Determination of S-adenosylmethionine, S-adenosylhomocysteine, and methylthioadenosine in urine using solvent-modified micellar electrokinetic chromatography

Alexander Vladimirovich Ivanov1*, Mariya Petrovna Kruglova2, Edward Danielevich Virus1, Polina Olegovna Bulgakova1, Sergei Vital’evich Grachev2, Aslan Amirkhanovich Kubatiev1,3
1 Department of Molecular and Cell Pathophysiology, Federal State Budgetary Scientific Institution “Institute for Pathology and Pathophysiology”, Moscow, Russia
2 I.M. Sechenov First Moscow State Medical University (Sechenov University), Moscow, Russia
3 Russian Medical Academy of Continuous Professional Education, Moscow, Russia
* for correspondence: Dr. Alexander Ivanov, ORCID ID 0000-0002-2424-6115, FSBSI “Institute of General Pathology and Pathophysiology”, Russia, Moscow, 125315, Baltiyskaya str.,8, phone +7-
499-151-1756, fax +7-495-601-2366, e-mail [email protected]; https://orcid.org/0000-0002-
2424-6115
Received: 27/09/2019; Revised: 22/11/2019; Accepted: 25/11/2019

Abstract

A new approach for direct determination of S-adenosylmethionine (SAM), S-adenosylhomocysteine (SAH), and methylthioadenosine (MTA) in urine was developed based on MEKC by using SDS modified with isobutanol in the presence of PEG-300. Analytes were first extracted with grafted phenylborononic acid. Using a 50 M internal diameter silica capillary of 32 cm total length filled with 0.05 M SDS, 0.05 M H3PO4, 5% (v/v) isobutanol, and 10% (v/v) PEG-300, LOQ of 0.15 M for SAM and SAH and 0.2 M for MTA were reached. Accuracy was 92% for MTA, 109% for SAH, and 105% for SAM, intra- and interday imprecision were <2.5 and 3%, respectively. The total time of analysis for one sample was 10 min. Analysis of 30 urine samples from healthy volunteers showed that the median SAM and SAH levels were 12.1 and 0.73 M, respectively. MTA levels, which were determined in urine for the first time (according to our data), were 0.43 M, and these values correlated well with the SAM level (r = 0.748, p <0.01).

Abbreviations: ATPA, S-(5-Adenosyl)-3-thiopropylamine; CMC, Critical micelle concentration; CV, Coefficient of variance; MTA, 5-Deoxy-5-(methylthio)adenosine; PBA, Perfluoroborate acid; SAH, S- Adenosylhomocysteine; SAM, S-Adenosylmethionine.

Keywords: methylthioadenosine, micellar electrokinetic chromatography, S-adenosylhomocysteine, S-adenosylmethionine, urine

