- Email: [email protected]
Photoaffinity labelling of a 33–35,000 dalton protein in cardiac, skeletal and smooth muscle membranes using a new 125I-labelled 1,4-dihydropyridine calcium channel antagonist
Life Sciences, Vol. 39, pp. 2401-2409 Printed in the U.S.A.
Pergamon Journals
PHOTOAFFINITY LABELLING OF A 33-35,000 DALTON PROTEIN IN CARDIAC, SKELETAL AND SMOOTHMUSCLEMEMBRANESUSING A NEW 125I-LABELLED 1,4-DIHYDROPYRIDINE CALCIUM CHANNELANTAGONIST 3.G. Sarmientol , P.M. Epstein2, W.A. Rowe2, D.W. Chesterl , H. Smilowitz 2, E. Wehinger3, and R.A. 3anis1,4 ]Department of Medicine, Division of Cardiology, and the 2Department of Pharmacology, University of Connecticut Health Center, Farmington, CT 06032, 3Chemical Research Laboratories, Pharma EP, Bayer AG, Wuppertal, West Germany, and 4Miles Institute for Preclinical Pharmacology, P.O. Box 1956, New Haven, CT 06509. (Received in final form September 16, 1986)
Summary The binding sites for Ca2+ channel antagonists were probed using Bay P 8857 [2-iodoethyl isopropyl 1,4-dihydropyridine-2,6dimethyl-4-(3-nitrophenyl)-pyridine-3,5-dicarboxylate] that has been radiolabelled with 1251. T h i s drug was shown to bind with high a f f i n i t y to cardiac, smooth, and skeletal muscle membranes, with a KD ~ 0.3 nM. A protein of molecular weight 33-35,000 daltons was specifically and irreversibly radiolabelled after irradiation of cardiac, skeletal and aortic smooth muscle membranes, incubated with the []25I]-Bay P 8857. The peptide labelled by 1,4-dihydropyridine binding therefore appears similar in size for cardiac, skeletal, and smooth muscle. T h i s data suggests that of the three peptide subunits which reportedly comprise the skeletal and cardiac muscle 1,4-dihydropyridine receptor complex, the 33-35,000 dalton peptide contains the dihydropyridine binding site. Bay P 8857 is a new Ca2+ channel antagonist that has been labelled with 1251 to provide an irreversible probe of h i g h specific activity for the 1,4-dihydropyridine binding sites. T h i s compound can be used not only for radioligand binding assays but also to identify the dihydropyridine binding site by photoaffinity labelling as we shall report in this communication. Campbell e t a ] . (l) have reported that [3H]nitrendipine binds irreversibly to a 32,000 Mr protein upon irradiation of the drug bound to isolated membranes from canine heart. Curtiss and Cattera]l (2), Borsotto et al. (3) and Rengasamy e t a ] . (4) recently purified the putative Ca2+ channel from skeletal muscle and heart and also found 33,000 Mr proteins, similar in molecular weight to that photolabelled by Campbell et al. In this report we describe the use of a novel iodinated 1,4-dihydropyridine Ca2+ channel blocking drug as an irreversible ligand for the dihydropyridine binding site. The results suggest that of the three peptide subunits which reportedly comprise the skeletal muscle and heart ],4-dihydropyridine receptor complex (2,3,4), the 33-35,000 dalton peptide contains the dihydropyridine binding site.
1Address a l l correspondence to 3.G. Sarmiento at B r i s t o l Myers Co., PRDD, P.O. Box 4755, Syracuse, NY 13221-4755. 0024-3205/86 $3.00 + .00 Copyright (c) 1986 Pergamon Journals Ltd.
2402
Photoaffinity Labeling of the Calcium Channel
Vol. 39, No. 25, 1986
Methods Calf hearts and bovine aortae were obtained from a local abattoir immediately after sacrifice and were placed in cold homogenization buffer (5); skeletal muscle was from adult chickens. P a r t i a l l y purified sarcolemmal membranes were prepared from heart muscle as described by 3ones et al. (6) and from aortic smooth muscle as described by Sarmiento et al. (7); transverse tubules from chicken skeletal muscle were prepared as described by Rosenblatt et al. (8). Binding assays were carried out in 2.5 ml volumes containing 0.1 - 0.2 mg membrane protein, between 20 and lO0 pM [1251]-Bay P 8857, 50 mM Tris-HCl, pH 7.4 at 24°C for 60 minutes. Nonspecific binding was defined as that which takes place in the presence of lO0 nM nisoldipine. Photoaffinity labelling of membrane proteins was done in the absence and presence of 50 ~M nicardipine which was used to define nonspecific binding. Membranes were incubated for l hour at 24°C in the presence or absence of 50 ~M nicardipine, irradiated with UV l i g h t From a Sperti lamp (425 watts) for lO minutes, then [1251]-Bay P 885l, 5 nM, was added and allowed to equilibrate for l hour, and irradiated again for lO minutes. In some experiments, the preirradiatlon was omitted, with no difference in the results. The reaction was terminated by the addition of an SDS stop solution containing 5 percent sodium dodecylsulfate, lO percent mercaptoethanol, 5 mM EDTA, and 0.5 percent bromphenol blue followed by heating of the mixture for 5 minutes at lO0°C. The solubilized samples were subjected to SDS-polyacrylamide gel electrophoresis as described by Laemmli et al. (g) on 12.5 percent or 5-15 percent gradient polyacrylamide gels. Autoradiography was done on dried gels using Kodak XAR-50 X-ray f i l m and developed on a Kodak automatic processor. Protein was measured u s i n g bovine serum albumin as standard (lO). [125I]-Bay P 8857 (2,200 Ci/mmole specific a c t i v i t y ) was custom radiolabelled by New England Nuclear, Boston, MA. Results The new iodinated Ca2÷ channel blocker Bay P 8857 was found to bind to cardiac membranes s p e c i f i c a l l y and with high a f f i n i t y as shown by the results of competition assays between [125I]-Bay P 8857 and nicardipine, nisoldipine or Bay P 8857 (Figures IA,B,C). The ICso values obtained from the figure were about 0.02-0.30 nM for these nonlabelled ligands. This suggests that this drug interacts at the 1,4-dihydropyridine binding site on cardiac membranes with an a f f i n i t y similar to that previously reported for nitrendipine (7,11,12) and nimodipine (13). The H i l l slope values from these data (Figures IA,B,C) were 1.22, o.gg and 1.23 for nicardipine, nisoldipine and Bay P 8857, respect i v e l y . In other separate experiments, these H i l l slopes ranged from 0.85 1.5, and there is no s t a t i s t i c a l evidence that these values are s i g n i f i c a n t l y different from one. Scatchard analysis of the binding (Figure 2) suggests further that this drug interacts with a homogeneous population of high a f f i n i t y binding sites with a Kd = 0.36 nM and a Bmax = 0.16 pmoles/mg membrane protein from cardiac tissue. Bay P 8857 like other 1,4-dihydropyridine Ca2÷ channel blocking drugs, is sensitive to l i g h t and w i l l decompose to reactive intermediates upon irradiation. Therefore in these experiments the drug was bound to membranes from cardiac, skeletal or smooth muscle, and the mixture was flashed with ultraviolet light. Irradiation resulted in the covalent linkage of the drug molecule to membrane proteins as shown by autoradiography and liquid s c i n t i l l a t i o n counting of solubilized membrane proteins.
