Anal. Chem. 2003, 75, 3859-3864
Simultaneous Determination of Tryptophan and Glutathione in Individual Rat Hepatocytes by Capillary Zone Electrophoresis with Electrochemical Detection at a Carbon Fiber Bundle-Au/Hg Dual Electrode Wenrui Jin,* Xiujun Li, and Ning Gao
School of Chemistry and Chemical Engineering, Shandong University, Jinan 250100, China
A method for single-cell analysis was developed by combining capillary zone electrophoresis (CZE) with electrochemical detection (ECD) using a dual electrode consisting of two different types of electrode material (carbon fiber and Au/Hg). In this method, the parallel mode was used. Different potentials were applied to both electrodes of the dual electrode. Tryptophan and glutathione, which could not be simultaneously detected by normal CZEECD, could be simultaneously and selectively detected by CZE-ECD at the dual electrode in one run, respectively. The CZE-ECD system with the dual electrode was applied to determine them in individual rat hepatocytes. Capillary zone electrophoresis (CZE) is a good method for single-cell analysis.1 The required sample volumes can go as low as nanoliters to femtoliters, and it can also detect analytes with high sensitivity. When electrochemical detection (ECD) or laserinduced fluorescence detection is used, femtomole to zeptomole components can be measured. The extremely high separation efficiency of CZE allows good resolution of many constituents present in a single cell. Species separations can be achieved on a relatively short time scale. Jankowski et al.2 and Yeung3 have reviewed the single-cell analysis, respectively. CZE-ECD has been used to detect electroactive neurotransmitters such as dopamine, serotonin, epinephrine, and norepinephrine in single cells.4-11 In our laboratory, single-cell analyses for other compounds such as * Corresponding author: (e-mail)
[email protected]; (fax) +86531-856-5167. (1) Ewing, A. G.; Mesaros, J. M.; Gavin, P. F. Anal. Chem. 1994, 66, 527A537A. (2) Jankowski, J. A.; Trach, S.; Sweedler, J. V. Trends Anal. Chem. 1995, 14 (4), 170-176. (3) Yeung, E. S. J. Chromatogr., A 1999, 830, 243-262. (4) Wallingford, R. A.; Ewing, A. G. Anal. Chem. 1988, 60, 1972-1975. (5) Chien, J. B.; Wallingford, R. A.; Ewing, A. G. J. Neurochem. 1990, 54, 633638. (6) Olefirowicz, T. M.; Ewing, A. G. J. Neurosci. Methods 1990, 34, 11-15. (7) Olefirowicz, T. M.; Ewing, A. G. Chimia 1991, 45, 106-108. (8) Bergquist, J.; Tarkowski, A.; Ekman, R.; Ewing, A. G. Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 12912-12916. (9) Kristensen, H. K.; Lau, Y.; Ewing, A. G. J. Neurosci. Methods 1994, 51, 183-188. (10) Hu, S.; Pang, D.; Wang, Z.; Cheng, J.; Li, Z.; Fan, Y.; Hu, H. Chem. J. Chin. Univ. 1996, 17, 1207-1209. (11) Swanek, F. D.; Chen, G.; Ewing, A. G. Anal. Chem. 1996, 68, 3912-3916. 10.1021/ac0207022 CCC: $25.00 Published on Web 06/26/2003
© 2003 American Chemical Society
glutathione,12,13 diclofenac (a drug),14 amino acids,15,16 and ascorbic acid17 have also been investigated. For those studies, only single electrodes were used. However, the electrochemical response varied depending upon the combination of the analytes and electrode materials. For example, the electrochemical response of glutathione (GSH) is sensitive at the mercury electrode but not at the carbon electrode. Neurotransmitters, ascorbic acid, uric acid, and some amino acids such as tryptophan (Trp), histidine, and tyrosine can be detected at the carbon electrode but not at the mercury electrode. Thus, they cannot be detected simultaneously in one run by CZE. When solution samples were applied, the substances that respond to different electrode materials could be measured with different electrodes in two runs. In this case, there will not be information lost for solution samples. However, this protocol cannot be applied to cell samples as duplicate cells cannot be obtained. To determine these compounds in single cells in one run, the dual electrodes constructed with different materials should be used in CZE. Dual electrodes with the same material such as Au/Hg,18,19 carbon fiber,20 Pt,21 Au,19,22-24 and carbon film25 have been described in CZE. However, there was no report concerning dual electrodes constructed with different materials. Tryptophan is an essential amino acid. Liver cells take it and use it to synthesize other important molecules such as niacin and serotonin.26 Glutathione is also produced in liver cells and acts as (12) Jin, W.; Li, W.; Xu, Q. Electrophoresis 2000, 21, 774-779. (13) Jin, W.; Dong, Q.; Ye, X.; Yu, D. Anal. Biochem. 