Cell cycle-dependent characterization of single MCF-7 breast cancer cells by 2-D CE

The composition of single MCF-7 breast cancer cells is characterized using 2-D CE. Individual MCF-7 cells were aspirated into a 30 mum inner diameter fused-silica capillary and lysed by contact with an SDS-containing buffer. Proteins and other primary amines were fluorescently labeled on-column using the fluorogenic dye 3-(2-furoyl)quinoline-2-carboxaldehyde. Labeled components were separated first according to molecular weight using capillary sieving electrophoresis (CSE) and then by MEKC. Analytes were detected in a sheath-flow cuvette using LIF. The expression profiles for MCF-7 cellular homogenate and a single MCF-7 cell are compared. As a proof-of-principle investigation, variation in expression was also compared within and between G1 and G2/M cell cycle phases for MCF-7 cells. Following their treatment with the viable nuclear stain Hoechst 33342, MCF-7 cells were sorted by flow cytometry on the basis of their ploidy. Sorted cells were then analyzed by 2-D CE. The degree of variability was >2.5 times larger between cells of different phases than between cells of the same phase. In typical 1 h 2-D CE separations using MCF-7 cells, over 100 components are resolved.     

Migration behavior and separation of active components in Glycyrrhiza uralensis Fisch and its commercial extract by micellar electrokinetic capillary chromatography

1-Anilinonaphthalene-8-sulfonic acid (1,8-ANS), 4,4'-dianilino-1,1'-binaphthyl-5,5'-disulfonic acid (bis-ANS) and 2-(p-toluidino)naphthalene-6-sulfonic acid (2,6-TNS) were evaluated as additives in different buffers for the detection of bovine whey proteins using laser-induced fluorescence (LIF) monitoring in capillary electrophoresis (CE). These N-arylaminonaphthalene sulfonates furnish a large fluorescence emission when associated to some proteins whereas their emission in aqueous buffers, such as those used in CE separations, is very small. To select the best detection conditions, the fluorescence of these probes was first compared using experiments carried out in a fluorescence spectrophotometer. Using bovine serum albumin (BSA) as a model protein, it was demonstrated that 2-(N-cyclohexylamino)ethanesulfonic acid (CHES) buffer (pH 8 and pH 10.2) and the fluorescent probe 2,6-TNS gave rise to the highest increase in fluorescence for BSA. When the composition of these separation buffers was optimized for the electrophoretic separations, CHES buffer, pH 10.2 was chosen as the most suitable buffer to detect bovine whey proteins. The limit of detection obtained for some whey proteins in CE separations was about 6.10(-8) M for BSA, 3.10(-7) M for beta-lactoglobulin A (beta-LGA), 3.10(-7) M for beta-lactoglobulin B (beta-LGB), and 3.10(-6) M for alpha-lactalbumin (alpha-LA). These detection limits were compared to those achieved using UV detection under the same separation conditions. The results showed that the detection limits of BSA, beta-LGA and beta-LGB were twice as good using LIF than with UV detection. However, the limit of detection for alpha-LA was better when UV was used. The applicability of LIF detection to CE separation of whey proteins in bovine milk samples was also demonstrated.     

Selective determination of cadmium(II) from divalent metal ions in environmental samples by capillary electrophoresis using in-capillary complexation with a lacunary Keggin-type [PW11O39]7- complex.

A novel capillary electrophoretic (CE) method, based on in-capillary complexation with [PW(11)O(39)](7-), was developed for the determination of cadmium(II) in natural water samples. When a sample solution is injected into a capillary containing 0.20 mM [PW(11)O(39)](7-) and 0.10 M malonate buffer (pH 3.0), the ternary Keggin-type complex, [P(Cd(II)W(11))O(39)](5-), which possesses high molar absorbtivities in the UV region, is formed in the capillary, and its migration toward the anode gives a well-defined migration peak in the electropherogram. An advantage of this method is that many divalent metal ions do not interfere. The proposed method was successfully applied to the determination of Cd(II) in environmental samples. The detection limits were 1 x 10(-7) and 5 x 10(-7) M for river-water and seawater samples, respectively (signal-to-noise ratio = 3).     

