331
Clinica Chimica Acta, 66 (1976) 331-343 0 Elsevier Scientific Publishing Company,
Amsterdam
- Printed in The Netherlands
CCA 7433
MET~ODOLOGIC~ SIMPLIFICATIONS IN RADIOIMMUNOASSAY OF URINARY ALDOSTERONE
R. MALVANO A. SALVETTI
*, S. ORLA~DINI,
P. COZZANI, P. D~RANTI,
N. S~O~INI
and
Laboratory of Clinical Physiology, C.N.R., Pisa, Faculty of Medicine, Institute of Chemistry. E.U.L.O., Brescia, S.O.R.Z.N. Biomedical Research, Saluggia, and Clinical Medicine, university of Pisa, Pisa ~Ztaiy~ (Received
July 14, 1975)
Summary Simplification of radioimmunoassay procedures of urinary aldosterone18-gluc~onide was attempted, taking into consideration the aspects implied by the hydrolysis of urine and the assay itself. The procedure standardized for the hydrolysis step (samples diluted with a two-fold volume of 0.2 N HCl and incubated at 30°C for 16-24 h) proved suitable in terms of practicability and accuracy. Aldosterone antisera, raised in the rabbit against an aldosterone3-bovine albumin conjugate, were selected according to their specificity towards competing steroids. Depending on the characteristics of the antisera used, an assay of extracts, or even direct measurements of hydrolyzed urines excluding any extraction, were found to yield reliable results. In the case of a high-quality antiserum, evidence for the adequacy of assay of non-hydrolyzed urine extracts for the measurement of the excretion of unconjugated aldosterone was provided by some preliminary data. The results of the experiments, directed at the methodological and clinical validation of the simplified procedures, are reported and discussed in this paper.
Introduction The increasing availab~ity of highly specific antisera for aldosterone radioimmunoassay (RIA) has largely helped in simplifying the procedures through the elimination of chromatographic purification [l-8] . However, direct assays of plasma extracts can remain subject to interfering effects difficult * Correspondence to: Dr. R. Malveno, LaboratorY of Clinical Physiology C.N.R., Via Savi 8,561OO Pie& Italy.
332
to control [9] , which are seemingly related to extractable substances of lipophilic nature present in the samples [lo]. In this connection, RIA of urinary aldosterone-1Sglucuronide is favoured with respect to plasma assay. Furthermore, the measurement of aldosterone excretion is less demanding in terms of both sensitivity and specificity: the concentration of urinary aldosterone glucuronide is in fact much higher than that of aldosterone in plasma, and the cross-reactivity of the main corticosteroid catabolites present in urine is expected to be lower than those due to steroids like cortisol and corticosterone circulating at relatively high levels. As for its clinical significance, the assay of urinary aldosterone excretion remains fully valid when compared to the measurement of circulating aldosterone. In fact these physiological parameters are closely related for both normal subjects and patients with essential hypertension [ 111. Moreover, the daily urinary excretion provides integrated information on adrenal secretion [12], while plasma aldosterone only reflects a single-point situation in time. The need for a 24-h collection of urine is a practical complication, which is, however, often also encountered for plasma assay as the daily sodium excretion is generally assumed as an index of the sodium balance [ 11-131. RIA of urinary aldosterone-glucuronide is somewhat particular, in that antagonist factors influencing accuracy can be involved. On one hand, depending on the individual characteristics of the antiserum used, systematic overestimation errors could take place as a result of an insufficient specificity to cross-reacting steroids; on the other hand, underestimations can be associated to both incomplete conversion of the conjugate into the aldosteronefree form and to the occurrence of aldosterone degradation during hydrolysis. Simplified procedures for the measurement of urinary aldosterone have been standardized at our laboratories, through the selection of antisera and the optimization of the conditions of hydrolysis. The results obtained, together with those concerning the validation of methods, are presented and discussed in this paper. Experimental Materials and reagents Unlabelled steroids were purchased from Steraloids (New York, U.S.A.); [1,2,6,7-3 H] aldosterone with a specific activity of about 250 mCi/mg was supplied by CEA-IRE-SORIN. The antisera were raised in rabbits against the aldosterone-18,21diacetate3-(0-carboxymethyl)oxime derivative conjugated to bovine serum albumin (BSA) according to the procedure reported by Bayard et al. [ 11. Two antisera out of five were selected for this study as potentially suitable for direct assays, on the basis of the specificity characteristics resulting from the multiple titration curves shown in Fig. 1 (antisera A and B). 0.04 M phosphate buffer at pH 7.4, containing 0.5% BSA, was used as a diluent in the RIA system. Suspension of 2 mg/ml of charcoal-dextran (10 : 1) (DCC) in phosphate buffer was employed to separate the free from the anti-
0 I
Fig. 1. Dose-response curves for aldosterone and cross-reactant steroids for the two selected antisera A and B. The antisera were chosen within a group of five obtained in a single immunization cycle, being other two antisera similar to B and tbe last with k poorer specificfty. 1. Aldosterone; 2. tetrabydrocortisol; 3, tetrabydrocortisone; 4, pregnanetaiol; 6, allotetrabydrocortiiol; 6. cortisone; 7. cortisol; 8, cortexalone; 9. corticosterone; 10, cortexone; 11. testosterone; 12, progesterone; 13. estradiol.
