0021-972X/78/4702-0391$02.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1978 by The Endocrine Society

Vol. 47, No. 2 Printed in U.S.A.

A Direct Microassay for Aldosterone in Plasma Extracts* JOSE SANCHOf AND EDGAR HABER WITH THE TECHNICAL ASSISTANCE OF ELIZABETH DIMOCK

Cardiac Unit, Massachusetts General Hospital, and the Department of Medicine, Harvard Medical School, Boston, Massachusetts 02114 ABSTRACT. We previously described an RIA for aldosterone based on highly specific antibodies elicited with aldosterone-3,20 dioxime bovine y-globulin (Poulsen, K., J. Sancho, and E. Haber, Clin Immunol Immunopathol 2: 373, 1974). We now report the development of higher affinity antibodies of similar specificity that allow the direct measurement of normal aldosterone concentrations in extracts of 100 JU.1 plasma. The detection limit at ± 2 SD of mean zero value was 0.2-0.3 pg. The assay was validated by 1) zero values in adrenalectomized plasma, 2) parallel displacement plots of aldosterone standards and plasma extracts, 3) 97.7 ±

T

HE VERY low concentrations of aldosterone in plasma in relation to other steroids and the lack of specificity of available antibodies has necessitated separation of aldosterone present in plasma from other steroids by paper (1-4), column (5), or thin layer chromatography (6). Recently, in order to simplify this assay, other approaches have been used such as the y lactone methods (7, 8) or those based on a preliminary antibody-binding exposure (9, 10). The number of separations that must be undertaken necessarily result in losses, for which accurate correction is at times difficult. Considerable simplification was achieved with the production of highly specific antibodies that allow the assay of aldosterone in crude extracts of plasma without further purification (11, 12). These assays required a relatively large volume of plasma because of the low binding affinity of these specific antibodies. For clinical studies, a method applicable to

-f-

y-

2.5% recovery of added tracer, 4) a correlation coefficient of 0.992 and a slope of 1.08 in a plot of added vs. recovered unlabeled aldosterone, 5) 10.2% interassay variation and 5.6% intraassay variation, and 6) failure of high concentrations of a variety of steroids added to plasma to perturb the measured aldosterone concentration. The simplicity of the assay results in a high degree of productivity, in that 195 plasma samples may be processed in duplicate, in addition to standard curves, in a single working day. (J Clin Endocrinol Metab 47: 391, 1978)

small volumes of plasma would be desirable in order to avoid artifactural stimulation of aldosterone secretion by volume depletion. Similar requirements exist for studies in children or in small animals. We now report an RIA for aldosterone that may be carried out directly on extracts of 50-100 ju.1 plasma utilizing antibodies of both high specificity and affinity. Materials and Methods Solvents and reagents

Received August 16, 1977. Address requests for reprints to: Dr. Edgar Haber, Professor of Medicine, Cardiac Unit, Massachusetts General Hospital, Boston, Massachusetts 02114. * This work was supported in part by USPHS Grant HL-14150. f Present address: Endocrine Unit, Hospital Ramon Y Cajal, Madrid, Spain.

Reagent grade dichloromethane (Fisher) was twice distilled [boiling point (BP) = 39.8 C] before use. Benzene (spectroquality), Tris (Sigma), dextran T-70 (Pharmacia), charcoal (Neutral Norit A Fisher Scientific Co.), and bovine serum albumin (BSA; Sigma) were used without further purification. Tris buffer. Tris-acetate buffer (0.15 M, pH 7.4) was prepared and, after 5 min of boiling, was stored frozen. Before use, BSA (1 mg/ml) and EDTA (3.72 mg/ml) were added. This buffer was kept at 4 C for a maximum of 10 days. Dextran-coated charcoal. Dextran T-70 (7.5 g) was dissolved in 1 liter 0.15 M Tris-acetate buffer, pH 7.4, and 75 g previously decanted charcoal was added. The solution was stirred overnight at 4 C and kept at 4 C.

391

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 22:07 For personal use only. No other uses without permission. . All rights reserved.

392

SANCHO AND HABER

JCE&M • 1978 Vo\ 47 • No 2

D-Aldosterone. D-Aldosterone (crystaline) was obtained from Sigma. It was purified by thin layer chromatography, by a system previously described (4), and stored in methanol (1 mg/ml). Working solutions were made in 0.15 M Tris buffer with BSA and EDTA.

All rabbits were bled weekly, and antisera were analyzed for titer, cross-reactivity, affinity, and the specific interference of extracted aldosterone-free plasma. All antisera were stored frozen undiluted or diluted (1:10) in assay buffer.

