- Email: [email protected]
[44] Transmembrane Ca2+ signaling and a new class of inhibitors
570
M E T H O D S FOR S T U D Y OF C a - B I N D I N G PROTEINS
[44]
shown to be more rapid in serum-free than in serum-supplemented medium. The restricted population of PV-positive neurons seen at later stages in culture is reminiscent to that observed in vioo. It may represent neurons expressing differentially Ca2+-dependent processes. Acknowledgments This work was supported by the Swiss National Science Foundation (3.559.083 and 3.147.085), the Wilhelm Sander, Hartmann M011er,EMDO, Sandoz, CIBA-Geigy,and Julius Klaus-Stiftung, Roche Foundation, and Jubil~umsspende fiir die Universit/RZfirich. We would like to thank Dr. A. M. Rowlerson for critical reading of this manuscript.
[44] T r a n s m e m b r a n e
C a 2+ S i g n a l i n g a n d a N e w C l a s s o f Inhibitors
By HIROYOSHI HIDAKA and TOSHIO TANAKA Introduction Calcium ion plays a critical and central role in various biological events as second messenger.l The intracellular calcium ion is involved in the mechanism of stimulus-induced cellular response such as muscle contraction, metabolic regulation, endo- and exocytosis, cell motility, cytoplasmic transport, cell proliferation, and fertilization. 2 Figure 1 shows a schematic representation of several processes that are involved in this flow of information in the Ca2÷-dependent regulatory system of cell function and selective inhibitors of each process. To investigate molecular mechanisms involved in the calcium messenger system, we developed potent specific inhibitors of each step of the calcium mesenger system, as shown in the model of Fig. 1. These selective inhibitors of each process in the calcium messenger system are discussed herein. N e w Aspects of Calmodulin Antagonists
The pharmacological action of naphthalenesulfonamides 3 and phenothiazines 4 have been well characterized. Since the discovery of the A. R. Means, J. S. Tash, and J. G. Chafouleas, Physiol. Rev. 62, 1 (1982). 2 A. K. Campbell, "Intracellular Calcium," Wiley, New York, 1983. 3 H. Hidaka and T. Tanaka, this series, Vol. 102, p. 185. 4 B. Weiss, this series, Vol. 102, p. 171. METHODS IN ENZYMOLOGY, VOL. 139
Copyright © 1987by Academic Press, Inc. All rights of reproduction in any form reserved.
[44]
INHIBITORS OF Ca 2+ SIGNALING
CalmodulinAntagonists CalciumChannelBlockers nifedipine NO.233 verapamllprenylamlne HT-74 diltiazem bepridi|
571
W-7
~
~
~
CaMBPInhibitors
~MLCK / CaM .,)
\
/
Inhibitor IML-e)
1 ~~ -
C(/~+-PDE~ ~mPDcEeltl~°hir~it
\\ C(a~HA~t~g~i~st
,- .., ~ PK_C~ pK_~Hi.n7~lbitor
Intracellular
~ m l
cell membrane
FIG. 1. Intracellular calcium messenger system and specific inhibitors. CaM, Calmodulin; CaMBP inhibitors, calmodulin-binding protein inhibitors; MLCK, myosin light chain kinase; Ca2+-PDE, Ca2+-dependent cyclic nucleotide phosphodiesterase; PK-C, protein kinase C.
Ca2+-dependent interaction of phenothiazines 5 and naphthalenesulfonamides 6 with calmodulin and the subsequent inhibition of calmodulin activity, a number of calmodulin antagonists have been developed. Included are melittin, 7 mastoparans, 8 calmidazolium, 9 compound 48/80,1° and others. H N~-Dansyl-L-arginine-4-t-butylpiperidine amide (No. 233) was found to be a selective and potent CaM antagonist possessing the fluorescence properties of the dansyl moiety, a useful probe for the study 5 B. Weiss and R. M. Levin, Ado. Cyclic Nucleotide Res. 9, 285 (1978). 6 H. Hidaka, M. Asano, S. Iwadare, I. Matsumoto, T. Totsuka, and N. Aoki, J. Pharmacol. Exp. Ther. 207, 8 (1978). 7 M. Comte, Y. Maulet, and J. A. Cox, Biochem. J. 209, 269 (1983). 8 D. A. Malencik and S. R. Anderson, Biochem. Biophys. Res. Commun. 114, 50 (1983). 9 K. Gietzen, A. Wuthrich, and H. t, ader, Biochem. Biophys. Res. Comrnun. 101, 418 (1981). 10 K. Gietzen, D. Sanchez, and H. Bader, IRCS Med. Sci. 11, 12 (1983). 11 H. Itoh, T. Tanaka, Y. Mitani, and H. Hidaka, Biochem. Pharmacol. 35, 217 (1986).
572
METHODS FOR STUDY OF Ca-BINDING PROTEINS
[44]
of CaM and its interactions. 12Among many of the calmodulin antagonists, naphthalenesulfonamide derivatives have been used for in vivo studies such as cell or tissue level experiments, as they have high specificity for calmodulin and a lower cytotoxicity compared with phenothiazines and also a higher bioavailability. Detailed pharmacological properties of naphthalenesulfonamides became evident in their specific binding to calmodulin, 13 structure-activity relationships, 14 affinity chromatography, 15 and their applications to enzymology~6: and cell biology, as The functions of calmodulin in various cell types have also been studied extensively, using calmodulin antagonists such as naphthalenesulfonamide derivatives. The involvement of calmodulin has been noted in various cellular processes, 19 including cell proliferation, tumor promotion, acrosome reaction of echinoderm sperm, oocyte maturation, neutrophil activation, arterial muscle contraction, platelet function, Ca 2+ transport of plasma membranes, biosyntheses of monoamine neurotransmitters, plant cell function, and others. Moreover, calmodulin antagonists are useful tools for studying the molecular mechanism of calmodulin action and to characterize the binding sites on calmodulin for its target proteins and its antagonists. W e 2° and Laporte et al. 21 independently proposed the possibility that calcium ion induces conformational changes in calmodulin that expose hydrophobic sites on the surface of the molecule and which may act as sites of interaction with calmodulin antagonists and its target proteins. 22 Subsequently, we demonstrated that the affinity of naphthalenesulfonamides23 or local anesthetics 24 for Ca2+-CaM correlated well with their hydrophobicity and \
12 T. Tanaka, M. Inagaki, and H. Hidaka, Arch. Biochem. Biophys. 220, 188 (1983). 13 H. Hidaka, T. Yamaki, M. Naka, T. Tanaka, H. Hayashi, and R. Kobayashi, Mol. Pharmacol. 17, 66 (1980). 14 H. Hidaka, M. Asano, and T. Tanaka, Mol. Pharmacol. 20, 571 (1981). 15T. Endo, T. Tanaka, T. Isobe, H. Kasai, T. Okuyama, and H. Hidaka, J. Biol. Chem. 256, 12485 (1981). 16 H. Hidaka, T. Yamald, T. Totsuka, and M. Asano, Mol. Pharmacol. 15, 49 (1979). 17T. Tanaka, T. Ohmura, T. Yamakado, and H. Hidaka, Mol. Pharmacol. 22, 408 (1982). 18 H. Hidaka, Y. Sasaki, T. Tanaka, T. Endo, S. Ohno, Y. Fujii, and T. Nagata, Proc. Natl. Acad. Sci. U.S.A. 78, 4354 (1981). 19 H. Hidaka and D. J. Hartshorne, "Calmodulin Antagonists and Cellular Physiology." Academic Press, Orlando, Florida, 1985. 2o T. Tanaka and H. Hidaka, J. Biol. Chem. 255, 11078 (1980). 21 D. C. LaPorte, B. M. Wierman, and D. R. Storm, Biochemistry 19, 3814 (1980). 22 H. Hidaka and T. Tanaka, in "Calmodulin and Intracellular Ca ++ Receptors" (S. Kakiuchi, H. Hidaka, and A. R. Means, eds.), p. 19. Plenum, New York, 1982. 23 T. Tanaka, T. Ohmura, and H. Hidaka, Mol. Pharmacol. 22, 403 (1982). ~4 T. Tanaka and H. Hidaka, Biochem. Biophys. Res. Commun. 101, 447 (1981).