Design, characterization, and utilization of a fast fluorescence derivatization reaction utilizing o-phthaldialdehyde coupled with fluorescent thiols

We have developed a chemical derivatization scheme for primary amines that couples the fast kinetic properties of o-phthaldialdehyde (OPA) with the photophysical properties of visible, high quantum yield, fluorescent dyes. In this reaction, OPA is used as a cross-linking reagent in the labeling reaction of primary amines in the presence of a fluorescent thiol, 5-((2-(and-3)-S-(acetylmercapto)succinoyl)amino)fluorescein (SAMSA fluorescein), thereby incorporating fluorescein (epsilon = 78 000 M(-1), quantum yield of 0.98) into the isoindole product. Detection is based on excitation and emission of the incorporated fluorescein using the 488 nm laser line of an Ar(+) laser rather than the UV-excited isoindole, thereby eliminating the UV light sources for detection. Using this method, we have quantitatively labeled biologically important primary amines in less than 10 s. Detection limits for analysis of glutamate, glycine, GABA, and taurine were less than 2 nM. We present the characterization of OPA/SAMSA-F reaction and the potential utility of the derivatization reaction for dynamic chemical monitoring of biologically relevant analytes using CE.     

Reaction of fluorogenic reagents with proteins III. Spectroscopic and electrophoretic behavior of proteins labeled with Chromeo P503

The spectroscopic and electrophoretic properties of proteins labeled with Chromeo P503 were investigated. Its photobleaching characteristics were determined by continually infusing Chromeo P503-labeled alpha-lactalbumin into a sheath-flow cuvette and monitored fluorescence as a function of laser power. The labeled protein is relatively photo-labile with an optimum excitation power of about 2 mW. The unreacted reagent is weakly fluorescent but present at much higher concentration than the labeled protein. The unreacted reagent undergoes photobleaching at a laser power more than an order of magnitude higher than the labeled protein. One-dimensional capillary electrophoresis analysis of Chromeo P503-labeled alpha-lactalbumin produced concentration detection limits (3sigma) of 12 pM and mass detection limits of 0.7 zmol, but with modest theoretical plate counts of 17,000. The reagent was employed for the two-dimensional capillary electrophoresis analysis of a homogenate prepared from a Barrett's esophagus cell line; the separation quality is similar to that produced by 3-(2-furoyl)quinoline-2-carboxaldehyde (FQ), a more commonly used reagent.     

Optimization of the separation and on-line sample concentration of phenethylamine designer drugs with capillary electrophoresis-fluorescence detection

Five 2C-series of phenethylamine designer drugs, including 2,5-dimethoxy-4-ethylthio-phenethylamine (2C-T-2), 2,5-dimethoxy-4-(n)-propylthiophenethylamine (2C-T-7), 4-chloro-2,5-dimethoxyphenethylamine (2C-C), 4-bromo-2,5-dimethoxy-phenethylamine (2C-B), 2,5-dimethoxy-4-iodo-phenethylamine (2C-I), were synthesized and standard GC/MS and fluorescence spectra are reported for them. A mixture of the five drugs was separated and detected by means of capillary electrophoresis (CE) with native fluorescence and light emitting diode (LED)-induced fluorescence (LIF) detection, respectively, for comparison. In the former case, exciting at a wavelength of 300 nm from a Xe lamp was used. The detection limits were found to be only in the range of approximately 10(-4) M by the micellar electrokinetic chromatography (MEKC) mode but were improved to approximately 10(-7) M when the sweeping-MEKC mode was used. For a highly sensitive analysis, LED-induced fluorescence detection was examined by derivatizing the compounds with a fluorescent dye, fluorescein isothiocyanate isomer I (FITC). A blue-LED (approximately 2 mW) was used as the fluorescence excitation source. The detection limits were improved to approximately 10(-7) and approximately 10(-8) M, respectively, when the MEKC and stacking-MEKC modes were applied. A mimic urine sample was obtained by spiking urine from a volunteer with the five standards, and after liquid-liquid extraction, the sample was examined by means of the MEKC-LIF mode. The extraction procedures used for the urine sample and the CE conditions for the separation were optimized.     

