Laminin enhances β₂-adrenergic receptor stimulation of L-type Ca²⁺ current via cytosolic phospholipase A₂ signalling in cat atrial myocytes

Introduction

In cardiac muscle both β₁-adrenergic receptor (AR) and β₂-AR signalling regulates L-type Ca²⁺ current (I_Ca,L), with β₁-ARs being the predominant signalling mechanism. In general, β₁-ARs are coupled exclusively to G_s/adenylate cyclase (AC)/cAMP-dependent PKA (cAMP/PKA) signalling whereas β₂-ARs are coupled to both G_s and G_i signalling pathways. β₂-ARs act via G_s to stimulate AC/cAMP/PKA signalling and via G_i to stimulate a variety of signalling pathways including cytosolic phospholipase A₂ (cPLA₂)/arachidonic acid (AA) signalling (Pavoine et al. 1999).

The extracellular matrix (ECM) is composed of a variety of proteins including laminin, fibronectin and collagen. Adult cardiomyocytes primarily attach to laminin (LMN) via low-affinity cell surface receptors known as integrins (Terracio et al. 1991). Along with its role in cell attachment, the ECM–β-integrin complex is an important modulator of intracellular signalling, i.e. so-called ‘outside-in’ signalling (Hynes, 1992; Schwartz et al. 1995; Dedhar, 1999). Atrial disease is commonly associated with atrial fibrosis, i.e. structural remodelling, which is associated with increases in ECM proteins.

We previously reported that attachment of atrial myocytes to LMN acts via β₁-integrin receptors to decrease β₁-AR and increase β₂-AR stimulation of I_Ca,L (Wang et al. 2000b). These changes in β-AR are similar to those that occur during remodelling of the failing heart in humans (Bristow et al. 1986) and experimental animals (Kiuchi et al. 1993). In addition, we reported that LMN decreases β₁-AR stimulation via inhibition of G_s/AC/cAMP signalling (Wang et al. 2009). The fact that LMN inhibits G_s/AC/cAMP signalling and yet increases β₂-AR signalling suggests that LMN stimulates β₂-AR signalling through a mechanism independent of G_s/AC/cAMP signalling. In this regard, Pavoine et al. (1999) reported that β₂-AR activation of cPLA₂ signalling may compensate for depressed AC/cAMP signalling. Therefore, the present study sought to determine whether atrial cell attachment to LMN enhances β₂-AR signalling via cPLA₂ signalling and whether this mechanism is activated by concomitant inhibition of cAMP/PKA signalling. The present findings provide new insights into the potential role of the ECM to remodel β₂-AR signalling in atrial muscle.

Methods

Adult cats of either sex were anaesthetized with sodium pentobarbital (50 mg kg⁻¹, i.p.). Once fully anaesthetized, a bilateral thoracotomy was performed, and the heart was rapidly excised and mounted on a Langendorff perfusion apparatus. After enzyme (collagenase; type II, Worthington Biochemical) digestion, atrial myocytes were isolated as previously reported (Wang et al. 2000a). The animal and experimental protocols used in this study were approved by the Institutional Animal Care and Use Committee (IACUC) of Loyola University Medical Center, Maywood, IL, USA. IACUC prescribed the rules for the animal care and supervised their enforcement. This study required 34 animals.

