7-Deaza-8-bromo-cyclic ADP-ribose, the First Membrane-permeant, Hydrolysis-resistant Cyclic ADP-ribose Antagonist*

Cyclic ADP-ribose (cADPR) is a putative second messenger that has been demonstrated to mobilize Ca2+in many cell types. Its postulated role as the endogenous regulator of ryanodine-sensitive Ca2+ release channels has been greatly supported by the advent and use of specific cADPR receptor antagonists such as 8-NH2-cADPR (Walseth, T. F., and Lee, H. C. (1993)Biochim. Biophys. Acta 1178, 235–242). However, investigations of the role of cADPR in physiological responses, such as fertilization, stimulus-secretion coupling, and excitation-contraction coupling, have been hindered by the susceptibility of cADPR receptor antagonists to hydrolysis and the need to introduce these molecules into cells by microinjection or patch clamp techniques. We have recently reported on the discovery of a poorly hydrolyzable analogue of cADPR, 7-deaza-cADPR (Bailey, V. C., Sethi, J. K., Fortt, S. M., Galione, A., and Potter, B. V. L. (1997) Chem. Biol. 4, 41–51) but this, like cADPR, is an agonist of ryanodine-sensitive Ca2+ release channels. We therefore explored the possibility of combining antagonistic activity with that of hydrolytic resistance and now report on the biological properties of the first hydrolysis-resistant cADPR receptor antagonist, 7-deaza-8-bromo-cADPR. In addition this compound has the advantage of being membrane-permeable. Together these properties make this hybrid molecule the most powerful tool to date for studying cADPR-mediated Ca2+ signaling in intact cells.

As is the case for the more established intracellular messengers (i.e. IP 3 , cAMP, and cGMP), cADPR-metabolizing enzymes are also present that can modulate cADPR levels (11). The synthetic activity of ADP-ribosyl cyclase and catabolic activity of cADPR hydrolase are often co-localized on the same polypeptide. In these cases, the hydrolase activity often exceeds that of cyclase (1). However, one notable exception is Aplysia ADPribosyl cyclase, which is isolated and purified from soluble ovotestis extracts of the sea hare Aplysia californica. The exceptionally high level of cyclase activity exhibited by this enzyme (1) has been well exploited to synthesize large quantities of cADPR. In addition, the finding that this cyclase exhibits loose substrate specificity has allowed the development of a chemoenzymatic synthesis of a number of cADPR analogues (12).
The first series of pharmacologically useful cADPR analogues to be synthesized was the 8-substituted analogues (13). These differ from cADPR by a substitution at the 8-position of the adenine ring. This single modification abolishes the agonistic activity of these compounds and produces instead specific competitive antagonists of cADPR-sensitive Ca 2ϩ release (13). Since its discovery, 8-NH 2 -cADPR has been used successfully to demonstrate the involvement of cADPR-mediated Ca 2ϩ signaling in sea urchin eggs during fertilization 2 (14) and NO-and cGMP-induced Ca 2ϩ release (15) in Purkinje neurons (16), hippocampal synaptic plasticity (36), permeabilized Jurkat T cells (6), intestinal smooth muscle during cholecystokinin-induced contractions (5), PC12 cells (17), and excitation contraction coupling in cardiac myocytes (18). However, like the parent compound, cADPR, the 8-substituted analogues are prone to hydrolysis by endogenous enzymes (13). Indeed this may explain the absence of an inhibitory effect on secretogogue-induced Ca 2ϩ release in rat pancreatic beta cells (19) during induction of long-term depression in Purkinje neurons (16) where a role for cADPR-mediated Ca 2ϩ signaling remains controversial.
These observations underscore the need for a stable, hydrolysis-resistant cADPR antagonist. Recently, we reported on the synthesis of another analogue of cADPR, 7-deaza-cADPR, and demonstrated that it is more stable during heat-induced hydrolysis and is also a poor substrate for cADPR hydrolase (20). These changes in stability were also brought about by a single modification, a replacement of the 7-position nitrogen with carbon (Fig. 1A). These findings then raised an intriguing question; what would be the biological activity of a compound * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18  modified at both the 7-and 8-positions of the adenosine ring? A "hybrid" analogue was successfully synthesized, namely 7-deaza-8-bromo-cADPR (Fig. 1A). Its biological properties were examined and are reported herein. Our findings show that 7-deaza-8-bromo-cADPR retains useful pharmacological properties; i.e. it is a hydrolysis-resistant antagonist of cADPRinduced Ca 2ϩ release. Furthermore, owing to the lipophilic nature of the bromo and CH moieties, we have explored its potential as a membrane-permeable analogue of cADPR, as has been established for 8-bromo-cGMP, cf. cGMP (21). A single molecular species exhibiting all three properties would be a very powerful pharmacological tool for investigations of cADPR-mediated Ca 2ϩ signaling in intact cells. We report here that 7-deaza-8-bromo-cADPR could be such a tool.
