- Email: [email protected]
A simple sensitive radioreceptor assay for calcium antagonist drugs
Life Sciences, Vol. 33, pp. 2665-2672 Printed in the U.S.A.
Pergamon Pres~
A SIMPLE SENSITIVE RADIORECEPTOR ASSAY FOR CALCIUM ANTAGONIST DRUGS
Robert J. Gould, Kenneth M.M. Murphy and Solomon H. Snyder* Departments of Neuroscience, Pharmacology and Experimental Therapeutics, Psychiatry and Behavioral Sciences Johns Hopkins University School of Medicine 725 North Wolfe Street Baltimore, Maryland 21205 (301) 955-3024 (Received in final form October 12, 1983)
Sun~nar~ A radioreceptor assay for calcium channel antagonist drugs described here is based on the a b i l i t y of these drugs to affect 3H-nitrendipine binding to calcium channels. All the known calcium channel antagonists may be assayed in this manner. The assay can detect I0-I00 nM ~4 - 40 ng/ml) nimodipine, 10-100 nM (3.5 35 ng/ml) nifedipine, 3-30 ~M (1.2 - 12 ~g/ml) prenylamine, 0.1 1.0 ,M (49 - 490 ng/ml) verapamil and 3-30 ~M (1.2 - 12 ,g/ml) diltiazem. These values cover the range of concentrations of calcium channel antagonists that are c l i n i c a l l y important. As the radioreceptor assay detects active metabolites as well as the parent drugs, i t should prove a useful adjunct in cardiovascular therapy. The method is more reproducible, simpler and less expensive than other methods such as high pressure liquid chromatography. Radioreceptor assays have proved valuable for measuring blood levels of a variety of drugs. A radioreceptor assay for neuroleptics depends on their a b i l i t y to compete for the binding of ~H-ligands to D2 dopamine receptors (1,2). Beta adrenergic antagonist drugs can be measured by their a b i l i t y to compete for the binding of ~H-ligands to beta receptors (3-5). Benzodiazepines can similarly be measured by competition for benzodiazepine receptor binding sites (6-8). Tricyclic antideRressant drugs can be measured by their a b i l i t y to compete for the binding of ~H-ligands to muscarinic cholinergic receptors (9). Radioreceptor assays possess a number of virtues as compared to other techniques for measuring drugs. For drugs such as neuroleptics, beta blockers and benzodiazepines, where the receptor employed represents the site of therapeutic action of the drugs, the monitored blood level will r e f l e c t not only the parent drug but any therapeutically active metabolite. Also, the assay will not be restricted to a single drug but to all agents which act at the same receptor. This is particularly important in the case of drugs such as the neuroleptics, beta blockers and benzodiazepines, where large numbers of drugs of these classes are marketed. Recently drugs which act by blocking voltage-sensitive calcium channels
0024-3205/83 $3.00 + .00 Copyright (c) 1983 Pergamon Press Ltd.
2666
Radioreceptor Assay of Calcium Blockers
Vol. 33, No. 26, 1983
have shown therapeutic value for angina, hypertension, migraine, subarachnoid hemorrhage, dysmenorhrea, premature labor, asthma and cardiac arrhythmias (for reviews, see 10-12). The major current therapeutic application of these drugs is for cardiac arrhythmias and angina (11). A simple technique for measuring blood levels of these drugs is desirable for several reasons. Most of the patients treated are susceptible to cardiovascular side-effects. Moreover, these patients are usually treated with several drugs, some of which may alter the metabolism of others. Besides metabolic interactions, drugs such as beta blockers may interact pharmacologically with calcium antagonists at sites such as the SA node of the heart where potentially l i f e threatening arrhythmias may result with drug concentrations higher than are therapeutically indicated. As nimodipine may be used prophylactically for migraine, a sensitive assay for measuring blood levels would allow precise individual adjustment of dosage levels (13). Verapamil is most effective i f plasma levels above 100 ng/ml are attained (14). Currently, the most widely used method for measuring calcium channel antagonist blood levels is by HPLC, which requires extensive handling of the samples (14-18). As these drugs are light-sensitive, manipulations are performed under a Na vapor lamp which is inconvenient.
Receptor sites for calcium antagonist drugs have been labeled with the dihydropyridine 3H-nitrendipine (19-27). All known dihydropyridine drugs compete with considerable potency at these binding s i t e s . Drugs such as verapamil and diltiazem do not compete d i r e c t l y for these binding sites but influence the binding of 3H-nitrendipine in an a l l o s t e r i c fashion (20-22). In the present study we describe a radioreceptor assay for calcium antagonist drugs which can be employed to monitor blood levels of a l l the calcium antagonist arugs presently employea c l i n i c a l l y including those of the dihydropyridine, verapamil and diltiazem classes, while this work was in progress, a similar assay has been developed f o r nitrendipine (28). Methods
Cerebral cortex membranes were used as a source of 3H-nitrendipine binding s i t e s , Bovine brain was obtained at a local abattoir and the cortex was immediately dissected. The cortex was weighed and homogenized in 10 vol. of ice-cold 50 n~M Tris-HCl buffer, pH 7.7 with a Brinkman Polytron (setting 7) for 30 sec. The homogenate was centrifuged at 4°C for 10 min. at 40,000 x g in a Sorvall RC5B centrifuge. The p e l l e t s were resuspended in 20 ml of ice-cold Tris-HCl, pH 7.7 and recentrifuged. This was repeated twice. The p e l l e t s were f i n a l l y resuspended into 10 vol. of ice-cold 50 ~ Tris-HCl, pH 7.7 and frozen at -70°C u n t i l use. When used, homogenates were thawed and adjusted to a f i n a l vol. of 400 ml by adding 380 ml of ice-cold 50 nW~ Tris-HCl, pH 7.7. Nifedipine ( P f i z e r , Groton, CT), nimodipine and n i t r e n d i p i n e (Miles, New Haven, CT) were dissolved in 100°/° ethanol at 10-2 M and diluted with 50 ~ I r i s HCI to the appropriate concentrations from this stock solution. Diltiazem (Marion, Kansas City, MO), prenylamine (Hoechst-Roussel, Sommerville, NJ), tiapamil (Hoffmann-LaRoche, Nutley, N~) and verapamil (Knoll AG, Ludwigshafen, West Germany) were dissolved in 50 /° 0.5 ~ HCI (vol/vol) at i0 -~ M and diluted with 50 n~i I r i s HCI to the appropriate concentrations~ Binaing was carried out in a total volume of i ml containing 0.15 - 0.20 nM OH-nitrendipine (70 Ci/mmol, New England Nuclear), 5 mg tissue (wet weight) and either serum or drug standards. After incubating for 1 hr at room temperature, the reaction was terminated by rapid vacuum f i l t r a t i o n over either Whatman GF/B glass fibre f i l t e r s or Schleicher and Schuell #32 glass fibre f i l t e r s . Equilibrium is attained by 0.5 - 0.75 hr. in the absence of any added inhibitors (19-27). Thus, even in the presence of inhibitors, a one hour incubation is sufficient to attain equilibrium. The f i l t e r s were rapidly washed four times with 2 ml of 50 m~i Tris-HCl, pH 7.7. Nonspecific binding was defined as that which occurred in the presence of 100 nM nifedipine and was subtracted from total binding to give the specific binding.
