Caffeine exerts a dual effect on capacitative calcium entry in Xenopus oocytes

Cellular Signalling 12 (2000) 31–35 http://www.elsevier.com/locate/cellsig

Caffeine exerts a dual effect on capacitative calcium entry in Xenopus oocytes Fre´de´ric Hague, Fabrice Matifat, Ge´rard Bruˆle´, Thibault Collin* Universite´ de Picardie Jules Verne, Faculte´ des Sciences, Laboratoire de Neurobiologie Cellulaire, 33 rue Saint Leu, 80039 Amiens cedex 1, France Received 20 July 1999; accepted 31 August 1999

Abstract Caffeine increases the amplitude of the Cl2 currents evoked by capacitative Ca21 entry (CCE) on thapsigargin-treated Xenopus oocytes. The caffeine-induced potentiation of the CCE process appears to rest on two distinct and additive components. The first component involves the cAMP second messenger system since it can be mimicked by either IBMX perfusion or cAMP microinjection into the oocyte and inhibited by the PKA inhibitory peptide i-PKA. The second component, although activatory, is dynamically related to the caffeine-evoked inhibition of InsP3-mediated Ca21 release and may arise from an interaction between caffeine and the InsP3 receptor in the context of a conformational coupling between the InsP3 receptor and the channels responsible for CCE.  2000 Elsevier Science Inc. All rights reserved. Keywords: Caffeine; Capacitative calcium entry; InsP3; Ca signalling; Protein kinase A; Xenopus oocytes

1. Introduction Activation of the phosphoinositide second messenger pathway elevates the cytoplasmic concentration of inositol (1,4,5) trisphosphate (InsP3), which releases Ca21 from intracellular stores [1,2]. Then, an as yet poorly understood mechanism termed “capacitative Ca21 entry” (CCE) is activated [3]. This permits Ca21 to enter cells through Ca21-selective Ca21 release-activated channels (CRACs), as well as through less selective store-operated channels [3,4]. Extracellular or intracellular applications of caffeine have been reported to inhibit InsP3-mediated Ca21 release (IMCR) in Xenopus oocytes through a mechanism distinct from the well-known action of caffeine as a phosphodiesterase inhibitor [5,6]. Although caffeine has been widely described and used as a blocker of IMCR [7,8], its effects toward CCE have not been intensively investigated. Petersen and Berridge [9] dem-

Abbreviations: InsP3, D-myo-inositol (1,4,5)-trisphosphate; InsP3-F, 3-deoxy-3-fluoro InsP3; Caged-InsP3, InsP3-P4(5)-1-(2nitrophenyl) ethyl ester; ICl(Ca), calcium-dependent chloride current. * Corresponding author. Tel.: 133-3-2282-7647; fax: 133-3-22827644. E-mail address: [email protected] (T. Collin)

onstrated that the G-protein regulation of CCE in Xenopus oocytes is mediated by protein kinases A and C. In this context, they studied the cAMP pathway in oocytes that had been treated with thapsigargin to activate CCE and reported that application of db-cAMP has a biphasic effect on the amplitude of the CCE-induced currents without altering their activation or inactivation kinetics. Using the same means to activate CCE, we found that caffeine exerts a potentiating effect on the CCE-evoked current amplitude. Roughly one third of the increase could be attributed to phosphodiesterase inhibition since the same level of potentiation was reached with either IBMX or cAMP. The other part of the potentiation was considered to be independent of the cAMP transduction process because it was not inhibited by the protein kinase A (PKA) inhibitory peptide i-PKA and it remained elicitable by caffeine upon IBMX- or cAMP-treated oocytes. Finally, it is postulated that caffeine may exert a dual effect on CCE, first by passively increasing the cAMP level and second by putatively interacting with the molecules responsible for the development of the CCE process. Because of the ability of caffeine specifically to bind the InsP3 receptor as demonstrated [10], we propose that the cAMP-independent component of the caffeine-induced increase of CCE is supported by the direct interaction between the drug and the InsP3 receptor.