[44]
INHIBITORS OF C a 2+ SIGNALING
573
their potency in inhibiting Ca 2+, CaM-dependent enzymes. Recently, Ca2+-dependent hydrophobic interactions between CaM and phenylSepharose 25 or a homologous series of to-aminoalkylagaroses were demonstrated. 26 Moreover, we found that some newly synthesized hydrophobic compounds activate Ca2+-dependent cyclic nucleotide phosphodiesterase. 27 These results suggest that hydrophobic interactions between the Ca2+-CaM complex and enzymes may be important for the enzyme activation by CaM. On the other hand, we found that Ca 2÷dependent cyclic nucleotide phosphodiesterase was activated by polyaspartic acid, an event which suggests that acidic amino acids in the CaM molecule may also play an important role in the stimulation of Ca2+-PDE. A covalent adduct of calmodulin with 1 mol of norchlorpromazine is prepared and this CAPPrCaM complex has the ability to bind to CaMdependent enzymes, with a high affinity, but loses the ability to activate phosphodiesterase and myosin light chain kinase. 28 However, it can activate the phosphatase activity of calcineurin.29 The interaction of calmodulin with Ca 2÷ channel blockers such as bepridil has been reported, n,3° We demonstrated that prenylamine interacts with CaM at sites differing from those related to the naphthalenesulfonamides and phenothiazines. 13,3~ Moreover, we found that a newly synthesized CaM antagonist, 3-(2-benzothiazolyl)-4,5-dimethoxy-N-[3-(4-phenylpiperidinyl)propyl]benzenesulfonamide (HT-74) may bind to CaM in a manner different than heretofore reported and have different effects on Ca2+-induced contractions of depolarized vascular smooth muscle from those of naphthalenesulfonamides. 32 The CaM antagonist R24571 and Ca 2÷ channel blockers prenylamine and diltiazem bind to CaM and potentiate felodipine binding. 33These results suggest that allosteric interactions occur among different drug-binding sites on CaM, hence these compounds should be useful tools for analyzing the functional domains of CaM.
25 R. Gopalakrishna and W. B. Anderson, Biochem. Biophys. Res. Commun. 104, 830 (1982). 26 T. Tanaka, H. Umekawa, T. Ohmura, and H. Hidaka, Biochim. Biophys. Acta 787, 158 (1984). 27 T. Tanaka, E. Yamada, T. Sone, and H. Hidaka, Biochemistry 22, 1030 (1983). 2s D. L. Newton, T. R. Burke, Jr., K. C. Rice, and C. B. Klee, Biochemistry 22, 5472 (1983). 29 D. L. Newton and C. B. Klee, FEBS Lett. 165, 269 (1984). 30 H. Itoh, T. Ishikawa, and H. Hidaka, J. Pharmacol. Exp. Ther. 230, 737 (1984). 31 M. Inagaki and H. Hidaka, Pharmacology 29, 75 (1984). 32 T. Tanaka, H. Umekawa, M. Saitoh, T. Ishikawa, T. Shin, M. Ito, Y. Kawamatsu, H. Sugihara, and H. Hidaka, Mol. Pharmacol. 29, 264 (1986). 33 j. D. Johnson, Biochem. Biophys. Res. Commun. 112, 787 (1983).
574
METHODS FOR STUDY OF Ca-BINDING PROTEINS
[44]
Inhibitors of Ca 2+-, CaM-Dependent Enzymes It is now generally accepted that calmodulin plays a general role in calcium ion signal transduction. W h e n calcium binds to calmodulin, it undergoes structural changes which generate specificinteraction site(s) recognized by various enzymes and proteins. As it is not feasible to use calmodulin antagonists alone for analyzing which calmodulin-depcndent enzyme is responsible for cach Ca2+-dependent cell function, we have continued to make effortsto acquire specificinhibitorsof Ca 2+, calmodufin-dependent enzymes. W e found that vinpocetine [14-ethoxycarbonyl(3a,16a-ethyl)-14,15-eburnamenine] is a potent selective inhibitor of CaE+-dependcnt cyclic nucleotide phosphodiestcrase34 and ML-9 [1-(5chloronaphtalene-l-sulfonyl)-IH-hexahydro-1,4-diazepine] is a specific inhibitor of myosin lightchain kinasc, respectively. W e investigatcd the pharmacological properties of these selectiveinhibitorsof Ca 2+, calmodufin-dependent enzymes and thc physiological rolc of each enzyme in various cell functions.
Modulators of Cae+-Dependent Cyclic Nucleotide Phosphodiesterase Recent studies suggested that cyclic AMP or cyclic GMP plays an important role in regulating Ca2+-dependent cell functions. Cyclic nucleotide phosphodiesterase is the only known pathway for the degradation of cyclic nucleotides) 5 Multiple forms of phosphodiesterase have been found in all the mammalian tissue investigated. 36 Although calmodulin was first found to be a Ca2+-dependent stimulator of cyclic nucleotide phosphodiesterase,37,38 the physiological significance of the calmodulindependent form of phosphodiesterase has yet to be established. Calmodulin antagonists such as W-7 and phenothiazine derivatives selectively inhibit Ca2+-dependent phosphodiesterase, by interacting with the Ca2+-calmodulin complex) :3 However, it has been demonstrated that calmodulin not only activates Ca2÷-dependent phosphodiesterase but also a number of enzymes, including myosin light chain kinase. Therefore, the utility of calmodulin antagonists as tools for elucidation of the physiological role of Ca2+-dependent phosphodiesterase in various cell functions is necessarily limited. Selective inhibitors and activators of Ca2+-dependent cyclic nucleotide phosphodiesterase should be appropriate tools for eluci34 M. Hagiwara, T. Endo, and H. Hidaka, Biochem. Pharmacol. 33, 453 (1984). 35 M. M. Appleman, W. J. Thompson, and T. R. Russell, Ado. Cyclic Nucleotide Res. 3, 65 (1973). 36 j. N. Wells and J. G. Hardman, Adv. Cyclic Nucleotide. Res. 8, 119 (1977). 37 S. Kakiuchi and R. Yamazaki, Biochem. Biophys. Res. Commun. 41, 1104 (1970). 38 W. Y. Cheung, Biochem. Biophys. Res. Commun. 38, 533 (1970).
[44]
INHIBITORS OF Ca 2+ SIGNALING
575
dating and properties and physiological significance of this enzyme. Vinpocetine inhibits selectively Ca2+-dependent phosphodiesterase not by interacting with CaM but rather by interacting with the enzyme directly. The effects of vinpocetine on rabbit aorta cyclic nucleotide phosphodiesterase, separated by DEAE-cellulose column chromatography, were investigated. The three forms of cyclic nucleotide phosphodiesterase observed were cyclic GMP phosphodiesterase (FI), Ca2÷-dependent phosphodiesterase (FII), and cyclic AMP phosphodiesterase (FIII). The results indicated a selective inhibition of Ca2+-dependent phosphodiesterase among these forms of the enzymes, and that the concentrations of vinpocetine which produced 50% inhibition of the Ca2+-dependent phosphodiesterase were about 21/zM, both in the presence and absence of the Ca2+-CaM complex. On the other hand, ICs0 values of this compound for cGMP phosphodiesterase and cAMP phosphodiesterase exceeded 500 /xM. Kinetic analysis by Dixon plots showed noncompetitive inhibition, with respect to cyclic GMP and a Ki value of 14/xM in the presence of the Ca2÷-CaM complex. Thus, vinpocetine may directly inhibit Ca2÷-dependent phosphodiesterase while CaM antagonists such as W-7 antagonize CaM-induced activation of the enzyme. This was tested further by examining the ability of vinpocetine and W-7 to alter the activation of the phosphodiesterase by CaM. Vinpocetine-induced inhibition of phosphodiesterase activity was not overcome by increasing the concentration of CaM and the extent of activation of CaM was greatly inhibited. On the other hand, the W-7-induced inhibition of the enzyme was overcome by increasing the concentration of CaM. These results suggested that vinpocetine inhibits Ca2+-dependent phosphodiesterase by a mechanism differing from those of CaM antagonists. The exposure of K+-depolarized rabbit aorta to 1, 10, and 100/xM vinpocetine resulted in a decrease in tension of aortic strips associated with a significant increase in cyclic GMP levels by 10 and 100/zM of vinpocetine and not statistically with 1 /.~M vinpocetine. There was no increase in cyclic AMP levels associated with relaxation and no increase in guanylate cyclase and adenylate activity in concentrations of vinpocetine ranging from 10 to 1000/xM. These results suggest that the increase in cyclic GMP level may be due to the selective inhibitory effect of vinpocetine on the Ca2+-dependent phosphodiesterase activity. 34 We synthesized HA-542 [1-(2,4-dipiperidino-6-quinazolinesulfonyl)-4(2-ethoxy-2-phenylethyl)piperazine] and HA-543 [1-(2,4-dipiperidino-6quinazolinesulfonyl)-4-cinnamylpiperazine] and found that these compounds are selective activators of Ca2+-dependent cyclic nucleotide phosphodiesterase. 27 As this activation was observed with 2,4-dipiperidino-6-quinazolinesulfonamides but not with 4-piperidino-6-quinazoline-