qol. 39, No. 25, 1986
Photoaffinity Labeling of the Calcium Channel
99
A 80 50 20
Z Z
k
I
I
I
I
I
i
99
~m
--I 0
I-Z 0 ¢-~
eO 50 ZO
, 99 80 50
I,~
12
!
i
I
II
I0
9
-LOG
DRUG,
._
I
8
M
FIG. l Displacement of [1251]-Bay P 8857 ( 0 . 0 2 nM) f r o m calf heart sarcolemmal membranes by non-radiolabelled nicardipine (Panel A), nisoldipine (Panel B) and Bay P 8857 (Panel C). The percentage of specific binding i s plotted versus the concentration of nonlabelled competing ligand. Nonspecific binding was determined at each concentration of nonradioactive ]igand and was defined as the amount of binding i n the presence of 0,] ~M nisoldipine. In these experiments, the nonspecific binding ranged from 51-65% of the total binding, The ordinate, which l i s t s the percentage of control binding, ,represents a probit plot calculated according to the equation p = lO0/(L F eL), where p = percentage and L = In p/(lOO-p).
2403
2404
Photoaffinity Labeling of the Calcium Channel
Vol. 39, No. 25, 1986
0"03 I
-,"~"' 0 " 0 2 I
g m
0.011
0L
I
i
0'.02
I
0.03
BOUND
(pmole)
FIG. 2 Scatchard analysis of the binding of [125I]-Bay P 8857 to cardiac membranes. Data from the competition assays were analyzed by the BDATA program from EMF Software, Nashville, IN. The Kd was 0.36 nM and 8maX was 0.16 pmoles/mg membrane protein.
The absorption spectra for nisoldipine and Bay P 885/ (Figures 3A,8) are shown to demonstrate the photolytic decay these drugs undergo upon exposure to the light conditions of the photolabelling assay. The sensitivity of nisoldipine to light was indicated by significant shifts in the absorption maximas after brief exposure to light (Fig. 3B), while the decomposition of Bay P 8857 was much less even after 60 minutes exposure (Fig. 3A). Because of the extreme light sensitivity of nisoldipine relative to Bay P 885l, i t was unsuitable as a ligand to define nonspecific binding in experiments where membranes were exposed to light. Bay P 8857 could not be used to define specific incorporation because its aqueous s o l u b i l i t y is very low relative to other dihydropyridines such as nisoldipine or nimodipine. Therefore, nicardipine, a dihydropyridine which has a light sensitivity similar to that of Bay P 8857 and high aqueous s o l u b i l i t y was used to define specific incorporation. The concentration of [125I]-Bay P 885l used for the labelling experiments was 5 nN, a concentration about 14-fold greater than its Kd value. The low light sensitivity of Bay P 8857 required that the receptor occupancy be at or near maximum, which is the case for this concentration of drug. Hence to define specific binding upon light activation relatively high concentrations (5-50 ~M) of nicardipine were used.
