Inhibition of cAMP-Dependent PKA Activates β₂-Adrenergic Receptor Stimulation of Cytosolic Phospholipase A₂ via Raf-1/MEK/ERK and IP₃-Dependent Ca²⁺ Signaling in Atrial Myocytes

Abstract

We previously reported in atrial myocytes that inhibition of cAMP-dependent protein kinase (PKA) by laminin (LMN)-integrin signaling activates β₂-adrenergic receptor (β₂-AR) stimulation of cytosolic phospholipase A₂ (cPLA₂). The present study sought to determine the signaling mechanisms by which inhibition of PKA activates β₂-AR stimulation of cPLA₂. We therefore determined the effects of zinterol (0.1 μM; zint-β₂-AR) to stimulate I_Ca,L in atrial myocytes in the absence (+PKA) and presence (-PKA) of the PKA inhibitor (1 μM) KT5720 and compared these results with atrial myocytes attached to laminin (+LMN). Inhibition of Raf-1 (10 μM GW5074), phospholipase C (PLC; 0.5 μM edelfosine), PKC (4 μM chelerythrine) or IP₃ receptor (IP₃R) signaling (2 μM 2-APB) significantly inhibited zint-β₂-AR stimulation of I_Ca,L in–PKA but not +PKA myocytes. Western blots showed that zint-β₂-AR stimulation increased ERK1/2 phosphorylation in–PKA compared to +PKA myocytes. Adenoviral (Adv) expression of dominant negative (dn) -PKCα, dn-Raf-1 or an IP₃ affinity trap, each inhibited zint-β₂-AR stimulation of I_Ca,L in + LMN myocytes compared to control +LMN myocytes infected with Adv-βgal. In +LMN myocytes, zint-β₂-AR stimulation of I_Ca,L was enhanced by adenoviral overexpression of wild-type cPLA₂ and inhibited by double dn-cPLA₂^S505A/S515A mutant compared to control +LMN myocytes infected with Adv-βgal. In–PKA myocytes depletion of intracellular Ca²⁺ stores by 5 μM thapsigargin failed to inhibit zint-β₂-AR stimulation of I_Ca,L via cPLA₂. However, disruption of caveolae formation by 10 mM methyl-β-cyclodextrin inhibited zint-β₂-AR stimulation of I_Ca,L in–PKA myocytes significantly more than in +PKA myocytes. We conclude that inhibition of PKA removes inhibition of Raf-1 and thereby allows β₂-AR stimulation to act via PKCα/Raf-1/MEK/ERK1/2 and IP₃-mediated Ca²⁺ signaling to stimulate cPLA₂ signaling within caveolae. These findings may be relevant to the remodeling of β-AR signaling in failing and/or aging heart, both of which exhibit decreases in adenylate cyclase activity.

Introduction

We previously reported that attachment of atrial myocytes to the extracellular matrix protein laminin (LMN) acts via β₁ integrin receptors to decrease β₁-adrenergic receptor (AR) and increase β₂-AR stimulation of L-type Ca²⁺ current (I_Ca,L) [1]. Cell attachment to LMN decreases β₁-AR signaling by inhibiting adenylate cyclase activity and diminishing cAMP levels via integrin-dependent activation of focal adhesion kinase (FAK)/phosphatidyinositol-3’ kinase (PI-3K)/protein kinase B (Akt) signaling [2]. We also reported that atrial cell attachment to LMN enhances β₂-AR signaling by activating Gi/ERK/cytosolic phospholipase A₂ (cPLA₂)/arachidonic acid (AA) stimulation of I_Ca,L [3]. β₂-AR activation of cPLA₂ signaling is dependent on concomitant LMN-mediated inhibition of adenylate cyclase/cAMP-dependent kinase (PKA) [3]. In other words, cell attachment to LMN acts via inhibition of adenylate cyclase/PKA to both inhibit β₁-AR signaling and enhance β₂-AR signaling through activation of cPLA₂. In embryonic chick ventricular myocytes [4] and rat ventricular myocytes [5] β₂-AR stimulation also activates cPLA₂/AA signaling. Moreover, these authors proposed that activation of β₂-AR/cPLA₂ signaling may compensate for depressed cAMP signaling [4]. Interestingly, in both of these studies by Pavoine et al. (1999) and Ait-Mamar et al., (2005) cardiomyocytes were cultured on LMN, supporting our findings that cell attachment to LMN may be responsible for inhibition of PKA and activation of β₂-AR/cPLA₂ signaling. However, the mechanism by which PKA inhibition activates β₂-AR/cPLA₂ signaling is not clear.