Determination of oxoanions in river water by capillary electrophoresis

A new capillary electrophoresis (CE) procedure was developed for simultaneous determination of ten oxoanions (CrO4(2-), SeO4(2-), MoO4(2-), WO4(2-), VO4(3-), SeO3(2-), As04(3-), TeO3(2-), TeO4(2-), and AsO3(3-)) which were baseline-separated from each other and from the interfering UV absorbing anions (NO3- and NO2-) commonly found in environmental water samples. The new background electrolyte system developed contained 5 mM potassium phosphate and 0.007 mM octadecyltrimethylammonium hydroxide, pH 11.2. The optimized working conditions were electrokinetic sampling at -5 kV for 10 s, running voltage at -15 kV with 5 microA current, and detection wavelength at 205 nm. No interference was observed for non-UV-absorbing anions and UV-absorbing anions up to 20 and 10 times higher concentrations respectively. The speed of analysis was fast, with a complete CE run within 6 min. Wide linear ranges (1-2,000 microg/L), good repeatability in migration time (relative standard deviation RSD 0.55-2.8%), satisfactory precision in peak area (RSD 3.8-5.6%) and peak height (RSD 3.9-5.3%) measurement, and detection limits (1-25 microg/L) sufficiently sensitive to detect oxoanions found in environmental water samples were obtained. The reliability of the CE procedure developed had been established by recovery test and parallel method determination using atomic absoprtion spectrophotometry for real river water sample.     

Determination of ebrotidine and its metabolites by capillary electrophoresis with UV and mass spectrometry detection

This study describes the application of capillary electrophoresis (CE) to the analysis of ebrotidine and its metabolites as an alternative analytical technique to liquid chromatography. Comparison between UV-diode array spectroscopy and mass spectrometry (MS) using an ion-trap system with electrospray ionization as detection systems has been performed. The quality parameters of the UV detection method were established, obtaining linear calibration curves over the range studied (8-200 mg ml(-1)), limits of detection between 3.4 and 4.3 microg ml(-1), and run-to-run and day-to-day precision lower than 14%. For these compounds the protonated species [M+H]+ and, in some cases, sodium adducts were observed in the MS spectra. Using MS coupled to CE, limits of detection were between 0.5 and 2.6 microg ml(-1).     

Low-cost, high-sensitivity laser-induced fluorescence detection for DNA sequencing by capillary gel electrophoresis.

A low cost, 0.75-mW helium neon laser, operating in the green region at 534.5 nm, is used to excite fluorescence from tetramethylrhodamine isothiocyanate-labelled DNA fragments that have been separated by capillary gel electrophoresis. The detection limit (3 sigma) for the dye is 500 ymol [1 yoctomole (1 ymol) = 10(-24) mol] or 300 analyte molecules in capillary zone electrophoresis; the detection limit for labeled primer separated by capillary gel electrophoresis is 2 zmol [1 zeptomole (1 zmol) = 10(-21) mol]. The Richardson-Tabor peak-height encoded sequencing technique is used to prepare DNA sequencing samples. In 6% T, 5% C acrylamide, 7 M urea gels, sequencing rates of 300 bases/hour are produced at an electric field strength of 200 V/cm; unfortunately, the data are plagued by compressions. These compressions are eliminated with addition of 20% formamide to the sequencing gel; the gel runs slowly and sequencing data are generated at a rate of about 70 bases/hour.     

Immunoassays using capillary electrophoresis laser induced fluorescence detection for DNA adducts

Human DNA is exposed to a variety of endogenous and environmental agents that may induce a wide range of damage. The critical role of DNA damage in cancer development makes it essential to develop highly sensitive and specific assays for DNA lesions. We describe here ultrasensitive assays for DNA damage, which incorporate immuno-affinity with capillary electrophoresis (CE) separation and laser induced fluorescence (LIF) detection. Both competitive and non-competitive assays using CE/LIF were developed for the determination of DNA adducts of benzo[a]pyrene diol epoxide (BPDE). A fluorescently labeled oligonucleotide containing a single BPDE adduct was synthesized and used as a fluorescent probe for competitive assay. Binding between this synthetic oligonucleotide and a monoclonal antibody (MAb) showed both 1:1 and 1:2 complexes between the MAb and the oligonucleotide. The 1:1 and 1:2 complexes were separated by CE and detected with LIF, revealing binding stoichiometry information consistent with the bidentate nature of the immunoglobulin G antibody. For non-competitive assay, a fluorescently labeled secondary antibody fragment F(ab′)₂ was used as an affinity probe to recognize a primary antibody that was specific for the BPDE-DNA adducts. The ternary complex of BPDE-DNA adducts with the bound antibodies was separated from the unbound antibodies using CE and detected with LIF for quantitation of the DNA adducts. The assay was used for the determination of trace levels of BPDE-DNA adducts in human cells. Analysis of cellular DNA from A549 human lung carcinoma cells that were incubated with low doses of BPDE (32 nM–1 μM) showed a clear dose–response relationship. BPDE is a potent environmental carcinogen, and the ultrasensitive assays for BPDE-DNA adducts are potentially useful for monitoring human exposure to this carcinogen and for studying cellular repair of DNA damage.     