2000, 285, 255-259. (14) Dong, Q.; Jin, W. Electrophoresis 2001, 22, 2786-2792. (15) Weng, Q.; Jin, W. Electrophoresis 2001, 22, 2797-2803. (16) Dong, Q.; Wang, X.; Zhu, L.; Jin, W. J. Chromatogr., A 2002, 959, 269279. (17) Jin, W.; Jiang, L. Electrophoresis 2002, 23, 2471-2476. (18) Lin, B. L.; Colo´n, L. A.; Zare, R. N. J. Chromatogr., A 1994, 680, 263-270. (19) Zhong, M.; Lunte, S. M. Anal. Chem. 1999, 71, 251-255. (20) Zhong, M.; Zhou, J.; Lunte, S. M.; Zhao, G.; Giolando, D. M.; Kirchhoff, J. R. Anal. Chem. 1996, 68, 203-207. (21) Holland, L. A.; Lunte, S. M. Anal. Chem. 1999, 71, 407-412. (22) Chen, D.; Chang, S.; Chen, C. Anal. Chem. 1999, 71, 3200-3206. (23) Matysik, F.-M.; Bjo¨refors, F.; Nyholm, L. Anal. Chim. Acta 1999, 385, 409415. (24) Niwa, O.; Kurita, R.; Liu, Z.; Horiuchi, T.; Torimitsu, K. Anal. Chem. 2000, 72, 949-955. (25) Liu, Z.; Niwa, O.; Kurita, R.; Horiuchi, T. Anal. Chem. 2000, 72, 13151321. (26) Fukuwatari, T.; Morikawa, Y.; Sugimoto, E.; Shibata, K. Biosci. Biotechnol. Biochem. 2002, 66, 1196-1204.
Analytical Chemistry, Vol. 75, No. 15, August 1, 2003 3859
a protection molecule for proper liver functions. Glutathione level is always evaluated in determining hepatotoxity.27 The amount of both Trp and GSH reflects liver cell function from different sides. If the amount of both Trp and GSH in liver cells could be obtained simultaneously, a more accurate conclusion on cell function would be drawn. It is very difficult to culture the primary liver cell in vitro, because a homogeneous population (cell line) obtained from one cell in a liver tissue by tissue culture techniques must be different from the primary liver cell and must have lost some of its characteristics. Therefore, simultaneous determination of Trp and GSH in single liver cells is important and significant work. In this work, we developed a method for simultaneous determination of Trp and GSH in individual rat hepatocytes by CZE-ECD at a carbon fiber bundle-Au/Hg dual electrode with a parallel mode. They were determined at both electrodes of the dual electrode, respectively. To our knowledge, this work is the first demonstration of the dual electrode being used for single-cell analysis. EXPERIMENTAL SECTION Chemicals and Reagents. Trp (chromatographically pure) was purchased from Shanghai Biochemical Reagents Co. (Shanghai, China). Trp was dissolved in water to make a 1.00 × 10-2 mol/L stock solution. Cysteine (Cys, content >98.5%) was obtained from Shanghai Kangda Amino Acids Co. (Shanghai, China). GSH (content >98%) was obtained from Acros Organic. The stock solutions were prepared at concentrations of 1.00 × 10-2 mol/L each for Cys and GSH in 1.00 × 10-3 mol/L Na2H2EDTA. They were stored at 4 °C. Dilute solutions were obtained by a serial dilution of the stock solutions with the running buffer. Sulfo-5-salicylic acid (SSA) was obtained from Shanghai First Reagents Factory (Shanghai, China). The physiological buffer saline (PBS) consisted of 0.135 mol/L NaCl and 0.02 mol/L NaH2PO4-NaOH (pH 7.4). The D-Hanks solution was prepared by dissolving 0.8 g of NaCl, 0.40 g of KCl, 0.06 g of Na2HPO4‚H2O, 0.06 g of KH2PO4, and 0.35 g of NaHCO3 in 1 L of water with 0.02 g of phenol red (95.8%, Beijing Chemical Reagents Factory, Beijing, China) as a pH indicator. A 0.25% (w/v) trypsin (>2500 units/mg, Shanghai Chemical Reagents Co., Shanghai, China) solution made in D-Hanks solution at 4 °C. pH 7.4, was adjusted with NaHCO3 after filtering. The trypsin solution was sealed and was stored at -20 °C. The running buffer was deaerated for 15 min with N2 before use. All other reagents were of analytical grade and were purchased from standard reagent suppliers. All solutions were prepared with double-distilled water. Preparation of Rat Hepatocytes. One rat (∼25 g) was killed by cervical dislocation. Its liver was washed with PBS and was cut into pieces. They were washed twice with PBS and then transferred into a 5-mL centrifuge tube. A 0.25% (w/v) trypsin solution (∼10-fold volume of the rat liver) was added to the centrifuge tube. After vibrating lightly, they were enzymolyzed for 20 min at 37 °C. The supernatant was discarded. The enzymolysis was stopped by addition of 3 mL of PBS, which was subsequently removed. This enzymolysis was repeated twice. Then 3 mL of PBS was added to the centrifuge tube, and the cell mixture was disrupted by pipetting. This was the hepatocyte suspension. Lysis and Extraction of Hepatocytes. The hepatocyte suspension was centrifuged for 10 min at 1000 rpm. The (27) Oak, S.; Choi, B. H. Exp. Mol. Pathol. 1998, 65 (1), 15-24.
