Bromo-eudistomin D, a novel inducer of calcium release from fragmented sarcoplasmic reticulum that causes contractions of skinned muscle fibers.

Bromo-eudistomin D induced a contraction of the chemically skinned fibers from skeletal muscle at concentrations of 10 microM or more. This contractile response to bromo-eudistomin D was completely blocked by 10 mM procaine. The extravascular Ca2+ concentrations of the heavy fractions of the fragmented sarcoplasmic reticulum (HSR) were measured directly by a Ca2+ electrode to examine the effect of bromo-eudistomin D on the sarcoplasmic reticulum. After the HSR was loaded with Ca2+ by the ATP-dependent Ca2+ pump, the addition of 10 microM bromo-eudistomin D caused Ca2+ release that was followed by spontaneous Ca2+ reuptake. In the presence of 2 microM ruthenium red or 4 mM MgCl2, no Ca2+ release was induced by 20 microM bromo-eudistomin D. The rate of 45Ca2+ efflux from HSR, which had been passively preloaded with 45Ca2+, was accelerated 7 times by 10 microM bromo-eudistomin D. The concentration of bromo-eudistomin D for half-maximum effect on the apparent efflux rate was 1.5 microM, while that of caffeine was 0.6 mM. The bromo-eudistomin D-evoked efflux of 45Ca2+ was abolished by 2 microM ruthenium red or 0.5 mM MgCl2. Bromo-eudistomin D was found to be 400 times more potent than caffeine in its Ca2+-releasing action but was similar in its action in other respects. These results indicate that bromo-eudistomin D may induce Ca2+ release from the sarcoplasmic reticulum through physiologically relevant Ca2+ channels.

increase in extravesicular Ca" induces Ca2+ release (1, 4, 5). One of the effective approaches to the resolution of the problem is the application of specific drugs which affect the releasing mechanism. A typical inducer of Ca" release is caffeine and a blocker is procaine. However, these are not satisfactory drugs for use in the identification or the purification of the molecules operating in Ca2+ release, because of their low affinity. Ruthenium red blocks Ca2+ release at a concentration as low as 0.1 FM (6). However, a labeled derivative of this compound is unavailable.
Bromo-eudistomin D is a monobrominated derivative of eudistomin D, a marine natural product with antiviral activity which has been isolated from the Caribbean tunicate EwEistoma oliuaceum (7). In the present paper, we show that bromoeudistomin D induces a contraction of chemically skinned fibers and a release of Ca2+ from a heavy fraction of the fragmented SR (HSR) at a concentration as low as 1 PM. Bromo-eudistomin D might become a powerful tool for investigation of the mechanism of Ca2+ release from the SR.

EXPERIMENTAL PROCEDURES
Matehls-Bromo-eudistomin D was synthesized by bromination of 6-methoxypyrido[3,4-b]indole (8), with bromine in acetic acid, followed by demethylation with BBr3. The structure of bromo-eudistomin D was confirmed by 'H and I3C nuclear magnetic resonance and fast atom bombardment mass spectra as shown in Fig. 1. 6-Hydroxy-@-carboline was prepared by demethylation of 6-methoxypyrido[3,4-b]indole described above. @-Carboline was synthesized from tryptamine and glyoxylic acid. The details of the synthetic procedures will be reported elsewhere.' Eudistomins A, D, G-J, M-0, and Q were isolated from extracts of E. olivaceum (7). These compounds were dissolved in ethanol or dimethyl sulfoxide prior to application.
The HSR was prepared by the method.of Kim et al. (9). The homogenate of rabbit white muscle in 5 volumes of 5 mM Trismaleate (pH 7.0) was fractionated by differential centrifugations, in which vesicles remaining in the supernatant after centrifugation at 5,000 X g for 5 min were pelleted at 12,000 X g for 30 min. The fraction was washed with buffer A containing 0.1 M KC1 and 5 mM Tris-maleate by centrifugation at 70,000 X g for 30 min. HSR was stored in buffer A at 0 "C and used within 4 days. The purified Ca2+-ATPase of SR (lo), myosin (ll), and creatine kinase (12) were also prepared from rabbit skeletal muscle. Na+,K+-ATPase from porcine cerebral cortex was purchased from Sigma. *5CaC12 was purchased from New England Nuclear, procaine HCI from Sigma, caffeine and ruthenium red from Wako Pure Chemical Industries, and saponin from ICN Pharmaceuticals.
Tension Measurement of Skinned Fibers-The isometric tension of chemically skinned fibers isolated from psoas muscle of guinea pigs was measured by the method of Endo and Kitazawa (13). A small bundle (-0.1 mm in diameter, -3 mm in length) of 4-5 fibers was set on the needles of a strain gauge (T7-8-240, Toyo Baldwin, Japan). The bundle was treated with 50 pg/ml saponin in a relaxation solution containing 2 mM EGTA, 4 mM ATP, 90 mM KMeS03, 5 mM MgS04, and 50 mM MOPS (pH 7.0) for 30 min in a 5-ml Petri dish with a magnetic stirrer at 20 "C. After flushing out saponin with the relaxation solution, 2 mM CaClz was added (-14 p~ free Ca2+) to check the tension development (usually 30-60 mg).