576
METHODS FOR STUDY OF Ca-BINDING PROTEINS
[44]
sulfonamides, the activation seems to be dependent on the piperidine residue at the 2 and 4 positions of the quinazoline ring and on the extent of hydrophobicity of each compound determined as retention index, using high-performance liquid chromatography. These 2,4-dipiperidino-6quinazoline-sulfonamides activate Ca2+-dependent phosphodiesterase in the absence of the Ca2+-CaM complex and do not further enhance the activity of the phosphodiesterase in the presence of the Ca2÷-CaM complex. These drugs are potent inhibitors of cyclic AMP and cyclic GMP phosphodiesterases. CaM antagonists such as N-(6-aminohexyl)-5chloro-l-naphthalenesulfonamide (W-7) and chlorpromazine inhibited selectively the quinazolinesulfonamide-induced activations of the phosphodiesterase. All these observations suggest that the quinazolinesulfonamides are calcium-independent activators of CaE÷-dependent phosphodiesterase and these compounds are proving to be most useful for studies on the molecular mechanism of CaM action and phosphodiesterase regulation, in vitro. Myosin Light Chain Kinase Inhibitor CaM antagonists such as W-7 or phenothiazine derivatives selectively inhibit myosin light chain kinase through their interaction with the Ca2÷CaM complex. However, CaM is involved in the stimulation of a variety of enzymes other than CaZ÷-dependent cyclic nucleotide phosphodiesterase and CaM seems to have multifunctional roles as a mediator of intracellular calcium ion signals. Therefore, CaM antagonists have limitations as tools for investigating the role of myosin light chain kinase. We found that a newly synthesized compound, 1-(5-chloronaphthalene-l-sulfonyl)-lH-hexahydro-l,4-diazepine (ML-9) is a potent selective inhibitor of myosin light chain kinase. Apparent Ki values of ML-9 for myosin light chain kinase from smooth muscle, cAMP-dependent protein kinase from bovine heart, and protein kinase C from rabbit brain were 3.8, 54, and 32 /xM, respectively. ML-9 also inhibited Caz+, CaM-independent activity of trypsin-treated myosin light chain kinase with a Ki value of 4.0 /xM. Kinetic analysis by the double-reciprocal curve revealed that the inhibition of myosin light chain kinase produced by ML-9 was competitive with respect to ATP and noncompetitive with respect to myosin light chain. Increasing the concentration of ML-9 inhibited selectively the Ca2÷-dependent phosphorylation of human platelet myosin light chain (20 kDa) with an IC50 value of 12/xM. However, ML-9 produced little inhibition of CaZ+-activated, phospholipid-dependent phosphorylation of 40-kDa peptides of human platelets. ML-9 is thus a potent selective inhibitor of myosin light chain kinase from smooth muscle and nonmuscle (platelet) and this compound will aid
[44]
INHIBITORS OF C a 2+ SIGNALING
577
in determining physiological functions of protein kinase in smooth muscle and nonmuscle cell functions. Selective Modulators of Protein Kinase C Calcium ion is an important intracellular messenger involved in the regulation of a variety of cell functions. Although the exact mechanism by which calcium ion exerts its influence has not been clarified, Ca2+-dependent protein phosphorylation is one of the major general mechanisms by which intracellular events in mammalian tissues are controlled by external physiological stimuli) 9 There are at least two different mechanisms related to the regulation of calcium-dependent protein kinases. Some Ca2÷-dependent protein kinases are stimulated by the Ca2÷-CaM complex and another type of Ca2+-dependent protein kinase, which requires phospholipid and diglyceride as cofactors, is a Ca2÷-activated, phospholipiddependent protein kinase (protein kinase C): ° However, the precise relationship between these two types of Ca2+-dependent protein kinases is unknown. CaM-dependent protein kinases such as myosin light chain kinase have been purified and charcterized.4~ We reported that myosin light chain kinase from chicken gizzard and rabbit skeletal muscle can be stimulated by limited proteolysis as well as by the Ca2+-CaM complex.42 Our data suggest that the limited proteolysis of CaM-dependent myosin kinase not only stimulates the activity of the kinase but also converts the kinase to the Ca2+-CaM-insensitive form. Protein kinase C has also been reported to be activated by limited proteolysis and converted to the Ca2÷phospholipid-insensitive form of this kinase. Myosin from smooth muscle and nonmuscle but not from skeletal and cardiac muscle can serve as a substrate for protein kinase C as well as for the Ca2+-CaM-dependent myosin light chain kinase: 3,44 When unphosphorylated smooth muscle heavy meromyosin (HMM) is phosphorylated by myosin light chain kinase, its actin-activated myosin ATPase activity increases. 45 We found that protein kinase C and myosin light chain kinase catalyzes the phosphorylation of different sites within the 20,000-Da light 39 H. Rasmussen, L Kojima, K. Kojima, W. Zawalich, and W. Apfeldorf, Adv. Cyclic Nucleotide Protein Phosphorylation Res. 18, 159 (1984). 4o y . Takai, A. Kishimoto, Y. Kawahara, R. Minakuchi, K. Sano, V. Kikkawa, T. Mori, B. Yu, K. Kaibuchi, and Y. Nishizuka, Adv. Cyclic Nucleotide Res. 14, 301 (1981). 41 C. B. Klee and T. C. Vanaman, Adv. Protein Chem. 15, 213 (1982). 42 T. Tanaka, M. Naka, and H. Hidaka, Biochem. Biophys. Res. Commun. 92, 313 (1980). 43 T. Endo, M. Naka, and H. Hidaka, Biochem. Biophys. Res. Commun. 105, 942 (1982). 44 M. Naka, M. Nishikawa, R. S. Adelstein, and H. Hidaka, Nature (London) 306, 490 (1983). 45 R. S. Adelstein, J. R. Sellers, M. A. Conti, M. D. Pato, and P. DeLanerolle, Fed. Proc., Fed. Am. Soc. Exp. Biol. 41, 2873 (1982).
578
METHODS FOR STUDY OF Ca-BINDING PROTEINS
[44]
[ C ,modu,n I MLCK
It
MEG @
Fro. 2. Regulation of actomyosin in smooth muscle and nonmuscle cells. PK-A, cAMPdependent protein kinase; MLCK, myosin light chain kinase; MLC, myosin light chain; PK-C, protein kinase C.
chain of smooth muscle HMM and that sequential phosphorylation of HMM by myosin light chain kinase and protein kinase C decreases the actin-activated Mg-ATPase activity, as compared to that following phosphorylation by myosin light chain kinase alone. 46,47 Moreover, we reported that protein kinase C incorporated phosphate into two sites of myosin light chain kinase and that this phosphorylation resulted in a reduced affinity for calmodulin.48 This dual regulation of the Ca2+-dependent myosin light chain phosphorylation system by Ca2+-phospholipid and Ca2+-CaM is complex to analyze (Fig. 2). Almost all CaM antagonists are hydrophobic and basic in property, thus it is not surprising that all of the CaM antagonists reported can inhibit the activity of protein kinase C. 49-51All these findings indicate that it is difficult to elucidate the physiological significance of protein kinase C and the interrelationship between the two putative Ca2÷-dependent protein phosphorylation systems using CaM antagonists. Now when the naphthalene ring of the CaM antagonists, naphthalenesulfonamides, is replaced by isoquinoline, the deriva46 M. Nishikawa, H. Hidaka, and R. S. Adelstein, J. Biol. Chem. 258, 14069 (1983). 47 M. Nishikawa, J. R. Sellers, R. S. Adelstein, and H. Hidaka, J. Biol. Chem. 259, 8808 (1984). M. Ikebe, M. Inagaki, K. Kanamaru, and H. Hidaka, J. Biol. Chem. 260, 4547 (1985). 49 T. Mori, Y. Takai, R. Minakuchi, B. Yu, and Y. Nishizuka, J. Biol. Chem. 255, 8378 (1980). 5o R. C. Schatzman, B. C. Wise, and J. F. Kuo, Biochem. Biophys. Res. Commun. 92, 313 (1980). 51 D e - F , Qi, R. C. Schatzman, G. J. Mazzei, R. S. Turner, R. L. Raylor, S. Liao, and H. Rasmussen, in "Calcium and Cell Function" (W. Y. Cheung, ed.), Vol. 4, p. 1. Academic Press, New York, 1983.