Electrophysiological recordings from myocytes were performed in the perforated (nystatin) patch whole-cell configuration at room temperature, as previously described (Wang et al. 2000a). L-type Ca²⁺ current (I_Ca,L) was activated by depolarizing pulses from a holding potential of −40 mV to 0 mV for 200 ms every 5 s and measured in relation to steady-state current. β₂-AR stimulation was achieved by 0.1 μm zinterol (zint-β₂-AR), a specific β₂-AR agonist that activates both G_i and G_s signalling (Wang et al. 2002). β₂-AR stimulation also was achieved by 0.1 μm fenoterol (fen-β₂-AR), a specific β₂-AR agonist that acts exclusively via G_s signalling (Dedkova et al. 2002). β₁-AR stimulation was achieved by 0.1 μm isoproterenol (isoprenaline; ISO-β₁-AR) plus 0.01 μm ICI 118,551, a specific β₂-AR antagonist. Each agonist was applied for approximately 4 min and the effects of each agonist on peak I_Ca,L amplitude was recorded at the maximum steady-state response. Generally, we compared two groups of freshly isolated atrial myocytes obtained from the same hearts: control cells plated on uncoated glass coverslips (−LMN), and test cells plated on glass coverslips coated with LMN (+LMN; 40 μg ml⁻¹) for at least 2 h, as previously described (Wang et al. 2000a). Control experiments have shown that atrial myocytes plated on poly-l-lysine, a non-specific substrate for cell attachment, fail to show changes in β-AR signalling similar to cells attached to LMN (Wang et al. 2000b). It is also important to note that freshly isolated cardiomyocytes retain LMN on their cell surface (Lundgren et al. 1988). Moreover, freshly isolated cardiomyocytes exhibit responses to cholinergic and β-AR stimulation which are similar to multicellular cardiac preparations. For example, freshly isolated atrial myocytes, like all cardiac muscle, exhibit predominantly β₁-AR over β₂-AR signalling (Wang et al. 2000b). Cell attachment to LMN decreases the β₁-/β₂-AR signalling ratio resulting in predominantly β₂-AR over β₁-AR signalling (Wang et al. 2000b). Therefore, atrial cell attachment to LMN is not restoring LMN-mediated signalling somehow lost during the cell isolation procedure. Our findings are consistent with the idea that LMN activates additional integrin receptors on atrial myocytes which alters normal β−AR function.

Drugs used in this study include: zinterol; arachidonyltrifluoromethyl ketone (AACOCF₃); H-89; KT5720; pertussis toxin (PTX); fenoterol; isoproterenol; ICI 118,551; 1,2-bis-(o-aminophenoxy)-ethane-N,N,N′,N′-tetraacetic acid, tetraacetoxymethyl ester (BAPTA-AM); cytochalasin D; U0126 (all from Sigma Chemical); arachidonic acid (AA) (Biomol).

Data are mean ± standard error of the mean (s.e.m.). Measurements were analysed using either Student's paired or unpaired t test for significance at P < 0.05. Multiple comparisons were performed by ANOVA followed by a Student–Newman–Keuls test with significance at P < 0.05.

Results

Figure 1A–B shows original traces of I_Ca,L recorded from −LMN (A) and +LMN (B) atrial myocytes and the effects of 0.1 μm zinterol (zint) in the absence and presence of 10 μm AACOCF₃, a selective inhibitor of cPLA₂ (Riendeau et al. 1994). The bar graphs (C and D) summarize the data. First, control zint-β₂-AR stimulation of I_Ca,L was significantly greater in +LMN (Ba and D; 230 ± 39%, N= 5) compared to −LMN (Aa and C; 138 ± 16%, N= 4) atrial myocytes (P < 0.05), confirming our previous report (Wang et al. 2000b). In −LMN myocytes AACOCF₃ had no significant effects on basal I_Ca,L amplitude (C; compare open bars) or zint-β₂-AR stimulation of I_Ca,L (Ab and C; 134 ± 8%, N= 5). In +LMN myocytes AACOCF₃ also had no significant effect on basal I_Ca,L amplitude (D; compare open bars), although AACOCF₃ significantly inhibited zint-β₂-AR stimulation of I_Ca,L (Bb and D; 107 ± 6%, N= 4, P < 0.02). As summarized in panels C and D, inhibition of cPLA₂ had no effect on β₂-AR stimulation in −LMN myocytes (C) while it significantly decreased β₂-AR stimulation in +LMN myocytes (D). These findings suggest that cPLA₂ plays little role in β₂-AR regulation of I_Ca,L in freshly isolated cat atrial myocytes. However, in atrial myocytes attached to LMN, β₂-AR signalling may be enhanced via activation of cPLA₂ signalling.