1 Introduction

S-Adenosylmethionine (SAM) is a key metabolite that enables transfer of methyl groups to a variety of substrates (proteins, DNA, steroids, etc.) and allows the synthesis of polyamines in mammalians to occur [1–3]. SAM is converted into the intermediate metabolites S- adenosylhomocysteine (SAH) and 5-deoxy-5-methylthioadenosine (MTA). Given that these biochemical pathways are involved in many vital processes, their disruption is often associated with pathological conditions. In particular, an increase in the level of SAH or a decrease in the SAM/SAH ratio is considered to be an independent marker of a number of cardiovascular diseases [4–7]. A key role of these metabolites, primarily SAH, in the development of endothelial dysfunction (the leading pathogenic process of most vascular diseases) was confirmed in experimental studies [8–10].
Most clinical and laboratory studies in this area have focused on the determination of SAM and SAH levels in the blood plasma, and, less often, in whole blood. However, high clearance of SAM (93%) and SAH (39%) [11] makes it possible to use alternative analytical strategies based on determination of these analytes in urine without derivatization or the use of expensive HPLC-MS/MS methods for large-scale clinical studies. They include approaches based on HPLC with electrochemical [12] and UV [13] detection. Later a simple and rapid approach based on the use of CE-UV [14] was developed in which the concentration of analytes was reached by CITP. However, this approach was unsuitable for blood and urine analysis and required significant modification. Previously, guided by this strategy, we proposed the use of CITP in combination with SPE for SAM and SAH determination in urine [15]. As a result of the concentration of analytes in the capillary and SPE, the quantification limit was 0.2 M. Nevertheless, despite the advantages the technique offers for large-scale clinical studies (general availability of the method, high separation efficiency, determination selectivity, performance, and avoidance of a derivatization stage), its widespread use in clinical laboratories is hampered by the lack of an internal standard. As a result, the reproducibility of this method has reached only 6–10%. Therefore, to improve the CE technique, it was necessary to find another approach to the concentration and separation of analytes.
MEKC, which combines chromatography and CE, provides additional opportunities for analyte separation. Crucially, separation of analytes by MEKC occurs not only because of the difference of their mobilities in the electric field but also because of the partition of analytes with micelles [16,17]. In addition, the possibility of concentrating multiple analytes because of their sorption on micelles (so-called sweeping) makes it possible to increase the sensitivity of CE by two to three orders of magnitude. Therefore, we explored the possibility of using MEKC for the determination of SAM, SAH, and MTA in urine and to optimize and validate the proposed approach.

2 Materials and Methods

2.1 Equipment
A capillary electrophoresis system CE 3D (Agilent, Waldbronn, Germany) was used with an unbound silica capillary of 50 M id and 32 cm total (23.5 cm effective) length. The absorption signal at 254 nm with 10 nm width was registered at a frequency of 5 s–1. Reference wavelength was 290 nm with 10 nm width. The temperature of the capillary was 30C. Bond Elute-perfluoroborate acid (PBA) 10 mg cartridges (Agilent Technologies, Lake Forest, CA, USA) were used for SPE.

2.2 Chemicals
5-Adenosyl-3-thiopropylamine (ATPA), HCOOH 98%, SAM >75%, MTA, SAH >98%, DMSO, H3PO4 >85%, serotonine creatinine sulfate monohydrate (Sigma-Aldrich, St. Louis, MO, USA); HCl (37%; Acros, New Jersey, USA); Na2HPO4, Na2HPO4 dihydrate (>98%; PanReac, Barcelona, Spain); NaOH purum (Dia-M, Moscow, Russia); ACN (99.99%; Fisher Scientific, Loughborough, UK), and PEG300 (Merck, Darmstadt, Germany). All solutions were prepared with deionized water and filtered through 0.2m filters (Whatman, Dassel, Germany). The ATPA solution (8 mM) was prepared in 50% DMSO. The analyte solutions were prepared in deionized water and stored at –80С.

2.3 Sample collection
Material was from 30 healthy volunteers (14 males and 16 females, 4–61 years old, median 30 years). All participants gave informed consent for this study, which was approved by the local ethics committee. Fasting morning urine samples were collected according to the classical Nechiporenko method [18] and were placed on ice immediately. Samples were frozen at –80C prior to SPE. Each sample was analyzed within two weeks from the date of the collection.

2.4 SPE
We used a previously proposed method, based on SPE with the use of grafted PBA [15] with some modifications: 70L of 400 mM Na-phosphate buffer (pH 8.0) and 30 L of 100 M ATPA/calibrate solution were added to 300 L of urine immediately prior to SPE; 10 mg of the Bond Elute-PBA phase (Agilent, USA) was washed with 30% ACN + 1% (w/w) HCOOH (0.2 mL), water (0.4 mL), and 50 mM Na phosphate buffer (pH 8.0; 0.4 mL). The sample (0.4 mL) was loaded, and the phase was flushed with 50 mM Na-phosphate buffer (pH 8.0; 0.4 mL) and 10 mM Na phosphate buffer (pH 8.0; 0.4 mL). Analytes were desorbed with HCl (0.1 M, 0.2 mL).

