- Email: [email protected]
Very fast and direct capillary zone electrophoresis method for the determination of creatinine and creatine in human urine
Analytica Chimica Acta 521 (2004) 53–59
Very fast and direct capillary zone electrophoresis method for the determination of creatinine and creatine in human urine J. Rodr´ıguez a,∗ , J.J. Berzas a , G. Castañeda a , N. Mora b , M.J. Rodr´ıguez b a
Department of Analytical Chemistry and Foods Technology, Univ. de Castilla-La Mancha, Avenida Camilo Jose Cela 10, Ciudad Real 13071, Spain b Department of Analytical Chemistry, University of Extremadura, Badajoz, Spain Received 4 March 2004; received in revised form 24 May 2004; accepted 24 May 2004 Available online 28 July 2004
Abstract A capillary zone electrophoresis (CZE) method was investigated for the determination of creatinine and creatine in human urine by using a fused-silica capillary (75 m ID × 30 cm total length, 10 cm effective length). The separation was performed using an hydrodynamic injection time of 3 s (0.5 psi), a voltage of −15 kV and a 60 mM phosphoric buffer solution at pH 2.2 as electrolyte separation. Under these conditions, the analysis takes less than 2.7 min. A linear response over the 3.0–120 mg L−1 concentration range was investigated for compounds. A dilution of the sample (water:urine) was the only step necessary before the electrophoresis analysis. Detection limits of 0.7 and 1.3 mg L−1 for creatinine and creatine (S/N = 3) were obtained. The developed method is easy, rapid and sensitive and has been applied to determine creatinine and creatine in different human urine samples. © 2004 Elsevier B.V. All rights reserved. Keywords: CZE; Creatinine; Creatine; Human urine
1. Introduction Diabetic nephropathy (DN) is a serious complication of diabetes mellitus [1]. However, when DN was diagnosed by the classical methods, little could be done to prevent the progressive downhill course to renal failure. It would be more significant, therefore, if DN could be detected at an even earlier stage, so that intervention could reverse the process. Nonprotein nitrogen compounds, including creatinine and creatine could serve as markers for renal function. Creatine is one of the good indicators of the glomerular filtration rate of the kidneys, and creatinine is the major inactive breakdown product of creatine and has no function in the body, it is simply excreted in urine, which is why it is monitored clinically as a marker of kidney function [2]. Creatine is synthesized in the body, and its main dietary source is red meat. Creatine has recently achieved great popularity as an ergogenic aid [3]. Creatine is converted to phosphocreatine in muscle in a reversible reaction with adenosine
∗
Corresponding author. Fax: +34-926-295318. E-mail address: [email protected] (J. Rodr´ıguez).
0003-2670/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.aca.2004.05.058
triphosphate (ATP), facilitated by creatine kinase [4]. When muscle contractions deplete the immediate supply of ATP, the phosphocreatine can rephosphorylate ADP to replenish the supply of ATP. The study of the uptake of creatine as a result of creatine supplementation, a practice increasingly common among bodybuilders and athletes, has lead to a need to measure urinary creatine concentration. The structures of creatine and creatinine are shown in Fig. 1. Numerous methods for the assay of these markers have been reported, such as FIA methods [5,6], by using biosensors [7], high-performance liquid chromatography [8–11] and capillary electrophoresis. The first two types of methods are limited due to the slow speed of analysis, the inability to test for multiple markers at one time, non-specificity for the analyte of interest and the consumption of large amounts of reagents and samples. On the other hand, the HPLC methods have relatively long analysis times the consume large amounts of solvents. However, capillary electrophoresis techniques, in many cases, have higher resolution, greater efficiency and smaller sample volumes [1]. There are a few articles reporting the determination of creatine and creatinine. Some include separations from
54
J. Rodr´ıguez et al. / Analytica Chimica Acta 521 (2004) 53–59
CH3 N
2.2. Apparatus and operating conditions NH
N H
O
Creatinine CH3 HN
O
N
OH
NH2
Creatine Fig. 1. Structures of creatinine and creatine.
other compounds. Yan et al. [12] employ sodium cholate at phosphate buffer pH 7.4 with voltage of 30 kV in the determination of urea, uric acid, creatine and creatinine. Burke et al. [13] make the determination by MEKC too but using 150 mM SDS in a 30 mM phosphate buffer pH 6, in this method creatinine and creatine were separated in 6 min without interferences. A novel capillary electrophoresis chip-based detection system is described by Wang and Chatrathi et al. [14]. Capillary zone electrophoresis have also been applied [1,15]. Kong et al. proposed a method for determining creatinine, creatine, urea and uric acid. The separation was achieved in 25 mM phosphate buffer pH 3.45; the separation takes less than 15 min and it was applied to urine and plasma samples. In this paper, we propose a very rapid and easy method using capillary zone electrophoresis (CZE) to determine creatine and creatinine in urine without any treatment of the biological sample. The separation take less than 2.5 min. The method is very short and sensitive enough to do this determination in human urine without the use of surfactants. 2. Materials and methods 2.1. Materials All solvents and reagents were of analytical-reagent grade unless indicated otherwise. Creatinine and creatine were obtained from Sigma. Standard solutions were prepared with deionized water (Milli-Q quality) and stored at 4 ◦ C. Diluted solutions of urine were prepared daily by diluting freshly collected human urine with purified water. One milliliter of urine was 10 times diluted using purified water in a calibrated flask. Triethanolamine solution was obtained from Fluka, Biochemika. Buffer solutions were prepared by dissolving an the adequate quantity of H3 PO4 in deionized water and then adjusting with NaOH solution to the required pH. The set of separation vials was changed after six runs.