Vol. 33, No. 26, 1983
Radioreceptor Assay of Calcium Blockers
2667
Standard curves were generated by adding known concentrations of nimodipine, nifedipine, or prenylamine, dissolved in human plasma, to incubation tubes containing JH-nitrendipine and bovine cerebral cortex membranes. The percent of control binding assayed in the absence of drug, was plotted on logit paper and was consequently linear. A diltiazem standard curve was generated by adding known concentrations of diltiazem, dissolved in human plasma, to tubes containing 3H-nitrendipine, 2.5 ~M tiapamil and bovine cerebral cortex membranes. As diltiazem reverses tiapamil inhibition of 3H-nitrendipine binding, this provides a quantitative and sensitive assay for diltiazem (20). Human out-dated platelet-rich plasma samples were obtained from the Johns Hopkins Hospital Blood Bank. Platelets were removed by centrifugation prior to assay. For determination of blood levels in rats, animals were injected i.p. with 0.5 mg/kg, 1 mg/kg or 2 mg/kg nitrendipine dissolved in polyethylene glycol 400. After 1 hr, animals were bled and the blood collected in tubes containing 0.10 ml of heparin sulfate (1 mg/ml). The plasma was isolated and the blood levels of nitrendi~ine then measured. Someanimals were injected i.p. with the same doses of ~H-nitrendipine (2.75 Ci/mole). Radioactivity in the plasma was measured by liquid s c i n t i l l a t i o n spectrometry. Results and Discussion As observed previously (19-27), dihydropyridines such as nimodipine and nifedipine reduce 3H-nitrendipine binding with Hill coefficients of approximately 1 (Figure i ) . When using log probit analysis, straight lines are obtained for these competition curves from which one can readily estimate the concentration of an unknown sample of drug. A number of calcium channel antagonists that are not dihydropyridines, such as prenylamine, also i n h i b i t 3H-nitrendipine binding completely (20,25). Thus, blood levels of these drugs may also be assayed by this radioreceptor assay. These drugs include prenylamine, tiapamil, lidoflazine, flunarizine and cinnarizine (20,25). Standard curves were prepared by adding known concentrations of nimodipine, nifedipine, prenylamine and diltiazem (Figure 1). Unknown samples were prepared by diluting drugs to a known concentration and conducting the experiments in a double-blind fashion. Figure i demonstrates the unknown drug sample concentrations as determined from the radioreceptor assay. Table I indicates the close correlation between the predicted and measured drug concentrations. With the exception of diltiazem, the measured concentration is always less than the actual concentration for the higher concentrations. This may be due to photolytic degradation of the compounds during addition to the plasma, as all these compounds are light sensitive. Alternatively, there may be components in the plasma which oxidize the compounds to an inactive form. In brain membranes verapamil only p a r t i a l l y reduces 3H-nitrendipine binding (20,21). However, in skeletal muscle we observe a 80-90"/° displacement of ~H-nitrendipine binding by verapamil (Figure 2). Thus, one can estimate verapamil concentrations directly from these competition curves provided that the source of receptor material is skeletal muscle. In contrast to the other calcium antagonist drugs, diltiazem actually stimulates mH-nitrendipine binding (25-27). The extent of this augmentation of binding is only about 20-30°/° which would not be adequate to permit an accurate estimation of concentrations of unknown samples. Diltiazem influences 3H-nitrendipine binding at the same site as verapamil and tiapamil (20,25). Diltiazem can compete with verapamil or tiapamil to block their inhibition of 3H-nitrendipine binding. Thus, i f ~H-nitrendipine binding is maximally reduced by tiapamil, then diltiazem can restore
2668
Radioreceptor Assay of Calcium Blockers
100
I
I
Vol. 33, No. 26, 1983
;
x= UNKNOWNS z
Z
~
~-~80
OILTIAZEM.
6o
.~ 40
nZ
2Q
-8
~ . . ~ _ . . . .
