301

ALDOSTERONE PLASMA RADIOIMMUNOASSAY INTERFERENCE BY A SPIROLACTONE METABOLITE W. Sad&e,(1) A. Michael8 Finn, P. Schmiedek, and A. BaethmaM Schools of Pharmacy, University of California, San Francisco and University of Southern California, Los Angeles, California, U.S.A. and Institut fur Chirurgische Forschung, Chirurgische Klinik der Universitaet MUnchen, Germany Received :

12/17/74 ABSTRACT

The plasma aldosterone radioimmunoassaydeveloped by Ito et al. was found to be non-specific for aldosterone following adminisGazon of the spirolactones, spironolactone and canrenoate-K,in rabbits, dogs and humans. The assay interfering principle was identified as a hydroxylated derivative ( ) of canrenone, which itself is a metabolite common to both spironolact "$ne and canrenoate-K. The metabolite MB possessed a high cross-reactivity to the 21-hemlsucclnatealdosterone antibody relative to other spirolactones. A modified procedure was developed specific for plasma aldosterone In the presence of MB. Following single doses of spironolactone and canrenoate-K, aldosterone plasma levels were unchanged in humans and in dogs and decreased in rabbits. INTRODUCTION The aldosterone antagonists spironolactone, canrenoate-K and their co-n

major metabolite canrenone (2) can affect endogenous aldosterone A reduction of aldosterone production by

kinetics in several ways.

competitive inhibition of steroidal C-11 and C-18 hydroxylation (3-S) was observed -in vitro

using rat adrenal sections at rather large

-4 and 6 x 1O"M) which is severconcentrations of canrenone (between 10 al times higher than therapeuticallyobtained plasma levels (6). Spironolactone treatment of whole rats reduced aldosterone plasma levels, possibly by induction of degradative hepatic enzymes -in vlvo (7). On the other hand pretreatment with spirolactones over several days stimulated -in vitro aldosterone production in rats (8) and resulted in anatomical changes in the adrenal zona glomerulosa in laboratory animals and men (8, 9, 10, 11, 12).

Volwne 25, Number 3

m

Plasma levels of aldosterone measured by

-x=D&OID-

March, 1975

S

302

TIIEOIDb

radloimmunoassayswere elevated in humans following spirolactone treatment over several days (13, 14). Displacement of aldosterone from its plasma protein binding sites was negligible at therapeutic splrolactone concentrations (15). Conflicting results may have been caused by the variety of assay techniques used to measure plasma aldosterone concentrations (15-27). We decided to investigate aldosterone kinetics during spirolactone treatment in several species utilizing the radiolnmunoassay of Ito, et al., (26) with an aldosterone Zl-hemissuccinyl BSA antibody. Interference of this aldosterone radioinrmunoassay by a yet unknown spirolactone metabolite and a modified specific procedure are reported here. EKPERIMRNTAL 1. Aldoaterone Plasma Radioimmunoassay The method of Ito et al., (26) was utilized, which consisted eesentially of the following steps: Plasma extraction Into C!R2C12,column chromatography on Sephadex LPI-20with CR2Cl2:CR3OH (98:2) as eluent, incubation with a aldosterone 21-hemisuccinyl BSA antibody (AntiAldos. Ewe Serum Batch 141-Serial No. aO196-09, Research Plus Laboratories, Inc., New Jersey), and measurement of the bound fraction of aldosterone. Recoveries from the Sephadex LH-20 column ranged within 30-60X, senaitivlty limit was about l-2 pgfincubation. 100 pg aldosterone standards were run through the procedures. The average amount of aldosterone found was 101 DQ with a standard deviation of 17% (n = 6). 2. Modified Aidosterone Plasma Radioimmunoassay in the Presence of Splrolactone Metabolites To 1 ml plasma, 0.2 ml of lN-NaOR was added to adjust the pH to approximately-13,and the samples were kept at room temperature for 10 min. This resulted in hydrolysis of the y-lactone ring of all steroidal spirolactones to yield y-hydroxycarboxylic acids, which are not extractable from aqueous medium into organic solvents above pR 6. After 10 min., the pH was readjusted to about pH 8 by adding 1 ml of 0.4 M pH 7.4 phosphate buffer to prevent decomposition of aldosterone at pH 13. The resulting solutions were then extracted with 15 ml CR2Cl2 and were carried through the same procedure as described by Ito et a1.(26). The recoveries of aldosterone from the Sephadex column rangzzthln 25-502. 100 pg aldosterone standards were run through the procedures. The average amount of aldosterone found was 94 pg with a standard deviation of 11% (n = 5). 3. Administration of Spironolactone and Canrenoate-K a. Ruman Subjects 400 mg spironolactone (AldactoneR)was given In tablets in a single dose and heparlnlzed plasma samples were obtained at 0, 3, 8, 24 and