body-bound radioactivity. The dichloromethane (DCM) used as extractant, as well as the other chemicals, was of reagent-grade purity.
methods The reference procedure adopted basically followed methodological schemes previously reported [ 2,7,8] , consisting of the successive steps of
334
pre-extraction of samples to remove unconjugated steroids, hydrolysis of urine, extraction and washing of extracts, and RIA itself. Simplified procedures, omitting some of the preliminary operations, were tested for validity as hereafter discussed. (a) Pre-extraction (urine washing). l-ml urine aliquots were extracted with 15 ml DCM by manual agitation during 1-2 min; the organic phase was discarded. (b) Hydrolysis (standardized procedure). 0.5 ml of urine, brought to 1.5 ml with 0.2 N-‘HCl, were incubated at 30°C for 16-24 h. (c) Extraction. 0.3 ml of hydrolyzate, corresponding to 100 ~1 of original urine, were extracted with 5 ml of DCM, by stirring on Vortex for 2 min; a mean recovery of 89 f 5% was evaluated using tritiated aldosterone. (d) Extract washing. 2 ml of extract were washed successively with 0.5 ml of 0.05 M NaOH, 0.5 ml of 0.5 M acetic acid and 2 ml of distilled water; 6.25-ml fractions of extracts (corresponding to 5 ~1 of original urine) were transferred in glass tubes and dried under a nitrogen stream. (e) Radioimmunoassay. 0.1 ml of the standard solutions were added to the tubes prepared for the calibration curve. To all the tubes, both standard tubes and tubes containing the extract residues, 0.1 ml of tracer solution (corresponding to 25 pg of tritiated aldosterone), 0.1 ml of antiserum (diluted to bind approximately 40% in the absence of unlabelled steroid), and phosphate buffer to an overall volume of 0.5 ml were added in turns. The tubes were kept at room temperature for approx. 30 min to ensure a complete solubilization of steroidal residues, and then incubated at 4°C during 2 h. 0.5 ml of DCC suspension were delivered keeping the bulk suspension under agitation, and both suspension and RIA tubes refrigerated in an ice-bath. After a lo-min waiting time, the tubes were centrifuged in the cold at approx. 2000 X g for 10 min, and 0.5-ml aliquots of each supernatant transferred for P-counting. Results Optimization
of hydrolysis
conditions
In order to take into consideration the different experimental parameters governing hydrolysis, aliquots of a pooled urine sample were incubated under various conditions of dilution, pH, temperature and time. Curves indicating the evolution of immunoassayable aldosterone with time, as those shown in Fig. 2, were obtained, and the data of Table I, referring to 16- and 24-h hydrolysis, were derived by interpolation. Both figure and table indicate how the measurable amount of aldosterone depends on the experimental conditions, being underestimations attributable either to low hydrolysis rates (higher pH values, greater dilutions, lower temperatures) or to high degradation rates (lower pH, higher temperatures). To account for these contrasting factors, an overnight incubation at 30°C of samples diluted with a two-fold volume of 0.2 M HCl was eventually adopted. In these conditions, in fact, pH was found to range approximately from 0.9 to 1.2 (mean value 1.01 L 0.07 S.D., 30 random samples with original pH from 5.0 to 8.5); for samples with pH values within this interval, virtually
335
?