Vol. 39, No. 25, 1986
Photoaffinity Labeling of the Calcium Channel
2405
A
Control ?.0 rain 60 min
Z
Z
t
I
0 Go
B~ ~L Orrl~rol
l
200
I
300
m
4OO WAVELENGTH (rim)
I
500
FIG. 3 The absorption spectra of 50 ~d4 Bay P 8857 (Panel A) and 50 n i s o l d i p i n e (Panel B) in 5 mR HEPES, pH 7.3, 25% ( d i s t i l l e d ) ethanol were recorded as a function of time of exposure at 20 cm and 4°C to broad band UV l i g h t . The absorption spectra were run from 200 to 500 nm (200-300 rim, f u l l scale = 2.0; 300-500 rim, f u l l s c a ] e = 0.5). The densitometric traces of the autoradiographs of photolabelled cardiac, smooth, and skeletal muscle membranes are shown in Figure 4. Several bends are observed on gels containing membranes incubated in the presence of [125I]-Bay P 8857, which indicates that the photoreactive species of the drug is capable of i n t e r a c t i n g with several membrane proteins during the course of the reaction. Membranes incubated with the radiolabelled drug plus an excess of unlabelled nicardipine which blocked s p e c i f i c binding (traces B, E, and G) show a s i g n i f i c a n t reduction p r i m a r i l y of a 33-35,000 Nr peak. Based on peak area, the reduction in l a b e l l i n g of the 33-35,000 Nr peak is 62%, 84%, and 69% in cardiac, smooth, and skeletal muscle membranes, respectively. In a o r t i c smooth muscle, some reduction in the l a b e l l i n g of 28,000 and 43,000 Nr bands is also seen (trace E). In some experiments, a b i f u r c a t i o n of the 33-35,000 Nr peak was observed upon scanning (traces B, F, and G); the reason f o r t h i s observed b i f u r c a t i o n is not known. Assays done with denatured membranes (trace C) showed few bands labelled by the drug; in binding assays with membranes denatured in the same manner no s p e c i f i c binding
2406
Photoaffinity
Labeling of the Calcium Channel
Vol. 39, No. 25, 1986
m
~
A B
k-
uL i
-..yc
i
I
|
J
J
(3 Z
94
67
43
30
203
m nO m m
&_o
UJ I-
I
i
*
I
I
I
i
*
94
67
43
30
20.1
t
I
130 93
I
43
,
,
J I
29 12
Mr (x aO"31 FIG. 4 Densitometric tracings of autoradiographs of polyacrylamide gels showing the radioactive peaks where [125I]-Bay P 8857 i r r e v e r sibly labelled the membrane proteins from calf heart (traces A-C), bovine aorta (traces D and E), and chicken skeletal muscle (traces F and G). Traces A, D and F represent samples of membranes incubated in the presence 5 nM [125I]-Bay P 8857 (total binding); traces B, E, and G are the same as A, D, and F except 50 ]~M nicardipine was added prior to incubation with [125I]-Bay P 8B5/ (nonspecific binding); in trace C, membranes were heat denatured at lO0°C for 2.5 min prior to incubation with [125I]-Bay P 8857. was observed. The Coomassie protein staining pattern of t h e gels, in the presence and absence of nicardipine, were identical and the specifically labelled band could not be detected visually. This is seen in Figure 5, where the Coomassie staining and autoradiographic results of a photolabelling experiment carried out with p a r t i a l l y purified transverse tubules prepared from chicken skeletal muscle (B pmol dihydropyridine binding sites/mg protein) are shown. I t can be seen d i r e c t l y from Figure 5 that the major labelled band displaced by nicardipine migrates at = 33-35,000 Mr , although other minor labelled bands at 26,000 Mr and 60,000 Mr show slight displacement as well.
Vol. 39, No. 25, 1986
Photoaffinity Labeling of the Calcium Channel
A
2407
B NONSPECIFIC TOTAL
._
alp
U
W
D
,.-
130K
In
66.3K
Q
43K
4--
2&7K
~
11.7K
--
m
1
NONSPECIFIC TOTAL
2
6
U
3
4
FIG. 5 Gradient polyacrylamide gel of skeletal muscle membranes photoa f f i n i t y labelled by [125I]-Bay P 8857. Panel A) Coomassie blue stained gel, lane l , membranes incubated with [1251]-Bay P 885/ plus nonlabelled nicardipine (nonspecific), lane 2, membranes incubated with [125I]-Bay P B85l only ( t o t a l ) . Panel B) autoradiograph of the stained gel in panel A. Discussion
The s p e c i f i c a l l y labelled protein, detected by the [125I]-Bay P 8857, is similar in molecular weight to that observed by Campbell (1) using [3H]nitrendipine; however, the iodinated Bay P 8857 compound offers several advantages to the t r i t i a t e d dihydropyridlnes. I t is labelled to higher specific a c t i v i t y (2,200 Ci/mmole) and the iodine label f a c i l i t a t e s autoradiographic analysis. Venter et a l . , using an isothiocyanate irreversible, dihydropyridine Ca2+ channel antagonist, ortho-NCS, reported the specific labelling of a protein of Mr 45,000 in cardiac membranes (13). Although we do see a small amount of specific labelling by [125I]-Bay P 885? of a protein of Mr 43,000 in aortic smooth muscle membranes (Figure 4, traces D and E), the most predominant specific labelling in cardiac and skeletal muscle membranes, as well as in aortic smooth muscle membranes is to a protein of Mr 33-35,000 (Figures 4 and 5).
