Radioimmunoassay of Arginine Vasopressin in Human Plasma
501
Furthermore its response resembles that obtained after administering 1 mg of tetracosactide which indicates that it possesses a potency some six times larger. This result corresponds with the findings obtained by other authors with different methods (Maier et al. 1971, Keenan et al. 1971). References
Appleby, J.I., G. Gibson, I.K. Norymberski, R.D. Stubbs: "The determination of 17-hydlOxycorticosteroids". Biochem. J. 60: 453460 (955) Bagazgoitill, F.J., A. Oriol Bosch: "Automatization por flujo continuo de la reaccion de Zimmermann para la determinacion de los 17 -cetosteroides y de los esteroides 17-cetogenicos totales urinarios". Archivos Facultad Medicina Madrid 26: 203-216 (1974) Bell, P.H.: "Purification and structure of ß-corticotropin" J. Amer. Chem. Soc. 76: 5565-5567 (1954)
tion of free II-hydroxycorticosteroids in human plasma". J. Clin. Path. 15: 374-379 (1962) Ramachandran, J., D. Chung, Ch. H. Li: "Adrenocorticotropins XXXIV. Aspects of structure-activity relationships of the ACTH molecule. Synthesis of a heptadecapeptide amide, an octadecapeptide amide and a nonadeca· peptide amide possessing high biological activities". J. Amer. Chem. Soc. 81: 2696-2708 (1965) Retiene, K., F. Schultz: "Human pharmacological studies with a new synthetic corticotrophic peptide consisting of 18 aminoacids". Acta Endocrinol. Suppl. 173: 33 (1973) Rittei, W.: "Techniques for the synthesis of ACTH and MSH peptides and analogues". "Pharmacology of hormonal polypeptides and proteins". N. Bach, L. Martini and R. Paoletti eds. Plenum Press, New York 1968,3547 Schwyzer, R.: "Programierte Molekein". Experientia 26: 577-587 (1970) Seelig, S., G. Sayers: "Structural activity relationship among ACTH". Fedn. PlOc. 30: 693 (1971)
Requests for reprints should be addressed to: Dr. A. OrioJ-Bosch, Dept. Exper. EndocrinoL. University Complutensis Medica1 School, Madrid (Spain).
Hormon. Metab. Res. 7 (1975) 501-507
© Georg Thieme Verlag Stuttgart Radioimmunoassay of Arginine Vasopressin in Human Plasma* E. Uhlich**, P. Weber, U. Gröschel-Stewart and T. Röschlau 11. Medizinische Klinik der Universität München, and Frauenklinik der Universität Würzburg, Germany
Summary Antibodies for the radioimmunoassay of arginine vasopressin (A VP) described here were produced in rabbits using synthetic AVP coupled to rabbit 'Y-globulin with carbodiimide. In three out of six rabbits, significant antibody titres were obtained. Using the best antisera produced' 40% of labeled AVP was bound at a final dilution of 1 :50.000. After iodination of synthetic AVP with 125 1 using the chloramin-T method, a gel filtration on Sephadex G-25 was performed •
A preliminary report of part of this work was published in Acta end ocr. (Kbh.) Suppl. 184 (1974) 52. .. Supported by the Deutsche Forschungsgemeinschaft.
Received: 22 Nov. 1974
Accepted: 30 Aug. 1975
to punfy the iodinated AVP. For separation of antibody bound and free hormone, a second antibody precipitation was used. There was no crossreactivity with oxytocin. AVP was extracted from plasma after ammoniumsulfate precipitation of the proteins by adsorption to Florisil. The recovery of AVP added to plasma in amounts between 5-25 pg/ml was 60 ± 15% (n=6). The minimum amount of AVP detectable was 1 pg per ml plasma. The plasma level in normal adults under standard conditions was 3.4 ± 2.2 pg/ml. This is in agreement with data recently published by other researchers. The applicability and reproducibility was further tested in measurements of sampies taken hourly during the entire day under water diuresis and after hormonal stimulation of AVP. Key-Words: Vasopressin - Radioimmunoassay - Secretion Pattern.
Downloaded by: Universite de Sherbrooke. Copyrighted material.
The more prolonged increase of II-OHCS following KeeNm, J., J.B., Thompson, M.A. Chamberlain, G.M. Besser: "Prolonged corticotrophic action of a synthetic substiintravenous administration of C-41795-Ba is also tuted 1-18ACTH". Brit. Med. J. I: 742-743 (1971) reflected in the higher urinary steroid excretion. This Landon, I., V.H. T. James, M.J. Wharton, M. Friedman: could be interpreted as a consequence of the rate of "Threshold adrenocortical sensitivity in man and its poslocal inactivation of the polypeptides following subsible application to corticotrophin bioassay". The Lancet 11: 697-700 (1967) cutaneous or intrarnuscular administration being sigLowry, P.J., C. McMartin, J. Peters: "Properties of a simplinificant in relation to their half-life in either the fied bioassay for adrenocroticotrophic activity using the plasmatic or adrenal compartments. steroidogenic responsed of isolated adrenal cells". J. Endocr. 59: 43-55 (1973) Finally one should consider that 0.125 mg doses of Maier, R., P.L. Barthe, L. Schenkel·Hulliger, P.A. Desaulles: C-41795-Ba, even though capable of a marked adre "The biological activity of (1-D-serine, 17-18-Dilysine) adrenal stimulation, do not interfere with the ß-corticotrophin-( 1-18)-Qctadecapeptide-amide". Acta circadian rhythm since the effect disappeared after Endocrinol. 68: 458466 (1971) Mattingly, D.; "A simple fluorimetric method for the stima8 hours.
E. Vhlich, P. Weber, V. Gröschel-Stewart and T. Röschlau
Introduction The bioassays of AVP that have been used up until now for studies of hormonal fluid regulation are not sensitive enough to detect minimal, although biologically significant, changes of this hormone. Therefore, in order to study mechanisms regulating the release of AVP, methods had to be designed which would permit the detection of slight fluctuations of the normally very low endogenous levels of AVP. This is also pertinent to various disease states in which the role of AVP should be characterized not only qualitatively but also quantitatively. The concept, for example, that AVP is excessively elevated in situations described as the syndrome of inappropriate ADH secretion (Schwartz, Bennet, Curelop and Bartter 1957) is by no means true in all instances. Recently, radioimmunological measurements of plasma AVP in such patients (Robertson and Mahr 1972) showed normal or even low plasma levels of AVP in some, thereby, indicating that a "syndrome of inappropriate secretion of AVP without inappropriately raised plasma levels of the hormone" does exist. Even though the presence of a hormone with a different radioimmunological characteristic is not excluded. Since some of the difficulties of such radioimmunologically AVP determinations are the production of specific antibodies and the extraction of AVP from plasma sampies, we wish to present a detailed description of our efforts in this field which led to an improvement of the assay. In order to test the applicability and reproducibility of the assay in clinical situations, plasma AVP was measured every 60 minutes in 6 normal human volunteers for aperiod of 24 hours. The nocturnal values were found to be higher (6.2 ± 1.6 pg/ml) when compared with mean plasma levels during day time (3.9 ± 0.8 pg/ml). An episodic rhythmicity of plasma AVP levels was evident in all individuals tested with peak levels (2-6 times the basal level) every 3-4 hours.
ear vein and tested for AVP antibody titer: Terminal coUecHon of the serum was performed by catheterization of the caro t id art ery .
