1962 Anna Makrl´ıkova´ 1 ˇ Frantisek Opekar1 2 ˚ Petr Tuma 1 Department
of Analytical Chemistry, Faculty of Science, Charles University in Prague, Prague, Czech Republic 2 Institute of Biochemistry Cell and Molecular Biology, Third Faculty of Medicine, Charles University in Prague, Prague, Czech Republic
Received December 30, 2014 Revised February 10, 2015 Accepted February 10, 2015
Electrophoresis 2015, 36, 1962–1968
Research Article
Pressure-assisted introduction of urine samples into a short capillary for electrophoretic separation with contactless conductivity and UV spectrometry detection A computer-controlled hydrodynamic sample introduction method has been proposed for short-capillary electrophoresis. In the method, the BGE flushes sample from the loop of a six-way sampling valve and is carried to the injection end of the capillary. A short pressure impulse is generated in the electrolyte stream at the time when the sample zone is at the capillary, leading to injection of the sample into the capillary. Then the electrolyte flow is stopped and the separation voltage is turned on. This way of sample introduction does not involve movement of the capillary and both of its ends remain constantly in the solution during both sample injection and separation. The amount of sample introduced to the capillary is controlled by the duration of the pressure pulse. The new sample introduction method was tested in the determination of ammonia, creatinine, uric acid, and hippuric acid in human urine. The determination was performed in a capillary with an overall length of 10.5 cm, in two BGEs with compositions 50 mM MES + 5 mM NaOH (pH 5.1) and 1 M acetic acid + 1.5 mM crown ether 18-crown-6 (pH 2.4). A dual contactless conductivity/UV spectrometric detector was used for the detection. Keywords: Creatinine / Human urine / Hydrodynamic sampling / Short capillary separation / Uric acid DOI 10.1002/elps.201400613
1 Introduction Because of its high separation efficiency, short separation time, and minimal requirements on the amounts of samples and reagents, CE is an ideal technique for the separation of complicated clinical samples [1–8]. In addition, when dual detection [9,10] based on the principle of capacitively coupled contactless conductivity (C4 D) [11,12] and UV spectrometry is employed, a wide range of application possibilities are available for use in bioanalysis without the necessity of performing derivatization of the sample; C4 D can be used to sensitively detect especially rapidly migrating small organic and inorganic ions and, with UV detection particularly of substances with aromatic structures. This technical approach is instrumentally simpler and cheaper than combination of CE with MS [13, 14] and can be constructed to order in the laboratory for a particular application.
˚ Correspondence: Dr. Petr Tuma, Institute of Biochemistry, Cell and Molecular Biology, Third Faculty of Medicine, Charles University in Prague, Ruska´ 87, 100 00 Prague 10, Czech Republic E-mail: [email protected] Fax: +420-267-102-460
Abbreviations: HAc, acetic acid; HV, high voltage
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Standard (commercially available) CE apparatuses employ capillaries with a length of few to several tens of centimeters and the separation time is usually 5 to 30 min [15]. When dealing with some analytical problems, for example in monitoring the kinetics of biochemical reactions [16] or as a subsequent separation method in multidimensional separation systems, it is necessary that the separation time be much shorter. Similarly, in the normal clinical monitoring of analytes in large sets of samples, even a small increase in the rate of separation contributes to shortening the analysis time, affecting the number of samples that can be analyzed in unit time and thus the price of the analysis. Shortening of the separation pathway is an effective technique for increasing the rate of separation. One of the options involves performance of electrophoresis on a glass or plastic microchip, where the length of the separation pathway is of the order of units of centimeters and the analysis time ranges from units to tens of seconds [17–19]. In practice, this means of separation has a disadvantage in the necessity of using a special separation system - an electrophoretic chip. In addition, contactless conductivity and UV spectrometry detection combined with a microchip do not achieve high sensitivity and the application potential required for clinical systems is limited [20,21]. From this point of view, the use of standard electrophoretic capillaries seems preferable, where their length, inner diameter, and material can be easily selected as required; their generally www.electrophoresis-journal.com
CE and CEC
Electrophoresis 2015, 36, 1962–1968
easy availability and substantially lower price are also important. In addition, the separation efficiency in channels with circular capillary cross-section is higher than in capillaries with rectangular cross-sections, familiar for microchips [22]. Short-end injection is a technically simple approach enabling shortening of the separation pathway, where the sample is injected into a site close to the detector; commercial instrumentation can frequently be used for this purpose [23]. This technique can be used to shorten the effective length of the capillary to approximately 8 to 10 cm with maintenance of minimal overall length of the capillary of about 30 cm, which is necessary when using commercial CE apparatuses. However, greater experimental variability can be attained using special apparatuses for separation directly in short capillaries with overall length of about 10 cm. A basic experimental problem when using separation in short capillaries lies in their limited mechanical flexibility. A short capillary cannot be handled in the same way as the capillaries used in standard electrophoretic systems. Consequently, the apparatus must be designed so that all the necessary experimental steps, especially sample injection and flushing of the capillary, can be performed without it being necessary to move the capillary. Injection of the sample into a separation capillary of normal length and especially into a short capillary is generally a critical point in achieving high-performance electrophoretic separation. A suitable injection strategy in separation in short capillaries permits elimination of one of the sources of uncertainty in the standard means of injection into capillaries of classical length, consisting in the repeated movement of the capillary injection end between sample and electrolyte solution; the capillary is thus physically disrupted from the solution. A detailed survey of various approaches to this method of separation can be found, for example, in the reviews [15,24,25]; a similar subject is discussed in relation to separation on an electrophoretic chip [26]. Another special injection technique is based on sucking the sample into the horizontally fixed capillary by contraction of the BGE inside the capillary that is achieved by lowering the temperature of the liquid surrounding the capillary [27]. Recently, rapidly developing fast separation systems using ESI-MS detection call on designing novel strategies and apparatus for exact sampling. The sampling systems are often complex and encompass many components including a micromanipulator controlling precise position of sampling and separation capillaries, and require multistep sampling protocol [28]. A special sampling microchip with integrated microvalves controlled pneumatically by nitrogen pressure [29] and a porous tip at the terminus of the capillary introduced for sheathless interfacing of CE to MS and for sample stacking during injection [30] are also described in literature. For less-demanding CE experimental arrangements, an easily available option lies in sample injection based on the principle described in the work [31]. The injection end of the capillary is loosely inserted into a tube through which the BGE is flowing during the injection, into which the sample is injected from the loop of a six-way valve. The sample can be injected into the capillary during the time that the sample C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1963
zone flows around the injection end of the capillary. Thus the injection end of the capillary is constantly submerged in the solution. This principle was utilized in the electrokinetic injection of samples of energy beverages [32, 33]. However, in electrokinetic sample injection, the amount of injected substance in the sample depends on its electrophoretic mobility, that is more ions with higher mobility than ions with lower mobility are added during the injection [34]. In addition, the amount of electrokinetically injected analyte is highly dependent on the total conductivity and composition of the sample matrix and, in clinical analyses, these factors depend on the content of majority sodium and chloride ions and on the preparation of the sample by dilution or the use of organic solvents. The above problems can be resolved by the use of pressure-assisted injection described in this work, where it is not necessary to move the capillary and its injection end is constantly immersed in a solution.
