Functional Coupling of Human L-Type Ca Channels and Angiotensin AT1A Receptors Coexpressed in Xenopus laevis Oocytes: Involvement of the Carboxyl-Terminal Ca Sensors

A human recombinant L-type Ca channel (a1C,77) was coexpressed with the rat angiotensin AT1A receptor in Xenopus laevis oocytes. In oocytes expressing only a1C,77 channels, application of human angiotensin II (1–10 mM) did not affect the amplitude or kinetics of Ba currents (IBa). In sharp contrast, in oocytes coexpressing a1C,77 channels and AT1A receptors, application of 1 nM to 1 mM angiotensin gradually and reversibly inhibited IBa, without significantly changing its kinetics. The inhibitory effect of angiotensin on IBa was abolished in oocytes that had been preincubated with losartan (an AT1A receptor antagonist) or thapsigargin or injected with 1,2-bis(o-aminophenoxy)ethane-N,N,N9,N9-tetraacetate, pertussis toxin, guanosine-59-O-(2-thio)diphosphate, or heparin, suggesting that the recombinant a1C channels were regulated by angiotensin through G protein-coupled AT1A receptors via activation of the inositol trisphosphate-dependent intracellular Ca release pathway. Consistent with this hypothesis, no crosssignaling occurred between the AT1A receptor and a splice variant of a1C lacking Ca 21 sensors (a1C,86). The data suggest that the regulation of recombinant L-type Ca channels by angiotensin is mediated by inositol trisphosphate-induced intracellular Ca release and occurs at the molecular motif responsible for the Cainduced inactivation of the channels. Voltage-gated Ca channels are a major route for Ca entry into cells in response to stimulation by hormones, neurotransmitters, or drugs. The resulting rise in cytoplasmic free Ca triggers a cascade of intracellular signaling events, which underlie a variety of cellular responses, ranging from contraction and secretion to growth and mitogenesis. Therefore, identification of the molecular basis for functional coupling between Ca channels and hormone or neurotransmitter receptors may provide critical information on cellular signaling mechanisms. The cardiac L-type Ca channel is composed of the poreforming a1C and auxiliary b and a2/d subunits (Catterall, 1995). In an artificial expression system, the a1Cba2/d complex is sufficient to give rise to Ca channels exhibiting all of the major electrophysiological properties observed in vivo. However, functional regulation of the recombinant Ca channel remains largely unknown. For example, in cardiac or vascular cells, the a1C channel is modulated by protein kinase Aand protein kinase C-dependent phosphorylation (McDonald et al., 1994). However, when all three recombinant subunits of the channel are coexpressed in Xenopus laevis oocytes or in eukaryotic systems (Chinese hamster ovary or human embryonic kidney cells), their modulation through phosphorylation is either strongly reduced or essentially absent (Bouron et al., 1995; Zong et al., 1995; Shuba et al., 1997), even though the expressed channels display the same voltage dependence, gating kinetics, unitary conductance, and pharmacological properties as the native a1C Ltype Ca channels. These findings demonstrate the complexity of molecular signaling involving the a1C Ca 21 channels; this complexity extends to the largely unexplored area of “cross-talk” between recombinant a1C channels and hormone receptors that are coexpressed in X. laevis oocytes. The coupling of a1C Ca 21 channels with angiotensin II AT1 receptors has attracted much attention. For example, the L-type Ca channel blockers verapamil, diltiazem, and nifedipine have been shown to block angiotensin II-mediated vascular contraction in vivo in humans (Andrawis et al., 1992). Activation of AT1 receptors seems to be associated with both immediate contractile and long term growth responses in vascular smooth muscle and cardiac myocytes (Baker et al., 1992; Sadoshima