5-
8 fz 4
I
........ ....... ........ ....... ........ ....... ........ ....... ........ .:s:.:.:.:.:.:. ........ ....... ........ ....... ........ ....... ........ ....... ............... ..............
I
I
60
40
20
0
HYDROLYSIS
TIME
(hr 1
Fig. 2. Amount of aldosterone measured in aIiquotS of pooled urine samples adjusted at different pH values, as a function of the time of hydrolysis at 3O’C. The shaded area indicates the time interval chosen for routine measurements.
constant estimates were observed in the hydrolysis time range from 16 to 24 h at 30°C (90-100% of the maximum estimate, see Fig. 2). Nevertheless the occurrence of aldosterone losses cannot be unequivocally excluded simply on the basis of these data, as the plateau values of Fig. 2 could result from the counterbalancing effects of low hydrolysis rates and degradation. Therefore a direct control of aldosterone stability was made through evaluation of the recovery of exogenous aldosterone added prior to hydrolysis, The results reported in Table II show that, even in the probably
TABLE
I
CHECK OF THE EFFECTIVENESS DOSTERONE-18-GLUCURONIDE
OF PARAMETERS
GOVERNING
THE
HYDROLYSIS
OF AL-
AIiquots of the same pool of urines were used. The complete procedure (see “Experimental”) was followed, using antiserum. B; the results were obtained by interpolating kinetics curves Iike those shown In Fig. 2. Urine dilution
l/1.2 l/3 l/20 l/3 l/3 l/3 l/3 l/3 l/3 l/3 l/3
PH
1.0 1.0 1.0 1.25 1.0 0.88 1.25 0.88 1.25 1.0 0.88
Temperature
25 25 25 20 20 20 25 25 30 30 30
(“C)
Aldosterone
measured
16 h
24h
21.1. 18.3 16.7 14.3 15.8 19.4 15.4 21.8 20.5 21.8 22.0
22.0 20.9 18.7 16.3 18.3 21.1 17.2 22.0 21.8 22.0 20.9
@g/ml)
_
336
TABLE
II
CHECK
OF
The
refer
data
STABILITY
OF
to the complete
Sample
EXOGENOUS procedure
Aldosterone
ALDOSTERONE
(see
Experimental)
DURING using
antiserum
HYDROLYSIS
(20
h,
3O’C)
B.
(pg)
% Recovery ____
Added 1 2
0
17
80
96
0
52
40 3 4
88
0
212
80
278
0
112
0
N HCI)
99
36
90
66
82
77
95
38
95
41
102
0
40 6 (0.1
79
35
80 5 (adrenalect.)
Recovered
Measured
38
0
0
40
41
~___
pessimistic situation of aldosterone being present since the beginning of hydrolysis, a 94% recovery was obtained as an average, thus demonstrating the substantial accuracy associated with the hydrolysis step. Evaluation
of assay accuracy
The selectiveness of measurement was checked by comparing the complete procedure (see Experimental) using the less specific antiserum B with gas-liquid chromatography [14] as a reference method. Despite the few samples assayed in parallel, the similarity of the results shown in Table III seems to exclude the occurrence of relevant systematic errors. An additional element in favour of the reliability of the RIA method was given by the equivalence of the analytical information provided by the two antisera with different specificity A and B, as shown in Fig. 3. In order to check the consistency of the analytical response with increasing doses and sample concentrations, dilution tests were carried out: the results
TABLE
III
COMPARISON Data
under
Sample
OF RIA
RESULTS
referring
OBTAINED
to the complete
Aldosterone
measured
WITH procedure @g/24
RIA
AND
(see
G.L.-CHROMATOGRAPHY
Experimental)
using
antiserum
B.
h) ___
RIA
GLC
1
5.8
5.8
2
7.8
8.8
3
12.7
13.5
4
21.2
23.8
5
40.2
37.3
337
y = 0.67+0.96 r
q
l
x
0.960
Nz25
40
20
”
ANTISERUM A, w/ml Fig. 3. Correlation of results obtained with antisera A and B. The data refer to the complete procedure (see Experimental) with random urine samples collected during the day. No significant difference is indicated between equivalence and calculated regression @ > 0.76).