The medium was changed to the relaxation solution and then to a solution containing the same reagents as the relaxation solution without EGTA. After 15 to 30 min, spontaneous cqntractions occurred repeatedly at constant intervals of 4 to 5 min.
Measurements by Cu" Electrode-The extravesicular CaZ+ concentration was monitored with a Ca2+ electrode prepared by the method of Tsien and Rink (14) with modifications. We used Pipetman polyethylene tips (C20, Gilson), instead of siliconized micro-glass pipettes. A small amount of the Ca2+ sensor, a mixture of 5 pl of Ca2+-cocktail (Fluka 21048) and 0.7 mg of polyvinyl chloride in 35 pl of tetrahydrofuran, was attached at the tip (-0.3 mm in diameter) of the polyethylene tips that had been filled with an internal solution containing 5 mM CaClz, 5 mM EGTA, 5 mM MgC12, 0.1 M KC1, and 50 mM MOPS (pH 7.0). The reference electrode of a glass pipette was filled with 1.5% agar in the internal solution without CaClz and EGTA. The electrodes were connected to a high impedance pH meter (Schott CG-822) with platinum terminals and dipped in a 1-ml sample kept in a thermostated bath with a magnetic stirrer. This Ca" electrode showed Nernstian response (slope: 27-29 mV/pCa unit) in the calibration solutions containing Ca-EGTA between pCa 3 and 6.5. The time for 90% response was approximately 0.6 s when pCa decreased from 6 to 4.
In a typical experiment shown in Fig Fig. 3A). The reaction of Ca" uptake was started by a simultaneous addition of 5 pl of 100 mM ATP and 5 pl of 20 mg/ml creatine kinase.
45Cu2+ Efflux Measurement-The "Ca2+ efflux from the HSR passively preloaded with *Ca2+ was measured according to the method of Kirino et ul. (15). After 12-h preincubation of 20 mg/ml HSR with 5 mM 45Ca2+ in solution B containing 90 mM KC1 and 50 mM MOPS at pH 7.0 and 0 "C, the HSR suspension was diluted with 100 volumes of reaction medium containing 0.4 mM CaClz and 1.9 mM EGTA in solution B. The aliquots (0.15 ml) of the diluted suspension were filtered through Millipore filters (HAWP), and washed three times with 0.5 ml of a solution containing 5 mM LaC13 and 5 mM MgClz in, solution B. The amount of 45Ca remaining in the HSR vesicles was measured by counting the radioactivity on the washed filters.
ATPase Actiuity-The enzyme activities of SR Ca2+-ATPase, myosin ATPase, and Na+,K+-ATPase were measured at pH 7.0 and 30 "C by the method of Martin and Doty (16) for the determination of Pi produced.

RESULTS AND DISCUSSION
The sarcoplasmic reticulum (SR) in skinned muscle fibers is known to be morphologically intact although connections between the transverse tubular systems and SR are disrupted. Endo et al. (4) proposed the "Ca-induced Ca release" hypothesis from the finding that skinned fibers showed spontaneous and oscillatory contractions in solutions containing submicro-molar Ca2+, which was attained with low concentration of EGTA. The contraction of skinned fibers is an adequate system to survey the effects of various drugs on Ca2+ release from the SR, possibly occurring through Ca2+ channels in SR membranes.
The skinned fibers showed spontaneous contractions repeatedly at constant intervals of 4 to 5 min (Fig. 2). The addition of 10 p~ bromo-eudistomin D just after relaxation induced a tension spike of the same or larger size as the spontaneous contractions. This contraction was followed by frequent and disordered contractions. Similar contractions were observed by the addition of 0.4 mM caffeine. In the presence of 10 mM procaine, tension did not develop following the addition of 20 p~ bromo-eudistomin D or 1 mM caffeine (data not shown). These facts suggest that bromo-eudistomin D activates Ca2+ channels in a manner similar to caffeine which is known to be an inducer of Ca2+ release from the SR.