3860
Analytical Chemistry, Vol. 75, No. 15, August 1, 2003
supernatant was discarded. The pellet was resuspended in 2 mL of PBS following by centrifugation at 1000 rpm for another 10 min. This was repeated several times until the supernatant was colorless and transparent. The cell number in the hepatocyte suspension was counted using a hemocytometer (Shanghai Medical Optical Instrument Plant, Shanghai, China). There were 2.0 × 106 cells in the 2-mL hepatocyte suspension. Then the hepatocyte suspension was centrifuged for 10 min at 1000 rpm. The supernatant was discarded again. The pellet of hepatocytes was resuspended in 200 µL of 4.0 × 10-4 mol/L Na2H2EDTA and were lysed by sonication for 10 min. A 700-µL aliquot of 0.025 mol/L Na2HPO4-NaH2PO4 running buffer (pH 7.5) and 200 µL of 5% SSA were added to remove proteins. After 5-min incubation, it was centrifuged for 20 min at 2500 rpm. The 1.10-mL supernatant from 2.0 × 106 cells was the hepatocyte extract. CZE-ECD System. The CZE system used in this work was similar to our previous system.28 A high-voltage power supply (model 9323-HVPS, Beijing Institute of New Technology, Beijing, China) provided a variable voltage of 0-30 kV across the capillary, with the outlet of the capillary at ground potential. Fused-silica capillaries (25-µm i.d., 375-µm o.d.), from Yongnian Optical Conductive Fiber Plant (Yongnian, China) were cut into a length of 49 cm and placed between two buffer reservoirs. A high voltage was applied at the injection end, while the reservoir containing the electrochemical detection cell was held at ground potential. Separations were carried out at an applied voltage of 20 kV. ECD at two constant potentials was performed with an electrochemical analyzer (model CHI802, CH Instruments, Austin, TX). The detection cell was housed in a Faraday cage in order to minimize the interference from noise from external sources. ECD was carried out with a four-electrode system that consisted of a carbon fiber bundle-Au/Hg dual electrode as the twice working electrode, a saturated calomel electrode (SCE), as the reference electrode, and a coiled Pt wire (0.5-mm diameter, 4 cm in length) placed at the bottom of the cell as the auxiliary electrode. The Pt wire also served as the ground for the high potential drop across the capillary. Construction of Carbon Fiber Bundle-Au/Hg Dual Electrode. First, the Au/Hg electrode was constructed using a 100µm-diameter, ∼3-cm-long gold wire. One end of the gold wire was bound to a 375-µm-diameter, ∼1-cm-long copper lead with a fine copper wire. Epoxy resin was then applied to the junction of the gold wire and the copper lead in order to isolate and protect the electrical junction (Figure 1A). The gold wire was washed with acetone and water. After drying, the gold wire was coated with epoxy resin and blown with hot air immediately. The surplus epoxy resin can be blown out, and the film of epoxy resin was formed on the gold surface. After drying, this process was repeated. The thickness of the film was ∼4 µm. The gold electrode was stored for use. Then ∼50 6-µm-diameter carbon fibers soaking up acetone were carefully inserted into a fused-silica capillary (∼250-µm i.d., 375-µm o.d., 0.6-cm length) (Figure 1B). The gold electrode was inserted into the fused-silica capillary with carbon fibers (Figure 1C). The gold wire and the carbon fibers were glued to the fused-silica capillary using a low-viscosity ethyl R-cyanoacrylate adhesive. The fused-silica capillary with the gold wire and the carbon fibers was then inserted into a glass capillary (∼2(28) Jin, W.; Weng, Q.; Wu, J. Anal. Chim. Acta 1997, 342, 67-74.
