- Email: [email protected]
Evidence for the presence of phospholamban in the endoplasmic reticulum of smooth muscle
258
Biochimica et Biopt~vsica Aeta 882 (1986) 258-265 Elsevier
BBA 22349
E v i d e n c e for the p r e s e n c e of p h o s p h o l a m b a n in the e n d o p l a s m i c reticulum of s m o o t h m u s c l e Luc Raeymaekers
a a n d L a r r y R. Jones b
" Laboratorium t,oor F)'siologie, University of Leuven, Campus Gasthuisberg, B-3000 Leuven (Belgium) and I, The Krannert Institute of Cardiolog},. Indiana Universi(v School of Medicine, Indianapolis. I N 46202 (U.S.A.)
(Received February 4th, 1986)
Key words: Phospholamban; Protein phosphorylation; (Smooth muscle)
In microsomal vesicles isolated from several smooth muscles many polypeptides were phosphorylated by the catalytic subunit of cyclic AMP-dependent protein kinase. In pig stomach and in rabbit and dog aorta components of M r 22 000 and 11000 were identified as forms of phospholamban. These polypeptides were, however, not observed in pig aorta. These phospholamban-like polypeptides presented the same electrophoretic mobility in sodium dodecyl sulphate gels as cardiac phospholamban, and the 22 000 M r form showed a similar reaction to heat treatment in sodium dodecyl sulphate. Antibodies against purified canine cardiac phospholamban cross-reacted with the 22000 and 11000 Mr phosphorylatable polypeptides from smooth muscle membranes. Subcellular fractionation of porcine stomach smooth muscle indicated that phospholamban was present in the membranes of the endoplasmic reticulum and not in the plasma membranes. Phospholamban was also phosphorylated by an endogenous calcium-caimodulin-dependent protein kinase and by an endogenous cyclic AMP-dependent kinase. It is concluded that the endoplasmic reticulum of many, but possibly not all, smooth muscles contains phospholamban. However, the physiological role of phospholamban in smooth muscle remains to be established.
Introduction Although the endoplasmic (sarcoplasmic) reticulum of smooth muscle cells is less developed than that of skeletal and cardiac muscle, a large body of evidence shows that it may play an important role in the regulation of the cellular calcium metabolism [1-4]. The study of Ca 2+ transport in isolated endoplasmic reticulum vesicles has been hampered because of the difficulty of isolating it from the large amount of plasma membrane, which moreover also presents ATP-dependent Ca 2+ transport [5-11]. Recently, two types of Ca 2+ transport ATPase have been Abbreviation: Hepes. 4-(2-hydroxyethyl)-l-piperaz,ineethanesutphonic acid.
identified in smooth muscle, one of M r 130000 that binds calmodulin and one of M r 100000 that does not bind calmodulin. The 100000 M r ATPase resembles the Ca 2+ transport ATPase of sarcoplasmic reticulum of skeletal and cardiac muscle, whereas the 130000 M r ATPase resembles the Ca 2+ extrusion p u m p of erythrocytes and of the sarcolemma of cardiac cells [8,9]. Using newly developed methods for the separation of endoplasmic reticulum and plasma membrane vesicles, it was shown that the 100000 M r ATPase was present in the endoplasmic reticulum, whereas the 130000 M r protein was confined to the plasma membrane [9-12]. The Ca 2+ transport ATPase of sarcoplasmic reticulum of cardiac muscle, but not that of fast skeletal muscle, is regulated by cyclic AMP-depen-
0304-4165/86/$03.50 ~) 1986 Elsevier Science Publishers B.V. (Biomedical Division)
259 dent and by calmodulin-dependent phosphorylation of the 22000-25000 M r protein phospholamban (for a review see Refs. 13, 14). Phospholamban is also present in slow skeletal muscle fibres, but not fast fibres [15], and phosphorylation of the protein appears to regulate Ca 2+ transport in slow skeletal muscle [16], as well as in heart. In the present study we have investigated whether phospholamban is present in smooth muscle. Membrane fractions isolated from several smooth muscles were analysed for phosphorylatable proteins and for binding of antibodies directed against phospholamban purified from canine cardiac muscle. The results reveal the presence in several smooth muscles of polypeptides with characteristics of cardiac phospholamban that bind phospholamban-specific antibodies. Phospholamban is present in the endoplasmic reticulum and is absent from the plasma membrane. Methods
Preparation of membrane fractions Sarcoplasmic reticulum from porcine or canine cardiac muscle was prepared according to procedure I of Jones et al. [17]. Membrane fractions enriched in endoplasmic reticulum or in plasma membranes were prepared from the smooth muscle of pig stomach or pig aorta (antral part) as described previously [11]. Briefly, digitonin was added to the postmitochondrial supernatant to increase the density of plasma membranes. 0.6 M KC1 was included to extract extrinsic proteins. The supernatant was applied below a sucrose density gradient which contained 0.6 M KC1 and was spun overnight at 105000 x gmax in a Kontron TZT32 zonal rotor. The endoplasmic reticulum membranes equilibrated between 18 and 23% sucrose (designated as fraction II), and the plasma membranes were collected between about 31 and 37% sucrose (designated as fractions V and VI). Also, fractions lI and III collected at 23-27% and 27-31% sucrose, respectively, which were of intermediate composition, were used in some experiments. For the preparation of microsomes from dog or rabbit aorta, homogenization was carried out as described for pig stomach [11]. A crude micro-
somal pellet was prepared by differential centrifugation between 20000 x gmax for 15 min and 200000 × g .... for 30 min. The pellet was resuspended in 50% sucrose/0.6 M KC1, applied in Beckman SW28 tubes below layers of 45% sucrose/0.6 M KC1 and 8% sucrose/0.6 M KC1, and floated by centrifugation at 110000 x g . . . . for 16 h. This floating of the vesicles efficiently extracts extrinsic proteins [11].
Phosphorylation The medium for phosphorylating porcine stomach membranes contained 50 mM Hepes (titrated to pH 6.9 with NaOH), 1 mM EGTA, 5 mM MgC1 z, 1 mM dithiothreitol and 500 t~g/ml membrane protein. The concentration of the catalytic subunit of cyclic AMP-dependent protein kinase, when present, was 10/~g/ml. Further additions to this solution were made as indicated in the results. Phosphorylation was started by the addition of 100/~M [~,-32p]ATP to the medium at 37°C. After 2 min of incubation, phosphorylation was stopped by the addition of an equal volume of the solubilization buffer used for SDS gel electrophoresis (31 mM Tris-HC1 (pH 6.8), 4% SDS, 20% glycerol, 2% mercaptoethanol, and 0.006% Bromophenol blue as a tracking dye).
Electrophoresis and autoradiography For the analysis of the phosphoproteins, the samples were either boiled for 4 min or warmed to 37°C for 10 min. 40 /~1 were applied to 0.75 mm-thick slab gels of 12% polyacrylamide, prepared according to the discontinuous system of Laemmli [18]. The gels were fixed, dried and autoradiographed as described by Wuytack et al. [8].