I
NIMODIPINE I
NIFEDIPIN:RENYLAMINE
I
-7 -6 -5 LOG DRUG CONCENTRATION {Molar)
-4
IN PLASMA
FIG. 1 Standard curves for assay of nifedipine, nimodipine, prenylamine and diltiazem. Nimodipine, nifed~pine and prenylamine standard curves were generated by inhibition of ~H-nitrendipine binding with known concentrations of the drug in human plasma. The diltiazem standard curve was generated by reversal of the inhibition of 3H-nitrendipine binding engendered by 2.5 ~M tiapamil. "Unknown": samples of human plasma with various drugs added. this binding to control levels. One can readily estimate diltiazem concentrations by monitoring the enhancement of ~H-nitrendipine binding which has been lowered by tiapamil (Figure 1). This increases the sensitivity of the assay by increasing the changes in specific binding of 3H-nitrendipine. To employ the radioreceptor assay with untreated plasma samples, the amount of plasma routinely employed should not inhibit 3H-nitrendipine binding i t s e l f . Accordingly, we evaluated the influence of human plasma on 3H-nitrendipine binding. In these experiments plasma of unmedicated adult males and females was employed. Samples of 10 ~I were added to incubations with 3H-nitrendipine in a total volume of i ml. To evaluate the possibility that inhibition of 3H-nitrendipine binding by human plasma may vary among individuals, we utilized plasma from 29 different subjects. To determine whether plasma effects on binding might vary within a subject, we utilized plasma samples from each subject on two consecutive days. The mean inhibition of bindin~ for the samples on day one was 15.7 ~ 1.4°I ° and on day two was 8.9 + 1.3 1°. The reasons for the lesser inhibition of samples from day two are not apparent. In studies of these plasma samples i t became apparent that a higher degree of inhibition of binding occurs with samples which are somewhat turbid due to high lipid content. This would f i t well with the known propensity of
Vol. 33, No. 26, 1983
Radioreceptor Assay of Calcium Blockers
2669
TABLE I Correlation Between Actual and Measured Drug Concentration in Plasma
Drug
Actual Concentration
Measured Concentration (,M)
Nifedipine
0.i0 0.33
0 . I 0 + 0.01 (.08-.13) 0.17 ¥ 0.01 (.13-.21)
Nimodipine
0.i0 0.33
0.08 ÷ 0.01 ( . 0 6 - . 1 ) 0.17 T 0.01 (.13-.21)
Prenylamine Diltiazem
10 30 15 30 50 100
9.25 + 2.32 ( 5 - 1 5 ) 22.75 ¥ 1.60 (20-26) 24 38 80 130
Standard curves for inhibition of 3H-nitrendipine binding by nifedipine, nimodipine, prenylamine and diltiazem were prepared as described in Figure 1. The concentrations of unknown samples, prepared and assayed in a double-blind fashion, were then determinated from the standard. Each value, except for diltiazem, represents the mean + standard error of 4 determinations, each performed in t r i p l i c a t e . - The range of values are indicated in parentheses. Diltiazem is a single experiment, performed in t r i p l i c a t e . 3H-nitrendipine to bind to plasma protein and f a t t y surfaces. Accordingly, one would anticipate less interference with assays i f samples were routinely obtained from subjects after an overnight fast. To evaluate the u t i l i t y of the radioreceptor assay after in vivo administration of drugs, we administered to rats i.p. several doses of 3H-nitrendipine mixed with unlabeled nitrendipine and monitored plasma levels 1 hour after drug administration, when blood levels are near maximal (Table I I , 17). While the radioreceptor assay only measures parent drug and active metabolites, the total radioactivity in the sample will include inactive as well as active metabolites (Table I I ) . The much lower values obtained by radioreceptor assay indicate that receptor-active materials only represent 2.9°/° - 3.6°/° of the total nitrendipine metabolites. This is in agreement with other reports showing substantial metabolism of nifedipine (15). The major metabolite of nifedipine, and presumably nitrendipine, is the de-esterified analog which is pharmacologically inactive (29). Another important aspect of any^assay is sensitivity. Does the potency of the various calcium antagonists at JH-nitrendipine binding sites suffice to permit measurement of therapeutic blood levels of these drugs? Studies monitoring blood levels of nifedipine and verapamil by high-pressure liquid chromatography, reveal mean blood levels at therapeutic doses of 50 - 100 ng/ml of plasma (14,18). The three calcium antagonists currently
2670
Radioreceptor Assay of Calcium Blockers
I
100
._c
80
-
"\
"o .c
="
Vol. 33, No. 26, 1983
~6o
,
,
,
"B
zI
~
40
20
I 9
I 8
I 7
-Log
I 6
[ve~apam,I (M)
FIG. 2 Complete inhibition by verapamil of 3H-nitrendipine binding to skeletal muscle membranes. Total and nonspecific binding to skeletal muscle membranes were determined as described l~nMethods in the absence or presence of various concentrations of verapamil. The inset shows the same oata plotted as the logit transformation. TABLE II Comparison of Plasma Radioactivity Following 3H-Nitrendipine and Nitrendipine Plasma Levels Determined by Radioreceptor Assay (RRA)
Dose
(mg/kg)
3H-Nitrendipine ~M
Nitrendipine by RRA ~M
0.5 mg/kg
2.27 + 0.16
0.0074 + 0.011
1
mg/kg
4.91 + 0.49
0.144
+ 0.027
2
mg/kg
10.83 + 0.63
0.392
+ 0.055
1
1
1
Rats were administered either nitrendipine alone or nitrendipine plus 3H-nitrendipine at the indicated aosages. After i hr, blood was collected by cardiac puncture into heparin (0.1 ml, 1 mg/ml). Plasma nitrendipine concentration was determined by the radioreceptor assay and total radioactivity assayed by liquid s c i n t i l l a t i o n spectrometry.
Vol. 33, No. 26, 1983
Radioreceptor Assay of Calcium Blockers
2671
marketea in the United States are nifedipine, verapamil and diltiazem, whose EC50 values at 3H-nitrendipine sites are approximately 1.0, 10.0 and 1000 nM (0.3, 0.5, 40 nglml). The minimal concentration of a drug that can be reliably detected is a concentration that alters binding about 15-I °. For nifedipine, verapamil and diltiazem, this would correspond to 0.1, 1.0 and 100 nM. The assay described in the present study requires a lO0-fold dilution of plasma samples. For nifedipine, patients with blood levels of 10 - 100 nglml would give rise to final concentrations of 0.3 - 3 nM which would be readily detected. For verapamil these concentrations would be 3 - 30 nM, which can also be readily monitored. For diltiazem the assay presently employed may not suffice to measure therapeutic blood levels. Sensitivity of the present assay can be readily enhanced by a number of techniques. Plasma samples must be diluted lO0-fold in the assay to reduce inhibition of 3H-nitrendipine binding. Presumably such inhibition is due to plasma protein which binds directly to the H-nitrendipine, reducing the free 3H-nitrendipine concentration accessible to receptor sites. A similar limitation was identified in assays of beta adrenergic antagonist drugs (3). Dialysis of plasma samples by r e l a t i v e l y simple techniques that are readily applied to large numbers of samples abolishes the inhibition by plasma protein of 3H-dihydroalprenolol binding to beta receptors so that much larger patient samples can be employed (3). Such a modification could easily increase the sensitivity of the present assay up to 50-fold. The d i f f i c u l t y with u t i l i z i n g the dialysates is that one will only monitor the free drug which may be less than 10°I ° of the total plasma level of drug. Nonetheless, one virtue of measuring free drug levels is that for drugs such as beta antagonists, therapeutic response correlates more closely with free than with total plasma levels of drug (3). The other principal means of increasing sensitivity is to concentrate the drug by techniques such as solvent extraction. This is typically done in the high performance liquid chromatographic techniques for measuring calcium channel antagonist level
(14-18).
