Interaction of &agr;1-Adrenoceptor Subtypes With Different G Proteins Induces Opposite Effects on Cardiac L-type Ca2+ Channel

We examined the effect of &agr;1-adrenoceptor subtype-specific stimulation on L-type Ca2+ current (ICa) and elucidated the subtype-specific intracellular mechanisms for the regulation of L-type Ca2+ channels in isolated rat ventricular myocytes. We confirmed the protein expression of &agr;1A- and &agr;1B-adrenoceptor subtypes at the transverse tubules (T-tubules) and found that simultaneous stimulation of these 2 receptor subtypes by nonsubtype selective agonist, phenylephrine, showed 2 opposite effects on ICa (transient decrease followed by sustained increase). However, selective &agr;1A-adrenoceptor stimulation (≥0.1 &mgr;mol/L A61603) only potentiated ICa, and selective &agr;1B-adrenoceptor stimulation (10 &mgr;mol/L phenylephrine with 2 &mgr; mol/L WB4101) only decreased ICa. The positive effect by &agr;1A-adrenoceptor stimulation was blocked by the inhibition of phospholipase C (PLC), protein kinase C (PKC), or Ca2+/calmodulin-dependent protein kinase II (CaMKII). The negative effect by &agr;1B-adrenoceptor stimulation disappeared after the treatment of pertussis toxin or by the prepulse depolarization, but was not attriburable to the inhibition of cAMP-dependent pathway. The translocation of PKC&dgr; and ϵ to the T-tubules was observed only after &agr;1A-adrenoceptor stimulation, but not after &agr;1B-adrenoceptor stimulation. Immunoprecipitaion analysis revealed that &agr;1A-adrenoceptor was associated with Gq/11, but &agr;1B-adrenoceptor interacted with one of the pertussis toxin-sensitive G proteins, Go. These findings demonstrated that the interactions of &agr;1-adrenoceptor subtypes with different G proteins elicit the formation of separate signaling cascades, which produce the opposite effects on ICa. The coupling of &agr;1A-adrenoceptor with Gq/11-PLC-PKC-CaMKII pathway potentiates ICa. In contrast, &agr;1B-adrenoceptor interacts with Go, of which the &bgr;&ggr;-complex might directly inhibit the channel activity at T-tubules.

T he ␣ 1 -adrenoceptor (AR) stimulation has an important role for the regulation of mammalian cardiac muscle contraction. [1][2][3][4] We have previously shown that ␣ 1 -AR stimulation modulates the function of voltage-gated L-type Ca 2ϩ channels (VLCC) which is one of the important regulatory factors in cardiac excitation-contraction coupling. 5 The effects of ␣ 1 -AR stimulation on cardiac Ca 2ϩ current through VLCC (I Ca ) can be classified into 2 opposite effects (negative and positive effects): the positive effect is dependent on protein kinase C (PKC) and Ca 2ϩ /calmodulin-dependent protein kinase II (CaMKII) activity, but the negative effect is not. 5 Although we have proposed this novel model for understanding the molecular mechanisms underlying the modulation of VLCC by ␣ 1 -AR stimulation, 2 important questions remain to be solved: (1) What is the molecular mechanism which simultaneously induces two opposite effects during ␣ 1 -AR stimulation?; (2) What are the molecular components for evoking the negative effect on I Ca by ␣ 1 -AR stimulation? We postulated that these 2 opposite effects simultaneously occur via (1) different ␣ 1 -AR subtypes, ␣ 1A and ␣ 1B , which are the dominant receptor subtypes in mammalian heart 1,4 and (2) subtype-specific intracellular signal transduction pathways. The aims of this study are to characterize the effects of ␣ 1 -AR subtype-selective stimulation on I Ca and to clarify the ␣ 1 -AR subtype-specific signaling pathway for the regulation of I Ca . Here, we show the direct evidence that cardiac ␣ 1 -AR signaling diverges at the level of the ␣ 1 -AR subtype and G protein, which produce the opposite effects on I Ca in rat ventricular myocyes. Alpha 1A -AR coupled with G q/11 and activated phospholipase C (PLC)-PKC-CaMKII pathway, which evoked the potentiation of I Ca . In contrast, ␣ 1B -AR interacted with G o , of which the ␤␥-complex could directly inhibit I Ca . These results represent the whole picture of intracellular mechanism for the unique regulation of VLCC by cardiac ␣ 1 -AR signaling and also provide the significant insight into the Figure 1. Detection and cellular localization of ␣ 1 -AR subtypes in rat ventricle. A, Detection of ␣ 1 -AR subtypes in rat ventricle by Western immunoblot (IB) using specific antibodies against ␣ 1A -(left), ␣ 1B -(middle) or ␣ 1D -AR (right). Each well contained 50-g membrane protein from rat ventricular myocytes (heart), urinary bladder (bladder), brain, or liver. B, Confocal images of isolated ventricular myocytes labeled with ␣ 1 -AR subtype-specific antibody (␣ 1A -or ␣ 1B -AR) (red, left) and the plasma membrane marker Wheat Germ Agglutinin-FITC (WGA) (green, middle). The overlay images are also shown (right). Barsϭ10 m. C, Immunoelectron microscopic images of ventricular tissue labeled with 15-nm gold-␣ 1A -AR (left) or ␣ 1B -AR (right). A high intensity of gold labeling was observed directly under T-tubule membranes (indicated by arrows). Mt indicates mitochondrion; Z, Z-line; Barϭ500 nm. regulation of cardiac excitation-contraction coupling by ␣ 1 -AR subtype-specific signaling.