2 Materials and methods 2.1 Apparatus The injection end of the electrophoretic apparatus, Fig. 1A, consists of commercially available plastic components used for connecting tubes of small dimensions (1, 2). The electrophoretic ground electrode (3) is a 1-cm-long piece of stainless-steel tube; plastic components (1) and (2) fix the electrode insulated using hot melt adhesive in a stable position. The separation capillary (4) is tightly fitted (under heat) into the auxiliary PTFE tube, ensuring its fixed position in the injection part. The injection end of the capillary is inserted to a depth of approximately 1 mm into the PTFE tube (5) through which the BGE solution with the sample flows during the injection. The shut-off valve (10) at the exit from the injection part of the apparatus closes the exit PTFE tube (11). This creates a pressure pulse in the apparatus, injecting the sample into the capillary. The hydrodynamic resistance (13) created from a 5-mm-long piece of capillary 75 m id causes a sharper increase in the pressure in the apparatus when the shut-off valve is closed. The separation was performed in a standard fused-silica capillary (Composite Metal Services, UK), 50 m id, 363 m od, with length total/to the detector 10.5/8 cm. The electrophoretic apparatus was laboratory made and is described in detail in a previous publication [33]. The separation capillary leads from the injection part to the dual C4 D/UV detector and to the end vessel with the high-voltage (HV) electrode of the HV source. Figure 1B depicts the time sequence of the injection. The injection is commenced by turning on the linear pump for the BGE and simultaneously switching the six-way valve (6) from the “load" position (filling the injection loop) to the “inject" position. Following a certain time, necessary for the sample zone in the BGE stream to reach the injection end of the capillary, the shut-off valve is activated for a defined time and the thus-generated pressure pulse injects the sample into the capillary. The linear pump is turned on for long enough www.electrophoresis-journal.com
1964
A. Makrl´ıkova´ et al.
Electrophoresis 2015, 36, 1962–1968
11
2 10
1
2
3
3
9 6
7
B
1
5 13
4
8
5
4
time
12
A
Figure 1. Scheme of the injection part of the electrophoretic apparatus (A) and scheme of the time dependence of the injection sequence (B). (A) 1, 2—barbed tubing fittings, 3/32 id Tee and 3/32 id Elbow (World Precision Instruments); 3—stainless-steel tube (ground electrode of the HV source) with soldered connection, insulated with hot melt adhesive; 4—separation capillary inserted in 1/16 od × 0.01 id PTFE tube; 5—PTFE tube 1/16 od × 0.031 id; 6—six-way sampling valve (Valco Instruments); 7— injection loop with volume of 40 L; 8—filling the injection loop; 9—BGE feed from the stock reservoir pumped by a linear pump ˇ (Labio, CR); 10—shut-off valve (Cole Parmer); 11—waste; 12—to the detector and end vessel; 13—5-mm-long piece of capillary 75 m id. (B) 1—initiation of injection; 2—time of operation of the linear pump; 3—delay of the shut-off valve; 4—time of activation of the shut-off valve, that is time of sample injection into the capillary; 5—turning on the HV source and commencement of the separation.
for the sample zone to reach a sufficient distance from the injection end of the capillary. After turning off the pump, the HV source is turned on and the separation is commenced. The actions of the individual parts of the apparatus were controlled through an interface by a program created in LabView (National Instruments).
2.2 Chemicals and tested materials All the chemicals used were of p.a. purity: acetic acid (HAc; Sigma, Steinheim, Germany), Ca(NO3 )2 (Lachema, Brno, CR), creatinine (Fluka, Buchs, Germany), crown ether 18-crown-6 (Fluka, Buchs, Germany), hippuric acid sodium salt (Aldrich, Milwaukee, WI, USA), histidine (His, Aldrich, Steinheim, Germany), KCl (Lachema, Brno, CR), 2-morpholinoethane sulfonic acid (MES, Sigma, Buchs, Hungary), NaOH (Fluka, Steinheim, Switzerland), (NH4 )2 SO4 (Fluka, Buchs, Germany), and uric acid (Fluka, Buchs, Switzerland). The parameters of the pressure-assisted injection were optimized using BGE 20 mM MES + 5 mM NaOH (pH 5.8), the test sample was an aqueous solution with equimolar concentration of 50 M K+ and His (i.e. a mixture formed by a small inorganic and larger organic ions). A separation voltage of 5 kV was used in all the analyses and the experiments were performed at laboratory temperature. Some of the components in human urine were determined to illustrate the practical applicability of the proposed C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
pressure-assisted injection. Urine samples obtained from a healthy volunteer were analyzed on the day of sampling. Prior to injection into the capillary, the urine was only filtered using a single-use injection-needle PVDF filter (Fisher Scientific, CR) with a pore size of 0.45 m and then diluted 50 or 100 times with deionized water. The separation of the urine was performed in two BGEs, 50 mM MES + 5 mM NaOH (pH 5.1) and 1 M HAc + 1.5 mM 18-crown-6 (pH 2.4). The detection was performed using both detection systems, laboratory-made C4 D operated with a sine-wave signal with a frequency of 450 kHz and an amplitude of 17 V (peak to peak) and UV operated at 214 nm.
3 Results and discussion 3.1 Optimization of the pressure-assisted injection parameters Optimization of the injection consisted of search for the time intervals of the individual steps of the injection sequence, see Fig. 1B (for illustration, only the results obtained for the potassium ion are presented, where not stated otherwise; similar results were obtained for His): (1) “Time of activation of the linear pump” so that the sample zone carried by the BGE stream reaches a sufficient distance from the injection end of the capillary, that is to the ground electrophoretic electrode, time (2). The following optimum pump parameters were determined experimentally (by monitoring the movement of colored zones in the sample): flow rate, 1 mL/min and activation time 10 s; under these conditions, the BGE consumption for one sample injection was 170 L. These parameters were used in subsequent measurements. (2) “Time of delayed activation of the shut-off valve” after commencing the injection sequence so that the sample zone passes around the injection end of the capillary, time (3). It can be seen from Fig. 2A that the greatest amount of substance is injected at a time 4 to 6 s after commencing the injection sequence, when the central part of the sample zone passes around the injection end of the capillary. A time of 5 s was chosen for the subsequent experiments. (3) “Time of activation of the shut-off valve,” that is, the time during which the pressure is increased in the injection system so that a suitable length of the sample zone is injected into the capillary, time (4). It can be seen from the peak area and the separation efficiency expressed as the number of theoretical plates on the length of the pressure pulse, Fig. 2B, that, under the given experimental conditions, the optimum duration of the pressure pulse is 0.5 s. A sufficient amount of sample is injected for problemfree detection without a reduction in the separation efficiency. Figure 3 depicts an illustrative electropherogram of the test mixture recorded under the optimized conditions by C4 D. www.electrophoresis-journal.com
CE and CEC
Electrophoresis 2015, 36, 1962–1968
1965
Figure 2. The effect of the time of delay of the activation pulse (A) and the duration of the activation pulse (B) of the shut-off valve on the peak area and separation efficiency for the potassium ion. Common injection parameters: time of activation of the linear pump; 10 s, BGE flow rate, 1 mL/min. Duration of the activation pulse in dependence (A), 0.5 s; time of delay of the activation pulse in dependence (B), 5 s.
Figure 3. Electropherogram of an equimolar mixture of 50 M potassium ion (1) and His (2) recorded under optimized conditions of the pressure-assisted injection. Buffer, 20 mM MES + 5 mM NaOH (pH 5.8), 5 kV/3 A. Injection parameters: time of activation of the linear pump, 10 s; BGE flow rate, 1 mL/min; delay and duration of the activation pulse of the shut-off valve, 5 and 0.5 s, respectively.
Figure 4. Time dependence of the pressure in the injection part of the apparatus during activation of the shut-off valve for a period of 0.5 s; BGE flow rate, 1 mL/min.
