Disulfonic Stilbene Derivatives Open the Ca2+ Release Channel of Sarcoplasmic Reticulum1

Transcription

1 J. Biochem. 106, (1989) Disulfonic Stilbene Derivatives Open the Ca2+ Release Channel of Sarcoplasmic Reticulum1 Takashi Kawasaki and Michiki Kasai Department of Biophysical Engineering, Faculty of Engineering Science, Osaka University, Toyonaka, Osaka 560 Received for publication, April 14, 1989 Ca2+ channels of isolated sarcoplasmic reticulum were incorporated into a planar lipid bilayer and their pharmacological properties were studied. The results show that the channel is a Ca2+ -induced Ca2+ release channel like that observed in skinned muscle fibers and isolated vesicles. (i) The open channel probability was increased by the addition of micromolar amounts of Ca2+ to the cis (myoplasmic) side and further increased by millimolar ATP. (ii) The channel was closed by millimolar Mg2+ and micromolar ruthenium red. We found that two disulfonic stilbene derivatives, 4,4'-diisothiocyanostilbene- 2,2'-disulfonic acid (DIDS) and 4-acetoamido-4'-isothiocyanostilbene-2,2'-disulfonic acid (SITS), when added to the cis side open the channel and lock it irreversibly at open without changing the single channel conductance. Ca2+ efflux from SR vesicles was also enhanced by SITS and DIDS, as monitored by a tracer assay. Further, Ag+ activated the channel transiently. These results suggest that certain amino and SH residues play important roles in gating the Ca2+ channel. Calcium ions released from sarcoplasmic reticulum (SR) trigger muscle contraction in excitation-contraction (E-C) coupling. The most probable candidate is the Ca2+ -induced Ca2+ release channel. In recent years studies of the Ca2+ channel have been far advanced by Meissner and his colleagues using planar bilayer systems (1-4). They have pharmacologically characterized the Ca2+ channel as being activated by micromolar Ca2+ and millimolar ATP and inhibited by millimolar Mg2+ and micromolar ruthenium red (2-4). Furthermore, ryanodine, a neutral alkaloid isolated from the plant Ryania speciosa Vahl, was found to bind irreversibly to the Ca2+ channel (5-7) and was used as a probe for the isolation of the channel molecules (8-10). Interestingly, ryanodine at low concentrations (<10ƒÊM) activates and locks the channel open, but decreases its single channel conductance to about half of the unit conductance obtained at activation by ATP (4, 6, 7, 11, 12). The mechanism, however, is not clear. Further increase of the ryanodine concentration (>100ƒÊM) closes the channel (5, 8, 13). In this study, we first confirmed the pharmacological properties of the Ca2+ channel as a Ca2+-induced Ca2+ release channel in a planar bilayer, and then studied extensively the effects of two disulfonic stilbene derivatives, 4,4'-diisothiocyanostilbene-2,2'-disulfonic acid (DIDS) and 4-acetoamido-4'-isothiocyanostilbene-2,2'-disulfonic acid (SITS), which are known as amino modifiers. It was found that the disulfonic stilbene derivatives opened the Ca2+ channel and locked it open without decreasing the 1 This work was supported in part by a Grant-in-Aid for Scientific Research on Priority Areas of Transmembrane Control to M. K. from the Ministry of Education, Science and Culture of Japan. Abbreviations: SR, sarcoplasmic reticulum; HSR, heavy fraction of fragmented SR; DIDS, 4,4'-diisothiocyanostilbene-2,2'-disulfonic acid; SITS, 4-acetoamido-4'-isothiocyanostilbene-2,2'-disulfonic acid. single channel conductance. This effect is different from that of ryanodine. The effect of the disulfonic stilbene derivatives was also confirmed by the tracer assay of45 Ca efflux from isolated SR vesicles. Further, we confirmed that Ag+ triggers Ca2+ release from SR by activating the Ca2+ -induced Ca2+ release channel (14). These results suggest that certain amino and SH residues play important roles in gating the Ca2+ channel. MATERIALS AND METHODS Isolation of HSR Since the Ca2+ channel is abundant in the heavy fraction of SR vesicles (HSR), HSR was used throughout these experiments. HSR, for planar bilayer measurement was prepared according to the method described by Meissner (15) with slight modification. It was ultimately obtained from the 36-45% region of sucrose gradients that contained membranes sedimented at 2,600-35,000 ~g. HSR for measurement of Ca2+ efflux was prepared as described previously (16). Reagents-DIDS and SITS were purchased from Pierce chemical, U.S.A.; AgNO3 from Wako Chemical Industries, Osaka and ATP, ruthenium red, asolectin from Sigma Chemical, U.S.A. 45Ca (CaCl2) was purchased from New England Nuclear, U.S.A. All other reagents were of reagent grade. Planar Bilayer Measurement Planar bilayer measurement to observe the current through the Ca2+ channel of SR was performed according to the method of Smith et al. (2) with slight modification. The procedure is as follows. A Mueller-Rudin planar lipid bilayer which contained asolectin was formed. Fusion events of SR vesicles were monitored by stpe-like current due to incorporation of Cl- channel co-existing with the Ca2+ channel in native SR vesicles under the following solvent conditions: 250mM choline-cl, 3mM CaCl2, 10mM HEPES-Tris (ph 7.4) in the cis Vol.106, No.3,