Radioiodination of Vasopressin Synthetic arginine vasopressin was iodinated using the chloramin-T method described by Greenwood. Hunter and Glover (1963): 7.5 U AVP in 20/.11 phosphate buffer were exposed for 10 seconds to 10/.11 Na/2 5 1 solution (specific activity about 100 mCi/ml) after addition of 10 /.11 chloramin-T solution (2.5 mg/mi). The reaction was stopped using 200 /.11 bovine serum albumin (pentex ) 4.2 mg/mI. The iodinated AVP solution was then applied to a 1.5 x 40 cm column of Sephadex G 25 (fine) and equilibrated with 0.25% acetic acid. Three radioactive peaks were eluted with the same solvent. The largest one (about 90% of the total radioactivity) contained the labeled hormone. One or two fractions (2 ml each) of the labeled AVP peak were diluted to a final radioactivity of about 1200 cpm per 50 /.11 and then added to the sampies used in the assay.
Assay Standard curves were obtained after mixing 200/.11 of diluent buffer (0.015 M NaH 2 P04 • 0.15 M NaCl, 0.02% thiomerosal, 2 mg/mi bovine serum albumin (Pentex). 0.001 M EDTA adjusted to pH 7.4), 250 /.11 of a freshly prepared AVP solution in varying concentrations and 50 /.11 of the antiserum which had been diluted with the above mentioned phosphate buffer to a concentration of 1 :5.500 (final concentration 1:50.000). For the measurement of plasma sam pies. the dried extracts (see below) were redissolved in diluent buffer and. after 500/.11 of diluent buffer had been added. 200 /.11 were used in the assay. After 24 hours preincubation of unlabeled standards or unknown sampies at 4°e. 50 /.11 of 125 1 AVP (1.200 cpm) were added to all of them, mixed and then reincubated for an additional 24 hours at 4 0 C, In order to separate the bound and free hormone, a second antibody (antirabbit ')'globuline from donkey, Wellcome) was added to the above incubates and after thorough mixing, left at 4°C far another 24 hours. It was then centrifuged at 4 0 C for 10 minutes at 2000 g. The supernatant containing the unbound hormone was discarded. The radioactivity of the precipitated antibodybound hormone was counted directly in an automatic gamma counter (Tracer Lab), connected to an off-line printer (Teletype).
To test the accuracy of the whole assay procedure vasopressinfree plasma sam pies were assayed with and without arginine vasopressin in three different concentrations (5, 10, 25 pg/ml) on six different occasions. To analyze the inter· Methods assay variation coefficient, a pooled plasma sampie with Preparation of antigens and A VP antisera known vasopressin content was taken as an "internal standard" The antigen complex was prepared in a slightly modified form in all assays perfarmed. as described by Wu and Rockey in 1969: 2 ml of a solution Extraction of Plasma Sampies containing 300 IV = 0.75 mg of synthetic arginine vasopresVenous blood sampies (at least 15 ml for duplicate assay) sin (Ferring, Malmö/Sweden), 40 mg rabbit ')'-globulin from both recumbent controls and patients were collectcd (Serva, Heidelberg) and 30 mg l-ethyl-3 (3-dimethyl-aminointo cooled (4°C) heparinized plastic tubes and immediately propyl)-carbodiimide-HCI (Sigma, SI. Louis, Mo.) were adcentrifuged at 4 oe. Plasma sampies were stored in plastic justed to a pH in the range of 7.0-7.5,using brom thymol tubes at -20°C and no detectable 10ss of AVP activity could blue as an indicator and the solution was kept at room be found over aperiod of several wecks. temperature for 16 to 18 hours. 2 ml Freund's complete adjuvant (Difco) and 1-3 mg mycobacterium tuberculosis After acidification to pH 5.8~.2 with 1 N HCI, equal were then added and the homogenized mixture injected amounts of saturated ammonium sulfate solution were added both subcutaneously into the fossa poplitea and intradermal- to 3 or 4 ml plasma aliquots. After mixing. centrifugation Iy between the shoulder blades. Each rabbit was givel} an at 3.000 rpm for 5 minutes at 4 0 C and decanting. the dear aliquot of the mixture containing 50 IV of the arginine supernatants were thoroughly mixed for 15 minutes with vasopressin. The animals were boostered at three week inter- 100 mg Florisil (100-200 mesh. Serva. Heidelberg) per vais with the above mixture alternating the intramuscular, sampie. The resin was washed 10 times with distilled water subcutaneous and intraperitoneal route. 8-10 days after and dried at about 40 0 C before use. The supernatant was each injection, a blood sampie was taken from the lateral discarded and the Florisil portion was washed twice with
Downloaded by: Universite de Sherbrooke. Copyrighted material.
502
Radioimmunoassay of Arginine Vasopressin in Human Plasma AB. TITRE AT
"I. BINDING OF LABELLED AVP
40·,. BINDING
so
, :70 000
........ •••••••
"
.....
......
. :
1:30000
:~:---------
.
I
3
4
..---
,,'
, .. ..
...7··,,••
IV
••••• 11
v I
5 6 MONTH CF IMMUNIZATION
Fig. I Response of rabbits (animal 11, IV, V) to immunization with 50 IV AVP coupled to rabbit ')"-globulin, given at 21 day intervals. . 1 ml each of 1 N HCI and double distilled water. The adsorbed AVP was finally eluted from the adsorbent by vigorous mixing for 15 minutes with 1 ml aceton water (80:20) v/v). After centrifugation and separation, the supernatant was washed with 1 ml of diethylether. The aqueous phase was Iyophilized and, without further storage, assayed after dissolving in 500 1'1 of diluent buffer, in duplicates of 200 1'1 each.
Calculations Calculations of radioimmunoassay standard curves of A VP were performed with the spline approximation method. This has been shown to be a better procedure for the calculation of radioimmunoassay standard curves than those previously used (Marschner, Dobry, Erhardt, Landersdorfer, Popp, Ringel and Scriba 1974). In this method, trial functions between the measure points of radioimmunologically obtained standard curves are calculated keeping the second derivative of the whole spline funchon continuous and the oscillation of this curve, depending upon the smoothing parameter, at aminimum. The advantages of this method compared with conventional approximations (e.g., logit transformation) have been discussed elsewhere (Marschner et al. 1974). A convenient Fortran IV program has been developed for a Siemens 404/3 computer.
Fig. 2 Crossreaction of various vasopressin analogues and oxytocin with anti-arginine-vasopressin (IV /4). PLV 2 = phe-Iys-vasopressin; POR 6 = ornithin vasopressin; Ls = Iys-vasopressin; DDAVP = desamin-(d-args ) vasopressindiacetat.
Crossreactivity with various AVP analogues has been tested with antibody IV /4. In contrast to most data reported in the literature (Edwards, Chard, Kitau, Forsling, and Landon 1972), the immunoreactivity with AVP was highly specific. In Figure 2, binding curves of AVP antiserum with different analogues (arginine vasopressin, arginine vasotocin, phenylalaninelysine vasopressin, ornithine vasopressin, desaminovasopressin and oxytocin) are shown. There was no crossreactivity with oxytocin, the binding - as compared to AVP - was less than I: 10.000.
Labeling The labeling of the hormone with 125 I using the chloramin-T method was time dependent (Fig. 3); a 10 second incubation time was found to be sufficient for the introduction of 90% of the iodine into the AVP molecules. A typical elution profile of radioactivity is depicted in Figure 4.