The pressure in the apparatus was measured with activated shut-off valve. For illustration, Fig. 4 depicts the time dependence of the pressure for duration of the activation pulse of 0.5 s. The volume of the sample injected into the capillary was estimated using the equation
Sampling pulse length (s)
Average pressure (mbar)
Volume sampled (nL)
Zone length (mm)
Analyte amount (pmol)
0.3 0.4 0.5 0.6 0.7 0.8
45.7 57.7 80.4 103.4 128.6 149.5
2.0 3.4 5.9 9.0 13.2 17.5
1.0 1.7 3.0 4.6 6.7 8.9
0.10 0.17 0.29 0.45 0.66 0.87
V = 7.81 × 10−11
⌬ P D4 t , L
employing the average pressure during the pulse. For the given numerical coefficient, V is the injected volume in nanoliters and the other parameters are ⌬ P (mbar) pressure, D (50 m) inner capillary diameter, t (s) injection time, (0.001 Pa·s) solution viscosity, and L (10.5 cm) total capillary length. The injected volume, length of the sample zone, and injected amount of analyte are given in Table 1 for various durations of the activation pulse. C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Table 1. Injected sample volume, length of the sample zone in the capillary, and injected amount of analyte (for a concentration of 50 M K+ ) for various durations of the injection pulse
It is possible to very simply control the amount of sample injected into the capillary by selecting the duration of the pulse activating the shut-off valve. It is necessary to use shorter times for greater demands on the separation efficiency; if the demands on the separation are not great, the www.electrophoresis-journal.com
1966
A. Makrl´ıkova´ et al.
Electrophoresis 2015, 36, 1962–1968
Figure 5. Part of the electropherograms of human urine recorded by C4 D and the UV detector. (A) Urine diluted with water 1:50, BGE 50 mM MES + 5 mM NaOH, separation voltage/current 5 kV/3 A. Identification of peaks: unseparated inorganic ions (1), His (2), creatinine (3), neutral substances absorbing UV radiation (4), uric acid (5), and hippuric acid (6). (B) Urine diluted with water 1:100, BGE 1 M HAc + crown ether 18-crown-6, separation voltage 5 kV/13 A. Identification of peaks: NH4 + (1), K+ (2), Ca2+ (3), Na+ (4), creatinine (5), and His (6).
Table 2. Parameters of the linear regression equation of the calibration dependences of the area of the peaks for the potassium ion and His in the concentration range 10 to 150 M
Analyte
K+ (C4 D)
His (C4 D)
Intercept (mV) Slope (mV.s.M−1 ) Coefficient of determination (R2 ) Standard error (mV.s) LOD (M)
0a) 0.073 (0.003) 0.993 0.350 14
0.728 (0.363) 0.358 (0.004) 0.999 0.511 4
Buffer, 20 mM MES + 5 mM NaOH (pH 5.8), 5 kV/3 A. For the injection parameters, see Fig. 3; SDs in parenthesis. a) The zero value of the intercept lies inside the reliability interval.
detected signal can be increased by increasing the injection time. The repeatability of the injection was determined from the areas of the peaks of ten consecutive measurements. RSD determined for 50 M K+ and 50 M His were 4.0 and 1.7%, respectively, and for 10 M K+ RSD equaled 7.4%. The repeatability given in the literature expressed as RSD is 5–10% for manual injection in laboratory systems and 1–3% for automated injection in commercial instruments [35]. The repeatability of the tested injection pressure is somewhere between these values. The dependence of the peak areas on the concentration was linear in the tested concentration interval for the potassium ion and His; the parameters of the calibration curves are given in Table 2. The LOD was calculated from the parameters of the calibration equation, LOD = 3 × Standard error/Slope. LOD for the potassium ion determined by this method is high; the coion in the buffer employed, the Na+ ion, is suitable for sensitive conductivity detection of His but not for detection of the K+ ion. For comparison, the same test sample was injected electrokinetically, that is the shut-off valve was not activated during the injection and the separation voltage was permanently turned on. The ratio of the slopes His/K+ of the calibration dependences in pressure-assisted injection was 4.9 and, for C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
electrokinetic injection, 1.9. The discrimination of large ions in electrokinetic injection is evident.
3.2 Separation of ammonium ions, histidine, creatinine, uric acid, and hippuric acid in human urine The determination of important biochemical analytes in human urine, ammonium ions, creatinine, His, and uric and hippuric acids [36,37] was selected for verification of the practical applicability of the tested injection method. The ammonia content is an indicator of disorders in urea synthesis and elevated protein catabolism. The amount of produced and excreted creatinine is used in correction of diuresis. An elevated uric acid level indicates a danger of the formation of bladder and kidney stones. Hippuric acid is the final product in the biotransformation of benzoic acid, toluene, and other substances containing a benzene ring. His is the main amino acid in urine and elevated excretion is connected with a congenital metabolic disorder. Separation of these substances was first tested in BGE with composition 50 mM MES + 5 mM NaOH, pH 5.1; the pH is lower than that for the separation model mixture of K+ and His because of the need to protonate creatinine and separate it from EOF (see Fig. 5A). In this BGE, EOF marker flows through the capillary in a time of 40 s; before EOF on the C4 D recording can be seen the positive peak of the mixed zone of inorganic ions, followed by the negative peaks of the organic cations His and creatinine. The EOF position in the UV recording is followed by the separated peaks of uric and hippuric acids, which migrate as anions in the opposite direction to EOF. Addition of 18-crown-6, which complexes K+ , to the BGE led to only partial separation of the NH4 + and K+ ions; at rapid EOF combined with a short capillary, there is not sufficient time for complete separation of the two ions with similar mobility values. For this reason, acidic BGE with composition 1 M HAc + 18-crown-6 (pH 2.4), in which EOF is greatly suppressed, was tested for the separation cations in urine. Under these conditions, the motion of the cations through the capillary is slowed and they move on the basis of their own electrophoretic www.electrophoresis-journal.com
CE and CEC
Electrophoresis 2015, 36, 1962–1968
1967
Table 3. Determination of creatinine, His, uric acid, and hippuric acid (detection method in parentheses)
Analyte
Migration time (s)
Concentration (mg/L)
His (C4 D) His (UV) Creatinine (C4 D) Creatinine (UV) Uric acid (UV) Hippuric acid (UV)
32 ± 1
67 61 1031 1209 392 539
36 ± 1 54 ± 3 69 ± 5
± ± ± ± ± ±
7 6 233 364 20 33
RSD (%)
S/Na)
5.1 4.6 10.2 13.7 2.3 2.7
10 5 160 75 34 30
Nb) (m−1 )
Range [39] (mg/L) 40–330
136 200 52 300 24 900 118 000 150 000
670–2150 40–670 50–1670
Sample: human urine/water 1:50, BGE 50 mM MES + 5 mM NaOH, separation voltage/current 5 kV/3 A. a) S/N, detector signal (peak height)/noise of baseline. b) The number of theoretical plates was calculated by the formula N = 5.54 (tM /w1/2 )2 , where tM is the migration time and w1/2 is the peak width at half-height.
mobility. This leads to complete separation of all the studied cations, yielding negative peaks in C4 D and, in addition, creatinine can be detected in the UV detector (see Fig. 5B). However, the acids cannot be determined in this acidic BGE. The obtained separation efficiencies (Tables 3 and 4) commonly attain values of about 100 000 theoretical plates/m; these values are fully comparable with the separation efficiencies for classical hydrodynamic injection of samples connected with exchange of vials. This indicates that the newly developed injection method ensures injection of discrete sample zones at the beginning of the capillary. The results of the determination of the monitored components in urine in both BGEs and detection by both detection systems are given in Tables 3 and 4. The standard addition method was used in the determination and the results are the median of three determinations. It can be seen that, within the limits of reliability, C4 D and UV detection yield the same results for those analytes that can be determined by both methods. Creatinine can be determined in both BGEs. However, the determination is more accurate in HAc-crown ether BGE, because the creatinine peak in MES-NaOH BGE is immediately next to the EOF peak, complicating its evaluation. In contrast, in HAc-crown BGE, His is poorly separated from the other components of the urine, probably from its methyl derivatives [38]; consequently, it was determined only in MES-NaOH BGE. The S/N parameter in Table 3 indicates slightly better performance of C4 D over UV detector for both creatinine and histidine. The last column in Table 3 gives the range of concentrations in which this component can be present in human urine [39]; the determined concentrations of all the test components fall within this range.