2.5 CE-UV
The capillary was rinsed daily (before commencing experiments), with 1 M NaOH (2 min), H2O (1 min), and BGE (0.05 M SDS, 0.05 M H3PO4, 5% (v/v) isobutanol and 10% (v/v) PEG-300; 3 min). The sample was injected over 120 s at 50 mbar, and then a plug of the micellar solution (215 mM SDS with 50 mM H3PO4 and 10% (v/v) PEG-300) was injected over 45 s at 50 mbar. A negative polarity (from –10 to –11 kV for the first 2.5 min and then –16 kV) was applied for CE for 6.5 min. The capillary was then washed with BGE for 0.5 min.

2.6 Creatinine determination
Creatinine determination in urine was performed as described previously [19].

2.7 Results processing
Primary processing of electropherograms (integration) was carried out with Data Analysis ChemStation software Rev. B.01.03 (Agilent, Germany). Calibration data plotting and statistical analysis were conducted using SPSS Statistics software, v.22 (IBM Corporation). Where appropriate, data are presented as the mean with the variation is expressed as SD and the coefficient of variance (CV, %) or as median [1st, 3rd quartile]. The Kolmogorov–Smirnov test was used for analysis of normality at =0.1. Correlations were performed using Spearman rank testing and the Holm– Bonferroni method was used for correction of p-values at multiple comparisons.

3 Results and discussion

3.1 MEKC optimization
Under acidic conditions, SAM and SAH, being cations, migrate to the cathode. This property is used for their determination using methods of zone CE and CITP [14,15]. Figure 1A shows an electropherogram of analyte solution using 50 mM H3PO4 in the normal polarity mode. When 5 mM SDS was added to BGE, analyte peaks were not detected in normal polarity mode (Figure 1B) but appeared in reversed-field mode (Figure 1C) because of the formation of complexes between individual negatively charged SDS molecules and positively charged analytes. With an increase in the SDS level from 10 to 50 mM (i.e., already in the MEKC mode), a strong concentration effect was observed (Figure 1D–F) because of an increase in the retention factor; that is, fractions of analytes adsorbed on the surface of micelles. However, as can be seen in Figure 1, the use of pure MEKC did not lead to the separation of SAM, SAH, and ATPA. Changing the pH of the BGE to the alkaline region by adding up to 15 mM NaOH did not make the separation of analytes possible. Replacing H3PO4 with either HCOOH (1 M) or HCl (10 mM) was also not successful (data not shown). Although the addition of PEG-300 increased the retention time of the analytes it also did not lead to the separation of the peaks of SAM, SAH, and ATPA (Figure 2A).
The inclusion of organic MEKC modifiers isopropanol (0–12% v/v), 1-butanol (0–6% v/v), and isobutanol (0–6% v/v) was also studied. The addition of isopropanol made it possible to separate the SAM and SAH peaks without significant loss of sensitivity, but did not allow the SAM and ATPA peaks (data not shown) to be separated. Butan-1-ol turned out to be a more effective modifier; its inclusion at 5% v/v made it possible to achieve a significant (but still incomplete) separation of SAM and ATPA. However, in this case, the sensitivity was greatly reduced because of the broadening of the peaks, and the separation time exceeded 10 min. When 2.5% v/v of isobutanol was added, partial separation of SAM and ATPA was observed (Figure 2B), while with 5% v/v isobutanol, their complete separation was observed (Figure 2C) without any noticeable blurring of the peaks. The presence of PEG-300 was also necessary; without it, the separation of the SAM and ATPA peaks could not be achieved (Figure 2D).
In addition, the use of postinjection inclusion of the SDS solution without isobutanol made it possible to reduce the analysis time without effecting the separation of analytes. By varying the volumes of injection and postinjection, the optimal parameters for the capillary (effective length of 23.5 cm and 50 M i.d.) were established to be a 6000 and 2250 mbar·s, correspondingly. With these parameters, the volume of the injected sample was 280 nL, which corresponded to filling 60% of the working length of the capillary. Although the introduction of an even larger sample volume led to an increase in signal intensity while maintaining complete separation of the SAM–ATPA pair, the reproducibility of their peak areas was lost.
The SAM, ATPA, SAH, and MTA retention times were 3.650.04, 3.920.05, 4.310.06, and 5.60.1 min, respectively, and the relative retention times were 0.9330.005 (SAM), 1.0990.003 (SAH), and 1.4290.008 (MTA), CV ca. 0.25–0.53%.