A Beckman P/ACE System MDQ (Fullerton, CA, USA) equipped with a diode-array detector was used. Beckman capillary electrophoresis software controlled the system. Separation was carried out on a 30 cm long (10 cm to the detector, short way) × 75 m ID fused-silica capillary housed in a cartridge with a detector window 100 m × 800 m. The wavelength selected for the electropherograms was 205 nm. The capillary was conditioned prior to its use by rising with 0.5 M NaOH for 20 min, water for 10 min and finally phosphate buffer. As the human urine samples contain creatinine and creatine, in each case the urine was analyzed after a dilution 1:10 (v/v) with Milli-Q grade water. The capillary was flushed between injections for 2 min with NaOH 0.1 min and 4 min with electrolyte. Electrophoretic separation was performed using as running buffer a 60 mM phosphoric buffer solution at pH 2.2, an hydrodynamic injection time of 3 s (0.5 psi) and a voltage of −15 kV (reversed polarity mode). Under the selected conditions the current was around −108 A. The electropherograms were recorded at 205 nm. Duplicate injections of the solutions were performed and average-corrected peak areas (CPA) (area/migration time) were used for quantitative analysis. The data generated from the first two injections of a sequence were not used on account of the necessary equilibration of the system.
3. Results and discussion The effect of pH of the running electrolyte had a significant impact on the ionization of the acidic silanols of the capillary wall and the electrophoretic mobility of the compounds studied. The buffer tested was phosphate (60 mM) and it was found that at pH 2 the migration times are shorter but the creatinine peak is overlapped with an unknown second peak. On the other hand at pH 3 the electrophoretic mobility decreases and the migration time for creatinine was 4.5 min. For this reason, after analyzing the effect of pH in this interval range a pH 2.2 was selected. At these low pH values our analytes are found before the migration time of EOF; this fact shows that creatinine and creatine are cationic forms. The effect of the composition of the separation electrolyte was studied by using the same pH value with the same ionic strength (50 mM) but different media. Besides phosphate buffer, an acetate and perchlorate buffer were tested but the highest signals and best peak shapes were found using phosphate buffer. In order to try to obtain symmetric and sharper peaks, a phosphate buffer (60 mM) was prepared, the pH was adjusted to 2.2 by addition of TEA (triethanolamine), and the separation was made as previously by applying −15 kV using a hydrodynamic injection of 3 s (0.5 psi), but no better
J. Rodr´ıguez et al. / Analytica Chimica Acta 521 (2004) 53–59
with TEA
55
without TEA
AU
Creatine
Creatinine
40000
20000
0
0.0
1.0
2.0
3.0
Migration Time(min) Fig. 2. Influence of the addition of triethanolamine (TEA). Operating conditions: 60 mM of phosphate buffer, −15 kV of voltage, 3 s injection time (0.5 psi).
behavior by addition of TEA of the peaks of creatinine and creatine was found, as it is possible to see in Fig. 2. The effect of the concentration of phosphate buffer solution (20–70 mM) on the migration time of the compounds
and the shape of the peaks was studied (Fig. 3). When the concentration of buffer increases the migration time decreases and the peaks become sharper. More than 60 mM was not used due to the high current density (to 70 mM of
40000
30000
20 mM
10000
40 mM
60 mM
AU
20000
0
0.0
1.0
2.0
3.0
Migration Time (min) Fig. 3. Influence of phosphate buffer concentration over a urine sample after the dilution process. Operating conditions: 60 mM of phosphate buffer, −15 kV of voltage, 3 s injection time (0.5 psi). Migration order: creatinine, unknown compound and creatine.
56
J. Rodr´ıguez et al. / Analytica Chimica Acta 521 (2004) 53–59 60000
AU
40000
___ 3s - - - 10s 20000
___ 1.8s - - - 5s
0
0.0
1.0
2.0
3.0
Migration Time (min) Fig. 4. Influence of amount of sample injected (injection time and pressure apply) over a urine sample after the dilution process. Operating conditions: 60 mM of phosphate buffer, −15 kV of voltage, 3 s injection time (0.5 psi). Migration order: creatinine, unknown compound and creatine.