0
ANTISERUM
A
.
ANTISERUM
B
6
400 0
0
/
/ 200 -
0
P
I
I
10
20 /I’
URINE
Fig. 4. Test of parallelism through the assay of different volumes of extxacts of a hydrolyzed urine %mPle using antisera A and B. The data refer to the complete procedure (see Experimental). The volumes of extracts are expressed as the corresponding volumes of original urines. Parallel responses are apparent up to volumes beyond those routinely used (corresponding to 5 ~1 of urine).
338 TABLE
IV
COMPARISON OF RESULTS All the data refer to antiserum Sample
Aldosterone
OBTAINED B. measured
Washed extract 11.1 17.0 92.5 20.4 27.5 20.3 8.3 16.7
*
See Experimental.
*
ON WASHED
AND UNTREATED
EXTRACTS
(ng/ml) Untreated
extract
9.5 17.1 99.6 22.5 29.3 21.9 7.7 14.6 step d.
reported in Fig. 4 indicate a parallel response for both antisera up to concentrations of samples higher than those routinely used (i.e. corresponding to 5 /.Aof original urine). Simplification of procedure No significant improvement
of accuracy was found to be associated with the washing of extract (see step d under Experimental) In fact, even
PRE -EXTRACTION
no/ml
Fig. 5. Correlation of the assays including avd omitting the step of dichloromethane extraction prior to hydrolysis, using antisera A and B. Random urine samples collected during the day were used. In the case of antiserum A no significant difference results between equivalence and calculated regression (p > 0.75). while the opposite was found for antiserum B @ < 0.01).
339
4oc
0
5
ANTISERUM
A oAc 1
ANTISERUM
6 A
10
-I 15
pl URINE Fig. 6. Test of parallelism through the assay of different volumes of hydrolyzed urine samples using antisera A and B. A non-parallel response is apparent for antiserum B, while a linearity of measured aldosterone with the sample volumel’is evident in all cases for antiserum A up to larger volumes than routinely used (5 ~1).
using the less specific antiserum B, similar data result for washed and untreated extracts according to the estimates of Table IV. The correlation of results obtained for urine fractions where pre-extraction (see step a under Experimental) was included or omitted, indicated a different behaviour according to the characteristics of the two antisera. As Fig. 5 shows, no significant difference resulted for the highly specific antiserum A, while a systematic overestimation of about 10% was associated with the use of antiserum B. Checks for the reliability of direct RIA of hydrolyzed urine by-passing extraction (see step c under Experimental) were also carried out. The dilution tests of Fig. 6 show a parallel response to increasing sample concentrations for antiserum A, whereas the poorer characteristics of antiserum B are reflected by a marked non-linearity. The absence of overestimation errors in the case of direct assay with antiserum A is confirmed in Fig. 7 by the equivalence of results observed between hydrolyzed samples and the relative extracts. Assessment of the methods The evaluation of within and between assay variability (see Table IV) put in evidence, as expected, a higher precision for the simplified procedtie allowed by antiserum A, with respect to the procedure including extraction. In any case, however, there resulted an acceptable repetition of data. The lower detection limit computed from the dose-response curves,
340
;r:,-.-.
SOI-
r = 0.666
l
!5
a
E I 0
0
40a
0
1 20
I 60
I 40 RIA
OF DRIED
w/ml
EXTRACT
Fig. 7. Correlation of the assays of hydrolyzed urine samples and the relative extracts, using antIaenun A. Random urine samples collected during the day were used. There was no significant difference between equivalence and calculated regression at a level of P = 0.8.
taking into account both uncertainty of the “zero” definition and of the curve slope [15] , was found to average approximately 3 pg for both antisera. This corresponds to approx. 0.6 pg/24 h under the standardized conditions. Whenever necessary, assay sensitivity was further improved by taking larger volumes of hydrolyzate for the measurement. The information provided by RIA of urinary aldosterone was validated by assaying samples corresponding to well-defined physiopathological situa-
TABLE
V
VARIABILITY
OF RIA OF URINARY
The data refer to antiserum
ALDOSTERONE
A. Aldosterone
@g/ml)
-Direct RIA
RIA of extract Mean
S.D.