2 402 T. Kawasaki and M. Kasai chamber and 53 mm Ba(OH)2, 250mM HEPES (ph 7.4) in the trans chamber. After the fusion the cis chamber was perfused with 125mM Tris, 250mM HEPES, and 2ƒÊM free Ca2+. The trans chamber was defined as electrical ground. Data were recorded on video tape using a modified audio processor and a video cassette recorder. All experiments were performed at room temperature. Measurement of Ca2+ Efflux-Ca+ efflux from SR vesicles was measured by the Millipore filtration method described previously (17): Briefly, concentrated SR vesicles were incubated for about 3h at room temperature in a solution containing 45Ca. The mixture was diluted more than 100 times with a dilution medium described in the legend to Fig.5. Then, at each measurement time, 1ml of the diluted mixture was rapidly filtered through a Millipore filter (HAWP 02500, Millipore, U.S.A.) and washed with 3 ml of the dilution medium. The radioactivity remaining on the filter was measured in a liquid scintillation counter (LS-500, Horiba, Kyoto). blocked one channel, but another channel remained open (Fig.3D). The channel blocked by ruthenium red might be the one activated by non-covalently bound SITS (see "DISCUSSION"). Blockers hardly affected the Ca2+ channels activated by SITS. DIDS was also effective in the same manner as SITS (Fig.4). These effects of SITS and DIDS were not reversed when the cis solution was perfused with a solution containing no disulfonic stilbene derivatives (Fig. 4B). The single channel conductance activated by DIDS was decreased by the addition of ryanodine (Fig.4C). Decrease of the single channel conductance by ryanodine was observed by Meissner et al. (4-7, 11, 12). The I-V relationship of the Ca2+ channel after being locked open by DIDS is shown in Fig. 2B. Although the reversal potential was the same as that in Fig. 2A, the slope conductance was slightly higher, 142 ps. In the case of Fig. 2A, the single channel RESULTS Effects of ATP, Mg2+, and Ruthenium Red-The effects of various reagents on the Ca2+ -induced Ca2+ release channel were examined in the bilayer system. After fusion of the vesicles with the lipid bilayer, the cis side was perfused with a solution containing 2ƒÊM free Ca2+. Open events of the Ca2+ channel (flickering) were observed frequently (Fig.1, Aa and Ba). The open probability was further increased by the addition of millimolar ATP (Ab and Bb). The fluctuation after the addition of ATP was such that it was easier to distinguish the open state from the closed state, because of the increase in the open probability. Mg2+ and/or ruthenium red inhibited the channel opened by ATP (Ac, Ad, Bc, and Bd). Mg2+ remarkably decreased the open probability at 500ƒÊM (Ac), and then the channel stayed closed most of the time at 5mM Mg2+ (Ad). Similarly, ruthenium red remarkably decreased the open probability at low micromolar concentrations: the mean number of open channels decreased from 2 to 1 at 2.5 (Bc) and the channel never opened at 7.5ƒÊM (Bd). The effect of ruthenium red could not be reversed by perfusion of the cis side with a solution containing 2ƒÊM free Ca2+ alone. All the agents used were effective only from the cis side, and did not affect the single channel conductance but rather the open probability. Figure 2A shows the I-V relationship of the single channel current recorded in the presence of 1mM ATP in the cis solution. The slope conductance of the channel in Fig. 2A was about 135 ps and the reversal potential was 30 mv. Since our condition was an asymmetric one, i. e., 125 mm Tris on the cis side and 53mM Ba2+ on the trans side, the permeability ratio could be obtained as PBa/PTris=8.3 (3). These results were consistent with those reported by Meissner and his colleagues (1-3). Effects of Disulfonic Stilbene Derivatives Figure 3 illustrates the effects of SITS on Ca2+ channels in the planar bilayer. Two Ca2+ channels were remarkably activated by 100ƒÊM SITS added to the cis solution. The open probability became nearly unity, but the single channel conductance was unchanged. Mg2+ (5mM) scarcely affected the open probability but reduced the single channel conductance (Fig.3C). Further, addition of 20ƒÊM ruthenium red Fig.1. Effects of ATP, ruthenium red, and Mg2+ on the Ca2+ channel. The trans solution contained 53mM Ba(OH)2 and 250mM HEPES (ph 7.4), and the cis solution contained 125mM Tris and 250 mm HEPES (ph 7.4). All effectors were added to the cis solution and their final concentrations are shown. Recordings (A) and (B) were obtained from different preparations. (Aa, Ba) 2ƒÊM free Ca2+; (Ab, Bb) 1mM ATP was added to Aa and Ba; (Ac) 500ƒÊM MgSO4 was added to Ab; (Ad) 5mM MgSO4 was added to Ab; (Bc) 2.5ƒÊM ruthenium red was added to Bb; (Bd) 7.5ƒÊM ruthenium red was added to Bb. The holding potential was 0mV in all cases. The arrow in each trace indicates the zero current level. The upward direction is negative current. J. Biochem.