Figure 1. Manufacturing process of the carbon fiber bundle-Au/ Hg dual electrode: 1, copper lead; 2, epoxy resin; 3, Au/Hg; 4, fusedsilica capillary; 5, carbon fiber; 6, ethyl R-cyanoacrylate adhesive; 7, copper lead; 8, glass capillary; 9, mercury.
mm i.d., 3-mm o.d., 3-cm length), into which a small amount of mercury had been drawn and protruded ∼2 cm from the glass capillary. The other end of carbon fibers was connected to a copper lead via the mercury junction by pushing the copper wire down (Figure 1D). Then, epoxy resin was applied to the two ends of the glass capillary to seal the fused-silica capillary and the two copper leads to it (Figure 1E). Finally, the gold wire and the carbon fibers protruding from the fused-silica capillary were cut (Figure 1F). Before use, the surface of the gold wire and the carbon fibers was ground with emery paper. It was cleaned in ethanol and water for 2 min, respectively, by an ultrasonicator. After drying with filter paper, the electrode was dipped into pure mercury for ∼2 min. Thus, the carbon fiber bundle-Au/Hg dual electrode was constructed. The cross-sectional view of the dual electrode is shown in Figure 1G. This type of design proved very convenient because the working electrodes are not easily broken or curved. Mercury is toxic and hence it should be covered with water during experiments and storage. CZE-ECD Procedure. The carbon fiber bundle-Au/Hg dual electrode was cemented onto a microscope slide, which was placed over a laboratory-made xyz micromanipulator and glued in place. The position of the carbon fiber bundle-Au/Hg dual electrode was adjusted (under a microscope) against the end of the capillary with a distance of 10 µm. This arrangement allowed one to easily remove and realign both the capillary and the electrode. The other end of the capillary was inserted into a plastic syringe tip (the
metal needle was previously removed) and glued in place with a small amount of epoxy glue. Before each run, the capillaries were flushed with double-distilled water, 0.1 mol/L NaOH, doubledistilled water, and the corresponding separation electrolyte, respectively, using a syringe. During the experiments, the separation voltage was applied across the capillary and the detection potential was applied at the working electrode. After the electroosmotic flow reached a constant value, the electromigration injection was carried out and the electropherogram was recorded. All potentials were measured versus SCE. In the end-column amperometric detection with the dual electrode, the current detected is critically dependent on the proper alignment of the dual-electrode center with the bore of the capillary. The alignment procedure is similar to that described in our previous work.30 To accomplish this, the working electrode center was aligned to the capillary center in the horizontal direction with a distance of 10 µm between the dual electrode and outlet of the capillary by using the xyz micromanipulator. With the aid of a mirror, then, the vertical alignment was carried out using the xyz micromanipulator. Since the inside diameter of the separation capillary (25 µm) is less than the diameter of the dual electrode (250 µm), the species eluting from the capillary have access to a part of both electrodes of the dual electrode. To guarantee reproducibility, the alignment must be checked before each experiment by paying attention to the following steps: The elution curves of 5.00 × 10-5 mol/L Cys at the dual electrode were recorded. The placement of the dual electrode was adjusted again using the xyz micromanipulator. The manipulation was repeated until the ratio of the two peak currents at the Au/Hg electrode and the carbon fiber bundle electrode of the dual electrode approached 1.5 ( 7%, which was obtained when the dualelectrode center and the bore of the capillary were aligned. Injection, Lysis, and Analysis of Single Cells. Since the electroosmotic flow rate for the running buffer is large enough to draw a cell at the capillary inlet into the capillary, the electromigration technique can be used for the injection of single cells. An ∼15 µL aliquot of the hepatocyte suspension was placed on a clean microscope slide. After the microscope slide was placed on the inverted biological microscope with a magnification of 400×, the injection end of the capillary filled with electrophoresis buffer was gently immersed in the droplet of hepatocyte suspension under the guidance of an xyz micromanipulator. To see the opening of the injection end, a ∼5-mm section of the polyimide coating at the injection end of the capillary was removed by burning before use. A platinum wire was placed in the cell suspension to serve as the electrophoresis anode. When a hepatocyte drifted toward the injection end under the field of vision of the microscope, an injection voltage of 2 kV was applied to transport the whole cell into the capillary tip. The entire process of cell injection typically took 30-150 s. Then the capillary was gently moved from the hepatocyte suspension into a droplet of 0.01 mol/L NaOH as the cell lysis solution. The electromigration injection of 0.01 mol/L NaOH was carried out at 2.0 kV for 5 s. After the cell was lysed, the capillary was manipulated up, out of the NaOH, and immersed to the CZE running buffer solution. (29) Ohmori, S.; Kawase, T.; Hiigashiura, M.; Chisaka, Y.; Nakata, K.; Yamasaki, Y. J. Chromatogr., B 2001, 762, 25-32. (30) Jin, W.; Wang, Y. Anal. Chim. Acta 1997, 343, 231-239.