Intracellular Free Ca 2ϩ Measurements in Intact Eggs-Ca 2ϩ imaging of intact cells was performed using unfertilized L. pictus eggs microinjected with 2 M fura-2 and 250 g/ml heparin as described previously (26,27).

RESULTS AND DISCUSSION
Since the elucidation of the structure of cADPR (28,29) and the development of a chemoenzymatic synthesis for cADPR and its analogues (12), interest has mounted concerning the structure-function relationships between the ligand and its endogenous receptor(s). To further aid these diagnostic studies, we synthesized 7-deaza-8-bromo-cADPR, a novel analogue of cADPR. Fig. 1A shows the chemical structure of 7-deaza-8bromo-cADPR as compared with that of cADPR and two previously reported analogues, 8-bromo-cADPR, a specific cADPR antagonist (13), and 7-deaza-cADPR, a hydrolysis-resistant partial agonist (20). Whereas the latter two compounds have a Modifications were made to cADPR at the 7-and 8-positions of the adenosine ring. The N 7 nitrogen atom is replaced with a carbon (and associated proton) to form 7-deaza-cADPR whereas 8-bromo-cADPR is formed by substituting the hydrogen on the C-8 position with a bromine atom. 7-Deaza-8-bromo-cADPR has both of these modifications. B, calcium-releasing action of cADPR, 7-deaza-cADPR, 8-bromo-cADPR, and 7-deaza-8-bromo-cADPR in sea urchin egg homogenates. L. pictus egg homogenates (2.5%, v/v) containing the Ca 2ϩ -sensitive dye, fluo-3 (3 M) was prepared as described under "Experimental Procedures." Aliquots (500 l) of these were challenged with 5 l of cyclic compounds to give a final cuvette concentration of 2 M. The Ca 2ϩ release profiles observed are shown and are representative of three separate experiments.
Whether 7-deaza-8-bromo-cADPR, like 8-bromo-cADPR, was also an antagonist of cADPR-sensitive Ca 2ϩ release was investigated next. Fig. 2 shows that this was indeed the case. Sea urchin egg homogenates pretreated with either 8-bromo-cADPR or 7-deaza-8-bromo-cADPR were markedly less responsive to 100 nM cADPR (Fig. 2A). These inhibitory actions were dependent on the antagonist concentration (Fig. 2B). Both Lower concentrations (31 nM) of either antagonist were also significant in preventing Ca 2ϩ release by 100 nM cADPR that did not exceed 85% of control values. Whether this is due to the presence of more than one population of receptors that exhibit different binding affinities and/or sensitizing properties remains to be investigated. Nonetheless, both 8-substituted analogues appeared to behave similarly with respect to these actions on the cADPR-induced Ca 2ϩ release channel in sea urchin egg homogenates. Since previous studies have shown that 8-substituted cADPR analogues (13, 30) and 7-deaza-cADPR (20) are able to displace cADPR binding, it is likely that 7-deaza-8-bromo-cADPR also interacts at the cADPR receptor in the same specific manner.
