A quantitative fluorescence enzyme immunoassay for plant cytokinins

ANALYTICAL

162, 301-308

BIOCHEMISTRY

A Quantitative

(1987)

Fluorescence

Enzyme Immunoassay

for Plant Cytokinins’

E. J. TRIONE,**~ G. M. BANOWETZ,* B. B. KRYGIER,~ J.M. KATHREIN,~AND L. SAYAVEDRA-SOTO? *USDA

Agricultural Oregon

Research Service, and TDepartment State University, Corvallis, Oregon

of Crop Science, 97331

Received October 14, 1986 An enzyme-linked immunosorbent assay (ELISA) which used 4-methylumbelliferyl phosphate as an enzyme substrate was used to quantify two plant cytokinins. This assay detected as little as 0.03 pmol (approximately 10 pg) of cytokinin in microplate wells coated with a cytokinin-ovalbumin conjugate. The method measured competition between free cytokinin and the bound conjugate for reaction with monoclonal anticytokinin antibodies and used a standard curve prepared by use of known amounts of free cytokinin to quantify hormone levels in unknown samples. Standard curves which consisted of logit/log plots of fluorescence units versus picomoles of competing cytokinin measured from 0.03 to 256 pmol (approximately 10-85000 picograms) of zeatin riboside (ZR) or isopentenyl adenosine. The fluorescence ELISA was compared with radioimmunoassay for the quantification of ZR in wheat (Triticum aestivum L., cultivar Stephens) seed samples. This fluorescence ELISA method is recommended for use in combination with a fractionation method, such as HPLC, to quantify cytokinins present in plant extracts. 0 1987 Academic Rest, Inc. KEY WORDS: fluorescence; immunoassay: ELISA; isopentenyl adenosine; zeatin riboside; monoclonal antibody.

The enzyme-linked immunosorbent assay (ELISA)* (1) has been modified to measure a wide range of antigens and antibodies (2). For convenience, safety, and cost of both equipment and materials, some workers prefer the ELISA over radioimmunoassay ’ The research was supported by the Agricultural Research Service, U.S. Department of Agriculture, Technical Paper No. 7691 of the Oregon Agricultural Experiment Station. Mention of a trade name, proprietary product, or specific equipment does not constitute a guarantee or warranty by the U.S. Department of Agriculture and does not imply approval to the exclusion of other products that may also be suitable. 2 Abbreviations used: BSA, bovine serum albumin; DIECA, sodium diethyldithiocarbamate; ELISA, enzyme-linked immunosorbent assay; F-ELISA, competition fluorescence ELISA, IgE, immunoglobulin E, IgG, immunoglobulin G, iPA, isopentenyl adenosine; IPABSA, isopentenyl adenosine-bovine serum albumin conjugate; iPA-OVA, isopentenyl adenosine-ovalbumin conjugate; OVA, ovalbumin; PBS, 0.15 M phosphate-buffered saline, pH 7.2; PBS-Tween, PBS supplemented with 0.05% Tween 20; RIA, radioimmunoassay; ZR, zeatin riboside.

(RIA) although certain RIAs are more sensitive than the ELISA. Enhanced sensitivity of ELISA techniques has been obtained by the use of biotinylated anti-immunoglobulin probes in conjunction with avidin-labeled enzymes (3), and by the use of fluorogenic enzyme substrates (4-6) although the latter three studies were hampered by the lack of instrumentation for the rapid handling of large numbers of samples. In 1984, Gillcash and Kamuzora (7) reported the use of a fluorescence ELISA and microplate fluorometer to quantitatively measure picomole quantities of allergen-specific IgE in human serum samples. This study describes a procedure which uses a competition fluorescence ELISA (FELISA) to quantify as little as 10 pg of isopentenyl adenosine (iPA) or zeatin riboside (ZR). The assay is conducted in 96-well microplates and fluorescence is measured automatically by a 96-well fluorometer. Sensitivity limits and measuring ranges for detection 301

0003-2697187

$3.00

Copyright 0 1987 by Academic Pres, Inc. All rights of reproduction in any form reserved.