Immunoblotting Non-phosphorylated membranes, which had been solubilized in SDS at 37°C or 100°C, were subjected to SDS gel electrophoresis and blotted onto nitrocellulose paper as described by Wuytack et al. [19]. The electrophoretic blots were treated with the anti-phospholamban antiserum made to phospholamban purified from dog heart, and bound antibodies were visualized by using ~25Ilabelled protein A and autoradiography as described [20].
260
The amount of antibodies bound to the nitrocellulose blot was determined by cutting out the appropriate bands, rrreasuring their surface area and counting the B-emission. Corrections for non-specifically bound radioactivity were made by subtracting the counts of pieces cut out from non-labelled areas of the blot.
polypeptides had an apparent molecular mass of about 22000 and 11000, respectively, and they presented the same mobility in SDS gels as the two forms of phospholamban of sarcoplasmic reticulum of cardiac muscle. We have therefore further characterized these phosphoproteins in order to determine whether they share other characteristics with cardiac phospholamban. In SDS gels, the relative amount of the 22000 M r and 11 000 M r forms of cardiac phospholamban depends on the pretreatment of the membranes. Both the 22000 M r and the 11000 M r form are seen when the membranes are treated with SDS at 37°C. Boiling of the sample in SDS induces a dissociation of the 22000 M r form into polypeptides of M r 11 000 [22,23]. Fig. 1 shows that both the 22000 M r phosphorylated polypeptide of endoplasmic reticulum from pig stomach and that of a crude membrane fraction from dog and rabbit aorta dissociate to the 11 000 M~ form at 100°C. It should be pointed out that the 22000 M r polypeptide is the only phosphorylated substrate affected by the boiling of the sample. In a plasma membrane fraction from pig stomach (fraction V of the density gradient) these polypeptides could hardly be detected (Fig. 1B, lane 2), indicating that these membranes do not contain phospholamban. This observation explains why the phosphorylation of phospholamban is much weaker in the crude fractions from
Materials
Catalytic subunit of cAMP-dependent protein kinase from bovine heart, trypsin (from bovine pancreas, type III), trypsin inhibitor (from bovine pancreas, type I-P), cyclic AMP and cyclic G M P were obtained from Sigma Chemical Co. Calmodulin prepared from bovine brain [21] was kindly supplied by Dr F. Wuytack. [y-32p]ATP and radioactive molecular mass markers for SDS gel electrophoresis were obtained from Amersham Corp. Results Phosphorylation of membrane proteins
In all our membrane preparations, several polypeptides were phosphorylated by the catalytic subunit of cyclic AMP-dependent protein kinase. In the endoplasmic reticulum fraction from pig stomach smooth muscle (fraction It of the density gradient), and also in crude microsomes from dog and rabbit aorta, two of these phosphorylatable
BOIL
SR
ER
+
÷
SR -
£R
Sf -
3
2
7 Mr /lO-
+
4"
-
4 4"
-
5 ÷
-
4.
200
92.5 69 46
30
"/4.3
!
/
Fig. 1. 32p autoradiograph of membranes phosphorylated by the catalytic subunit of cAMP-dependent protein kinase. The samples were boiled prior to the electrophoresis or warmed to 37°C, as indicated. (A) Pig stomach endoplasmic reticulum (ER) and pig cardiac sarcoplasmic reticulum (SR). (B) 1, pig stomach endoplasmic-reticulum; 2, pig stomach plasma-membranes; 3, crude microsomes from rabbit aorta; 4, crude microsomes from dog aorta; 5, sarcoplasmic reticulum from pig heart. The arrows indicate the 22000 and 11000 M r components. St, standard proteins.
261 dog and rabbit aorta (Fig. 1B, lanes 3 and 4), than in endoplasmic reticulum from stomach (Fig. 1B, lanes 1), since the crude preparations largely consist of plasma membranes. The absence of phosp h o l a m b a n from plasma membranes was confirmed by using antibodies (see below). P h o s p h o l a m b a n of sarcoplasmic reticulum of cardiac muscle is a substrate for an endogenous c a l c i u m - c a l m o d u l i n - d e p e n d e n t protein kinase oLJ qa "-~
ro
ro ka
~
m
u
M~/ld
[23-25]. Such a protein kinase is also present in the endoplasmic reticulum of smooth muscle, since addition of Ca z+ and calmodulin induced the phosphorylation of several polypeptides, including the 22000 and 11000 M r bands. Calmodulin was required for this phosphorylatiom whereas Ca 2+ alone was not sufficient (Fig. 2). This phosphorylation was, however, less intense than that induced by the addition of the catalytic subunit of cyclic A M P - d e p e n d e n t protein kinase. The endoplasmic reticulum fraction of pig stomach smooth muscle contains an endogenous cyclic A M P - d e p e n d e n t kinase as well, since addition of cyclic A M P results in the appearance of phosphorylated polypeptides. The substrates include the 11000 M r protein. Phosphorylation of the 22 000 M r polypeptide was not seen, possibly because it remained below the detection limit. Addition of cyclic G M P did not induce any phosphorylation (Fig. 2, lane 6). Phosphorylation by the catalytic subunit of c A M P - d e p e n d e n t protein kinase of polypeptides at positions corresponding to those of phosp h o l a m b a n could not be detected in an endoplasmic reticulum-enriched fraction isolated from the smooth muscle of pig aorta (data not shown). B i n d i n g o f antibodies
7 2 3 4 5 6
789
Fig. 2. Phosphorylation of endoplasmic reticulum (ER) vesicles from pig stomach smooth muscle by endogenous kinases. The basic phosphorylation medium (lane 1) was supplemented with Ca2+ at a calculated free concentration of 10 -5 M (lane 3), Ca2+ plus 10 t~g/ml calmodulin (lanes 4, 7 and 8), 5 ~M cyclic AMP (lane 5) or 5 t~M cyclic GMP (lane 6). The samples were treated at 37°C, except lane 8 which was treated at 100°C. For comparison, phosphorylation was also induced with 10 /~g/ml exogenous catalytic subunit of cyclic AMP-dependent protein kinase (A kinase) shown in lane 2. Lanes 7-9 were obtained from a separate gel. Pig cardiac sarcoplasmic reticulum (SR) (lane 9) was included to allow identification of the phospholamban components in the smooth muscle membranes. Lane 8 shows that the polypeptides phosphorylated in the presence of Ca2+ plus calmodulin include the 22000 Mr form of phospholamban, since it is dissociated by boiling of the sample (arrow).