2408
Photoaffinity
Labeling of the Calcium Channel
Vol. 39, No. 25, 1986
A recent report by Curtiss and Catterall (2) indicated that the solubilized and purified Ca2+ channel complex From skeletal muscle was composed of three subunits, having molecular weights of 33,000, 50,000, and 130,000 Mr. Similarly, Borsotto et al. (3) and Rengasamy et al. (4) have observed a 33-34,000 dalton peptide subunit in t h e i r purified Ca2+ channel preparations from skeletal and cardiac tissue, respectively. Our photolabelling experiments suggest that of the peptide subunits which comprise the multimeric Ca2+ channel from skeletal and heart muscle, the 33-35,000 Mr peptide is the 1,4-dihydropyridine drug binding site. These results suggest further that this site is similar in cardiac, skeletal, and smooth muscle membranes. We have previously reported that the binding characteristics of several 1,4-dihydropyridines are the same for cardiac and smooth muscle membranes (5), although i t is known that there is a lO-fold lower a f f i n i t y , as well as other differences, in the binding of dihydropyridines to skeletal muscle as compared to these other tissues (15,16,17). In contrast to smooth muscle, where a close correlation exists between the a f f i n i t y of binding of dihydropyridines and t h e i r pharmacological efficacy, i n i t i a l experiments on dihydropyridine binding to cardiac membranes showed a poor correlation between these parameters, suggesting that the high a f f i n i t y dihydropyridine binding site in these two tissues may d i f f e r (5). I t appears now, however, that this discrepancy in cardiac muscle can be accounted for by the voltage dependence of binding (18,19). Nevertheless, in skeletal muscle, the characteristics of the binding of dihydropyridines clearly d i f f e r from that of other tissues, and differences in the dihydropyridine receptor could s t i l l exist between cardiac and smooth muscle. Hence, using a sensitive photoaffinity labelling technique employing a highly labelled iodinated dihydropyridine probe, we compared the dihydropyridine receptor in cardiac, skeletal, and smooth muscle membranes. Based on our results shown here, the dihydropyridine binding site appears to be similar in a l l three of these tissues. Although these studies suggest a similar dihydro~yridine binding site for these tissues, regulation or modulation of the CaZ+ ~,annels ~L in these tissues may nevertheless d i f f e r . Horne et al. (20) have recently reported that nitrendipine stimulated the endogenous phosphorylation of a 42,000 Mr protein that comigrates with the specific incorporation of the [3H]o-NCS compound in cardiac membranes. Whether the Ca2+ channel is regulated by a phosphorylation of one of the channel complex subunits remains to be established, but the differences in s e n s i t i v i t y that these drugs display between skeletal muscle and other tissues may be related to some biochemical modification of the channel of this type. Note Added Subsequent to the submission of this manuscript, a paper appeared by Galizzi et al. (21) in which the dihydropyridine, (÷)-[3H] PN 200-II0 was shown to photoaffinity label a protein of 170,000 Mr in rabbit skeletal muscle T-ruble membranes under non-reducing conditions. However, under reducing conditions, the 170,000 Mr protein migrates as two peptides of 140,000 Mr and 33,000 Mr (22 and H. Smilowitz, unpublished observations). Hence, i t is possible that the 33-35,000 Mr peptide labelled by [125I]-Bay P 8857 is a component of a 170,000 Mr complex which is released under reducing conditions. Acknowledgments This work was supported in part by grants HL-21812, HL-22135 and HL-33026, from the National Institutes of Health and grants to P.M.E. from the University of Connecticut Research Foundation and Miles Laboratories. The authors would l i k e to thank C.A. Kostecki for her expert technical assistance and Dr. A.M. Katz for his helpful discussions during the preparation of this paper.
Vol. 39, No. 25, 1986
Photoaffinity Labeling of the Calcium Channel
2409
References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. IB. 19. 20. 21. 22.
K.P. CAMPBELL, G.M. LIPSHUTZ and G . H . DENNEY, 3. B i o l . Chem. 259 5384-5387 (1984). B.N. CURTISS and W.A. CATTERALL, Biochemistry 23 2113-2118 (1984). M. BORSOTTO, J. BARHANIN, R.I. NORMAN and M. LAZOUNSKI, Biochem. 8iophys. Res. Commun. 122 1357-1366 (1984). A. RENGASAMY, 3. PTASIENSKI and M. HOSEY, Biochem. Biophys. Res. Commun. 126 1-7 (1985). J.G. SARMIENTO, R.A. JANIS, A . M . KATZ and O.J. TRIGGLE, Biochem. Pharmacol. 33 3119-3123 (1984). L.R. JONES, S.W. MADDOCH and H.R. 8ESCH, 3. Biol. Chem. 255 997]-9980 (1980). 3.G. SARMIENTO, R.A. OANIS, D. 3ENKINS and D.J. TRIGGLE, in Nitrendipine, Eds. A. Scriabine, S. Vanov and K. Deck, pp. 153-160 Urban-Schwarzenberg Baltimore NO (1984). M. ROSENBLATT, C. HIDALGO, C. VERGARA and N. IKEMOTO, J. 8 i o l . Chem. 256 8140-8148 (1981). U.K. LAEMMLI, Nature 227 580-685 (1970). M.M. BRADFORD, Anal. 8iochem. 72 248-253 (1976). R.A. JANIS, 3.G. SARMIENTO, S.C. MAURER, G.T. 80LGER and O.J. TRIGGLE, 3. Phamacol. Exp. Ther. 231 8-15 (1984). H.A. CAMERON, P . M . EPSTEIN, 3.3. LAMBERT and G . A . LYLES, 8 r i t . 3. Pharmacol. 83 444P (1984). R.A. 3ANIS, S.C. MAURER, J.G. SARMIENTO, G.T. BOLGER and 0.3. TRIGGLE, Eur. 3. Pharmacol. 82 191-194 (1982). 3.C. VENTER, C.M. FRASER, J.S. SCHABER, C.Y. JUNG, G. BOLGER and O.J. TRIGGLE, J. 8 i o l . Chem. 258 9344-9348 (1984). M. FOSSET, E. JAIMOVICH, E. DELPONT, and M. LAZDUNSKI, Eur. J. Pharmacol. 86 141-142 (1983). H. GLOSSMAN, D.R. FERRY, A. GOLL, and M. ROMBUSCH, J. Card. Pharmacol. 6 5608-5621 (1984). R . J . GOULD, K.M.M. MURPHY, and S.H. SNYDER, Mol. Pharmacol. 25 235-241 (1984). M.C. SANGUINETTI and R.S. KASS, Circ. Res. 55 336-348 (1984). B.P. BEAN, Proc. Natl. Acad. Sci. U.S.A. 81 6388-6392 (1984). P. HORNE, D . J . TRIGGLE and J.C. VENTER, Biochem. Biophys. Res. Commun. 12l 890-898 (1984). J.-P. GALIZZI, M. BORSOTTO, J. BARHANIN, M. FOSSET, and M. LAZOUNSKI, J. Biol. Chem. 26l 1393-1397 (1986). M.M. HOSEY, M. BORSOTTO, and M. LAZDUNSKI, Proc. Natl. Acad. Sci. U.S.A. 83 3733-3737 (1986).