[44]
INHIBITORS OF C a 2+ SIGNALING
579
tives are no longer CaM- and phospholipid-interacting agents but rather suppress protein kinase C, directly. Among them, l-(5-isoquinolinesulfonyl)-2-methylpiperazine, refered to as H-7, is a relatively selective inhibitor of protein kinase C.52 The Ki value of H-7 for protein kinase C is 6.0 /zM. On the other hand, N-[2-(methylamino)ethyl]-5-isoquinolinesulfonamide (H-8) seems to be the most active of the inhibitors examined and inhibits more markedly the cGMP-dependent and cAMP-dependent protein kinases than do the other kinases and ATPases examined. The Ki values of H-8 for cGMP-dependent and cAMP-dependent protein kinases were 0.48 and 1.2/zM, respectively. When the aminoethyl-sulfonamide residue in the H-8 molecule was replaced by sulfonylpiperazine, derivatives such as H-7 became more potent inhibitors of protein kinase C, compared to other derivatives. The inhibitory effect of the compounds on each protein kinase could be overcome by increasing the amounts of ATP added. These findings indicate that the interaction between isoquinolinesulfonamide and the enzyme is freely reversible, and that these compounds increase the apparent Kd value for ATP. Kinetic analysis by double-reciprocal plots revealed that the inhibition of each protein kinase produced by each of the compounds was competitive with respect to ATP and noncompetitive with respect to the phosphate acceptor. Thus, it has been more clearly demonstrated that isoquinolinesulfonamides potently and selectively inhibit cyclic nucleotide-dependent protein kinases and/or protein kinase C. Among them, H-8 was a relatively selective inhibitor of cyclic nucleotide-dependent protein kinase, and H-7 was the most potent inhibitor of protein kinase C. 52 Moreover, it has been established that H-8 penetrates the cell membrane effectively, determined using a sensitive enzyme immunoassay system. The antigen detected in the cells increased in accord with increases in the H-8 concentration, and the apparent Km value was 38/zM. Therefore, this compound can be used for in vivo studies. 52 H-7, which proved to be a potent inhibitor of protein kinase C in vitro, is used to elucidate the physiological significance of protein kinase C and the molecular mechanism of tumor-promoting agents such as 12-O-tetradecanoylphorbol- 13-acetate (TPA). We found that H-7 enhanced serotonin release from human platelets, as induced by TPA, and correspondingly decreased incorporation of radioactive phosphate into a 20,000-Da protein) 3 H-7 has no effect on protein phosphorylation or on serotonin secretion in the unstimulated platelets. A phosphopeptide with a molecular weight of 20,000 has been identified as a light chain (LC20) of platelet myosin and both protein kinase 52 H. Hidaka, M. Inagaki, S. Kawamoto, and Y. Sasaki, Biochemistry 23, 5036 (1984). 53 M. lnagaki, S. Kawamoto, and H. Hidaka, J. Biol. Chem. 259, 14321 (1984).
580
METHODS FOR STUDY OF Ca-BINDING PROTEINS
[44]
C and Ca2+-CaM-dependent myosin light chain kinase have been shown to catalyze the phosphorylation. Two-dimensional peptide mapping following tryptic hydrolysis revealed that H-7 selectively inhibited the protein kinase C-catalyzed phosphorylation of myosin light chain. CaE+-acti vated, phospholipid-dependent myosin light chain phosphorylation may play an inhibitory role in the release reaction of human platelets. Moreover, the addition of TPA to the mast cell suspension prior to stimulation with compound 48/80 was found to prevent full secretion. When mast cells were preincubated in the presence of TPA (10 and 100 ng/ml) for 5 min at 37°, histamine secretion induced by compound 48/80 (0.5/zg/ml) was inhibited to 78 and 65% of the maximal response observed without TPA pretreatment, respectively. There were no significant changes in the secretory activity by treatment with H-7, an inhibitor of protein kinase C. The addition of TPA to the cells previously exposed to H-7 produced an enhanced histamine release compared to the controls with TPA alone. 54 Recently, it was found that H-7 suppresses the phosphorylation of a protein of relative Mr 80,000, which is stimulated by the tumor-promoting phorbol ester, 12-O-tetradecanoylphorbol-13-acetate, in permeable mouse lymphocytes. 55 The phosphorylation of the 80-kDa protein was demonstrated to be dependent both on exogenous added calcium and on the concentration of concanavalin A. Maximal phosphorylation of this protein occurs at 20/.~M added calcium and 4/.Lg/ml concanavalin A and this protein is phosphorylated at a serine residue. These results suggest that it is a substrate for protein kinase C and this phosphorylation may be an element of the concanavalin A signal transduction mechanism. Phosphorylation of a 36,000-Da protein [which may be the fl subunit protein of the immunoglobulin E receptor of rat basopholic leukemia (RBL-2H3) cell membranes] was stimulated by phosphatidylserine and inhibited by H-7. Moreover, H-7 suppressed the release of serotonin from RBL-2H3 cells stimulated with an antigen and TPA. 56 H-7 was found to suppress TPA-induced HTLV-I p19 expression but did not suppress multinucleated cell formation. On the other hand vitamin D3 analogs inhibited both TPA-induced HTLV-I pl9 antigen expression and multinucleotide cell formation with TPA. 57 H-7 selectively restored the proliferation of TPA-treated HL-60 cells and inhibited TPA-induced phenotypic differentiation. H-7 also suppressed 1,25-(OH)ED3- or retinoic acid-induced phenotypic changes of HL-60 cells, but did not eliminate the 54 y . Okano, H. Takagi, S. Nakashima, T. Tohmatsu, and Y. Nozawa, Biochem. Biophys. Res. Commun. 132, l l 0 (1985). 55 G. V. Denis, S. Toyoshima, and T. Osawa, Biochim. Biophys. Acta 885, 136 (1986). 56 R. Teshima, K. Suzuki, H. Ikebuchi, and T. Terao, Mol. lmmunol. 23, 279 (1986). 57 y . Nakao, S. Matsuda, T. Matsui, T. Koizumi, and Y. Ito, Int. J. Cancer 37, 911 (1986).
[44]
INHIBITORS OF C a 2+ SIGNALING
581
growth inhibition by these inducers. These results show that protein kinase C is involved in the phorbol ester-induced phenotypic differentiation of HL-60 cells) 8 The possibility that H-7 can be used as a protein kinase C inhibitor at cellular levels can thus be considered. We developed a simple procedure for the purification of protein kinase C, using affinity chromatography with a new affinity ligand. The adsorbent was synthesized by attaching the amino residue of N-(2-aminoethyl)5-isoquinolinesulfonamide (H-9) to cyanogen bromide-activated Sepharose. 59 H-9 is a potent competitive inhibitor of protein kinase C for ATP. 52A 960-fold purification was achieved by a two-step procedure with DEAE-cellulose and affinity chromatography. The resultant preparation was essentially homogeneous, as indicated by polyacrylamide gel electrophoresis, under conditions of denaturation with sodium dodecyl sulfate. The affinity of protein kinase C for the H-9-Sepharose was high, and the enzyme could not be eluted, either by high concentrations of sodium chloride or by 40% glycerol. H-9 coupled to Sepharose retained protein kinase C and this enzyme could be eluted by the buffer containing Larginine, an inhibitor of the enzyme. Although the dual regulation of the Ca2+-dependent smooth muscle myosin phosphorylation system by Ca2÷-phospholipid and Ca2+-CaM is difficult to analyze, some derivatives of naphthalenesulfonamide were found useful for studies on the molecular mechanism of Ca2÷-dependent myosin light chain phosphorylation by protein kinase C and myosin light chain kinase. Analogs including N-(6-phenylhexyl)-5-chloro-l-naphthalenesulfonamide (SC-9), which has a hydrophobic residue at the end of the hydrocarbon chain, activated Ca2÷-activated, phospholipid-dependent myosin light chain phosphorylation in a calcium-dependent manner. There was no significant effect of these compounds on Ca2÷-CaM-depen dent myosin phosphorylation. 6° SC-9 was similar to phosphatidylserine with regard to stimulation, and the apparent Km values for Ca 2÷ of the enzyme with this activator and phosphatidylserine were 40 and 80/zM, respectively. 6° These findings suggest that SC-9 is a novel and potent synthetic activator of protein kinase C. 6° Recently, it was found that hexose uptake of 3T3 cells was stimulated by SC-9 as well as by TPA. 61 These naphthalenesulfonamide derivatives are proving to be most useful 58 T. Matsui, Y. Nakao, T. Koizumi, Y. Katakami, and T. Fujita, Cancer Res. 46, 583 (1986). 59 M. Inagaki, M. Watanabe, and H. Hidaka, J. Biol. Chem. 260, 2922 (1985). 6o M. Ito, T. Tanaka, M. Inagaki, K. Nakanishi, and H. Hidaka, Biochemistry 25, 4179 (1986). 6~ H. Nishino, K. Kitagawa, A. Iwashima, M. Ito, T. Tanaka, and H. Hidaka, Biochim. Biophys. Acta (in press).
582
METHODS FOR STUDY OF Ca-BINDING PROTEINS
[44]
for elucidating the m e c h a n i s m of activation of protein kinase C. Differentiation b e t w e e n Ca2+-activated, phospholipid-dependent and Ca2+-CaM dependent myosin light chain phosphorylation of smooth muscle is also feasible w h e n these derivatives, including H-7, ML-9, and SC-9, are used. Acknowledgments We thank M. Ohara of Kyushu University for comments on the manuscript. This work was supported in part by a Grant-In-Aid for Scientific Research from the Ministry of Education, Science and Culture, Japan.