4 Concluding remarks A simple apparatus is proposed for pressure-assisted injection into a short separation capillary, which is an advantageous alternative to more frequently used electrokinetic injection, previously described in [32, 33]. It is not necessary to move the capillary and its injection end is constantly in contact with the solution. For optimized working conditions, it is possible to obtain injection reproducibility expressed as C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Table 4. Determination of ammonia ions and creatinine (detection method in parentheses)
Analyte
Migration time (s)
Concentration (mg/L)
RSD (%)
N (m−1 )
NH4 + Creatinine (C4 D) Creatinine (UV)
34 ± 0 68 ± 0
515 ± 95 1666 ± 75 1717 ± 159
8.3 2.0 4.2
132 000 95 600 62 700
Sample: human urine/water 1:100, BGE 1 M HAc + 1.5 mM crown ether 18-crown-6, separation voltage 5 kV/13 A.
RSD of about 5%. The equipment facilitates easy changes in the injection conditions by changing a single parameter, the duration of the pressure pulse. The apparatus was tested by determining several components of human urine. Separation on a short pathway has the general advantage of speed. Using the described apparatus, the separation time of the monitored components, ammonium ions, creatinine, His, uric and hippuric acids, was about 70 s. However, attempts to determine all the test substances in a single BGE were not successful and two different separation buffers had to be used. The apparatus can be used for rapid screening of the contents of the other inorganic ions; the resolution parameters for the NH4 + /K+ , K+ /Ca2+ , and Ca2+ /Na+ ion pairs equaled 1.3, 3.4, and 1.3,respectively. This work was supported by Charles University in Prague, Projects SVV, PRVOUK 31, and UNCE 204015/2012. The authors have declared no conflict of interest.
5 References [1] Ramautar, R., Demirci, A., deJong, G. J., Trends Anal. Chem. 2006, 25, 455–466. [2] Jabeen, R., Payne, D., Wiktorowcz, J., Mohammad, A., Petersen, J., Electrophoresis 2006, 27, 2413–2438. [3] Li, S. F. Y., Kricka, L. J., Clin. Chem. 2006, 52, 37–45. [4] Stalmach, A., Albalat, A., Mullen, W., Mischak, H., Electrophoresis 2013, 34, 1452–1464. [5] de Macedo, A. N., Jiwa, M. I. Y., Macri, J., Belostotsky, V., Hill, S., Britz-McKibbin, P., Anal. Chem. 2013, 85, 11112–11120.
www.electrophoresis-journal.com
1968
A. Makrl´ıkova´ et al.
Electrophoresis 2015, 36, 1962–1968
[6] Guihen, E., Electrophoresis 2014, 35, 138–146.
[25] Matysik, F. M., Anal. Bioanal. Chem. 2010, 397, 961–965.
[7] Moser, A. C., Willicott, C. W., Hage, D. S., Electrophoresis 2014, 35, 937–955.
[26] Karlinsey, J. M., Anal. Chim. Acta 2012, 725, 1–13.
[8] Lucangioli, S., Tripodi, V., Curr. Anal. Chem. 2014, 10, 183–183. ˇ ık, K., Electrophoresis 2011, 32, 795–810. [9] Opekar, F., Stul´
[28] Grundmann, M., Matysik, F. M., Anal. Bioanal. Chem. 2012, 404, 1713–1721.
´ M., Preisler, J., Foret, F., Hauser, P. C., Krasen[10] Ryvolova, sky, P., Paull, B., Macka, M., Anal. Chem. 2010, 82, 129–135. ˇ P., Hauser, P. C., Electrophoresis 2013, 34, 55–69. [11] Kuba´ n, ˇ P., Hauser, P. C., Electrophoresis 2011, 32, 30–42. [12] Kuba´ n, ´ ık, K., Electrophoresis 2013, 34, 70–85. [13] Kleparn´ [14] Ramautar, R., Somsen, G. W., deJong, G. J., Electrophoresis 2013, 34, 86–98. ˇ ık, K., Chem. Rev. 2009, 109, [15] Opekar, F., Coufal, P., Stul´ 4487–4499. [16] van Dyck, S., Kaale, E., Novakova, S., Glatz, Z., Hoogmartens, J., vanSchepdael, A., Electrophoresis 2003, 24, 3868–3878.
[27] Hjerten, S., Electrophoresis 1990, 11, 665–690.
[29] Kelly, R. T., Wang, C. C., Rausch, S. J., Lee, C. S., Tang, K. Q., Anal. Chem. 2014, 86, 6723–6729. [30] Moini, M., Martinez, B., Rapid Commun. Mass Spectrom. 2014, 28, 305–310. ˚ [31] Tuma, P., Opekar, F., Jel´ınek, I., J. Chromatogr. A 2000, 883, 223–230. ˇ ık, K., Anal. ´ B., Opekar, F., Tuma, ˚ [32] Vochyanova, P., Stul´ Bioanal. Chem. 2012, 404, 1549–1554. ´ ´ B., Opekar, F., Tuma, ˚ [33] Vochyanov a, P., Electrophoresis 2014, 35, 1660–1665. [34] Lauer, H. H., Rozing, G. P., High Performance Capillary Electrophoresis, A Primer, Agilent Technologies, Germany 2010, pp. 59–72.
[17] Shang, F. J., Guihen, E., Glennon, J. D., Electrophoresis 2012, 33, 105–116.
ˇ P., Seiman, A., Kaljurand, M., Capillary Elec[35] Kuba´ n, trophoresis and Microchip Capillary Electrophoresis, John Wiley & Sons, New Jersey 2013, pp. 41–65.
[18] Guihen, E., O’Connor, W. T., Electrophoresis 2010, 31, 55–64.
[36] Devlin, T. M., Biochemistry with Clinical Correlations, Wiley-Liss, New York 1992, pp. 475–573.
[19] Roman, G. T., Kennedy, R. T., J. Chromatogr. A 2007, 1168, 170–188.
[37] Murray, R. K., Bender, D. A., Botham, K. M., Kennelly, P. J., Rodwell, V. W., Weil, P. A., Harper’s Illustrated Biochemistry, McGraw-Hill Lange, China 2007, pp. 271–306, 331–342. ˇ ık, K., ˚ ´ E., Opekar, F., Jurka, V., Stul´ [38] Tuma, P., Samcova, Electrophoresis 2007, 28, 2174–2180.
ˇ P., Hauser, P. C., Lab Chip 2008, 8, 1829–1836. [20] Kuba´ n, [21] Abad-Villar, E. M., Kuban, P., Hauser, P. C., Electrophoresis 2005, 26, 3609–3614. ˇ B., Kenndler, E., Electrophoresis 2002, 23, 3817– [22] Gas, 3826. [23] Glatz, Z., Electrophoresis 2013, 34, 631–642. ´ M., Preisler, J., Brabazon, D., Macka, M., [24] Ryvolova, Trends Anal. Chem. 2010, 29, 339–353.