Pergamon Journals
PHOTOAFFINITY LABELLING OF A 33-35,000 DALTON PROTEIN IN CARDIAC, SKELETAL AND SMOOTHMUSCLEMEMBRANESUSING A NEW 125I-LABELLED 1,4-DIHYDROPYRIDINE CALCIUM CHANNELANTAGONIST 3.G. Sarmientol , P.M. Epstein2, W.A. Rowe2, D.W. Chesterl , H. Smilowitz 2, E. Wehinger3, and R.A. 3anis1,4 ]Department of Medicine, Division of Cardiology, and the 2Department of Pharmacology, University of Connecticut Health Center, Farmington, CT 06032, 3Chemical Research Laboratories, Pharma EP, Bayer AG, Wuppertal, West Germany, and 4Miles Institute for Preclinical Pharmacology, P.O. Box 1956, New Haven, CT 06509. (Received in final form September 16, 1986)
Summary The binding sites for Ca2+ channel antagonists were probed using Bay P 8857 [2-iodoethyl isopropyl 1,4-dihydropyridine-2,6dimethyl-4-(3-nitrophenyl)-pyridine-3,5-dicarboxylate] that has been radiolabelled with 1251. T h i s drug was shown to bind with high a f f i n i t y to cardiac, smooth, and skeletal muscle membranes, with a KD ~ 0.3 nM. A protein of molecular weight 33-35,000 daltons was specifically and irreversibly radiolabelled after irradiation of cardiac, skeletal and aortic smooth muscle membranes, incubated with the []25I]-Bay P 8857. The peptide labelled by 1,4-dihydropyridine binding therefore appears similar in size for cardiac, skeletal, and smooth muscle. T h i s data suggests that of the three peptide subunits which reportedly comprise the skeletal and cardiac muscle 1,4-dihydropyridine receptor complex, the 33-35,000 dalton peptide contains the dihydropyridine binding site. Bay P 8857 is a new Ca2+ channel antagonist that has been labelled with 1251 to provide an irreversible probe of h i g h specific activity for the 1,4-dihydropyridine binding sites. T h i s compound can be used not only for radioligand binding assays but also to identify the dihydropyridine binding site by photoaffinity labelling as we shall report in this communication. Campbell e t a ] . (l) have reported that [3H]nitrendipine binds irreversibly to a 32,000 Mr protein upon irradiation of the drug bound to isolated membranes from canine heart. Curtiss and Cattera]l (2), Borsotto et al. (3) and Rengasamy e t a ] . (4) recently purified the putative Ca2+ channel from skeletal muscle and heart and also found 33,000 Mr proteins, similar in molecular weight to that photolabelled by Campbell et al. In this report we describe the use of a novel iodinated 1,4-dihydropyridine Ca2+ channel blocking drug as an irreversible ligand for the dihydropyridine binding site. The results suggest that of the three peptide subunits which reportedly comprise the skeletal muscle and heart ],4-dihydropyridine receptor complex (2,3,4), the 33-35,000 dalton peptide contains the dihydropyridine binding site.
1Address a l l correspondence to 3.G. Sarmiento at B r i s t o l Myers Co., PRDD, P.O. Box 4755, Syracuse, NY 13221-4755. 0024-3205/86 $3.00 + .00 Copyright (c) 1986 Pergamon Journals Ltd.
2402
Photoaffinity Labeling of the Calcium Channel
Vol. 39, No. 25, 1986
Methods Calf hearts and bovine aortae were obtained from a local abattoir immediately after sacrifice and were placed in cold homogenization buffer (5); skeletal muscle was from adult chickens. P a r t i a l l y purified sarcolemmal membranes were prepared from heart muscle as described by 3ones et al. (6) and from aortic smooth muscle as described by Sarmiento et al. (7); transverse tubules from chicken skeletal muscle were prepared as described by Rosenblatt et al. (8). Binding assays were carried out in 2.5 ml volumes containing 0.1 - 0.2 mg membrane protein, between 20 and lO0 pM [1251]-Bay P 8857, 50 mM Tris-HCl, pH 7.4 at 24°C for 60 minutes. Nonspecific binding was defined as that which takes place in the presence of lO0 nM nisoldipine. Photoaffinity labelling of membrane proteins was done in the absence and presence of 50 ~M nicardipine which was used to define nonspecific binding. Membranes were incubated for l hour at 24°C in the presence or absence of 50 ~M nicardipine, irradiated with UV l i g h t From a Sperti lamp (425 watts) for lO minutes, then [1251]-Bay P 885l, 5 nM, was added and allowed to equilibrate for l hour, and irradiated again for lO minutes. In some experiments, the preirradiatlon was omitted, with no difference in the results. The reaction was terminated by the addition of an SDS stop solution containing 5 percent sodium dodecylsulfate, lO percent mercaptoethanol, 5 mM EDTA, and 0.5 percent bromphenol blue followed by heating of the mixture for 5 minutes at lO0°C. The solubilized samples were subjected to SDS-polyacrylamide gel electrophoresis as described by Laemmli et al. (g) on 12.5 percent or 5-15 percent gradient polyacrylamide gels. Autoradiography was done on dried gels using Kodak XAR-50 X-ray f i l m and developed on a Kodak automatic processor. Protein was measured u s i n g bovine serum albumin as standard (lO). [125I]-Bay P 8857 (2,200 Ci/mmole specific a c t i v i t y ) was custom radiolabelled by New England Nuclear, Boston, MA. Results The new iodinated Ca2÷ channel blocker Bay P 8857 was found to bind to cardiac membranes s p e c i f i c a l l y and with high a f f i n i t y as shown by the results of competition assays between [125I]-Bay P 8857 and nicardipine, nisoldipine or Bay P 8857 (Figures IA,B,C). The ICso values obtained from the figure were about 0.02-0.30 nM for these nonlabelled ligands. This suggests that this drug interacts at the 1,4-dihydropyridine binding site on cardiac membranes with an a f f i n i t y similar to that previously reported for nitrendipine (7,11,12) and nimodipine (13). The H i l l slope values from these data (Figures IA,B,C) were 1.22, o.gg and 1.23 for nicardipine, nisoldipine and Bay P 8857, respect i v e l y . In other separate experiments, these H i l l slopes ranged from 0.85 1.5, and there is no s t a t i s t i c a l evidence that these values are s i g n i f i c a n t l y different from one. Scatchard analysis of the binding (Figure 2) suggests further that this drug interacts with a homogeneous population of high a f f i n i t y binding sites with a Kd = 0.36 nM and a Bmax = 0.16 pmoles/mg membrane protein from cardiac tissue. Bay P 8857 like other 1,4-dihydropyridine Ca2÷ channel blocking drugs, is sensitive to l i g h t and w i l l decompose to reactive intermediates upon irradiation. Therefore in these experiments the drug was bound to membranes from cardiac, skeletal or smooth muscle, and the mixture was flashed with ultraviolet light. Irradiation resulted in the covalent linkage of the drug molecule to membrane proteins as shown by autoradiography and liquid s c i n t i l l a t i o n counting of solubilized membrane proteins.