0898-6568/00/$ – see front matter  2000 Elsevier Science Inc. All rights reserved. PII: S0898-6568(99)00074-1

32

F. Hague et al. / Cellular Signalling 12 (2000) 31–35

2. Materials and methods 2.1. Oocytes preparation Adult Xenopus laevis were anaesthetized in tricaine methane sulphonate (0.2%; MS 222). Pieces of ovary were surgically removed and placed in ND96 medium of the following composition (in mM): NaCl 96; CaCl2 1.8; KCl 2; MgCl2 2; HEPES 5 titrated to pH 7.4 with NaOH. Oocytes were then treated for 2 to 3 h with collagenase (2 mg/ml, type IA) in calcium-free medium to remove follicular cells. Stages V and VI oocytes were selected and maintained in ND96 supplemented with gentamycin (50 mg/ml) at 1208C for up to 5 days. This medium was renewed daily. 2.2. Electrophysiological measurements Changes in cytosolic calcium concentration were monitored by electrophysiological recording of the calcium-activated chloride current [ICl(Ca)] as described elsewhere [11]. Briefly, oocytes were impaled by two microelectrodes filled with 3 M KCl (0.5–1 MV) and voltage-clamped to 280 mV using a Geneclamp 500 voltage-clamp amplifier (Axon Instruments, Burlingame, CA, USA). Data acquisition and analysis were conducted using the pClamp software package (v. 5.7.1, Axon Instruments, Burlingame, CA, USA). Voltage monitoring was performed by using solely the voltage electrode of the dual-microelectrodes setup described above. Experiments were carried out at room temperature (18–228C). cAMP and i-PKA were injected using a third additional micropipette whose tip diameter did not exceed 10 mm. 2.3. Flash-photolysis experiments The optical arrangement for light flash experiments (Optoelektronik, Hartkro¨gen, Hamburg, Germany) is described by Miledi and Woodward [12]. The wavelengths most efficient for photolysis experiments (300– 400 nm) were selected from the spectrum by a UG11 filter. Light was focused onto the cell with a spot diameter of about 6 mm. Flashes were iteratively triggered by the “trig” output of the Digidata 1200 data acquisition system (Axon Instruments, Burlingame CA, USA) using the time base of the computer. The caged-InsP3 was pressure-injected 1 h before the experiments. The final intraoocyte concentration was estimated to 10 mM. 3. Results and discussion CCE is activated by inhibiting the cellular Ca21 uptake mechanism through application of the endoplasmic Ca21/ATPase inhibitor thapsigargin. Fig. 1A shows the iterative application of 2 mM of CaCl2 in 2-min pulses, each separated by an interval of 3 min on a voltage-clamped oocyte that had been pretreated with thapsigargin. Readmission of Ca21 to the bathing solu-

tion evokes inward-directed Ca21-dependent Cl21 current (ICl(Ca)) transients that have been reported to reflect the Ca21 entry process [9,11]. Bath application of caffeine (5 mM) transiently potentiates the Ca21 entryevoked currents (328 6 52% of control, n 5 8). Because caffeine belongs to the alkylxanthine family, its effects are classically attributed to the cAMP phosphodiesterase (PDE) inhibition [13]. This is especially true for cells that do not express adenosine receptors, such as defolliculated Xenopus oocytes [12]. In that context, we investigated the effects of one of the more potent PDE inhibitor of the alkylxanthine family: IBMX. The results depicted in Fig. 1B clearly demonstrate that IBMX (5 mM) increases the amplitude of the CCE-elicited currents. Nevertheless, this potentiation (176 6 61% of control, n 5 6) is not as strong as that obtained with caffeine when both drugs are used at the same concentration of 5 mM. It has been reported that activation of PKA exerts a biphasic effect on the amplitude of the Ca21 entryevoked currents [9]. More precisely, application of dbcAMP at 1 to 10 mM to thapsigargin-treated oocytes inhibited the currents evoked by pulsed application of 2 mM Ca21, whereas the latter were potentiated by application of 1 to 10 mM db-cAMP [9]. Surprisingly, we did not observe any biphasic effect with caffeine, IBMX, or intracellular injection of cAMP. As summarized in Fig. 2, these reagents all have a potentiating effect on CCE. To assess the link between the alkyxanthine and CCE, we measured the impact of caffeine and IBMX superfusion on oocytes that had been microinjected with the PKA inhibitory peptide i-PKA (10 mg/ml, final intraoocyte concentration) [14]. Data presented in Fig. 2 show that most of the IBMX-induced increase in the CCE-evoked currents is abolished by i-PKA (117 6 10% of control, n 5 5). In the case of caffeine, however, i-PKA did not completely abolish the potentiation (214 6 40% of control, n 5 5 vs. 328 6 52% of control, n 5 8), which suggests that the effects of caffeine are not totally mediated by its PDE inhibitory activity. This guess was further corroborated by experiments based on the additive effect of caffeine upon a preliminary treatment of the cell with either cAMP or IBMX. An “on-line” cAMP microinjection into the oocyte clearly caused an increase in the CCE amplitude (183 6 37% of control, n 5 3), which was further increased by caffeine superfusion (5 mM; 356 6 68% of control, n 5 5; Fig. 3A). The same result was observed with IBMX (5 mM) treatment followed by caffeine application in the continuous presence of IBMX (174 6 28% of control, n 5 5 with IBMX and 367 6 73% of control, n 5 5 with IBMX 1 caffeine; Fig. 3B). These results tend to demonstrate that caffeine may exert a dual effect on the CCE by: (1) inhibiting the PDE and thus passively raising the cAMP level and (2) another means that is clearly independent of cAMP.

