Microtiter plate assay for the measurement of glutathione and glutathione disulfide in large numbers of biological samples

ANALYTICAL

BIOCHEMISTRY

190,360-365 (1990)

Microtiter Plate Assay for the Measurement of Glutathione and Glutathione Disulfide in Large Numbers of Biological Samples Margaret

A. Baker,l George J. Cerniglia, and Aziza Zaman

Department

of Radiation Oncology, University of Pennsylvania, Philadelphia, Pennsylvania 19104

Received

June

7,199O

By combining the least complicated and expedient methods of sample handling with the sensitivity and specificity of the GSH assay by enzymatic recycling and the small volumes and software capabilities of microtiter plate technology we have devised a rapid, sensitive, and easy assay for GSH and GSSG in biological samples. The assay is sensitive to 5 pmol in sample volumes of 50 pl, although other volumes could be used. The use of a computer-driven microplate with software capable of linear kinetic data storage and analysis on each well, Maxline series microplate readers and Softmax software, enables the user not only to assay large numbers of samples per day but also to have immediate calculated results. We suggest by examples that measurements of total GSH as well as changes in GSH:GSSG in vitro and in viuo are feasible with this technology. 0 1990 Academic Press, Inc.

Glutathione is present in all mammalian cells at millimolar concentrations and serves several cellular functions including amino acid transport, maintenance of protein sulfhydryls reduction status, and as defense against oxidizing molecules and electrophilic xenobiotits. These functions are performed through hydrogen donation from the thiol moiety or by conjugation reactions. Depletion of intracellular GSH seems not to be, strictly speaking, lethal. However, depletion of GSH followed by or in addition to a challenge by other oxidative reactions can lead to irreversib!e cell damage. The latter is presently understood as the mechanism of acetaminophen-induced liver failure (1). Tissue defense

1 To whom correspondence and reprint dressed at: Division of Oncology Research, Philadelphia, PA 19104-6072.

request& 195 John

should Morgan

be adBldg.,

against reactive oxygen species plays a primary role in a number of disease states including arthritis, inflammation, reperfusion injury, and aging as well as defense against ionizing radiation and therapeutic drugs. The measurement of GSH and its disulfide form, GSSG, therefore, provides information about cellular defense and cellular response to these challenges. Numerous procedures have been developed for the measurement of GSH: chemical, enzymatic, high-pressure chromatographic, and more recently, flow cytometric. These assays use a variety of detection systems including ultraviolet, visible, fluorescent, and electrochemical. Several methods are quite sensitive but suffer from complicated sample preparation procedures and/ or require specialized equipment. We have adapted the enzymatic recycling method first described by Owens and Belcher (2) and later modified by Tietze (3) to the microtiter plate reader, providing an assay which is specific for GSH and/or GSSG and sensitive and enables up to hundreds of assays to be performed in a day. The use of a microplate reader with compatible software having linear kinetic calculation in each well allows for immediate data analysis. Through the coupling of a spin through oil technique to separate intracellular and extracellular GSH pools in tissue culture cells suspensions with the microplate assay, we have studied the rapid changes in intracellular GSH/GSSG under oxidizing conditions (4). We have used the assay to study GSH in nude mouse tissues and suggest that the assay is sensitive enough to be used for appropriately processed biopsy samples. MATERIALS

AND

METHODS

Unless otherwise noted all reagents were purchased from Sigma Chemical Co. (St. Louis, MO). GSH refers to the reduced form of glutathione and GSSG the disulfide dimer. GSx refers to the tripeptide unit of unspeci-

360 All

Copyright 0 1990 rights of reproduction

0003~2697/90 $3.00 by Academic Press, Inc. in any form resewed.

MICROTITER

PLATE

ASSAY

FOR

GLUTATHIONE

tied oxidation state (GSH equivalents). According to the methods recommended by Anderson (5), we have used sulfosalicylic acid (SSA)2 to precipitate cell macromolecules and extract GSH from both cells and tissues.