In atrial muscle, β₂-ARs are coupled to both G_s and G_i signalling (Kilts et al. 2000; Wang et al. 2002) and cPLA₂ signalling is coupled to β₂-ARs via pertussis toxin (PTX)-sensitive G_i (Pavoine et al. 1999). We therefore determined the effects of zint-β₂-AR stimulation of I_Ca,L in −LMN and +LMN atrial myocytes either untreated (control) or treated with PTX (5 μg ml⁻¹ for 3 h at 36°C). Figure 2A shows that in −LMN myocytes zint-β₂-AR stimulation of I_Ca,L was significantly larger in cells pre-treated with PTX compared to control (control, 124 ± 4%, N= 5 vs. PTX, 147 ± 7%, N= 4; P < 0.05), consistent with an inhibitory influence of G_i on β₂-AR/G_s signalling. In contrast, in +LMN myocytes zint-β₂-AR stimulation of I_Ca,L was significantly smaller in cells pre-treated with PTX compared to control (control, 176 ± 10%, N= 3 vs. PTX, 99 ± 13%, N= 3; P < 0.02). These findings indicate that LMN enhances β₂-AR stimulation via a G_i signalling mechanism, consistent with cPLA₂ activation.

Another way of testing whether LMN enhances β₂-AR signalling through a G_i-coupled signalling pathway is to determine the effects of fenoterol (fen), a specific β₂-AR agonist that acts exclusively via G_s signalling (Dedkova et al. 2002). If LMN enhances β₂-AR signalling through a G_i-mediated pathway, then LMN should inhibit rather than enhance fen-β₂-AR stimulation of I_Ca,L. In fact, as shown in Fig. 2B, 0.1 μm fen-β₂-AR stimulation of I_Ca,L was significantly smaller in +LMN than in −LMN atrial myocytes (−LMN, 169 ± 14%, N= 5 vs.+LMN, 95 ± 7%, N= 4; P < 0.002). These findings are consistent with our previous reports that LMN decreases G_s/AC/cAMP activity (Wang et al. 2000a, 2009) and strongly supports the idea that LMN enhances β₂-AR signalling via a G_i-coupled signalling pathway, consistent with cPLA₂ signalling.

Activation of cPLA₂ is Ca²⁺ dependent (Leslie, 2004). Therefore, if LMN is enhancing β₂-AR stimulation via cPLA₂ signalling then chelation of intracellular Ca²⁺ should prevent the effects of LMN to enhance zint-β₂-AR signalling. As shown in Fig. 2C, we determined the effects of zint-β₂-AR stimulation of I_Ca,L in −LMN and +LMN myocytes in control and test cells pre-treated with 10 μm BAPTA-AM (15 min), a cell-permeable Ca²⁺ chelator. In −LMN myocytes, basal I_Ca,L was slightly but not statistically larger in BAPTA-treated cells compared to control (open bars) and zint-β₂-AR stimulation of I_Ca,L was unaffected by BAPTA (control; 119 ± 3%, N= 3 vs. BAPTA; 114 ± 17%; N= 4). In +LMN myocytes, basal I_Ca,L was unaffected by BAPTA (open bars). However, zint-β₂-AR stimulation of I_Ca,L was significantly inhibited in BAPTA-treated cells compared to control responses (control; 177 ± 10%; N= 4 vs. BAPTA; 57 ± 9%, N= 4, P < 0.0001). In other words, chelation of intracellular Ca²⁺ prevented the effects of LMN from enhancing β₂-AR signalling. Additional experiments measuring intracellular [Ca²⁺] confirmed that 15 min exposure to BAPTA-AM abolished electrically stimulated Ca²⁺ transients (authors’ unpublished observations). These findings support the idea that Ca²⁺-dependent cPLA₂ plays no role in −LMN myocytes (see Fig. 1) and that LMN enhances zint-β₂-AR signalling via Ca²⁺-dependent cPLA₂ signalling.

Cytosolic PLA₂ hydrolyzes the sn-2 fatty acyl ester bonds of membranous phospholipids, leading to the release of arachidonic acid (AA). We therefore determined whether 5 μm AA stimulates I_Ca,L in −LMN and +LMN atrial myocytes. AA stimulated I_Ca,L to a similar extent in −LMN (57 ± 10%) and +LMN (65 ± 12%) myocytes (N= 8). These findings indicate that AA stimulates I_Ca,L in atrial myocytes and that LMN does not enhance β₂-AR signalling by increasing the sensitivity of I_Ca,L channels to AA.