Analytical Chemistry, Vol. 75, No. 15, August 1, 2003
3861
Table 1. Values of ip, tm, W1/2, and N for Trp and GSH at Different Concentrations of Running Buffer, CBa CB (mol/L)
ip,Trp (nA)
ip,GSH (pA)
tm,Trp (min)
tm,GSH (min)
W1/2,Trp (s)
W1/2,GSH (s)
NTrp (104)
NGSH (104)
0.015 0.020 0.025 0.030
0.907 1.23 1.32 1.35
11.7 24.1 40.9 40.2
4.00 4.22 4.45 4.93
6.50 6.93 7.30 8.30
2.4 2.5 2.6 2.7
5.9 4.6 3.1 2.9
5.5 5.7 5.8 6.7
2.4 4.5 11 16
a Conditions: 5.00 × 10-5 mol/L Trp and 5.00 × 10-5 mol/L GSH; 4.0 × 10-4 mol/L Na H EDTA; capillary, 25-µm i.d., 375-µm o.d., 49-cm 2 2 length; separation voltage, 20 kV; injection, 5.0 kV for 10 s; detection potential, 1.10 V for the carbon fiber bundle electrode, 0.18 V for the Au/Hg electrode.
Then a separation voltage of 20 kV was applied and the electropherogram was recorded. RESULTS AND DISCUSSION Detection of Trp and GSH. Since Cys is present in hepatocytes29 and it can be oxidized at both the carbon fiber bundle electrode and Au/Hg electrode, the separation of Cys and GSH or Cys and Trp should be investigated. In the previous work,30,31 it was found that GSH and Cys can be oxidized at the Au/Hg electrode and separated at pH 2.4-8.0. These characteristics have been used for their simultaneous determination in CZE. In addition, when a constant detection potential more positive than 0 V versus SCE is applied to the Au/Hg electrode, deoxygenation is not necessary. Since the electrochemical response of GSH is not sensitive at the carbon fiber electrode, a low concentration of GSH cannot be detected at this electrode. Therefore, when the carbon fiber electrode was used, only one peak corresponding to Cys was depicted on the electropherogram. This means that there is no problem for simultaneous detection of Cys and GSH by using the dual electrode. It was noted that both Trp and Cys can be detected at the carbon fiber bundle electrode in weak acid and neutral solutions and Trp cannot be oxidized at the Au/Hg electrode. Therefore, the separation and detection conditions of Trp and Cys were investigated. Trp and Cys can be separated in 0.035 mol/L Na2HPO4-NaH2PO4 of pH 7.0 and 7.5, but not in the buffer of pH 6.1 and 6.5. In the pH 7.5 running buffer, the resolution of Trp and Cys is 9.65. The electrophoretic peak current, ip, the migration time, tm, the width at the peak half-height, W1/2, and the number of theoretical plates, N, on the electropherograms of Trp and GSH detected by the dual electrode at different concentrations of the running buffer, CB, are summarized in Table 1. tm of the both Trp and GSH increases with increasing CB. The migration velocity of the substance depends mainly on the electroosmotic velocity, veo, of the buffer, which is proportional to the ζ potential. With increasing buffer concentration, the thickness of the electrical double layer becomes thinner and the ζ potential becomes smaller. Therefore, veo decreases and tm increases with increasing CB. ip of both Trp and GSH increases with increasing CB. A possible reason is the reduction of iR drop in the electrochemical system, when CB increases. When CB g 0.025 mol/L, ip of both is a constant value, achieving a limit value. Therefore, 0.025 mol/L Na2HPO4-NaH2PO4 was selected in our experiments. Since the reaction of GSH with oxygen dissolved in the solution is catalyzed by trace metal ions, the stock solution of GSH was prepared in 0.001 mol/L Na2H2EDTA, which can form (31) Jin, W.; Wang, Y. J. Chromatogr., A 1997, 769, 307-314.