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
[39] Putnam, D. F., Composition and concentrative properties of human urine, Report No. NASA CR1802, McDonnell Douglas Astronautic company, Huntington Beach, California, http://ntrs.nasa.gov/ archive/nasa/casi.ntrs.nasa.gov/19710023044.pdf.
www.electrophoresis-journal.com
of Analytical Chemistry, Faculty of Science, Charles University in Prague, Prague, Czech Republic 2 Institute of Biochemistry Cell and Molecular Biology, Third Faculty of Medicine, Charles University in Prague, Prague, Czech Republic
Received December 30, 2014 Revised February 10, 2015 Accepted February 10, 2015
Electrophoresis 2015, 36, 1962–1968
Research Article
Pressure-assisted introduction of urine samples into a short capillary for electrophoretic separation with contactless conductivity and UV spectrometry detection A computer-controlled hydrodynamic sample introduction method has been proposed for short-capillary electrophoresis. In the method, the BGE flushes sample from the loop of a six-way sampling valve and is carried to the injection end of the capillary. A short pressure impulse is generated in the electrolyte stream at the time when the sample zone is at the capillary, leading to injection of the sample into the capillary. Then the electrolyte flow is stopped and the separation voltage is turned on. This way of sample introduction does not involve movement of the capillary and both of its ends remain constantly in the solution during both sample injection and separation. The amount of sample introduced to the capillary is controlled by the duration of the pressure pulse. The new sample introduction method was tested in the determination of ammonia, creatinine, uric acid, and hippuric acid in human urine. The determination was performed in a capillary with an overall length of 10.5 cm, in two BGEs with compositions 50 mM MES + 5 mM NaOH (pH 5.1) and 1 M acetic acid + 1.5 mM crown ether 18-crown-6 (pH 2.4). A dual contactless conductivity/UV spectrometric detector was used for the detection. Keywords: Creatinine / Human urine / Hydrodynamic sampling / Short capillary separation / Uric acid DOI 10.1002/elps.201400613
1 Introduction Because of its high separation efficiency, short separation time, and minimal requirements on the amounts of samples and reagents, CE is an ideal technique for the separation of complicated clinical samples [1–8]. In addition, when dual detection [9,10] based on the principle of capacitively coupled contactless conductivity (C4 D) [11,12] and UV spectrometry is employed, a wide range of application possibilities are available for use in bioanalysis without the necessity of performing derivatization of the sample; C4 D can be used to sensitively detect especially rapidly migrating small organic and inorganic ions and, with UV detection particularly of substances with aromatic structures. This technical approach is instrumentally simpler and cheaper than combination of CE with MS [13, 14] and can be constructed to order in the laboratory for a particular application.
˚ Correspondence: Dr. Petr Tuma, Institute of Biochemistry, Cell and Molecular Biology, Third Faculty of Medicine, Charles University in Prague, Ruska´ 87, 100 00 Prague 10, Czech Republic E-mail: [email protected] Fax: +420-267-102-460
Abbreviations: HAc, acetic acid; HV, high voltage
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Standard (commercially available) CE apparatuses employ capillaries with a length of few to several tens of centimeters and the separation time is usually 5 to 30 min [15]. When dealing with some analytical problems, for example in monitoring the kinetics of biochemical reactions [16] or as a subsequent separation method in multidimensional separation systems, it is necessary that the separation time be much shorter. Similarly, in the normal clinical monitoring of analytes in large sets of samples, even a small increase in the rate of separation contributes to shortening the analysis time, affecting the number of samples that can be analyzed in unit time and thus the price of the analysis. Shortening of the separation pathway is an effective technique for increasing the rate of separation. One of the options involves performance of electrophoresis on a glass or plastic microchip, where the length of the separation pathway is of the order of units of centimeters and the analysis time ranges from units to tens of seconds [17–19]. In practice, this means of separation has a disadvantage in the necessity of using a special separation system - an electrophoretic chip. In addition, contactless conductivity and UV spectrometry detection combined with a microchip do not achieve high sensitivity and the application potential required for clinical systems is limited [20,21]. From this point of view, the use of standard electrophoretic capillaries seems preferable, where their length, inner diameter, and material can be easily selected as required; their generally www.electrophoresis-journal.com
CE and CEC
Electrophoresis 2015, 36, 1962–1968
easy availability and substantially lower price are also important. In addition, the separation efficiency in channels with circular capillary cross-section is higher than in capillaries with rectangular cross-sections, familiar for microchips [22]. Short-end injection is a technically simple approach enabling shortening of the separation pathway, where the sample is injected into a site close to the detector; commercial instrumentation can frequently be used for this purpose [23]. This technique can be used to shorten the effective length of the capillary to approximately 8 to 10 cm with maintenance of minimal overall length of the capillary of about 30 cm, which is necessary when using commercial CE apparatuses. However, greater experimental variability can be attained using special apparatuses for separation directly in short capillaries with overall length of about 10 cm. A basic experimental problem when using separation in short capillaries lies in their limited mechanical flexibility. A short capillary cannot be handled in the same way as the capillaries used in standard electrophoretic systems. Consequently, the apparatus must be designed so that all the necessary experimental steps, especially sample injection and flushing of the capillary, can be performed without it being necessary to move the capillary. Injection of the sample into a separation capillary of normal length and especially into a short capillary is generally a critical point in achieving high-performance electrophoretic separation. A suitable injection strategy in separation in short capillaries permits elimination of one of the sources of uncertainty in the standard means of injection into capillaries of classical length, consisting in the repeated movement of the capillary injection end between sample and electrolyte solution; the capillary is thus physically disrupted from the solution. A detailed survey of various approaches to this method of separation can be found, for example, in the reviews [15,24,25]; a similar subject is discussed in relation to separation on an electrophoretic chip [26]. Another special injection technique is based on sucking the sample into the horizontally fixed capillary by contraction of the BGE inside the capillary that is achieved by lowering the temperature of the liquid surrounding the capillary [27]. Recently, rapidly developing fast separation systems using ESI-MS detection call on designing novel strategies and apparatus for exact sampling. The sampling systems are often complex and encompass many components including a micromanipulator controlling precise position of sampling and separation capillaries, and require multistep sampling protocol [28]. A special sampling microchip with integrated microvalves controlled pneumatically by nitrogen pressure [29] and a porous tip at the terminus of the capillary introduced for sheathless interfacing of CE to MS and for sample stacking during injection [30] are also described in literature. For less-demanding CE experimental arrangements, an easily available option lies in sample injection based on the principle described in the work [31]. The injection end of the capillary is loosely inserted into a tube through which the BGE is flowing during the injection, into which the sample is injected from the loop of a six-way valve. The sample can be injected into the capillary during the time that the sample C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1963
zone flows around the injection end of the capillary. Thus the injection end of the capillary is constantly submerged in the solution. This principle was utilized in the electrokinetic injection of samples of energy beverages [32, 33]. However, in electrokinetic sample injection, the amount of injected substance in the sample depends on its electrophoretic mobility, that is more ions with higher mobility than ions with lower mobility are added during the injection [34]. In addition, the amount of electrokinetically injected analyte is highly dependent on the total conductivity and composition of the sample matrix and, in clinical analyses, these factors depend on the content of majority sodium and chloride ions and on the preparation of the sample by dilution or the use of organic solvents. The above problems can be resolved by the use of pressure-assisted injection described in this work, where it is not necessary to move the capillary and its injection end is constantly immersed in a solution.