Peak 1 (fractions 7-10) contains 125 Ibound to protein measured at 280 nm (Zeiss MK2), peak 2 (fraction 15-20) represents free iodine (measured titriResults metrically). 90% of the total amount of the radioAntibodies activity was covered by peak 3 (fractions 20-40) representing iodinated AVP loosely adsorbed to the gel. Sufficiently high antibody titers were found in 3 This elution pattern is exactly the same as has been out of 6 rabbits; 2 developed diabetes insipidus shown by Oyama, Kagan and GUck (1971) and during the period of immunization. The titers measured in 3 rabbits and expressed as 40% bind- Robertson, Roth, Beardwell, Klein, Petersen and ing of a constant amount of labeled hormone are Gorden (1973). Biologically active, noniodinated AVP shown in Figure I. All results described here were was found between peak 2 and peak 3 (measured in rat bioassay). The two fractions following the top of obtained with the antiserum IV /4 in a final dilution of 1 :50.000 except where other specifications peak 3 were used as tracer in the assay. They were stored in diluted form at -20o C and gave reproducible are noted. results for up to about 8 weeks.
Downloaded by: Universite de Sherbrooke. Copyrighted material.
,: 50000
1:10000
503
E. Uhlich, P. Weber, U. Gröschel-Stewart and T. Röschlau
504
PROTEIN·PEAK ~.
100
90
Fig. 3 Time dependent effect of chloramin-T on the iodination of AVP with tracer amounts of 125 1, separation: Sephadex column K 9/15; G-25 fine; ordinate: percentage of total amount of 125 I.
ethylalcohol, etc.) is usually followed by an adsorption step. Florisil was used and found to give adequate The effect of calcium ions (0 to 0.001 M) or EDT A and reproducible results. However, we found it neces(0 to 0.02 M) on binding ability of different antisary to test each batch of Florisil to get an "individu· vasopressins is shown in Figure 5. Higher concentraal" adsorptive characterization since the recovery of tions of either EDT A or Ca inhibited the binding of added AVP varied from one batch to another. The antisera to tracer amounts of AVP. highest recovery measured in this study using syn· The reaction between AVP and the antibodies was thetic AVP was 75% on the average 60 ± 15%. Flori· also modified by increasing amounts of aserum si! batches with recoveries below 45% were discarded. albumin/globulin-mixture (prepared freshly every two It should be noted that the use of 125 1 labeled AVP to three months from normal rabbit serum and addin recovery studies is of very limited value since ed in amounts of 0 to 0.4 J.tl/ml incubation fluid). The maximum of binding was achieved after addition of 0.1 Ill/ml as outlined in Figure 6. Results of experiments designed to test the optimal and most practicable conditions in respect to time and temperature of the incubation with the first and second antibody are depicted in Figure 7.
ANTI· SERUM AI/I,
• 11/2 "111/4 o IV/I, o V /1
Extraction method For assaying AVP in plasma sampIes, an extraction procedure is an essential prerequisite in alm ost all assay methods described up until now with only few exceptions (Oyama et al. 1971; Wagner, Maier, He"mann and Franz 1975). The extraction (acetone,
75
-VI/I,
FINAL DILUTION 1 1 1. 1 1 1
10000 50000 30000 50000 60000 20000
50
I 1
I I
25
II
o ••••••• I
o
\ ....,..'
~/
I
I
I
10
20
30
o
--_.
\.
POcbufler + EDTA 0001M
+
+
+
(-)
(- )
0.02 M
CaCI2
(-)
1O- 3 M
(- )
(-)
I
I
40
50
fractlons a'3 ml
Fig. 4 Separation of arginine-vasopressin (AVP) and AVP on Sephadex G-25 fine.
125 1_
Fig. 5 Effect of varying concentrations of EDTA and Caions (given bclow) on the binding capacity of several anti· arginine vasopressins. Ordinate: percentage of constant amounts of I2S I·AVP bound.
Downloaded by: Universite de Sherbrooke. Copyrighted material.
Characterization of antibody antigen reaction
505
Radioimmunoassay of Arginine Vasopressin in Human Plasma ANTISERUM
.11/2
ANTIFINAL SERUM DILUTION
FINAL DILUTION
.11/2
'20000 .20000 . 20000 . 20000 20000 20000
• V/2
aIV/2 oll/I c VII "IV/I
I 20000 150000 I 100000 I; 200000
011/2 .11/2
.11/2
75
50
Fig. 6 Effect of increasing amounts of protein on the binding of AVP to various anti-argininevasopressins (6a) and to anti-arginine vasopressin 11/2 in various final dilutions (6b). Ordinate: percentage of constant amounts of 1251_AVP bound.
I
oLr-----r-I--,,----r,- , I o .1 .2 .3 .4(~I/ml)
these "recoveries" had more arte facts than those from unlabeled AVP and, therefore, were not comparable. Standard CU11Jes A ty~ical stand.ard ~urve obtained with ~y~thetic A ~ IS shown m !Igure 8a; the lowest 1~lt of de-
tectlon was 1.10 - 0.5 pg/ml. The 50% mtercept was 9.5 ± 0.4 pg/ml in 7 standard curves (not shown). Immunoidentity of synthetic A VP given as a bolus of I mU/kg body weight i.v. and of endogenous A VP is documented by the parallelism of diIuted sampies with the standard curve (Fig. 8b). The interassay variation coefficient of identical sampies (n = 6) containing 5, 10, 25 pg/ml plasma, each measured at 6 different occasions, was found to be 12%. When plasma was made vasopressin-free by storing it for 24 hours at room temperature (n = 12), virtually no vasopressin could be detected. Application In addition to experimental data already published (Uhlich, Weber, Eigler and Gröschel-Stewart 1975), we would like to present results (with respect to the applicability of the assay) which are part of a study concerning the secretion pattern of AVP (Uhlich, Weber, and Haslbeck 1975). Plasma levels measured in normal adults ( n = 28) of both sexes (9 a.m., recumbent position, free access to salt, water and food) were 3.4 ± 2.2. pg/ml plasma. The mean values of all AVP plasma levels taken from 6 healthy volunteers at 60 minutes intervals over a 24 hour period ( n = 144) was 4.5 ± 1.0 pg/ml. A marked difference in the mean levels of AVP
o'-,--"",r--"",---. .1
.2
throughout the entire day was found between the individuals tested in this study (independent of the actual plasma osmolality measured at the same time). The lowest mean value in one individual was 2.2 ± 1.6 pg/ml, the highest in another was 8.6 ±4.1 pg/ml. In alI subjects investigated, peaks of plasma A VP 2~ times the basal level were measured every 3-4 hours. Changes in plasma osmolality were not '/, BINDING OF 1251-AVP
50
~...-.-
40
.6~---
30
,!F."-"~ /'
[ANTI-VASOPRESSIN-ABI
:~.
20 "11
r~--~--O~'------
10
e,oe
_ _ _ _---22"C ~A
~
1
___
~37"C
I
I
o
I I,/,
I
50 40
~==e/~4"C ~22·C
f
30 20
//
@li:nLO-'""N'A.~
10
o
I
T., ,,,
01020 JO' 1 MIN.
I
I
,
B
-L
1
12
16
HRS
1
I
20 2'
I
I
2
3
.
...L..----J
4
5
DAYS
Fig. 7 Dependence of primary (abo~e) and secondary (below) antibody reaction on duration and temperature of incubation.
Downloaded by: Universite de Sherbrooke. Copyrighted material.
25
25
E. Uhlich, P. Weber, U. Gröschel-Stewart and T. Röschlau B ['1.1 100
75
50
25
\
\
.,
•
\
\
\
6
\
10
o
OLl----'----.JL-----',10':------',~0 AVP pg/ml
o oL..'- - - - - - ' - - - - - - , " ' " ' 0 - - - - - ' " " 0 0 pg/ml Fig. 8 a) Standard curve of arginine vasopressin. b) Immunoidentity of exogenous (open symbols) and endogenous (block symbols) AVP, as indicated by parallelism of standard curves.
detectable. This periodicity also persisted during the night but at a higher level.
reproducible procedure to extract AVP from plasma sampies.