Determination of different forms of human interferon-gamma in single natural killer cells by capillary electrophoresis with on-capillary immunoreaction and laser-induced fluorescence detection

A novel method for determining different forms of human interferon-gamma (IFN-gamma) in single natural killer cells was developed by capillary electrophoresis (CE) with on-capillary immunoreaction and laser-induced fluorescence (LIF) detection. Cells were perforated with digitonin and one single cell was electrokinatically introduced into the front end of a separation capillary. The monoclonal antibody labeled with fluorescein isothiocyanate of IFN-gamma was hydrodynamically injected into the front end of the capillary around the cell introduced. After the cell was lysed by ultrasonication, the front end of the capillary was used as a microreactor to allow different forms of IFN-gamma to process the immunoreaction with their labeled antibody. Finally, the complexes of different forms of IFN-gamma with their labeled antibody were separated and detected by CE with LIF detection with a limit of detection of zeptomoles (10(-21) mol).     

High-sensitivity detection of biological amines using fast Hadamard transform CE coupled with photolytic optical gating

Here, we report the first utilization of Hadamard transform CE (HTCE), a high-sensitivity, multiplexed CE technique, with photolytic optical gating sample injection of caged fluorescent labels for the detection of biologically important amines. Previous implementations of HTCE have relied upon photobleaching optical gating sample injection of fluorescent dyes. Photolysis of caged fluorescent labels reduces the fluorescence background, providing marked enhancements in sensitivity compared to photobleaching. Application of fast Hadamard transform CE (fHTCE) for fluorescein-based dyes yields a ten-fold higher sensitivity for photolytic injections compared to photobleaching injections, due primarily to the reduced fluorescent background provided by caged fluorescent dyes. Detection limits as low as 5 pM (ca. 18 molecules per injection event) were obtained with on-column LIF detection using fHTCE in less than 25 s, with the capacity for continuous, online separations. Detection limits for glutamate and aspartate below 150 pM (1-2 amol/injection event) were obtained using photolytic sample injection, with separation efficiencies exceeding 1 x 10(6) plates/m and total multiplexed separation times as low as 8 s. These results strongly support the feasibility of this approach for high-sensitivity dynamic chemical monitoring applications.     

Design and development of microfabricated capillary electrophoresis devices with electrochemical detection

Our efforts have been focused on developing a self-contained and transportable microfabricated electrophoresis (CE) system with integrated electrochemical detection (ED). The current prototype includes all necessary electrodes “on-chip” and utilizes miniaturized CE and ED supporting electronics custom designed for this purpose. State-of-the-art design/modeling tools and novel microfabrication procedures were used to create recessed platinum electrodes with complex geometries and the CE/ED device from two patterned ultra-flat glass substrates. The electrodes in the bottom substrate were formed by a self-aligned etch and deposition technique followed by a photolithographic lift-off process. The microchannels (20 μm deep×65 μm wide (average)) were chemically etched into the top substrate followed by thermal bonding to complete the microchip device. CE/ED experiments were performed using 0.02 M phosphate buffer (pH 6), an analyte/buffer solution (2.2 mM dopamine, 2.3 mM catechol) and varying separation voltages (0-500 V) with a custom electronics unit interfaced to a laptop computer for data acquisition. Detection limits (S/N=3) were found to be at the micromolar level and a linear detection response was observed up to millimolar concentrations for dopamine and catechol. The microchip CE/ED system injected 50 pl volumes of sample, which corresponded to mass detection limits on the order of 200 amol. For the first time, an integrated “on-chip” multi-electrode array CE/ED device was successfully designed, fabricated and tested. The majority of the electrodes (six out of eight) in the array were capable of detecting dopamine with the amplitude of the signal (i.e., peak heights) decreasing as the electrode distance from the channel exit increased.     

Capillary electrophoretic determination of phosphate based on the formation of a Keggin-type [PMo12O40]3- complex.