Materials and Methods
For details, please see the Data Supplement (available online at http://circres.ahajournals.org). Single ventricular myocytes and papillary muscles were prepared from adult male Wistar rats (300 to 400 g; Sankyo Labo Service, Tokyo, Japan). 5 The measurement of I Ca using a perforated patch clamp, 5 Western immunoblot, 5 immunoprecipitation, 6 cAMP determination using an enzyme immunoassay, 7 and immunofluorescence microscopy 5 were performed on freshly isolated ventricular myocytes. Papillary muscles were used for immunoelectron microscopy. 5 All results are shown as meanϮSD. Bars in the graphs indicate SD. Paired data were evaluated by Student t test. For multiple comparisons, 1-way or 1-way repeated ANOVA followed by Bonferroni post hoc test with the significance level set at PϽ0.05.

Detection and Cellular Localization of ␣ 1 -AR Subtypes in Cardiomyocytes
The protein expression of ␣ 1 -AR subtypes in isolated adult rat ventricular myocytes was confirmed by Western immunoblot with the commercially available antibodies against ␣ 1A -, ␣ 1B -, and ␣ 1D -AR ( Figure 1A). In the membrane proteins from cardiomyocytes and urinary bladder, single bands were detected with the expected molecular size for glycosylated ␣ 1A -AR (68 kDa) 8 using specific antibody against human ␣ 1A -AR ( Figure 1A, left). However, in the parallel measurement with membrane proteins from rat brain, no positive band was observed. The specific antibody against human ␣ 1B -AR showed a major band with the expected molecular size for glycosylated ␣ 1B -AR (Ϸ80 kDa) 9 in rat cardiomyocytes, liver, and brain ( Figure 1B, middle). The ␣ 1D -AR (60 kDa) was only found in rat brain; no significant bands were observed in cardiomyocytes and liver cells using the specific antibody against rat ␣ 1D -AR ( Figure 1A, right). These results show that ␣ 1A -and ␣ 1B -AR (but not ␣ 1D -AR) were detectable at the protein level in our preparation of cardiomyocytes. Thus, we focused on the role of these 2 subtypes of ␣ 1 -AR (␣ 1A and ␣ 1B ) in native cardiomyocytes in the following experiments.
We determined the cellular localization of ␣ 1A -AR and ␣ 1B -AR in cardiac cells using an immunofluorescence microscope ( Figure 1B). In ventricular myocytes, ␣ 1A -AR was detectable at the plasmalemma and along the Z-lines, which coincides with the sarcolemmal invaginations termed transverse tubules (T-tubules) where the majority of VLCC are located. 10 On the other hand, ␣ 1B -AR was not detectable at the plasmalemma; rather it was localized at T-tubules and intercalated disks. The light microscopic images obtained from papillary muscle also showed a similar tendency of the localization of ␣ 1A -AR and ␣ 1B -AR as observed in the isolated cells (supplemental Figure I).
To confirm the detailed subcellular localization of ␣ 1A -AR and ␣ 1B -AR, ultrathin cryosections of the left ventricular papillary muscles were incubated with these receptor subtype-specific antibodies ( Figure 1C). The membranes of T-tubules were specifically labeled with the antibodies against ␣ 1A -AR and ␣ 1B -AR. These results suggest that 2 ␣ 1 -AR subtypes (␣ 1A and ␣ 1B ) are detectable at the protein level in cardiac membrane, and they are preferentially localized at the T-tubules.