phosphate buffer concentration correspond −124 A). The buffer concentration of 60 mM was thus considered to be optimal. Sample components from the diluted urine can become adsorbed onto the capillary surface and change the effective charge on the wall. So, in order to prevent this effect the capillary is flushed between injections for 2 min with NaOH 0.1 M and then 4 min with separation electrolyte in order to obtain an stable EOF (electroosmotic flow). Minor times of the washing step give worse electropherograms. In order to decrease the detection limits in urine, the injection time was varied between 0.6 and 10 s at 0.5 psi (1 psi = 6894.76 Pa), and it was found that up to 0.5 psi and 3 s the peak became higher, but if we continued increasing the injected sample quantity the electrophoretic peaks became double (as it is possible to observe in Fig. 4). Then it was selected as optimum an injection condition of 0.5 psi during 3 s. The effect of varying the voltage from −5 to −20 kV was investigated (Fig. 5). A potential of −15 kV yielded the best compromise in terms of run time, current generated and efficiency of separation. As consequence, this potential was used in subsequent stages of method development. Selected conditions. From the studies carried out before, the procedure summarized below was convenient to separate creatinine and creatine in diluted urine samples: • separation electrolyte: 60 mM phosphate buffer pH 2.2; • voltage: −15 kV (reverse polarity mode), current: −108 A;
• capillary: fused-silica 30 cm long (10 cm to the detector, short way) × 75 m ID; • injection: hydrodynamic, 3 s (0.5 psi) (reverse polarity mode); • detection wavelength: 205 nm. Under these conditions, the migration times were 1.0 and 2.4 min for creatinine and creatine, respectively. In all cases the urine samples were diluted with deionized water (a dilution of urine:water (1:10) was the only step necessary before the electrophoresis analysis). 3.1. Quantitative aspects 3.1.1. Precision Two different urine samples (from a middle-aged woman and a pregnant woman) were analyzed (eight repetition of each one). The precision of the proposed method is expressed in terms of relative standard deviation (R.S.D.). The reproducibility was evaluated between-day using the urine from the middle-aged woman. The results showed that the repeatability for every component in each day is satisfactory. In terms of reproducibility, the comparison of the standard deviations with the Fisher test did not provide any significant difference between both days series, for a significance level of 0.05. The repeatability of the migration times and corrected peak area were good in both cases, with a %R.S.D. less than 2% in all cases (Table 1).
J. Rodr´ıguez et al. / Analytica Chimica Acta 521 (2004) 53–59
57
40000
15 kV 10 kV 5 kV 30000
AU
20000
10000
0
0.0
2.0
4.0
6.0
8.0
Migration Time(min) Fig. 5. Influence of separation voltage on migration times and resolution, over a urine sample after the dilution process. Operating conditions: 60 mM of phosphate buffer, −15 kV of voltage, 3 s injection time (0.5 psi). Migration order: creatinine, unknown compound and creatine.
3.2. Specificity
3.3. Limits of detection and quantification
Specificity can be determined by measurement of peak homogeneity. Because of the different techniques available in a DAD (diode-array detector) are not equally effective for the detection of possible impurities or interferences in an electrophoretic peak, the use of several techniques is recommended [16]. In this work the techniques used to validate the peak purity of the studied compounds present in urine samples were [17]:
Limits of detection (LODs) and quantification (LOQs) were estimated in the usual way. The LOD was obtained as the drug concentration that caused a peak height 3 times the baseline noise level and the LOQ was calculated as 10 times the baseline noise level [18]. The LODs were 0.7 mg L−1 for creatinine and 1.3 mg L−1 for creatine, and the LOQs were 2.3 mg L−1 for creatinine and 4.3 mg L−1 for creatine.
• normalization and comparison of spectra from different peak sections; • absorbance at two wavelengths. Both techniques showed that the purity of the peaks corresponding to the compounds studied in urine present a high level of purity. Therefore, no interference of other urine compounds were observed. Table 1 Repeatability
All results were obtained by using CPA (corrected peak areas) for calculations (to obtain CPA, peak area was divided by its corresponding migration time). A calibration graph was obtained in water by using six spiked water samples. It was tested over the range from 3 to 120 mg L−1 for each analyte. The linear regression equations obtained using the least-squares method and coefficients of correlation are: Creatinine : CPA = (−497.9 ± 902.1) + (992.2 ± 23.2)c,
R.S.D. (%)
r 2 = 0.995,
Creatinine
Migration time Corrected area
3.4. Linearity range and calibration curves
Creatine
creatine :
Middle-aged woman
Pregnant woman
Middle-aged woman
Pregnant woman
0.18 1.97
0.21 0.67
0.97 1.53
0.65 0.55
CPA = (951.7 ± 535.4) + (981.4 ± 17.4)c,
r = 0.998, 2
CPA = (a ± Sa ) + (b ± Sb )c
where a is the intercept, Sa the standard deviation of intercept, b the slope, Sb the standard deviation of slope.
58
J. Rodr´ıguez et al. / Analytica Chimica Acta 521 (2004) 53–59 60000
Creatinine
Creatine
AU
40000
20000
0
0.0
1.0
2.0
3.0
Migration Time (min) Fig. 6. CZE electropherogram of creatinine and creatine in urine sample with the final selected conditions. Migration order: creatinine, unknown compound and creatine.