3.2 Validation

Sensitivity. The detection limit (LOD, S/N = 3) was 0.04 M for SAM and SAH, and 0.05 M for ATPA and MTA. The limit of quantification (LOQ, S/N = 10) was 0.15 M for SAM and SAH, and 0.2M for ATPA and MTA. Extracts of model solutions were used to determine these indicators.
Linearity. A pool of 6 urine samples was mixed and divided into 3 aliquots of 6 ml for independent calibration of each analyte. Calibration solutions of SAM (0, 1.25, 2.5, 5, 10 and 20 mkM), SAH or MTA (0, 0.65, 1.25, 2.5, 5 and 10 mkM) and an internal standard were added to these aliquots before carrying out SPE as described in section 2.4. Each sample was analyzed 3 times and the measurement results were averaged. The calibration equations have the following form: y=[1.144±0.002]·x+[2.338±0.019] (SAM, R2=1.0), y=[1.209±0.039]·x+[0.745±0.182] (SAH, R2=0.996) and y=[2.478±0.031]·x+[0.464±0.147] (MTA, R2=0.999), where x is the analyte concentration and y is 10 × analyte / ATPA (10 M is the ATPA concentration in terms of the sample volume).
Repeatability. The intraday precision was examined by analyzing four samples prepared as a series from a single urine sample. CV for SAM, SAH, and MTA were 1.4, 2.4, and 3.5%, respectively. The evaluation of the between-run precision was performed by fivefold independent sample preparation on different days. Each sample was analyzed three times. CV for SAM, SAH, and MTA were 2.0, 3.0, and 2.6%, correspondingly.
Accuracy. Accuracy was investigated as recovery by independently adding SAM (0, 10, and 20 M), SAH (0, 0.67, and 2 M), and MTA (0, 0.5, and 2 M) to urine samples. Then all samples were then processed as described in section 2.4 and analyzed three times. Corresponding recovery levels are shown in Table 1. Verification. Thirty urine samples from healthy donors were analyzed to verify our method. The MTA level in 5 samples and the SAH level in two samples were lower than LOQ. Since urine analyte levels, including creatinine did not obey the law of normal distribution, the measured concentrations of analytes are presented as medians and quartiles (Table 2). Depending on many factors the concentration of primary urine in the kidneys can undergo strong changes, this fact affects the variability of the content of metabolites in it. So in addition to the absolute content of SAM, SAH and MTA, the table also contains the analyte / creatinine ratios, whereas creatinine is an endogenous compound which is practically not reabsorbed by the kidneys is easily determined. Therefore it is traditionally used to measure the concentrating ability of the kidneys in determining proteins, amino acids and other metabolites in the urine. In addition, SAM is required as a methyl group donor for the synthesis of creatine, a creatinine precursor. Although the analysis of these samples showed that SAM and SAH levels correlate well with creatinine levels ( = 0.72 and 0.56 correspondigly, p <0.01), the magnitude of the analyte / creatinine ratio for SAH and MTA was higher than the magnitude of their absolute concentrations in urine. It was previously shown that the level of SAM or SAM / SAH ratio, but not SAM / creatinine or SAH / creatinine, decreased significantly in patients with CKD [20]. Urine SAM and SAH correlated with each other ( = 0.479, p=0.029). Significant correlation was also observed between urine SAM and MTA ( = 0.602, p=0.07).
Figure 3 shows electropherograms of extracts of the model solution; i.e., urine samples with and without the addition of analytes. The levels of analytes in the urine of healthy volunteers obtained in this work are generally consistent with previous results [15,20,21]. It was also previously found that the distribution of urine SAM, SAH was different from normal, characterized by pronounced asymmetry and kurtosis [15]. The association of SAM with SAH levels was also noted earlier [21] not only in urine but also in blood plasma. The above-mentioned high level of excretion of SAM and SAH gives reason to use it in urine as an alternative to the determination of these metabolites in blood plasma. Although it was previously shown that SAM (and SAH) levels in urine and plasma correlate positively, this association was weaker than between SAM and SAH in plasma or urine separately [21]. Therefore, additional investigation is required in the level association of these metabolites in plasma and urine of healthy people and in patients with various pathologies.
To our knowledge, there are no data available about MTA levels in urine and no reference values are known. Earlier it was shown that the ratio of MTA/SAM in plasma was about 0.15 [22], which is much higher than in urine (~0.04). This may be because of MTA absorption by the tubules of the kidneys (however, there are no data on the clearance of the MTA) or because of peculiarities of the chemical derivatization of the method during which a portion of SAM may be spontaneously converted into SAH.