Alpha 1A -AR Stimulation Showed Only a Positive Effect on I Ca Without a Negative Effect
Alpha 1 -AR stimulation by the nonsubtype selective agonist, 10 mol/L phenylephrine (Phe), showed a biphasic change in I Ca measured using the perforated patch clamp (a transient decrease followed by a sustained increase) in the presence of ␤-AR antagonist, 1 mol/L bupranolol, which we used previously 5 (Figure 2A). Similar results were obtained when we used another ␤-AR antagonist, 1 mol/L propranolol, as shown in supplemental Figure II. Following experiments were all performed in the presence of 1 mol/L bupranolol.
Next we observed the effect of selective ␣ 1A -AR stimulation on I Ca by using the selective ␣ 1A -AR agonist A61603. Fifteen-minute treatment with A61603 (0.1 mol/L) evoked only potentiation of I Ca ( Figure 2B) without changing the current-voltage relationship (supplemental Figure III), and there was no negative effect in the initial period, which was observed in the presence of nonsubtype selective ␣ 1 -AR stimulation by Phe (see Figure 2A). This positive effect after 15-minute treatment with A61603 was saturated at 1 mol/L A61603 (0.1 mol/L, 35.36Ϯ14.37%, nϭ6; 1 mol/L, 42.64Ϯ27.49%, nϭ8; Pϭ1.00) and was blocked by the selective ␣ 1A -AR antagonist, 2 mol/L WB4101 (nϭ5, data not shown). All concentrations of A61603 (0.1 to 1 mol/L) used showed only a positive without a negative effect ( Figure 2C).
As we previously reported that the positive effect of ␣ 1 -AR stimulation on I Ca is evoked through a PKC-and CaMKIIdependent mechanism, 5 next we investigated the involvement of PKC and CaMKII in the signaling pathways which evoke the potentiation of I Ca during ␣ 1A -AR stimulation. In the presence of a PKC inhibitor chelerythrine, the positive effect of A61603 was not observed. CaMKII inhibition by KN-93 or autocamtide-2 inhibitory peptide (AIP; a membranepermeable and a highly specific peptide type inhibitor of CaMKII) also abolished the potentiation of I Ca by A61603 ( Figure 2D). Moreover, in the presence of a PLC inhibitor, U73122, the positive effect of A61603 completely disappeared ( Figure 2D). These results suggest that ␣ 1A -AR stimulation shows only a positive effect on I Ca and this effect is mediated through the PLC-PKC-CaMKII pathway.