After that, and with the object to know if it is possible to determine the concentration of creatinine and creatine in an urine sample by using directly this calibration graph obtaining with the water samples spiked or if it is necessary to apply the standard addition method to each urine sample, a mixture of several human urine samples was analyzed by standard addition. The slopes of both graphs (aqueous medium and standard addition over a diluted urine samples) were compared. First of all the variances of both data groups were compared and it was found no significant differences between them. Then it was possible to compare the slopes and it was found no significant differences between the slopes of both linear graphs (aqueous and urine medium). A significance level of 0.05 was used. Thus it is possible to determine creatinine and creatine in human urine sample by applying directly the calibration graph in water.
In order to compare the obtained results by the use of the calibration graph, the addition standard method was also applied over three of the urine samples analyzed, the results obtained have totally satisfactory. An electropherogram of an urine sample (from the middle aged women) is presented in Fig. 6. Urine creatinine levels can be used as a screening test to evaluate kidney function. Normal values are highly dependent on the age and lean body mass of the person the urine is being collected from. Urine creatinine values may therefore be quite variable and can range from 300 to 1300 mg L−1 , as can be seen in Table 2 except to the sportive young man,
Table 2 Determination of creatinine and creatine in different urine samples Type of urine sample
3.5. Applications In order to apply the proposed method, six different types of urine samples (middle aged woman, middle aged man, sportive young man, boy, girl and pregnant woman) were analyzed. Two samples of each type of urine were analysed in duplicate. The concentration of creatinine and creatine were determine after dilution with water and the values obtained for each original urine samples are presented in Table 2.
Girl Boy Sportive young man (with urea problems)a Middle aged woman Middle aged man Pregnant woman
Concentration in original urine sample (mg L−1 ) Creatinine
Creatine
896 1179 2042
397 392 Not detected
614 554 1050
890 75 406
a For this sample, a dilution factor of 50 was necessary (1.0 mL of urine was 50 times diluted with purified water).
J. Rodr´ıguez et al. / Analytica Chimica Acta 521 (2004) 53–59
due to the exercise may caused increased creatinine clearance and also his high urea level. (For this analysis a 25 mL dilution factor was applied.) Also the glomerular filtration rate is substantially increased in pregnant women (see pregnant woman in the mentioned table). Creatine levels are of the same order of magnitude as creatinine levels only when subjects have recently ingested creatine, while somewhat elevated urinary creatine concentrations in non-supplementing subjects can be an indication of a degenerative disease of the muscle (see the high level of creatine in middle aged woman in the table, may be due to her main dietary source is red meat).
4. Conclusions In this work, a fast and direct capillary zone electrophoresis method for the determination of creatinine and creatine in human urine sample is described. The proposed method is faster than those previously proposed. The parameters calculated for this method are satisfactory.
Acknowledgements The authors are grateful to the DGES of the Ministerio de Educación y Ciencia (project BQU 2001-1190) and to the Junta de Extremadura for the financial support.
59
References [1] Y. Kong, N. Zheng, Z. Zhang, R. Gao, J. Chromatogr. A 987 (2003) 477–483. [2] A.D. Baines, Disorders of the kidney and urinary tract, in: A.G. Gornall (Ed.), Applied Biochemistry of Clinical Disorders, 2nd ed., JB Lippincott, Philadelphia, PA, 1986, pp. 139–171. [3] W.J. Kramer, J.S. Volek, Clin. Sports Med. 18 (1999) 651. [4] G. Zubay, Biochemistry, Wm.C. Brown, Dubuque, IA, 1998, p. 39. [5] T. Yao, K. Kotegawa, Anal. Chim. Acta 462 (2) (2002) 283–291. [6] G. Del Campo, A. Irastorza, J. Alfonso-Casado, Fresen. J. Anal. Chem. 352 (6) (1995) 557–561. [7] R.J. Stefan, R.G. Bokretsion, J.F. Van-Staden, H.Y. Aboul-Enein, Talanta 60 (6) (2003) 1223–1228. [8] T. Seki, Y. Orita, K. Yamaji, A. Shinoda, J. Pharm. Biomed. Anal. 15 (1997) 1621–1626. [9] P. Ellerbe, A. Cohen, M.J. Weleh, V.E. White, Anal. Chem. 62 (1990) 2173–2177. [10] G. Werner, V. Schneider, J. Emmet, J. Chromatogr. 525 (1990) 265– 275. [11] S.D. Yang, J.M. Xu, L. Yang, Y.N. Ma, F. Bai, J. Liq. Chromatogr. 12 (1989) 1791–1807. [12] S.L. Yan, P.Z. Lin, M.W. Hsiao, J. Chromatogr. Sci. 37 (2) (1999) 45–50. [13] D.G. Burke, P.G. MacLean, R.A. Walker, P.J. Dewar, T. Smith-Palmer, J. Chromatogr. B 732 (1999) 479–485. [14] J. Wang, M.P. Chatrathi, Anal. Chem. 75 (3) (2003) 525–529. [15] E.A. Clark, J.C. Fanguy, C.S. Henry, J. Pharm. Biomed. Anal. 25 (2001) 795–801. [16] H. Farbe, A. LeBris, M.D.J. Blanchin, J. Chromatogr. A 697 (1–2) (1995) 81–88. [17] L. Huber, Applications of Diode-array Experience, vol. 12, Hewlett-Packard Publication, 1989, p. 5953 [18] K.D. Altria, Chromatographia 35 (3–4) (1993) 177–182.