M E T H O D S FOR S T U D Y OF C a - B I N D I N G PROTEINS
[44]
shown to be more rapid in serum-free than in serum-supplemented medium. The restricted population of PV-positive neurons seen at later stages in culture is reminiscent to that observed in vioo. It may represent neurons expressing differentially Ca2+-dependent processes. Acknowledgments This work was supported by the Swiss National Science Foundation (3.559.083 and 3.147.085), the Wilhelm Sander, Hartmann M011er,EMDO, Sandoz, CIBA-Geigy,and Julius Klaus-Stiftung, Roche Foundation, and Jubil~umsspende fiir die Universit/RZfirich. We would like to thank Dr. A. M. Rowlerson for critical reading of this manuscript.
[44] T r a n s m e m b r a n e
C a 2+ S i g n a l i n g a n d a N e w C l a s s o f Inhibitors
By HIROYOSHI HIDAKA and TOSHIO TANAKA Introduction Calcium ion plays a critical and central role in various biological events as second messenger.l The intracellular calcium ion is involved in the mechanism of stimulus-induced cellular response such as muscle contraction, metabolic regulation, endo- and exocytosis, cell motility, cytoplasmic transport, cell proliferation, and fertilization. 2 Figure 1 shows a schematic representation of several processes that are involved in this flow of information in the Ca2÷-dependent regulatory system of cell function and selective inhibitors of each process. To investigate molecular mechanisms involved in the calcium messenger system, we developed potent specific inhibitors of each step of the calcium mesenger system, as shown in the model of Fig. 1. These selective inhibitors of each process in the calcium messenger system are discussed herein. N e w Aspects of Calmodulin Antagonists
The pharmacological action of naphthalenesulfonamides 3 and phenothiazines 4 have been well characterized. Since the discovery of the A. R. Means, J. S. Tash, and J. G. Chafouleas, Physiol. Rev. 62, 1 (1982). 2 A. K. Campbell, "Intracellular Calcium," Wiley, New York, 1983. 3 H. Hidaka and T. Tanaka, this series, Vol. 102, p. 185. 4 B. Weiss, this series, Vol. 102, p. 171. METHODS IN ENZYMOLOGY, VOL. 139
Copyright © 1987by Academic Press, Inc. All rights of reproduction in any form reserved.
[44]
INHIBITORS OF Ca 2+ SIGNALING
CalmodulinAntagonists CalciumChannelBlockers nifedipine NO.233 verapamllprenylamlne HT-74 diltiazem bepridi|
571
W-7
~
~
~
CaMBPInhibitors
~MLCK / CaM .,)
\
/
Inhibitor IML-e)
1 ~~ -
C(/~+-PDE~ ~mPDcEeltl~°hir~it
\\ C(a~HA~t~g~i~st
,- .., ~ PK_C~ pK_~Hi.n7~lbitor
Intracellular
~ m l
cell membrane
FIG. 1. Intracellular calcium messenger system and specific inhibitors. CaM, Calmodulin; CaMBP inhibitors, calmodulin-binding protein inhibitors; MLCK, myosin light chain kinase; Ca2+-PDE, Ca2+-dependent cyclic nucleotide phosphodiesterase; PK-C, protein kinase C.
Ca2+-dependent interaction of phenothiazines 5 and naphthalenesulfonamides 6 with calmodulin and the subsequent inhibition of calmodulin activity, a number of calmodulin antagonists have been developed. Included are melittin, 7 mastoparans, 8 calmidazolium, 9 compound 48/80,1° and others. H N~-Dansyl-L-arginine-4-t-butylpiperidine amide (No. 233) was found to be a selective and potent CaM antagonist possessing the fluorescence properties of the dansyl moiety, a useful probe for the study 5 B. Weiss and R. M. Levin, Ado. Cyclic Nucleotide Res. 9, 285 (1978). 6 H. Hidaka, M. Asano, S. Iwadare, I. Matsumoto, T. Totsuka, and N. Aoki, J. Pharmacol. Exp. Ther. 207, 8 (1978). 7 M. Comte, Y. Maulet, and J. A. Cox, Biochem. J. 209, 269 (1983). 8 D. A. Malencik and S. R. Anderson, Biochem. Biophys. Res. Commun. 114, 50 (1983). 9 K. Gietzen, A. Wuthrich, and H. t, ader, Biochem. Biophys. Res. Comrnun. 101, 418 (1981). 10 K. Gietzen, D. Sanchez, and H. Bader, IRCS Med. Sci. 11, 12 (1983). 11 H. Itoh, T. Tanaka, Y. Mitani, and H. Hidaka, Biochem. Pharmacol. 35, 217 (1986).
572
METHODS FOR STUDY OF Ca-BINDING PROTEINS
[44]
of CaM and its interactions. 12Among many of the calmodulin antagonists, naphthalenesulfonamide derivatives have been used for in vivo studies such as cell or tissue level experiments, as they have high specificity for calmodulin and a lower cytotoxicity compared with phenothiazines and also a higher bioavailability. Detailed pharmacological properties of naphthalenesulfonamides became evident in their specific binding to calmodulin, 13 structure-activity relationships, 14 affinity chromatography, 15 and their applications to enzymology~6: and cell biology, as The functions of calmodulin in various cell types have also been studied extensively, using calmodulin antagonists such as naphthalenesulfonamide derivatives. The involvement of calmodulin has been noted in various cellular processes, 19 including cell proliferation, tumor promotion, acrosome reaction of echinoderm sperm, oocyte maturation, neutrophil activation, arterial muscle contraction, platelet function, Ca 2+ transport of plasma membranes, biosyntheses of monoamine neurotransmitters, plant cell function, and others. Moreover, calmodulin antagonists are useful tools for studying the molecular mechanism of calmodulin action and to characterize the binding sites on calmodulin for its target proteins and its antagonists. W e 2° and Laporte et al. 21 independently proposed the possibility that calcium ion induces conformational changes in calmodulin that expose hydrophobic sites on the surface of the molecule and which may act as sites of interaction with calmodulin antagonists and its target proteins. 22 Subsequently, we demonstrated that the affinity of naphthalenesulfonamides23 or local anesthetics 24 for Ca2+-CaM correlated well with their hydrophobicity and \
12 T. Tanaka, M. Inagaki, and H. Hidaka, Arch. Biochem. Biophys. 220, 188 (1983). 13 H. Hidaka, T. Yamaki, M. Naka, T. Tanaka, H. Hayashi, and R. Kobayashi, Mol. Pharmacol. 17, 66 (1980). 14 H. Hidaka, M. Asano, and T. Tanaka, Mol. Pharmacol. 20, 571 (1981). 15T. Endo, T. Tanaka, T. Isobe, H. Kasai, T. Okuyama, and H. Hidaka, J. Biol. Chem. 256, 12485 (1981). 16 H. Hidaka, T. Yamald, T. Totsuka, and M. Asano, Mol. Pharmacol. 15, 49 (1979). 17T. Tanaka, T. Ohmura, T. Yamakado, and H. Hidaka, Mol. Pharmacol. 22, 408 (1982). 18 H. Hidaka, Y. Sasaki, T. Tanaka, T. Endo, S. Ohno, Y. Fujii, and T. Nagata, Proc. Natl. Acad. Sci. U.S.A. 78, 4354 (1981). 19 H. Hidaka and D. J. Hartshorne, "Calmodulin Antagonists and Cellular Physiology." Academic Press, Orlando, Florida, 1985. 2o T. Tanaka and H. Hidaka, J. Biol. Chem. 255, 11078 (1980). 21 D. C. LaPorte, B. M. Wierman, and D. R. Storm, Biochemistry 19, 3814 (1980). 22 H. Hidaka and T. Tanaka, in "Calmodulin and Intracellular Ca ++ Receptors" (S. Kakiuchi, H. Hidaka, and A. R. Means, eds.), p. 19. Plenum, New York, 1982. 23 T. Tanaka, T. Ohmura, and H. Hidaka, Mol. Pharmacol. 22, 403 (1982). ~4 T. Tanaka and H. Hidaka, Biochem. Biophys. Res. Commun. 101, 447 (1981).