In summary, this radioreceptor assay offers a potential for ready application to large numbers of patients. The procedure is adequately sensitive for most calcium antagonist drugs u t i l i z i n g plasma samples directly. The assay is selective for calcium antagonist drugs. Of large numbers of drugs screened, the only ones which influence 3H-nitrendipine binGing are those which possess pharmacologically relevant calcium antagonist properties (19-27). Another virtue of the assay is that i t w i l l detect therapeutically active metabolites. For instance, after treatment with therapeutic doses of verapamil, plasma concentrations of norverapamil, an active metabolite, are similar to those of verapamil (14). Thus, simply measuring levels of the administered drug is not sufficient. The radioreceptor assay avoids the d i f f i c u l t y of not knowing what metabolites may be active. Norverapamil and verapamil may not have the same IC50 for 3H-nitrendipine binding sites. The measured concentration of norverapamil may, therefore, not r e f l e c t the actual concentration of norverapamil, but rather the concentration in "verapamil equivalents". That is, in terms of the amount of verapamil that would give the same amount of displaced binding. The measured concentration for verapamil, and indeed any of the calcium channel antagonists, is consequently a complex mixture of administered drug and active metabolites. I t need not necessarily reflect the actual concentration of each. The virtue of the assay is in detecting all these therapeutically active metabolites. The radioreceptor assay described here is simple to perform. As many as 200 assays can be readily conducted in a morning.
2672
Radioreceptor Assay of Calcium Blockers
Vol. 33, No. 26, 1983
Acknowledgements Supported by USPHSgrants MH-18501, NS-16375, RSA Award DA-O0074 to S.H.S., training grant GM-O73090 to K.M.M.M. and training grant MH-15330 to R.J .G. References 1. I. CREESE and S.H. SNYDER Nature 270 180-182 (1977). 2. L.E. TUNE, I. CREESE, J. COYLE, G. ~RLSON and S.H. SNYDER Am. J. Psychiatry 137 80-82 (1982). 3. R.B. INNIS, D.B. BYLUNDand S.H. SNYDER Life Sci. 23 2031-2038 (1978). 4. S.R. NAHORSKI, M.I. BATTA and D.B. BARNETT Eur. J. ~armacol. 52 393-396 (1978). 5. J.P. BILEZIKIAN, D.E. GAMMON, C.L. ROCHESTERand D.G. SHAND Clin. Pharmacol. Ther. 26, 173-180 (1979). 6. P. SKOLNICK, F.K. ~-~[ODWIN and S.M. PAUL Arch. Gen. Psychiat. 36 78-80 (1979). 7. P. HUNT, J.-M. HUSSONAND J.-P. ILAYMOND J. Pharm. Pharmacol. 31 448-451 (1979). 8. F. OWEN, R. LOFTHOUSEand R.C. BOURNE Clin. Chim. Acta. 93 305-310 (1979). 9. R.B. INNIS, L. TUNE, R. ROCK, R. DEPAULO, D.C. U'PRICHARDand S.H. SNYDER Eur. J. Pharmacol. 58 473-477 (1979). 10. A. FLECKENSTEIN Ann.'--Rev. Pharmacol. Toxicol. 17 149-166 (1977). 11. P.D. HENRY Am. J. Cardiol. 46 1047-1058 (1980)-~12. R.A. JANIS and D.J. TRIGGLE ~ M e d . Chem. 26, 775-785 (1983). 13. H.J. GELMERS Headache 23 106-109 (1983). 14. W. FRISHF~AN, E. KIRSTEIN,'--M. KLEIN, M. PINE, S.M. JOHNSON, L.D. HILLIS, M. PACKER and R. KATES Am. J. Cardiol. 50 1180-1184 (1982). 15. P. PIETTA, A. P~AUAand P. BIONDI J. C~omatog. 210 516-521 (1981). 16. T. SADANAGA, K. HIKIDA, K. TAMETO, Y. ~TSUSHIF~A and Y. OHKURA Chem. Pharm. Bull. 30 3807-3809 (1982). 17. R. TESTA, E. DOLFINI, C. ROSCHIOTTO, C. SECCHI and P.A. BIONDI II Farmaco Ed. Pr. 34 463473 (1979). 18. K. AOKI, ~ SATO, Y. K~AWAGUCHIand M. YAMAMOTO Eur. J. Clin. Pharmaco]. 23 197-201 (1982). 19. ~ J . GOULD, K.M.M. MURPHYand S.H. SNYDER Proc. Natl. Acad. Sci. USA 79 3656-3660 (1982). 20. K.M.M. MURPHY, R.J. GOULD, B.L. LARGENTand S.H. SNYDER Proc. Natl. Acad. Sci. USA 80 860464 (1983). 21. G.T. BOLGE~P. GENGO, R. KLOCKOWSKI, E. LUCHOWSKI, H. SIEGEL, R.A. JANIS, A.M. TRIGGLE and D.J. TRIGGLE J. Pharmacol. Exp. Ther. 225 291-309 (1983) 22. F.J. EHLERT, W.R. ROESKE, E. ITOGA and H.I. YAMAMURA Life Sci. 30 2191-2202 (1982). 23. L.T. WILLIAMS and P. TREMBLE J. Clin. Invest. 70 209-212 (1982). 24. P. BELLEFtANN, D. FERRY, F. LUBBACKEand H. GLOSSFtANN Arzneim.-Forsch./Drug Res. 31 2064-2067 (1981). 25. K.M.M. MURPHY, R.J. GOULDand--S.H. SNYDER in Nitrendipine (A. Scriabine, S. Vanor and K. Deck, eds.), Urban and Schwarzenberg, Baltimore-Munich, in press, (1983). 26. H.I. YAJ~A~IURA, H. SCHOE~LAKER,R.G. BOLES and W.R. ROESKE Biochem. Biophys. Res. Commun. 108 640~46 (1982). 27. A. DE POVER, M.A. MATLIB, S.W. LEE, G.P. DUBE, I.L. GRUPP, G. GRUPPand A. SCHWARTZ Biochem. Biophys. Res. Commun. 108 110-117 (1982). 28. R.A. JANIS, G.J. KROHL, A.J. NOE AND M. PAN J. Clin. Pharmacol., 2__3 266-273 (1983). 29. H. MEDENWALD, K. SCHLOSSF~ANNand C. WINSCHE Arznein.-Forsch./Drug Res. 22 53-56 (1972).