C.V.
Mean
S.D.
Within assay (8 replicates)
5.8 26.4
1.1 3.4
18.4 12.8
5.6 27.0
0.4 1.6
7.9 5.9
Between assay (8 replicates)
6.1 24.8
1.5 3.9
25.1 15.7
5.9 27.0
0.8 2.7
13.1 10.0
C.V.
l
l
l
l l
le
l
l
l
l l l
l
l l *
88
le
se lee
l
l l l e
l
l
l
l
l 100
l
l
3000
200
Na EXCRETION t meq/24
2
hr 1
6
4
PRA tng angiotensin
6
I/ml,hr 1
Fig. 8. Corzelation of urinary excretion of aldosterone-18-gIucuronide with sodium excretion (A) and plasma renin activity (upright position) (B) in normal subjects. The aldosterone assay was carried out directly in urine samples using antiserum A.
tions and by establishing the correlations with some other physiological parameters. Fig. 8 shows the correlations obtained with sodium excretion and plasma renin activity [16] in normal subjects; for 5 subjects affected by aldosteroneproducing adenoma (sodium intake 50-150 mequivJ24 h) the excretion rate of aldosterone-l&glucuronide was found to range from 36 to 63 pg/ TABLE
VI
CORRELATION URINE
OF
ALDOSTERONE
ESTIMATES
The data refer to assay of extracts and direct RIA. Subject *
Normal
Aldosterone
1 2 3 4 5 6
Aldosterone-producing adenoma *On uncontrolled
diet.
IN
NON-HYDROLYZED
respectively,
AND
HYDROLIZED
using antiserum A.
measured @g/24 h)
bla (%)
a. hydrolyzed
b. non-hydrolyzed
16.0 27.6 13.7 16.7 7.9 18.0
0.34 0.68 0.27 0.32 0.14 0.35
2.12 2.10 1.97 2.03 1.77 1.04
60.1
1.18
1.96
342
)I’
URINE
Fig. 9. Test of parallelism through the assay of different volumes of extracts of a non-hydrolyzed urine sample using antisera A and B. The extract volumes are expressed as the corresponding volumes of original urine. Parallel response is apparent for antiserum A up to volumes larger than those used for RIA (100 ~1). while non-parallelism and higher estimates result for antiserum B.
24 h (56.3 f 10.4 mean and S.D.), while for a single case of bilaterally adrenalectomized patient no measurable amount of aldosterone (GO.2 pg/24 h) resulted. All these data are consistent with the ones previously reported [2,7, 12,13,14,17,18,19]. Applicability
to assay of unconjugated
aldosterone
The possibility of measuring unconjugated aldosterone excretion by directly assaying the extracts of non-hydrolyzed urine was checked. The results of the dilution tests shown in Fig. 9 indicate a non-parallelism for antiserum B; parallel response and lower estimates, on the contrary, were obtained in the case of antiserum A. RIA, using the latter antiserum, gave for some samples the results reported in Table VI. Both levels and ratios with the excretion of 18-glucuronide are in good agreement with previously published data [ 18,201. Discussion In this work, aimed at simplifying the methodology of the assay of aldosterone excretion without affecting the assay quality two aspects have been considered. For the former, concerning the hydrolysis step, a compromise solution b,etween the antagonist factors involved was adopted, standardizing a procedure ‘which allowed a sufficient accuracy to be retained without any need for the individual adjustment of the urine pH. In terms of RIA, on the other hand, the results obtained confirm the crucial importance of the individual characteristics of the antiserum used, on which the possibility of any methodological simplification ultimately depends.