Our previous work indicates that LMN-mediated enhancement of β₂-AR signalling is accompanied by inhibition of cAMP signalling (Wang et al. 2000a,b;). The following experiments were designed to determine whether β₂-AR/cPLA₂ signalling is activated by inhibition of cAMP/PKA. In other words, does pharmacological inhibition of cAMP/PKA in −LMN myocytes mimic the effects of +LMN myocytes to enhance β₂-AR signalling. As shown in Fig. 3A–C and summarized in panel D, we determined the effects of zint-β₂-AR stimulation in −LMN myocytes in the absence and presence of 5 μm H-89, an inhibitor of cAMP/PKA (Chijiwa et al. 1990). In control cells (A) zint-β₂-AR stimulation significantly increased I_Ca,L (D; 130 ± 13%; N= 11). In another group of −LMN myocytes from the same hearts (B), 5 μm H-89 decreased basal I_Ca,L by −21 ± 2% (D; open bar, N= 4, P < 0.01), consistent with inhibition of basal cAMP/PKA activity. Interestingly, in the presence of H-89 (B), zint-β₂-AR stimulation of I_Ca,L (D, 185 ± 13%; N= 4) was significantly enhanced compared to control zint-β₂-AR stimulation (130 ± 13%, P < 0.05). In a third group of −LMN myocytes (C), 10 μm AACOCF₃ plus H-89 decreased basal I_Ca,L by −31 ± 3% (D; open bar, N= 4; P < 0.01) and significantly inhibited zint-β₂-AR stimulation of I_Ca,L (D; 59 ± 8%, N= 4) compared to either zinterol alone or zinterol plus H-89. These findings are in contrast to those in −LMN myocytes in the absence of H-89, in which AACOCF₃ had no effect on zint-β₂-AR stimulation of I_Ca,L (Fig. 1).

Similar results were obtained with KT5720 (KT), another cAMP/PKA inhibitor (Kase et al. 1987). As shown in Fig. 4A–C and summarized in panel D, in control (A), zint-β₂-AR stimulation significantly increased I_Ca,L (D; 107 ± 8%, N = 5). In another group of cells from the same hearts (B), 1 μm KT decreased basal I_Ca,L by 21 ± 2% (D; open bar; N= 6; P < 0.01) and in the presence of KT (B), zint-β₂-AR stimulation of I_Ca,L (D, 220 ± 11%, N= 6) was significantly enhanced compared to control zint-β₂-AR stimulation (107 ± 8%, P < 0.05). In a third group of −LMN myocytes (C), 10 μm AACOCF₃ plus KT decreased basal I_Ca,L by −16 ± 1% (D; open bar, N= 4; P < 0.005) and significantly inhibited zint-β₂-AR stimulation of I_Ca,L (D; 67 ± 12%, N= 4) compared to either zinterol alone or zinterol plus KT. Again, these findings are in contrast to those in −LMN myocytes in the absence of KT, in which AACOCF₃ had no effect on zint-β₂-AR stimulation of I_Ca,L (Fig. 1). Together, these results indicate that pharmacological inhibition of cAMP/PKA mimics the effects of +LMN myocytes to enhance β₂-AR stimulation of I_Ca,L via activation of cPLA₂ signalling.

To confirm that H-89 was acting to inhibit cAMP/PKA activity, we determined the effect of H-89 on β₁-AR stimulation (see Methods) which is mediated exclusively through cAMP/PKA. In −LMN cells, 5 μm H-89 decreased basal I_Ca,L by 22 ± 4% (N= 3). In the presence of H-89, ISO-β₁-AR stimulation of I_Ca,L was almost completely blocked (control, 206 ± 25, N= 4 vs. H-89, 37 ± 3%, N= 3; P < 0.005).

If inhibiting cAMP/PKA activity enhances β₂-AR/G_i/cPLA₂ signalling then fen-β₂-AR stimulation, which acts exclusively via G_s/cAMP/PKA signalling, should not be enhanced by inhibition of cAMP/PKA. As shown in Fig. 5A–C and summarized in panel D, in control −LMN myocytes (A) fen-β₂-AR stimulation significantly increased I_Ca,L (D; 134 ± 16%, N= 4). In another group of −LMN myocytes from the same hearts (B), 5 μm H-89 decreased basal I_Ca,L by −31 ± 4% (D, open bar), consistent with inhibition of basal cAMP/PKA. In contrast to zint-β₂-AR stimulation, H-89 almost abolished fen-β₂-AR stimulation of I_Ca,L (D; 27 ± 6%, N= 4). This finding is similar to the effects of H-89 to block ISO-β₁-AR stimulation, which also is mediated exclusively via G_s/cAMP/PKA signalling. In a third group of −LMN myocytes (C), the addition of AACOCF₃ plus H-89 had no significant additional effect on basal I_Ca,L (D; open bar) or fen-β₂-AR stimulation of I_Ca,L (D; 37 ± 7%, N= 4). These experiments further confirm that H-89 acts to block cAMP/PKA and support the conclusion that inhibition of cAMP/PKA enhances zint-β₂-AR stimulation through G_i-mediated cPLA₂ signalling.