Pergamon Pres~
A SIMPLE SENSITIVE RADIORECEPTOR ASSAY FOR CALCIUM ANTAGONIST DRUGS
Robert J. Gould, Kenneth M.M. Murphy and Solomon H. Snyder* Departments of Neuroscience, Pharmacology and Experimental Therapeutics, Psychiatry and Behavioral Sciences Johns Hopkins University School of Medicine 725 North Wolfe Street Baltimore, Maryland 21205 (301) 955-3024 (Received in final form October 12, 1983)
Sun~nar~ A radioreceptor assay for calcium channel antagonist drugs described here is based on the a b i l i t y of these drugs to affect 3H-nitrendipine binding to calcium channels. All the known calcium channel antagonists may be assayed in this manner. The assay can detect I0-I00 nM ~4 - 40 ng/ml) nimodipine, 10-100 nM (3.5 35 ng/ml) nifedipine, 3-30 ~M (1.2 - 12 ~g/ml) prenylamine, 0.1 1.0 ,M (49 - 490 ng/ml) verapamil and 3-30 ~M (1.2 - 12 ,g/ml) diltiazem. These values cover the range of concentrations of calcium channel antagonists that are c l i n i c a l l y important. As the radioreceptor assay detects active metabolites as well as the parent drugs, i t should prove a useful adjunct in cardiovascular therapy. The method is more reproducible, simpler and less expensive than other methods such as high pressure liquid chromatography. Radioreceptor assays have proved valuable for measuring blood levels of a variety of drugs. A radioreceptor assay for neuroleptics depends on their a b i l i t y to compete for the binding of ~H-ligands to D2 dopamine receptors (1,2). Beta adrenergic antagonist drugs can be measured by their a b i l i t y to compete for the binding of ~H-ligands to beta receptors (3-5). Benzodiazepines can similarly be measured by competition for benzodiazepine receptor binding sites (6-8). Tricyclic antideRressant drugs can be measured by their a b i l i t y to compete for the binding of ~H-ligands to muscarinic cholinergic receptors (9). Radioreceptor assays possess a number of virtues as compared to other techniques for measuring drugs. For drugs such as neuroleptics, beta blockers and benzodiazepines, where the receptor employed represents the site of therapeutic action of the drugs, the monitored blood level will r e f l e c t not only the parent drug but any therapeutically active metabolite. Also, the assay will not be restricted to a single drug but to all agents which act at the same receptor. This is particularly important in the case of drugs such as the neuroleptics, beta blockers and benzodiazepines, where large numbers of drugs of these classes are marketed. Recently drugs which act by blocking voltage-sensitive calcium channels
0024-3205/83 $3.00 + .00 Copyright (c) 1983 Pergamon Press Ltd.
2666
Radioreceptor Assay of Calcium Blockers
Vol. 33, No. 26, 1983
have shown therapeutic value for angina, hypertension, migraine, subarachnoid hemorrhage, dysmenorhrea, premature labor, asthma and cardiac arrhythmias (for reviews, see 10-12). The major current therapeutic application of these drugs is for cardiac arrhythmias and angina (11). A simple technique for measuring blood levels of these drugs is desirable for several reasons. Most of the patients treated are susceptible to cardiovascular side-effects. Moreover, these patients are usually treated with several drugs, some of which may alter the metabolism of others. Besides metabolic interactions, drugs such as beta blockers may interact pharmacologically with calcium antagonists at sites such as the SA node of the heart where potentially l i f e threatening arrhythmias may result with drug concentrations higher than are therapeutically indicated. As nimodipine may be used prophylactically for migraine, a sensitive assay for measuring blood levels would allow precise individual adjustment of dosage levels (13). Verapamil is most effective i f plasma levels above 100 ng/ml are attained (14). Currently, the most widely used method for measuring calcium channel antagonist blood levels is by HPLC, which requires extensive handling of the samples (14-18). As these drugs are light-sensitive, manipulations are performed under a Na vapor lamp which is inconvenient.
Receptor sites for calcium antagonist drugs have been labeled with the dihydropyridine 3H-nitrendipine (19-27). All known dihydropyridine drugs compete with considerable potency at these binding s i t e s . Drugs such as verapamil and diltiazem do not compete d i r e c t l y for these binding sites but influence the binding of 3H-nitrendipine in an a l l o s t e r i c fashion (20-22). In the present study we describe a radioreceptor assay for calcium antagonist drugs which can be employed to monitor blood levels of a l l the calcium antagonist arugs presently employea c l i n i c a l l y including those of the dihydropyridine, verapamil and diltiazem classes, while this work was in progress, a similar assay has been developed f o r nitrendipine (28). Methods
Cerebral cortex membranes were used as a source of 3H-nitrendipine binding s i t e s , Bovine brain was obtained at a local abattoir and the cortex was immediately dissected. The cortex was weighed and homogenized in 10 vol. of ice-cold 50 n~M Tris-HCl buffer, pH 7.7 with a Brinkman Polytron (setting 7) for 30 sec. The homogenate was centrifuged at 4°C for 10 min. at 40,000 x g in a Sorvall RC5B centrifuge. The p e l l e t s were resuspended in 20 ml of ice-cold Tris-HCl, pH 7.7 and recentrifuged. This was repeated twice. The p e l l e t s were f i n a l l y resuspended into 10 vol. of ice-cold 50 ~ Tris-HCl, pH 7.7 and frozen at -70°C u n t i l use. When used, homogenates were thawed and adjusted to a f i n a l vol. of 400 ml by adding 380 ml of ice-cold 50 nW~ Tris-HCl, pH 7.7. Nifedipine ( P f i z e r , Groton, CT), nimodipine and n i t r e n d i p i n e (Miles, New Haven, CT) were dissolved in 100°/° ethanol at 10-2 M and diluted with 50 ~ I r i s HCI to the appropriate concentrations from this stock solution. Diltiazem (Marion, Kansas City, MO), prenylamine (Hoechst-Roussel, Sommerville, NJ), tiapamil (Hoffmann-LaRoche, Nutley, N~) and verapamil (Knoll AG, Ludwigshafen, West Germany) were dissolved in 50 /° 0.5 ~ HCI (vol/vol) at i0 -~ M and diluted with 50 n~i I r i s HCI to the appropriate concentrations~ Binaing was carried out in a total volume of i ml containing 0.15 - 0.20 nM OH-nitrendipine (70 Ci/mmol, New England Nuclear), 5 mg tissue (wet weight) and either serum or drug standards. After incubating for 1 hr at room temperature, the reaction was terminated by rapid vacuum f i l t r a t i o n over either Whatman GF/B glass fibre f i l t e r s or Schleicher and Schuell #32 glass fibre f i l t e r s . Equilibrium is attained by 0.5 - 0.75 hr. in the absence of any added inhibitors (19-27). Thus, even in the presence of inhibitors, a one hour incubation is sufficient to attain equilibrium. The f i l t e r s were rapidly washed four times with 2 ml of 50 m~i Tris-HCl, pH 7.7. Nonspecific binding was defined as that which occurred in the presence of 100 nM nifedipine and was subtracted from total binding to give the specific binding.