343
In this respect, the adequacy of direct assay of extracts of urine hydrolyzates, proved for an “average’‘-quality antiserum like B, seems of practical interest. Also the elimination of the pre-washing operation, intended to remove interferences from unconjugated steroids could be accepted, at least for screening measurements, when slight overestimations as those found for antiserum B are involved. The availability of antisera with an exceptional quality, which could be the case of antiserum A, seems to remain a matter of chance rather than the result of particular choices in immunization. The high quality of antisera can be exploited for further improvements of the practicability of assay, such as the avoidance of any extraction step. On one hand this enables the operation time to be decidedly shortened, which is a practical advantage especially in a routine context. On the other hand, the lower number of pipetting operations involved and the absence of uncertainty related to the extraction yield are reflected by a higher level of overall precision. As far as the excretion rate of unconjugated aldosterone can be regarded as a useful physiopathological parameter [X3], the feasibility of simple measurements of extracts of non-hydrolyzed urine appears interesting. The reliability of such an assay, which seems confirmed by the preliminary results reported above, is therefore to be considered as a further advantage implied by the availability of high-quality antisera. References 1 Bayard, F.. Beitkins. I.Z., Kowarski, A. and M&eon. C.J. (1972) in Techniques radioimmunologique, p. 489. INSERM, Paris 2 Vetter. W., Vetter, H. and Siegenthaler, W. (1973) Acta Endocrinol. 74. 548 3 Vetter. W., Vetter, H. and Siegenthaler. W. (1973) Acta Endocrinol. 74, 558 4 Jowett. T.P., Slate& J.D.H.. Piyasena, R.D. and Ekins, R.P. (1973) Clin. Sci. Molec. Med. 45, 607 5 McKenzie, J.K. and Clements, J.A. (1974) J. Clin. Endocrinol. Metab. 38,622 6 Pham-Hun-Tnmg, M.T. and Corvol, P. (1974) Steroids 24. 587 7 Langan, J.. Jackson, R., Adlin. E.V. and Channick, B.J. (1974) J. Clin. Endocrinol. Metab. 38, 189 8 Deck, K.A. and Eberlein, L. (1974) Z. Klin. Chem. Klin. Biochem. 12,504 9 Giannotti, P.. Mannelli, M.. Fiorelli, G. and Serio, M. (1974) J. Nucl. Biol. Med. 18,104 10 Malvano, R. and Rolleri. E. (1975) in Radioimmunoassay of steroid hormones (Gupta. D., ed.), Verlag Chemie, Veinheim 11 Buhler, F.R.. Laragh, J.H.. Sealey, J.E. and Brunner, H.R. (1973) Am. J. Cardiol. 32. 554 12 Laragh, J.H.. Sealey, J.E. and Sommers. S.C. (1966) Circ. Res. 18.158 13 Sealey, J.E., Buhler, F.R., Laragh. J.H., Manning, E.L. and Brunner. H.R. (1972) Circ. Res. 31. 367 14 Bravo, E.L. and Travis, R.H. (1967) J. Lab. Cli. Med. 70,831 15 Rolleri, E., Novario. P.G. and Pagliano. B. (1973) J. Nucl. Biol. Med. 17.128 16 Malvano. R., Zucchelli. G.C., Gasser, D. and Bartolini, V. (1974) Clin. Chim. Acta 50,161 17 Salvet,ti, A., Arzilli, F., Simonini. N., Maltinti, G. and Baccini, C. (1972) Arch. Sot. It. Chir., 74O Congresso, Roma, 15-18 Ottobre. 1972, p. 37 18 Deck, K.A., Champion, Jr., P.K. and Corm. J.W. (1973) J. Clin. Endocrinol. Metab. 36. 756 F. (1973) in Research 19 Salvetti, A., Arzilli, F.. Simonini. N.. Baccini. C.. Maltinti. G. and Tronchetti, on Steroids (Breuer, H.. Conti, C.. kughes. A., Jungblut, P.. Klopper. A. and Lerner. L., eds.), Vol. VI, P. 421, North-Holland, Amsterdam 20 Vecsei, P.. Penke, B. and Joumaah. A. (1972) Experientia 28,622