F. Hague et al. / Cellular Signalling 12 (2000) 31–35

33

Fig. 1. Pulsed extracellular application of 2 mM Ca21 evoked Ca21 influx-triggered Cl2 currents in thapsigargin-treated oocytes under voltage clamp conditions at 280 mV. Ca21 pulses of 2 min are indicated by stripped bars above the current traces. (A) The magnitude of the Ca21-evoked Cl2 current is markedly enhanced by caffeine superfusion (5 mM; indicated by a dark bar). (B) IBMX superfusion (5 mM; indicated by a dark bar) clearly causes a raise in CCE amplitude.

Fig. 2. The bar graph represents maximal CCE-evoked Cl2 current amplitude measured in the presence of caffeine, IBMX (both at 5 mM in the superfusate), or cAMP (50 mM, final intraoocyte concentration) on control or i-PKA-injected oocytes (10 mg/ml, final concentration).

Aside from its role in this latter second messenger system, caffeine has been reported to block the InsP3mediated liberation of Ca21 [5–8]. This blocking effect, which is not observed with other alkylxanthines, correlates with the ability of caffeine specifically to bind the InsP3 receptor [15]. A particular feature of the caffeineevoked blockade of IMCR is its somehow transient nature. As shown in Fig. 4A, caffeine transiently inhibits InsP3-induced potential oscillations in Xenopus oocytes. As reported by Berridge [5], caffeine stimulation elicits a large membrane hyperpolarisation together with an evident blockade of the InsP3-evoked oscillations. At first, caffeine abolishes the InsP3 response but during continued application of caffeine, the potential goes back to its original value and the oscillations recover. These results indicate that the IMCR mechanism can escape from its blockade by caffeine. To confirm these data, we performed experiments with flash photolysis of caged InsP3 to liberate precisely controlled amounts of InsP3 into the oocyte. Because each oocyte

34

F. Hague et al. / Cellular Signalling 12 (2000) 31–35

Fig. 3. (A) Injection of cAMP (1 mM, final intracellular concentration; indicated by an arrow) increases the amplitude of the CCE-evoked currents that is further augmented by caffeine superfusion (5 mM, indicated by a dark bar). (B) The same results are obtained with IBMX superfusion (5 mM, indicated by a dark bar) followed by caffeine application (5 mM) in the continuous presence of IBMX (indicated by a dotted line).