302

TRIONE

of iPA by hormone-specific monoclonal antibodies were compared to those obtained with RIA. The F-ELISA also was used to quantify ZR in HPLC fractions of wheat seed samples and its accuracy was compared to that of a radioimmunoassay. Because the monoclonal antibodies used in this study cross-reacted with other plant cytokinins, we recommend that the F-ELISA serve as a quantitative method in cytokinin analysis in combination with an analytical separation procedure such as HPLC. MATERIALS

AND

METHODS

Chemicals ZR, iPA, 4-methylumbelliferyl phosphate, diethanolamine, ovalbumin (OVA), and bovine serum albumin (BSA) were purchased from Sigma Chemical Co. (St. Louis, MO). Dulbecco’s minimal essential medium (DME-0) was purchased from Flow Labs (McLean, VA). The fetal bovine serum was obtained from Hyclone Labs (Logan, UT). Freund’s complete and incomplete adjuvants were purchased from VWR Scientific (San Francisco, CA). The tritiated cytokinin dialcohols were provided by Dr. R. 0. Morris, Department of Agricultural Chemistry, Oregon State University (Corvallis, OR). All chemicals and reagents were of the highest purity available.

Preparation of Cytokinin-Protein Conjugates Cytokinins were conjugated to bovine serum albumin and ovalbumin as described in (8) with modifications (9). The ZR-BSA and iPA-BSA conjugates were used for mouse immunizations while the ZR-OVA and iPA-OVA conjugates served as immunosorbents for the F-ELISA.

Preparation of Monoclonal Antibodies Hybridomas were prepared by fusing Sp2.0 myeloma cells with splenocytes harvested from BALB/c mice previously immunized with either iPA-BSA or ZR-BSA con-

ET AL.

jugates. Mice received an initial injection (600 pg of iPA-BSA or 300 pg of ZR-BSA) intraperitoneally (ip) in Freund’s complete adjuvant and additional ip injections after 3 and 5 weeks of the appropriate conjugate in Freund’s incomplete adjuvant. Four days prior to the fusion, each mouse received an additional ip injection (either 600 pg of iPA or 300 pg of ZR) in PBS, and on the following day, received intravenously 200 pg of conjugate in PBS. The myeloma cells and splenocytes were fused in the presence of Fisher 1000 polyethylene glycol according to (lo), and the resulting hybridoma culture fluids were screened for anticytokinin activity by an ELISA. Cultures were cloned twice and frozen in liquid nitrogen for storage or injected into BALB/c mice to induce ascites tumor formation ( 11).

Quantification of Plant Cytokinin Hormones Competition fluorescence ELISA. The competition assay (F-ELISA) developed to quantify either iPA or ZR in unknown samples, was conducted in a black 96-well Microfluor plate (Dynatech Laboratories, Alexandria, VA) coated with 100 &well of the appropriate cytokinin-OVA conjugate at 10 pg (for iPA) or 0.1 pg (for ZR) protein/ml of PBS for 2 h at 37°C. After the initial coating, the wells were blocked with 200 ~1 of a 1% BSA solution to prevent nonspecific protein adsorption to the polystyrene. A standard curve was prepared as follows; each of a series of wells received 50 ~1 of a solution which contained a known quantity of the competing cytokinin in PBS. Typically, this included a range of 0.03-512 pmol/50 ~1. After all wells had received competing cytokinin, 50 ~1 of appropriately diluted anti-iPA or anti-ZR antibody were added and the plate was covered with a Mylar cover sheet (Flow Labs, McLean, VA) and incubated in a moist chamber at 37°C for 30 min. The appropriate antibody dilutions were determined empirically in competition experiments; the selected dilution gave the maximum difference between the fluorescence at