Alpha 1B -AR Stimulation Showed Only a Negative Effect on I Ca Without a Positive Effect
We investigated the effect of ␣ 1B -AR stimulation on I Ca by the application of a nonsubtype selective ␣ 1 -AR agonist (Phe) with a selective ␣ 1A -AR antagonist (WB4101), because no selective ␣ 1B -AR agonist is available at present. 4 Ten-minute exposure to 2 mol/L WB4101 significantly decreased I Ca without changing the shape of the current-voltage relationship (supplemental Figure III) and reached another steady state ( Figure 2F, red diamonds). In the continuous presence of WB4101, 10 mol/L Phe only decreased I Ca ( Figure 2E and 2F) without changing the shape of the current-voltage relationship (supplemental Figure III). In contrast, 1 mol/L Phe (no negative effect on I Ca was observed at this concentra-

PKC Was Activated After ␣ 1A -AR Stimulation, but Not After ␣ 1B -AR Stimulation
We previously showed that the positive effect of ␣ 1 -AR stimulation on I Ca is dependent on PKC activity. 5 Therefore, we examined the involvement of PKC in the signaling pathway after ␣ 1A -or ␣ 1B -AR stimulation. One of the hallmarks of PKC activation is the translocation of soluble enzymes to particle fractions, presumably near their protein substrates that include sarcolemmal proteins. 11 We determined the isoform-specific PKC translocation to the membrane by calculating the membrane-to-cytosolic (M/C) ratio before and after selective ␣ 1 -AR-subtype stimulation ( Figure  3A to 3F). In our preparation, one of the Ca 2ϩ -dependent PKC isoforms, PKC␣ was not significantly translocated by nonsubtype selective ␣ 1 -AR stimulation or by subtype-selective ␣ 1 -AR stimulations, although endogenous PKC activator, phorbol 12-myristate 13-acetate (PMA) did translocate PKC␣ from cytosol to membrane and filament fractions ( Figure 3A and 3D). On the contrary, significant translocation of the Ca 2ϩ -independent PKCs (␦ and ) from cytosol to membrane fraction was found 15 minutes after ␣ 1A -AR stimulation ( Figure 3B, 3C, 3E, and 3F). However, no remarkable translocation of PKC after ␣ 1B -AR stimulation was observed ( Figure 3E and 3F).
Phosphorylation of PKC itself was also measured because it is another hallmark of PKC activation. 11 Immunoreactivity of PKC phosphorylation at the hydrophobic motif and PKC␦ phosphorylation at the activation loop significantly increased after ␣ 1A -AR stimulation or PMA treatment in the membrane fraction. However, the amount of phosphorylated PKC␣ in the membrane fraction did not increase after ␣ 1A -or ␣ 1B -AR stimulation (supplemental Figure V).
We identified the subcellular localization of the activated PKC isozymes to elucidate their roles in the regulation of VLCC before and after selective ␣ 1 -AR-subtype stimulation by using an immunofluorescence microscope ( Figure 4A to 4C). Significant translocation of PKC␣ was not observed after the treatment with Phe as shown in Western immunoblot ( Figure 4A). Most of PKC␦ was localized in the nucleus or at the nuclear membrane, but the remainder was diffusely distributed in the cytosol at rest ( Figure 4B). After Phe treatment, a striated pattern also became visible, which was in accordance with the location of T-tubules ( Figure 4B). PKC was diffusely distributed in the cytosol before stimulation ( Figure 4C). After Phe treatment, PKC was accumulated at the T-tubules and intercalated disks ( Figure 4C).
These results suggest that Ca 2ϩ -independent PKCs were activated and the activated PKCs were redistributed to the membrane fraction, presumably to the T-tubules after ␣ 1A -AR stimulation. However, there was no obvious involvement of PKC in the ␣ 1B -AR signaling pathway.