and Izumo, 1993; Miyata and Haneda, 1994). Supporting the possibility of interactions between the G protein-coupled AT1 receptors (Anand-Srivastava, 1983; Ohya and Spereliakis, 1991) and voltage-activated Ca channels is the regulation of neuronal (Scott and This work was supported in part by a grant-in-aid from the American Heart Association, Nation’s Capital Affiliate (to N.M.S.), and National Institutes of Health Grants HL16152 (to M.M.) and AG08226 and GM08386 (to D.R.A.). ABBREVIATIONS: PTX, pertussis toxin; IP3, inositol trisphosphate; BAPTA, 1,2-bis(o-aminophenoxy)ethane-N,N,N9,N9-tetraacetate; GDPbS, guanosine-59O-(2-thio)diphosphate; ICl(Ca), Ca -activated Cl current; IBa, Ba 21 current; ICl, Cl 2 current; HEPES, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid. 0026-895X/98/061106-07$3.00/0 Copyright © by The American Society for Pharmacology and Experimental Therapeutics All rights of reproduction in any form reserved. MOLECULAR PHARMACOLOGY, 54:1106–1112 (1998). 1106 at A PE T Jornals on O cber 8, 2017 m oharm .aspeurnals.org D ow nladed from Dolphin, 1987) and cardiac (Yatani et al., 1987) L-type Ca channels by PTX-sensitive or -insensitive G proteins. Similar interactions have been suggested for angiotensin II activation of L-type Ca currents in rat portal vein myocytes (Macrez-Lepretre et al., 1996) and T-type Ca currents in adrenal zona glomerulosa cells (Lu et al., 1996). In this study, we have used the X. laevis oocyte expression system to study the functional coupling between recombinant rat AT1A receptors and splice variants of recombinant human a1C Ca 21 channels with or without the molecular motif responsible for Ca-dependent inactivation of the channel. We show that heterogeneously expressed Ca channels and AT1A receptors are functionally coupled via the G protein/ IP3-mediated Ca 21 signaling cascade. Additionally, we report that the molecular locus for the angiotensin-induced modulation of the a1C Ca 21 channel is independent of permeation of Ca through the pore and is confined to the carboxyl-terminal cytoplasmic motif (positions 1572–1651), which contains multiple Ca sensors of the channel. Materials and Methods Preparation of mRNAs. Template a1C,77 (Soldatov et al., 1995) and a1C,86 (Soldatov et al., 1997) cDNAs were linearized by digestion with BamHI. Capped transcripts were synthesized in vitro with T7 RNA polymerase, using the mRNA cap kit (Stratagene, La Jolla, CA). mRNAs were dissolved in water (0.5 mg/ml). Rat angiotensin AT1A receptor (Murphy et al., 1991) transcripts were kindly provided by Kathryn Sandberg (Georgetown University). Oocyte preparation and injection. Mature female X. laevis frogs were purchased from Xenopus I (Ann Arbor, MI). Clusters of oocytes were defolliculated by shaking for 2 hr at room temperature in 25 ml of medium containing 82.5 mM NaCl, 2 mM KCl, 1 mM MgCl2, 5 mM HEPES, pH 7.5 (adjusted with NaOH), and 0.2% collagenase A (Boehringer Mannheim, Indianapolis, IN). Oocytes were injected with 50–100 nl of a1C,77 or a1C,86 mRNA premixed with mRNAs coding for auxiliary b1 (Ruth et al., 1989) and a2d subunits (Singer et al., 1991) and the AT1A receptor (in a 1:1:1:0.05 molar ratio). Injected oocytes were incubated at 18° in sterile Barth’s medium supplemented with 10,000 units/liter penicillin, 10 mg/liter streptomycin, 50 mg/liter gentamicin, and 0.5 mM theophylline (all from Sigma Chemical Co., St. Louis, MO). Electrophysiological measurements. Whole-cell ion currents were recorded at room temperature (20–22°) by a two-electrode, voltage-clamp method, as previously described (Soldatov et al., 1998). Current traces were elicited at 30-sec intervals by 1-sec (current-voltage relationships) or 250-msec test pulses to 120 mV, from a holding potential of 290 mV. The Ba extracellular (bath) solution contained 50 mM NaOH, 1 mM KOH, 10 mM HEPES, and 40 mM Ba(OH)2 (pH