FLUORESCENCE

IMMUNOASSAY

the lowest and highest competitor concentrations, Typical dilutions for culture fluids ranged from 1:75 to 1:300 while those for ascites fluid ranged from 1800 to 1:64,000. Control wells which contained no competing cytokinin were included, and all assays were replicated (3-8X). After incubation, the plates were emptied, washed three times with PBS-Tweet-i, and each well received 100 ~1 of a 1: 1000 dilution of a goat anti-mouse IgGalkaline phosphatase conjugate. The plates were covered and returned to the moist chamber at 37°C. After 30 min, the plates were emptied and washed three times with PBS-Tween. Finally, after each well received 200 ~1 of enzyme substrate solution (25.6 pg/ml of 4-methylumbelliferyl phosphate prepared in 10% diethanolamine, pH 9.8,0.5 mM MgCl& the plates were covered and placed in a moist chamber at 37°C for 30 min. After this final incubation, each plate was uncovered and placed in a fluorometer (3M Diagnostics, Mt. View, CA) where the fluorescence of the solutions in each well was determined and recorded automatically. This fluorometer was fitted with a 365 nm excitation filter and 450 nm emission filter. Solutions which contained unknown quantities of either ZR or iPA were used in place of the competing cytokinin and were quantified via the standard curve which consisted of a logit/log plot of fluorescence units versus picomoles of competing cytokinin. RIA. An RIA was conducted for comparison of the measuring range and sensitivity with the F-ELISA. Briefly, 50 ~1 of antibody were added to a 1.5-ml polypropylene microcentrifuge tube which contained tritiated cytokinin dialcohol (approximately 700 cpm) and a selected level of unlabeled cytokinin in a total volume of 400 ~1 of PBS containing 0.1% OVA and 1.0 mM EDTA buffer (pH 7.0). After the contents of the tube were mixed, the tube was incubated at 20°C for 20 min. Antigen-antibody complexes were precipitated by addition of 600 Cal of 90% saturated ammonium sulfate, incubation at 20°C for 15 min, and centrifugation. The precipitates were washed once in

FOR CYTOKININS

303

1000 ~1 of 50% saturated ammonium sulfate solution and then mixed in 130 ~1 of methanol at 20°C for at least 30 min. Radioactivity was determined using a standard cocktail (NEF 963; New England Nuclear, Boston, MA) and a Packard Tricarb Model 4530 scintillation counter. The number of counts of tritiated cytokinin precipitated in the absence of any unlabeled cytokinin was designated as &; the number of counts of tritiated cytokinin precipitated in the presence of a given level of unlabeled cytokinin was designated B. A plot of B/B0 against log of picomoles of unlabeled hormone was drawn to illustrate this relationship, and a logit-log transformation of this plot was used to linearize these data. The log&-log transformed plot was used as a standard curve to determine the effective measuring range and sensitivity of the assay.

Analysis of ZR Levels in Wheat Seed A separate study used monoclonal antibody ZR3 to quantify ZR levels in samples prepared from a wheat seed extract. The extract was prepared as follows: developing wheat seeds (harvested l-5 days after anthesis) were homogenized in methanol (4-6 ml/g seed) which contained 100 mg of the antioxidant, sodium diethyldithiocarbamate (DIECA), and then centrifuged at 20,OOOg for 15 min. The supematant was collected, concentrated approximately fourfold under a stream of nitrogen, and diluted with 9 vol of ammonium acetate buffer (40 mM, pH 6.5, supplemented with 10 mM DIECA and 30 mM 2-mercaptoethanol). Wheat germ acid phosphatase (0.04 units/ml) was added to the sample followed by a 20-min incubation period to convert the cytokinin-5’-phosphate moieties to free nucleosides. The extract was passed through a DE-52 cellulose column (2.5 cm (i.d.) X 4 cm) and the cytokinin-like compounds were adsorbed onto an octadecylsilica column (1 cm (id.) X 2 cm). After the column was washed with 2 vol of the ammonium acetate buffer, the cytokinins were eluted with methanol and dried

TRIONE

304

ET AL.

- 3

:321

- 4 -2.0

-1.0 LOG PICOMOLES

OF’FREE

1.0

2.0

IPA

FIG. 1. Competition fluorescence immunoassay in which selected amounts of free isopentenyl adenosine (iPA) competed with iPA-ovalbumin conjugate bound to wells of a 96-well microplate for reaction with iPA-specific monoclonal antibody. The straight line represents the theoretical line defined by the slope and intercept determined by linear regression analysis (r = 0.98) of the experimental data pairs. Brackets indicate f one standard deviation.