[1,2,6,7-3HJAldosterone. [l,2,6,7-3H]Aldosterone was obtained from New England Nuclear (54 Ci/mM). Purity was checked by the same thin layer chromatography system used in the purification of unlabeled aldosterone; it was stored in benzeneethanol (9:1) at 4 C in the dark.

Assay

Other steroids. Other steroids used for cross-reactivity studies were obtained from Sigma, except 18OH-deoxycorticosterone that was kindly donated by Dr. J. Melby (Boston University). Other materials For the extraction step, glass disposable test tubes (12 X 75 mm) were used. The assay was performed in plastic tubes (12 X 75 mm) (Falcon). Samples and scintillation cocktail (Instagel, Packard) were placed in glass vials and counted in a Packard Tricap liquid scintillation spectrometer (model 3320).

Immunization The antibody used was elicited by the aldosterone-3,20 dioxime bovine y-globulin antigen previously described (13). Sixteen New Zealand White rabbits were divided into three groups. Group 1 (six rabbits) received intradermal injections at 20-30 sites on the back with a total of 0.5 ml complete Freund's adjuvant containing 10 mg/ml killed Tubercule bacilli and 375 jtig antigen. Immunization was repeated monthly for 3 consecutive months. Group 2 (six rabbits) was immunized as in group 1, except that the amount of antigen used per injection was 75 |xg. After a 5-month rest period, both groups 1 and 2 received toe pad injections of 2 mg antigen dissolved in 1 ml complete Freund's adjuvant that contained 10 mg/ml killed Tubercule bacilli. Subsequently, rabbits of both groups 1 and 2 received toe pad injections at monthly intervals with 1 mg antigen. Group 3 (four rabbits) was primarily immunized as in group 1, except that 2.87 mg antigen was used. Eight weeks later, toe pad injections with 2 mg antigen suspended in complete Freund's adjuvant were administered. A second toe pad injection of 1 mg antigen was given 23 weeks after the primary immunization; this was repeated at 3 weekly intervals thereafter.

Extraction. Depending on the expected concentration of aldosterone in the unknown sample, 100 or 50 jttl, or a smaller volume of EDTA plasma or serum were extracted in 12 X 75-mm glass disposable tubes with 2 ml methylene chloride. The tubes were vortexed at high speed for 10 sec, centrifuged 5 min at 1200 rpm, and the aqueous layer was removed by suction. The solvent fraction was dried in a water bath at 37 C under a nitrogen stream. RIA The dry samples were reconstituted with 600 /xl buffer and vortexed. Two hundred microliters in duplicate were pipetted into assay tubes. A standard curve was constructed by the addition in triplicate of 200 /xl D-aldosterone diluted in assay buffer at concentrations of 0, 0.2, 0.5,1, 2, 3, 4, 6, 8,12,16, and 24 pg/200 j j . To all tubes was added 300 jul freshly made solution of antibody (1:1800) in assay buffer containing 3100 cpm [l,2,6,7-3H]aldosterone. At this dilution of antibody (final dilution, 1:3000), approximately 50% of the tracer was bound in the absence of unlabeled aldosterone. Tubes were incubated at 4 C for 16 h. The separation of bound and free hormone was achieved by the rapid addition of 0.5 ml cold dextran-coated charcoal suspension, followed by centrifugation at 2300 rpm for 20 min at 4 C. The addition of charcoal to the assay tubes was done in groups of 168 tubes, the capacity of the centrifuge used. Assay tubes were kept in an ice bath while the charcoal was being added. The time interval between the addition of charcoal to the first tube until the last was always less than 8 min. A total of 13 min elapsed from the start of charcoal addition to the beginning of centrifugation. Strict adherence to elapsed time during charcoal exposure resulted in reproducible standard curves. The supernatant was decanted into vials containing 14 ml counting solution and 10,000 counts were allowed to accumulate. A standard curve was constructed by plotting concentrations of aldosterone standards vs. the percentage of binding. The concentration of the unknown samples was read from the standard curve. Final results were expressed in picograms per ml.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 22:07 For personal use only. No other uses without permission. . All rights reserved.

-f -4

MICROASSAY FOR ALDOSTERONE IN PLASMA EXTRACTS

393

of them also having a very high affinity for aldosterone.

