A; 50:50 DI water/2-propanol + 0.1% formic acid and B; acetonitrile + 0.1% formic acid. Gradient: 0.0 min 90% B, 0–5 min to 30% B, 5–8 min 30% B. Flow rate 0.4 mL/min.
Fig. 5. Analysis of ADMA and SDMA in urine samples utilizing the parent ion (m/z 203) and the in-source fragment ions of m/z 46 for ADMA and m/z 172 for SDMA. Conditions same as Fig. 3.
curve was created for ADMA over the expected concentration range for urine samples on the DH column. The least squares correlation coefficient (R2 ) for this calibration curve is 0.99. The result for a combined ADMA/SDMA analysis in the urine of three healthy individuals was in the range of 0.91 to 0.96 g/mL thus establishing
a level to be used for diagnostic purposes using the same sample preparation procedure. This protocol was tested with a set of urine samples obtained from patients diagnosed with tuberculosis and another set from a control group. Fig. 4 shows typical results obtained when urine samples of healthy individuals are compared to those from TB
J.J. Pesek et al. / J. Chromatogr. A 1441 (2016) 52–59
Fig. 6. In source fragmentation extracted ion chromatograms for glucose-1- and glucose-6-phosphate. A: m/z 241 and B m/z 96.7. Mobile phase: A; DI water + 10 mM ammonium acetate and B; 97:3 acetonitrile/water + 10 mM ammonium acetate. Gradient: 0.0 min 90% B, 0–6 min to 20% B, 6–8 min 20%B. Column: DH packed in PEEK tubing. Flow rate 0.4 mL/min.
patients. There is a dramatic increase in the total ADMA/SDMA concentration for the TB urine samples. The average value in the control group was 0.93 g/mL, consistent with the results reported above, while that in the TB group was 12.8 g/mL. The same figure also shows a comparison for another potential compound to be used for better understanding the disease process, arginine. The green traces represent the range of results obtained for the control sample while the blue trace is a typical result from the TB group. Based on this limited data, it would appear that arginine blood urine levels for TB patients are only about 20% of those found in healthy individuals. The aqueous component of the mobile phase contained 2-propanol as means of preventing column contamination from the urine samples. A sample from another TB patient and a control were investigated using the in-source parent ion fragmentation protocol. The results for these analyses are shown in Fig. 5. For this particular sample the combined ADMA/SDMA concentration is about two times the concentration for the TB sample (Fig. 5A) in comparison to the control (Fig. 5B). There is no significant difference between the relative concentrations of ADMA and SDMA when comparing the TB patient samples to the control group. This result is found for the parent ion m/z 203 as well as analysis of the fragment ions (m/z 46 and 172). In any case, further investigations with a larger sample set are needed to verify and/or refine the results reported here. 3.5. Analysis of other hydrophilic metabolites with in-source fragmentation As a further verification of the utility of in-source fragmentation, two other isobaric compound pairs were tested. Fig. 6 shows the
57
Fig. 7. In source fragment ion chromatograms for leucine and isoleucine. A: leucine with fragment ion m/z 30; and B isolueucine with fragment ions m/z 30 and m/z 68. Mobile phase: A; DI water + 0.1% acetic acid and B; acetonitrile + 0.1% acetic acid. Gradient same as Fig. 2C. AMPS column. Flow rate 0.4 mL/min.
result of using in-source fragmentation for the chromatographic analysis of glucose-1-phosphate and glucose-6-phosphate. The stationary phase is the same DH material used in the ADMA/SDMA studies but packed in PEEK tubing to control the peak distortions that often occur with phosphate compounds when eluted through stainless steel columns and frits. Under the conditions used in this experiment, there is minimal separation of the two isobaric compounds. However, each of the two compounds has a unique fragmentation ion that can be potentially used for both identification and quantitation. Fig. 6A shows the extracted fragment ion chromatogram for m/z 241 which is unique for glucose-1phosphate (black trace) as can be seen by the lack of any signal when glucose-6-phosphate (red trace) is injected with detection at the same mass-to-charge ratio. Fig. 6B shows an identical experiment with detection at m/z 96.7. In this case a peak for glucose-6phosphate (red trace) is readily detected and there is no signal for glucose-1- phosphate (black trace). Thus each compound can be readily identified and quantified through their unique fragment ions even though there is little or no chromatographic separation and without the use of an MS/MS instrument. Another possible utilization of in-source fragmentation occurs when only one of the non- or partially separated pairs of isobaric compounds has a unique fragment. This situation is illustrated by the example of leucine and isoleucine shown in Fig. 7. One distinguishable fragment at m/z 30 is obtained for leucine (Fig. 7A). However, isoleucine has two readily identified fragments by insource fragmentation, m/z 68 and m/z 30 (Fig. 7B). Thus a calibration curve for isoleucine can be created based on the unique fragment at m/z 68. A calibration curve for both leucine and isoleucine at m/z 30 can be prepared separately using standard solutions of the two
58
J.J. Pesek et al. / J. Chromatogr. A 1441 (2016) 52–59
compounds. For a mixture, the concentration of iosleucine is determined from the m/z 68 peak. Once this is determined, the signal corresponding to isoleucine can be subtracted from the total signal at m/z 30, with the remainder due to leucine. From this difference, then the concentration of leucine can be determined.