under vacuum. The samples were rediseluted by increasing the concentration of solved in a 1: 1 mixture of methanol and acetonitrile from 10 to 15% over a period of triethylammonium acetate buffer (40 mM, 15 min, and then to 35% over an additional pH 3.5). All glassware used in the extraction 15 min. Absorbance of the eluate was monitored at 254 nm, and 0.5-ml fractions were process was silanized (9). The wheat seed preparation was divided collected, dried under vacuum, and quantiinto equal portions to provide the contents of fied by immunoassay. 1 g (fresh weight) of extracted seed per aliquot for further analyses. The first aliquot RESULTS received an additional 60 ng of free ZR prior Assay Conditions to HPLC. The second received no additional The procedure described for the F-ELISA cytokinins in order to provide an estimate of the amount of the respective cytokinins was selected from empirical trials which varpresent in native wheat seed. A mixture of 30 ied the experimental conditions reported ng each of the following cytokinins served as here. Both PBS (pH 7.2) and carbonate (0.05 an HPLC reference solution; zeatin, ZR, di- M, pH 9.6) were tested as plate-coating hydrozeatin, dihydrozeatin riboside, cis-zea- buffers with either ZR-BSA or iPA-BSA. tin, and iPA. All HPLC analyses were per- Similar amounts of coating protein were adformed with an octadecylsilica column (250 sorbed to the plates with each buffer as X 4.6 mm, 5-Km particle size) equilibrated in judged by immunoassay. Assay conditions 90% triethlyammonium acetate (40 mM, pH were optimized by testing selected levels of 3.4)- 10% acetonitrile. Cytokinins were coating proteins, goat anti-mouse IgG-alka-

FLUORESCENCE

IMMUNOASSAY

08. 97Q6-

95-

QO-

305

FOR CYTOKININS

-+ 4

-*3 -+2

SOW'O Z60;50w400 a30-

:20-

FIG. 2. Competition fluorescence immunoassay in which selected amounts of free zeatin riboside competed with zeatin riboside-ovalbumin conjugate bound to wells of a 96well microplate for reaction with zeatin riboside-specific monoclonal antibody. The straight line represents the theoretical line defined by the slope and intercept determined by linear regression analysis (r = 0.98) of the experimental data pairs. Brackets indicate f one standard deviation.

line phosphatase probes, methylumbelliferyl phosphate, incubation length, and temperature.

Quantification of Plant Cytokinin Hormones Competition j7uorescenceELBA. The FELBA detected as little as 0.03 pmol (approximately 10 pg) of free cytokinin, and measured over a range of 0.03-256 pmol with high linearity (Figs. 1, 2, and Table 1); average coefficients of variation were less than 10% in both iPA and ZR assays. These detection limits and effective measuring ranges were reproducible from trial to trial, and background fluorescence was less than 5% of the highest values obtained in each assay. When solutions which contained known amounts of either ZR or iPA were measured in triplicate in assays with the appropriate monoclonal antibody, estimated hormone quantities fell within 95% confidence limits of the expected values.

Radioimmunoassay. To compare the FELISA with an RIA, monoclonal antibody iPA was used in a system which contained selected amounts of tritiated iPA-dialcohol and free unlabeled iPA. A logit plot of B/B0 against the log of the picomoles of unlabeled iPA added to the system was linear (r = -0.99) between the detection limits of 0.5 and 10 pmol of iPA (Fig. 3; Table 1). The plotted logit values represented the mean of six replicate samples; average coefficient of variation was 7.5% (range of 1.98-12.21%). Quantification of ZR in Wheat SeedSamples Monoclonal antibody ZR3 was used in the two immunoassays to quantify ZR in HPLC-fractionated wheat seed samples (Table 2). Estimates of ZR in native wheat seed ranged from 2.55 to 4.99 rig/g wheat seed; the F-ELBA estimates were lower than those provided by RIA. Estimates of ZR in

306

TRlOh’E

ET AL.

-4

as-

-3

go-

-2

m” SO2

70-

-1

; SO; so= 40-

-0

0 30E P 20-

--1

!I cl 2

--2

IO-

5-

- -3

LOG.

PICOMOLES

0.2 FREE

0.4 0.8 ISOPENTENYL

0.8 ADENOSINE

1.0

FIG. 3. Radioimmunoassay of monoclonal antibody iPA in which selected amounts of free isopentenyl adenosine (IPA) competed with tritiated iPA-dialcohol. Data are presented as a logit/log transformation plot. The straight line represents the theoretical line defined by the slope and intercept determined by linear regression analysis (r = 0.99) of the experimental data pairs. Brackets indicate + one standard deviation.

wheat seed sample spiked with 60 ng of additional ZR prior to HPLC fractionation ranged from approximately 56 to 62 rig/g wheat seed with RIA providing the highest estimate. Both methods showed greater than 90% recovery of expected ZR values.