3862 Analytical Chemistry, Vol. 75, No. 15, August 1, 2003
complexes with the metal ions. It was found that GSH was not oxidized during runs because of the short running time and the presence of Na2H2EDTA.30 When the concentrations of Na2H2EDTA were more than 2.0 × 10-4 mol/L, ip of GSH was a constant value. A 4.0 × 10-4 mol/L Na2H2EDTA concentration was used in our experiments. ip of both Trp and GSH increases with increasing Ed in the range of 0.7-1.20 V for the carbon fiber bundle electrode and in the range of 0-0.20 V for the Au/Hg electrode. In our experiments, the detection potentials of 1.10 and 0.18 V were applied to the carbon fiber bundle electrode and Au/ Hg electrode of the dual electrode, respectively, because of less noise and higher ip. The linear ranges of Trp and GSH at the dual electrode are 5.00 × 10-7-5.00 × 10-5 mol/L with a correlation coefficient of 0.9998 for Trp and 5.00 × 10-6-1.00 × 10-4 mol/L with a correlation coefficient of 0.9995 for GSH. Their limits of detection (LODs) calculated from the electrophoretic peak currents obtained for the concentration at the low end of their linear range are 3.0 × 10-7 mol/L for Trp and 2.3 × 10-6 mol/L for GSH, respectively, when the signal-to-noise ratio is 3. Their mass LODs are 0.68 fmol for Trp and 3.2 fmol for GSH. The relative standard deviations of the method for determination of Trp are 0.50% for tm and 3.6% for ip, and for determination of GSH are 0.56% for tm and 1.6% for ip. Detection of Trp and GSH in Hepatocyte Extract. To determine whether electroactive compounds such as Cys, tyrosine (Tyr), histidine (His), dopa, dopamine (DA), serotonin (5-HT), epinephrine (E), and norepinephrine (NE), which can be directly oxidized at the dual electrode, interfered with the determination of Trp and GSH, the electropherograms of a solution containing them with Trp and GSH at the dual electrode were recorded and are shown in Figure 2. It was found that no peak of GSH for the concentration of 5.00 × 10-5 mol/L was detected and the peak of Trp could be well separated from the peaks of 5-HT, DA, E, NE, dopa, and Cys at the carbon fiber bundle electrode of the dual electrode. The overlapped peak of Tyr and His is close to the peak of Trp. If the amount Tyr and His in the real samples is higher than their detection limit, they will affect the quantification of Trp. However, the peak of Tyr and His was not found in the hepatocyte extract and individual hepatocytes. Therefore, the separation of Trp, Tyr, and His was not investigated. The electropherogram at the Au/Hg electrode of the dual electrode is much simpler. Only two well-separated peaks are observed, which correspond to Cys and GSH, respectively. This means that these substances do not interfere with the determination of Trp and GSH when the dual electrode is used. Due to other amino acids besides Cys, Tyr and
Figure 2. Electropherograms of a solution containing 3.00 × 10-6 mol/L serotonin (5-HT), dopamine (DA), epinephrine (E), and norepinephrine (NE), 1.70 × 10-5 mol/L dopa, 1.43 × 10-5 mol/L Trp, 3.40 × 10-5 mol/L tyrosine (Tyr), 5.00 × 10-5 mol/L cysteine (Cys), histidine (His), and GSH detected by the dual electrode. (1) at the carbon fiber bundle electrode; (2) at the Au/Hg electrode. 0.025 mol/L Na2HPO4-NaH2PO4 containing 4.0 × 10-4 mol/L Na2H2EDTA (pH 7.5). Other conditions are the same as in Table 1.
Figure 3. Electropherograms of (1, 3) the blank solution consisted of running buffer and SSA and (2, 4) the hepatocyte extract detected by the dual electrode: (1, 2) at the carbon fiber bundle electrode; (3, 4) at the Au/Hg electrode. U, unknown peaks. 0.025 mol/L Na2HPO4NaH2PO4 containing 4.0 × 10-4 mol/L Na2H2EDTA (pH 7.5). Other conditions are the same as in Table 1.