Mean values of hourly measurements (n = 36) were 3.4 ± 1.1 pg/ml for the 2-7 a.m. period; 3.9 ± 0.5 pg/ml for the 8 a.m. - 1 p.m. period; 4.4 ± 0.9 pg/ml for the 2-7 p.m. period and 6.2 ± 1.6 pg/ml for the 8 p.m. - 1 a.m. period, respectively.
The antibody described here may be used in a dilution of 1:50.000, indicative of a high titer. Even more important, in contrast to most of the immunoassays published thus far, no interfering crossreactivity with lysine vasopressin has been observed. Crossreactivity with oxytocin, vasotocin or other analogues tested was negligible. Therefore, the antibody used must be highly specific (see Fig. 2).
In an additional series, a water diuresis of at least 15 mi/rn in was induced and maintained over a 4 hour per iod in 6 healthy male adult volunteers. The prediuretic value of AVP was 3.7 ± 3.1 pg/ml and it dropped to a mean of 2.4 ± 1.3 pg/ml, 1.6 ± l.l pg/ml and 1.4 ± 0.8 pg/ml at 2, 3 and 4 hours after beginning the water diuresis, respectively. The results correspond weil with previous observations by Robertson et al. (I 973) as weil as Husain, Fernando, Shapiro, Kagan and GUck (1973). Discussion In developing a radioimmunoassay fo AVP, two major difficulties are encountered: 1. The production of both a highly sensitive and specific antibody, 2. the development of a sufficiently selective and
In contrast to the method reported by Edwards et al. in 1972, a satisfactory label could not be produced by desalting the iodination mixture with ion exchange resin. It was, however, obtained after filtration on Sephadex G 25 fine. In order to acquire optimal standard curves, different protein concentrations and varying calcium concentrat ions have been tested in the incubation medium. The results of these studies are shown in Figures 5 and 6. They are comparable to data published by Hussain et al. (1973) and Robertson et al. (1973). In addition, aseries of experiments was performed where equilibration time and temperature were varied
Downloaded by: Universite de Sherbrooke. Copyrighted material.
506
Radioimmunoassay of Arginine Vasopressin in Human Plasma References
Beardwell, e. G.: Radioimmunoassay of arginine vasopressin in human plasma. J. ehn. Endocr. 33: 254-260 (1971) Edwards, e.R. W.. T. Chard, M.J. Kitau, M.L. Forsling, J. Landon: The development of a radioimmunoassay for In regard to the extraction of AVP from plasma, the arginine vasopressin: productlon of antisera and labelIed hormone; separation techniques; specifity and sensitivity method described by Beardwell - after minor modiof the assay in aqueous solution. J. Endocr. 52: 279fications - seemed to fulfill all criteria required. 288 (1972) Florisil will easily and rapidly remove AVP from Greenwood, F.e., w'M. Hunter, J.s. Glover: Thc preparaplasma. Most of the batches we tested removed ation of 131 I-labelled human growth hormone of high bout 60% of the exogenous AVP from plasma and specific radioactivity. Biochem. J. 8q: 114-123 (1963) some even more. The concentration of AVP rcquired Husain, M.K., N. Fernando, M. Shapzro, A. J\.agan, S.M. lilick: Radioimmunoassay of arginine vasopressin in human for the assay procedure was easily achieved by Iyoplasma. J. Clin. Endocr. 37: 616-623 (1973) philization of the eluted hormone. Marschner, I., H. Dobry, F. Erhardt, T. Landersdor[er, B. Popp, e. Ringel, P. Scriba: Berechnung radioimmunologiIn both bioassay and radioimmunoassay, measurescher Meßwerte mittels Spline-Funktionen. Ärzt\. Lab. ments of the recovery rates of chemically pure AVP 20: 184-191 (1974) added to plasma sam pIes or phosphate buffer yielded Oyama, S.N., A, Kagan, S.M. Glick: Radioimmunoassay of identical results. It can reasonably be assumed that, vasopressin: Application to unextracted human urine. J. Clin. Endocr. 33: 739-744 (1971) due to the Florisil extraction step, only biologically active material is determined (Uhlich,personal obser- Robertson, G.L., E.A. Mahr: Inappropriate antidiuresis without inappropriate vasopressin secretion. 4th Int. Cong. vation 1973). Endocr., Washington (1972) Since the volume of plasma required for one determi- Robertson, G.L., E.A. Mahr, S. Athar, T. Sinha: Development and c1inical application of a new method for the nation is small (3 to 4 ml) the assay mayaIso be radioimmunoassay of arginine vasopressin in human applied to various laboratory animals. plasma. J. Clin. Invest. 52: 2340-2351 (1973) Robertson, G.L., J. Roth, C. Beardwell, L.A. Klein, M.J. We feel that the necessity of a detailed and exact Petersen, P. Gorden: Radioimmunoassay of vasopressin description of AVP radioimmunoassay is of utmost in man. In: Methods in investigative and diagnostic endoimportance because general agreement has not been crinology. Berson, S.A. and R. Yalow, editors. Amsterdam North-Holland Publishing Co., II A: 656-668 (1973) reached even at the so called "basal levels" of AVP. Schwartz, W.B., W, Bennet, S. Curelop, F.e. Bartter: SynThe figures given vary between 4 pg/ml (Robertson drome of renal sodium loss and hyponatremia probably et al. 1973, Oyama et al. 1971) and 50 pg/ml resulting from inappropriate secretion of antidiuretic (Wagner et al. 1975), i.e., by a factor of 10. hormone. Amer. J. Med. 23: 529-542 (1957) Uhlich, E.,P., Weber, J. Eigler, U. Gräschel·Stewart: AngioTo test the applicability (including sensitivity, specitensin stimulated AVP-release in humans. Klin. Wschr. ficity and reproducibility) of the method we refer 53: 177-180 (1975) to already published data (Uhlich et al. 1975) and Uhlich, E., P. Weber, R. Haslbeck: in preparation to the results represented. The latter are part of an ex- Wagner, H., V. Maier, H..J. Herrmann, H.E. Franz: Direct measurement of arginine-vasopressin in human serum tensive study on secretion pattern of AVP in vitro, in without extraction proccdure. (Abstr.) Acta Endocr. vivo, under control conditions and after stimulation (Kbh.) 193 (1975) (Uhlich et al. 1975). Wu, w'.H., J.H. Rockey: Antivasopressin antibody. Characterization of high-affinity rabbit antibody with limited asAcknow ledgem ents sociation constant heterogeneity. Biochemistry 8: 27 192728 (1969) We would like to thank Professor Dr. H. Schwalm for his encouragement and advice, Drs. J. Eigler, K. v. Werder, F. Erhardt and J. Marschner (Munich), Drs. J. Landon and M. Forsling (London), Dr. C. Beardwell (Manchester) and Dr. G. Robertson (\ndianapoh~) for many helpful ~uggestions and R. Herbst and H. Ulbrich for the skillful technical assistance. Preliminary studies of this project were planned in 1972/73 with a British Council Scholarship. Requests for reprints should be addressed to: Dr. E. Uhlich, 11. Medizinische Universitätsklinik, 0-8000 München (Germany)
Downloaded by: Universite de Sherbrooke. Copyrighted material.
in order to achieve maximal sensitivity in the radioimmunoassay procedure. Having thus established optimal assay conditions, it was possible to detect as little as I pg of AVP per ml plasma.