Based on the formation of a Keggin-type [PMo12O40]3- complex, a sensitive capillary electrophoresis (CE) method was developed for the determination of P(V) with direct UV detection at 220 nm. A mixture of alpha- and beta-Keggin-type [PMo12O40]3- complexes was readily formed in a sample solution consisting of a trace amount of P(V), 2.5 mM Mo(VI), 0.050 M p-C6H3(CH3)-2-SO3H (XSA), and 60% v/v CH3CN. When a 0.05 M HCl and 60% v/v CH3CN solution was used as a migration electrolyte, the Keggin complexes exhibited a sharp and well-defined peak in the electropherogram. The peak area was linearly dependent on the P(V) concentration in the range of 5 x 10(-7)-5 x 10(-5) M; a detection limit of 1 x 10(-7) M was achieved. In comparison with indirect UV detection, the direct UV detection is about ten times more sensitive, because the Keggin complexes possess high molar absorptivities. The developed CE method was applied to the determination of P(V) in river water, and the results were in good agreement with those obtained by ion chromatography (IC) and colorimetry (COL) based on the formation of mixed-valence heteropoly blue species.     

Sensitive determination of sulfonamides in environmental water by capillary electrophoresis coupled with both silvering detection window and in‐capillary optical fiber light‐emitting diode‐induced fluorescence detector

A new detector, silvering detection window and in‐capillary optical fiber light‐emitting diode‐induced fluorescence detector (SDW‐ICOF‐LED‐IFD), is introduced for capillary electrophoresis (CE). The strategy of the work was that half surface of the detection window was coated with silver mirror, which could reflect the undetected fluorescence to the photomultiplier tube to be detected, consequently enhancing the detection sensitivity. Sulfonamides (SAs) are important antibiotics that achieved great applications in many fields. However, they pose a serious threat on the environment and human health when they enter into the environment. The SDW‐ICOF‐LED‐IFD‐CE system was used to determine fluorescein isothiocyanate (FITC)‐labeled sulfadoxine (SDM), sulfaguanidine (SGD) and sulfamonomethoxine sodium (SMM‐Na) in environmental water. The detection results obtained by the SDW‐ICOF‐LED‐IFD‐CE system were compared to those acquired by the CE with in‐capillary optical fiber light‐emitting diode‐induced fluorescence detection (ICOF‐LED‐IFD‐CE). The limits of detection (LODs) of SDW‐ICOF‐LED‐IFD‐CE and ICOF‐LED‐IFD‐CE were 1.0–2.0 nM and 2.5–7.7 nM (S/N = 3), respectively. The intraday (n = 6) and interday (n = 6) precision of migration time and corresponding peak area for both types of CE were all less than 0.86% and 3.68%, respectively. The accuracy of the proposed method was judged by employing standard addition method, and recoveries obtained were in the range of 92.5–102.9%. The results indicated that the sensitivity of the SDW‐ICOF‐LED‐IFD‐CE system was improved, and that its reproducibility and accuracy were satisfactory. It was successfully applied to analyze SAs in environmental water.     

Electrochemiluminescence quenching as an indirect method for detection of dopamine and epinephrine with capillary electrophoresis

An electrochemiluminescence (ECL) inhibition method was developed as an indirect detection method for the determination of dopamine and epinephrine separated by capillary electrophoresis (CE). When the concentration of Ru(bpy)(3) (2+) was 50 muM diluted by 50 mM phosphate (pH 8.5) in the cell and 0.5 M tripropylamine (TPA) was added to the running buffer (10 mM phosphate, pH 9.0), an inhibition of ECL of the Ru(bpy)(3) (2+)/TPA system by the analytes was observed. Under the optimized conditions, the relative standard deviations of migration time and negative peak area were less than 1% and 3%, respectively, for 1 microM dopamine or 1 microM epinephrine (n = 10). Linear ranges of 0.1-10 microM for both analytes and the detection limits (signal-to-noise ratio S/N = 3) of 10 nM for dopamine and 30 nM for epinephrine were obtained.     

高效毛细管电泳-电化学检测对多羟基抗生素分离测定的研究

用自制的高效毛细管电泳-电化学检测装置,以350μm铜圆盘电极为工作电极,在碱性溶液中,对7种多羟基抗生素的分离进行了研究。结果表明:浓度线性范围可达2至3个数量级,最低检测限为5.0×10^-7mol/L.在8h内连续测定,峰电流和迁移时间的相对标准偏差分别为7.6%和2.2%.实际样品(丁胺卡那霉素)测定回收率为99.6%.     