Fractions from pig stomach smooth muscle enriched in plasma membranes or in endoplasmic reticulum were subjected to SDS gel electrophoresis and transferred to nitrocellulose. I m m u n o b l o t ting showed that antibodies against purified phosp h o l a m b a n from dog heart cross-reacted with 22000 and 11 000 M r polypeptides in these membrane preparations (Fig. 3A). As observed for the phosphorylated polypeptides, the 22000 M r comp o n e n t was dissociated when the sample was boiled in SDS. A m o n g the smooth muscle m e m b r a n e fractions, the endoplasmic reticulum fraction II b o u n d the most antibodies, whereas the plasma m e m b r a n e fraction V had the weakest binding capacity. Fractions III and IV, which were intermediate in composition as shown by marker enzyme activities and by their content of 100000 and 130000 M r (Ca2+ + Mg2+)-ATPases [ll], also b o u n d an intermediate quantity of anti-phosp h o l a m b a n antibodies. The relative a m o u n t of antibodies b o u n d to the
262
22000 M r component of each fraction is shown below each lane in Fig. 3A. The radioactivity bound to the 11 000 M r band was not sufficiently above background level to allow its quantification. The 22000 M r component of the endoplasmic reticulum fraction II bound nearly 10-times as much anti-phospholamban antibodies than that of the plasma membrane fraction V. Since the plasma membrane fraction contains 10-30% endoplasmic reticulum membranes [11], the amount of antiphospholamban binding observed in the plasma membrane fraction can be accounted for completely by contaminating endoplasmic reticulum. As can be seen in Fig. 3, the quantity of antiphospholamban antibodies bound by the 22000 and 11000 M r polypeptides of the endoplasmic reticulum fraction II of smooth muscle was less
than that bound by the 22 000 M r phospholamban form of sarcoplasmic reticulum of pig cardiac muscle. Quantification of the binding ratio was not attempted since the cardiac and smooth muscle forms of phospholamban may have different affinities for the antibodies. The binding of anti-phospholamban antibodies to a crude membrane fraction from canine aorta is shown in Fig. 3B. Also in this tissue the molecular mass of the binding component, and the sensitivity to heat treatment of the 22 000 M r polypeptide are similar to that seen in cardiac sarcoplasmic reticulum. No binding of antibodies was observed in the endoplasmic reticulum fraction from pig aorta, in agreement with the negative result of the phosphorylation experiments (data not shown).
BOlL
7
2
3
4
-
÷
_
+
FRACTIONI IIIIIIIIV vlsR | BOIL
+
]
-
- I
+
+
-I
-
-i ~ ~i~~iiii~i ! ~!ii~~!iii~iilli~ii!i~~:~iii~i/ii!!~iiii)i!~i~i~i!~i
I
A
~!~ !!~= ~
i
~
2942
~
B
7750
7630
302
i~i~iii~i~~
34 550
Fig. 3. lmmunoautoradiograph of antibodies against purified canine cardiac phospholamban. Membrane proteins of different fractions were separated by SDS gel electrophoresis and blotted onto nitrocellulose paper. Samples were treated at 37°C or boiled as indicated. Bound antibodies were visualized using 125l-labelled protein A and autoradiography. (A) Membranes isolated from porcine tissues. Fractions I f - V : stomach smooth muscle membranes fractionated on a sucrose density gradient. Fraction II is enriched in endoplasmic reticulum, fraction V in plasma membranes. Fractions llI and IV are of intermediate composition. 50 #g membrane protein was applied per lane. The numbers below the figure give for each fraction the cpm in the 22000 M r band. (B) Membranes from dog. Lanes ] and 2, cardiac sarcoplasmic reticulum; lanes 3 and 4, aortic microsomes. SR, cardiac sarcoplasmic reticulum.
263 Discussion
Analysis of membranes from different smooth muscles for 32p incorporation from [7-32p]ATP revealed the presence of many phosphorylatable polypeptides. In this study we have focussed our experiments on two phosphorylatable polypeptides which presented the same mobility in SDS gels as the 22000 and 11000 M r forms of phospholamban of cardiac sarcoplasmic reticulum. A more detailed comparison of these polypeptides with cardiac phospholamban was carried out mainly on an endoplasmic reticulum-enriched fraction from pig stomach. The following similarities were found: (1) the oligomeric proteins in cardiac and smooth muscle exhibited identical molecular masses of about 22000; (2) the 22000 M r polypeptides were decomposed into the same 11000 M r peptides when the membranes were boiled in the presence of SDS; (3) the phospholamban-like components were phosphorylated by an endogenous calcium-calmodulin-dependent protein kinase in both tissues, as well as by exogenous cyclic AMP-dependent protein kinase; (4) antibodies against purified cardiac phospholamban cross-reacted with the 22000 and 11000 M r phosphorylatable membrane proteins from smooth muscle. Among other smooth muscles examined for phosphorylatable phospholamban-like polypeptides, rabbit and dog aorta were positive, whereas pig aorta was negative. Immunoblotting confirmed the presence of phospholamban in dog aorta and confirmed the negative result for pig aorta. Also in bovine aorta, Chiesi et al. [26] did not find any evidence for the presence of phospholamban. On the other hand, phosphate incorporation into polypeptides of about 22000 and 11000 M,. has been observed previously by Brockbank and England [27] in rat aorta, and by Thorens [28] in rabbit mesenteric artery. These substrates were not further characterized, but in view of the present results they could correspond to phospholamban. Both the antibody binding experiments and the phosphorylation experiments on subfractioned membranes from pig stomach smooth muscle indicated that phospholamban is present in the membranes of the endoplasmic reticulum and that