Results Immunization During the first 5 months after the primary immunization of groups 1 and 2, no significant aldosterone binding by the antisera was observed. Three weeks after secondary toe pad injection, antigen binding was first noted. In group 3, no binding was observed for 8 weeks after the initial immunization, but all bleeding showed significant binding by the 3rd week after the second toe pad injection (11 weeks after primary immunization). The antibody utilized in the assay described was obtained from a rabbit in group 3 (886), 16 weeks after primary immunization. Four consecutive weekly bleedings showed similar characteristics with respect to binding affinity, titer, and cross-reactivity. Table 1 describes relevant characteristics of antisera found to have the greatest specificity for aldosterone. Five animals of group 1 produced specific antisera with appropriately low cross-reactivity with other steroids, but each of these antisera had a lower affinity for aldosterone than the serum produced by rabbit 886; in four animals, there was nonspecific interference of binding by extracts of adrenalectomized plasma. In group 2, two animals produced specific antibodies, one with a very high affinity constant. This antiserum also showed interference by adrenalectomized plasma. In group 3, two animals produced very specific antibodies, one

Assay procedure Equilibration was generally allowed to proceed for 16 h at 4 C. In some experiments, a 15-min equilibration at 37 C followed by 4 h at 4 C resulted in similar displacement curves. Serum samples gave similar results to plasma samples anticoagulated with 0.8 ng/ml Na EDTA. A typical assay in our laboratory includes three consecutive groups of 168 tubes, each with its own standard curve (approximately 200-225 samples in duplicate). Sensitivity and antibody affinity A typical standard curve is shown in Fig. 1. When antibody binding data was replotted according to the method of Scatchard, a curvilinear plot resulted, indicating heterogeneity of association constants. Approximately xh of the sites were of high affinity, the average dissociation constant being 1.7 X 10"u M. The detection limit of the assay at ±2 SD from the mean zero value is 0.2-0.3 pg. This corresponds to 6-9 pg/ml"1 for 100 JU.1 plasma, or 3-4.5 pg/mT 1 for 200 /xl plasma. The absence of nonspecific interference by plasma was

TABLE 1. Characteristics of antibodies with highest affinity for aldosterone

Rabbit no.

Group 1 865 866 867 868 870 Group 2 879 882 Group 3 885 886

Relative displacement of Relative dis- Interference [l,2,6,7-3H]al- placement by by adrenaldosterone by 5 100 ng corti- ectomized pg aldosterone sol plasma

19 11 13 7 13

0 0 0 0 0

Yes Yes Yes No Yes

30 5

0 0

Yes No

14 35

0 0

No No

15

20

ALDOSTERONE (pg)

FIG. 1. Typical standard curve showing the displacement of [l,2,6,7-3H]aldosterone by unlabeled aldosterone; both were purified by the same thin layer chromatographical system. Antibody dilution was 1:3000. Each point is the mean of triplicate determinations.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 22:07 For personal use only. No other uses without permission. . All rights reserved.

JCE&M • 1978 Vol47 • No 2

SANCHO AND HABER

394

demonstrated by the extraction of different aliquots (200, 100, 50, and 25 jul) of a pooled plasma sample that had an aldosterone concentration of 400 pg/mT 1 . Results of an RIA of these extracts paralled a standard curve, indicating immunological identity of aldosterone, as extracted, with the standard. Antibody seems to have a higher affinity for unlabeled aldosterone than for the 1,2,6,7-3Hlabeled steroid. Relative interaction of labeled and unlabeled steroid with antibody is detailed in Table 2. Final dilution of the antiserum was 1:3000; the mass of labeled aldosterone was calculated from the specific activity. It is apparent that greater amounts of 1,2,6,73H-labeled aldosterone than of unlabeled aldosterone are required to produce equal displacement of the initial aliquot of tracer. Blanks The blank value obtained by assaying the dry residue of 2 ml freshly distilled methylene chloride always corresponded to the zero value of a standard curve constructed in buffer. Recovery To 100 /xl adrenalectomized plasma, 1500 cpm labeled aldosterone was added. The plasma was extracted with methylene chloride, dried, and then reconstituted with buffer. The recovery of radioactivity was 97.7 ± 2.5% (mean ±SD; n = 22).

quired to displace labeled aldosterone from its binding site. In order to test interference by a variety of steroids under actual assay conditions, a physiological or supraphysiological concentration was added to a plasma standard. No significant alteration in the expected plasma concentration of aldosterone was observed upon addition of any of the steroids tested (Fig. 2). Recovery of added aldosterone Aldosterone was added to adrenalectomized plasma and the concentration was determined by RIA. Assayed values were plotted against added aldosterone (Fig. 3). A correlation coefficient of 0.992, a slope approaching 1, and an TABLE 3. Relative amounts of steroid required to effect "*•* 50% displacement of [3H]aldosterone Aldosterone Cortisone Progesterone 17-OH-Progesterone Deoxycorticosterone Corticosterone Cortisol Testosterone Pregnenolone 11-Deoxycortisol 18-OH-Deoxycorticosterone 18-OH-Corticosterone