3 Ca2+ Channel of SR 403 Fig.2. I-V relationships of the Ca2+ channel. The data were taken from single channel recordings similar to Figs. 1 and 4. (A) In the presence of 2ƒÊM free Ca2+ and 1mM ATP in the cis solution. (B) In the presence of 100ƒÊM DIDS in the cis solution. The slope conductance was 135pS in A and 142 ps in B. The reversal potentials were about 30mV in both cases. Fig.3. Effects of SITS on Ca2+ channel. Experiments similar to those in Fig.1 were carried out. (A) One hundred micromolar SITS was added to the cis solution containing 2ƒÊM free Ca2+ at the point indicated by the arrow. (B) Twenty minutes after the addition of 100 M SITS. The two levels of the open channels are indicated by the dotted lines. (C) Five millimolar MgSO4 was added to the cis solution of B. A decrease in the single channel conductance was seen. (D) Twenty micromolar ruthenium red was added to C. Fig.4. Effects of ryanodine on the Ca2+ channel activated by DIDS. Experiments similar to those of Fig.1 were carried out. (A) One hundred micromolar DIDS was added to the cis solution containing 2ƒÊM free Ca2+ at the point indicated by the arrow. (B) Recording after perfusion of the cis side with a solution containing 2ƒÊM free Ca2+, indicating that the activation by DIDS could not be reversed by perfusion. (C) Ten micromolar ryanodine was added to the cis solution of B. The single channel conductance decreased about 30s after the addition of ryanodine. The dotted lines indicate the zero current levels. conductance might have been underestimated because of the flickering. These results indicate that SITS and DIDS locked the channel in the fully open state. They were not effective when added to the trans side. Next the effects of the disulfonic stilbene derivatives on Ca2+ permeability of isolated membrane vesicles were studied by the tracer assay. Figure 5 shows the time course of Ca+ efflux from HSR in the presence of various concentrations of SITS. SITS was added to the dilution medium of the flux measurement in Fig. 5A and was incubated with samples for 3h before the flux measurement in Fig. 5B. In both cases, Ca2+ release was accelerated with the increase of SITS concentration. Similar effects were observed with DIDS (data not shown). Effect of Ag+-In the previous paper (14), we showed that Ag+ increased Ca2+ release from isolated HSR. In this paper the effect of Ag+ was examined in a bilayer system. Ca2+ channels were transiently activated by the addition of Ag+ to the cis solution (Fig.6). The activated state persisted for a few minutes; then the channels spontaneously closed. The effect of Ag+ was irreversible. The channels closed by Ag+ were not reopened by ATP. These effects of Ag+ were observed at more than 0.3ƒÊM Ag+. In addition, the channels activated by disulfonic stilbene derivatives were not closed by the addition of Ag+ (data not shown). DISCUSSION Various pharmacological properties of the Ca2+ release channel of SR were studied in a lipid bilayer system. Vol.106, No.3, 1989

4 404 T. Kawasaki and M. Kasai Fig.5. Increase of Ca2+ efflux of SR vesicles by SITS. (A) The effect of the addition of SITS to the Ca2+ releasing medium. SR vesicles were incubated in 0.1M KCl, 1mM CaCl2, 5mM K-HEPES (ph 6.5), a small amount of 45CaCl2, and 20mg protein/ml for 3h at room temperature. The mixture was then diluted 150 times with a dilution medium containing 0.1M KCl, 2mM MgCl2, 5mM K-HEPES (ph 6.5), and various concentrations of SITS, and the Ca2+ efflux was followed. The SITS concentration of the dilution medium was as follows: ( œ ) 0, ( ) 30ƒÊM, ( ) 100ƒÊM, ( ) 300ƒÊM, ( ) 