How does inhibition of cAMP/PKA enhance zint-β₂-AR/cPLA₂ signalling? Phosphorylation of cPLA₂ by mitogen-activated protein kinases (MAPKs) contributes to the activation of cPLA₂ (see Gijon & Leslie, 1999). In addition, cAMP/PKA can inhibit MAPKs ERK1/2 signalling (see Gerits et al. 2008). Therefore, inhibition of the inhibitory influence of cAMP/PKA may enhance zint-β₂-AR/cPLA₂ via activation of ERK1/2 signalling. In Fig. 6A and B we therefore determined whether inhibition of ERK1/2 could prevent the effect of cAMP/PKA inhibition from enhancing zint-β₂-AR stimulation in −LMN myocytes. ERK1/2 was inhibited by U0126, a potent and selective inhibitor of mitogen-activated protein kinase kinase (MEK1/2) (Favata et al. 1998). MEK1/2 is the immediate upstream activator of ERK1/2. As shown in Fig. 6A and B, U0126 was tested in control −LMN myocytes (A) and −LMN myocytes in which cAMP/PKA was inhibited by 1 μm KT5720 (KT) (B). Cells were exposed to U0126 for at least 15 min before recordings were obtained. U0126 had no significant effects on basal I_Ca,L amplitude (control, 4.2 ± 0.1 pA pF⁻¹; N= 7 vs. U0126, 4.6 ± 0.1 pA pF⁻¹; N= 9). As shown in Fig. 6A, in control −LMN myocytes zint-β₂-AR stimulation of I_Ca,L was not significantly affected by U0126 (control, 115 ± 8%; N= 3 vs. U0126, 106 ± 15%; N= 5). However, as shown in Fig. 6B, U0126 significantly inhibited the effect of cAMP/PKA inhibition (KT) from enhancing zint-β₂-AR stimulation of I_Ca,L (KT, 171 ± 7%; N= 5 vs. KT + U0126, 54 ± 5%; N= 4, P < 0.0002). As shown in Fig. 6C, U0126 also prevented +LMN myocytes from enhancing zint-β₂-AR stimulation of I_Ca,L (control, 147 ± 15%; N= 3 vs. U0126, 70 ± 12%; N= 4, P < 0.002). These findings indicate that LMN-mediated inhibition of cAMP/PKA acts, at least in part, via activation of ERK1/2 to enhance zint-β₂-AR signalling.