3.3 Discussion

The S-adenosyl derivatives of amines and amino acids chosen in this work possess affinity to SDS, both to free molecules and to micelles. At the same time, based on the retention times of analytes (SAM>ATPA>SAH>MTA), the charge of the amino acid residue plays an important role in the stability of the analyte–SDS complex. However, the use of pure MEKC prevents determination of these analytes because of the lack of sufficient selectivity. It was previously shown that the main factor influencing the selectivity of MEKC is the nature of the cosurfactant. Indeed, the addition of a cosurfactant to the micelle solutions increased separation efficiency [23]. In this case, amphiphilic or moderately hydrophobic compounds, such as ACN, tetrahydrofuran, methanol, ethanol, propanol, and butanol are often used. They have an effect on critical micelle concentration (CMC), aggregation number, the size of the micelles, and their mobility. So, for example, butan-1-ol and propan-1-ol at a concentration of 1–5% reduce CMC from 8.2 to 2–5 mM, cause swelling of SDS micelles, increase their hydrodynamic radius, and decrease both their -potential and surface charge density. Methanol and ACN, in contrast, cause an increase in CMC [24,25]. We studied the effect of several cosurfactants on the selectivity of MEKC, and isobutanol turned out to be the best choice for resolving the problem. The introduction of isopropanol or butan-1-ol also significantly improved the separation of analyte peaks, but their use was not effective in the complete separation of SAM and ATPA.
In addition, to achieve high selectivity, it was necessary to use PEG-300 in the BGE and postinjection solution. Typically, PEG is used in CE to suppress EOF, thus making peak separation possible. However, it cannot be ruled out that PEG-300 also affects the selectivity of micelles, although we have not found studies on this issue.
Although formally the number of theoretical plates of MEKC (1.7×105) was an order of magnitude lower than that in the previously developed transient isotachophoresis (1.6×106)[15], at approximately equal analysis time, the sensitivity of these approaches turned out to be similar. The LOD of the latter approach was even lower than that of temporary isotachophoresis (40–50 nM vs. 70 nM in [15]). This was because of the possibility of injecting double the volume (compared with the previous approach) and because concentration of analytes could be achieved using sweeping.
The introduction of an internal standard made it possible to improve the CE reproducibility by reducing the CV of between-run reproducibility from 6–10% [10] to 2–3% and decreasing intraday precision from 4–6% to 1.5–3.5%, respectively.

4 Concluding remarks

This is the first time that a solvent-modified MEKC method has been developed for the determination of SAM, SAH, and MTA levels in urine. In the new approach, instead of CITP [15], we used another mechanism of concentration (sweeping), which made it possible to double the injection volume. The introduction of cosurfactant isobutanol and neo-ion detergent PEG-300 resolved the problem of analyte separation. The use of an internal standard greatly improved the reproducibility of the method. Given its accessibility, ease of use, reliability, and noninvasive nature, this approach can be used for large-scale investigations of samples from large groups of patients.

