Production of a monodonal antibody
to cortisol: application to a direct enzyme-linked immunosorbent assay of plasma John G. Lewis, Laurette Manley, Joanna C. Whitlow, and Peter A. Elder Steroid Unit, Department o f Clinical Biochemistry, Christchurch Hospital, Christchurch, N e w Zealand
Cortisol mouse monocional antibodies were produced and characterized. Of the four clones studied, supernatant from one clone (A2), compared with other cortisol monoclonal antibodies, showed minimal cross-reactivity to other C2t steroids and was suitable for the direct determination of cortisol in plasma by enzyme-linked immunosorbent assay using a standard 96-well microtiter plate. The enzymelinked immunosorbent assay uses the immobilized antigen approach, in which cortisol in plasma samples or standards competes with immobilized steroid for antibody-binding sites. After washing, the cortisol antibody bound to the wells of the microtiter plate is detected with antimouse immunoglobulin conjugated to horseradish peroxidase. Following further washing, o-phenylenediamine substrate is added. The enzyme-linked immunosorbent assay is robust and semiautomated. The mean +- SD recovery from plasma was 97% +- 6%. Precision studies on three different plasma pools showed mean coefficients of variation of 7.6% and 8.6% for within- and between-assay variation, respectively. The satisfactory performance criteria allow its use in the routine laboratory. (Steroids 57:82-85, 1992) Keywords: cortisol; monoclonal; ELISA; antibody; steroid; cortisol monoclonal antibody; plasma cortisol; monoclonal antibody to cortisol
Introduction Although many laboratories engaged in steroid hormone assays have produced monoclonal antibodies suitable for use in routine assays, L2 including our o w n ) ,4 surprisingly few have published data on monoclonal antibodies to cortisol. Previously, our laboratory attempted to raise a cortisol monoclonal antibody but generated a monoclonal antibody with exceptionally high cross-reactivity to prednisone (> 1,200%) and 11-deoxycortisol ( > 100%). 5 Other groups have also reported monoclonal antibodies with high cross-reactivity to 11-deoxycortisol and prednisolone in rats and mice immunized with a cortisol-bovine serum albumin (BSA) conjugate. 6 We report the production of a monoclonal antibody to cortisol following immunization of R B F / D N mice with cortisol-21-acetate-3CMOBSA. Although there is significant cross-reactivity to Address reprint requests to Dr. John G. Lewis at the Steroid Unit, Department of Clinical Biochemistry, Christchurch Hospital, Christchurch, New Zealand. Received April 25, 1991; accepted August 15, 1991.
82
Steroids, 1992, vol. 57, February
l l-deoxycortisol (19%), cross-reactivity with other glucocorticoids is low and comparable to many commercial kit assays. The supernatant from hybridoma cultures of this antibody-secreting clone has been used, without further treatment, for the direct determination o f cortisol in plasma by enzyme-linked immunosorbent assay (ELISA). Experimental
Materials Steroids were obtained from Sigma Chemical Co. (St. Louis, MO, USA), and all other chemicals were of analytic grade. Cortisol-3-(O-carboxymethyl)oxime was synthesized from cortisoF and coupled to bovine thyroglobulin by the mixed anhydride method) Cortisol-21-acetate-3-CMO-BSA was obtained from Steraloids Inc. (Wilton, NH, USA). Goat antimouse immunoglobulin (Ig)-peroxidase was obtained from Amersham Corp. (Amersham, UK). I m m u n i z a t i o n a n d cell f u s i o n Four male RBF/DN mice were each immunized intraperitoneally with 10 to 100 /zg of cortisol-21-acetate-3CMO-BSA in
© 1992 Butterworth-Heinemann
Monoclonal antibody to cortisol: Lewis et al. Freunds complete adjuvant (100 p,l) at four 4-week intervals. Following the fourth injection and 1 week prior to fusion, the mice were immunized a further three times on alternate days with 10 to 100/~g of cortisol-21-acetate-3CMO-BSA in normal saline (100 t~l). The spleens were then excised and spleen cells fused with the mouse myeloma cell line FOX-NY as previously described. 3