[44]
INHIBITORS OF C a 2+ SIGNALING
573
their potency in inhibiting Ca 2+, CaM-dependent enzymes. Recently, Ca2+-dependent hydrophobic interactions between CaM and phenylSepharose 25 or a homologous series of to-aminoalkylagaroses were demonstrated. 26 Moreover, we found that some newly synthesized hydrophobic compounds activate Ca2+-dependent cyclic nucleotide phosphodiesterase. 27 These results suggest that hydrophobic interactions between the Ca2+-CaM complex and enzymes may be important for the enzyme activation by CaM. On the other hand, we found that Ca 2÷dependent cyclic nucleotide phosphodiesterase was activated by polyaspartic acid, an event which suggests that acidic amino acids in the CaM molecule may also play an important role in the stimulation of Ca2+-PDE. A covalent adduct of calmodulin with 1 mol of norchlorpromazine is prepared and this CAPPrCaM complex has the ability to bind to CaMdependent enzymes, with a high affinity, but loses the ability to activate phosphodiesterase and myosin light chain kinase. 28 However, it can activate the phosphatase activity of calcineurin.29 The interaction of calmodulin with Ca 2÷ channel blockers such as bepridil has been reported, n,3° We demonstrated that prenylamine interacts with CaM at sites differing from those related to the naphthalenesulfonamides and phenothiazines. 13,3~ Moreover, we found that a newly synthesized CaM antagonist, 3-(2-benzothiazolyl)-4,5-dimethoxy-N-[3-(4-phenylpiperidinyl)propyl]benzenesulfonamide (HT-74) may bind to CaM in a manner different than heretofore reported and have different effects on Ca2+-induced contractions of depolarized vascular smooth muscle from those of naphthalenesulfonamides. 32 The CaM antagonist R24571 and Ca 2÷ channel blockers prenylamine and diltiazem bind to CaM and potentiate felodipine binding. 33These results suggest that allosteric interactions occur among different drug-binding sites on CaM, hence these compounds should be useful tools for analyzing the functional domains of CaM.
25 R. Gopalakrishna and W. B. Anderson, Biochem. Biophys. Res. Commun. 104, 830 (1982). 26 T. Tanaka, H. Umekawa, T. Ohmura, and H. Hidaka, Biochim. Biophys. Acta 787, 158 (1984). 27 T. Tanaka, E. Yamada, T. Sone, and H. Hidaka, Biochemistry 22, 1030 (1983). 2s D. L. Newton, T. R. Burke, Jr., K. C. Rice, and C. B. Klee, Biochemistry 22, 5472 (1983). 29 D. L. Newton and C. B. Klee, FEBS Lett. 165, 269 (1984). 30 H. Itoh, T. Ishikawa, and H. Hidaka, J. Pharmacol. Exp. Ther. 230, 737 (1984). 31 M. Inagaki and H. Hidaka, Pharmacology 29, 75 (1984). 32 T. Tanaka, H. Umekawa, M. Saitoh, T. Ishikawa, T. Shin, M. Ito, Y. Kawamatsu, H. Sugihara, and H. Hidaka, Mol. Pharmacol. 29, 264 (1986). 33 j. D. Johnson, Biochem. Biophys. Res. Commun. 112, 787 (1983).
574
METHODS FOR STUDY OF Ca-BINDING PROTEINS
[44]
Inhibitors of Ca 2+-, CaM-Dependent Enzymes It is now generally accepted that calmodulin plays a general role in calcium ion signal transduction. W h e n calcium binds to calmodulin, it undergoes structural changes which generate specificinteraction site(s) recognized by various enzymes and proteins. As it is not feasible to use calmodulin antagonists alone for analyzing which calmodulin-depcndent enzyme is responsible for cach Ca2+-dependent cell function, we have continued to make effortsto acquire specificinhibitorsof Ca 2+, calmodufin-dependent enzymes. W e found that vinpocetine [14-ethoxycarbonyl(3a,16a-ethyl)-14,15-eburnamenine] is a potent selective inhibitor of CaE+-dependcnt cyclic nucleotide phosphodiestcrase34 and ML-9 [1-(5chloronaphtalene-l-sulfonyl)-IH-hexahydro-1,4-diazepine] is a specific inhibitor of myosin lightchain kinasc, respectively. W e investigatcd the pharmacological properties of these selectiveinhibitorsof Ca 2+, calmodufin-dependent enzymes and thc physiological rolc of each enzyme in various cell functions.
Modulators of Cae+-Dependent Cyclic Nucleotide Phosphodiesterase Recent studies suggested that cyclic AMP or cyclic GMP plays an important role in regulating Ca2+-dependent cell functions. Cyclic nucleotide phosphodiesterase is the only known pathway for the degradation of cyclic nucleotides) 5 Multiple forms of phosphodiesterase have been found in all the mammalian tissue investigated. 36 Although calmodulin was first found to be a Ca2+-dependent stimulator of cyclic nucleotide phosphodiesterase,37,38 the physiological significance of the calmodulindependent form of phosphodiesterase has yet to be established. Calmodulin antagonists such as W-7 and phenothiazine derivatives selectively inhibit Ca2+-dependent phosphodiesterase, by interacting with the Ca2+-calmodulin complex) :3 However, it has been demonstrated that calmodulin not only activates Ca2÷-dependent phosphodiesterase but also a number of enzymes, including myosin light chain kinase. Therefore, the utility of calmodulin antagonists as tools for elucidation of the physiological role of Ca2+-dependent phosphodiesterase in various cell functions is necessarily limited. Selective inhibitors and activators of Ca2+-dependent cyclic nucleotide phosphodiesterase should be appropriate tools for eluci34 M. Hagiwara, T. Endo, and H. Hidaka, Biochem. Pharmacol. 33, 453 (1984). 35 M. M. Appleman, W. J. Thompson, and T. R. Russell, Ado. Cyclic Nucleotide Res. 3, 65 (1973). 36 j. N. Wells and J. G. Hardman, Adv. Cyclic Nucleotide. Res. 8, 119 (1977). 37 S. Kakiuchi and R. Yamazaki, Biochem. Biophys. Res. Commun. 41, 1104 (1970). 38 W. Y. Cheung, Biochem. Biophys. Res. Commun. 38, 533 (1970).
[44]
INHIBITORS OF Ca 2+ SIGNALING
575
dating and properties and physiological significance of this enzyme. Vinpocetine inhibits selectively Ca2+-dependent phosphodiesterase not by interacting with CaM but rather by interacting with the enzyme directly. The effects of vinpocetine on rabbit aorta cyclic nucleotide phosphodiesterase, separated by DEAE-cellulose column chromatography, were investigated. The three forms of cyclic nucleotide phosphodiesterase observed were cyclic GMP phosphodiesterase (FI), Ca2÷-dependent phosphodiesterase (FII), and cyclic AMP phosphodiesterase (FIII). The results indicated a selective inhibition of Ca2+-dependent phosphodiesterase among these forms of the enzymes, and that the concentrations of vinpocetine which produced 50% inhibition of the Ca2+-dependent phosphodiesterase were about 21/zM, both in the presence and absence of the Ca2+-CaM complex. On the other hand, ICs0 values of this compound for cGMP phosphodiesterase and cAMP phosphodiesterase exceeded 500 /xM. Kinetic analysis by Dixon plots showed noncompetitive inhibition, with respect to cyclic GMP and a Ki value of 14/xM in the presence of the Ca2÷-CaM complex. Thus, vinpocetine may directly inhibit Ca2÷-dependent phosphodiesterase while CaM antagonists such as W-7 antagonize CaM-induced activation of the enzyme. This was tested further by examining the ability of vinpocetine and W-7 to alter the activation of the phosphodiesterase by CaM. Vinpocetine-induced inhibition of phosphodiesterase activity was not overcome by increasing the concentration of CaM and the extent of activation of CaM was greatly inhibited. On the other hand, the W-7-induced inhibition of the enzyme was overcome by increasing the concentration of CaM. These results suggested that vinpocetine inhibits Ca2+-dependent phosphodiesterase by a mechanism differing from those of CaM antagonists. The exposure of K+-depolarized rabbit aorta to 1, 10, and 100/xM vinpocetine resulted in a decrease in tension of aortic strips associated with a significant increase in cyclic GMP levels by 10 and 100/zM of vinpocetine and not statistically with 1 /.~M vinpocetine. There was no increase in cyclic AMP levels associated with relaxation and no increase in guanylate cyclase and adenylate activity in concentrations of vinpocetine ranging from 10 to 1000/xM. These results suggest that the increase in cyclic GMP level may be due to the selective inhibitory effect of vinpocetine on the Ca2+-dependent phosphodiesterase activity. 34 We synthesized HA-542 [1-(2,4-dipiperidino-6-quinazolinesulfonyl)-4(2-ethoxy-2-phenylethyl)piperazine] and HA-543 [1-(2,4-dipiperidino-6quinazolinesulfonyl)-4-cinnamylpiperazine] and found that these compounds are selective activators of Ca2+-dependent cyclic nucleotide phosphodiesterase. 27 As this activation was observed with 2,4-dipiperidino-6-quinazolinesulfonamides but not with 4-piperidino-6-quinazoline-