Very fast and direct capillary zone electrophoresis method for the determination of creatinine and creatine in human urine J. Rodr´ıguez a,∗ , J.J. Berzas a , G. Castañeda a , N. Mora b , M.J. Rodr´ıguez b a
Department of Analytical Chemistry and Foods Technology, Univ. de Castilla-La Mancha, Avenida Camilo Jose Cela 10, Ciudad Real 13071, Spain b Department of Analytical Chemistry, University of Extremadura, Badajoz, Spain Received 4 March 2004; received in revised form 24 May 2004; accepted 24 May 2004 Available online 28 July 2004
Abstract A capillary zone electrophoresis (CZE) method was investigated for the determination of creatinine and creatine in human urine by using a fused-silica capillary (75 m ID × 30 cm total length, 10 cm effective length). The separation was performed using an hydrodynamic injection time of 3 s (0.5 psi), a voltage of −15 kV and a 60 mM phosphoric buffer solution at pH 2.2 as electrolyte separation. Under these conditions, the analysis takes less than 2.7 min. A linear response over the 3.0–120 mg L−1 concentration range was investigated for compounds. A dilution of the sample (water:urine) was the only step necessary before the electrophoresis analysis. Detection limits of 0.7 and 1.3 mg L−1 for creatinine and creatine (S/N = 3) were obtained. The developed method is easy, rapid and sensitive and has been applied to determine creatinine and creatine in different human urine samples. © 2004 Elsevier B.V. All rights reserved. Keywords: CZE; Creatinine; Creatine; Human urine
1. Introduction Diabetic nephropathy (DN) is a serious complication of diabetes mellitus [1]. However, when DN was diagnosed by the classical methods, little could be done to prevent the progressive downhill course to renal failure. It would be more significant, therefore, if DN could be detected at an even earlier stage, so that intervention could reverse the process. Nonprotein nitrogen compounds, including creatinine and creatine could serve as markers for renal function. Creatine is one of the good indicators of the glomerular filtration rate of the kidneys, and creatinine is the major inactive breakdown product of creatine and has no function in the body, it is simply excreted in urine, which is why it is monitored clinically as a marker of kidney function [2]. Creatine is synthesized in the body, and its main dietary source is red meat. Creatine has recently achieved great popularity as an ergogenic aid [3]. Creatine is converted to phosphocreatine in muscle in a reversible reaction with adenosine
∗
Corresponding author. Fax: +34-926-295318. E-mail address: [email protected] (J. Rodr´ıguez).
0003-2670/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.aca.2004.05.058
triphosphate (ATP), facilitated by creatine kinase [4]. When muscle contractions deplete the immediate supply of ATP, the phosphocreatine can rephosphorylate ADP to replenish the supply of ATP. The study of the uptake of creatine as a result of creatine supplementation, a practice increasingly common among bodybuilders and athletes, has lead to a need to measure urinary creatine concentration. The structures of creatine and creatinine are shown in Fig. 1. Numerous methods for the assay of these markers have been reported, such as FIA methods [5,6], by using biosensors [7], high-performance liquid chromatography [8–11] and capillary electrophoresis. The first two types of methods are limited due to the slow speed of analysis, the inability to test for multiple markers at one time, non-specificity for the analyte of interest and the consumption of large amounts of reagents and samples. On the other hand, the HPLC methods have relatively long analysis times the consume large amounts of solvents. However, capillary electrophoresis techniques, in many cases, have higher resolution, greater efficiency and smaller sample volumes [1]. There are a few articles reporting the determination of creatine and creatinine. Some include separations from
54
J. Rodr´ıguez et al. / Analytica Chimica Acta 521 (2004) 53–59
CH3 N
2.2. Apparatus and operating conditions NH
N H
O
Creatinine CH3 HN
O
N
OH
NH2
Creatine Fig. 1. Structures of creatinine and creatine.
other compounds. Yan et al. [12] employ sodium cholate at phosphate buffer pH 7.4 with voltage of 30 kV in the determination of urea, uric acid, creatine and creatinine. Burke et al. [13] make the determination by MEKC too but using 150 mM SDS in a 30 mM phosphate buffer pH 6, in this method creatinine and creatine were separated in 6 min without interferences. A novel capillary electrophoresis chip-based detection system is described by Wang and Chatrathi et al. [14]. Capillary zone electrophoresis have also been applied [1,15]. Kong et al. proposed a method for determining creatinine, creatine, urea and uric acid. The separation was achieved in 25 mM phosphate buffer pH 3.45; the separation takes less than 15 min and it was applied to urine and plasma samples. In this paper, we propose a very rapid and easy method using capillary zone electrophoresis (CZE) to determine creatine and creatinine in urine without any treatment of the biological sample. The separation take less than 2.5 min. The method is very short and sensitive enough to do this determination in human urine without the use of surfactants. 2. Materials and methods 2.1. Materials All solvents and reagents were of analytical-reagent grade unless indicated otherwise. Creatinine and creatine were obtained from Sigma. Standard solutions were prepared with deionized water (Milli-Q quality) and stored at 4 ◦ C. Diluted solutions of urine were prepared daily by diluting freshly collected human urine with purified water. One milliliter of urine was 10 times diluted using purified water in a calibrated flask. Triethanolamine solution was obtained from Fluka, Biochemika. Buffer solutions were prepared by dissolving an the adequate quantity of H3 PO4 in deionized water and then adjusting with NaOH solution to the required pH. The set of separation vials was changed after six runs.