plasma membranes most probably do not contain this protein. The small amount of binding of anti-phospholamban antibodies to the plasma membrane fraction could be accounted for entirely by contaminating endoplasmic reticulum. These results are similar to those obtained on cardiac muscle. Although earlier studies had suggested that phospholamban is present in cardiac sarcolemma [29,30], more recent evidence shows that this protein is confined to the sarcoplasmic reticulum [31]. The amount of phosphate incorporation into phospholamban, and the amount of binding of anti-phospholamban antibodies was much smaller in smooth muscle endoplasmic reticulum than in cardiac sarcoplasmic reticulum, This observation is not unexpected, in view of the fact that phospholamban is suggested to be closely associated with the sarcoplasmic reticulum Ca 2+-pump ATPase in a 1 : 1 molar ratio [13], and because the content of Ca 2+ transport enzyme is much lower in smooth muscle endoplasmic reticulum than in cardiac sarcoplasmic reticulum [11]. In sarcoplasmic reticulum of cardiac muscle, phosphorylation of phospholamban stimulates active Ca 2+ transport mediated by the (Ca2++ Mg2+)-ATPase. In endoplasmic reticulum of smooth muscle, phospholamban could serve a similar function, although this contention needs further experimental evidence, considering the fact that we did not observe a consistent stimulation of the Ca 2+ uptake by phosphorylation of pig stomach endoplasmic reticulum (data not shown). However, in the smooth muscle of the guinea pig Taenia coli, a ,8-adrenergic-mediated increase of the Ca 2+ uptake in an intracellular store has been suggested [32]. This intracellutar Ca 2+ store is probably the endoplasmic reticulum [2,33]. In contrast, in the rabbit ear artery such a B-effect was very small [34]. Therefore, it would be interesting to look for differences in phospholamban content between tissues which differ in their reaction to ,8-agonists, and vice versa. A stimulatory effect of cyclic AMP-dependent protein kinase on the Ca 2+ uptake in isolated membranes from smooth muscle has been reported by others (for a review see Refs. 35 and 36), but in most cases the purity of the preparation was not assessed. It probably consisted of a
264
mixture of endoplasmic reticulum and plasma membranes, and Ca 2+ uptake ascribed to plasma membrane vesicles has also been reported to be stimulated by cyclic AMP-dependent protein kinase [37,38]. It should also be pointed out that in endoplasmic reticulum of smooth muscle, unlike sarcoplasmic reticulum of cardiac muscle, many substrates for protein kinases are present. In principle, each of these phosphorylatable proteins could serve as an effector on the Ca 2+ transport ATPase. The results of this study strongly suggest that phospholamban is present in many but not all smooth muscles, and that this protein is confined to the endoplasmic reticulum and is absent from plasma membranes. The amount of phospholamban in the endoplasmic reticulum is much smaller than in sarcoplasmic reticulum of cardiac muscle, as is the amount of Ca 2+ transport enzyme, which is compatible with the idea that the presence of phospholamban is related to the Ca 2+ transport ATPase. Since evidence for the presence of a protein similar to cardiac phospholamban has now been obtained for slow skeletal muscle [15,16], smooth muscle and possibly also for blood platelets [39], the distribution of this membrane protein may be more widespread than previously thought. References 1 Johansson, B. and Somlyo, A.P. (1980) in Handbook of Physiology: Vascular Smooth Muscle (Bohr, D.F., Somlyo, A.P. and Sparks, H.V., eds.), pp. 301-324, American Physiological Society, Baltimore, MD 2 Bond, M., Kitazawa, T., Somlyo, A.P. and Somlyo, A.V. (1984) J. Physiol. 355, 677-695 3 Brading, A.F. and Sneddon, P. (1980) Br. J. Pharmacol. 70. 229 240 4 Casteels, R. and Raeymaekers, U (1979) J. Physiol. 294, 51 68 5 Daniel, E.E., Grover, A.K. and Kwan, C.Y. (1982) Fed. Proc. 41, 2898-2904 6 Wibo, M., Morel, N. and Godfraind, T. (1981) Biochim. Biophys. Acta 649, 651-660 7 Raeymaekers, L., Wuytack, F., Eggermont, J., De Schutter, G. and Casteels, R. (1983) Biochem. J. 210, 315-322 8 Wuytack, F., Raeymaekers, L., De Schutter, G. and Casteels, R. (1982) Biochim. Biophys. Acta 693, 45-52 9 Wuytack, F., Raeymaekers, L., Verbist, J,, De Smedt, H. and Casteels, R. (1984) Biochem. J. 224, 445-451
10 Wuytack, F., Raeymaekers, L. and Casteels, R. (1985) Experienta 41,900-905 11 Raeymaekers, L., Wuytack, F. and Casteels, R. (1985) Biochim. Biophys. Acta 815, 441-454 12 Carsten, M.E. and Miller, J.D. (1984) Arch. Biochem. Biophys. 232, 616-623 13 Tada, M. and Katz, A.M. (1982) Annu. Rev. Physiol. 44, 401-423 14 Tada, M. and Inui, M. (1983) J. Mol. Cell. Cardiol. 15, 565-575 15 Jorgensen, A.O. and Jones, UR. (1986) J. Biol. Chem. 261, 3775-3781 16 Kirchberger, M.A. and Tada, M. (1976) J. Biol. Chem. 251, 725-729 17 Jones, L.R., Besch, H.R., Jr., Fleming, J.W., McConnaughey, M.M. and Watanabe, A.M. (1979) J. Biol. Chem. 254, 530-539 18 Laemmli, U.K. (1970) Nature (Lond.) 227, 670-685 19 Wuytack, F., De Schutter, G,, Verbist, J. and Casteels, R. (1983) FEBS Lett. 154, 191-195 20 Jones, L.R., Simmerman, H.K.B., Wilson, W.W., Gurd, F.R.N. and Wegener, A.D. (1985) J. Biol. Chem. 260, 7721-7730 21 Sharma, R. and Wang, J. (1979) in Advances in Cyclic Nucleotide Research, Vol. 10, (Brooker, G., Greengard, P. and Robison, G.A., eds.), pp. 187-198, Raven Press, New York 22 Kirchberger, M.A. and Antonetz, T. (1982) Biochem. Biophys. Res. Commun. 105, 152-156 23 Plank, B., Wyskowsky, W., Hellmann, G. and Suko, J. (1983) Biochim. Biophys. Acta 732, 99-109 24 Le Peuch, C.J., Haiech, J. and Demaille, J.G. (1979) Biochemistry 18, 5150-5157 25 Kirchberger, M.A. and Antonetz, T. (1982) J. Biol. Chem. 257, 5685-5691 26 Chiesi, M., Gasser, J. and Carafoli, E. (1984) Biochem. Biophys. Res. Commun. 124, 797 806 27 Brockbank, K.J. and England, P.J. (1980) FEBS Lett. 122, 67 71 28 Thorens, S. (1982) J. Muscle Res. Cell Motility 3, 41%436 29 Rinaldi, M., Le Peuch, C. and Demaille, J.G. (1981) FEBS Lett. 129, 277-281 30 Huggins, J.P. and England, P.J. (1983) FEBS Lett. 163, 297-302 31 Presti, C.F., Scott, B.T. and Jones, L.R. (1985) J. Biol. Chem. 260, 13879-13889 32 Casteels, R. and Raeymaekers, L. (1979) J. Physiol. 294, 51-68 33 Raeymaekers, L. (1982) Z. Naturforsch. 37c, 481-488 34 Van Eldere, J., Raeymaekers, L. and Casteels, R. (1982) Pfliigers Arch. 395, 81-83 35 Hardman, J.G. (1981) in Smooth Muscle: An Assessment of Current Knowledge (Bulbring, E., Bradding, A., Jones, A.W. and Tomita, T., eds.), pp. 249-262, Edward Arnold, London
265 36 Hardman, J.G. (1984) J. Cardiovasc. Pharmacol. 6, $639-$645 37 Kattenburg, D.M. and Daniel, E.E. (1984) Blood Vessels 21,257-266
38 Suematsu, E., Hirata, M. and Kuriyama, H. (1984) Biochim. Biophys. Acta 773, 83-90 39 K~iser-Glanzmann, R., Gerber, E. and Luscher, E.F. (1979) Biochim. Biophys. Acta 558~ 344-347
Biochimica et Biopt~vsica Aeta 882 (1986) 258-265 Elsevier
BBA 22349
E v i d e n c e for the p r e s e n c e of p h o s p h o l a m b a n in the e n d o p l a s m i c reticulum of s m o o t h m u s c l e Luc Raeymaekers
a a n d L a r r y R. Jones b
" Laboratorium t,oor F)'siologie, University of Leuven, Campus Gasthuisberg, B-3000 Leuven (Belgium) and I, The Krannert Institute of Cardiolog},. Indiana Universi(v School of Medicine, Indianapolis. I N 46202 (U.S.A.)