Screening of supernatants ELISA plates (Falcon 3912 Microtest III, Becton Dickinson, Oxnard, CA, USA) were coated overnight at 4C with 100/~1 of cortisol-thyroglobulin in conjugate/well (1 t~g/ml) in 6 M aqueous guanidine hydrochloride. The following day the plates were washed four times with a solution of phosphate-buffered saline (PBS), 0.05 M NaH2PO4, 0.15 M NaCI adjusted to pH 7.4 with 5 M NaOH, containing 0.1% Tween 20 (v/v). To prevent further adsorption of protein, the plates were then blocked with assay buffer, PBS containing 0.1% Tween 20 (v/v), and 0. I% gelatin (w/v) (150 p~l/well) for 1 hour at 20C. After emptying the plates, 50 p.I of assay buffer was added to each well followed by 50 p.l of supernatant for a further 2-hour incubation at 20C. The plates were washed four times and sheep anti-mouse Ig-peroxidase was added (100 t~l/well at I : 2,000 dilution in assay buffer) for I hour at 20C, after which the plates were washed and substrate solution was added. Substrate was freshly prepared by the addition of 40 mg o-phenylenediamine to 100 ml of 50 mM Na~_HPO4 and 25 mM citric acid buffer, pH 5.0. Color development was terminated by the addition of 100 ftl of 1.25 M H2SOJwell and the absorbance read at 492 nm. The use of a Behring ELISA Processor II (Hoechst, Germany) allowed total automation with the exception of dispensing samples and standards. Cross-reactivity studies were performed by incubating eight different steroid doses from 0 to 5,000 ng in assay buffer (50/zl) with supernatant from the appropriate clone diluted in assay buffer (50 td) and subsequent processing as described.
Plasma cortisol ELISA Assay standards were prepared by appropriate dilution of cortisol primary standard (1 mg/ml in ethanol) in 10% stripped human plasma (Amberlite XAD-2), diluted in assay buffer, to yield six different doses from 0 to 280 nmoi/L. After the addition of standards (50/~1) to the wells of a coated and blocked plate, 45/~1 assay buffer was added to the sample wells followed by 5/~1 of sample plasma. Supernatant from clone A2, diluted 1 : 50 with assay buffer, was added to each well (50 p,l) for a 2-hour incubation at 200C. The plate was then washed four times, anti-mouse lg-peroxidase was added for 2 hours ~tt 20C, and the plate further processed as described in the previous section.
Results Cell fusion, screening, and cloning by limiting dilution resulted in four clones secreting cortisol antibody, which could be displaced from the immobilized cortisol-thyroglobulin conjugate by increasing doses of cortisol in assay buffer. Cross-reactivity profiles of these four antibodies are shown in Table 1. Supernatant from clone A2 was the most acceptable, and additional cross-reactivity results are shown in Table 2. A standard curve in 10% human stripped plasma is shown in Figure 1. Samples displayed parallelism when diluted in stripped plasma and the sensitivity at two standard deviations from zero was less than 55 nmol/L. Comparison of plasma cortisol levels with our previous
ELISA 9 are shown in Figure 2. Repeated analysis of three plasma controls provided an assessment of precision. Within-assay variations (mean -- SD) for six sets of duplicates for three different plasma pools were 108 -- 14; 446 -- 28, and 804 -- 27 nmol/L. Betweenassay variation for six sets of duplicates for the same plasma pools were 113 -- 19, 429 -- 19, and 786 ± 37 nmol/L. Recovery was estimated by adding known amounts of cortisol to a normal plasma and subsequent assay. Results are shown in Table 3. Plasma cortisol adult normal ranges for 0900, 1600, and 2400 hours are 250 to 800, 200 to 330, and less than 300 nmol/L, respectively. In normal subjects, the short synacthen test (0.25 mg synacthen intramuscularly) resulted in a 1hour postsynacthen increment of at least 200 nmol/L cortisol or a postsynacthen value exceeding 500 nmol/ L. The overnight dexamethasone suppression test, 1 mg at midnight and sampling for plasma cortisol the following morning at 0800 hours, showed normal subjects had plasma cortisol levels lower than 150 nmol/ Table 1 Cross-reactivity (%) of supernatants as determined by ELISA Supernatant Steroid
A2
Cortisol 11-Deoxycortisol Deoxycorticosterone 11-Dehydrocorticosterone Corticosterone Progesterone
A6
B5
C1
100
100
100
100
19
200