Vol. 33, No. 26, 1983
Radioreceptor Assay of Calcium Blockers
2667
Standard curves were generated by adding known concentrations of nimodipine, nifedipine, or prenylamine, dissolved in human plasma, to incubation tubes containing JH-nitrendipine and bovine cerebral cortex membranes. The percent of control binding assayed in the absence of drug, was plotted on logit paper and was consequently linear. A diltiazem standard curve was generated by adding known concentrations of diltiazem, dissolved in human plasma, to tubes containing 3H-nitrendipine, 2.5 ~M tiapamil and bovine cerebral cortex membranes. As diltiazem reverses tiapamil inhibition of 3H-nitrendipine binding, this provides a quantitative and sensitive assay for diltiazem (20). Human out-dated platelet-rich plasma samples were obtained from the Johns Hopkins Hospital Blood Bank. Platelets were removed by centrifugation prior to assay. For determination of blood levels in rats, animals were injected i.p. with 0.5 mg/kg, 1 mg/kg or 2 mg/kg nitrendipine dissolved in polyethylene glycol 400. After 1 hr, animals were bled and the blood collected in tubes containing 0.10 ml of heparin sulfate (1 mg/ml). The plasma was isolated and the blood levels of nitrendi~ine then measured. Someanimals were injected i.p. with the same doses of ~H-nitrendipine (2.75 Ci/mole). Radioactivity in the plasma was measured by liquid s c i n t i l l a t i o n spectrometry. Results and Discussion As observed previously (19-27), dihydropyridines such as nimodipine and nifedipine reduce 3H-nitrendipine binding with Hill coefficients of approximately 1 (Figure i ) . When using log probit analysis, straight lines are obtained for these competition curves from which one can readily estimate the concentration of an unknown sample of drug. A number of calcium channel antagonists that are not dihydropyridines, such as prenylamine, also i n h i b i t 3H-nitrendipine binding completely (20,25). Thus, blood levels of these drugs may also be assayed by this radioreceptor assay. These drugs include prenylamine, tiapamil, lidoflazine, flunarizine and cinnarizine (20,25). Standard curves were prepared by adding known concentrations of nimodipine, nifedipine, prenylamine and diltiazem (Figure 1). Unknown samples were prepared by diluting drugs to a known concentration and conducting the experiments in a double-blind fashion. Figure i demonstrates the unknown drug sample concentrations as determined from the radioreceptor assay. Table I indicates the close correlation between the predicted and measured drug concentrations. With the exception of diltiazem, the measured concentration is always less than the actual concentration for the higher concentrations. This may be due to photolytic degradation of the compounds during addition to the plasma, as all these compounds are light sensitive. Alternatively, there may be components in the plasma which oxidize the compounds to an inactive form. In brain membranes verapamil only p a r t i a l l y reduces 3H-nitrendipine binding (20,21). However, in skeletal muscle we observe a 80-90"/° displacement of ~H-nitrendipine binding by verapamil (Figure 2). Thus, one can estimate verapamil concentrations directly from these competition curves provided that the source of receptor material is skeletal muscle. In contrast to the other calcium antagonist drugs, diltiazem actually stimulates mH-nitrendipine binding (25-27). The extent of this augmentation of binding is only about 20-30°/° which would not be adequate to permit an accurate estimation of concentrations of unknown samples. Diltiazem influences 3H-nitrendipine binding at the same site as verapamil and tiapamil (20,25). Diltiazem can compete with verapamil or tiapamil to block their inhibition of 3H-nitrendipine binding. Thus, i f ~H-nitrendipine binding is maximally reduced by tiapamil, then diltiazem can restore
2668
Radioreceptor Assay of Calcium Blockers
100
I
I
Vol. 33, No. 26, 1983
;
x= UNKNOWNS z
Z
~
~-~80
OILTIAZEM.
6o
.~ 40
nZ
2Q
-8
~ . . ~ _ . . . .