Cell Sample Preparation A single cell suspension is required if the spinthrough oil technique is to be used. An aliquot of resuspending medium is required to prepare a blank for the assay. Note that cell counts, protein, or DNA content must be determined on an aliquot of the cell suspension which has not been acid treated. For cell suspensions, intracellular and extracellular pools are separated by using a spin through oil technique in 1.5ml microcentrifuge tubes. The tubes are prepared by adding 0.2 ml 4% (w/v) SSA in 20% glycerol and overlaying that with 0.3 ml 8515 silicon oil (Dow-Corning 550 fluid, Midland MI):heavy parafin oil (Fisher Scientific, Fair Lawn, NJ). The density of the oil mixture is temperature dependent. The needs for a particular cell type, size, and environment must be individually determined. The total GSH and/or GSSG can be measured in the cell sample by treating the cell suspension with 0.2 vol of 5% SSA on ice for 15 min and assaying the supernatant fractions after centrifugation. We used the human lung adenocarcinoma cell line (ATCC-A549) grown in RPM1 1640 supplemented with 10% fetal bovine serum as previously described (4). Cells were harvested from flasks using 0.05% trypsin plus EDTA (0.5 mM) and resuspended in growth media, PBS, or saline as required by the experimental protocol. The intracellular GSx content averaged 27.6 + 6.4 (n = 15) nmol/106 cells and the extracellular 8.7 + 3.2 nmol/106 cells for cells suspended in phosphate-buffered saline supplemented with 5 mM glucose as previously described (4). The samples were processed by gently overlaying 1 ml of cell suspension on the oil layer. The tubes were centrifuged at 12,000-13,000g for 5 min in a swing-out rotor such as Fisher Model 59A. A 0.8-ml aliquot of the extracellular, top layer, was acid treated by adding 0.2 vol of 5% SSA on ice for 15 min before centrifuging any precipitated material. After the remaining extracellular material and most of the oil were removed from the tube, the acid/glycerol layer was left for l-2 h at room temperature or refrigerated overnight to extract. A 150-~1 aliquot was removed by piercing a micropipet tip under the remaining oil and avoiding the cell pellet. The separated acid-treated fractions can be stored at -80°C up to 2 months.

’ Abbreviations used: SSA, sulfosalicylic sulfoximine; DTNB, 5,5’-dithiobis(2-nitrobenzoic 2-nitrobenzoic acid.

acid;

BSO, buthionine acid); TNB, B-thio-

AND

GLUTATHIONE

DISULFIDE

361

Tissue Samples Because of the possible rapid loss of or change in oxidation state in intracellular GSH, tissue samples must be processed as quickly as possible (5). Athymic mice (NCE/SED, Nu,Nu) were injected with a colon carcinoma cell line LS-174T subcutaneously on the left flank 2 weeks before treatment with the GSH synthesis inhibitor buthionine sulfoximine (BSO) (6). Nine mice weighing an average of 20.9 f. 1.5 g were injected with 0.5 ml of 0.2 M DL-BSO (5 mmol/kg) i.p. and then given 20 mM DL-BSO in the drinking water, ad libitum. Three mice were sacrificed immediately and three at 24,48, and 72 h post-i.p. injection. The mice were anesthetized with pentabarbital and bled by cardiac puncture. Whole blood was treated with 5 vol/vol of ice-cold 5% SSA. The liver, kidney, and tumors were removed quickly, rinsed in cold saline, and minced into ice-cold 5% SSA (5 vol/g wet tissue). The tissue samples were further processed by sonication (Heat Systems-Ultrasonics Model W-385, Farmingdale, NY, equipped with a microprobe) followed by a 15- to 20-min period of extraction on ice. The samples are then centrifuged at 13,000-20,OOOg for 10 min to remove precipitated material and the supernatant fractions were stored frozen at -80°C. Enzymatic