qol. 39, No. 25, 1986
Photoaffinity Labeling of the Calcium Channel
99
A 80 50 20
Z Z
k
I
I
I
I
I
i
99
~m
--I 0
I-Z 0 ¢-~
eO 50 ZO
, 99 80 50
I,~
12
!
i
I
II
I0
9
-LOG
DRUG,
._
I
8
M
FIG. l Displacement of [1251]-Bay P 8857 ( 0 . 0 2 nM) f r o m calf heart sarcolemmal membranes by non-radiolabelled nicardipine (Panel A), nisoldipine (Panel B) and Bay P 8857 (Panel C). The percentage of specific binding i s plotted versus the concentration of nonlabelled competing ligand. Nonspecific binding was determined at each concentration of nonradioactive ]igand and was defined as the amount of binding i n the presence of 0,] ~M nisoldipine. In these experiments, the nonspecific binding ranged from 51-65% of the total binding, The ordinate, which l i s t s the percentage of control binding, ,represents a probit plot calculated according to the equation p = lO0/(L F eL), where p = percentage and L = In p/(lOO-p).
2403
2404
Photoaffinity Labeling of the Calcium Channel
Vol. 39, No. 25, 1986
0"03 I
-,"~"' 0 " 0 2 I
g m
0.011
0L
I
i
0'.02
I
0.03
BOUND
(pmole)
FIG. 2 Scatchard analysis of the binding of [125I]-Bay P 8857 to cardiac membranes. Data from the competition assays were analyzed by the BDATA program from EMF Software, Nashville, IN. The Kd was 0.36 nM and 8maX was 0.16 pmoles/mg membrane protein.
The absorption spectra for nisoldipine and Bay P 885/ (Figures 3A,8) are shown to demonstrate the photolytic decay these drugs undergo upon exposure to the light conditions of the photolabelling assay. The sensitivity of nisoldipine to light was indicated by significant shifts in the absorption maximas after brief exposure to light (Fig. 3B), while the decomposition of Bay P 8857 was much less even after 60 minutes exposure (Fig. 3A). Because of the extreme light sensitivity of nisoldipine relative to Bay P 885l, i t was unsuitable as a ligand to define nonspecific binding in experiments where membranes were exposed to light. Bay P 8857 could not be used to define specific incorporation because its aqueous s o l u b i l i t y is very low relative to other dihydropyridines such as nisoldipine or nimodipine. Therefore, nicardipine, a dihydropyridine which has a light sensitivity similar to that of Bay P 8857 and high aqueous s o l u b i l i t y was used to define specific incorporation. The concentration of [125I]-Bay P 885l used for the labelling experiments was 5 nN, a concentration about 14-fold greater than its Kd value. The low light sensitivity of Bay P 8857 required that the receptor occupancy be at or near maximum, which is the case for this concentration of drug. Hence to define specific binding upon light activation relatively high concentrations (5-50 ~M) of nicardipine were used.