507
501
Furthermore its response resembles that obtained after administering 1 mg of tetracosactide which indicates that it possesses a potency some six times larger. This result corresponds with the findings obtained by other authors with different methods (Maier et al. 1971, Keenan et al. 1971). References
Appleby, J.I., G. Gibson, I.K. Norymberski, R.D. Stubbs: "The determination of 17-hydlOxycorticosteroids". Biochem. J. 60: 453460 (955) Bagazgoitill, F.J., A. Oriol Bosch: "Automatization por flujo continuo de la reaccion de Zimmermann para la determinacion de los 17 -cetosteroides y de los esteroides 17-cetogenicos totales urinarios". Archivos Facultad Medicina Madrid 26: 203-216 (1974) Bell, P.H.: "Purification and structure of ß-corticotropin" J. Amer. Chem. Soc. 76: 5565-5567 (1954)
tion of free II-hydroxycorticosteroids in human plasma". J. Clin. Path. 15: 374-379 (1962) Ramachandran, J., D. Chung, Ch. H. Li: "Adrenocorticotropins XXXIV. Aspects of structure-activity relationships of the ACTH molecule. Synthesis of a heptadecapeptide amide, an octadecapeptide amide and a nonadeca· peptide amide possessing high biological activities". J. Amer. Chem. Soc. 81: 2696-2708 (1965) Retiene, K., F. Schultz: "Human pharmacological studies with a new synthetic corticotrophic peptide consisting of 18 aminoacids". Acta Endocrinol. Suppl. 173: 33 (1973) Rittei, W.: "Techniques for the synthesis of ACTH and MSH peptides and analogues". "Pharmacology of hormonal polypeptides and proteins". N. Bach, L. Martini and R. Paoletti eds. Plenum Press, New York 1968,3547 Schwyzer, R.: "Programierte Molekein". Experientia 26: 577-587 (1970) Seelig, S., G. Sayers: "Structural activity relationship among ACTH". Fedn. PlOc. 30: 693 (1971)
Requests for reprints should be addressed to: Dr. A. OrioJ-Bosch, Dept. Exper. EndocrinoL. University Complutensis Medica1 School, Madrid (Spain).
Hormon. Metab. Res. 7 (1975) 501-507
© Georg Thieme Verlag Stuttgart Radioimmunoassay of Arginine Vasopressin in Human Plasma* E. Uhlich**, P. Weber, U. Gröschel-Stewart and T. Röschlau 11. Medizinische Klinik der Universität München, and Frauenklinik der Universität Würzburg, Germany
Summary Antibodies for the radioimmunoassay of arginine vasopressin (A VP) described here were produced in rabbits using synthetic AVP coupled to rabbit 'Y-globulin with carbodiimide. In three out of six rabbits, significant antibody titres were obtained. Using the best antisera produced' 40% of labeled AVP was bound at a final dilution of 1 :50.000. After iodination of synthetic AVP with 125 1 using the chloramin-T method, a gel filtration on Sephadex G-25 was performed •
A preliminary report of part of this work was published in Acta end ocr. (Kbh.) Suppl. 184 (1974) 52. .. Supported by the Deutsche Forschungsgemeinschaft.
Received: 22 Nov. 1974
Accepted: 30 Aug. 1975
to punfy the iodinated AVP. For separation of antibody bound and free hormone, a second antibody precipitation was used. There was no crossreactivity with oxytocin. AVP was extracted from plasma after ammoniumsulfate precipitation of the proteins by adsorption to Florisil. The recovery of AVP added to plasma in amounts between 5-25 pg/ml was 60 ± 15% (n=6). The minimum amount of AVP detectable was 1 pg per ml plasma. The plasma level in normal adults under standard conditions was 3.4 ± 2.2 pg/ml. This is in agreement with data recently published by other researchers. The applicability and reproducibility was further tested in measurements of sampies taken hourly during the entire day under water diuresis and after hormonal stimulation of AVP. Key-Words: Vasopressin - Radioimmunoassay - Secretion Pattern.
Downloaded by: Universite de Sherbrooke. Copyrighted material.
The more prolonged increase of II-OHCS following KeeNm, J., J.B., Thompson, M.A. Chamberlain, G.M. Besser: "Prolonged corticotrophic action of a synthetic substiintravenous administration of C-41795-Ba is also tuted 1-18ACTH". Brit. Med. J. I: 742-743 (1971) reflected in the higher urinary steroid excretion. This Landon, I., V.H. T. James, M.J. Wharton, M. Friedman: could be interpreted as a consequence of the rate of "Threshold adrenocortical sensitivity in man and its poslocal inactivation of the polypeptides following subsible application to corticotrophin bioassay". The Lancet 11: 697-700 (1967) cutaneous or intrarnuscular administration being sigLowry, P.J., C. McMartin, J. Peters: "Properties of a simplinificant in relation to their half-life in either the fied bioassay for adrenocroticotrophic activity using the plasmatic or adrenal compartments. steroidogenic responsed of isolated adrenal cells". J. Endocr. 59: 43-55 (1973) Finally one should consider that 0.125 mg doses of Maier, R., P.L. Barthe, L. Schenkel·Hulliger, P.A. Desaulles: C-41795-Ba, even though capable of a marked adre "The biological activity of (1-D-serine, 17-18-Dilysine) adrenal stimulation, do not interfere with the ß-corticotrophin-( 1-18)-Qctadecapeptide-amide". Acta circadian rhythm since the effect disappeared after Endocrinol. 68: 458466 (1971) Mattingly, D.; "A simple fluorimetric method for the stima8 hours.
E. Vhlich, P. Weber, V. Gröschel-Stewart and T. Röschlau
Introduction The bioassays of AVP that have been used up until now for studies of hormonal fluid regulation are not sensitive enough to detect minimal, although biologically significant, changes of this hormone. Therefore, in order to study mechanisms regulating the release of AVP, methods had to be designed which would permit the detection of slight fluctuations of the normally very low endogenous levels of AVP. This is also pertinent to various disease states in which the role of AVP should be characterized not only qualitatively but also quantitatively. The concept, for example, that AVP is excessively elevated in situations described as the syndrome of inappropriate ADH secretion (Schwartz, Bennet, Curelop and Bartter 1957) is by no means true in all instances. Recently, radioimmunological measurements of plasma AVP in such patients (Robertson and Mahr 1972) showed normal or even low plasma levels of AVP in some, thereby, indicating that a "syndrome of inappropriate secretion of AVP without inappropriately raised plasma levels of the hormone" does exist. Even though the presence of a hormone with a different radioimmunological characteristic is not excluded. Since some of the difficulties of such radioimmunologically AVP determinations are the production of specific antibodies and the extraction of AVP from plasma sampies, we wish to present a detailed description of our efforts in this field which led to an improvement of the assay. In order to test the applicability and reproducibility of the assay in clinical situations, plasma AVP was measured every 60 minutes in 6 normal human volunteers for aperiod of 24 hours. The nocturnal values were found to be higher (6.2 ± 1.6 pg/ml) when compared with mean plasma levels during day time (3.9 ± 0.8 pg/ml). An episodic rhythmicity of plasma AVP levels was evident in all individuals tested with peak levels (2-6 times the basal level) every 3-4 hours.
ear vein and tested for AVP antibody titer: Terminal coUecHon of the serum was performed by catheterization of the caro t id art ery .