Our initial experiments indicated that in atrial myocytes β₂-AR activation of cPLA₂ is Ca²⁺-dependent and mediated via ERK1/2 signaling [3]. This is consistent with studies in embryonic chick ventricular myocytes (cultured on LMN) in which β₂-AR stimulation acts via ERK1/2 signaling to activate cPLA₂ [6]. Moreover, in a variety of cell systems Raf-1 activates downstream ERK1/2 and PKA inhibits Raf-1 [7, 8]. Therefore, inhibition of PKA should remove inhibition of Raf-1, thereby allowing β₂-AR stimulation to act via Raf-1/MEK/ERK1/2 signaling. Moreover, protein kinase C (PKC) activates Raf-1 [9, 10]. In other words, PKA inhibits and PKC activates Raf-1/MEK/ERK1/2 signaling. Based on these considerations we sought to determine whether inhibition of PKA facilitates β₂-AR stimulation to act via PKC/Raf-1/MEK/ERK1/2 to activate cPLA₂. These findings may be relevant to the remodeling of β₂-AR signaling in the failing and/or aging heart, both of which exhibit decreases in adenylate cyclase activity.

Materials and Methods

Ethics Statement

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. IACUC prescribed the rules for the animal care and supervised their enforcement. Animals were obtained from a licensed vendor (R & R Research, Howard City, MI., USA), and housed and fed in our AAALAC approved Comparative Medicine Department. Adult cats of either sex (n = 32 cats) were anesthetized with sodium pentobarbital (50 mg/kg, IP).

Isolation of atrial myocytes

Once fully anesthetized, 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 [11].

Perforated patch clamp experiments

Electrophysiological recordings from atrial myocytes were performed in the perforated (nystatin) patch whole-cell configuration at room temperature, as previously described [11]. 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 [12]. Agonist was applied for approximately 4 min and the effects on peak I_Ca,L amplitude were recorded at the steady-state response.

Plating of atrial myocytes on LMN coated glass coverslips

Generally, we compared freshly isolated atrial myocytes obtained from the same hearts, as previously described [3]; atrial myocytes on uncoated glass cover-slips in the absence of PKA inhibitor (+PKA) and atrial myocytes on uncoated glass coverslips exposed to the specific PKA inhibitor (1 μM) KT5720 (-PKA). Because pharmacological inhibition of PKA elicits signaling mechanisms that are similar to those elicited by LMN-integrin signaling, we performed some experiments on atrial myocytes attached to glass cover-slips coated with laminin (+LMN; 40 ug/ml) for at least 2 hrs, as previously described [11]. Inhibition of PKA by KT5720 typically decreases basal I_Ca,L amplitude by 15–20% [3], consistent with the relatively high endogenous PKA activity in cat atrial myocytes [11]. In addition, a variety of experimental results indicate that atrial cell attachment to LMN is not restoring LMN-mediated signaling somehow lost during the cell isolation procedure. For example, control experiments have shown that atrial myocytes plated on poly-L-lysine, a non-specific substrate for cell attachment, fail to exhibit changes in β-AR signaling similar to cells attached to LMN [1]. Moreover, freshly isolated cardiomyocytes not plated on LMN exhibit responses to β-AR stimulation which are similar to multicellular cardiac preparations, i.e. exhibit predominantly β₁-AR over β₂-AR signaling [1]. However, cell attachment to LMN decreases the β₁-/β₂-AR signaling ratio resulting in predominantly β₂-AR over β₁-AR signaling [1]. Moreover, pharmacological inhibition of PKA in myocytes not attached to LMN mimics the effects of cell attachment to LMN [3].