Vol. 39, No. 25, 1986
Photoaffinity Labeling of the Calcium Channel
2405
A
Control ?.0 rain 60 min
Z
Z
t
I
0 Go
B~ ~L Orrl~rol
l
200
I
300
m
4OO WAVELENGTH (rim)
I
500
FIG. 3 The absorption spectra of 50 ~d4 Bay P 8857 (Panel A) and 50 n i s o l d i p i n e (Panel B) in 5 mR HEPES, pH 7.3, 25% ( d i s t i l l e d ) ethanol were recorded as a function of time of exposure at 20 cm and 4°C to broad band UV l i g h t . The absorption spectra were run from 200 to 500 nm (200-300 rim, f u l l scale = 2.0; 300-500 rim, f u l l s c a ] e = 0.5). The densitometric traces of the autoradiographs of photolabelled cardiac, smooth, and skeletal muscle membranes are shown in Figure 4. Several bends are observed on gels containing membranes incubated in the presence of [125I]-Bay P 8857, which indicates that the photoreactive species of the drug is capable of i n t e r a c t i n g with several membrane proteins during the course of the reaction. Membranes incubated with the radiolabelled drug plus an excess of unlabelled nicardipine which blocked s p e c i f i c binding (traces B, E, and G) show a s i g n i f i c a n t reduction p r i m a r i l y of a 33-35,000 Nr peak. Based on peak area, the reduction in l a b e l l i n g of the 33-35,000 Nr peak is 62%, 84%, and 69% in cardiac, smooth, and skeletal muscle membranes, respectively. In a o r t i c smooth muscle, some reduction in the l a b e l l i n g of 28,000 and 43,000 Nr bands is also seen (trace E). In some experiments, a b i f u r c a t i o n of the 33-35,000 Nr peak was observed upon scanning (traces B, F, and G); the reason f o r t h i s observed b i f u r c a t i o n is not known. Assays done with denatured membranes (trace C) showed few bands labelled by the drug; in binding assays with membranes denatured in the same manner no s p e c i f i c binding
2406
Photoaffinity
Labeling of the Calcium Channel
Vol. 39, No. 25, 1986
m
~
A B
k-
uL i
-..yc
i
I
|
J
J
(3 Z
94
67
43
30
203
m nO m m
&_o
UJ I-
I
i
*
I
I
I
i
*
94
67
43
30
20.1
t
I
130 93
I
43
,
,
J I
29 12
Mr (x aO"31 FIG. 4 Densitometric tracings of autoradiographs of polyacrylamide gels showing the radioactive peaks where [125I]-Bay P 8857 i r r e v e r sibly labelled the membrane proteins from calf heart (traces A-C), bovine aorta (traces D and E), and chicken skeletal muscle (traces F and G). Traces A, D and F represent samples of membranes incubated in the presence 5 nM [125I]-Bay P 8857 (total binding); traces B, E, and G are the same as A, D, and F except 50 ]~M nicardipine was added prior to incubation with [125I]-Bay P 8B5/ (nonspecific binding); in trace C, membranes were heat denatured at lO0°C for 2.5 min prior to incubation with [125I]-Bay P 8857. was observed. The Coomassie protein staining pattern of t h e gels, in the presence and absence of nicardipine, were identical and the specifically labelled band could not be detected visually. This is seen in Figure 5, where the Coomassie staining and autoradiographic results of a photolabelling experiment carried out with p a r t i a l l y purified transverse tubules prepared from chicken skeletal muscle (B pmol dihydropyridine binding sites/mg protein) are shown. I t can be seen d i r e c t l y from Figure 5 that the major labelled band displaced by nicardipine migrates at = 33-35,000 Mr , although other minor labelled bands at 26,000 Mr and 60,000 Mr show slight displacement as well.
Vol. 39, No. 25, 1986
Photoaffinity Labeling of the Calcium Channel
A
2407
B NONSPECIFIC TOTAL
._
alp
U
W
D
,.-
130K
In
66.3K
Q
43K
4--
2&7K
~
11.7K
--
m
1
NONSPECIFIC TOTAL
2
6
U
3
4
FIG. 5 Gradient polyacrylamide gel of skeletal muscle membranes photoa f f i n i t y labelled by [125I]-Bay P 8857. Panel A) Coomassie blue stained gel, lane l , membranes incubated with [1251]-Bay P 885/ plus nonlabelled nicardipine (nonspecific), lane 2, membranes incubated with [125I]-Bay P B85l only ( t o t a l ) . Panel B) autoradiograph of the stained gel in panel A. Discussion
The s p e c i f i c a l l y labelled protein, detected by the [125I]-Bay P 8857, is similar in molecular weight to that observed by Campbell (1) using [3H]nitrendipine; however, the iodinated Bay P 8857 compound offers several advantages to the t r i t i a t e d dihydropyridlnes. I t is labelled to higher specific a c t i v i t y (2,200 Ci/mmole) and the iodine label f a c i l i t a t e s autoradiographic analysis. Venter et a l . , using an isothiocyanate irreversible, dihydropyridine Ca2+ channel antagonist, ortho-NCS, reported the specific labelling of a protein of Mr 45,000 in cardiac membranes (13). Although we do see a small amount of specific labelling by [125I]-Bay P 885? of a protein of Mr 43,000 in aortic smooth muscle membranes (Figure 4, traces D and E), the most predominant specific labelling in cardiac and skeletal muscle membranes, as well as in aortic smooth muscle membranes is to a protein of Mr 33-35,000 (Figures 4 and 5).