2 Materials and methods 2.1 Apparatus The injection end of the electrophoretic apparatus, Fig. 1A, consists of commercially available plastic components used for connecting tubes of small dimensions (1, 2). The electrophoretic ground electrode (3) is a 1-cm-long piece of stainless-steel tube; plastic components (1) and (2) fix the electrode insulated using hot melt adhesive in a stable position. The separation capillary (4) is tightly fitted (under heat) into the auxiliary PTFE tube, ensuring its fixed position in the injection part. The injection end of the capillary is inserted to a depth of approximately 1 mm into the PTFE tube (5) through which the BGE solution with the sample flows during the injection. The shut-off valve (10) at the exit from the injection part of the apparatus closes the exit PTFE tube (11). This creates a pressure pulse in the apparatus, injecting the sample into the capillary. The hydrodynamic resistance (13) created from a 5-mm-long piece of capillary 75 m id causes a sharper increase in the pressure in the apparatus when the shut-off valve is closed. The separation was performed in a standard fused-silica capillary (Composite Metal Services, UK), 50 m id, 363 m od, with length total/to the detector 10.5/8 cm. The electrophoretic apparatus was laboratory made and is described in detail in a previous publication [33]. The separation capillary leads from the injection part to the dual C4 D/UV detector and to the end vessel with the high-voltage (HV) electrode of the HV source. Figure 1B depicts the time sequence of the injection. The injection is commenced by turning on the linear pump for the BGE and simultaneously switching the six-way valve (6) from the “load" position (filling the injection loop) to the “inject" position. Following a certain time, necessary for the sample zone in the BGE stream to reach the injection end of the capillary, the shut-off valve is activated for a defined time and the thus-generated pressure pulse injects the sample into the capillary. The linear pump is turned on for long enough www.electrophoresis-journal.com
1964
A. Makrl´ıkova´ et al.
Electrophoresis 2015, 36, 1962–1968
11
2 10
1
2
3
3
9 6
7
B
1
5 13
4
8
5
4
time
12
A
Figure 1. Scheme of the injection part of the electrophoretic apparatus (A) and scheme of the time dependence of the injection sequence (B). (A) 1, 2—barbed tubing fittings, 3/32 id Tee and 3/32 id Elbow (World Precision Instruments); 3—stainless-steel tube (ground electrode of the HV source) with soldered connection, insulated with hot melt adhesive; 4—separation capillary inserted in 1/16 od × 0.01 id PTFE tube; 5—PTFE tube 1/16 od × 0.031 id; 6—six-way sampling valve (Valco Instruments); 7— injection loop with volume of 40 L; 8—filling the injection loop; 9—BGE feed from the stock reservoir pumped by a linear pump ˇ (Labio, CR); 10—shut-off valve (Cole Parmer); 11—waste; 12—to the detector and end vessel; 13—5-mm-long piece of capillary 75 m id. (B) 1—initiation of injection; 2—time of operation of the linear pump; 3—delay of the shut-off valve; 4—time of activation of the shut-off valve, that is time of sample injection into the capillary; 5—turning on the HV source and commencement of the separation.
for the sample zone to reach a sufficient distance from the injection end of the capillary. After turning off the pump, the HV source is turned on and the separation is commenced. The actions of the individual parts of the apparatus were controlled through an interface by a program created in LabView (National Instruments).
2.2 Chemicals and tested materials All the chemicals used were of p.a. purity: acetic acid (HAc; Sigma, Steinheim, Germany), Ca(NO3 )2 (Lachema, Brno, CR), creatinine (Fluka, Buchs, Germany), crown ether 18-crown-6 (Fluka, Buchs, Germany), hippuric acid sodium salt (Aldrich, Milwaukee, WI, USA), histidine (His, Aldrich, Steinheim, Germany), KCl (Lachema, Brno, CR), 2-morpholinoethane sulfonic acid (MES, Sigma, Buchs, Hungary), NaOH (Fluka, Steinheim, Switzerland), (NH4 )2 SO4 (Fluka, Buchs, Germany), and uric acid (Fluka, Buchs, Switzerland). The parameters of the pressure-assisted injection were optimized using BGE 20 mM MES + 5 mM NaOH (pH 5.8), the test sample was an aqueous solution with equimolar concentration of 50 M K+ and His (i.e. a mixture formed by a small inorganic and larger organic ions). A separation voltage of 5 kV was used in all the analyses and the experiments were performed at laboratory temperature. Some of the components in human urine were determined to illustrate the practical applicability of the proposed C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
pressure-assisted injection. Urine samples obtained from a healthy volunteer were analyzed on the day of sampling. Prior to injection into the capillary, the urine was only filtered using a single-use injection-needle PVDF filter (Fisher Scientific, CR) with a pore size of 0.45 m and then diluted 50 or 100 times with deionized water. The separation of the urine was performed in two BGEs, 50 mM MES + 5 mM NaOH (pH 5.1) and 1 M HAc + 1.5 mM 18-crown-6 (pH 2.4). The detection was performed using both detection systems, laboratory-made C4 D operated with a sine-wave signal with a frequency of 450 kHz and an amplitude of 17 V (peak to peak) and UV operated at 214 nm.
3 Results and discussion 3.1 Optimization of the pressure-assisted injection parameters Optimization of the injection consisted of search for the time intervals of the individual steps of the injection sequence, see Fig. 1B (for illustration, only the results obtained for the potassium ion are presented, where not stated otherwise; similar results were obtained for His): (1) “Time of activation of the linear pump” so that the sample zone carried by the BGE stream reaches a sufficient distance from the injection end of the capillary, that is to the ground electrophoretic electrode, time (2). The following optimum pump parameters were determined experimentally (by monitoring the movement of colored zones in the sample): flow rate, 1 mL/min and activation time 10 s; under these conditions, the BGE consumption for one sample injection was 170 L. These parameters were used in subsequent measurements. (2) “Time of delayed activation of the shut-off valve” after commencing the injection sequence so that the sample zone passes around the injection end of the capillary, time (3). It can be seen from Fig. 2A that the greatest amount of substance is injected at a time 4 to 6 s after commencing the injection sequence, when the central part of the sample zone passes around the injection end of the capillary. A time of 5 s was chosen for the subsequent experiments. (3) “Time of activation of the shut-off valve,” that is, the time during which the pressure is increased in the injection system so that a suitable length of the sample zone is injected into the capillary, time (4). It can be seen from the peak area and the separation efficiency expressed as the number of theoretical plates on the length of the pressure pulse, Fig. 2B, that, under the given experimental conditions, the optimum duration of the pressure pulse is 0.5 s. A sufficient amount of sample is injected for problemfree detection without a reduction in the separation efficiency. Figure 3 depicts an illustrative electropherogram of the test mixture recorded under the optimized conditions by C4 D. www.electrophoresis-journal.com
CE and CEC
Electrophoresis 2015, 36, 1962–1968
1965
Figure 2. The effect of the time of delay of the activation pulse (A) and the duration of the activation pulse (B) of the shut-off valve on the peak area and separation efficiency for the potassium ion. Common injection parameters: time of activation of the linear pump; 10 s, BGE flow rate, 1 mL/min. Duration of the activation pulse in dependence (A), 0.5 s; time of delay of the activation pulse in dependence (B), 5 s.
Figure 3. Electropherogram of an equimolar mixture of 50 M potassium ion (1) and His (2) recorded under optimized conditions of the pressure-assisted injection. Buffer, 20 mM MES + 5 mM NaOH (pH 5.8), 5 kV/3 A. Injection parameters: time of activation of the linear pump, 10 s; BGE flow rate, 1 mL/min; delay and duration of the activation pulse of the shut-off valve, 5 and 0.5 s, respectively.
Figure 4. Time dependence of the pressure in the injection part of the apparatus during activation of the shut-off valve for a period of 0.5 s; BGE flow rate, 1 mL/min.