1mM. (B) Effect of SITS treatment of SR vesicles. SR vesicles were first treated for 10min at 37 Ž and incubated for 3h at room temperature with various concentrations of SITS in 0.1M KCl, 1mM CaCl2, 5mM K-HEPES (ph 6.5), a small amount of 45CaCl 2, and 20mg protein/ml. The incubated mixture was diluted 110 times with a dilution medium containing 0.1M KCl, 2mM MgCl2, and 5mM K-HEPES (ph 6.5), and the Ca2+ efflux was followed. The SITS concentration during treatment was as follows: ( œ) 0, ( ) 0.53mM, ( ) 1.6mM. SITS concentrations during flux measurement were 110 times lower than these values. Fig.6. Effect of Ag+ on the Ca2+ channel. Experiments similar to those in Fig.1 were carried out. (A) Five micromolar and (B) 3ƒÊM Ag+ were added at the points indicated by the arrows to the cis solutions containing 2ƒÊM free Ca2+. After activation of the Ca2+ channels, they closed spontaneously. The solid lines indicate the zero current levels. Almost all the properties of the Ca2+ -induced Ca2+ release channel observed in isolated vesicles and in skinned muscle fibers were confirmed. However, some discrepancies were seen. One is the dependence of the activation on Ca2+ concentration. In the lipid bilayer system, only slight activation due to 2ƒÊM Ca2+ was observed. Up to a few hundred micromolar concentrations the greater the concentration of Ca2+, the more the channel was activated (data not shown). The optimum concentration of Ca2+ in the bilayer system (a few hundred ƒêm, data not shown) was more than that obtained from experiments in native vesicles (14, 17) or in skinned fibers (18). One noteworthy point about the bilayer system is that the experimental results were not always reproducible. For example, when we used the same concentrations of Ca2+ or ATP, the degree of activation differed from preparation to preparation, and in some cases no activation by Ca2+ and ATP was observed. Even in the latter cases, activation by disulfonic stilbene derivatives or ryanodine was observed. It is probable that the channel proteins or their environment were damaged or some important components were lost in the bilayer experiments. Since we were able to establish the bilayer system to monitor the Ca2+ channel activity, the effects of disulfonic stilbene derivatives, which are known as amino modifiers and blockers of anion exchanger in red blood cells (19), were examined. Because we had previously found that these compounds block the channel in SR (20-22), open the K+ channel in SR (23), and increase the Ca2+ flux from SR vesicles (21), we expected that they might modify the gate of the Ca2+ channel. As expected, SITS and DIDS were effective as openers of the Ca2+ channel. When added to the cis side, they increased the open probability of the channel nearly to unity and locked it open without changing the single channel conductance. Moreover, the channel opened by the disulfonic stilbene derivatives was not easy closed by blockers. They opened even the channel that could not be activated by ATP. Although an increase in Ca2+ permeability due to disulfonic stilbene derivatives was observed in a vesicle system as well as in the lipid bilayer system, some differences were seen between them. Firstly, it should be noted that the effect of SITS was instantaneous in Fig. 5A, while some time-lag was observed in the bilayer experiments as in Figs.3 and 4. The delay might be attributable to the time required for diffusion of the added SITS to the vicinity of the bilayer. Secondly, in the vesicle system, we carried out two types of experiments: dilution (Fig.5A) and incubation J. Biochem.