90
4.8 1.3 10.0
to cortisol: application to a direct enzyme-linked immunosorbent assay of plasma John G. Lewis, Laurette Manley, Joanna C. Whitlow, and Peter A. Elder Steroid Unit, Department o f Clinical Biochemistry, Christchurch Hospital, Christchurch, N e w Zealand
Cortisol mouse monocional antibodies were produced and characterized. Of the four clones studied, supernatant from one clone (A2), compared with other cortisol monoclonal antibodies, showed minimal cross-reactivity to other C2t steroids and was suitable for the direct determination of cortisol in plasma by enzyme-linked immunosorbent assay using a standard 96-well microtiter plate. The enzymelinked immunosorbent assay uses the immobilized antigen approach, in which cortisol in plasma samples or standards competes with immobilized steroid for antibody-binding sites. After washing, the cortisol antibody bound to the wells of the microtiter plate is detected with antimouse immunoglobulin conjugated to horseradish peroxidase. Following further washing, o-phenylenediamine substrate is added. The enzyme-linked immunosorbent assay is robust and semiautomated. The mean +- SD recovery from plasma was 97% +- 6%. Precision studies on three different plasma pools showed mean coefficients of variation of 7.6% and 8.6% for within- and between-assay variation, respectively. The satisfactory performance criteria allow its use in the routine laboratory. (Steroids 57:82-85, 1992) Keywords: cortisol; monoclonal; ELISA; antibody; steroid; cortisol monoclonal antibody; plasma cortisol; monoclonal antibody to cortisol
Introduction Although many laboratories engaged in steroid hormone assays have produced monoclonal antibodies suitable for use in routine assays, L2 including our o w n ) ,4 surprisingly few have published data on monoclonal antibodies to cortisol. Previously, our laboratory attempted to raise a cortisol monoclonal antibody but generated a monoclonal antibody with exceptionally high cross-reactivity to prednisone (> 1,200%) and 11-deoxycortisol ( > 100%). 5 Other groups have also reported monoclonal antibodies with high cross-reactivity to 11-deoxycortisol and prednisolone in rats and mice immunized with a cortisol-bovine serum albumin (BSA) conjugate. 6 We report the production of a monoclonal antibody to cortisol following immunization of R B F / D N mice with cortisol-21-acetate-3CMOBSA. Although there is significant cross-reactivity to Address reprint requests to Dr. John G. Lewis at the Steroid Unit, Department of Clinical Biochemistry, Christchurch Hospital, Christchurch, New Zealand. Received April 25, 1991; accepted August 15, 1991.
82
Steroids, 1992, vol. 57, February
l l-deoxycortisol (19%), cross-reactivity with other glucocorticoids is low and comparable to many commercial kit assays. The supernatant from hybridoma cultures of this antibody-secreting clone has been used, without further treatment, for the direct determination o f cortisol in plasma by enzyme-linked immunosorbent assay (ELISA). Experimental
Materials Steroids were obtained from Sigma Chemical Co. (St. Louis, MO, USA), and all other chemicals were of analytic grade. Cortisol-3-(O-carboxymethyl)oxime was synthesized from cortisoF and coupled to bovine thyroglobulin by the mixed anhydride method) Cortisol-21-acetate-3-CMO-BSA was obtained from Steraloids Inc. (Wilton, NH, USA). Goat antimouse immunoglobulin (Ig)-peroxidase was obtained from Amersham Corp. (Amersham, UK). I m m u n i z a t i o n a n d cell f u s i o n Four male RBF/DN mice were each immunized intraperitoneally with 10 to 100 /zg of cortisol-21-acetate-3CMO-BSA in
© 1992 Butterworth-Heinemann
Monoclonal antibody to cortisol: Lewis et al. Freunds complete adjuvant (100 p,l) at four 4-week intervals. Following the fourth injection and 1 week prior to fusion, the mice were immunized a further three times on alternate days with 10 to 100/~g of cortisol-21-acetate-3CMO-BSA in normal saline (100 t~l). The spleens were then excised and spleen cells fused with the mouse myeloma cell line FOX-NY as previously described. 3