[13]
[14]
4. Conclusions
[15]
Aqueous normal phase chromatography was shown to be an excellent alternative for the analysis of the two clinically important compounds ADMA and SDMA. No derivatization or mobile phase additives that are incompatible with mass spectrometery are needed. Partial separation of the two compounds was achieved with the DH stationary phase and baseline resolution was obtained on the experimental AMPS column. The combined ADMA/SDMA levels as well as those for arginine in a limited number of TB patients were measured and it was determined that the former were considerably higher and the latter considerably lower than those of normal subjects. It was also demonstrated that ADMA and SDMA could be analyzed individually without complete separation through the use of “in-source fragmentation” where each of the compounds has a unique daughter ion. This protocol could potentially be used for a number of analyses thus avoiding the need for more expensive MS/MS instrumentation. Finally, the in-source approach was shown to be viable for two other pairs of isobaric compounds, glucose-1- and glucose-6-phosphate, and leucine and isoleucine.
[16]
Acknowledgments The authors would like to thank Microsolv Technology Corp. for donation of the HPLC columns used in this study and Dr. Kyu Rhee of Weill Cornell Medical Center for the urine samples.
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
References [26] [1] D. Fliser, Asymmetric dimethylarginine (ADMA): the silent transition froman ‘uraemic toxin’ to a global cardiovascular risk molecule, Eur. J. Clin. Invest. 35 (2005) 71–79. [2] R.H. Böger, Asymmetric dimethylarginine (ADMA) and cardiovascular disease: insights from prospective clinical trials, Vasc. Med. 10 (2005) 19–25. [3] C. Fleck, A. Janz, F. Schweitzer, E. Karge, M. Schwertfeger, G. Stein, Serum concentrations of asymmetric (ADMA) and symmetric (SDMA) dimethyl arginine in renal failure patients, Kidney Int. Suppl. 78 (2001) S14–S18 (1974–2011). [4] A. Sasaki, T. Mabuchi, K. Serizawa, I. Takasaki, T. Andoh, K. Shiraki, S. Ito, Y. Kuraishi, Different roles of nitric oxide synthase-1 and -2 between herpetic and postherpetic allodynia in mice, Neuroscience 150 (2007) 459–466. [5] S. Moncada, E.A. Higgs, The discovery of nitric oxide and its role in vascular biology, Br. J. Pharmacol. 147 (2006) S193–201. [6] J.T. Kielstein, R. Boger, S.M. Bode-Böger, J. Schaffer, M. Barbey, K.M. Koch, J.C. Frolich, Asymmetric dimethylargine plasma concentrations differ in patients with end-stage renal disease: relationship to treatment methods and atherosclerotic disease, J. Am. Soc. Nephrol. 10 (1999) 594–600. [7] H. Worthmann, S. Chen, J. Martens-Lobenhoffer, N. Li, M. Deb, A.B. Tryc, A. Goldbecker, Q. Dong, J.T. Kielstein, J. Thomas, S.M. Bode-Böger, K. Weissenborn, High plasma dimethylarginine levels are associated with adverse clinical outcome after stroke, J. Atheroscler. Thromb. 18 (2011) 753–761. [8] R.J. Nijveldt, T. Teerlink, P.A. van Leeuwen, The asymmetric dimethylarginine (ADMA)-multiple organ failure hypothesis, Clin. Nutr. 22 (2003) 99–104. [9] R.J. Nijveldt, T. Teerlink, R. Van Der Hoven, M.P. Siroen, D.J. Kuik, A. Rauwerda, P.A. van Leeuwen, Asymmetrical dimethylarginine (ADMA) in critically ill patients: high plasma ADMA concentration is an independent risk factor of ICU mortality, Clin. Nutr. 22 (2003) 23–30. [10] M.J. O’Dwyer, F. Dempsey, V. Crowley, D.P. Kelleher, R. McManus, R. Ryan, Septic shock is correlated with asymmetrical dimethyl arginine levels, which may be influenced by a polymorphism in the dimethylarginine dimethylaminohydrolase II gener: a prospespective observational study, Crit. Care 10 (2006) R139. [11] T. Teerlink, HPLC analysis of ADMA and other methylated l-arginine analogs in biological fluids, J. Chromatogr. B 851 (2007) 21–29. [12] Y. Kakimoto, S. Akazawa, Isolation and identification of NG ,NG - and NG ,N G -dimethylarginine, N -mono, di, and trimethyllysine, and