2408
Photoaffinity
Labeling of the Calcium Channel
Vol. 39, No. 25, 1986
A recent report by Curtiss and Catterall (2) indicated that the solubilized and purified Ca2+ channel complex From skeletal muscle was composed of three subunits, having molecular weights of 33,000, 50,000, and 130,000 Mr. Similarly, Borsotto et al. (3) and Rengasamy et al. (4) have observed a 33-34,000 dalton peptide subunit in t h e i r purified Ca2+ channel preparations from skeletal and cardiac tissue, respectively. Our photolabelling experiments suggest that of the peptide subunits which comprise the multimeric Ca2+ channel from skeletal and heart muscle, the 33-35,000 Mr peptide is the 1,4-dihydropyridine drug binding site. These results suggest further that this site is similar in cardiac, skeletal, and smooth muscle membranes. We have previously reported that the binding characteristics of several 1,4-dihydropyridines are the same for cardiac and smooth muscle membranes (5), although i t is known that there is a lO-fold lower a f f i n i t y , as well as other differences, in the binding of dihydropyridines to skeletal muscle as compared to these other tissues (15,16,17). In contrast to smooth muscle, where a close correlation exists between the a f f i n i t y of binding of dihydropyridines and t h e i r pharmacological efficacy, i n i t i a l experiments on dihydropyridine binding to cardiac membranes showed a poor correlation between these parameters, suggesting that the high a f f i n i t y dihydropyridine binding site in these two tissues may d i f f e r (5). I t appears now, however, that this discrepancy in cardiac muscle can be accounted for by the voltage dependence of binding (18,19). Nevertheless, in skeletal muscle, the characteristics of the binding of dihydropyridines clearly d i f f e r from that of other tissues, and differences in the dihydropyridine receptor could s t i l l exist between cardiac and smooth muscle. Hence, using a sensitive photoaffinity labelling technique employing a highly labelled iodinated dihydropyridine probe, we compared the dihydropyridine receptor in cardiac, skeletal, and smooth muscle membranes. Based on our results shown here, the dihydropyridine binding site appears to be similar in a l l three of these tissues. Although these studies suggest a similar dihydro~yridine binding site for these tissues, regulation or modulation of the CaZ+ ~,annels ~L in these tissues may nevertheless d i f f e r . Horne et al. (20) have recently reported that nitrendipine stimulated the endogenous phosphorylation of a 42,000 Mr protein that comigrates with the specific incorporation of the [3H]o-NCS compound in cardiac membranes. Whether the Ca2+ channel is regulated by a phosphorylation of one of the channel complex subunits remains to be established, but the differences in s e n s i t i v i t y that these drugs display between skeletal muscle and other tissues may be related to some biochemical modification of the channel of this type. Note Added Subsequent to the submission of this manuscript, a paper appeared by Galizzi et al. (21) in which the dihydropyridine, (÷)-[3H] PN 200-II0 was shown to photoaffinity label a protein of 170,000 Mr in rabbit skeletal muscle T-ruble membranes under non-reducing conditions. However, under reducing conditions, the 170,000 Mr protein migrates as two peptides of 140,000 Mr and 33,000 Mr (22 and H. Smilowitz, unpublished observations). Hence, i t is possible that the 33-35,000 Mr peptide labelled by [125I]-Bay P 8857 is a component of a 170,000 Mr complex which is released under reducing conditions. Acknowledgments This work was supported in part by grants HL-21812, HL-22135 and HL-33026, from the National Institutes of Health and grants to P.M.E. from the University of Connecticut Research Foundation and Miles Laboratories. The authors would l i k e to thank C.A. Kostecki for her expert technical assistance and Dr. A.M. Katz for his helpful discussions during the preparation of this paper.
Vol. 39, No. 25, 1986
Photoaffinity Labeling of the Calcium Channel
2409
References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. IB. 19. 20. 21. 22.
K.P. CAMPBELL, G.M. LIPSHUTZ and G . H . DENNEY, 3. B i o l . Chem. 259 5384-5387 (1984). B.N. CURTISS and W.A. CATTERALL, Biochemistry 23 2113-2118 (1984). M. BORSOTTO, J. BARHANIN, R.I. NORMAN and M. LAZOUNSKI, Biochem. 8iophys. Res. Commun. 122 1357-1366 (1984). A. RENGASAMY, 3. PTASIENSKI and M. HOSEY, Biochem. Biophys. Res. Commun. 126 1-7 (1985). J.G. SARMIENTO, R.A. JANIS, A . M . KATZ and O.J. TRIGGLE, Biochem. Pharmacol. 33 3119-3123 (1984). L.R. JONES, S.W. MADDOCH and H.R. 8ESCH, 3. Biol. Chem. 255 997]-9980 (1980). 3.G. SARMIENTO, R.A. OANIS, D. 3ENKINS and D.J. TRIGGLE, in Nitrendipine, Eds. A. Scriabine, S. Vanov and K. Deck, pp. 153-160 Urban-Schwarzenberg Baltimore NO (1984). M. ROSENBLATT, C. HIDALGO, C. VERGARA and N. IKEMOTO, J. 8 i o l . Chem. 256 8140-8148 (1981). U.K. LAEMMLI, Nature 227 580-685 (1970). M.M. BRADFORD, Anal. 8iochem. 72 248-253 (1976). R.A. JANIS, 3.G. SARMIENTO, S.C. MAURER, G.T. 80LGER and O.J. TRIGGLE, 3. Phamacol. Exp. Ther. 231 8-15 (1984). H.A. CAMERON, P . M . EPSTEIN, 3.3. LAMBERT and G . A . LYLES, 8 r i t . 3. Pharmacol. 83 444P (1984). R.A. 3ANIS, S.C. MAURER, J.G. SARMIENTO, G.T. BOLGER and 0.3. TRIGGLE, Eur. 3. Pharmacol. 82 191-194 (1982). 3.C. VENTER, C.M. FRASER, J.S. SCHABER, C.Y. JUNG, G. BOLGER and O.J. TRIGGLE, J. 8 i o l . Chem. 258 9344-9348 (1984). M. FOSSET, E. JAIMOVICH, E. DELPONT, and M. LAZDUNSKI, Eur. J. Pharmacol. 86 141-142 (1983). H. GLOSSMAN, D.R. FERRY, A. GOLL, and M. ROMBUSCH, J. Card. Pharmacol. 6 5608-5621 (1984). R . J . GOULD, K.M.M. MURPHY, and S.H. SNYDER, Mol. Pharmacol. 25 235-241 (1984). M.C. SANGUINETTI and R.S. KASS, Circ. Res. 55 336-348 (1984). B.P. BEAN, Proc. Natl. Acad. Sci. U.S.A. 81 6388-6392 (1984). P. HORNE, D . J . TRIGGLE and J.C. VENTER, Biochem. Biophys. Res. Commun. 12l 890-898 (1984). J.-P. GALIZZI, M. BORSOTTO, J. BARHANIN, M. FOSSET, and M. LAZOUNSKI, J. Biol. Chem. 26l 1393-1397 (1986). M.M. HOSEY, M. BORSOTTO, and M. LAZDUNSKI, Proc. Natl. Acad. Sci. U.S.A. 83 3733-3737 (1986).