Acknowledgements
This study was supported by a grant from the Russian Science Foundation (project Nu. 16-15-10340).
Conflict of interest statement
The authors have declared no conflict of interest.

5 References
[1] Lu, S. C., Mato, J. M., Physiol. Rev. 2012, 92, 1515-1542.
[2] Soda, K., Int. J. Mol. Sci. 2018, 19, E3106. doi: 10.3390/ijms19103106.
[3] Janke, R., Dodson, A. E., Rine, J., Annu. Rev. Cell. Dev. Biol. 2015, 31, 473-496. doi: 10.1146/annurev-cellbio-100814-125544.
[4] Zhang, H., Liu, Z., Ma, S., Zhang, H., Kong, F., He, Y., Yang, X., Wang, Y., Xu, H., Yang, A., Tian, J., Zhang, M., Cao, J., Jiang, Y., Guo, X., Mol. Med. Rep. 2016, 14, 289-300.
[5] Huang, X., Lv, X., Song, H., Yang, Q., Sun, Y., Zhang, W., Yu, X., Dong, S., Yao, W., Li, Y., Wang, Q., Wang, B., Ma, L., Huang, G., Gao, Y., Clin. Chim. Acta 2017, 471, 314-320.
[6] Bharatkumar, V. P., Rudreshkumar, K. J., Nagaraja, D., Christopher, R., Clin. Chim. Acta 2016, 458, 44-48.
[7] Xiao, Y., Zhang, Y., Wang, M., Li, X., Su, D., Qiu, J., Li, D., Yang, Y., Xia, M., Ling, W., Am. J. Clin. Nutr. 2013, 98, 1162-1169.
[8] Sipkens, J. A., Hahn, N. E., Blom, H. J., Lougheed, S. M., Stehouwer, C. D., Rauwerda, J. A., Krijnen,
P. A., van Hinsbergh, V. W., Niessen, H. W., Atherosclerosis 2012, 221, 48-54.
[9] Barroso, M., Florindo, C., Kalwa, H., Silva, Z., Turanov, A. A., Carlson, B. A., de Almeida, I. T., Blom, H. J., Gladyshev, V. N., Hatfield, D. L., Michel, T., Castro, R., Loscalzo, J., Handy D. E., J. Biol. Chem. 2014, 289, 15350-15362.
[10] Xiao, Y., Xia, J., Cheng, J., Huang, H., Zhou, Y., Yang, X., Su, X., Ke, Y., Ling, W., Circulation 2019, 139, 2260-2277.
[11] Garibotto, G., Valli, A., Anderstam, B., Eriksson, M., Suliman, M. E., Balbi, M., Rollando, D., Vigo, E., Lindholm, B., Kidney Int. 2009, 76, 293-296.
[12] Melnyk, S., Pogribna, M., Pogribny, I. P., Yi, P., James, S.J., Clin. Chem. 2000, 46, 265–272.
[13] Luippold, G., Delabar, U., Kloor, D., Mühlbauer, B., J. Chromatogr. B 1999, 724, 231–238.
[14] Uthus, E. O., Electrophoresis 2003, 24, 1221-1226.
[15] Ivanov, A. V., Virus, E. D., Nikiforova, K. A., Kushlinskii, N. E., Luzyanin, B. P., Maksimova, M. Y., Piradov, M. A., Kubatiev, A. A., Electrophoresis 2016, 37, 2663-2669.
[16] Lin, C. H., Kaneta, T., Electrophoresis 2004, 25, 4058-4073.
[17] Terabe, S., Annu. Rev. Anal. Chem. 2009, 2, 99-120.
[18] Nechiporenko, A. Z., Pediatriia 1967, 46, 58-60.
[19] Liotta, E., Gottardo, R., Bonizzato, L., Pascali, J. P., Bertaso, A., Tagliaro, F., Clin. Chim. Acta.
2009, 409, 52-55.
[20] Kruglova, M. P., Grachev, S. V., Bulgakova, P. O., Ivanov, A. V., Virus, E. D., Nikiforova, K. A., Fedoseev, A. N., Savina, G. D., Kubatiev, A. A., Lab. Med. 2019, lmz035. doi: 10.1093/labmed/lmz035.
[21] Stabler, S. P., Allen, R. H., Clin. Chem. 2004, 50, 365-372.
[22] Ivanov, A. V., Dubchenko, E. A., Kruglova, M. P., Virus, E. D., Bulgakova, P. O., Alexandrin, V. V., Fedoseev, A. N., Boyko, A. N., Grachev, S. V., Kubatiev, A. A.. J. Chromatogr. B: Anal. Technol. Biomed. Life Sci. 2019, 1124, 366-374.
[23] Hansen, S. H., Gabel-Jensen, C., El-Sherbiny, D. T. M., Pedersen-Bjergaard S., TrAC Trends Anal. Chem. 2001, 11, 614-619.
[24] Anderson, J. L., Analytical Separation Science, Wiley-VCH, Weinheim 2015.
[25] Cao, Y., Ni, X., Sheng, J., J. Chromatogr. A 2011, 1218, 2598-2603.