[27]
[28]
[29]
[30]
[31]
[32]
[33] [34]
[35]
[36]
[37]
glucosylgalactosyl- and galactosyl-␦-hydroxylysine from human urine, J. Biol. Chem. 245 (1970) 5751–5758. R.J. McAllister, M.H. Rambausek, P. Vallance, D. Williams, K.-H. Hoffman, E. Ritz, Concentration of dimethyl-l-arginine in the plasma of patients with end-stage renal failure, Nephrol. Dial. Transplant 11 (1996) 2449–2452. M. Ivanova, C. Artusi, G.M. Boffa, M. Zaninotto, M. Plebani, HPLC determination of plasma dimethylarginines: method validation and preliminary clinical application, Clin. Chim. Acta 411 (2010) 1632–1636. C.E. Jones, C.J. Darcy, T. Woodberry, N.M. Anstey, Y.R. McNeil, HPLC analysis of asymmetric dimethylarginine in small plasma volumes using a Gemini-NC column at high pH, J. Chromatogr. B 878 (2010) 8–12. M. Davids, E. Swieringa, F. Palm, D.E.C. Smith, Y.M. Smulders, P.G. Scheffer, H.J. Blom, T. Teerlink, Simultaneous determination of asymmetric and symmetric dimethylarginine, l-monomethylarginine, l-arginine, and l-homoarginine in biological samples using stable isotope dilution liquid chromatography tandem mass spectrometry, J. Chromatogr. B 900 (2012) 38–47. S. Blackwell, D. St. J. O’Reilly, D.K. Talwar, HPLC analysis of asymmetric dimethylarginine (ADMA) and related arginine metabolites in human plasma using a novel non-endogenous internal standard, Clin. Chim. Acta 401 (2009) 14–19. I.M. Di Gangi, L. Chiandetti, A. Gucciardi, V. Moret, M. Naturale, G. Giordano, Simultaneous quantitative determination of NG,NG-dimethyl-l-arginine or asymmetric dimethylarginine and related pathway’s metabolites in biological fluids by ultrahigh-performance liquid chromatography/electrospray ionization-tandem mass spectrometry, Anal Chim. Acta 677 (2010) 140–148. J. Martens-Lobenhoffer, S.M. Bode-Böger, Quantification of l-arginine, asymmetric dimethylarginine and symmetric dimethylarginine in human plasma: a step improvement in precision by stable isotope dilution mass spectrometry, J. Chromatogr. B 904 (2012) 140–143. J.J. Pesek, M.T. Matyska, Hydrophilic Interaction Chromatography (HILIC) and Advanced Applications, in: P.G. Wang, W. He (Eds.), CRC Press, Boca Raton, 2011, pp. 1–26. J.J. Pesek, M.T. Matyska, R.I. Boysen, Y. Yang, M.T.W. Hearn, Aqueous normal phase chromatography using silica hydride-based stationary phases, Trends Anal. Chem. 42 (2013) 64–73. J.J. Pesek, R.I. Boysen, M.T.W. Hearn, M.T. Matyska, Hydride-based stationary phases: A rapidly evolving technology for the development of new bio-analytical methods, Anal. Methods 6 (2014) 4496–4503. J. Soukup, P. Janas, P. Jandera, Gradient elution in aqueous normal-phase liquid chromatography on hydrosilated silica-based stationary phases, J. Chromatogr. A 1286 (2013) 111–118. S. Jenkins, S.M. Fischer, L. Chen, T.R. Sana, Global LC/MS metabolomics profiling of calcium stressed and immunosuppressant drug treated Saccharomyces cerevisiae, Metabolites 3 (2013) 1102–1117. H. Eoh, K.Y. Rhee, Multifunctional essentiality of succinate metabolism in adaptation to hypoxia in Mycobacterium tuberculosis, PNAS 110 (2013) 6554–6559. N. Putluria, A. Shojaieg, V.T. Vasub, et al., Metabolomic profiling reveals potential