Determination of ascorbic acid in individual rat hepatocyte cells based on capillary electrophoresis with electrochemiluminescence detection

A novel method to detect ascorbic acid (AA) in individual rat hepatocyte cells was developed by combining CE with electrochemiluminescence (ECL) based on tris(2,2'-bipyridine) ruthenium(II) (Ru(bpy)(3)2+). A single cell, followed by 0.1% SDS as cell lysis solution, was injected into the inlet of the separation capillary by electromigration. After optimizing the analytical conditions, the RSDs of migration time and peak height were 0.38% and 2.6% for 1.0x10(-5) M AA (n=10), respectively. The linear range of AA was from 1.0x10(-8) to 5.0x10(-5) M with a correlation coefficient of 0.9979 and the LOD (S/N=3) was estimated to be 1.0x10(-8) M. This method has been successfully applied to determine AA in single rat hepatocytes and the amount of AA in seven rat hepatocytes ranged from 16 to 62 fmol. The above results demonstrated that CE coupled with ECL is convenient, sensitive, and will become an attractive alternative method for single-cell analysis.     

Capillary sieving electrophoresis-micellar electrokinetic chromatography fully automated two-dimensional capillary electrophoresis analysis of Deinococcus radiodurans protein homogenate

We report the one- and two-dimensional (1-D and 2-D) capillary electrophoresis separation of Deinococcus radiodurans protein homogenate. Proteins are labeled with the fluorogenic reagent 3-(2-furoyl)quinoline-2-carboxaldehyde (FQ), which reacts with lysine residues and creates a highly fluorescent product. Detection is by laser-induced fluorescence. 1-D capillary sieving electrophoresis (CSE) produces over 150,000 plates and micellar electrokinetic capillary chromatography (MEKC) produces over 900,000 plates for components in a D. radiodurans protein homogenate. In a 2-D separation, proteins are first separated by CSE. Fractions are repetitively transferred to a second capillary for further separation based on MEKC. The 2-D separation has a approximately 550 spot capacity. Over 150 components are partially resolved from the homogenate. Resolution is limited in the first dimension by diffusion of proteins during the long separation period and in the second dimension by the combination of a long fraction-transfer time and short separation period.     

Integration of single cell injection, cell lysis, separation and detection of intracellular constituents on a microfluidic chip

A microfluidic system was developed for the analysis of single biological cells, with functional integration of cell sampling, single cell loading, docking, lysing, and capillary electrophoretic (CE) separation with laser induced fluorescence (LIF) detection in microfabricated channels of a single glass chip. Channels were 12 microm deep and 48 microm wide, with a simple crossed-channel design. The effective separation channel length was 35 mm. During sampling with a cell suspension (cell population 1.2 x 10(5) cells per mL in physiological salt solution), differential hydrostatic pressure (created by adjusting liquid levels in the four reservoirs) was used to control cell flow exclusively through the channel crossing. Single cell loading into the separation channel was achieved by electrophoretic means by applying a set of potentials at the four reservoirs, counteracting the hydrostatic flow. A special docking (adhering) procedure for the loaded cell was applied before lysis by repeatedly connecting and disconnecting a set of low potentials, allowing precise positioning of the cell within the separation channel. Cell lysis was then effected within 40 ms under an applied CE separation voltage of 1.4 kV (280 V cm(-1)) within the working electrolyte (pH 9.2 borate buffer) without additional lysates. The docked lysing approach reduced dispersion of released intracellular constituents, and significantly improved the reproducibility of CE separations. Glutathione (GSH) was used as a model intracellular component in single human erythrocyte cells. NDA derivatized GSH was detected using LIF. A throughput of 15 samples h(-1), a retention time precision of 2.4% RSD was obtained for 14 consecutively injected cells. The average cellular concentration of GSH in human erythrocytes was found to be 7.2 [times] 10(-4)+/- 3.3 x 10(-4) M (63 +/- 29 amol per cell). The average separation efficiency for GSH in lysed cells was 2.13 x 10(6)+/- 0.4 x 10(6) plates per m, and was about a factor of 5 higher than those obtained with GSH standards using pinched injection.