I
NIMODIPINE I
NIFEDIPIN:RENYLAMINE
I
-7 -6 -5 LOG DRUG CONCENTRATION {Molar)
-4
IN PLASMA
FIG. 1 Standard curves for assay of nifedipine, nimodipine, prenylamine and diltiazem. Nimodipine, nifed~pine and prenylamine standard curves were generated by inhibition of ~H-nitrendipine binding with known concentrations of the drug in human plasma. The diltiazem standard curve was generated by reversal of the inhibition of 3H-nitrendipine binding engendered by 2.5 ~M tiapamil. "Unknown": samples of human plasma with various drugs added. this binding to control levels. One can readily estimate diltiazem concentrations by monitoring the enhancement of ~H-nitrendipine binding which has been lowered by tiapamil (Figure 1). This increases the sensitivity of the assay by increasing the changes in specific binding of 3H-nitrendipine. To employ the radioreceptor assay with untreated plasma samples, the amount of plasma routinely employed should not inhibit 3H-nitrendipine binding i t s e l f . Accordingly, we evaluated the influence of human plasma on 3H-nitrendipine binding. In these experiments plasma of unmedicated adult males and females was employed. Samples of 10 ~I were added to incubations with 3H-nitrendipine in a total volume of i ml. To evaluate the possibility that inhibition of 3H-nitrendipine binding by human plasma may vary among individuals, we utilized plasma from 29 different subjects. To determine whether plasma effects on binding might vary within a subject, we utilized plasma samples from each subject on two consecutive days. The mean inhibition of bindin~ for the samples on day one was 15.7 ~ 1.4°I ° and on day two was 8.9 + 1.3 1°. The reasons for the lesser inhibition of samples from day two are not apparent. In studies of these plasma samples i t became apparent that a higher degree of inhibition of binding occurs with samples which are somewhat turbid due to high lipid content. This would f i t well with the known propensity of
Vol. 33, No. 26, 1983
Radioreceptor Assay of Calcium Blockers
2669
TABLE I Correlation Between Actual and Measured Drug Concentration in Plasma
Drug
Actual Concentration
Measured Concentration (,M)
Nifedipine
0.i0 0.33
0 . I 0 + 0.01 (.08-.13) 0.17 ¥ 0.01 (.13-.21)
Nimodipine
0.i0 0.33
0.08 ÷ 0.01 ( . 0 6 - . 1 ) 0.17 T 0.01 (.13-.21)
Prenylamine Diltiazem
10 30 15 30 50 100
9.25 + 2.32 ( 5 - 1 5 ) 22.75 ¥ 1.60 (20-26) 24 38 80 130
Standard curves for inhibition of 3H-nitrendipine binding by nifedipine, nimodipine, prenylamine and diltiazem were prepared as described in Figure 1. The concentrations of unknown samples, prepared and assayed in a double-blind fashion, were then determinated from the standard. Each value, except for diltiazem, represents the mean + standard error of 4 determinations, each performed in t r i p l i c a t e . - The range of values are indicated in parentheses. Diltiazem is a single experiment, performed in t r i p l i c a t e . 3H-nitrendipine to bind to plasma protein and f a t t y surfaces. Accordingly, one would anticipate less interference with assays i f samples were routinely obtained from subjects after an overnight fast. To evaluate the u t i l i t y of the radioreceptor assay after in vivo administration of drugs, we administered to rats i.p. several doses of 3H-nitrendipine mixed with unlabeled nitrendipine and monitored plasma levels 1 hour after drug administration, when blood levels are near maximal (Table I I , 17). While the radioreceptor assay only measures parent drug and active metabolites, the total radioactivity in the sample will include inactive as well as active metabolites (Table I I ) . The much lower values obtained by radioreceptor assay indicate that receptor-active materials only represent 2.9°/° - 3.6°/° of the total nitrendipine metabolites. This is in agreement with other reports showing substantial metabolism of nifedipine (15). The major metabolite of nifedipine, and presumably nitrendipine, is the de-esterified analog which is pharmacologically inactive (29). Another important aspect of any^assay is sensitivity. Does the potency of the various calcium antagonists at JH-nitrendipine binding sites suffice to permit measurement of therapeutic blood levels of these drugs? Studies monitoring blood levels of nifedipine and verapamil by high-pressure liquid chromatography, reveal mean blood levels at therapeutic doses of 50 - 100 ng/ml of plasma (14,18). The three calcium antagonists currently
2670
Radioreceptor Assay of Calcium Blockers
I
100
._c
80
-
"\
"o .c
="
Vol. 33, No. 26, 1983
~6o
,
,
,
"B
zI
~
40
20
I 9
I 8
I 7
-Log
I 6
[ve~apam,I (M)
FIG. 2 Complete inhibition by verapamil of 3H-nitrendipine binding to skeletal muscle membranes. Total and nonspecific binding to skeletal muscle membranes were determined as described l~nMethods in the absence or presence of various concentrations of verapamil. The inset shows the same oata plotted as the logit transformation. TABLE II Comparison of Plasma Radioactivity Following 3H-Nitrendipine and Nitrendipine Plasma Levels Determined by Radioreceptor Assay (RRA)
Dose
(mg/kg)
3H-Nitrendipine ~M
Nitrendipine by RRA ~M
0.5 mg/kg
2.27 + 0.16
0.0074 + 0.011
1
mg/kg
4.91 + 0.49
0.144
+ 0.027
2
mg/kg
10.83 + 0.63
0.392
+ 0.055
1
1
1
Rats were administered either nitrendipine alone or nitrendipine plus 3H-nitrendipine at the indicated aosages. After i hr, blood was collected by cardiac puncture into heparin (0.1 ml, 1 mg/ml). Plasma nitrendipine concentration was determined by the radioreceptor assay and total radioactivity assayed by liquid s c i n t i l l a t i o n spectrometry.
Vol. 33, No. 26, 1983
Radioreceptor Assay of Calcium Blockers
2671
marketea in the United States are nifedipine, verapamil and diltiazem, whose EC50 values at 3H-nitrendipine sites are approximately 1.0, 10.0 and 1000 nM (0.3, 0.5, 40 nglml). The minimal concentration of a drug that can be reliably detected is a concentration that alters binding about 15-I °. For nifedipine, verapamil and diltiazem, this would correspond to 0.1, 1.0 and 100 nM. The assay described in the present study requires a lO0-fold dilution of plasma samples. For nifedipine, patients with blood levels of 10 - 100 nglml would give rise to final concentrations of 0.3 - 3 nM which would be readily detected. For verapamil these concentrations would be 3 - 30 nM, which can also be readily monitored. For diltiazem the assay presently employed may not suffice to measure therapeutic blood levels. Sensitivity of the present assay can be readily enhanced by a number of techniques. Plasma samples must be diluted lO0-fold in the assay to reduce inhibition of 3H-nitrendipine binding. Presumably such inhibition is due to plasma protein which binds directly to the H-nitrendipine, reducing the free 3H-nitrendipine concentration accessible to receptor sites. A similar limitation was identified in assays of beta adrenergic antagonist drugs (3). Dialysis of plasma samples by r e l a t i v e l y simple techniques that are readily applied to large numbers of samples abolishes the inhibition by plasma protein of 3H-dihydroalprenolol binding to beta receptors so that much larger patient samples can be employed (3). Such a modification could easily increase the sensitivity of the present assay up to 50-fold. The d i f f i c u l t y with u t i l i z i n g the dialysates is that one will only monitor the free drug which may be less than 10°I ° of the total plasma level of drug. Nonetheless, one virtue of measuring free drug levels is that for drugs such as beta antagonists, therapeutic response correlates more closely with free than with total plasma levels of drug (3). The other principal means of increasing sensitivity is to concentrate the drug by techniques such as solvent extraction. This is typically done in the high performance liquid chromatographic techniques for measuring calcium channel antagonist level
(14-18).