Figure legends

Figure 1. Influence of SDS on analyte concentration in CE. Sample: 0.5 mM SAM, SAH, ATPA and MTA in 25 mM HCl, 250 mbar·s injection. Voltage +12 kV (A, B) or –12 kV (C–F). BGE: (A) 50 mM H3PO4; (B, C) 50 mM H3PO4 with 5 mM SDS; (D) 50 mM H3PO4 with 10 mM SDS; (E) 50 mM H3PO4 with 25 mM
SDS; (F) 50 mM H3PO4 with 50 mM SDS. Peaks are labeled: 1: SAM, 2: ATPA, 3: SAH, and 4: MTA.
Figure 2. Optimization of MEKC. Sample: 5 M SAM, SAH, ATPA, and MTA in 0.1 M HCl, 6000 mbar·s injection and 2250 mbar·s postinjection (50 mM H3PO4 + 0.215 M SDS + 10% v/v PEG-300). Peaks are labeled: 1: SAM, 2: ATPA, 3: SAH and 4: MTA.
(A) BGE: 50 mM H3PO4 with 50 mM SDS and 10% v/v PEG-300. (B) BGE + 2.5% v/v isobutanol. (C) BGE PEG300
+ 5% v/v isobutanol. (D) BGE + 5% v/v isobutanol but without PEG-300.
Figure 3. Electropherograms of (A) model solution (5 M SAM, SAH, ATPA and MTA in 0.1M HCl), urine extract (B) and urine spiked with 10 M SAM and 5 M SAH, MTA (C). Peaks are labeled: 1: SAM, 2: ATPA, 3: SAH, and 4: MTA.

Tables
Table 1. Recovery of the method
Analyte Added level (M) Recovery (%)
SAM 10 105.03.8
20 101.64.2
SAH 0.67 109.45.4
2 107.37.1
MTA 0.5 92.46.2
2 91.16.9

Table 2. SAM, SAH, and MTA levels in human urine samples
Analyte Median 1st quartile 3rd quartile Min–max
SAM (M), n=30 12.06 8.5 15.05 2.8–21.4
SAH (M), n=28 0.73 0.56 1.13 0.24–1.66
MTA (M), n=25 0.49 0.22 0.74 0.22–1.1
Creatinine (mM),
n=30 8.62 6.38 14.94 3.05–22.7
SAM/SAH, n=28 13.1 10.7 21 5.5–28.0
MTA/SAM, n=25 0.041 0.030 0.051 0.017–0.082
SAM/Creatinine
(M/mM) 1.17 0.91 1.49 0.68–2.71
SAH/Creatinine
(M/mM) 0.083 0.069 0.102 0.029–0.350
MTA/Creatinine
(M/mM) 0.045 0.037 0.069 0.018–0.137