Negative Change in I Ca During ␣ 1 -AR Stimulation Was Mediated via the Pertussis Toxin-Sensitive G Protein Pathway
We showed that the positive effect on I Ca caused by ␣ 1A -AR stimulation was dependent on PKC, but the negative effect of ␣ 1B -AR stimulation was independent of PKC activation (Figures 2 and 3). Several reports demonstrated that ␣ 1 -AR couples not only with G q/11 which in turns leads to activation of PLC and PKC, but also with the pertussis toxin (PTX)sensitive G proteins, and it shows diverse physiological effects in cardiomyocytes. 1,4,12 Therefore, we hypothesized that ␣ 1B -AR functionally couples with other G proteins, and we examined the involvement of PTX-sensitive G protein in the regulation of I Ca by ␣ 1 -AR stimulation.
Inhibition of G i/o -protein by PTX in our preparations was confirmed by the ability of PTX to block the muscarinic inhibition of I Ca in the presence of ␤-AR stimulant (supplemental Figure VI). Treatment of PTX significantly inhibited the negative effect by 10 mol/L Phe at 2 minutes and then enhanced the positive effect at 15 minutes ( Figure 5A). Moreover, we separately investigated the effects of ␣ 1 -AR subtype-selective stimulation on I Ca in PTX-treated cells. We confirmed that the negative effect by ␣ 1B -AR stimulation was blocked by PTX ( Figure 5C), but the magnitude of the positive effect by ␣ 1A -AR stimulation did not alter after PTX treatment ( Figure 5B). These results indicate that the negative phase of I Ca during ␣ 1B -AR stimulation (Figure 2A) was produced through PTX-sensitive G protein (G i/o ) pathways. In adult rat cardiomyocytes at least 3 subtypes of PTXsensitive G ␣ (G ␣i-2 , G ␣i-3 , G ␣o ) are expressed at the mRNA level and are detectable at the protein level. 13 Therefore, the possibility that ␣ 1 -AR subtypes directly couple with these PTX-sensitive G proteins was examined by coimmunoprecipitation of these G ␣ -subunits with anti-␣ 1A -or anti-␣ 1B -AR antibody. The immunoprecipitants were analyzed by Western immunoblot probing with the antibodies against G ␣ -subunits, of which specificities were checked by using the recombinant G ␣ -subunits (see supplemental Figure VII). The ␣ 1A -AR antibody coimmunoprecipitated G ␣q/11 , whereas the ␣ 1B -AR antibody coimmunoprecipitated G ␣o ( Figure 5D). Moreover, immunofluorescence images with the specific antibodies against ␣ 1B -AR and G ␣o showed that G ␣o was colocalized with ␣ 1B -AR at T-tubules ( Figure 5E). Thus, these results indicate that the ␣ 1A -AR couples with G q/11 -protein in a classical coupling mode, which activates the PLC-diacylglycerol-PKC pathway and that ␣ 1B -AR is linked to G o at the T-tubules and evokes the negative phase in I Ca .