Recycling Assay

The buffer was 100 mM sodium phosphate and 1 mM EDTA, pH 7.5. For preparation of standards for both the total GSH plus GSSG and GSSG only assays, the same stock solution of 0.2 mg GSSG/ml was used. The stock solution was diluted loo-fold in the appropriate background to give a solution containing 2 rg/ml, 3.27 nmol/ml, GSSG, or 6.54 nmol GSx/ml. Samples were diluted to contain less than 0.15% residual SSA or the pH was adjusted to 7.0 by addition of triethanolamine. We routinely diluted the 20 ~1 of intracellular A549 samples 50-fold, giving 0.08% SSA final. Standards containing from 3 to 320 pmol of GSH or GSSG in 50 ~1 were prepared in a background of buffer and acid identical with the samples and then treated simultaneously with the samples. The samples and standards were kept on ice until being loaded in the microtiter plate. The microtiter plate was prepared by pipetting 50 ~1 of standards, samples and blanks per well. The Softmax software template shows the placement of the blanks (six), standards described in picomoles GSH/well, and samples. The microtiter plate reader (uv,,, Molecular Devices, Palo Alto, CA) instrument settings are single wavelength kinetics using the 405-nm filter, read for 2.0 min, with one initial mixing. At this time the following freshly prepared reagents are mixed at room temperature in a reservoir: 2.8 ml of 1 mM 5,5’-dithiobis(2-nitrobenzoic acid) (DTNB), 3.75 ml of 1 mM NADPH, 5.85 ml of buffer, and 20 U of GSH reductase. Immediately, 0.1 ml was pipetted into each well, the plate placed into

362

BAKER,

CERNIGLIA,

AND

ZAMAN

content of the sample and compared to the appropriate standard curve of slope versus GSx. RESULTS

GSx

(pmollwell)

FIG. 1. A typical

standard curve used to quantitate intracellular or tissue GSH plus GSSG. Each well contained 50 ~1 of standard solution containing 0.08% SSA and the indicated amount of GSSG expressed as GSH equivalents.

the microplate drawer, and the read initialized. The final concentrations of reagents were 0.15 mM DTNB, 0.2 mM NADPH, and 1.0 U GSH reductase/ml. Oxidized Glutathione (GSSG)

The rate of the 5-thio-2-nitrobenzoic acid chromophore (TNB) formed per minute for standards containing GSSG over two orders of magnitude is shown in Fig. 1. The wells contain 50 ~1 of standard solution in 0.08% SSA. The final pH is 7.5 and the final GSH reductase concentration is 0.15 U/well. The regression of rate on GSSG, expressed as GSH equivalents, was AmOD/min = 8.98 + 0.570 (pmol GSH/well) and R2 = 0.995. The rate of TNB chromophore formed by samples processed for assay of GSSG only is shown in Fig. 2. The lower curve was produced from standards made up in 4% SSA, which, therefore, contain more triethanolamine (0.24 M final). The upper curve was produced from standards containing 1% SSA and 0.07 M triethanolamine final. The GSH reductase concentration was 0.15 U per well in both assays. The presence of excess 2-vinylpyridine decreased the reaction rate for GSSG by only about 10% while giving 99% conjugation of GSH (not shown). We also found that the decrease in rate is proportional to the concentration of SSA in the sample as has been noted previously (7). The effect of GSH reductase concentration on the reaction rates for total GSH plus GSSG samples containing 0.1% SSA (slope of the AmOD at 405 nm/min versus pmol GSH/well) is shown in Fig. 3 and the corresponding change in the Y-intercept is shown in Fig. 4. We assayed 24 cell samples, 12 intracellular samples containing 85 to 160 pmol GSH per well, and 12 extracellular samples containing 14 to 40 pmol GSH/well at four GSH reductase concentrations from 0.08 to 0.48 U/well. The resulting sample values had coefficients of