A Beckman P/ACE System MDQ (Fullerton, CA, USA) equipped with a diode-array detector was used. Beckman capillary electrophoresis software controlled the system. Separation was carried out on a 30 cm long (10 cm to the detector, short way) × 75 m ID fused-silica capillary housed in a cartridge with a detector window 100 m × 800 m. The wavelength selected for the electropherograms was 205 nm. The capillary was conditioned prior to its use by rising with 0.5 M NaOH for 20 min, water for 10 min and finally phosphate buffer. As the human urine samples contain creatinine and creatine, in each case the urine was analyzed after a dilution 1:10 (v/v) with Milli-Q grade water. The capillary was flushed between injections for 2 min with NaOH 0.1 min and 4 min with electrolyte. Electrophoretic separation was performed using as running buffer a 60 mM phosphoric buffer solution at pH 2.2, an hydrodynamic injection time of 3 s (0.5 psi) and a voltage of −15 kV (reversed polarity mode). Under the selected conditions the current was around −108 A. The electropherograms were recorded at 205 nm. Duplicate injections of the solutions were performed and average-corrected peak areas (CPA) (area/migration time) were used for quantitative analysis. The data generated from the first two injections of a sequence were not used on account of the necessary equilibration of the system.
3. Results and discussion The effect of pH of the running electrolyte had a significant impact on the ionization of the acidic silanols of the capillary wall and the electrophoretic mobility of the compounds studied. The buffer tested was phosphate (60 mM) and it was found that at pH 2 the migration times are shorter but the creatinine peak is overlapped with an unknown second peak. On the other hand at pH 3 the electrophoretic mobility decreases and the migration time for creatinine was 4.5 min. For this reason, after analyzing the effect of pH in this interval range a pH 2.2 was selected. At these low pH values our analytes are found before the migration time of EOF; this fact shows that creatinine and creatine are cationic forms. The effect of the composition of the separation electrolyte was studied by using the same pH value with the same ionic strength (50 mM) but different media. Besides phosphate buffer, an acetate and perchlorate buffer were tested but the highest signals and best peak shapes were found using phosphate buffer. In order to try to obtain symmetric and sharper peaks, a phosphate buffer (60 mM) was prepared, the pH was adjusted to 2.2 by addition of TEA (triethanolamine), and the separation was made as previously by applying −15 kV using a hydrodynamic injection of 3 s (0.5 psi), but no better
J. Rodr´ıguez et al. / Analytica Chimica Acta 521 (2004) 53–59
with TEA
55
without TEA
AU
Creatine
Creatinine
40000
20000
0
0.0
1.0
2.0
3.0
Migration Time(min) Fig. 2. Influence of the addition of triethanolamine (TEA). Operating conditions: 60 mM of phosphate buffer, −15 kV of voltage, 3 s injection time (0.5 psi).
behavior by addition of TEA of the peaks of creatinine and creatine was found, as it is possible to see in Fig. 2. The effect of the concentration of phosphate buffer solution (20–70 mM) on the migration time of the compounds
and the shape of the peaks was studied (Fig. 3). When the concentration of buffer increases the migration time decreases and the peaks become sharper. More than 60 mM was not used due to the high current density (to 70 mM of
40000
30000
20 mM
10000
40 mM
60 mM
AU
20000
0
0.0
1.0
2.0
3.0
Migration Time (min) Fig. 3. Influence of phosphate buffer concentration over a urine sample after the dilution process. Operating conditions: 60 mM of phosphate buffer, −15 kV of voltage, 3 s injection time (0.5 psi). Migration order: creatinine, unknown compound and creatine.
56
J. Rodr´ıguez et al. / Analytica Chimica Acta 521 (2004) 53–59 60000
AU
40000
___ 3s - - - 10s 20000
___ 1.8s - - - 5s
0
0.0
1.0
2.0
3.0
Migration Time (min) Fig. 4. Influence of amount of sample injected (injection time and pressure apply) over a urine sample after the dilution process. Operating conditions: 60 mM of phosphate buffer, −15 kV of voltage, 3 s injection time (0.5 psi). Migration order: creatinine, unknown compound and creatine.