Negative Effect of ␣ 1 -AR Stimulation on I Ca Is Mediated Through ␤␥-Complex of G Protein
Our biochemical and electrophysiological results indicated that ␣ 1B -AR interacted with one of the PTX-sensitive G proteins, G o . However, the functional roles of G o -protein in native cardiomyocytes have not been clarified. We postulated here that ␣ 1B -AR-G o interaction could inhibit the VLCC activity through (1) the decrease of basal cAMP level (eg, by the inhibition of adenylyl-cyclase activity as in the case of G i 4 ), or (2) stimulation of protein phosphatase activity, followed by the reduction of basal phosphorylation level of the VLCC. However, we found that the basal cAMP level in our preparations did not significantly change during ␣ 1 -AR stimulation as described previously 14 (Figure 6A), and negative effect of I Ca by Phe was clearly observed even in the presence of cAMP-dependent protein kinase (PKA) inhibitor (1 mol/L H-89; Figure 6B). Thus, the inhibition of cAMP-PKA signaling is not involved in the mechanism for evoking negative phase of I Ca . Moreover, we pretreated the cells with a protein phosphatase inhibitor, calyculin A in the presence of H-89 15 and then investigated the effects of Phe in the continuous presence of calyculin A and H-89. Under this condition, 15 we still observed the negative phase of I Ca , suggesting that activation of phosphatases followed by the reduction of basal VLCC phosphorylation is not involved in the negative phase (supplemental Figure VIII).
Several reports stated that the ␤␥-complex of heterotrimetric G o -protein directly interacts with N-type or L-type Ca 2ϩ channels to inhibit their activity. 16,17 Moreover, a depolarization pulse applied to the membrane before channel activation is known to counteract this inhibition. 16 Therefore, we next observed the effect of a nonsubtype selective ␣ 1 -AR agonist (10 mol/L Phe) on I Ca using this prepulse depolarization protocol ( Figure 7A).
Recording with this prepulse depolarization, there was no significant transient decrease of I Ca for up to 2 minutes after the application of 10 mol/L Phe ( Figure 7A and 7B). Fifteen minutes after the application of Phe, I Ca was significantly increased ( Figure 7A and 7B). Thus, the current inhibition at Figure 6. Inhibition of cAMP-PKA signaling was not involved in the mechanism for evoking negative phase of I Ca during ␣ 1 -AR stimulation. A, cAMP concentration in isolated rat ventricular myocytes treated with 10 mol/L phenylephrine (Phe) for 2 or 15 minutes (nϭ3 or 4). cAMP concentration after treatment of 100 nmol/L isoproterenol (Iso) for 15 minutes in the absence of ␤-AR antagonist was also shown as the positive control (nϭ4). *PϽ0.05, compared to the control (nontreated cells; CTR). B, The effect of 10 mol/L Phe on I Ca in the presence of the selective PKA inhibitor, 1 mol/L H-89 (nϭ6). The amplitudes of the currents at each period were normalized by the current before the application of Phe. *PϽ0.05, compared to the current before stimulation (0 minutes).  the initial stage (Ϸ2 minutes) induced by ␣ 1 -AR stimulation was not attributable to the reduction of basal phosphorylation level of VLCC, but was possibly produced by the direct interaction of ␤␥-complex of G o with VLCC.

Discussion
In this study, we elucidated the differences between cardiac ␣ 1A -and ␣ 1B -AR signaling pathways and provide direct evidence indicating that different G proteins (and kinases) are involved in the respective subtype-specific signaling pathway and induce opposite changes in I Ca in native cardiomyocytes ( Figure 8). We showed that ␣ 1A -and ␣ 1B -AR were functionally expressed at T-tubules where VLCC is concentrated, 10 but ␣ 1D -AR was not detected at protein level and was not functionally expressed in our preparations (supplemental Figure IV). Furthermore, we clearly separated the effect of ␣ 1A -or ␣ 1B -AR stimulation from that of nonsubtype selective stimulation on I Ca by pharmacological procedure and clarified the detail of each signaling pathway by biochemical and morphological techniques. Alpha 1A -AR and ␣ 1B -AR signaling pathways couple with different G proteins, G q/11 and G o , respectively and produce different functional outcomes; ␣ 1A -AR stimulation activates G q/11 -PLC-diacylglycerol-PKC-CaMKII pathway and increases I Ca . On the contrary, ␣ 1B -AR interacts with G o and inhibits the VLCC activity.