5 Ca2+ Channel of SR 405 (Fig. 5B). We considered that the effect of non-covalently bound SITS was dominant in the dilution experiment since the effect of SITS was instantaneous and was partially reversed by further dilution of the diluted samples (data not shown). On the other hand, we considered that the effect of covalently bound SITS was realized in the incubation experiment because the concentrations of SITS during the flux measurement was much smaller than those in Fig.5A and the effect could not be reversed by further dilution (data not shown). Thus, the data show that the disulfonic stilbene derivatives could open the channel in both the non-covalently and covalently bound states. Similar effectiveness of disulfonic stilbene derivatives in both states was reported for blocking of the anion channel in SR (21). The fact that the effects could not be reversed by perfusion of the cis solution in the bilayer experiment suggests that disulfonic stilbene derivatives bind covalently to the gate of the channel, probably to amino residues of the channel protein. The activated state of the channel observed in Fig. 3C, which could be closed by ruthenium red, may be attributable to the non-covalent binding of SITS. However, no further quantitative comparison between two systems was done. It should be noted that although the activation by SITS and DIDS resembles that by ryanodine, the reduction of the single channel conductance was not accompanied in the former cases. Disulfonic stilbene derivatives may be the strongest opener of the Ca2+ release channel of SR. It was reported that disulfonic stilbene derivatives open cation channels in tissues other than SR, such as the K+ channel in squid giant axon (24) and the K+ channel in the vacuolar membrane of yeast (25). The results suggest contributions of important amino residues to the gating of the cation channels. Disulfonic stilbene derivatives will be good tools for analysis of the gating mechanism. In the previous papers, we showed that Ag+ increased the Ca2+ efflux from SR vesicles by affecting the Ca2+-induced Ca2+ release channel (14) and that some SH reagents, such as 4,4'-dithiodipyridine, activated the Ca2+ release channel (16). Ag+ is thought to interact with SH residues in the Ca2+ release channel (14, 26, 27). In the present study, Ag+ was shown to activate the Ca2+ channel transiently (Fig.6). Since the SH reagent, 4,4'-dithiodipyridine, did not block the channel, it is probable that Ag+ binds further to a different site and closes the channel. Previous results together with the above findings suggest that certain amino and SH residues play important roles in gating the Ca2+ release channel in SR. REFERENCES 1. Smith, J. S., Coronado, R., & Meissner, G. (1985) Nature 316, Smith, J. S., Coronado, R., & Meissner, G. (1986) J. Gen. Physiol. 88, Smith, J. S., Coronado, R., & Meissner, G. (1986) Biophys. J. 50, Rousseau, E., Smith, J. S., & Meissner, G. (1987) Am. J. Physiol. 253, C364-C Feher, J. J. & Lipford, G. B. (1985) Biochim. Biophys. Acta 813, Fleischer, S., Ogunbunmi, E. M., Dixon, M. K., & Fleer, E. A. M. (1985) Proc. Natl. Acad. Sci. U. S. 82, Lattanzio, F. A., Schlatterer, R. G., Nicar, M., Campbell, K. P., & Sutko, J. L. (1987) J. Biol. Chem. 262, Inui, M., Saito, A., & Fleischer, S. (1987) J. Biol. Chem. 262, Imagawa, T., Smith, J. S., Coronado, R., & Campbell, K. P. (1987) J. Biol. Chem. 262, Lai, F. A., Erickson, H. P., Rousseau, E., Liu, Q.-Y., & Meissner, G. (1988) Nature 331, Meissner, G. (1986) J. Biol. Chem. 261, Pessah, I. N., Waterhouse, A. L., & Casida, J. E. (1985) Biochem. Biophys. Res. Commun. 128, Seiler, T., Wegener, A. D., Whang, D. D., Hathaway, D. R., & Jones, L. R. (1985) J. Biol. Chem. 259, Tatsumi, S., Suzuno, M., Taguchi, T., & Kasai, M. (1988) J. Biochem. 104, Meissner, G. (1984) J. Biol. Chem. 259, Nagura, S., Kawasaki, T., Taguchi, T., & Kasai, M. (1988) J. Biochem. 104, Nagasaki, K. & Kasai, M. (1981) J. Biochem. 90, Endo, M. (1985) in Structure and Function of Sarcoplasmic Reticulum (Fleischer, S. & Tonomura, Y., eds.) pp , Academic Press, Orlando 19. Cabantchik, Z.I., Knauf, P. A., & Rothstein, A. (1979) Biochim. Biophys. Acta 515, Kasai, M. & Taguchi, T. (1981) Biochim. Biophys. Acta 643, Kasai, M. (1981) J. Biochem. 89, Tanifuji, M., Sokabe, M., & Kasai, M. (1986) J. Membr. Biol. 99, Sokabe, M. (1986) in Int. Symp. Ion Channels and Electrogenic Pumps in Biomembranes(Kasai, M., ed.) pp. L39-50, Osaka University, Osaka 24. Inoue, I. (1986) J. Gen. Physiol. 88, Tanifuji, M., Sato, M., Wada, Y., Anraku, Y., & Kasai, M. (1988) J. Membr. Biol. 106, Abramson, J. J., Trimm, J. L., Weden, L., & Salama, G. (1983) Proc. Natl. Acad. Sci. U. S. 80, Salama, G. & Abramson, J. (1984) J. Biol. Chem. 259, Vol.106, No.3, 1989