Radioiodination of Vasopressin Synthetic arginine vasopressin was iodinated using the chloramin-T method described by Greenwood. Hunter and Glover (1963): 7.5 U AVP in 20/.11 phosphate buffer were exposed for 10 seconds to 10/.11 Na/2 5 1 solution (specific activity about 100 mCi/ml) after addition of 10 /.11 chloramin-T solution (2.5 mg/mi). The reaction was stopped using 200 /.11 bovine serum albumin (pentex ) 4.2 mg/mI. The iodinated AVP solution was then applied to a 1.5 x 40 cm column of Sephadex G 25 (fine) and equilibrated with 0.25% acetic acid. Three radioactive peaks were eluted with the same solvent. The largest one (about 90% of the total radioactivity) contained the labeled hormone. One or two fractions (2 ml each) of the labeled AVP peak were diluted to a final radioactivity of about 1200 cpm per 50 /.11 and then added to the sampies used in the assay.
Assay Standard curves were obtained after mixing 200/.11 of diluent buffer (0.015 M NaH 2 P04 • 0.15 M NaCl, 0.02% thiomerosal, 2 mg/mi bovine serum albumin (Pentex). 0.001 M EDTA adjusted to pH 7.4), 250 /.11 of a freshly prepared AVP solution in varying concentrations and 50 /.11 of the antiserum which had been diluted with the above mentioned phosphate buffer to a concentration of 1 :5.500 (final concentration 1:50.000). For the measurement of plasma sam pies. the dried extracts (see below) were redissolved in diluent buffer and. after 500/.11 of diluent buffer had been added. 200 /.11 were used in the assay. After 24 hours preincubation of unlabeled standards or unknown sampies at 4°e. 50 /.11 of 125 1 AVP (1.200 cpm) were added to all of them, mixed and then reincubated for an additional 24 hours at 4 0 C, In order to separate the bound and free hormone, a second antibody (antirabbit ')'globuline from donkey, Wellcome) was added to the above incubates and after thorough mixing, left at 4°C far another 24 hours. It was then centrifuged at 4 0 C for 10 minutes at 2000 g. The supernatant containing the unbound hormone was discarded. The radioactivity of the precipitated antibodybound hormone was counted directly in an automatic gamma counter (Tracer Lab), connected to an off-line printer (Teletype).
To test the accuracy of the whole assay procedure vasopressinfree plasma sam pies were assayed with and without arginine vasopressin in three different concentrations (5, 10, 25 pg/ml) on six different occasions. To analyze the inter· Methods assay variation coefficient, a pooled plasma sampie with Preparation of antigens and A VP antisera known vasopressin content was taken as an "internal standard" The antigen complex was prepared in a slightly modified form in all assays perfarmed. as described by Wu and Rockey in 1969: 2 ml of a solution Extraction of Plasma Sampies containing 300 IV = 0.75 mg of synthetic arginine vasopresVenous blood sampies (at least 15 ml for duplicate assay) sin (Ferring, Malmö/Sweden), 40 mg rabbit ')'-globulin from both recumbent controls and patients were collectcd (Serva, Heidelberg) and 30 mg l-ethyl-3 (3-dimethyl-aminointo cooled (4°C) heparinized plastic tubes and immediately propyl)-carbodiimide-HCI (Sigma, SI. Louis, Mo.) were adcentrifuged at 4 oe. Plasma sampies were stored in plastic justed to a pH in the range of 7.0-7.5,using brom thymol tubes at -20°C and no detectable 10ss of AVP activity could blue as an indicator and the solution was kept at room be found over aperiod of several wecks. temperature for 16 to 18 hours. 2 ml Freund's complete adjuvant (Difco) and 1-3 mg mycobacterium tuberculosis After acidification to pH 5.8~.2 with 1 N HCI, equal were then added and the homogenized mixture injected amounts of saturated ammonium sulfate solution were added both subcutaneously into the fossa poplitea and intradermal- to 3 or 4 ml plasma aliquots. After mixing. centrifugation Iy between the shoulder blades. Each rabbit was givel} an at 3.000 rpm for 5 minutes at 4 0 C and decanting. the dear aliquot of the mixture containing 50 IV of the arginine supernatants were thoroughly mixed for 15 minutes with vasopressin. The animals were boostered at three week inter- 100 mg Florisil (100-200 mesh. Serva. Heidelberg) per vais with the above mixture alternating the intramuscular, sampie. The resin was washed 10 times with distilled water subcutaneous and intraperitoneal route. 8-10 days after and dried at about 40 0 C before use. The supernatant was each injection, a blood sampie was taken from the lateral discarded and the Florisil portion was washed twice with
Downloaded by: Universite de Sherbrooke. Copyrighted material.
502
Radioimmunoassay of Arginine Vasopressin in Human Plasma AB. TITRE AT
"I. BINDING OF LABELLED AVP
40·,. BINDING
so
, :70 000
........ •••••••
"
.....
......
. :
1:30000
:~:---------
.
I
3
4
..---
,,'
, .. ..
...7··,,••
IV
••••• 11
v I
5 6 MONTH CF IMMUNIZATION
Fig. I Response of rabbits (animal 11, IV, V) to immunization with 50 IV AVP coupled to rabbit ')"-globulin, given at 21 day intervals. . 1 ml each of 1 N HCI and double distilled water. The adsorbed AVP was finally eluted from the adsorbent by vigorous mixing for 15 minutes with 1 ml aceton water (80:20) v/v). After centrifugation and separation, the supernatant was washed with 1 ml of diethylether. The aqueous phase was Iyophilized and, without further storage, assayed after dissolving in 500 1'1 of diluent buffer, in duplicates of 200 1'1 each.
Calculations Calculations of radioimmunoassay standard curves of A VP were performed with the spline approximation method. This has been shown to be a better procedure for the calculation of radioimmunoassay standard curves than those previously used (Marschner, Dobry, Erhardt, Landersdorfer, Popp, Ringel and Scriba 1974). In this method, trial functions between the measure points of radioimmunologically obtained standard curves are calculated keeping the second derivative of the whole spline funchon continuous and the oscillation of this curve, depending upon the smoothing parameter, at aminimum. The advantages of this method compared with conventional approximations (e.g., logit transformation) have been discussed elsewhere (Marschner et al. 1974). A convenient Fortran IV program has been developed for a Siemens 404/3 computer.
Fig. 2 Crossreaction of various vasopressin analogues and oxytocin with anti-arginine-vasopressin (IV /4). PLV 2 = phe-Iys-vasopressin; POR 6 = ornithin vasopressin; Ls = Iys-vasopressin; DDAVP = desamin-(d-args ) vasopressindiacetat.
Crossreactivity with various AVP analogues has been tested with antibody IV /4. In contrast to most data reported in the literature (Edwards, Chard, Kitau, Forsling, and Landon 1972), the immunoreactivity with AVP was highly specific. In Figure 2, binding curves of AVP antiserum with different analogues (arginine vasopressin, arginine vasotocin, phenylalaninelysine vasopressin, ornithine vasopressin, desaminovasopressin and oxytocin) are shown. There was no crossreactivity with oxytocin, the binding - as compared to AVP - was less than I: 10.000.
Labeling The labeling of the hormone with 125 I using the chloramin-T method was time dependent (Fig. 3); a 10 second incubation time was found to be sufficient for the introduction of 90% of the iodine into the AVP molecules. A typical elution profile of radioactivity is depicted in Figure 4.