Alpha 1A -AR-Gq-PKC Signaling Pathway Induces Potentiation of I Ca
In this study, we showed that ␣ 1A -AR was expressed at T-tubules and also demonstrated that ␣ 1A -AR stimulation did affect the VLCC activity which was confirmed by using ␣ 1A -AR selective agonist, A61603. Alpha 1A -AR pathway potentiated I Ca in native cardiomyocytes, which is mediated through a PKC-dependent mechanism (Figure 8). PKC is a phospholipid-dependent Ser-Thr kinase, and most isoforms of the PKC are activated as a result of receptor-dependent activation of PLC and the hydrolysis of membrane phosphoinositides. 1 Although all cloned subtypes of ␣ 1 -AR can induce PLC activation and inositol phosphate formation, 18 receptor subtype-specific activation or downregulation of PKC has been reported in cultured neonatal cardiomyocytes. 19 In cardiac tissue, the isoforms of 1 Ca 2ϩ -dependent PKC (PKC␣) and 2 Ca 2ϩ -independent PKCs (novel PKCs; PKC␦ and PKC) are at least detectable at the protein level. 20 In our preparations, we demonstrated that only ␣ 1A -AR stimulation induces the activation of novel PKCs and translocates them to the cell membrane structure called T-tubules, and that ␣ 1B -AR stimulation did not show any activation of PKC. This result is consistent with the previous report that ␣ 1B -or ␣ 1D -AR signaling pathway does not have any influences on PKC activity. 19 We did not detect any significant activation of PKC␣ after ␣ 1 -AR stimulation, indicating that PKC␣ was not involved in ␣ 1 -AR signaling 19 in our preparations. Moreover, we directly showed the interaction of ␣ 1A -AR and G ␣q/11 which activates PLC. 1,6 The ␣ 1A -AR-G q/11 -PLC-diacylglycerol pathway activated novel PKCs and translocated them to T-tubules where VLCC and CaMKII are prevalent. 5,21 The translocated PKC could activate CaMKII at T-tubules, 5 and then the activated CaMKII could directly potentiate I Ca through the phosphorylation of ␣ 1c and /or ␤ subunits of the channel. 21,22 Thus, ␣ 1A -AR signaling components preferentially localized at the T-tubules and efficiently regulated cardiac VLCC, as in the case of cardiac endothelinreceptor signaling. 23

Alpha 1B -AR-G o Interaction Induces the Inhibition of I Ca
We showed that ␣ 1B -AR was expressed at T-tubules and also demonstrated that ␣ 1B -AR stimulation did affect the VLCC activity confirmed by pharmacological procedures. Alpha 1B -AR stimulation inhibited I Ca , which is mediated through PKC-independent mechanisms (Figure 8). Our biochemical studies indicated that ␣ 1B -AR stimulation shows less influence on the PKC activity than ␣ 1A -AR stimulation. This result is compatible with the previous reports that ␣ 1B -AR subtype has less potency for the stimulation of phosphoinositide hydrolysis than ␣ 1A -AR. 24,25 We found that ␣ 1B -AR coupled with G o instead of G q/11 and this pathway brought about negative modulation of I Ca , which was not attributable to the decline of basal phosphorylation level of VLCC by the inhibition of cAMP-PKA pathway or activation of protein phosphatases ( Figure 6 and supplemental Figure VIII). Our result is consistent with the previous results obtained using a constitutively active mutant of ␣ 1B -AR or the technique of the overexpression of wild-type ␣ 1B -AR, which shows the possibility that only ␣ 1B -AR (but not ␣ 1A -AR) couples with the PTX-sensitive pathway. 26,27 I Ca inhibition through G o was reported in secretory cells 17 and cardiac cells from genetically engineered mice, 28 but the functional role of G o in native cardiomyocytes is still poorly understood. Ivanina et al showed that the direct binding of G ␤␥ subunit of G o to VLCC inhibits the channel activity in heterologous expression systems. 29 They also reported that the basal intracellular Ca 2ϩ level is essential for this inhibition, which is consistent with our previous data. 5 We also showed that ␣ 1B -AR and G o colocalize with VLCC at T-tubules, and we propose the working model that ␤␥-complex of G o protein could directly inhibit VLCC.

Conclusion
In conclusion, our results represent the evidence that the unique combinations of ␣ 1 -AR subtypes and specific G proteins form subtype-specific signal transduction pathways, which induce the opposite effects on VLCC in native cardiac cells. The coupling of specific ␣ 1 -AR subtypes with PTX-sensitive G protein could exhibit the negative feedback response to ␣ 1 -AR stimulation, and this mechanism would contribute to the protection of the heart from Ca 2ϩ overload as in the relation between ␤ 1 -and ␤ 2 -AR. The approach of characterizing the receptor subtype-specific interacting G protein will provide new insight to elucidate the whole picture of the subtype-specific signaling pathway in native cardiomyocytes, and further could lead to an understanding of the functional roles of each ␣ 1 -AR subtype under physiological and pathophysiological condition.