(Received February 4th, 1986)
Key words: Phospholamban; Protein phosphorylation; (Smooth muscle)
In microsomal vesicles isolated from several smooth muscles many polypeptides were phosphorylated by the catalytic subunit of cyclic AMP-dependent protein kinase. In pig stomach and in rabbit and dog aorta components of M r 22 000 and 11000 were identified as forms of phospholamban. These polypeptides were, however, not observed in pig aorta. These phospholamban-like polypeptides presented the same electrophoretic mobility in sodium dodecyl sulphate gels as cardiac phospholamban, and the 22 000 M r form showed a similar reaction to heat treatment in sodium dodecyl sulphate. Antibodies against purified canine cardiac phospholamban cross-reacted with the 22000 and 11000 Mr phosphorylatable polypeptides from smooth muscle membranes. Subcellular fractionation of porcine stomach smooth muscle indicated that phospholamban was present in the membranes of the endoplasmic reticulum and not in the plasma membranes. Phospholamban was also phosphorylated by an endogenous calcium-caimodulin-dependent protein kinase and by an endogenous cyclic AMP-dependent kinase. It is concluded that the endoplasmic reticulum of many, but possibly not all, smooth muscles contains phospholamban. However, the physiological role of phospholamban in smooth muscle remains to be established.
Introduction Although the endoplasmic (sarcoplasmic) reticulum of smooth muscle cells is less developed than that of skeletal and cardiac muscle, a large body of evidence shows that it may play an important role in the regulation of the cellular calcium metabolism [1-4]. The study of Ca 2+ transport in isolated endoplasmic reticulum vesicles has been hampered because of the difficulty of isolating it from the large amount of plasma membrane, which moreover also presents ATP-dependent Ca 2+ transport [5-11]. Recently, two types of Ca 2+ transport ATPase have been Abbreviation: Hepes. 4-(2-hydroxyethyl)-l-piperaz,ineethanesutphonic acid.
identified in smooth muscle, one of M r 130000 that binds calmodulin and one of M r 100000 that does not bind calmodulin. The 100000 M r ATPase resembles the Ca 2+ transport ATPase of sarcoplasmic reticulum of skeletal and cardiac muscle, whereas the 130000 M r ATPase resembles the Ca 2+ extrusion p u m p of erythrocytes and of the sarcolemma of cardiac cells [8,9]. Using newly developed methods for the separation of endoplasmic reticulum and plasma membrane vesicles, it was shown that the 100000 M r ATPase was present in the endoplasmic reticulum, whereas the 130000 M r protein was confined to the plasma membrane [9-12]. The Ca 2+ transport ATPase of sarcoplasmic reticulum of cardiac muscle, but not that of fast skeletal muscle, is regulated by cyclic AMP-depen-
0304-4165/86/$03.50 ~) 1986 Elsevier Science Publishers B.V. (Biomedical Division)
259 dent and by calmodulin-dependent phosphorylation of the 22000-25000 M r protein phospholamban (for a review see Refs. 13, 14). Phospholamban is also present in slow skeletal muscle fibres, but not fast fibres [15], and phosphorylation of the protein appears to regulate Ca 2+ transport in slow skeletal muscle [16], as well as in heart. In the present study we have investigated whether phospholamban is present in smooth muscle. Membrane fractions isolated from several smooth muscles were analysed for phosphorylatable proteins and for binding of antibodies directed against phospholamban purified from canine cardiac muscle. The results reveal the presence in several smooth muscles of polypeptides with characteristics of cardiac phospholamban that bind phospholamban-specific antibodies. Phospholamban is present in the endoplasmic reticulum and is absent from the plasma membrane. Methods
Preparation of membrane fractions Sarcoplasmic reticulum from porcine or canine cardiac muscle was prepared according to procedure I of Jones et al. [17]. Membrane fractions enriched in endoplasmic reticulum or in plasma membranes were prepared from the smooth muscle of pig stomach or pig aorta (antral part) as described previously [11]. Briefly, digitonin was added to the postmitochondrial supernatant to increase the density of plasma membranes. 0.6 M KC1 was included to extract extrinsic proteins. The supernatant was applied below a sucrose density gradient which contained 0.6 M KC1 and was spun overnight at 105000 x gmax in a Kontron TZT32 zonal rotor. The endoplasmic reticulum membranes equilibrated between 18 and 23% sucrose (designated as fraction II), and the plasma membranes were collected between about 31 and 37% sucrose (designated as fractions V and VI). Also, fractions lI and III collected at 23-27% and 27-31% sucrose, respectively, which were of intermediate composition, were used in some experiments. For the preparation of microsomes from dog or rabbit aorta, homogenization was carried out as described for pig stomach [11]. A crude micro-
somal pellet was prepared by differential centrifugation between 20000 x gmax for 15 min and 200000 × g .... for 30 min. The pellet was resuspended in 50% sucrose/0.6 M KC1, applied in Beckman SW28 tubes below layers of 45% sucrose/0.6 M KC1 and 8% sucrose/0.6 M KC1, and floated by centrifugation at 110000 x g . . . . for 16 h. This floating of the vesicles efficiently extracts extrinsic proteins [11].
Phosphorylation The medium for phosphorylating porcine stomach membranes contained 50 mM Hepes (titrated to pH 6.9 with NaOH), 1 mM EGTA, 5 mM MgC1 z, 1 mM dithiothreitol and 500 t~g/ml membrane protein. The concentration of the catalytic subunit of cyclic AMP-dependent protein kinase, when present, was 10/~g/ml. Further additions to this solution were made as indicated in the results. Phosphorylation was started by the addition of 100/~M [~,-32p]ATP to the medium at 37°C. After 2 min of incubation, phosphorylation was stopped by the addition of an equal volume of the solubilization buffer used for SDS gel electrophoresis (31 mM Tris-HC1 (pH 6.8), 4% SDS, 20% glycerol, 2% mercaptoethanol, and 0.006% Bromophenol blue as a tracking dye).
Electrophoresis and autoradiography For the analysis of the phosphoproteins, the samples were either boiled for 4 min or warmed to 37°C for 10 min. 40 /~1 were applied to 0.75 mm-thick slab gels of 12% polyacrylamide, prepared according to the discontinuous system of Laemmli [18]. The gels were fixed, dried and autoradiographed as described by Wuytack et al. [8].