phosphate buffer concentration correspond −124 A). The buffer concentration of 60 mM was thus considered to be optimal. Sample components from the diluted urine can become adsorbed onto the capillary surface and change the effective charge on the wall. So, in order to prevent this effect the capillary is flushed between injections for 2 min with NaOH 0.1 M and then 4 min with separation electrolyte in order to obtain an stable EOF (electroosmotic flow). Minor times of the washing step give worse electropherograms. In order to decrease the detection limits in urine, the injection time was varied between 0.6 and 10 s at 0.5 psi (1 psi = 6894.76 Pa), and it was found that up to 0.5 psi and 3 s the peak became higher, but if we continued increasing the injected sample quantity the electrophoretic peaks became double (as it is possible to observe in Fig. 4). Then it was selected as optimum an injection condition of 0.5 psi during 3 s. The effect of varying the voltage from −5 to −20 kV was investigated (Fig. 5). A potential of −15 kV yielded the best compromise in terms of run time, current generated and efficiency of separation. As consequence, this potential was used in subsequent stages of method development. Selected conditions. From the studies carried out before, the procedure summarized below was convenient to separate creatinine and creatine in diluted urine samples: • separation electrolyte: 60 mM phosphate buffer pH 2.2; • voltage: −15 kV (reverse polarity mode), current: −108 A;
• capillary: fused-silica 30 cm long (10 cm to the detector, short way) × 75 m ID; • injection: hydrodynamic, 3 s (0.5 psi) (reverse polarity mode); • detection wavelength: 205 nm. Under these conditions, the migration times were 1.0 and 2.4 min for creatinine and creatine, respectively. In all cases the urine samples were diluted with deionized water (a dilution of urine:water (1:10) was the only step necessary before the electrophoresis analysis). 3.1. Quantitative aspects 3.1.1. Precision Two different urine samples (from a middle-aged woman and a pregnant woman) were analyzed (eight repetition of each one). The precision of the proposed method is expressed in terms of relative standard deviation (R.S.D.). The reproducibility was evaluated between-day using the urine from the middle-aged woman. The results showed that the repeatability for every component in each day is satisfactory. In terms of reproducibility, the comparison of the standard deviations with the Fisher test did not provide any significant difference between both days series, for a significance level of 0.05. The repeatability of the migration times and corrected peak area were good in both cases, with a %R.S.D. less than 2% in all cases (Table 1).
J. Rodr´ıguez et al. / Analytica Chimica Acta 521 (2004) 53–59
57
40000
15 kV 10 kV 5 kV 30000
AU
20000
10000
0
0.0
2.0
4.0
6.0
8.0
Migration Time(min) Fig. 5. Influence of separation voltage on migration times and resolution, over a urine sample after the dilution process. Operating conditions: 60 mM of phosphate buffer, −15 kV of voltage, 3 s injection time (0.5 psi). Migration order: creatinine, unknown compound and creatine.
3.2. Specificity
3.3. Limits of detection and quantification
Specificity can be determined by measurement of peak homogeneity. Because of the different techniques available in a DAD (diode-array detector) are not equally effective for the detection of possible impurities or interferences in an electrophoretic peak, the use of several techniques is recommended [16]. In this work the techniques used to validate the peak purity of the studied compounds present in urine samples were [17]:
Limits of detection (LODs) and quantification (LOQs) were estimated in the usual way. The LOD was obtained as the drug concentration that caused a peak height 3 times the baseline noise level and the LOQ was calculated as 10 times the baseline noise level [18]. The LODs were 0.7 mg L−1 for creatinine and 1.3 mg L−1 for creatine, and the LOQs were 2.3 mg L−1 for creatinine and 4.3 mg L−1 for creatine.
• normalization and comparison of spectra from different peak sections; • absorbance at two wavelengths. Both techniques showed that the purity of the peaks corresponding to the compounds studied in urine present a high level of purity. Therefore, no interference of other urine compounds were observed. Table 1 Repeatability
All results were obtained by using CPA (corrected peak areas) for calculations (to obtain CPA, peak area was divided by its corresponding migration time). A calibration graph was obtained in water by using six spiked water samples. It was tested over the range from 3 to 120 mg L−1 for each analyte. The linear regression equations obtained using the least-squares method and coefficients of correlation are: Creatinine : CPA = (−497.9 ± 902.1) + (992.2 ± 23.2)c,
R.S.D. (%)
r 2 = 0.995,
Creatinine
Migration time Corrected area
3.4. Linearity range and calibration curves
Creatine
creatine :
Middle-aged woman
Pregnant woman
Middle-aged woman
Pregnant woman
0.18 1.97
0.21 0.67
0.97 1.53
0.65 0.55
CPA = (951.7 ± 535.4) + (981.4 ± 17.4)c,
r = 0.998, 2
CPA = (a ± Sa ) + (b ± Sb )c
where a is the intercept, Sa the standard deviation of intercept, b the slope, Sb the standard deviation of slope.
58
J. Rodr´ıguez et al. / Analytica Chimica Acta 521 (2004) 53–59 60000
Creatinine
Creatine
AU
40000
20000
0
0.0
1.0
2.0
3.0
Migration Time (min) Fig. 6. CZE electropherogram of creatinine and creatine in urine sample with the final selected conditions. Migration order: creatinine, unknown compound and creatine.
After that, and with the object to know if it is possible to determine the concentration of creatinine and creatine in an urine sample by using directly this calibration graph obtaining with the water samples spiked or if it is necessary to apply the standard addition method to each urine sample, a mixture of several human urine samples was analyzed by standard addition. The slopes of both graphs (aqueous medium and standard addition over a diluted urine samples) were compared. First of all the variances of both data groups were compared and it was found no significant differences between them. Then it was possible to compare the slopes and it was found no significant differences between the slopes of both linear graphs (aqueous and urine medium). A significance level of 0.05 was used. Thus it is possible to determine creatinine and creatine in human urine sample by applying directly the calibration graph in water.