We previously reported in atrial myocytes that inhibition of actin polymerization by cytochalasin D prevents LMN-mediated changes in β-AR signalling (Wang et al. 2000b). If our current hypothesis is correct, then inhibition of actin polymerization also should prevent the enhancement of β₂-AR signalling induced by pharmacological inhibition of cAMP/PKA in −LMN myocytes. We therefore determined the effects of cytochalasin D (cyto D), an agent that prevents actin polymerization (Carter, 1967), on zint-β₂-AR stimulation of I_Ca,L in the absence (Fig. 7A) and presence (B) of 5 μm H-89 in −LMN myocytes. As shown in Fig. 7A, in control −LMN myocytes zint-β₂-AR stimulation significantly increased I_Ca,L (94 ± 13%, N= 4). In −LMN myocytes treated with 10 μm cyto D (30 min), basal I_Ca,L was slightly reduced (–6 ± 1%; open bar, N= 3) and zint-β₂-AR stimulation of I_Ca,L was significantly enhanced (169 ± 10%, N= 3) compared with control responses (94 ± 13%, P < 0.006). This is consistent with the idea that in cardiac myocytes, actin filaments restrict G-protein-coupled receptor-activated AC/cAMP formation (Head et al. 2006). In Fig. 7B we repeated the experiment in −LMN myocytes exposed to 5 μm H-89. As expected, H-89 decreased basal I_Ca,L by 21 ± 2% (N= 3) and enhanced zint-β₂-AR stimulation of I_Ca,L (157 ± 5%, N= 4) compared to control responses in the absence of H-89 (A; 94 ± 13%, N= 4; P < 0.01). In −LMN myocytes pre-treated with cyto D (B), exposure to H-89 produced a typical inhibition of basal I_Ca,L by 20 ± 5% (N= 3). However, in cyto D-treated cells H-89 inhibited (rather than enhanced) zint-β₂-AR stimulation of I_Ca,L (B; 28 ± 13%, N= 3) compared to control responses in the presence of H-89 alone (B; 157 ± 5%; P < 0.002). In other words, cyto D prevented inhibition of cAMP/PKA (H-89) from enhancing zint-β₂-AR stimulation of I_Ca,L, suggesting that β₂-AR/cPLA₂ signalling is dependent on the organization of the actin cytoskeleton. Moreover, the fact that H-89 blocked the effects of cyto D to enhance zint-β₂-AR stimulation of I_Ca,L indicates that the effects of cyto D to enhance zint-β₂-AR stimulation were, in fact, mediated via cAMP/PKA activity.

Discussion

Our previous work indicates that attachment of atrial myocytes to LMN inhibits β₁-AR/G_s/AC/cAMP signalling (Wang et al. 2000a, 2009) and at the same time enhances β₂-AR signalling (Wang et al. 2000b). This implies that LMN is enhancing β₂-AR signalling through a cAMP-independent pathway that is receptor coupled to G_i. In fact, the present study indicates that LMN acts via inhibition of cAMP/PKA to enhance β₂-AR stimulation of I_Ca,L via activation of G_i/ERK1/2/cPLA₂/AA signalling.

The present study shows that PTX significantly inhibited zint-β₂-AR stimulation of I_Ca,L in +LMN myocytes, consistent with the coupling of β₂-ARs via G_i to cPLA₂ signalling (Pavoine et al. 1999). Moreover, in contrast to zint-β₂-AR stimulation, cell attachment to LMN decreased fen-β₂-AR stimulation of I_Ca,L, which is similar to the effects of LMN to decrease β₁-AR signalling (Wang et al. 2000b, 2009). This is consistent with the fact that fen-β₂-AR and β₁-AR signalling are both coupled exclusively via G_s and lends strong support to the idea that LMN enhances β₂-AR signalling via a G_i-mediated mechanism. It seems unlikely that LMN is enhancing G_i because LMN does not enhance muscarinic receptor/G_i-mediated inhibition of basal I_Ca,L in atrial myocytes (Wang et al. 2000a). The role of cPLA₂ is further supported by the present finding that inhibition of cPLA₂ by AACOCF₃ (Riendeau et al. 1994) prevented LMN-mediated enhancement of zint-β₂-AR signalling. In contrast, AACOCF₃ had no effect on zint-β₂-AR stimulation in −LMN myocytes, indicating that cPLA₂ plays little role in β₂-AR signalling in freshly isolated atrial myocytes (Fig. 1). The present study also shows that AA stimulates I_Ca,L, consistent with the stimulatory effects of zint-β₂-ARs acting via cPLA₂ signalling. Moreover, the stimulatory effects of AA were similar in −LMN and +LMN myocytes, indicating that the effects of LMN to enhance zint-β₂-AR signalling are not due to changes in the sensitivity of the I_Ca,L channel to AA. AA also stimulates I_Ca,L in guinea pig ventricular myocytes (Huang et al. 1992) and Ca²⁺ transients in embryonic chick ventricular myocytes (Pavoine et al. 1999), but AA inhibits I_Ca,L and intracellular Ca²⁺ release in rat ventricular myocytes (Xiao et al. 1997). Finally, chelation of intracellular [Ca²⁺] by BAPTA had no significant effects on zint-β₂-AR signalling in −LMN myocytes but in +LMN myocytes BAPTA significantly inhibited zint-β₂-AR stimulation, consistent with the Ca²⁺ dependence of cPLA₂ (Leslie, 2004). The fact that BAPTA abolished electrically stimulated elevation of intracellular [Ca²⁺], i.e. Ca²⁺ transients (authors’ unpublished observations), but had no affect on basal Ca²⁺ influx via I_Ca,L (Fig. 2), suggests that cPLA₂ activation in atrial myocytes is primarily dependent on intracellular Ca²⁺ release rather than Ca²⁺ influx.