576
METHODS FOR STUDY OF Ca-BINDING PROTEINS
[44]
sulfonamides, the activation seems to be dependent on the piperidine residue at the 2 and 4 positions of the quinazoline ring and on the extent of hydrophobicity of each compound determined as retention index, using high-performance liquid chromatography. These 2,4-dipiperidino-6quinazoline-sulfonamides activate Ca2+-dependent phosphodiesterase in the absence of the Ca2+-CaM complex and do not further enhance the activity of the phosphodiesterase in the presence of the Ca2÷-CaM complex. These drugs are potent inhibitors of cyclic AMP and cyclic GMP phosphodiesterases. CaM antagonists such as N-(6-aminohexyl)-5chloro-l-naphthalenesulfonamide (W-7) and chlorpromazine inhibited selectively the quinazolinesulfonamide-induced activations of the phosphodiesterase. All these observations suggest that the quinazolinesulfonamides are calcium-independent activators of CaE÷-dependent phosphodiesterase and these compounds are proving to be most useful for studies on the molecular mechanism of CaM action and phosphodiesterase regulation, in vitro. Myosin Light Chain Kinase Inhibitor CaM antagonists such as W-7 or phenothiazine derivatives selectively inhibit myosin light chain kinase through their interaction with the Ca2÷CaM complex. However, CaM is involved in the stimulation of a variety of enzymes other than CaZ÷-dependent cyclic nucleotide phosphodiesterase and CaM seems to have multifunctional roles as a mediator of intracellular calcium ion signals. Therefore, CaM antagonists have limitations as tools for investigating the role of myosin light chain kinase. We found that a newly synthesized compound, 1-(5-chloronaphthalene-l-sulfonyl)-lH-hexahydro-l,4-diazepine (ML-9) is a potent selective inhibitor of myosin light chain kinase. Apparent Ki values of ML-9 for myosin light chain kinase from smooth muscle, cAMP-dependent protein kinase from bovine heart, and protein kinase C from rabbit brain were 3.8, 54, and 32 /xM, respectively. ML-9 also inhibited Caz+, CaM-independent activity of trypsin-treated myosin light chain kinase with a Ki value of 4.0 /xM. Kinetic analysis by the double-reciprocal curve revealed that the inhibition of myosin light chain kinase produced by ML-9 was competitive with respect to ATP and noncompetitive with respect to myosin light chain. Increasing the concentration of ML-9 inhibited selectively the Ca2÷-dependent phosphorylation of human platelet myosin light chain (20 kDa) with an IC50 value of 12/xM. However, ML-9 produced little inhibition of CaZ+-activated, phospholipid-dependent phosphorylation of 40-kDa peptides of human platelets. ML-9 is thus a potent selective inhibitor of myosin light chain kinase from smooth muscle and nonmuscle (platelet) and this compound will aid
[44]
INHIBITORS OF C a 2+ SIGNALING
577
in determining physiological functions of protein kinase in smooth muscle and nonmuscle cell functions. Selective Modulators of Protein Kinase C Calcium ion is an important intracellular messenger involved in the regulation of a variety of cell functions. Although the exact mechanism by which calcium ion exerts its influence has not been clarified, Ca2+-dependent protein phosphorylation is one of the major general mechanisms by which intracellular events in mammalian tissues are controlled by external physiological stimuli) 9 There are at least two different mechanisms related to the regulation of calcium-dependent protein kinases. Some Ca2÷-dependent protein kinases are stimulated by the Ca2÷-CaM complex and another type of Ca2+-dependent protein kinase, which requires phospholipid and diglyceride as cofactors, is a Ca2÷-activated, phospholipiddependent protein kinase (protein kinase C): ° However, the precise relationship between these two types of Ca2+-dependent protein kinases is unknown. CaM-dependent protein kinases such as myosin light chain kinase have been purified and charcterized.4~ We reported that myosin light chain kinase from chicken gizzard and rabbit skeletal muscle can be stimulated by limited proteolysis as well as by the Ca2+-CaM complex.42 Our data suggest that the limited proteolysis of CaM-dependent myosin kinase not only stimulates the activity of the kinase but also converts the kinase to the Ca2+-CaM-insensitive form. Protein kinase C has also been reported to be activated by limited proteolysis and converted to the Ca2÷phospholipid-insensitive form of this kinase. Myosin from smooth muscle and nonmuscle but not from skeletal and cardiac muscle can serve as a substrate for protein kinase C as well as for the Ca2+-CaM-dependent myosin light chain kinase: 3,44 When unphosphorylated smooth muscle heavy meromyosin (HMM) is phosphorylated by myosin light chain kinase, its actin-activated myosin ATPase activity increases. 45 We found that protein kinase C and myosin light chain kinase catalyzes the phosphorylation of different sites within the 20,000-Da light 39 H. Rasmussen, L Kojima, K. Kojima, W. Zawalich, and W. Apfeldorf, Adv. Cyclic Nucleotide Protein Phosphorylation Res. 18, 159 (1984). 4o y . Takai, A. Kishimoto, Y. Kawahara, R. Minakuchi, K. Sano, V. Kikkawa, T. Mori, B. Yu, K. Kaibuchi, and Y. Nishizuka, Adv. Cyclic Nucleotide Res. 14, 301 (1981). 41 C. B. Klee and T. C. Vanaman, Adv. Protein Chem. 15, 213 (1982). 42 T. Tanaka, M. Naka, and H. Hidaka, Biochem. Biophys. Res. Commun. 92, 313 (1980). 43 T. Endo, M. Naka, and H. Hidaka, Biochem. Biophys. Res. Commun. 105, 942 (1982). 44 M. Naka, M. Nishikawa, R. S. Adelstein, and H. Hidaka, Nature (London) 306, 490 (1983). 45 R. S. Adelstein, J. R. Sellers, M. A. Conti, M. D. Pato, and P. DeLanerolle, Fed. Proc., Fed. Am. Soc. Exp. Biol. 41, 2873 (1982).
578
METHODS FOR STUDY OF Ca-BINDING PROTEINS
[44]
[ C ,modu,n I MLCK
It
MEG @
Fro. 2. Regulation of actomyosin in smooth muscle and nonmuscle cells. PK-A, cAMPdependent protein kinase; MLCK, myosin light chain kinase; MLC, myosin light chain; PK-C, protein kinase C.
chain of smooth muscle HMM and that sequential phosphorylation of HMM by myosin light chain kinase and protein kinase C decreases the actin-activated Mg-ATPase activity, as compared to that following phosphorylation by myosin light chain kinase alone. 46,47 Moreover, we reported that protein kinase C incorporated phosphate into two sites of myosin light chain kinase and that this phosphorylation resulted in a reduced affinity for calmodulin.48 This dual regulation of the Ca2+-dependent myosin light chain phosphorylation system by Ca2+-phospholipid and Ca2+-CaM is complex to analyze (Fig. 2). Almost all CaM antagonists are hydrophobic and basic in property, thus it is not surprising that all of the CaM antagonists reported can inhibit the activity of protein kinase C. 49-51All these findings indicate that it is difficult to elucidate the physiological significance of protein kinase C and the interrelationship between the two putative Ca2÷-dependent protein phosphorylation systems using CaM antagonists. Now when the naphthalene ring of the CaM antagonists, naphthalenesulfonamides, is replaced by isoquinoline, the deriva46 M. Nishikawa, H. Hidaka, and R. S. Adelstein, J. Biol. Chem. 258, 14069 (1983). 47 M. Nishikawa, J. R. Sellers, R. S. Adelstein, and H. Hidaka, J. Biol. Chem. 259, 8808 (1984). M. Ikebe, M. Inagaki, K. Kanamaru, and H. Hidaka, J. Biol. Chem. 260, 4547 (1985). 49 T. Mori, Y. Takai, R. Minakuchi, B. Yu, and Y. Nishizuka, J. Biol. Chem. 255, 8378 (1980). 5o R. C. Schatzman, B. C. Wise, and J. F. Kuo, Biochem. Biophys. Res. Commun. 92, 313 (1980). 51 D e - F , Qi, R. C. Schatzman, G. J. Mazzei, R. S. Turner, R. L. Raylor, S. Liao, and H. Rasmussen, in "Calcium and Cell Function" (W. Y. Cheung, ed.), Vol. 4, p. 1. Academic Press, New York, 1983.