His are electroinactive for the dual electrode, and they should not interfere with the determination of Trp and GSH. Figure 3 shows the electropherograms of the blank solution consisted of the running buffer and SSA and the hepatocyte extract. Two peaks appear on both electropherograms of the blank solution shown in curves 1 and 3 at both electrodes of the dual electrode. There are seven peaks on the electropherogram of the hepatocyte extract recorded at the carbon fiber bundle electrode (curve 2). By comparing curve 2 with curve 1 and Figure 2, curve 1, the peak on curve 2, eluting at 4.45 min, can be identified as the peak of Trp on the basis of the migration time. In the hepatocyte extract, four electrophoretic peaks were detected by the Au/Hg electrode (curve 4). By comparing curve 4 with curve 3 and Figure 2, curve 2, the peak on curve 4, eluting at 7.30 min, can be identified as the peak of GSH on the basis of the migration time. To affirm the peaks of Trp and GSH, the standard solution
containing Trp and GSH was added to the hepatocyte extract. The peak current of Trp and GSH detected by the dual electrode was increased. This can also prove that these two peaks correspond to Trp and GSH, respectively. In addition, the two electrophoretic peaks shown in curves 2 and 4, eluting at 4.9 min, can be identified as the peak of Cys based on their migration time and the ratio of their peak current on both electrodes of the dual electrode. Other peaks noted U on the two electropherograms are not identified, because identifying and quantifying them are difficult for such a complex biologic system without enough biochemical knowledge and standard reagents. The concentrations of Trp and GSH in the hepatocyte extract obtained by the standard calibration method are 7.47 × 10-6 mol/L for Trp and 1.63 × 10-5 mol/L for GSH. To prove the reliability of the method, a certain amount of standard Trp and GSH was added to the hepatocyte extract and then the extract was measured. From the detected concentrations in the hepatocyte extract with and without the standard Trp and GSH, the recoveries calculated are 106% for Trp and 103% for GSH, respectively. The amount of Trp and GSH in the hepatocyte extract of 1.10 mL containing 2.0 × 106 cells can be calculated to be 8.2 × 10-9 and 1.8 × 10-8 mol, respectively. Thus, the mean mass of Trp and GSH in a single hepatocyte was 4.1 and 9.0 fmol, respectively. Detection of Trp and GSH in an Individual Hepatocyte. Since hepatocytes are suspended in PBS, PBS is injected into the separation capillary with a hepatocyte during the cell injection. It was found that when the separation and detection conditions for analysis of hepatocyte extract mentioned above were used, the peaks of the blank solution containing PBS and Trp overlapped. To separate both peaks, 0.025 mol/L Na2HPO4 (pH 8.0) containing 1.0 × 10-4 mol/L Na2H2EDTA was used. In this running buffer, both peaks could be well separated and the mass linear ranges of Trp and GSH at the dual electrode are 2.29-115 fmol with a correlation coefficient of 0.999 for Trp and 7.15-143 fmol with a correlation coefficient of 0.998 for GSH. The electropherograms of the blank solution (injecting the cell suspension containing PBS and then 0.010 mol/L NaOH) and one individual hepatocyte at the dual electrode are shown in Figure 4. Four peaks appear on the electropherogram of the hepatocyte (curve 2) detected by the carbon fiber bundle electrode of the dual electrode. The first peak has the same migration time as that of the blank peak shown in curve 1. The second peak, eluting at 4.37 min, can be identified as Trp according to its migration time as compared with the electropherogram of the standard Trp. The third peak is the peak of Cys on the basis of its migration time. However, the fourth peak cannot be identified. There are three peaks on the electropherogram of the hepatocyte (curve 4) detected by the Au/Hg electrode of the dual electrode. The first peak is the blank peak, by comparison with curve 3. The second peak can be identified as Cys on the basis of its migration time and the ratio of its peak current at the Au/Hg electrode to the peak current of Cys detected at the carbon fiber bundle electrode shown in curve 2. GSH is responsible for the third peak, eluting at 7.02, as its migration time is the same as that of the peak of the standard GSH. In almost all experiments of single-cell analysis for whole-cell injection, the migration time of the analytes detected was prolonged and the number of theoretical plates decreased with increasing run numbers. The accumulation of the substances in Analytical Chemistry, Vol. 75, No. 15, August 1, 2003
3863
Table 2. Values of ip, tm, and N on the Electropherograms and Amount of Trp and GSH in the Six Hepatocytesa
a
cell
ip,Trp (pA)
ip,GSH (pA)
tm,Trp (min)
tm,GSH (min)
NTrp (104)
NGSH (104)
amounttrp (fmol)
amountgsh (fmol)
1 2 3 4 5 6
106 79.4 41.0 85.3 61.2 63.9
13.1 12.7 8.33 12.0 14.4 15.2
4.38 4.42 4.42 4.52 4.38 4.37
7.03 7.07 7.08 7.25 7.07 7.02
12 12 12 11 12 11
9.6 9.7 9.2 9.1 9.7 9.2
7.39 5.47 2.71 5.90 4.16 4.36
11.5 11.2 8.23 10.7 12.4 12.9
Conditions: 0.025 mol/L Na2HPO4 containing 4.0 × 10-4 mol/L Na2H2EDTA (pH 8.0). Other conditions as in Table 1.