Since a modification on the 8-position does not alter the stability of the molecule (13) but a substitution of N 7 with a carbon has been shown to render the cyclic compound more resistant to hydrolysis (20), we investigated whether 7-deaza-8-bromo-cADPR could differ from 8-bromo-cADPR but resemble 7-deaza-cADPR in this respect. We subjected standard solutions of both antagonists to heat-induced hydrolysis. This treatment has previously been shown to strip unstable cyclic compounds such as cADPR and 8-NH 2 -cADPR of their biological activity (13,18,20). Fig. 3 shows the effect of heat treatment on the antagonistic actions of both 7-deaza-8-bromo-cADPR and 8-bromo-cADPR. Whereas 8-bromo-cADPR is stripped of its antagonistic activity, 7-deaza-8-bromo-cADPR remains an effective antagonist of cADPR-induced Ca 2ϩ release (Fig. 3, B compared with A). HPLC analysis of the same samples confirmed that this loss of activity was due to degra-
It has been demonstrated that the presence of a lipophilic bromide moiety in cGMP affords greater membrane permeability to 8-bromo-cGMP (21), and the replacement of a nitrogen with a CH-group also offers greater hydrophobicity (32). Therefore, the novel cADPR analogue, 7-deaza-8-bromo-cADPR, should have greater hydrophobic character than any that were previously synthesized. We investigated this by testing the effect of extracellular applications of 7-deaza-8-bromo-cADPR on fertilization-induced Ca 2ϩ mobilization in intact sea urchin eggs. Eggs were co-injected with the IP 3 receptor antagonist, heparin (250 g/ml), and Ca 2ϩ -sensitive fluorochrome fura-2 (2 M). Fig. 4A shows that upon sperm addition to control heparinized eggs, a propagating Ca 2ϩ wave was produced (Fig. 4A, open squares). This Ca 2ϩ wave had an average amplitude of 1705 Ϯ 119 nM Ca 2ϩ (S.E., n ϭ 11) and took 41.9 Ϯ 5.8 s (S.E., n ϭ 11) to reach this peak. These data are consistent with previously reported observations of sperm-induced Ca 2ϩ mobilization from IP 3 -insensitive Ca 2ϩ stores (14,33), which suggests inhibition of the redundant cADPR-sensitive Ca 2ϩ release mechanism (14,33). At a concentration of 50 M in the Despite the presence of heparin, an IP 3 receptor antagonist, the wave properties of the sperm-induced Ca 2ϩ rise were intact, and activation envelopes were formed (14,33). The peak Ca 2ϩ rise following the addition of sperm was 1705 Ϯ 119 nM Ca 2ϩ (S.E., n ϭ 11). B, in the presence of 50 M 7-deaza-8-bromo-cADPR in the bathing medium (5 min prior to sperm addition; similar results were observed whether preincubations lasted 5, 10, or 15 min), the amplitude of the Ca 2ϩ transient was significantly reduced, and the propagation of the Ca 2ϩ rise across the egg was significantly slowed compared with control. Numbers refer to the points annotated in panel C. Neither antagonist mobilized Ca 2ϩ in the eggs during the 5-15-min preincubation period. C, accompanying data from panels A and B show the average rise in Ca 2ϩ following sperm addition at t ϭ 0. Note the slower increase in Ca 2ϩ following sperm addition in eggs pretreated with 50 M 7-deaza-8-bromo-cADPR and the reduced amplitude. Open squares represent control data, and the filled squares indicate the points represented by the images in panel A. Closed circles represent the response of an egg pretreated with 50 M 7-deaza-8-bromo-cADPR, and open circles indicate the points represented by the images in panel B. Note also the presence of a small initial Ca 2ϩ rise after sperm addition and prior to the full-blown Ca 2ϩ rise associated with fertilization a feature often observed in eggs pretreated with 7-deaza-8-bromo-cADPR. At higher concentrations of 7-deaza-8-bromo-cADPR the fertilizationinduced Ca 2ϩ transient was completely abolished as shown by the crossed symbols.
bathing solution the 7-deaza-8-bromo-cADPR successfully reduced the fertilization-induced Ca 2ϩ transient in heparinized eggs (Fig. 4, A and B). The amplitude of the Ca 2ϩ transient was significantly reduced in the treated eggs compared with the heparinized controls (p Ͻ 0.01, Student's t test) as seen in Fig.  4C, open squares versus closed circles (mean value, 988 Ϯ 81 nM Ca 2ϩ (S.E., n ϭ 6)). As can also be observed in Fig. 4B, the propagation of the Ca 2ϩ wave across the egg was slowed significantly in the eggs treated with the 7-deaza-8-bromo-cADPR. The time to peak of the Ca 2ϩ rise at fertilization was also reduced in eggs pretreated with the antagonist compared with the heparinized controls (p Ͻ 0.01, Student's t test; mean value, 98.8 Ϯ 12.1 s; S.E., n ϭ 6). This is also apparent in Fig.  4C. At a higher concentration of 100 M 7-deaza-8-bromo-cADPR in the bathing medium the sperm-induced intracellular Ca 2ϩ transients were completely abolished (Fig. 4C, crosses). The dose dependence of the effects of the 7-deaza-8-bromo-cADPR is shown in Fig. 5A, filled symbols, right-hand axis. Comparison with a preincubation with 8-bromo-cADPR showed that at 100 M it also reduced the amplitude of the Ca 2ϩ transient following sperm addition and to a similar extent as observed with 50 M 7-deaza-8-bromo-cADPR (see Fig. 5B cf. Fig. 5A, closed symbols). This result indicates that 7-deaza-8bromo-cADPR appears to be the more effective antagonist. Neither antagonist released Ca 2ϩ in the eggs during a 5-15min preincubation period (data not shown), which is consistent with the absence of agonistic activity observed in vitro (Fig. 1B).