We have measured GSSG in cells treated with a variety of agents known to produce oxidative stress using the enzymatic recycling assay. GSSG produces the identical rate of DTNB reduction as does GSH. Thus, GSSG can be measured only by removal of GSH. We used 2vinylpyridine (7) to conjugate GSH before assay of residual GSSG. The addition of excess 2-vinylpyridine (Aldrich Chemical Co., Milwaukee, WI), 0.35 M final, followed by neutralization of the sample with triethanolamine ensures that oxidation of GSH in the sample does not occur. The standards were made up with the same acid solution as the samples and carried through the conjugation and neutralization process simultaneously. Both sample and standard volumes were 130 ~1 E 60 AmODImin = - 0.0082 + 0.127 to which 5 ~1 of undiluted 97% 2-vinylpyridine was 3 60 added and sufficient 1:4 dilution of triethanolamine in distilled water to bring the pH to 6.7. As above, 50 ~1 of ti; 40 each standard or sample was added to each well and 100 ~1 of the reagent mix as above added to start the reacg 20 tion. The final pH of the mixture in the well for these assays is less than 7.5 depending on the buffer strength z in the sample. Nevertheless, all of the samples can be 0 a in = -0.209 + 0.035 (pmol GSx) assayed in the same plate by using the same reagent R = 0.998 I I I mixture. -20 0 200 400 600 If a plate reader without data storage and analysis GSx (pmollwell) capacity is to be used the linear kinetics for each well can be calculated on optical density data from succes- FIG. 2. Typical standard curves for GSSG processed from stansive reads. The rates are linear over several minutes. dard solutions containing either 1% SSA (e) or 4% SSA (0) and the The slopes only are used to calculate the resulting GSx indicated amount of GSSG expressed as GSH equivalents.

MICROTITER

PLATE

ASSAY

FOR

GLUTATHIONE

AND

GLUTATHIONE

363

DISULFIDE

i : 0.0 0.00

I

0.50

GSH

Reductase

0.0

I

I

0.25

0.1

0.75 (U/well)

FIG. 3. The effect of GSH reductase concentration on the reaction rates of the standard curves (as in Fig. 1). The slopes of the standard curves were plotted against enzyme units per well (added in 0.1 ml).

variation for the four determinations ranging from 1.0 for one sample containing 102 pmol GSH to 20.2 for a sample containing 14.2 pmol GSH. Analysis of the high and low values for each sample showed no relationship to the enzyme concentration used. The effect of GSH reductase concentration on samples containing 1 or 4% SSA and processed with 2-vinylpyridine for GSSG only assay is shown in Fig. 5. The effect of SSA inhibition on the GSH reductase is evident as all samples contained the same 2-vinylpyridine concentration and were adjusted to pH 6.7 with triethanolamine. The Y-intercept for the curves did not increase with increasing GSH reductase used (not shown). To illustrate the rapidity of changes in intracellular GSH:GSSG we show the effect of 300 pM diamide added to an A549 cell suspension (1 X lo6 cell/ml in phosphate buffered saline) on intracellular GSH and GSSG and

0.2 GSH

Reductase

0.3

0.4

0.5

(U/well)

FIG. 5. The effect of GSH reductase concentration on the reaction rates of the standard curves for GSSG only assays carried out with samples treated with 1 or 4% SSA (as in Fig. 2). All of the standards were treated with 0.35 M Z-vinylpyridine but contained no GSH.

extracellular GSH and GSSG (Fig. 6).3 Cells were rapidly spun through oil at the times indicated as shown in Fig. 6. The intracellular GSH was rapidly oxidized and immediately thereafter, GSSG content in the extracellular space increased. The total GSH plus GSSG in the intracellular plus extracellular space was 33.8 f 1.9 nmol per lo6 cells for eight time points sampled between 0 and 120 min. These accounting data show that diamide caused a major shift in the oxidation state and localization of the glutathione species but no loss in total GSx. The effect of DL-BSO treatment on nude mice tissues is shown in Table 1. The tissue extracts were diluted 50to 200-fold to achieve values within the standard curve range shown in Fig. 1. The data show that liver and kidney are maximally affected within 24 h and remain at approximately 33 and 18% of pretreatment values, respectively. The tumor tissue GSH was much lower than the organ concentrations and continued to decline over the three days on treatment reaching 23% of pretreatment values. The tumor tissue weights averaged 173 t- 14 mg at time 0 (6 tumors), 133 + 55 mg (3) at 24 h, 133 +_87 mg (3) at 48 h, and 190 +_106 mg (3) at 72 h.