Peak 1 (fractions 7-10) contains 125 Ibound to protein measured at 280 nm (Zeiss MK2), peak 2 (fraction 15-20) represents free iodine (measured titriResults metrically). 90% of the total amount of the radioAntibodies activity was covered by peak 3 (fractions 20-40) representing iodinated AVP loosely adsorbed to the gel. Sufficiently high antibody titers were found in 3 This elution pattern is exactly the same as has been out of 6 rabbits; 2 developed diabetes insipidus shown by Oyama, Kagan and GUck (1971) and during the period of immunization. The titers measured in 3 rabbits and expressed as 40% bind- Robertson, Roth, Beardwell, Klein, Petersen and ing of a constant amount of labeled hormone are Gorden (1973). Biologically active, noniodinated AVP shown in Figure I. All results described here were was found between peak 2 and peak 3 (measured in rat bioassay). The two fractions following the top of obtained with the antiserum IV /4 in a final dilution of 1 :50.000 except where other specifications peak 3 were used as tracer in the assay. They were stored in diluted form at -20o C and gave reproducible are noted. results for up to about 8 weeks.
Downloaded by: Universite de Sherbrooke. Copyrighted material.
,: 50000
1:10000
503
E. Uhlich, P. Weber, U. Gröschel-Stewart and T. Röschlau
504
PROTEIN·PEAK ~.
100
90
Fig. 3 Time dependent effect of chloramin-T on the iodination of AVP with tracer amounts of 125 1, separation: Sephadex column K 9/15; G-25 fine; ordinate: percentage of total amount of 125 I.
ethylalcohol, etc.) is usually followed by an adsorption step. Florisil was used and found to give adequate The effect of calcium ions (0 to 0.001 M) or EDT A and reproducible results. However, we found it neces(0 to 0.02 M) on binding ability of different antisary to test each batch of Florisil to get an "individu· vasopressins is shown in Figure 5. Higher concentraal" adsorptive characterization since the recovery of tions of either EDT A or Ca inhibited the binding of added AVP varied from one batch to another. The antisera to tracer amounts of AVP. highest recovery measured in this study using syn· The reaction between AVP and the antibodies was thetic AVP was 75% on the average 60 ± 15%. Flori· also modified by increasing amounts of aserum si! batches with recoveries below 45% were discarded. albumin/globulin-mixture (prepared freshly every two It should be noted that the use of 125 1 labeled AVP to three months from normal rabbit serum and addin recovery studies is of very limited value since ed in amounts of 0 to 0.4 J.tl/ml incubation fluid). The maximum of binding was achieved after addition of 0.1 Ill/ml as outlined in Figure 6. Results of experiments designed to test the optimal and most practicable conditions in respect to time and temperature of the incubation with the first and second antibody are depicted in Figure 7.
ANTI· SERUM AI/I,
• 11/2 "111/4 o IV/I, o V /1
Extraction method For assaying AVP in plasma sampIes, an extraction procedure is an essential prerequisite in alm ost all assay methods described up until now with only few exceptions (Oyama et al. 1971; Wagner, Maier, He"mann and Franz 1975). The extraction (acetone,
75
-VI/I,
FINAL DILUTION 1 1 1. 1 1 1
10000 50000 30000 50000 60000 20000
50
I 1
I I
25
II
o ••••••• I
o
\ ....,..'
~/
I
I
I
10
20
30
o
--_.
\.
POcbufler + EDTA 0001M
+
+
+
(-)
(- )
0.02 M
CaCI2
(-)
1O- 3 M
(- )
(-)
I
I
40
50
fractlons a'3 ml
Fig. 4 Separation of arginine-vasopressin (AVP) and AVP on Sephadex G-25 fine.
125 1_
Fig. 5 Effect of varying concentrations of EDTA and Caions (given bclow) on the binding capacity of several anti· arginine vasopressins. Ordinate: percentage of constant amounts of I2S I·AVP bound.
Downloaded by: Universite de Sherbrooke. Copyrighted material.
Characterization of antibody antigen reaction
505
Radioimmunoassay of Arginine Vasopressin in Human Plasma ANTISERUM
.11/2
ANTIFINAL SERUM DILUTION
FINAL DILUTION
.11/2
'20000 .20000 . 20000 . 20000 20000 20000
• V/2
aIV/2 oll/I c VII "IV/I
I 20000 150000 I 100000 I; 200000
011/2 .11/2
.11/2
75
50
Fig. 6 Effect of increasing amounts of protein on the binding of AVP to various anti-argininevasopressins (6a) and to anti-arginine vasopressin 11/2 in various final dilutions (6b). Ordinate: percentage of constant amounts of 1251_AVP bound.
I
oLr-----r-I--,,----r,- , I o .1 .2 .3 .4(~I/ml)
these "recoveries" had more arte facts than those from unlabeled AVP and, therefore, were not comparable. Standard CU11Jes A ty~ical stand.ard ~urve obtained with ~y~thetic A ~ IS shown m !Igure 8a; the lowest 1~lt of de-
tectlon was 1.10 - 0.5 pg/ml. The 50% mtercept was 9.5 ± 0.4 pg/ml in 7 standard curves (not shown). Immunoidentity of synthetic A VP given as a bolus of I mU/kg body weight i.v. and of endogenous A VP is documented by the parallelism of diIuted sampies with the standard curve (Fig. 8b). The interassay variation coefficient of identical sampies (n = 6) containing 5, 10, 25 pg/ml plasma, each measured at 6 different occasions, was found to be 12%. When plasma was made vasopressin-free by storing it for 24 hours at room temperature (n = 12), virtually no vasopressin could be detected. Application In addition to experimental data already published (Uhlich, Weber, Eigler and Gröschel-Stewart 1975), we would like to present results (with respect to the applicability of the assay) which are part of a study concerning the secretion pattern of AVP (Uhlich, Weber, and Haslbeck 1975). Plasma levels measured in normal adults ( n = 28) of both sexes (9 a.m., recumbent position, free access to salt, water and food) were 3.4 ± 2.2. pg/ml plasma. The mean values of all AVP plasma levels taken from 6 healthy volunteers at 60 minutes intervals over a 24 hour period ( n = 144) was 4.5 ± 1.0 pg/ml. A marked difference in the mean levels of AVP
o'-,--"",r--"",---. .1
.2
throughout the entire day was found between the individuals tested in this study (independent of the actual plasma osmolality measured at the same time). The lowest mean value in one individual was 2.2 ± 1.6 pg/ml, the highest in another was 8.6 ±4.1 pg/ml. In alI subjects investigated, peaks of plasma A VP 2~ times the basal level were measured every 3-4 hours. Changes in plasma osmolality were not '/, BINDING OF 1251-AVP
50
~...-.-
40
.6~---
30
,!F."-"~ /'
[ANTI-VASOPRESSIN-ABI
:~.
20 "11
r~--~--O~'------
10
e,oe
_ _ _ _---22"C ~A
~
1
___
~37"C
I
I
o
I I,/,
I
50 40
~==e/~4"C ~22·C
f
30 20
//
@li:nLO-'""N'A.~
10
o
I
T., ,,,
01020 JO' 1 MIN.
I
I
,
B
-L
1
12
16
HRS
1
I
20 2'
I
I
2
3
.
...L..----J
4
5
DAYS
Fig. 7 Dependence of primary (abo~e) and secondary (below) antibody reaction on duration and temperature of incubation.
Downloaded by: Universite de Sherbrooke. Copyrighted material.
25
25
E. Uhlich, P. Weber, U. Gröschel-Stewart and T. Röschlau B ['1.1 100
75
50
25
\
\
.,
•
\
\
\
6
\
10
o
OLl----'----.JL-----',10':------',~0 AVP pg/ml
o oL..'- - - - - - ' - - - - - - , " ' " ' 0 - - - - - ' " " 0 0 pg/ml Fig. 8 a) Standard curve of arginine vasopressin. b) Immunoidentity of exogenous (open symbols) and endogenous (block symbols) AVP, as indicated by parallelism of standard curves.
detectable. This periodicity also persisted during the night but at a higher level.
reproducible procedure to extract AVP from plasma sampies.