Control Serum

Specificity

1

3 x 1.6 x >5 x >5 x 2 x 2 x 1.2 x >5 X >1.5 x 1.6 x 1.2 x

105 10 f >

107 107 105 106 106 107 107 105 105

• DOC o Cortisone

-

3

• Cortisol • Testosterone

Cross-reactivities of a variety of steroids are listed in Table 3 in relation to aldosterone. Clearly, very high concentrations were re-

• 11 Deoxycortisol ^ Progesterone ^ Pregnenolone ? 18 OH DOC 0 Corticosterone • 17a OH Progesterone

TABLE 2. Competition of [1,2,6,7-3H]aldosterone and unlabeled aldosterone for antibody-binding sites Increment" [1,2,% Relative disUnlabeled aldos3 6,7H]aldosterone placement terone (pg/assay) (pg/assay) 15 24 38 50

3.4 5 10.6 16

1 1.9* 4 7

" Initial tracer added was 6.4 pg [l,2,6,7-3H]aldosterone/assay. 6 Interpolated point.

240 STEROID ADDED ( ng/ml)

FIG. 2. Direct determination of interference in the aldosterone assay by various steroids. Mean aldosterone concentration (heavy line) of a pooled sample was determined in 24 replicate assays. Shaded area represents 2 SD from the mean. The pooled serum was then reassayed after addition of a steroid compound. Points represent the mean of triplicate determinations.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 22:07 For personal use only. No other uses without permission. . All rights reserved.

MICROASSAY FOR ALDOSTERONE IN PLASMA EXTRACTS

395

from 39-353 pg/ml, were assayed by both methods. When the values obtained by the two methods were plotted against each other, a correlation coefficient of 0.992 was obtained.

600 -

Discussion

100

200

300

400

500

600

ALDOSTERONE ADDED (pg/ml) FIG. 3. Correlation of aldosterone recovered and aldosterone added to adrenalectomized plasma. Each point is the mean of triplicate determinations.

intercept close to 0 indicate adequate recovery and an absence of nonspecific interference by plasma. Interassay and intraassay variation Interassay and intraassay variation was determined by 12 independent assays of a plasma pool. The coefficient of variation was 10.2% (68.9 pg/jiil ± 7.06 SD). Intraassay variation was determined by the measurement of a plasma pool in sextuplicate in the same assay. The coefficient of variation was 5.6% (185 pg//il ± 4.8 SD). Normal values In Table 4 are shown the values obtained in normal subjects studied in a metabolic ward after 4-days equilibration on three different Na intakes. K intake was 100 meq/day in each instance. Blood samples were obtained at 0800 h while subjects were in a supine position and at 1000 h after 2 h of upright posture. Correlation with other methods Although the normal values obtained agree with those obtained with other published methods, we have also compared our method with those of St. Cyr et al. (6). Their method includes a chromatographic step. Eleven samples, ranging in aldosterone concentration

The production of highly specific aldosterone antibodies (13, 14) has allowed the assay of aldosterone in plasma extracts without the necessity of prior fractionation (1-9). Thus, losses are avoided and the possibility of error is reduced. Because of the relatively low affinity of antibodies previously available, 1.3 (11) or 2 (12) ml plasma were required in order to measure values within the normal range. The volume required is likely to perturb aldosterone concentration in studies requiring multiple samples or in those in which children or small animals are the subjects. Utilizing an antigen that previously gave rise to specific but lower affinity antibodies (13), we have been able to elicit a high affinity antibody that retained its specificity for aldosterone. This antibody allowed for the development of an assay 10-20 times more sensitive. A high percentage of animals immunized produced very specific antibodies. Nine of 16 animals had sera demonstrating cross-reactivity with cortisol less than 0.0002%. Two animals produced antibodies that had both very high affinity, but undiminished specificity for aldosterone (879 and 886, Table 1). One of these antisera, however, was unsuitable for assay (879) because of nonspecific interference by adrenalectomized plasma. Serum 886 retained its characteristics for 4 successive weeks. Thereafter, antibody concentration in serum decreased. After a booster injection, the antiserum showed lower specificity and affinity. This assay is limited, not by the affinity of the antibody, but by the available specific TABLE 4. Plasma aldosterone levels in five normal male subjects in balance at different levels of Na intake Normal values (pg/ml) meq Na/24 h Supine 210 110 10

48 ±29 107 ± 45 175 ± 75

Upright (2 h) 65 ±23 239 ± 123 532 ± 228

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 22:07 For personal use only. No other uses without permission. . All rights reserved.