Immunoblotting Non-phosphorylated membranes, which had been solubilized in SDS at 37°C or 100°C, were subjected to SDS gel electrophoresis and blotted onto nitrocellulose paper as described by Wuytack et al. [19]. The electrophoretic blots were treated with the anti-phospholamban antiserum made to phospholamban purified from dog heart, and bound antibodies were visualized by using ~25Ilabelled protein A and autoradiography as described [20].
260
The amount of antibodies bound to the nitrocellulose blot was determined by cutting out the appropriate bands, rrreasuring their surface area and counting the B-emission. Corrections for non-specifically bound radioactivity were made by subtracting the counts of pieces cut out from non-labelled areas of the blot.
polypeptides had an apparent molecular mass of about 22000 and 11000, respectively, and they presented the same mobility in SDS gels as the two forms of phospholamban of sarcoplasmic reticulum of cardiac muscle. We have therefore further characterized these phosphoproteins in order to determine whether they share other characteristics with cardiac phospholamban. In SDS gels, the relative amount of the 22000 M r and 11 000 M r forms of cardiac phospholamban depends on the pretreatment of the membranes. Both the 22000 M r and the 11000 M r form are seen when the membranes are treated with SDS at 37°C. Boiling of the sample in SDS induces a dissociation of the 22000 M r form into polypeptides of M r 11 000 [22,23]. Fig. 1 shows that both the 22000 M r phosphorylated polypeptide of endoplasmic reticulum from pig stomach and that of a crude membrane fraction from dog and rabbit aorta dissociate to the 11 000 M~ form at 100°C. It should be pointed out that the 22000 M r polypeptide is the only phosphorylated substrate affected by the boiling of the sample. In a plasma membrane fraction from pig stomach (fraction V of the density gradient) these polypeptides could hardly be detected (Fig. 1B, lane 2), indicating that these membranes do not contain phospholamban. This observation explains why the phosphorylation of phospholamban is much weaker in the crude fractions from
Materials
Catalytic subunit of cAMP-dependent protein kinase from bovine heart, trypsin (from bovine pancreas, type III), trypsin inhibitor (from bovine pancreas, type I-P), cyclic AMP and cyclic G M P were obtained from Sigma Chemical Co. Calmodulin prepared from bovine brain [21] was kindly supplied by Dr F. Wuytack. [y-32p]ATP and radioactive molecular mass markers for SDS gel electrophoresis were obtained from Amersham Corp. Results Phosphorylation of membrane proteins
In all our membrane preparations, several polypeptides were phosphorylated by the catalytic subunit of cyclic AMP-dependent protein kinase. In the endoplasmic reticulum fraction from pig stomach smooth muscle (fraction It of the density gradient), and also in crude microsomes from dog and rabbit aorta, two of these phosphorylatable
BOIL
SR
ER
+
÷
SR -
£R
Sf -
3
2
7 Mr /lO-
+
4"
-
4 4"
-
5 ÷
-
4.
200
92.5 69 46
30
"/4.3
!
/
Fig. 1. 32p autoradiograph of membranes phosphorylated by the catalytic subunit of cAMP-dependent protein kinase. The samples were boiled prior to the electrophoresis or warmed to 37°C, as indicated. (A) Pig stomach endoplasmic reticulum (ER) and pig cardiac sarcoplasmic reticulum (SR). (B) 1, pig stomach endoplasmic-reticulum; 2, pig stomach plasma-membranes; 3, crude microsomes from rabbit aorta; 4, crude microsomes from dog aorta; 5, sarcoplasmic reticulum from pig heart. The arrows indicate the 22000 and 11000 M r components. St, standard proteins.
261 dog and rabbit aorta (Fig. 1B, lanes 3 and 4), than in endoplasmic reticulum from stomach (Fig. 1B, lanes 1), since the crude preparations largely consist of plasma membranes. The absence of phosp h o l a m b a n from plasma membranes was confirmed by using antibodies (see below). P h o s p h o l a m b a n of sarcoplasmic reticulum of cardiac muscle is a substrate for an endogenous c a l c i u m - c a l m o d u l i n - d e p e n d e n t protein kinase oLJ qa "-~
ro
ro ka
~
m
u
M~/ld
[23-25]. Such a protein kinase is also present in the endoplasmic reticulum of smooth muscle, since addition of Ca z+ and calmodulin induced the phosphorylation of several polypeptides, including the 22000 and 11000 M r bands. Calmodulin was required for this phosphorylatiom whereas Ca 2+ alone was not sufficient (Fig. 2). This phosphorylation was, however, less intense than that induced by the addition of the catalytic subunit of cyclic A M P - d e p e n d e n t protein kinase. The endoplasmic reticulum fraction of pig stomach smooth muscle contains an endogenous cyclic A M P - d e p e n d e n t kinase as well, since addition of cyclic A M P results in the appearance of phosphorylated polypeptides. The substrates include the 11000 M r protein. Phosphorylation of the 22 000 M r polypeptide was not seen, possibly because it remained below the detection limit. Addition of cyclic G M P did not induce any phosphorylation (Fig. 2, lane 6). Phosphorylation by the catalytic subunit of c A M P - d e p e n d e n t protein kinase of polypeptides at positions corresponding to those of phosp h o l a m b a n could not be detected in an endoplasmic reticulum-enriched fraction isolated from the smooth muscle of pig aorta (data not shown). B i n d i n g o f antibodies
7 2 3 4 5 6
789
Fig. 2. Phosphorylation of endoplasmic reticulum (ER) vesicles from pig stomach smooth muscle by endogenous kinases. The basic phosphorylation medium (lane 1) was supplemented with Ca2+ at a calculated free concentration of 10 -5 M (lane 3), Ca2+ plus 10 t~g/ml calmodulin (lanes 4, 7 and 8), 5 ~M cyclic AMP (lane 5) or 5 t~M cyclic GMP (lane 6). The samples were treated at 37°C, except lane 8 which was treated at 100°C. For comparison, phosphorylation was also induced with 10 /~g/ml exogenous catalytic subunit of cyclic AMP-dependent protein kinase (A kinase) shown in lane 2. Lanes 7-9 were obtained from a separate gel. Pig cardiac sarcoplasmic reticulum (SR) (lane 9) was included to allow identification of the phospholamban components in the smooth muscle membranes. Lane 8 shows that the polypeptides phosphorylated in the presence of Ca2+ plus calmodulin include the 22000 Mr form of phospholamban, since it is dissociated by boiling of the sample (arrow).