The present experiments also indicate that cell attachment to LMN enhances β₂-AR signalling via a mechanism activated by inhibition of cAMP/PKA signalling. Indeed, inhibition of cAMP/PKA by two different pharmacological agents enhanced zint-β₂-AR stimulation of I_Ca,L via cPLA₂ in −LMN myocytes, much the same as cell attachment to LMN. The finding that fen-β₂-AR stimulation of I_Ca,L was significantly inhibited (rather than enhanced) by inhibition of cAMP/PKA (H-89) also confirms that inhibition of cAMP/PKA enhances a G_i-coupled β₂-AR signalling mechanism. The fact that pharmacological inhibition of cAMP/PKA activates β₂-AR/cPLA₂ signalling is consistent with our previous findings that LMN acts via phosphatidylinositol 3′-kinase (PI-(3)K) signalling to inhibit AC/cAMP signalling (Wang et al. 2009) and elicits a concomitant increase in β₂-AR signalling (Wang et al. 2000b). In embryonic chick ventricular myocytes (Pavoine et al. 1999) and rat ventricular myocytes (Ait-Mamar et al. 2005), β₂-AR stimulation activates cPLA₂ signalling. Moreover, β₁-AR/cAMP signalling exerts a negative constraint on β₂-AR/cPLA₂ signalling leading to the idea that activation of β₂-AR/cPLA₂ may potentially compensate for depressed cAMP signalling (Pavoine et al. 1999). This idea is entirely consistent with the present results and may explain the present finding that cPLA₂ plays no role in freshly isolated atrial myocytes, which exhibit significant endogenous cAMP synthesis (Wang et al. 2000a). Interestingly, in the studies of Pavoine et al. (1999) and Ait-Mamar et al. (2005) myocytes were cultured on LMN, raising the possibility that LMN may have activated β₂-AR/cPLA₂ signalling via a concomitant decrease in cAMP/PKA. This suggests that experimental design may contribute importantly to the results reported in different studies of cPLA₂ signalling. The present study also indicates that inhibition of cAMP/PKA may enhance β₂-AR/cPLA₂ signalling by activation of ERK1/2 signalling. This finding is consistent with the fact that cAMP/PKA typically inhibits ERK1/2 signalling (see Gerits et al. 2008) and therefore inhibition of cAMP/PKA removes this inhibitory control, resulting in the activation of ERK1/2 signalling. Moreover, ERK1/2 contributes, along with other signalling mechanisms, to activation of cPLA₂ (see Hirabayashi et al. 2004). In embryonic chick cardiomyocytes (cultured on LMN) β₂-AR stimulation acts via ERK1/2 to stimulate cPLA₂ (Magne et al. 2001). Therefore, the schematic diagram in Fig. 8A and B summarizes our previous (Wang et al. 2000a, 2009) and present results. Zinterol stimulates β₂-ARs coupled to both G_s and G_i signalling pathways. Panel A shows that in −LMN myocytes, zint-β₂-ARs act via G_s/AC/cAMP/PKA to stimulate I_Ca,L. In addition, stimulation of cAMP/PKA inhibits ERK1/2 and thereby prevents activation of cPLA₂ signalling. Panel B shows that in +LMN myocytes, laminin binding to β₁-integrins acts via focal adhesion kinase (FAK)/PI-(3)K/protein kinase B (Akt) signalling to inhibit AC/cAMP/PKA (Wang et al. 2009). As discussed above, the inhibition of cAMP/PKA activates ERK1/2 and thereby enhances β₂-AR stimulation of I_Ca,L via G_i-coupled cPLA₂/AA signalling.