SANCHO AND HABER

396

activity of [l,2,6,7-3H]aldosterone. It is of interest that during tritiation, [l,2,6,7-3H]aldosterone seems to undergo a change that results in lower affinity for antibody than unlabeled aldosterone. It is likely that this represents a chemical rather than an isotope effect, even though migration of the labeled and unlabeled compounds was identical on several thin layer chromatographical systems. The use of a labeled compound of different affinity for antibody than the unlabeled ligand has been used previously for the measurement of steroids (15). In the aldosterone assay described here, it may contribute to enhanced sensitivity. The small plasma volume used has simplified the extraction step, allowing the use of disposable tubes and a vortex mixer. The high recovery obtained obviates the need to correct for losses during extraction and manipulation. Even though much smaller plasma volumes are used, precision is comparable to other reported RIA methods for aldosterone. Normal values compare with those reported by other RIA methods. An added benefit of simplification has been an increase in the number of samples processed per technician. Typically, a single technician processes 195 plasma samples as well as several standard curves in an 8-h working day, exclusive of counting time. Acknowledgment We wish to thank Dr. James Melby, Boston University Medical Center, Boston, MA, for providing assayed plasma samples.

JCE&M • 1978 Vol47 • No 2

References 1. MAYES, D., S. FURUYAMA, D. C. KEM, AND C. A. NUGENT,

Radioimmunoassay for plasma aldosterone, J. Clin Endocrinol Metab 30: 682, 1970. 2. BAYARD, F., I. BEITIUS, A. KOWARSKY, AND C. MEGEON,

Measurement of plasma aldosterone by radioimmunoassay, J Clin Endocrinol Metab 31: 1, 1970. 3. UNDERWOOD, R. H., AND G. H. WILLIAMS, The simultaneous

measurement of aldosterone, cortisol, and corticosterone in human peripheral plasma by displacement analysis, J Lab Clin Med 79: 848, 1972. 4. EKINS, R. P., G. B. NEWMAN, R. PIYASENA, P. BANKS, AND J.

D. H. SLATER, The radioimmunoassay of aldosterone in serum and urine: theoretical and practical aspects, J Steroid Biochem 3: 289, 1972. 5. ITO, T., J. Woo, R. HANING, AND R. HORTON, A radioimmu-

noassay for aldosterone in human peripheral plasma including a comparison of alternate techniques, J Clin Endocrinol Metab 34: 106, 1972. 6. ST. CYB, M., J. SANCHO, AND J. MELBY, Quantitation of

plasma aldosterone by radioimmunoassay, J Clin Chem 18: 1395 1972 7. FARMER, R. W., D. H. BROWN, D. Y. HOWARD, AND L.F.

FABRE, A radioimmunoassay for plasma aldosterone without chromatography, J Clin Endocrinol Metab 36: 461, 1973. 8. VARSANO-AHARON, N., AND S. ULICK, A simplified radioim-

munoassay of plasma aldosterone, J Clin Endocrinol Metab 37: 372, 1973. 9. GOMEZ-SANCHEZ, C. D., C. KEM, AND N. KAPLAN, A radioim-

munoassay for plasma aldosterone by immunologic purification, J Clin Endocrinol Metab 36: 795, 1973. 10. MARTIN, B. T., AND C. A. NUGENT, A nonchromatographic

RIA for plasma aldosterone, Steroids 21: 169, 1973. 11. POULSEN, K., J. SANCHO, AND E. HABER, A simplified radioim-

munoassay for plasma aldosterone employing an antibody of unique specificity, Clin Immunol Immunopathol 2: 373,1974. 12. MCKENZIE, J. K., AND J. A. CLEMENTS, Simplified radioim-

munoassay for serum aldosterone utilizing increased antibody specificity, J Clin Endocrinol Metab 38: 622, 1974. 13. VETTER, W., E. FREEDLENDER, AND E. HABER, Specificity of

antialdosterone antibodies in relation to the site of coupling, Clin Immunol Immunopathol. 2: 361, 1974. 14. SLAUNWHITE, W. R., AND A. A. SANDBERG, Analysis of cor-

ticosteroids, Ada Endocrinol (Suppl 147) 64: 144, 1970. 15. MURPHY, B. E., Some studies of the protein-binding of steroids and their application to the routine micromeasurement of various steroids in body fluids by competitive protein-binding radioassay, J Clin Endocrinol Metab 27: 973, 1967.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 22:07 For personal use only. No other uses without permission. . All rights reserved.

A direct microassay for aldosterone in plasma extracts.

0021-972X/78/4702-0391$02.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1978 by The Endocrine Society Vol. 47, No. 2 Printed in U...
613KB Sizes 0 Downloads 0 Views