The pressure in the apparatus was measured with activated shut-off valve. For illustration, Fig. 4 depicts the time dependence of the pressure for duration of the activation pulse of 0.5 s. The volume of the sample injected into the capillary was estimated using the equation
Sampling pulse length (s)
Average pressure (mbar)
Volume sampled (nL)
Zone length (mm)
Analyte amount (pmol)
0.3 0.4 0.5 0.6 0.7 0.8
45.7 57.7 80.4 103.4 128.6 149.5
2.0 3.4 5.9 9.0 13.2 17.5
1.0 1.7 3.0 4.6 6.7 8.9
0.10 0.17 0.29 0.45 0.66 0.87
V = 7.81 × 10−11
⌬ P D4 t , L
employing the average pressure during the pulse. For the given numerical coefficient, V is the injected volume in nanoliters and the other parameters are ⌬ P (mbar) pressure, D (50 m) inner capillary diameter, t (s) injection time, (0.001 Pa·s) solution viscosity, and L (10.5 cm) total capillary length. The injected volume, length of the sample zone, and injected amount of analyte are given in Table 1 for various durations of the activation pulse. C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Table 1. Injected sample volume, length of the sample zone in the capillary, and injected amount of analyte (for a concentration of 50 M K+ ) for various durations of the injection pulse
It is possible to very simply control the amount of sample injected into the capillary by selecting the duration of the pulse activating the shut-off valve. It is necessary to use shorter times for greater demands on the separation efficiency; if the demands on the separation are not great, the www.electrophoresis-journal.com
1966
A. Makrl´ıkova´ et al.
Electrophoresis 2015, 36, 1962–1968
Figure 5. Part of the electropherograms of human urine recorded by C4 D and the UV detector. (A) Urine diluted with water 1:50, BGE 50 mM MES + 5 mM NaOH, separation voltage/current 5 kV/3 A. Identification of peaks: unseparated inorganic ions (1), His (2), creatinine (3), neutral substances absorbing UV radiation (4), uric acid (5), and hippuric acid (6). (B) Urine diluted with water 1:100, BGE 1 M HAc + crown ether 18-crown-6, separation voltage 5 kV/13 A. Identification of peaks: NH4 + (1), K+ (2), Ca2+ (3), Na+ (4), creatinine (5), and His (6).
Table 2. Parameters of the linear regression equation of the calibration dependences of the area of the peaks for the potassium ion and His in the concentration range 10 to 150 M
Analyte
K+ (C4 D)
His (C4 D)
Intercept (mV) Slope (mV.s.M−1 ) Coefficient of determination (R2 ) Standard error (mV.s) LOD (M)
0a) 0.073 (0.003) 0.993 0.350 14
0.728 (0.363) 0.358 (0.004) 0.999 0.511 4
Buffer, 20 mM MES + 5 mM NaOH (pH 5.8), 5 kV/3 A. For the injection parameters, see Fig. 3; SDs in parenthesis. a) The zero value of the intercept lies inside the reliability interval.
detected signal can be increased by increasing the injection time. The repeatability of the injection was determined from the areas of the peaks of ten consecutive measurements. RSD determined for 50 M K+ and 50 M His were 4.0 and 1.7%, respectively, and for 10 M K+ RSD equaled 7.4%. The repeatability given in the literature expressed as RSD is 5–10% for manual injection in laboratory systems and 1–3% for automated injection in commercial instruments [35]. The repeatability of the tested injection pressure is somewhere between these values. The dependence of the peak areas on the concentration was linear in the tested concentration interval for the potassium ion and His; the parameters of the calibration curves are given in Table 2. The LOD was calculated from the parameters of the calibration equation, LOD = 3 × Standard error/Slope. LOD for the potassium ion determined by this method is high; the coion in the buffer employed, the Na+ ion, is suitable for sensitive conductivity detection of His but not for detection of the K+ ion. For comparison, the same test sample was injected electrokinetically, that is the shut-off valve was not activated during the injection and the separation voltage was permanently turned on. The ratio of the slopes His/K+ of the calibration dependences in pressure-assisted injection was 4.9 and, for C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
electrokinetic injection, 1.9. The discrimination of large ions in electrokinetic injection is evident.
3.2 Separation of ammonium ions, histidine, creatinine, uric acid, and hippuric acid in human urine The determination of important biochemical analytes in human urine, ammonium ions, creatinine, His, and uric and hippuric acids [36,37] was selected for verification of the practical applicability of the tested injection method. The ammonia content is an indicator of disorders in urea synthesis and elevated protein catabolism. The amount of produced and excreted creatinine is used in correction of diuresis. An elevated uric acid level indicates a danger of the formation of bladder and kidney stones. Hippuric acid is the final product in the biotransformation of benzoic acid, toluene, and other substances containing a benzene ring. His is the main amino acid in urine and elevated excretion is connected with a congenital metabolic disorder. Separation of these substances was first tested in BGE with composition 50 mM MES + 5 mM NaOH, pH 5.1; the pH is lower than that for the separation model mixture of K+ and His because of the need to protonate creatinine and separate it from EOF (see Fig. 5A). In this BGE, EOF marker flows through the capillary in a time of 40 s; before EOF on the C4 D recording can be seen the positive peak of the mixed zone of inorganic ions, followed by the negative peaks of the organic cations His and creatinine. The EOF position in the UV recording is followed by the separated peaks of uric and hippuric acids, which migrate as anions in the opposite direction to EOF. Addition of 18-crown-6, which complexes K+ , to the BGE led to only partial separation of the NH4 + and K+ ions; at rapid EOF combined with a short capillary, there is not sufficient time for complete separation of the two ions with similar mobility values. For this reason, acidic BGE with composition 1 M HAc + 18-crown-6 (pH 2.4), in which EOF is greatly suppressed, was tested for the separation cations in urine. Under these conditions, the motion of the cations through the capillary is slowed and they move on the basis of their own electrophoretic www.electrophoresis-journal.com
CE and CEC
Electrophoresis 2015, 36, 1962–1968
1967
Table 3. Determination of creatinine, His, uric acid, and hippuric acid (detection method in parentheses)
Analyte
Migration time (s)
Concentration (mg/L)
His (C4 D) His (UV) Creatinine (C4 D) Creatinine (UV) Uric acid (UV) Hippuric acid (UV)
32 ± 1
67 61 1031 1209 392 539
36 ± 1 54 ± 3 69 ± 5
± ± ± ± ± ±
7 6 233 364 20 33
RSD (%)
S/Na)
5.1 4.6 10.2 13.7 2.3 2.7
10 5 160 75 34 30
Nb) (m−1 )
Range [39] (mg/L) 40–330
136 200 52 300 24 900 118 000 150 000
670–2150 40–670 50–1670
Sample: human urine/water 1:50, BGE 50 mM MES + 5 mM NaOH, separation voltage/current 5 kV/3 A. a) S/N, detector signal (peak height)/noise of baseline. b) The number of theoretical plates was calculated by the formula N = 5.54 (tM /w1/2 )2 , where tM is the migration time and w1/2 is the peak width at half-height.
mobility. This leads to complete separation of all the studied cations, yielding negative peaks in C4 D and, in addition, creatinine can be detected in the UV detector (see Fig. 5B). However, the acids cannot be determined in this acidic BGE. The obtained separation efficiencies (Tables 3 and 4) commonly attain values of about 100 000 theoretical plates/m; these values are fully comparable with the separation efficiencies for classical hydrodynamic injection of samples connected with exchange of vials. This indicates that the newly developed injection method ensures injection of discrete sample zones at the beginning of the capillary. The results of the determination of the monitored components in urine in both BGEs and detection by both detection systems are given in Tables 3 and 4. The standard addition method was used in the determination and the results are the median of three determinations. It can be seen that, within the limits of reliability, C4 D and UV detection yield the same results for those analytes that can be determined by both methods. Creatinine can be determined in both BGEs. However, the determination is more accurate in HAc-crown ether BGE, because the creatinine peak in MES-NaOH BGE is immediately next to the EOF peak, complicating its evaluation. In contrast, in HAc-crown BGE, His is poorly separated from the other components of the urine, probably from its methyl derivatives [38]; consequently, it was determined only in MES-NaOH BGE. The S/N parameter in Table 3 indicates slightly better performance of C4 D over UV detector for both creatinine and histidine. The last column in Table 3 gives the range of concentrations in which this component can be present in human urine [39]; the determined concentrations of all the test components fall within this range.