In summary, this radioreceptor assay offers a potential for ready application to large numbers of patients. The procedure is adequately sensitive for most calcium antagonist drugs u t i l i z i n g plasma samples directly. The assay is selective for calcium antagonist drugs. Of large numbers of drugs screened, the only ones which influence 3H-nitrendipine binGing are those which possess pharmacologically relevant calcium antagonist properties (19-27). Another virtue of the assay is that i t w i l l detect therapeutically active metabolites. For instance, after treatment with therapeutic doses of verapamil, plasma concentrations of norverapamil, an active metabolite, are similar to those of verapamil (14). Thus, simply measuring levels of the administered drug is not sufficient. The radioreceptor assay avoids the d i f f i c u l t y of not knowing what metabolites may be active. Norverapamil and verapamil may not have the same IC50 for 3H-nitrendipine binding sites. The measured concentration of norverapamil may, therefore, not r e f l e c t the actual concentration of norverapamil, but rather the concentration in "verapamil equivalents". That is, in terms of the amount of verapamil that would give the same amount of displaced binding. The measured concentration for verapamil, and indeed any of the calcium channel antagonists, is consequently a complex mixture of administered drug and active metabolites. I t need not necessarily reflect the actual concentration of each. The virtue of the assay is in detecting all these therapeutically active metabolites. The radioreceptor assay described here is simple to perform. As many as 200 assays can be readily conducted in a morning.
2672
Radioreceptor Assay of Calcium Blockers
Vol. 33, No. 26, 1983
Acknowledgements Supported by USPHSgrants MH-18501, NS-16375, RSA Award DA-O0074 to S.H.S., training grant GM-O73090 to K.M.M.M. and training grant MH-15330 to R.J .G. References 1. I. CREESE and S.H. SNYDER Nature 270 180-182 (1977). 2. L.E. TUNE, I. CREESE, J. COYLE, G. ~RLSON and S.H. SNYDER Am. J. Psychiatry 137 80-82 (1982). 3. R.B. INNIS, D.B. BYLUNDand S.H. SNYDER Life Sci. 23 2031-2038 (1978). 4. S.R. NAHORSKI, M.I. BATTA and D.B. BARNETT Eur. J. ~armacol. 52 393-396 (1978). 5. J.P. BILEZIKIAN, D.E. GAMMON, C.L. ROCHESTERand D.G. SHAND Clin. Pharmacol. Ther. 26, 173-180 (1979). 6. P. SKOLNICK, F.K. ~-~[ODWIN and S.M. PAUL Arch. Gen. Psychiat. 36 78-80 (1979). 7. P. HUNT, J.-M. HUSSONAND J.-P. ILAYMOND J. Pharm. Pharmacol. 31 448-451 (1979). 8. F. OWEN, R. LOFTHOUSEand R.C. BOURNE Clin. Chim. Acta. 93 305-310 (1979). 9. R.B. INNIS, L. TUNE, R. ROCK, R. DEPAULO, D.C. U'PRICHARDand S.H. SNYDER Eur. J. Pharmacol. 58 473-477 (1979). 10. A. FLECKENSTEIN Ann.'--Rev. Pharmacol. Toxicol. 17 149-166 (1977). 11. P.D. HENRY Am. J. Cardiol. 46 1047-1058 (1980)-~12. R.A. JANIS and D.J. TRIGGLE ~ M e d . Chem. 26, 775-785 (1983). 13. H.J. GELMERS Headache 23 106-109 (1983). 14. W. FRISHF~AN, E. KIRSTEIN,'--M. KLEIN, M. PINE, S.M. JOHNSON, L.D. HILLIS, M. PACKER and R. KATES Am. J. Cardiol. 50 1180-1184 (1982). 15. P. PIETTA, A. P~AUAand P. BIONDI J. C~omatog. 210 516-521 (1981). 16. T. SADANAGA, K. HIKIDA, K. TAMETO, Y. ~TSUSHIF~A and Y. OHKURA Chem. Pharm. Bull. 30 3807-3809 (1982). 17. R. TESTA, E. DOLFINI, C. ROSCHIOTTO, C. SECCHI and P.A. BIONDI II Farmaco Ed. Pr. 34 463473 (1979). 18. K. AOKI, ~ SATO, Y. K~AWAGUCHIand M. YAMAMOTO Eur. J. Clin. Pharmaco]. 23 197-201 (1982). 19. ~ J . GOULD, K.M.M. MURPHYand S.H. SNYDER Proc. Natl. Acad. Sci. USA 79 3656-3660 (1982). 20. K.M.M. MURPHY, R.J. GOULD, B.L. LARGENTand S.H. SNYDER Proc. Natl. Acad. Sci. USA 80 860464 (1983). 21. G.T. BOLGE~P. GENGO, R. KLOCKOWSKI, E. LUCHOWSKI, H. SIEGEL, R.A. JANIS, A.M. TRIGGLE and D.J. TRIGGLE J. Pharmacol. Exp. Ther. 225 291-309 (1983) 22. F.J. EHLERT, W.R. ROESKE, E. ITOGA and H.I. YAMAMURA Life Sci. 30 2191-2202 (1982). 23. L.T. WILLIAMS and P. TREMBLE J. Clin. Invest. 70 209-212 (1982). 24. P. BELLEFtANN, D. FERRY, F. LUBBACKEand H. GLOSSFtANN Arzneim.-Forsch./Drug Res. 31 2064-2067 (1981). 25. K.M.M. MURPHY, R.J. GOULDand--S.H. SNYDER in Nitrendipine (A. Scriabine, S. Vanor and K. Deck, eds.), Urban and Schwarzenberg, Baltimore-Munich, in press, (1983). 26. H.I. YAJ~A~IURA, H. SCHOE~LAKER,R.G. BOLES and W.R. ROESKE Biochem. Biophys. Res. Commun. 108 640~46 (1982). 27. A. DE POVER, M.A. MATLIB, S.W. LEE, G.P. DUBE, I.L. GRUPP, G. GRUPPand A. SCHWARTZ Biochem. Biophys. Res. Commun. 108 110-117 (1982). 28. R.A. JANIS, G.J. KROHL, A.J. NOE AND M. PAN J. Clin. Pharmacol., 2__3 266-273 (1983). 29. H. MEDENWALD, K. SCHLOSSF~ANNand C. WINSCHE Arznein.-Forsch./Drug Res. 22 53-56 (1972).