The reproducible peak currents, together with the large linear dynamic range for standard Trp and GSH, made it suitable to use external standardization for the quantification of both in individual hepatocytes. Table 2 shows the amounts of Trp and GSH determined consecutively in six individual hepatocytes. It is observed that the amounts of Trp and GSH in single hepatocytes differ from cell to cell. The amounts of Trp and GSH determined in the six cells are 2.71-7.39 fmol with a mean value of 5.00 fmol for Trp and 8.23-12.9 fmol with a mean value of 11.2 fmol for GSH, respectively. The corresponding values in the hepatocyte extract (4.1 fmol for Trp and 9.0 mol/ for GSH) are within the range of values found in the present single-cell analysis. The relative mean level of GSH to Trp determined for the six cells is 2.2:1, which is identical to that found in the hepatocyte extract. The amount of GSH detected here is somewhat lower than the macroscopic value (14 fmol) reported in the literature.32 Naturally, the cells from different ethnic groups and different rats can lead to different results.
Figure 4. Electropherograms of (1, 3) the blank solution and (2, 4) one hepatocyte detected by the dual electrode: (1, 2) at the carbon fiber bundle electrode; (3, 4) at the Au/Hg electrode. U, unknown peaks. 0.025 mol/L Na2HPO4 containing 4.0 × 10-4 mol/L Na2H2EDTA (pH 8.0). Other conditions are the same as in Table 1.
the cells adsorbed on the inner surface of the capillary is probably responsible for this. To solve this problem, after analyzing each cell, the injection end of the capillary was treated as follows: The injection end was cleaned for 3 min in water by the ultrasonicator. Then a small amount of 0.1 mol/L NaOH was injected into the capillary with a syringe. After 3 min, the NaOH solution was drawn out from the injection end. Finally, the capillary was flushed with the running buffer for ∼5 min by using the syringe. The results of analysis for six single hepatocytes are listed in Table 2. It can be found that the migration time and the number of theoretical plates do not change further with increasing run numbers. It is noted that the migration times of Trp and GSH for the fourth cell are longer than that of the other cells. A possible reason is that the position of the cell in the capillary before lysis is different from other cells. 3864 Analytical Chemistry, Vol. 75, No. 15, August 1, 2003
CONCLUSION Capillary electrophoresis with electrochemical detection at a dual electrode is a valuable technique for the analysis of biochemical substances that are oxidized or reduced at different kinds of electrode material. The carbon fiber bundle-Au/Hg dual electrode constructed here is a good example. In this scheme, different potentials are applied to the two electrodes of the dual electrode. Trp and GSH can be identified and quantified simultaneously by the dual electrode in one run. Other compounds such as Cys, Tyr, His, dopa, DA, 5-HT, epinephrine, and norepinephrine can be determined simultaneously with Trp and GSH at the dual electrode. This dual electrode can be used simultaneously to determine Trp and GSH in single hepatocytes, which cannot be carried out by using single electrodes. We believe that the dual electrode in capillary electrophoresis with electrochemical detection will be a useful tool for single-cell analysis. ACKNOWLEDGMENT This project was supported by the National Natural Science Foundation of China (No. 20235010). The authors thank Dr. Hao Li for assistance in preparing the manuscript. Received for review November 12, 2002. Accepted February 7, 2003. AC0207022 (32) Mertens, K.; Rogiers, V.; Sonck, W.; Vercruysse, A. J. Chromatogr.. B 1991, 565, 149-157.