48 hours follwing administration. Three patients received rapid intravenous infusions over 10 min. of 1 g canrenoate-K @oldactoneK)/day at 9 a.m. on three consecutive days as part of their medical treatment prior to neurosurgery and plasma samples were collected at 9, 12, 15 and 18* each day. 25-150 mg canrenoate-K/kgwere infused intravenously over l-5 minutes. Heparinized plasma samples were obtained 45 min. prior to administration until 100 minutes following administration in 15-30 minute intervala. Two dogs were adrenalectomized and corticoeteroidswere substituted by daily intramuscular injection of0.4 mg aldosterone and 10 mg prednisolone-21-hemisuccinatesodium for three days. A doee of 100 mgj kg canrenoate-K was given I.V. two days after termination of corticosteroid substitution and again plasma samples were collected as above. c. Rabbits 40 ag Spironolaetone/kgwere administered orally in aqueous suspension and 20 mg canrenoate-K/kg intravenously in aqueous solution. Heparinized plasma samples were obtained 24 hours prior to administration until 72 hours fo lwing administration at appropriate intervals. 4. Administration of & -Canrenoate-K to Rabbits and Isolation of Metabolites 20,21-3H-canrenoate-Kwith a specific activity of 860 nCi/mg was provided by G.D. Searle Company. 20 mg 3H-canrenoate-K/kgwith a 3Hactivity of 50 uCi were injected intravenously. The rabbits were sacrificed at 2 hours following administration, and total plasma was obFor isolation of larger amounts of tained for metabolfte analysis. spirolactone metabolites, total liver was homogenized in a 5 fold volume of methanol, centrifuged and the methanolic extract concentrated & vacua to remove most of the methanol. The plasma samples and liver extracts were extracted with CH2Cl.2and the CH2Cl2 layer was evaporated. The residue was dissolved in CH30H and diluted withO. N-NaOH. After 10 minutes, the aqueous layer was washed with CH2Cl.2,acidified with LN-HCl to pH 1 and extracted with CH2C12. The organic layer was then washed with 1% NaHC03 solution and with water and was evaporated at 50. under N2. This procedure results in an effective purification of spirolactone metabolftes, including separation from aldosterone. Control experiments by column chromatography confirmed that the pH 13 and pH 1 treatment did not result in chemical changes of canrenoate-K metabolites. The resulting extract was subjected to thin layer chromatography (Silica CH2Cl2:CH OH (90:10)), and radioGel KG Merck F 254, solvent Sy8iXm active bands were eluted from the plates wit2 methanol. These bands were rechromatographed on a Sephadex LH-20 column with CU2C12:CH30H (98:2) as solvent system, and appropriate fractions were collected (cancounted for 3?lcontent and subjected to fluorescence, renone, MA, g>, W and mass svectral analysis. 5. Cross-raaitivity of S&rolactones to Aldosterone Antibody Tritiated aldosterone (3 DQ) with a soecific activity of 300 dnm/ pg (Amersham-SearleCorporation;.Chicago) ;as added to solutions of'its aldosterone 21-hem1 uccinyl BSA antibody sufficiently concentrated to bind 60-652 of the $I-aldosterone (titer 1:SOOOO). Displacement of 3Haldosterone from its antibody was measured with amounts of spirolactones ranging from I-200 ng. The following spirolactones were utilized: spironolactone, its major metabolite canrenone (l), canrenoate-K, and