statistical precision, routine assays could use duplicate samples with satisfactory precision. ELISA methods with fluorescent substrates have been used to detect low levels of ferritin (5) (100 pg/ml), insulin (6) (2 1 pg/ml), and the herbicide diclofopmethyl(4) (45 rig/ml). Although the sensitivities and ease of these techniques represented significant improveDISCUSSION ments over previous methods, none could The F-ELISA was a rapid and sensitive process the large number of samples possible technique for quantification of free cyto- in a 96-well microplate system with an autokinins and effectively measured these plant mated fluorescence detector. growth regulators over a wide range of conThe F-ELISA used either iPA or ZR concentrations. The abilities to perform these jugated to OVA for the immunosorbent to assays in 96-well microplates and to deter- avoid detection of antibodies against the immine final fluorescence directly with an au- munizing carrier protein, BSA. All the antitomatic microplate fluorometer enabled the bodies detected in this manner had specifichandling of a large number of samples with ities for the appropriate cytokinin and not precision, ease, and speed. One person could the carrier; an additional test for specificities likely process 4-6 plates (384-576 wells) in 1 against either carrier protein was conducted day. Although up to eight replicates of each by use of OVA (with no attached cytokisample were used in this study for enhanced nin) as an immunosorbent in a fluores-

FLUORESCENCE TABLE 1

IMMUNOASSAY

FOR CYTOKININS

307

another analytical technique for routine cytokinin quantification (e.g., purification on A COMPARISON OF Two COMPETITION IMMUNOASimmunoaffinity columns and separation by SAYS WHICH USED MON~CLONAL ANTIBODY iPA FOR THE MEASUREMENT OF THE CYTOKININ, ISOPENTENYL HPLC prior to quantitative immunoassay) ADENOSINE as suggested by others (9,15,16). Initial purification steps may not be required if monoAverage Competition Measuring clonal antibodies which do not cross-react rb C.V. (%)’ immunoassay range (PgY with other cytokininlike molecules can be Fluorescence produced. The antibodies used in the present -0.98 8.10 ELISAd IO-85,000 study, however, are known to cross-react RIA’ 170- 3,300 -0.99 7.50 with certain other cytokinins (( 17), unpublished data). Monoclonal antibodies have a The range of picograms over which a correlation been reported which showed remarkable coefficient 3 0.98 was obtained with linear regression analysis of data pairs which consisted of a logit transforspecificity for iPA and minimal cross-reactimation of fluorescence/absorbance/cpm and log pmol. vity with structurally related molecules ( 18). b r = Correlation coefficient (P = 0.001). Nevertheless, a quantitative immunoassay, ’ Coefficient of variation. such as the one described here, is used most ’ A fluorescence enzyme-linked immunosorbent assay appropriately in combination with an HPLC in which the competition between free isopentenyl adenosine (iPA) and an iPA-ovalbumin conjugate for fractionation method. binding to monoclonal antibody iPA was quantified The range of ZR (approximately 2.5-5 with the fluorescent enzyme substrate Cmethylumbellirig/g) measured in developing wheat seeds by feryl phosphate. the two immunoassays was similar to pre’ A radioimmunoassay which measured the competiviously reported levels (19) which peaked at tion between unlabeled iPA and tritiated iPA-dialcohol for binding to the monoclonal antibody iPA3. approximately 60 rig/g 2 days past anthesis with a sharp drop to less than 10 rig/g at 5 days past anthesis. The wheat seeds used in cence ELISA. Although it is likely that antiour study were collected at 5 days past anthBSA antibody-producing hybridomas were esis, although younger seeds were also present among the original cell fusion products, neither of the antibodies used in these TABLE 2 studies cross-reacted with the OVA immunosorbent. The F-ELISA provided greater sen- A COMPARISON OF Two IMMUNOASSAYS FOR QUANTIsitivity and a broader measuring range than FICATION OF ZEATIN RIBOSIDE (ZR) IN DEVELOPING WHEAT SEEDS (rig/g FRESH WEIGHT) the RIA (Table 1) when these assays were conducted using the same antibody (iPA3). Immunoassay An RIA with enhanced sensitivity was reported ( 12) which used polyclonal rabbit anSample” F-ELISAb RIA’ tiserum to detect 40 to 6708 pg of iPA while another (13) used RIA to measure from 250 Seed 2.55 + 0.16 4.99 + 1.51 Seed/ZR 56.70 f 0.26 61.96 f 3.85 to 10000 pg of this plant growth regulator. A competition standard ELISA was developed ’ The seed sample consisted of extracted wheat seed (14) to measure the cytokinins zeatin and fractionated by HPLC prior to immunoassay. The seed/ zeatin riboside in maize stems over the range ZR sample consisted of a l-g aliquot of the seed sample of 100 to 10,000 pg. The F-ELISA described which received an additional 60 ng of ZR prior to HPLC and immunoassay. in the present study should prove useful, b A competition fluorescence ELISA which used 4especially to plant hormone research that methylumbelliferyl phosphate for alkaline phosphatase deals with tissues available only in minute substrate. quantities, e.g., apical meristems. The ’ A radioimmunoassay which measured competition method should be used in combination with between tritiated ZR-dialcohol and free ZR.