[44]
INHIBITORS OF C a 2+ SIGNALING
579
tives are no longer CaM- and phospholipid-interacting agents but rather suppress protein kinase C, directly. Among them, l-(5-isoquinolinesulfonyl)-2-methylpiperazine, refered to as H-7, is a relatively selective inhibitor of protein kinase C.52 The Ki value of H-7 for protein kinase C is 6.0 /zM. On the other hand, N-[2-(methylamino)ethyl]-5-isoquinolinesulfonamide (H-8) seems to be the most active of the inhibitors examined and inhibits more markedly the cGMP-dependent and cAMP-dependent protein kinases than do the other kinases and ATPases examined. The Ki values of H-8 for cGMP-dependent and cAMP-dependent protein kinases were 0.48 and 1.2/zM, respectively. When the aminoethyl-sulfonamide residue in the H-8 molecule was replaced by sulfonylpiperazine, derivatives such as H-7 became more potent inhibitors of protein kinase C, compared to other derivatives. The inhibitory effect of the compounds on each protein kinase could be overcome by increasing the amounts of ATP added. These findings indicate that the interaction between isoquinolinesulfonamide and the enzyme is freely reversible, and that these compounds increase the apparent Kd value for ATP. Kinetic analysis by double-reciprocal plots revealed that the inhibition of each protein kinase produced by each of the compounds was competitive with respect to ATP and noncompetitive with respect to the phosphate acceptor. Thus, it has been more clearly demonstrated that isoquinolinesulfonamides potently and selectively inhibit cyclic nucleotide-dependent protein kinases and/or protein kinase C. Among them, H-8 was a relatively selective inhibitor of cyclic nucleotide-dependent protein kinase, and H-7 was the most potent inhibitor of protein kinase C. 52 Moreover, it has been established that H-8 penetrates the cell membrane effectively, determined using a sensitive enzyme immunoassay system. The antigen detected in the cells increased in accord with increases in the H-8 concentration, and the apparent Km value was 38/zM. Therefore, this compound can be used for in vivo studies. 52 H-7, which proved to be a potent inhibitor of protein kinase C in vitro, is used to elucidate the physiological significance of protein kinase C and the molecular mechanism of tumor-promoting agents such as 12-O-tetradecanoylphorbol- 13-acetate (TPA). We found that H-7 enhanced serotonin release from human platelets, as induced by TPA, and correspondingly decreased incorporation of radioactive phosphate into a 20,000-Da protein) 3 H-7 has no effect on protein phosphorylation or on serotonin secretion in the unstimulated platelets. A phosphopeptide with a molecular weight of 20,000 has been identified as a light chain (LC20) of platelet myosin and both protein kinase 52 H. Hidaka, M. Inagaki, S. Kawamoto, and Y. Sasaki, Biochemistry 23, 5036 (1984). 53 M. lnagaki, S. Kawamoto, and H. Hidaka, J. Biol. Chem. 259, 14321 (1984).
580
METHODS FOR STUDY OF Ca-BINDING PROTEINS
[44]
C and Ca2+-CaM-dependent myosin light chain kinase have been shown to catalyze the phosphorylation. Two-dimensional peptide mapping following tryptic hydrolysis revealed that H-7 selectively inhibited the protein kinase C-catalyzed phosphorylation of myosin light chain. CaE+-acti vated, phospholipid-dependent myosin light chain phosphorylation may play an inhibitory role in the release reaction of human platelets. Moreover, the addition of TPA to the mast cell suspension prior to stimulation with compound 48/80 was found to prevent full secretion. When mast cells were preincubated in the presence of TPA (10 and 100 ng/ml) for 5 min at 37°, histamine secretion induced by compound 48/80 (0.5/zg/ml) was inhibited to 78 and 65% of the maximal response observed without TPA pretreatment, respectively. There were no significant changes in the secretory activity by treatment with H-7, an inhibitor of protein kinase C. The addition of TPA to the cells previously exposed to H-7 produced an enhanced histamine release compared to the controls with TPA alone. 54 Recently, it was found that H-7 suppresses the phosphorylation of a protein of relative Mr 80,000, which is stimulated by the tumor-promoting phorbol ester, 12-O-tetradecanoylphorbol-13-acetate, in permeable mouse lymphocytes. 55 The phosphorylation of the 80-kDa protein was demonstrated to be dependent both on exogenous added calcium and on the concentration of concanavalin A. Maximal phosphorylation of this protein occurs at 20/.~M added calcium and 4/.Lg/ml concanavalin A and this protein is phosphorylated at a serine residue. These results suggest that it is a substrate for protein kinase C and this phosphorylation may be an element of the concanavalin A signal transduction mechanism. Phosphorylation of a 36,000-Da protein [which may be the fl subunit protein of the immunoglobulin E receptor of rat basopholic leukemia (RBL-2H3) cell membranes] was stimulated by phosphatidylserine and inhibited by H-7. Moreover, H-7 suppressed the release of serotonin from RBL-2H3 cells stimulated with an antigen and TPA. 56 H-7 was found to suppress TPA-induced HTLV-I p19 expression but did not suppress multinucleated cell formation. On the other hand vitamin D3 analogs inhibited both TPA-induced HTLV-I pl9 antigen expression and multinucleotide cell formation with TPA. 57 H-7 selectively restored the proliferation of TPA-treated HL-60 cells and inhibited TPA-induced phenotypic differentiation. H-7 also suppressed 1,25-(OH)ED3- or retinoic acid-induced phenotypic changes of HL-60 cells, but did not eliminate the 54 y . Okano, H. Takagi, S. Nakashima, T. Tohmatsu, and Y. Nozawa, Biochem. Biophys. Res. Commun. 132, l l 0 (1985). 55 G. V. Denis, S. Toyoshima, and T. Osawa, Biochim. Biophys. Acta 885, 136 (1986). 56 R. Teshima, K. Suzuki, H. Ikebuchi, and T. Terao, Mol. lmmunol. 23, 279 (1986). 57 y . Nakao, S. Matsuda, T. Matsui, T. Koizumi, and Y. Ito, Int. J. Cancer 37, 911 (1986).
[44]
INHIBITORS OF C a 2+ SIGNALING
581
growth inhibition by these inducers. These results show that protein kinase C is involved in the phorbol ester-induced phenotypic differentiation of HL-60 cells) 8 The possibility that H-7 can be used as a protein kinase C inhibitor at cellular levels can thus be considered. We developed a simple procedure for the purification of protein kinase C, using affinity chromatography with a new affinity ligand. The adsorbent was synthesized by attaching the amino residue of N-(2-aminoethyl)5-isoquinolinesulfonamide (H-9) to cyanogen bromide-activated Sepharose. 59 H-9 is a potent competitive inhibitor of protein kinase C for ATP. 52A 960-fold purification was achieved by a two-step procedure with DEAE-cellulose and affinity chromatography. The resultant preparation was essentially homogeneous, as indicated by polyacrylamide gel electrophoresis, under conditions of denaturation with sodium dodecyl sulfate. The affinity of protein kinase C for the H-9-Sepharose was high, and the enzyme could not be eluted, either by high concentrations of sodium chloride or by 40% glycerol. H-9 coupled to Sepharose retained protein kinase C and this enzyme could be eluted by the buffer containing Larginine, an inhibitor of the enzyme. Although the dual regulation of the Ca2+-dependent smooth muscle myosin phosphorylation system by Ca2÷-phospholipid and Ca2+-CaM is difficult to analyze, some derivatives of naphthalenesulfonamide were found useful for studies on the molecular mechanism of Ca2÷-dependent myosin light chain phosphorylation by protein kinase C and myosin light chain kinase. Analogs including N-(6-phenylhexyl)-5-chloro-l-naphthalenesulfonamide (SC-9), which has a hydrophobic residue at the end of the hydrocarbon chain, activated Ca2÷-activated, phospholipid-dependent myosin light chain phosphorylation in a calcium-dependent manner. There was no significant effect of these compounds on Ca2÷-CaM-depen dent myosin phosphorylation. 6° SC-9 was similar to phosphatidylserine with regard to stimulation, and the apparent Km values for Ca 2÷ of the enzyme with this activator and phosphatidylserine were 40 and 80/zM, respectively. 6° These findings suggest that SC-9 is a novel and potent synthetic activator of protein kinase C. 6° Recently, it was found that hexose uptake of 3T3 cells was stimulated by SC-9 as well as by TPA. 61 These naphthalenesulfonamide derivatives are proving to be most useful 58 T. Matsui, Y. Nakao, T. Koizumi, Y. Katakami, and T. Fujita, Cancer Res. 46, 583 (1986). 59 M. Inagaki, M. Watanabe, and H. Hidaka, J. Biol. Chem. 260, 2922 (1985). 6o M. Ito, T. Tanaka, M. Inagaki, K. Nakanishi, and H. Hidaka, Biochemistry 25, 4179 (1986). 6~ H. Nishino, K. Kitagawa, A. Iwashima, M. Ito, T. Tanaka, and H. Hidaka, Biochim. Biophys. Acta (in press).
582
METHODS FOR STUDY OF Ca-BINDING PROTEINS
[44]
for elucidating the m e c h a n i s m of activation of protein kinase C. Differentiation b e t w e e n Ca2+-activated, phospholipid-dependent and Ca2+-CaM dependent myosin light chain phosphorylation of smooth muscle is also feasible w h e n these derivatives, including H-7, ML-9, and SC-9, are used. Acknowledgments We thank M. Ohara of Kyushu University for comments on the manuscript. This work was supported in part by a Grant-In-Aid for Scientific Research from the Ministry of Education, Science and Culture, Japan.