Screening of supernatants ELISA plates (Falcon 3912 Microtest III, Becton Dickinson, Oxnard, CA, USA) were coated overnight at 4C with 100/~1 of cortisol-thyroglobulin in conjugate/well (1 t~g/ml) in 6 M aqueous guanidine hydrochloride. The following day the plates were washed four times with a solution of phosphate-buffered saline (PBS), 0.05 M NaH2PO4, 0.15 M NaCI adjusted to pH 7.4 with 5 M NaOH, containing 0.1% Tween 20 (v/v). To prevent further adsorption of protein, the plates were then blocked with assay buffer, PBS containing 0.1% Tween 20 (v/v), and 0. I% gelatin (w/v) (150 p~l/well) for 1 hour at 20C. After emptying the plates, 50 p.I of assay buffer was added to each well followed by 50 p.l of supernatant for a further 2-hour incubation at 20C. The plates were washed four times and sheep anti-mouse Ig-peroxidase was added (100 t~l/well at I : 2,000 dilution in assay buffer) for I hour at 20C, after which the plates were washed and substrate solution was added. Substrate was freshly prepared by the addition of 40 mg o-phenylenediamine to 100 ml of 50 mM Na~_HPO4 and 25 mM citric acid buffer, pH 5.0. Color development was terminated by the addition of 100 ftl of 1.25 M H2SOJwell and the absorbance read at 492 nm. The use of a Behring ELISA Processor II (Hoechst, Germany) allowed total automation with the exception of dispensing samples and standards. Cross-reactivity studies were performed by incubating eight different steroid doses from 0 to 5,000 ng in assay buffer (50/zl) with supernatant from the appropriate clone diluted in assay buffer (50 td) and subsequent processing as described.
Plasma cortisol ELISA Assay standards were prepared by appropriate dilution of cortisol primary standard (1 mg/ml in ethanol) in 10% stripped human plasma (Amberlite XAD-2), diluted in assay buffer, to yield six different doses from 0 to 280 nmoi/L. After the addition of standards (50/~1) to the wells of a coated and blocked plate, 45/~1 assay buffer was added to the sample wells followed by 5/~1 of sample plasma. Supernatant from clone A2, diluted 1 : 50 with assay buffer, was added to each well (50 p,l) for a 2-hour incubation at 200C. The plate was then washed four times, anti-mouse lg-peroxidase was added for 2 hours ~tt 20C, and the plate further processed as described in the previous section.
Results Cell fusion, screening, and cloning by limiting dilution resulted in four clones secreting cortisol antibody, which could be displaced from the immobilized cortisol-thyroglobulin conjugate by increasing doses of cortisol in assay buffer. Cross-reactivity profiles of these four antibodies are shown in Table 1. Supernatant from clone A2 was the most acceptable, and additional cross-reactivity results are shown in Table 2. A standard curve in 10% human stripped plasma is shown in Figure 1. Samples displayed parallelism when diluted in stripped plasma and the sensitivity at two standard deviations from zero was less than 55 nmol/L. Comparison of plasma cortisol levels with our previous
ELISA 9 are shown in Figure 2. Repeated analysis of three plasma controls provided an assessment of precision. Within-assay variations (mean -- SD) for six sets of duplicates for three different plasma pools were 108 -- 14; 446 -- 28, and 804 -- 27 nmol/L. Betweenassay variation for six sets of duplicates for the same plasma pools were 113 -- 19, 429 -- 19, and 786 ± 37 nmol/L. Recovery was estimated by adding known amounts of cortisol to a normal plasma and subsequent assay. Results are shown in Table 3. Plasma cortisol adult normal ranges for 0900, 1600, and 2400 hours are 250 to 800, 200 to 330, and less than 300 nmol/L, respectively. In normal subjects, the short synacthen test (0.25 mg synacthen intramuscularly) resulted in a 1hour postsynacthen increment of at least 200 nmol/L cortisol or a postsynacthen value exceeding 500 nmol/ L. The overnight dexamethasone suppression test, 1 mg at midnight and sampling for plasma cortisol the following morning at 0800 hours, showed normal subjects had plasma cortisol levels lower than 150 nmol/ Table 1 Cross-reactivity (%) of supernatants as determined by ELISA Supernatant Steroid
A2
Cortisol 11-Deoxycortisol Deoxycorticosterone 11-Dehydrocorticosterone Corticosterone Progesterone
A6
B5
C1
100
100
100
100
19
200
90
4.8 1.3 10.0