Fractions from pig stomach smooth muscle enriched in plasma membranes or in endoplasmic reticulum were subjected to SDS gel electrophoresis and transferred to nitrocellulose. I m m u n o b l o t ting showed that antibodies against purified phosp h o l a m b a n from dog heart cross-reacted with 22000 and 11 000 M r polypeptides in these membrane preparations (Fig. 3A). As observed for the phosphorylated polypeptides, the 22000 M r comp o n e n t was dissociated when the sample was boiled in SDS. A m o n g the smooth muscle m e m b r a n e fractions, the endoplasmic reticulum fraction II b o u n d the most antibodies, whereas the plasma m e m b r a n e fraction V had the weakest binding capacity. Fractions III and IV, which were intermediate in composition as shown by marker enzyme activities and by their content of 100000 and 130000 M r (Ca2+ + Mg2+)-ATPases [ll], also b o u n d an intermediate quantity of anti-phosp h o l a m b a n antibodies. The relative a m o u n t of antibodies b o u n d to the
262
22000 M r component of each fraction is shown below each lane in Fig. 3A. The radioactivity bound to the 11 000 M r band was not sufficiently above background level to allow its quantification. The 22000 M r component of the endoplasmic reticulum fraction II bound nearly 10-times as much anti-phospholamban antibodies than that of the plasma membrane fraction V. Since the plasma membrane fraction contains 10-30% endoplasmic reticulum membranes [11], the amount of antiphospholamban binding observed in the plasma membrane fraction can be accounted for completely by contaminating endoplasmic reticulum. As can be seen in Fig. 3, the quantity of antiphospholamban antibodies bound by the 22000 and 11000 M r polypeptides of the endoplasmic reticulum fraction II of smooth muscle was less
than that bound by the 22 000 M r phospholamban form of sarcoplasmic reticulum of pig cardiac muscle. Quantification of the binding ratio was not attempted since the cardiac and smooth muscle forms of phospholamban may have different affinities for the antibodies. The binding of anti-phospholamban antibodies to a crude membrane fraction from canine aorta is shown in Fig. 3B. Also in this tissue the molecular mass of the binding component, and the sensitivity to heat treatment of the 22 000 M r polypeptide are similar to that seen in cardiac sarcoplasmic reticulum. No binding of antibodies was observed in the endoplasmic reticulum fraction from pig aorta, in agreement with the negative result of the phosphorylation experiments (data not shown).
BOlL
7
2
3
4
-
÷
_
+
FRACTIONI IIIIIIIIV vlsR | BOIL
+
]
-
- I
+
+
-I
-
-i ~ ~i~~iiii~i ! ~!ii~~!iii~iilli~ii!i~~:~iii~i/ii!!~iiii)i!~i~i~i!~i
I
A
~!~ !!~= ~
i
~
2942
~
B
7750
7630
302
i~i~iii~i~~
34 550
Fig. 3. lmmunoautoradiograph of antibodies against purified canine cardiac phospholamban. Membrane proteins of different fractions were separated by SDS gel electrophoresis and blotted onto nitrocellulose paper. Samples were treated at 37°C or boiled as indicated. Bound antibodies were visualized using 125l-labelled protein A and autoradiography. (A) Membranes isolated from porcine tissues. Fractions I f - V : stomach smooth muscle membranes fractionated on a sucrose density gradient. Fraction II is enriched in endoplasmic reticulum, fraction V in plasma membranes. Fractions llI and IV are of intermediate composition. 50 #g membrane protein was applied per lane. The numbers below the figure give for each fraction the cpm in the 22000 M r band. (B) Membranes from dog. Lanes ] and 2, cardiac sarcoplasmic reticulum; lanes 3 and 4, aortic microsomes. SR, cardiac sarcoplasmic reticulum.
263 Discussion
Analysis of membranes from different smooth muscles for 32p incorporation from [7-32p]ATP revealed the presence of many phosphorylatable polypeptides. In this study we have focussed our experiments on two phosphorylatable polypeptides which presented the same mobility in SDS gels as the 22000 and 11000 M r forms of phospholamban of cardiac sarcoplasmic reticulum. A more detailed comparison of these polypeptides with cardiac phospholamban was carried out mainly on an endoplasmic reticulum-enriched fraction from pig stomach. The following similarities were found: (1) the oligomeric proteins in cardiac and smooth muscle exhibited identical molecular masses of about 22000; (2) the 22000 M r polypeptides were decomposed into the same 11000 M r peptides when the membranes were boiled in the presence of SDS; (3) the phospholamban-like components were phosphorylated by an endogenous calcium-calmodulin-dependent protein kinase in both tissues, as well as by exogenous cyclic AMP-dependent protein kinase; (4) antibodies against purified cardiac phospholamban cross-reacted with the 22000 and 11000 M r phosphorylatable membrane proteins from smooth muscle. Among other smooth muscles examined for phosphorylatable phospholamban-like polypeptides, rabbit and dog aorta were positive, whereas pig aorta was negative. Immunoblotting confirmed the presence of phospholamban in dog aorta and confirmed the negative result for pig aorta. Also in bovine aorta, Chiesi et al. [26] did not find any evidence for the presence of phospholamban. On the other hand, phosphate incorporation into polypeptides of about 22000 and 11000 M,. has been observed previously by Brockbank and England [27] in rat aorta, and by Thorens [28] in rabbit mesenteric artery. These substrates were not further characterized, but in view of the present results they could correspond to phospholamban. Both the antibody binding experiments and the phosphorylation experiments on subfractioned membranes from pig stomach smooth muscle indicated that phospholamban is present in the membranes of the endoplasmic reticulum and that