DISCUSSION

The overall sensitivity of the assay described here is ultimately an order of magnitude greater than that described by Tietze in 1969 (3). Both methods produce a 25-50% increase in the rate of optical density change over background at GSH concentrations as low as 30-50 GSH

FIG. 4. The effect cept of the standard Fig. 3.

Reductase

of GSH reductase curve (background

(U/well)

concentration on the Y-interrate) for the data shown in

3 Some published

of the data figure (4).

shown

were

presented

as part

of a previously

364

BAKER,

-

“0

20

Time

after

GSHin

-h-

GSSGin

-

GSHex

-t-

GSSGer

40

Diamide

CERNIGLIA,

Addition

60

(min)

FIG. 6. The results of the mean for seven experiments showing the changes in intracellular GSH, GSSG, and extracellular GSH and GSSG after challenge of A549 cells in suspension at 1 X lo6 cells/ml with 300 pM diamide. Cells were spun through oil into 4% SSA as described under Materials and Methods at the times indicated.

pmol/ml of reaction mixture. In using a microtiter plate and reader, we have decreased the volume of reagents and more importantly sample needed for measurements as well. Thus, we can measure 5-10 pmol of GSx in 0.15 ml of reaction mixture. As little as 50 mg of biopsy tissue containing 50 nmol of GSx or less could be processed with 5 vol of SSA yielding 300 ~1 of extract at a concentration of at least 167 pmol/pl, which, in this protocol, would be sufficient for multiple determinations. The use of microvolumes may spare other precious samples such as mouse bone marrow cells. The greatest advantage of using the microplate is the inherent ability to assay a large number of samples simultaneously. As many as 80 or 90 samples per plate can be assayed depending on the number of wells used for blanks and standards. Therefore, once the samples have been appropriately diluted or otherwise prepared and the reagent stock solutions made up, hundreds of assays can be performed in an afternoon. The accuracy of the assay depends largely on the accuracy in sample processing and pipetting. Figure 1 shows that the linear fit for 12 standards ranging over two orders of magnitude in GSH concentration is good (R2 = 0.995). The photometric performance of the plate reader used (uv,,) is given as t-+1.0% at the optical density and wavelength used. Therefore, for samples in which GSH is labile such as kidney (5) and the extract must be diluted several hundredfold, processing errors predominate. Similarly, the accuracy of the GSSG only determination depends both on the sensitivity of the assay and on the error introduced by manipulation of the sample. Our preference is for the 2-vinylpyridine derivitization followed by neutralization with triethanolamine because only two additions are required and

AND

ZAMAN

no further dilution of the sample is made. This is especially important since GSSG concentrations are normally at least two orders of magnitude lower than GSH. Evaluation of the reaction kinetics as a function of GSH reductase, pH, EDTA concentration, and DTNB concentration have been discussed elsewhere (8). We use nonsaturating levels of GSH reductase, 1.0 U/ml, with respect to the DTNB concentration, 0.15 mM (refer to Fig. 3). These conditions minimize the ratio of Y-intercept to slope (refer to Fig. 4). Even so, for low GSH measurements the error introduced by the variation around the Y-intercept is probably the largest source of inaccuracy in the determination. The Softmax software allows the linear fit of the 25 data points collected per well in a 2-min period to be displayed and the individual correlation coefficient per well shown. We have found that the reaction rate for individual samples has an R value of 0.999 in all of the cell samples examined. If a gradual decrease in rate occurs, suggesting either inhibition of GSH reductase or consumption of the NADPH, the sample should be diluted and reassayed. For cell studies, the use of a spin through oil into acid for separating cells from their extracellular environment provides a rapid metabolic stop, a concentrating step, and an acid extraction and denaturation method all in one. Furthermore, excess drug is removed from the sample. Many agents used in studying thiols such as N-ethylmaleimide can inhibit GSH reductase (8). We have successfully used the method to assay changes in intracellular GSHGSSG in cells within minutes after addition of diamide (4) or an analog (9). The results demonstrate that the cellular response to a thiol oxidative challenge can be extremely rapid. The export of GSSG apparently begins within 5 to 10 min after diamide addition (Fig. 5) as evidenced in the increase in extracellular and the decrease in intracellular GSH plus GSSG. The same GSSG stock solution was used to create the standards for both the GSH plus GSSG and GSSG only curves, thereby eliminating the error introduced by us-