304

S

TDEOXDI

two new metabolite fractions, MA and Mb, isolated from rabbit plasma (see 4). 6. Fluorescence Measurements of Canrenone and Canrenoate The fluorescence procedures were described recently (28). The new spirolactone metabolite fractions, MA and isolated from rabbit plasma and liver (see 4) do not interfere with"Bhe assay of canrenone and canrenoate. 7. Apparatus UV spectra were taken on a Beckman UV Spectrophotometer.Fluorescence was read on an Aminco-Bowman Spectrophotofluorimeter. Mass spectra were obtained by direct insertion on a Varian CH-7 Mass Spectrometer using 70eV electron impact and on an AK1 MS 902 high resolution mass spectrometer using chemical ionization with isobutane. 8. Chemicals Spironolactone: 17-hydroxy-7a-acetylthio-&oxo-4-androsten-17a-y1 B-propionic acid lactone. Canrenone: 17-hydroxy-3-oxo-4,6-androstadien17a-yl B-propionic acid lactone. Canrenoate-K: Potassium 17-hydroxy-3oxo-4,6-androstadien-17a-yl8-propionate. Amounts of isolated met bolites MA and MB were calculatgd from their &activity and the specific $activity of the parent H-canrenoate-K. RESULTS Following single doses of spironolactone and canrenoate-K in rabbits, dogs, and man, we have found increased aldosterone plasma values using the radioimmunoassay by Ito, -et al. (26). However, high aldosterone values after canrenoate-K administration were also observed in adrenalectomized dogs, which suggests that the spirolactones or their metabolites interfere with the aldosterone radioimmunoassay. Spironolactone, Its major metabolite canrenone (21, and canrenoate did not contribute to this assay interference, since canrenoate was not extractable by CU2C12 from aqueous medium above pH 6 and splronolactone and canrenone were completely separated from aldosterone during Sephadex LH-20 column chromatography even in a lo6 fold excess.

The

erratic aldosterone values were equally not dependent on sulfur-retaining metabolites specific to spironolactone (29). since splronolactone as well as canrenoate-K without the 7a-thioacetyl substltuent led to equivalent increases in false aldosterone readings in rabbits. Thus,

S high aldoeterone spironolactone

readings

TDEOID-

305

were caused by a yet unknown metabolite

of

and canrenoate-K.

3

0

CM& ..*

JZXP c

0 Y

Spironolactone Spirolactone

2 hours following

point

the highest

by differential trography. eluted

were isolated

i.v.

from rabbit

administration

aldoeterone

from aldosterone

was observed. and other

of unknown metabolites

1

konsisted

were detectable

Column chromatography

from rabbit plasma (II:

of

by chroma-

of 3H-canrenoate-K

of canrenone.

metabolitea

volume of aldos-

Two further

(MA and sg>,

The

corticosteroids

at pH 13 and pH 1 and purified

Figure 1 shows the 3H-activity

One 31i-fraction

plasma and

of 311-canrenoate-K at which

assay interference

were separated extractions

canrenoate

from a Sephadex LH-20 column and the retention

terone.

Fig.

canrenone

metabolitee

liver

spirolactones

’

fractions

one of which (Mb)

3 H-canrenoate-K metabolites extracted

canrenone).

S

TIlXSO

SD=

with a retention volume similar to that of aldosterone. The metabolic fraction Mg was therefore studied by UV and mass spectrometry.

The

UV maximum at 284 mm in methanol suggested an intact 4,6-dien-3-one function. Electron impact and chemical ionization mass spectrometry gave parent ions at m/e 356 and m/e 357 (M+ + H), respectively, which indicated a molecular weight of 356 for Mb.

An abundant ion fragment

at m/e 338 (M+-H,O) and m/e 339 (