Since intracellular Ca 2ϩ mobilization is a prerequisite for the cortical reaction (34,35), we also monitored the formation of activation envelopes following fecundation and directly compared the actions of 7-deaza-8-bromo-cADPR and 8-bromo-cADPR in the bathing medium. Treatment of eggs with either antagonist prevented the cortical reaction (in heparinized eggs) in a concentration-dependent manner (Fig. 5, A and B, open  symbols). This is in keeping with previous reports that demonstrate inhibition of the cortical reaction only when both redundant mechanisms have been blocked (14,33). As indicated by the fertilization-induced Ca 2ϩ transients, we observed a greater effect of the 7-deaza-8-bromo-cADPR compared with the 8-bromo-cADPR at both 50 and 100 M concentrations (p Ͻ 0.05, Student's t test in both cases). Eggs that were treated with antagonist only (i.e. not micro-injected with heparin) were also scored for activation envelopes in the same experiments. These consistently showed Ͼ95% activation as shown in Fig. 5, A and B, by the dotted lines and cross symbols. This suggested that neither 7-deaza-8-bromo-cADPR nor 8-bromo-cADPR acted as spermicides; rather they are able to permeate the sea urchin egg plasma membrane and specifically compete for endogenous cADPR-binding sites and inhibit agonist-induced Ca 2ϩ mobilization. Presumably, since the net charge on 7-deaza-8-bromo-cADPR is only 1 at physiological pH (Fig. 1A), this compares well with 8-bromo-cGMP, which has the same net charge and where 8-substitution confers membrane permeability. This property now eliminates the need for potentially disruptive protocols such as cell permeabilization or microinjection methods to introduce cADPR antagonists into whole cells. This advancement should greatly aid investigations of the role of cADPR in physiological responses to extracellular stimuli.
In conclusion, we have demonstrated that, unlike 8-bromo-cADPR, 7-deaza-8-bromo-cADPR is a stable hydrolysis-resistant, cADPR antagonist. This is the first report of such a compound. In addition, we have exploited the lipophilic nature of the bromo moiety to produce a compound that is also sufficiently membrane-permeable. In all, this makes 7-deaza-8-bromo-cADPR a very powerful pharmacological tool for investigations of cADPR-mediated Ca 2ϩ signaling in intact cells.
FIG. 5. Dose-dependent action of 7-deaza-8-bromo-cADPR and 8-bromo-cADPR on peak fertilization-induced Ca 2؉ rise and egg activation. For each concentration of both antagonists, 7-deaza-8bromo-cADPR (A) and 8-bromo-cADPR (B), 4 -9 eggs in a single dish were co-injected with heparin (250 g/ml) and fura-2 (2 M). Sperm were added following incubation in artificial sea water that contained either antagonist for 5 min, and the maximum change in intracellular free Ca 2ϩ (primary y axis, filled symbols) was monitored on 1 of the preinjected eggs. These data are taken from the experiments done in 1 day (1 egg at each concentration). Following recovery of the Ca 2ϩ transient we scored the presence or absence of an activation envelope in all injected eggs (secondary y axis, open symbols). Egg activation results represent the mean Ϯ S.E. for 3-4 separate determinations of 4 -12 eggs when concurrent Ca 2ϩ measurements were not always made. In separate fields within the same dish we scored the percentage of fertilization in uninjected, non-heparinized eggs (dotted lines). All were consistently activated, which indicated that sperm activity was unaffected by increasing concentrations of either antagonist. Note that the reduction of the fertilization-induced Ca 2ϩ transient with 100 M 8-bromo-cADPR (panel B, filled triangles) was similar to that seen using 50 M 7-deaza-8-bromo-cADPR (panel A, filled circle) supporting the greater effectiveness of the 7-deaza-8-bromo-cADPR compound. Further support for this comes from the far superior percent reduction in egg activation observed at 100 M concentration of the 7-deaza-8-bromo-cADPR compared with the same concentration of 8-bromo-cADPR (p ϭ 0.02, Student's t test).