TABLE

1

GSH in Nude Mouse Tissues Bearing LS-174T Tumors Treated with BSO Tissue Blood Kidney Liver Tumor

Pre-BSO 0.36 1.41 a.77 0.57

+ * & +

24 h 0.07 0.28 0.50 0.10

0.24 0.28 2.71 0.18

k + f +-

0.13 0.04 0.50 0.06

43h 0.26 0.24 2.93 0.14

k f f +

72 h 0.09 0.04 1.14 0.02

0.29 0.24 3.10 0.13

f + * f

0.06 0.01 0.41 0.01

Note. Mean and standard deviations for tissues from three mice expressed in gmol/G wet weight. Blood values are amol/ml whole blood.

MICROTITER

PLATE

ASSAY

FOR

GLUTATHIONE

ing two different stock solutions. The ability to follow changes in the GSH:GSSG disulfide status is thereby facilitated but the absolute quantitation will depend on the purity of the GSSG used. The determinations described here are variations on the basic recycling assay. It must be emphasized that appropriate standards must be treated and assayed in parallel with samples. That is, all dilutions, acid strength, and treatment times must be simulated and a set of the standards assayed in the same plate with the samples. By combining the least complicated and expedient methods of sample handling with the sensitivity and specificity of the GSH assay by enzymatic recycling and the small volumes and software capabilities of microtiter plate technology we have devised a rapid, sensitive, and relatively easy assay for GSH and GSSG in large numbers of biological samples. One application would be testing of pharmaceuticals or xenobiotics suspected of causing either GSH depletion or oxidation. Furthermore, we have successfully applied the assay to the determination of hydroperoxides using a GSH peroxidase plus GSH to convert peroxide to GSSG as recently described (10).

AND

GLUTATHIONE

365

DISULFIDE

ACKNOWLEDGMENT This work NCI-DHHS

was supported in part to Dr. John E. Biaglow.

by Grant

CA 4498-03

from

the

REFERENCES 1. Flower, R. J., Mondada, S., and cological Basis of Therapeutics and Gilman, Eds.), 6th ed., pp. 2. Owens, C. W. I., and Belcher, 711. 3. Tietze, F. (1969) Anal. Biochem. B. 4. Baker, M. A., and Hagner,

Vane, J. R. (1980) in The Pharma(Gilman, A. G., Goodman, L. S., 703-704, Macmillan, New York. R. V. (1965) Biochem. J. 94, 70527,502-522. A. (1990) Biochim.

Biophys.

Acta

1037,39-47. 5. Anderson, Ed.),

Vol.

M. A. (1985) in Methods in Enzymology (Meister, 113, pp. 548-555, Academic Press, Orlando, FL.

A.,

6. Griffith, 0. W., and Meister, A. (1979) J. Biol. Chem. 264, 755% 7560. 7. Griffith, 0. W. (1980) Anal. Biochem. 106,207-212. 8. Eyer, P., and Podhradsky, D. (1986) Anal. Biochem. 153,57-66. 9. Baker, M. A., Taylor, Y. C., and Brown, J. M. (1988) Radiut. Res. 113,346-355. 10. Allen,

K. G., Huang,

186, 108-111.

C.-J.,

and Morin,

C. L. (1990)

And.

Biochem.