In order to compare the obtained results by the use of the calibration graph, the addition standard method was also applied over three of the urine samples analyzed, the results obtained have totally satisfactory. An electropherogram of an urine sample (from the middle aged women) is presented in Fig. 6. Urine creatinine levels can be used as a screening test to evaluate kidney function. Normal values are highly dependent on the age and lean body mass of the person the urine is being collected from. Urine creatinine values may therefore be quite variable and can range from 300 to 1300 mg L−1 , as can be seen in Table 2 except to the sportive young man,
Table 2 Determination of creatinine and creatine in different urine samples Type of urine sample
3.5. Applications In order to apply the proposed method, six different types of urine samples (middle aged woman, middle aged man, sportive young man, boy, girl and pregnant woman) were analyzed. Two samples of each type of urine were analysed in duplicate. The concentration of creatinine and creatine were determine after dilution with water and the values obtained for each original urine samples are presented in Table 2.
Girl Boy Sportive young man (with urea problems)a Middle aged woman Middle aged man Pregnant woman
Concentration in original urine sample (mg L−1 ) Creatinine
Creatine
896 1179 2042
397 392 Not detected
614 554 1050
890 75 406
a For this sample, a dilution factor of 50 was necessary (1.0 mL of urine was 50 times diluted with purified water).
J. Rodr´ıguez et al. / Analytica Chimica Acta 521 (2004) 53–59
due to the exercise may caused increased creatinine clearance and also his high urea level. (For this analysis a 25 mL dilution factor was applied.) Also the glomerular filtration rate is substantially increased in pregnant women (see pregnant woman in the mentioned table). Creatine levels are of the same order of magnitude as creatinine levels only when subjects have recently ingested creatine, while somewhat elevated urinary creatine concentrations in non-supplementing subjects can be an indication of a degenerative disease of the muscle (see the high level of creatine in middle aged woman in the table, may be due to her main dietary source is red meat).
4. Conclusions In this work, a fast and direct capillary zone electrophoresis method for the determination of creatinine and creatine in human urine sample is described. The proposed method is faster than those previously proposed. The parameters calculated for this method are satisfactory.
Acknowledgements The authors are grateful to the DGES of the Ministerio de Educación y Ciencia (project BQU 2001-1190) and to the Junta de Extremadura for the financial support.
59
References [1] Y. Kong, N. Zheng, Z. Zhang, R. Gao, J. Chromatogr. A 987 (2003) 477–483. [2] A.D. Baines, Disorders of the kidney and urinary tract, in: A.G. Gornall (Ed.), Applied Biochemistry of Clinical Disorders, 2nd ed., JB Lippincott, Philadelphia, PA, 1986, pp. 139–171. [3] W.J. Kramer, J.S. Volek, Clin. Sports Med. 18 (1999) 651. [4] G. Zubay, Biochemistry, Wm.C. Brown, Dubuque, IA, 1998, p. 39. [5] T. Yao, K. Kotegawa, Anal. Chim. Acta 462 (2) (2002) 283–291. [6] G. Del Campo, A. Irastorza, J. Alfonso-Casado, Fresen. J. Anal. Chem. 352 (6) (1995) 557–561. [7] R.J. Stefan, R.G. Bokretsion, J.F. Van-Staden, H.Y. Aboul-Enein, Talanta 60 (6) (2003) 1223–1228. [8] T. Seki, Y. Orita, K. Yamaji, A. Shinoda, J. Pharm. Biomed. Anal. 15 (1997) 1621–1626. [9] P. Ellerbe, A. Cohen, M.J. Weleh, V.E. White, Anal. Chem. 62 (1990) 2173–2177. [10] G. Werner, V. Schneider, J. Emmet, J. Chromatogr. 525 (1990) 265– 275. [11] S.D. Yang, J.M. Xu, L. Yang, Y.N. Ma, F. Bai, J. Liq. Chromatogr. 12 (1989) 1791–1807. [12] S.L. Yan, P.Z. Lin, M.W. Hsiao, J. Chromatogr. Sci. 37 (2) (1999) 45–50. [13] D.G. Burke, P.G. MacLean, R.A. Walker, P.J. Dewar, T. Smith-Palmer, J. Chromatogr. B 732 (1999) 479–485. [14] J. Wang, M.P. Chatrathi, Anal. Chem. 75 (3) (2003) 525–529. [15] E.A. Clark, J.C. Fanguy, C.S. Henry, J. Pharm. Biomed. Anal. 25 (2001) 795–801. [16] H. Farbe, A. LeBris, M.D.J. Blanchin, J. Chromatogr. A 697 (1–2) (1995) 81–88. [17] L. Huber, Applications of Diode-array Experience, vol. 12, Hewlett-Packard Publication, 1989, p. 5953 [18] K.D. Altria, Chromatographia 35 (3–4) (1993) 177–182.