TRIONE

present in the sample. Immunoassays performed on seed samples spiked with additionai ZR showed that both RIA and the FELBA provided greater than 90% recovery of added and native cytokinin. ACKNOWLEDGMENTS We thank Dr. R. 0. Morris and E. M. S. MacDonald for providing the iPA and ZR-OVA and BSA conjugates, the tritiated iPA, and for their advice concerning the radioimmunoassay.

ET

AL.

9. MacDonald, E. M. S., Akiyoshi, D. E., and Morris, R. 0. (198l)J. Chromatogr. 214, 101-109. 10. Oi, V. T., and Herzenberg, L. A. (1980) in Selected Methods in Cellular Immunology (Mishell, B. B., and Shiigi, S. M., Eds.), pp. 351-372, Freeman, San Francisco. 11. Zola, H., and Brooks, D. (1982) in Monoclonal Hybridoma Antibodies: Techniques and Apphcations (Hurrell, J. G. R., Ed.), pp. 4-54, CRC Press, Boca Raton. FL. 12. Weiler, E. W., and Spanier, K. (1981) Pluntn 153, 326-337.

Ernst, D., Schafer, W., and Oesterhelt, D. (1983) Planta 159, 2 16-22 1. 14. Hansen, C. E., Wenzler, H., and Meins, F. (1984) 13.

REFERENCES 1. Engvall, E., and Perlmann, P. (197 1) Immunochemistry 8, 87 l-874. 2. Clark, M. F. (1981) Annu. Rev. Phytopathol. 19, 83-106. 3. Plebani, A., Avanzani, M. A., Massa, M., and Ugazio, A. G. (1984) J. Immunol. Methods 71, 133-140. 4. Schwalbe, M., Dorn, E., and Beyermann. K. (1984) J. Agric. Food Chem. 32,734-74 1. 5. Konijn, A. M., Levy, R., Link, G., and Hershko, C. (1982)

J. Immunol.

Methods

54,297-307.

6. Hinsberg, W. D., Milby, K. H., and Zare, R. N. (198l)Anal. Chem. 53, 1.509-1512. 7. Gillcash, R. S., and Kamuzora, H. (1984) Corm. Med.

48,73-77.

8. Erlanger, B. F., and Beiser, S. M. (1964) Proc. Nut/. Acad. Sci. USA 52,68-74.

Plant Physiol.

75, 959-963.

15. Morris, R. O., Akiyoshi, D. E., MacDonald, E. M. S., Morris, J. W., Regier, D. A., and Zaerr, J. B. (1982) in Plant Growth Substances (Waring, P. F., Ed.), pp. 175-183, Academic Press, New York. 16. Badenoch-Jones. J.. Letham. D. S., Parker, C. W., and Rolfe, B. G. (1984) Plant Physiol. 75, 1117-1125. 17. Trione, E. J., Krygier, B. B., Banowetz, G. M., and Kathrein, J. M. (I 985) J. Plant Growth Reguf. 4, 101-109. 18. Woodsworth, M. L., Latimer, L. J. P.. Janzer, J. J., McLennan, B. D., and Lee, J. S. (1983) Biochem. Biophys.

Rex

Commun.

114, 79 l-796.

19. Durley, R. C., and Morris, R. 0. (1983) Plant Physiol. 72, 114 (Abstract).