4 Concluding remarks A simple apparatus is proposed for pressure-assisted injection into a short separation capillary, which is an advantageous alternative to more frequently used electrokinetic injection, previously described in [32, 33]. It is not necessary to move the capillary and its injection end is constantly in contact with the solution. For optimized working conditions, it is possible to obtain injection reproducibility expressed as C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Table 4. Determination of ammonia ions and creatinine (detection method in parentheses)
Analyte
Migration time (s)
Concentration (mg/L)
RSD (%)
N (m−1 )
NH4 + Creatinine (C4 D) Creatinine (UV)
34 ± 0 68 ± 0
515 ± 95 1666 ± 75 1717 ± 159
8.3 2.0 4.2
132 000 95 600 62 700
Sample: human urine/water 1:100, BGE 1 M HAc + 1.5 mM crown ether 18-crown-6, separation voltage 5 kV/13 A.
RSD of about 5%. The equipment facilitates easy changes in the injection conditions by changing a single parameter, the duration of the pressure pulse. The apparatus was tested by determining several components of human urine. Separation on a short pathway has the general advantage of speed. Using the described apparatus, the separation time of the monitored components, ammonium ions, creatinine, His, uric and hippuric acids, was about 70 s. However, attempts to determine all the test substances in a single BGE were not successful and two different separation buffers had to be used. The apparatus can be used for rapid screening of the contents of the other inorganic ions; the resolution parameters for the NH4 + /K+ , K+ /Ca2+ , and Ca2+ /Na+ ion pairs equaled 1.3, 3.4, and 1.3,respectively. This work was supported by Charles University in Prague, Projects SVV, PRVOUK 31, and UNCE 204015/2012. The authors have declared no conflict of interest.
5 References [1] Ramautar, R., Demirci, A., deJong, G. J., Trends Anal. Chem. 2006, 25, 455–466. [2] Jabeen, R., Payne, D., Wiktorowcz, J., Mohammad, A., Petersen, J., Electrophoresis 2006, 27, 2413–2438. [3] Li, S. F. Y., Kricka, L. J., Clin. Chem. 2006, 52, 37–45. [4] Stalmach, A., Albalat, A., Mullen, W., Mischak, H., Electrophoresis 2013, 34, 1452–1464. [5] de Macedo, A. N., Jiwa, M. I. Y., Macri, J., Belostotsky, V., Hill, S., Britz-McKibbin, P., Anal. Chem. 2013, 85, 11112–11120.
www.electrophoresis-journal.com
1968
A. Makrl´ıkova´ et al.
Electrophoresis 2015, 36, 1962–1968
[6] Guihen, E., Electrophoresis 2014, 35, 138–146.
[25] Matysik, F. M., Anal. Bioanal. Chem. 2010, 397, 961–965.
[7] Moser, A. C., Willicott, C. W., Hage, D. S., Electrophoresis 2014, 35, 937–955.
[26] Karlinsey, J. M., Anal. Chim. Acta 2012, 725, 1–13.
[8] Lucangioli, S., Tripodi, V., Curr. Anal. Chem. 2014, 10, 183–183. ˇ ık, K., Electrophoresis 2011, 32, 795–810. [9] Opekar, F., Stul´
[28] Grundmann, M., Matysik, F. M., Anal. Bioanal. Chem. 2012, 404, 1713–1721.
´ M., Preisler, J., Foret, F., Hauser, P. C., Krasen[10] Ryvolova, sky, P., Paull, B., Macka, M., Anal. Chem. 2010, 82, 129–135. ˇ P., Hauser, P. C., Electrophoresis 2013, 34, 55–69. [11] Kuba´ n, ˇ P., Hauser, P. C., Electrophoresis 2011, 32, 30–42. [12] Kuba´ n, ´ ık, K., Electrophoresis 2013, 34, 70–85. [13] Kleparn´ [14] Ramautar, R., Somsen, G. W., deJong, G. J., Electrophoresis 2013, 34, 86–98. ˇ ık, K., Chem. Rev. 2009, 109, [15] Opekar, F., Coufal, P., Stul´ 4487–4499. [16] van Dyck, S., Kaale, E., Novakova, S., Glatz, Z., Hoogmartens, J., vanSchepdael, A., Electrophoresis 2003, 24, 3868–3878.
[27] Hjerten, S., Electrophoresis 1990, 11, 665–690.
[29] Kelly, R. T., Wang, C. C., Rausch, S. J., Lee, C. S., Tang, K. Q., Anal. Chem. 2014, 86, 6723–6729. [30] Moini, M., Martinez, B., Rapid Commun. Mass Spectrom. 2014, 28, 305–310. ˚ [31] Tuma, P., Opekar, F., Jel´ınek, I., J. Chromatogr. A 2000, 883, 223–230. ˇ ık, K., Anal. ´ B., Opekar, F., Tuma, ˚ [32] Vochyanova, P., Stul´ Bioanal. Chem. 2012, 404, 1549–1554. ´ ´ B., Opekar, F., Tuma, ˚ [33] Vochyanov a, P., Electrophoresis 2014, 35, 1660–1665. [34] Lauer, H. H., Rozing, G. P., High Performance Capillary Electrophoresis, A Primer, Agilent Technologies, Germany 2010, pp. 59–72.
[17] Shang, F. J., Guihen, E., Glennon, J. D., Electrophoresis 2012, 33, 105–116.
ˇ P., Seiman, A., Kaljurand, M., Capillary Elec[35] Kuba´ n, trophoresis and Microchip Capillary Electrophoresis, John Wiley & Sons, New Jersey 2013, pp. 41–65.
[18] Guihen, E., O’Connor, W. T., Electrophoresis 2010, 31, 55–64.
[36] Devlin, T. M., Biochemistry with Clinical Correlations, Wiley-Liss, New York 1992, pp. 475–573.
[19] Roman, G. T., Kennedy, R. T., J. Chromatogr. A 2007, 1168, 170–188.
[37] Murray, R. K., Bender, D. A., Botham, K. M., Kennelly, P. J., Rodwell, V. W., Weil, P. A., Harper’s Illustrated Biochemistry, McGraw-Hill Lange, China 2007, pp. 271–306, 331–342. ˇ ık, K., ˚ ´ E., Opekar, F., Jurka, V., Stul´ [38] Tuma, P., Samcova, Electrophoresis 2007, 28, 2174–2180.
ˇ P., Hauser, P. C., Lab Chip 2008, 8, 1829–1836. [20] Kuba´ n, [21] Abad-Villar, E. M., Kuban, P., Hauser, P. C., Electrophoresis 2005, 26, 3609–3614. ˇ B., Kenndler, E., Electrophoresis 2002, 23, 3817– [22] Gas, 3826. [23] Glatz, Z., Electrophoresis 2013, 34, 631–642. ´ M., Preisler, J., Brabazon, D., Macka, M., [24] Ryvolova, Trends Anal. Chem. 2010, 29, 339–353.
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
[39] Putnam, D. F., Composition and concentrative properties of human urine, Report No. NASA CR1802, McDonnell Douglas Astronautic company, Huntington Beach, California, http://ntrs.nasa.gov/ archive/nasa/casi.ntrs.nasa.gov/19710023044.pdf.
www.electrophoresis-journal.com