We previously reported that cytochalasin D prevents the effects of LMN to both inhibit AC/cAMP (Wang et al. 2000a) and enhance β₂-AR signalling (Wang et al. 2000b). In the present experiments cytochalasin D also prevented activation of β₂-AR/cPLA₂ signalling induced by pharmacological inhibition of cAMP/PKA in −LMN cells. The fact that cytochalasin D similarly affected β₂-AR signalling enhanced by either LMN or inhibition of cAMP/PKA further supports a causal relationship between inhibition of cAMP/PKA (either by LMN or pharmacological agents) and activation of β₂-AR/cPLA₂ signalling. In addition, these findings indicate that activation of β₂-AR/cPLA₂ by inhibition of cAMP/PKA (either by LMN or pharmacological agents) is dependent on an intact actin cytoskeleton in atrial myocytes. In endothelial cells the cAMP/PKA pathway is regulated by integrin-mediated cell adhesion, and inhibition of cAMP/PKA signalling enhances actin synthesis and stress fibre formation (Whelan & Senger, 2003; Howe, 2004). Moreover, in vascular smooth muscle cells intact actin filaments are required for cPLA₂ translocation (Fatima et al. 2005). We therefore speculate that LMN-mediated inhibition of cAMP/PKA alters the actin architecture to facilitate the translocation of cPLA₂ to intracellular membrane sites (Gijon & Leslie, 1999). Furthermore, the fact that β₂-ARs are localized within caveolae (Rybin et al. 2000) suggests that the actin cytoskeleton may facilitate cPLA₂ translocation to caveolae membranes. In fact, in rat ventricular myocytes (cultured on LMN) zinterol induces translocation of cPLA₂ to caveolin-3-enriched membrane fractions associated with β₂-AR signalling components (Ait-Mamar et al. 2005). We therefore conclude that β₂-AR/cPLA₂ signalling is dependent on the actin cytoskeleton, which may facilitate the translocation of cPLA₂ to β₂-AR-containing caveolae. We have previously reported electron micrographs showing prominent subsarcolemmal caveolae in cat atrial myocytes (Wang et al. 2005).

In the failing heart, the ratio of β₁-/β₂-AR signalling decreases primarily as a result of a decrease in β₁-AR signalling (Bristow et al. 1986). The heart therefore becomes more dependent on β₂-AR signalling. The present results suggest that increases in extracellular matrix protein, as in fibrosis and/or hypertrophy, may enhance β₂-AR signalling such that β₂-AR signalling relies less on G_s/AC/cAMP/PKA and more on G_i/cPLA₂/AA signalling. In this regard, β₂-AR stimulation activates cPLA₂ signalling in human diseased left ventricular biopsies and right atrial appendages, and this signalling mechanism is favoured in those tissues with altered AC activity (Pavoine et al. 2003). Based on the present findings, fibrosis of these diseased tissues may act via integrin-mediated signalling to inhibit cAMP/PKA and enhance β₂-AR/cPLA₂ signalling. It remains to be determined whether LMN-mediated changes in β₂-AR signalling are beneficial or deleterious. β₁-AR and β₂-AR signalling exert opposing effects on cardiac cell survival where chronic stimulation of β₁-ARs promote and β₂-ARs inhibit cell apoptosis (Communal et al. 1999; Chesley et al. 2000). The anti-apoptotic effects of β₂-AR stimulation are mediated via G_i/PI−(3)K/Akt signalling (Kennedy et al. 1997). In the present study, β₂-AR/G_i activates cPLA₂/AA signalling, which is involved in numerous physiological and pathophysiological mechanisms (Hirabayashi et al. 2004). In terms of pathology, cPLA₂/AA signalling is associated with membrane degradation and inflammatory responses, and metabolites of AA can cause vasoconstriction, and hence compromise tissue blood flow (Bonventre, 1999). We therefore propose that in the setting of atrial fibrosis, chronic β₂-AR/cPLA₂ signalling may lead to atrial cell damage and inflammation. In fact, AA is reported to slow atrial conduction and has been implicated in the development of postoperative atrial fibrillation (Tselentakis et al. 2006).

Author contributions

All experiments were performed in the Department of Physiology, Loyola University Medical Center, Maywood, IL 60153. Conception and design of Experiments: M.R.P., A.M.S. & S.L.L. Collection, analysis and interpretation of data: M.R.P., X.J., A.M.S. & S.L.L. Drafting the article or revising it critically for important intellectual content: M.R.P., A.M.S. & S.L.L. All authors approved the final version to be published.