adjusted to 7.4 with methanesulfonic acid). Voltageclamped oocytes were continuously perfused with control experimental solutions at the rate of ;10 ml/min (bath volume, ;150 ml). Human angiotensin II (Sigma) was applied extracellularly. In some experiments, oocytes were injected with 50 nl of PTX (5 mg/ml), 10 mM GDPbS, 94 mM Cs4BAPTA (pH 7.4), or 10 mM heparin (molecular weight, ;3000; Sigma) approximately 1 hr before the experiment. In other experiments, oocytes were incubated at 18° overnight in Cafree Barth’s solution containing 10 nM thapsigargin (RBI, Natick, MA), to deplete their intracellular Ca stores. Results are shown as mean 6 standard error. IBa, determined in the presence of 5 mM (6)-PN200–110 to block the L-type current, did not exceed 3–5% of the total current. Results Coexpression of the a1C,77 channel with the AT1A receptor allows regulation of Ca channels by angiotensin. Coinjection into X. laevis oocytes of cRNAs coding for the conventional a1C,77 channel and auxiliary b1 and a2d subunits gave rise to the expression of well defined, slowly inactivating currents through Ca channels 2–3 days after the injection of cRNAs (Soldatov et al., 1995). With Ba as a charge carrier, step depolarization to 120 mV from a holding potential of 290 mV activated a slowly inactivating, L-type Fig. 1. Effect of angiotensin on Ca currents. A, Representative traces of IBa through a1C,77 channels, elicited by stepwise depolarization to 120 mV from a holding potential of 290 mV, before (F) and 4 min after (E) application of 1 mM angiotensin. B, Traces of IBa through a1C,77 channels coexpressed with AT1A receptors, recorded before (F) and 1.5, 2, and 3 min after application of 1 mM angiotensin. C, Time dependence and reversibility of the angiotensin effect on IBa through a1C,77 channels expressed alone (f) or coexpressed with AT1A receptors (M). IBa amplitudes were measured in response to 250-msec test pulses to 120 mV, applied at 30-sec intervals, and were normalized to maximal IBa in the absence of angiotensin. Arrows, times of application of bath solutions containing the indicated concentrations of angiotensin. D, Current-voltage relationships for IBa through a1C,77 channels coexpressed with AT1A receptors, before treatment (F), 5 min after treatment with 1 mM angiotensin (E), and after a 10-min perfusion with bath medium (M), obtained using the same oocyte as in B and C (M). Experiments were performed at room temperature (;21°). Regulation of Recombinant Ca Channels via AT1A Receptors 1107 at A PE T Jornals on O cber 8, 2017 m oharm .aspeurnals.org D ow nladed from IBa (mean amplitude, 21.64 6 0.33 mA, n 5 9). Application of 0.5–10 mM angiotensin to oocytes expressing only a1C,77 Ca 21 channels produced little or no change in the magnitude or the kinetics of the current, at all voltages examined (Fig. 1, A and C; Table 1). In sharp contrast, in oocytes coexpressing the human a1C,77 channel and rat AT1A receptor (Murphy et al., 1991), angiotensin (0.1–1 mM) inhibited IBa by ;54% (n 5 12), in a timeand concentration-dependent manner (Table 1). The suppressive effect of angiotensin developed within 3–4 min of the hormone exposure, but the effect slowly (20–30 min) reversed even in the presence of the hormone. Fig. 1C shows that 57.6% inhibition of IBa by 1 mM angiotensin reversed spontaneously and washout of the hormone did not accelerate the recovery of the current (Fig. 2). In the presence of angiotensin, IBa recovered by 90.2 6 4.0% (n 5 7) within 20–30 min. The voltage dependence of IBa at the peak of the hormone effect was often shifted by approximately 110 mV (Fig. 1D). These results suggest that the time course of the hormone effect is not critically dependent on the continued presence of the hormone. Fig. 3A illustrates the concentration dependence of the angiotensin effect on a1C,77 channels. Under our experimental conditions, the