Mean values of hourly measurements (n = 36) were 3.4 ± 1.1 pg/ml for the 2-7 a.m. period; 3.9 ± 0.5 pg/ml for the 8 a.m. - 1 p.m. period; 4.4 ± 0.9 pg/ml for the 2-7 p.m. period and 6.2 ± 1.6 pg/ml for the 8 p.m. - 1 a.m. period, respectively.
The antibody described here may be used in a dilution of 1:50.000, indicative of a high titer. Even more important, in contrast to most of the immunoassays published thus far, no interfering crossreactivity with lysine vasopressin has been observed. Crossreactivity with oxytocin, vasotocin or other analogues tested was negligible. Therefore, the antibody used must be highly specific (see Fig. 2).
In an additional series, a water diuresis of at least 15 mi/rn in was induced and maintained over a 4 hour per iod in 6 healthy male adult volunteers. The prediuretic value of AVP was 3.7 ± 3.1 pg/ml and it dropped to a mean of 2.4 ± 1.3 pg/ml, 1.6 ± l.l pg/ml and 1.4 ± 0.8 pg/ml at 2, 3 and 4 hours after beginning the water diuresis, respectively. The results correspond weil with previous observations by Robertson et al. (I 973) as weil as Husain, Fernando, Shapiro, Kagan and GUck (1973). Discussion In developing a radioimmunoassay fo AVP, two major difficulties are encountered: 1. The production of both a highly sensitive and specific antibody, 2. the development of a sufficiently selective and
In contrast to the method reported by Edwards et al. in 1972, a satisfactory label could not be produced by desalting the iodination mixture with ion exchange resin. It was, however, obtained after filtration on Sephadex G 25 fine. In order to acquire optimal standard curves, different protein concentrations and varying calcium concentrat ions have been tested in the incubation medium. The results of these studies are shown in Figures 5 and 6. They are comparable to data published by Hussain et al. (1973) and Robertson et al. (1973). In addition, aseries of experiments was performed where equilibration time and temperature were varied
Downloaded by: Universite de Sherbrooke. Copyrighted material.
506
Radioimmunoassay of Arginine Vasopressin in Human Plasma References
Beardwell, e. G.: Radioimmunoassay of arginine vasopressin in human plasma. J. ehn. Endocr. 33: 254-260 (1971) Edwards, e.R. W.. T. Chard, M.J. Kitau, M.L. Forsling, J. Landon: The development of a radioimmunoassay for In regard to the extraction of AVP from plasma, the arginine vasopressin: productlon of antisera and labelIed hormone; separation techniques; specifity and sensitivity method described by Beardwell - after minor modiof the assay in aqueous solution. J. Endocr. 52: 279fications - seemed to fulfill all criteria required. 288 (1972) Florisil will easily and rapidly remove AVP from Greenwood, F.e., w'M. Hunter, J.s. Glover: Thc preparaplasma. Most of the batches we tested removed ation of 131 I-labelled human growth hormone of high bout 60% of the exogenous AVP from plasma and specific radioactivity. Biochem. J. 8q: 114-123 (1963) some even more. The concentration of AVP rcquired Husain, M.K., N. Fernando, M. Shapzro, A. J\.agan, S.M. lilick: Radioimmunoassay of arginine vasopressin in human for the assay procedure was easily achieved by Iyoplasma. J. Clin. Endocr. 37: 616-623 (1973) philization of the eluted hormone. Marschner, I., H. Dobry, F. Erhardt, T. Landersdor[er, B. Popp, e. Ringel, P. Scriba: Berechnung radioimmunologiIn both bioassay and radioimmunoassay, measurescher Meßwerte mittels Spline-Funktionen. Ärzt\. Lab. ments of the recovery rates of chemically pure AVP 20: 184-191 (1974) added to plasma sam pIes or phosphate buffer yielded Oyama, S.N., A, Kagan, S.M. Glick: Radioimmunoassay of identical results. It can reasonably be assumed that, vasopressin: Application to unextracted human urine. J. Clin. Endocr. 33: 739-744 (1971) due to the Florisil extraction step, only biologically active material is determined (Uhlich,personal obser- Robertson, G.L., E.A. Mahr: Inappropriate antidiuresis without inappropriate vasopressin secretion. 4th Int. Cong. vation 1973). Endocr., Washington (1972) Since the volume of plasma required for one determi- Robertson, G.L., E.A. Mahr, S. Athar, T. Sinha: Development and c1inical application of a new method for the nation is small (3 to 4 ml) the assay mayaIso be radioimmunoassay of arginine vasopressin in human applied to various laboratory animals. plasma. J. Clin. Invest. 52: 2340-2351 (1973) Robertson, G.L., J. Roth, C. Beardwell, L.A. Klein, M.J. We feel that the necessity of a detailed and exact Petersen, P. Gorden: Radioimmunoassay of vasopressin description of AVP radioimmunoassay is of utmost in man. In: Methods in investigative and diagnostic endoimportance because general agreement has not been crinology. Berson, S.A. and R. Yalow, editors. Amsterdam North-Holland Publishing Co., II A: 656-668 (1973) reached even at the so called "basal levels" of AVP. Schwartz, W.B., W, Bennet, S. Curelop, F.e. Bartter: SynThe figures given vary between 4 pg/ml (Robertson drome of renal sodium loss and hyponatremia probably et al. 1973, Oyama et al. 1971) and 50 pg/ml resulting from inappropriate secretion of antidiuretic (Wagner et al. 1975), i.e., by a factor of 10. hormone. Amer. J. Med. 23: 529-542 (1957) Uhlich, E.,P., Weber, J. Eigler, U. Gräschel·Stewart: AngioTo test the applicability (including sensitivity, specitensin stimulated AVP-release in humans. Klin. Wschr. ficity and reproducibility) of the method we refer 53: 177-180 (1975) to already published data (Uhlich et al. 1975) and Uhlich, E., P. Weber, R. Haslbeck: in preparation to the results represented. The latter are part of an ex- Wagner, H., V. Maier, H..J. Herrmann, H.E. Franz: Direct measurement of arginine-vasopressin in human serum tensive study on secretion pattern of AVP in vitro, in without extraction proccdure. (Abstr.) Acta Endocr. vivo, under control conditions and after stimulation (Kbh.) 193 (1975) (Uhlich et al. 1975). Wu, w'.H., J.H. Rockey: Antivasopressin antibody. Characterization of high-affinity rabbit antibody with limited asAcknow ledgem ents sociation constant heterogeneity. Biochemistry 8: 27 192728 (1969) We would like to thank Professor Dr. H. Schwalm for his encouragement and advice, Drs. J. Eigler, K. v. Werder, F. Erhardt and J. Marschner (Munich), Drs. J. Landon and M. Forsling (London), Dr. C. Beardwell (Manchester) and Dr. G. Robertson (\ndianapoh~) for many helpful ~uggestions and R. Herbst and H. Ulbrich for the skillful technical assistance. Preliminary studies of this project were planned in 1972/73 with a British Council Scholarship. Requests for reprints should be addressed to: Dr. E. Uhlich, 11. Medizinische Universitätsklinik, 0-8000 München (Germany)
Downloaded by: Universite de Sherbrooke. Copyrighted material.
in order to achieve maximal sensitivity in the radioimmunoassay procedure. Having thus established optimal assay conditions, it was possible to detect as little as I pg of AVP per ml plasma.
507