We examined the effect of bromo-eudistomin D on SR by monitoring directly extravesicular concentrations of Ca2+ of HSR with a Ca2+ electrode (Fig. 3). The reaction medium contained 1.75 mg/ml HSR, 50 p~ CaC12 (added), ATPregenerating system, i.e 10 mM creatine phosphate and 0.1 mg/ml creatine kinase, and 0.4 mM (a-c) or 4 mM (d and e ) MgC12 (see "Experimental Procedures" for details). The semilogarithmic graph (Fig. 3) demonstrates that, upon the addition of 0.5 mM ATP, free Ca2+ concentrations decreased rapidly due to the formation of Ca-ATP complexes and further decreased gradually due to Ca2+ uptake by the HSR. This Ca2+ uptake phase was almost linear in an antilogarithm plot (not shown). Ca2+ uptake was slowed when the concentration of Ca2+ was reduced to submicromolar levels. The addition of 10 p~ bromo-eudistomin D to the Ca2+-filled HSR induced a rapid Ca2+ release followed by a Ca2+ reuptake. The rate of Ca2+ reuptake was almost the same as that before the addition of bromo-eudistomin D. The transient Ca2+ release was observed repeatedly upon addition of 10 pM bromo-eudistomin D. A similar Ca2+ release was induced by the addition of 2 mM caffeine (Fig. 3b). Ca2+ release by 20 p M bromo-eudistomin D was completely blocked after the addition of 2 p~ ruthenium red (Fig. 3c). In the presence of 4 mM MgC12, the Ca2+ uptake rate increased significantly but no Ca2+ release was induced by 20 p~ bromo-eudistomin D or 4 mM caffeine (Fig. 3, d and e ) . These results suggest that bromo-eudistomin D induces Caz+ release through specific Ca2+ channels, which are blocked by ruthenium red and high concentrations of M P . It is also implied that bromo-eudistomin D-sensitive channels are the same as or similar to channels activated by caffeine, although the affinity of the channels for bromoeudistomin D is much higher than that for caffeine.
The mechanism which is responsible for the reuptake of Ca2+ after its release is not clear. However, two possibilities are present: 1) the activated channels close after a certain period and the released Ca2+ is reincorporated into vesicles or 2) the activated channels are kept open and the released Ca" is taken up into vesicles which have no Ca2+ channels. Ca2+ release from the SR was induced by the addition of A-23187, a Ca2+ ionophore, whereas Ca2+ reuptake was not observed (data not shown). The presence of the Ca2+ reuptake phase, thus, excludes the possibility that bromo-eudistomin D acts on the vesicles as a Ca2+ ionophore.
We measured the concentration-dependent effect of bromoeudistomin D and caffeine on Caz+ efflux from HSR under the conditions in which the Ca2+ pump did not work. The suspension of HSR preloaded passively with 5 mM %aC12 was transferred into 100 volumes of a reaction medium containing various concentrations of bromo-eudistomin D or caffeine, 0.4 mM CaC12, and 1.9 mM EGTA at pH 7.0 and 0 "C. The amount of &Ca remaining in the vesicles decreased and reached a constant basal level of 9.5 k 1.0 nmol/mg 20 min after the dilution irrespective of drug concentrations. The time courses of the change in 45Ca content of the HSR were plotted logarithmically and the plots were linear as shown in Fig. 4. The rate of Ca2+ efflux was markedly accelerated from 0.11 to 0.75 min-' by 10 p~ bromo-eudistomin D. Caffeine also increased the rate of Ca2+ efflux, but required much higher concentrations in the range of several hundred micromolar. In the absence of drugs, the rate of Ca2+ efflux was slightly higher in A than in B, indicating that 1% ethanol (in A) raised the efflux rate. The effect of ethanol was reported by Ohnishi et al. (17). Hill plots (not shown) of the apparent rate constants, which were determined from the slopes in Fig.  4, showed nH = 1.9 and K = 1.5 p~ for bromo-eudistomin D and n~ = 1.9 and K = 0.62 mM for caffeine. 45Ca2+ efflux, accelerated by 3 p~ bromo-eudistomin D, was completely blocked in the presence of 2 p~ ruthenium red or 0.5 mM MgClz (data not shown).
We also examined the effect on Ca" release of 12 compounds related to bromo-eudistomin D, eudistomin A, D, G-J, M-O, and Q (7) as well as @-carboline and &hydroxy-@carboline using a Ca" electrode. Among them, only three compounds, eudistomins A, D, and J, stimulated Ca2+ release from the HSR, although severalfold higher concentrations than bromo-eudistomin D were required (data not shown). These results suggest that 5-Br, 6-OH, and 7-Br in the benzenoid ring are important for the activity.
In addition, the effects of bromo-eudistomin D were examined on other enzymatic reactions which play important roles in muscle contraction mechanisms. Bromo-eudistomin D (up to 30 p~) had no effect on the myosin ATPase activity in the presence of 5 mM EDTA and 0.5 M KC1. It also had no effect on the activity of the Ca2+-ATPase purified from the SR or on the Na+,K+-ATPase.
The present study thus indicates that a low concentration of bromo-eudistomin D induces contractions of skinned muscle fibers by triggering Ca2+ release from intact SR through Ca2+ channels without affecting other muscular enzymes. Isolation of BED-sensitive components in HSR by using bromo-eudistomin D as the reporter group is currently in progress in our laboratory.