maximal inhibitory effect (;60% suppression) was reached with 1 mM angiotensin. In none of the cells tested (n 5 12) did the inhibitory effect on IBa exceed 60%. The estimated IC50 value for angiotensin was 33 6 8 nM (n 5 4), with a Hill coefficient of approximately 0.85. Angiotensin failed to suppress the Ca channels in the presence of the reversible AT1A receptor antagonist losartan. Fig. 3B shows that 1 mM losartan had no effect by itself on IBa in an oocyte coexpressing AT1A receptors and a1C channels but completely blocked the angiotensin effect. Replacement of losartan-containing solution with one containing 1 mM angiotensin, however, produced up to 40% (n 5 3) inhibition of IBa. The time course of the inhibition of IBa was slower than in control experiments (Figs. 1C, 4, and 5), which might have been partly caused by the slow dissociation of losartan from the AT1A receptor sites. Taken together, these data suggest that the suppression of IBa through a1C,77 channels by angiotensin is mediated through the direct interaction of angiotensin with AT1A receptors. Angiotensin activates a transient ICl. The rapid application of the hormone in Cl–free solutions was often but not always accompanied by activation of a large, transient, inward current lasting ;2 min. The activation of this inward holding current, measured at 290 mV in Cl2free extracellular solution (Fig. 4, lower), preceded the decrease in IBa. This current had properties similar to those previously identified (Hartzell, 1996; Gomez-Hernandez et al., 1997) for ICl(Ca). During the activation of ICl, IBa often exhibited decreased inactivation kinetics, producing large, slowly deactivating, tail currents (Fig. 4, upper, traces 2 and 3). Interestingly, the angiotensin-induced, transient suppression of IBa outlasted the activation of ICl(Ca) by 2–3 min (Fig. 4), suggesting either different affinities of Ca channels and Caactivated Cl channels for Ca or differences in the spatial distribution of the two channels with respect to the intracellular Ca pools. Lower affinity of ICl(Ca) for activation by Ca, compared with Ca-induced inactivation of Ca channels, and variations in the Ca contents of intracellular Ca pools of the oocytes might be partly responsible for the variations in the magnitude of ICl in different oocytes. The IP3/Ca 21 signaling pathway is involved in channel regulation by angiotensin. Ca stores in X. laevis oocytes are known to be regulated through the activation of IP3-sensitive Ca 21 release channels (Berridge and Irvine, 1989; Putney et al., 1989). These channels are thought to be involved in receptor-mediated Ca signaling, and their activation is known to evoke ICl(Ca) in oocytes (Yao and Parker, 1993; Hartzell, 1996). Consistent with this idea, in oocytes bathed in Barth’s solution and expressing only AT1A receptors, a transient (2–3-min) ICl was activated upon rapid application of angiotensin (data not shown). To further characterize the steps in the regulation of recombinant a1C channels by AT1A receptors, when coexpressed in oocytes, we probed the various steps of the IP3-mediated Ca 21 signaling cascade by inhibiting the G proteins, blocking the IP3 receptor, and interfering with the rise in intracellular Ca levels. Release of intracellular Ca mediates the angiotensin-induced effects. The depletion of intracellular Ca stores by overnight incubation of oocytes with 10 nM thapsigargin (Thastrup et al., 1990) completely abolished the effect of 1 mM angiotensin on IBa (Fig. 6, A and B). No significant difference in the amplitude of IBa in control and thapsigargin-incubated oocytes was observed (Table 1). Similarly, oocytes injected with Ca buffers failed to respond to angiotensin. Fig. 6, C and D, shows data recorded from an oocyte that was injected with 50 nl of 94 mM Cs4BAPTA solution 30 min before measurements of IBa. The data (n 5 4) showed