was loaded with a large excess of caged InsP3, it was possible to evoke many tens of responses. The currents elicited by this means were more transient than those evoked either by metabotropic stimulation or free InsP3, but arise similarly through Ca21 liberation from intracellular stores. Fig. 4B depicts the transience of the caffeine block observed on caged-InsP3-induced currents. These results are in agreement with those reported by Parker and Ivorra [6] using the same technique. It has been demonstrated that the inhibitory effect of caffeine on IMCR requires a specific binding site of the drug on the InsP3 receptor [15,16]. It is therefore conceivable that such interaction induces a conformational change of the receptor that affects the gating properties,

resulting in an inhibition of the Ca21 release. In other respects, the CCE channel protein is now established to be involved in a direct interaction with a regulatory protein such as the InsP3 receptor on the intracellular Ca21 stores [17]. The emergence of this model, referred to as the “conformational coupling” [18], allows us to postulate that the conformational change of the InsP3 receptor due to its caffeine binding site occupancy may also affect the CCE process. Interestingly, the transience of the IMCR inhibition parallels the transient character of the cAMP-independent component of the caffeine-induced potentiation of CCE (Fig. 3), suggesting that both actions may be supported by the same molecular interaction. The effects of

F. Hague et al. / Cellular Signalling 12 (2000) 31–35

35

Fig. 4. Injection of InsP3 (1 mM, final intracellular concentration) induces an immediate depolarizing spike followed by a burst of membrane potential oscillations. The latter was transiently abolished by perfusing the oocyte with caffeine. (B) Current traces obtained by flash photolysis (each transient is triggered by a light signal) of caged-InsP3 in an oocyte voltage-clamped at 280 mV. Caffeine superfusion (5 mM; indicated by a dark bar) immediately but transiently inhibited the flash-evoked ICl(Ca) transients.

caffeine on Xenopus oocyte can take place globally at two distinct levels: the cAMP PDE or the InsP3 receptor. In that respect, we propose that the cAMP-independent facilitating effect of caffeine on CCE could take place on the InsP3 receptor itself. Finally, these data may bring a new insight in the effects of caffeine as a well-known stimulant. Since ingestion of a few cups of coffee can result in plasma concentrations as high as 100 mM, it is likely that actions on phosphoinositide/Ca21 signalling pathway in neurons may contribute to the behavioural effects of caffeine. Acknowledgments F.H. is supported by a fellowship from Le Biopoˆle Ve´ge´tal de Picardie.

[3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16]

References [1] Berridge MJ. Nature (London) 1993;361:315–25. [2] Clapham DE. Cell 1995;80:259–68.

[17] [18]

Hoth M, Penner R. Nature (London) 1992;355:353–6. Shuttleworth TJ. Cell Calcium 1999;25:237–46. Berridge MJ. Proc R Soc Lond B 1991;244:57–62. Parker I, Ivorra I. J Physiol (London) 1991;433:229–40. Bezprozvanny I, Bezprozvannaya S, Ehrlich BE. Mol Biol Cell 1994;5:97–103. Humez S, Collin T, Matifat F, Guilbault P, Fournier F. Cell Signal 1996;8:375–9. Petersen CCH, Berridge MJ. Biochem J 1995;307:663–8. Missiaen L, Parys JB, De Smet H, Himpens B, Casteels R. Biochem J 1994;300:81–4. Matifat F, Fournier F, Lorca T, Bruˆle´ G, Collin T. Biochem J 1997;322:267–72. Miledi R, Woodward RM. J Physiol (London) 1989;416:601–21. Butcher RW, Sutherland EW. J Biol Chem 1962;237:1244–50. Knighton DR, Zheng JH, Ten Eyck LF, Xuong NH, Taylor SS, Sowadski JM. Science 1991;253:414–20. Maes K, Missiaen L, Parys JB, Sienaert I, Bultynck G, Zizi M, De Smet P, Casteels R, De Smet H. Cell Calcium 1999;25:143–52. Missiaen L, Parys JB, De Smet H, Sienaert I, Bootman MD, Casteels R. In: Biswas BB, Biswas S, editors. Subcellular Biochemistry, Volume 26: myo-Inositol Phosphates, Phosphoinositides, and Signal Transduction. New York: Plenum Press, 1996. pp. 59–95. Petersen CCH, Berridge MJ. Pflu¨gers Arch 1996;432:286–92. Berridge MJ. Biochem J 1995;312:1–11.