markers and bioprocesses altered in bladder cancer progression, Cancer Res. 71 (2011) 7376–7386. K.R. Chalcraft, B.E. McCarry, Tandem LC columns for the simultaneous retention of polar and nonpolar molecules in comprehensive metabolomics analysis, J. Sep. Sci. 36 (2013) 3478–3485. J.J. Pesek, M.T. Matyska, J.A. Loo, S.M. Fischer, T.R. Sana, Analysis of hydrophilic metabolites in physiological fluids by HPLC-MS using a silica hydride-based stationary phase, J. Sep. Sci. 32 (2009) 2200–2208. J.J. Pesek, M.T. Matyska, M.T.W. Hearn, R.I. Boysen, Aqueous normal-phase retention of nucleotides on silica hydride columns, J. Chromatogr. A 1216 (2009) 1140–1146. J.J. Pesek, M.T. Matyska, J. Duley, M. Zamzami, S.M. Fischer, Aqueous normal phase (ANP) retention of nucleotides on silica hydride-based columns. method development strategies for analytes relevant in clinical analysis, J. Sep. Sci. 33 (2010) 930–938. J. Ge, F. Liu, E.H. Holmes, G.K. Ostrander, Q.X. Li, Aqueous normal phase liquid chromatography coupled with tandem time-of-flight quadrupole mass spectrometry for determination of zanamivir in human serum, J. Chromatogr. B 906 (2012) 58–62. J.J. Pesek, M.T. Matyska, A. Dang, Analysis of cycloserine and related compounds using aqueous normal phase chromatography/mass spectrometry, J. Pharm. Biomed. Anal. 64-65 (2012) 72–76. J.J. Pesek, M.T. Matyska, A. Kim, Evaluation of silica hydride-based stationary phases for the analysis of drugs of abuse, J. Sep. Sci. 36 (2013) 2760–2766. J.J. Pesek, M.T. Matyska, A. Dang, Analysis of ethyl glucuronide and ethyl sulfate using aqueous normal phase chromatography/mass spectrometry, J. Sep. Sci. 38 (2015) 1515–1520. J.E. Young, M.T. Matyska, J.J. Pesek, Liquid chromatography/mass spectrometry compatible approaches for the quantitation of folic acid in fortified juices and cereals using aqueous normal phase conditions, J. Chromatogr. A 1218 (2011) 2121–2126. A. Dang, M.T. Matyska, J.J. Pesek, The use of aqueous normal phase chromatography as an analytical tool for food analysis. Determination of histamine as a model system, Food Chem. 141 (2013) 4226–4230. R. Le, J.E. Young, J.J. Pesek, M.T. Matyska, Separation of 1,3-dimethylamylamine and other polar compounds in dietary supplement formulation using aqueous normal phase chromatography with mass spectrometry, J. Sep. Sci. 36 (2013) 2578–2583.
J.J. Pesek et al. / J. Chromatogr. A 1441 (2016) 52–59 [38] Y. Nolvachai, C. Kulsing, R.I. Boysen, M.T. Matyska, J.J. Pesek, P.J. Marriott, M.T.W. Hearn, Comparison of the performance of different silica hydride particles for the solid-phase extraction of non-volatile analytes from dark chocolate with analysis by gas chromatography-quadrupole mass spectrometry, Food Chem. 174 (2015) 434–439. [39] J.E. Young, J.J. Pesek, M.T. Matyska, LC–MS compatible approaches for quantitation of limonin in citrus juice, LCGC 33 (2015) 192–199. [40] R. MacNeill, R. Stromeyer, B. Urbanowicz, V. Acharya, M. Moussaille, J.J. Pesek, Silica hydride-based chromatography of LC–MS response altering compounds native to human plasma, Bioanalysis 4 (2012) 2877–2886.
59
[41] S. Rocchi, A. Rocco, J.J. Pesek, M.T. Matyska, D. Capitani, S. Fanali, Enantiomers separation by nano-liquid chromatography: Use of a novel-sub-2 m vancomycin silica hydride stationary phase, J. Chromatogr. A 1381 (2015) 149–159. [42] L. Abranko, J.F. Garcia-Reyes, A. Molina-Diaz, In-source fragmentation and accurate mass analysis of multiclass flavonoid conjugates by electrospray ionization time-of-flight mass spectrometry, J. Mass Spec. 46 (2011) 478–488.