plasma membranes most probably do not contain this protein. The small amount of binding of anti-phospholamban antibodies to the plasma membrane fraction could be accounted for entirely by contaminating endoplasmic reticulum. These results are similar to those obtained on cardiac muscle. Although earlier studies had suggested that phospholamban is present in cardiac sarcolemma [29,30], more recent evidence shows that this protein is confined to the sarcoplasmic reticulum [31]. The amount of phosphate incorporation into phospholamban, and the amount of binding of anti-phospholamban antibodies was much smaller in smooth muscle endoplasmic reticulum than in cardiac sarcoplasmic reticulum, This observation is not unexpected, in view of the fact that phospholamban is suggested to be closely associated with the sarcoplasmic reticulum Ca 2+-pump ATPase in a 1 : 1 molar ratio [13], and because the content of Ca 2+ transport enzyme is much lower in smooth muscle endoplasmic reticulum than in cardiac sarcoplasmic reticulum [11]. In sarcoplasmic reticulum of cardiac muscle, phosphorylation of phospholamban stimulates active Ca 2+ transport mediated by the (Ca2++ Mg2+)-ATPase. In endoplasmic reticulum of smooth muscle, phospholamban could serve a similar function, although this contention needs further experimental evidence, considering the fact that we did not observe a consistent stimulation of the Ca 2+ uptake by phosphorylation of pig stomach endoplasmic reticulum (data not shown). However, in the smooth muscle of the guinea pig Taenia coli, a ,8-adrenergic-mediated increase of the Ca 2+ uptake in an intracellular store has been suggested [32]. This intracellutar Ca 2+ store is probably the endoplasmic reticulum [2,33]. In contrast, in the rabbit ear artery such a B-effect was very small [34]. Therefore, it would be interesting to look for differences in phospholamban content between tissues which differ in their reaction to ,8-agonists, and vice versa. A stimulatory effect of cyclic AMP-dependent protein kinase on the Ca 2+ uptake in isolated membranes from smooth muscle has been reported by others (for a review see Refs. 35 and 36), but in most cases the purity of the preparation was not assessed. It probably consisted of a
264
mixture of endoplasmic reticulum and plasma membranes, and Ca 2+ uptake ascribed to plasma membrane vesicles has also been reported to be stimulated by cyclic AMP-dependent protein kinase [37,38]. It should also be pointed out that in endoplasmic reticulum of smooth muscle, unlike sarcoplasmic reticulum of cardiac muscle, many substrates for protein kinases are present. In principle, each of these phosphorylatable proteins could serve as an effector on the Ca 2+ transport ATPase. The results of this study strongly suggest that phospholamban is present in many but not all smooth muscles, and that this protein is confined to the endoplasmic reticulum and is absent from plasma membranes. The amount of phospholamban in the endoplasmic reticulum is much smaller than in sarcoplasmic reticulum of cardiac muscle, as is the amount of Ca 2+ transport enzyme, which is compatible with the idea that the presence of phospholamban is related to the Ca 2+ transport ATPase. Since evidence for the presence of a protein similar to cardiac phospholamban has now been obtained for slow skeletal muscle [15,16], smooth muscle and possibly also for blood platelets [39], the distribution of this membrane protein may be more widespread than previously thought. References 1 Johansson, B. and Somlyo, A.P. (1980) in Handbook of Physiology: Vascular Smooth Muscle (Bohr, D.F., Somlyo, A.P. and Sparks, H.V., eds.), pp. 301-324, American Physiological Society, Baltimore, MD 2 Bond, M., Kitazawa, T., Somlyo, A.P. and Somlyo, A.V. (1984) J. Physiol. 355, 677-695 3 Brading, A.F. and Sneddon, P. (1980) Br. J. Pharmacol. 70. 229 240 4 Casteels, R. and Raeymaekers, U (1979) J. Physiol. 294, 51 68 5 Daniel, E.E., Grover, A.K. and Kwan, C.Y. (1982) Fed. Proc. 41, 2898-2904 6 Wibo, M., Morel, N. and Godfraind, T. (1981) Biochim. Biophys. Acta 649, 651-660 7 Raeymaekers, L., Wuytack, F., Eggermont, J., De Schutter, G. and Casteels, R. (1983) Biochem. J. 210, 315-322 8 Wuytack, F., Raeymaekers, L., De Schutter, G. and Casteels, R. (1982) Biochim. Biophys. Acta 693, 45-52 9 Wuytack, F., Raeymaekers, L., Verbist, J,, De Smedt, H. and Casteels, R. (1984) Biochem. J. 224, 445-451
10 Wuytack, F., Raeymaekers, L. and Casteels, R. (1985) Experienta 41,900-905 11 Raeymaekers, L., Wuytack, F. and Casteels, R. (1985) Biochim. Biophys. Acta 815, 441-454 12 Carsten, M.E. and Miller, J.D. (1984) Arch. Biochem. Biophys. 232, 616-623 13 Tada, M. and Katz, A.M. (1982) Annu. Rev. Physiol. 44, 401-423 14 Tada, M. and Inui, M. (1983) J. Mol. Cell. Cardiol. 15, 565-575 15 Jorgensen, A.O. and Jones, UR. (1986) J. Biol. Chem. 261, 3775-3781 16 Kirchberger, M.A. and Tada, M. (1976) J. Biol. Chem. 251, 725-729 17 Jones, L.R., Besch, H.R., Jr., Fleming, J.W., McConnaughey, M.M. and Watanabe, A.M. (1979) J. Biol. Chem. 254, 530-539 18 Laemmli, U.K. (1970) Nature (Lond.) 227, 670-685 19 Wuytack, F., De Schutter, G,, Verbist, J. and Casteels, R. (1983) FEBS Lett. 154, 191-195 20 Jones, L.R., Simmerman, H.K.B., Wilson, W.W., Gurd, F.R.N. and Wegener, A.D. (1985) J. Biol. Chem. 260, 7721-7730 21 Sharma, R. and Wang, J. (1979) in Advances in Cyclic Nucleotide Research, Vol. 10, (Brooker, G., Greengard, P. and Robison, G.A., eds.), pp. 187-198, Raven Press, New York 22 Kirchberger, M.A. and Antonetz, T. (1982) Biochem. Biophys. Res. Commun. 105, 152-156 23 Plank, B., Wyskowsky, W., Hellmann, G. and Suko, J. (1983) Biochim. Biophys. Acta 732, 99-109 24 Le Peuch, C.J., Haiech, J. and Demaille, J.G. (1979) Biochemistry 18, 5150-5157 25 Kirchberger, M.A. and Antonetz, T. (1982) J. Biol. Chem. 257, 5685-5691 26 Chiesi, M., Gasser, J. and Carafoli, E. (1984) Biochem. Biophys. Res. Commun. 124, 797 806 27 Brockbank, K.J. and England, P.J. (1980) FEBS Lett. 122, 67 71 28 Thorens, S. (1982) J. Muscle Res. Cell Motility 3, 41%436 29 Rinaldi, M., Le Peuch, C. and Demaille, J.G. (1981) FEBS Lett. 129, 277-281 30 Huggins, J.P. and England, P.J. (1983) FEBS Lett. 163, 297-302 31 Presti, C.F., Scott, B.T. and Jones, L.R. (1985) J. Biol. Chem. 260, 13879-13889 32 Casteels, R. and Raeymaekers, L. (1979) J. Physiol. 294, 51-68 33 Raeymaekers, L. (1982) Z. Naturforsch. 37c, 481-488 34 Van Eldere, J., Raeymaekers, L. and Casteels, R. (1982) Pfliigers Arch. 395, 81-83 35 Hardman, J.G. (1981) in Smooth Muscle: An Assessment of Current Knowledge (Bulbring, E., Bradding, A., Jones, A.W. and Tomita, T., eds.), pp. 249-262, Edward Arnold, London
265 36 Hardman, J.G. (1984) J. Cardiovasc. Pharmacol. 6, $639-$645 37 Kattenburg, D.M. and Daniel, E.E. (1984) Blood Vessels 21,257-266
38 Suematsu, E., Hirata, M. and Kuriyama, H. (1984) Biochim. Biophys. Acta 773, 83-90 39 K~iser-Glanzmann, R., Gerber, E. and Luscher, E.F. (1979) Biochim. Biophys. Acta 558~ 344-347