Adenoviral infection of atrial myocytes

In some experiments, atrial myocytes were attached to laminin (2h) and then infected (100 moi, 24h) with replication-defective adenovirus (Adv) prior to electrophysiological recording. PKCα was inhibited by infection with an Adv expressing kinase-inactive mouse PKCα [13], kindly provided by Dr. Trevor Biden, Garvan Institute of Medical Research, St. Vincent’s Hospital, Sydney Australia. An Adv expressing a dominant-negative (dn) mutant of rabbit PKCε [14] was kindly provided by Dr. Peipei Ping, University of California-Los Angeles. Adv expressing wild type, and a non-phosphorylatable mutant of human cPLA₂ (S515A/S505A double mutant) [15] were generously provided by Dr. K.U. Malik, University of Tennessee Health Science Center, Memphis, TN. An Adv expressing a dn mutant of human Raf-1 [9] was kindly provided by Dr. Dan Kuppuswamy, Medical University of South Carolina, Charleston, SC. A control Adv expressing nuclear-encoded β-galactosidase (Adv-βgal) was used to control for nonspecific effects of adenoviral infection [16]. Adenoviruses were amplified and purified using HEK293 cells, and the multiplicity of infection (moi) for each virus was determined by dilution assay in HEK293 cells grown in 96 well clusters, as previously described [16]. Preliminary experiments using 5-bromo-4-chloro-3-indolyl-beta- D-galactopyranoside (X-gal) staining of Adv-βgal infected cells determined that a concentration of 100 moi infected 93±3% (n = 3 expts, 400–700 cells/expt) of cultured myocytes. The IP₃ affinity trap consists of the ligand binding domain of the rat type 1 IP₃R. The construction of this vector and subsequent production of adenovirus was previously described in detail [17]. Freshly isolated atrial myocytes were infected with Adv-IP₃ affinity trap or the Adv-βgal (control) for 1 h, followed by 18 h short-term culture at 37°C.

Western blot

Atrial myocytes were plated on poly-L-lysine, a biologically inactive substrate and therefore represent +PKA myocytes. Atrial myocytes were either untreated or treated with Zinterol, KT5720 alone (PKA inhibitor) and KT combined with Zinterol. A positive control showing ERK1/2 and p38MAPK phosphorylation in A7r5 cells is used after stimulation with 1 μM angiotensin II (5 min). Briefly atrial homogenates were centrifuged at 10,000 g for 2 min and the supernatant was collected. The protein concentrations of the samples were determined using the Bradford protein assay (Bio-Rad). Aliquots of the samples (40 μg) were dissolved in a Laemmli Sample buffer containing: 60 mM Tris-HCl, 2% SDS, 20% Glycerol, 5% β-mercaptoethanol, 0.01% bromophenol blue (pH 6.8) and the proteins were separated on a 4–15% Mini-PROTEAN TGX Gel (Bio-Rad). After transfer, the membrane was blocked with TBS-T Buffer containing 20 mM Tris pH 7.5, 150 mM NaCl, 0.05% Tween, and a 5% blocking powder (Bio Rad) at 4°C for 1 hr. The blot was probed with primary and secondary antibodies for phorpho-ERK1/2 and phosphor-p38^MAPK that were purchased from cell signaling technologies (Danvers, MA). The blot was subsequently probed with GAPDH primary antibody for 2 hrs, followed by HRP-conjugated goat anti-mouse secondary antibody (1:10,000, Santa Cruz Biotechnology) at 4°C for 1 hr. Specific binding was visualized by chemiluminescence (Immun-Star Western C Kit, Bio-Rad) using a ChemiDoc XRS imager (Bio-Rad Life Science Research, Hercules, CA). The intensities of the bands corresponding to each protein were quantified using ImageLab software (Bio-Rad). The relative intensity for each band was normalized to the intensity of the GAPDH staining.

Chemicals

Zinterol, AACOCF₃, GW5074, edelfosine (ET-18-OCH3), chelerythrine, 2-aminoethyl diphenyl borate (2-APB), KT5720, methyl-β-cyclodextrin, ryanodine, thapsigargin (Sigma Chemical).

Statistics

Data are mean ± standard error (SE) of the mean. Measurements were analyzed using either paired or unpaired Student’s 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

Role of Raf-1 in +PKA and –PKA atrial myocytes

As shown in Fig 1we determined the effects of 10 μM GW5074, a potent Raf-1 inhibitor [8] on zint-β₂-AR stimulation of I_Ca,L in +PKA and–PKA atrial myocytes. In +PKA myocytes (A) 0.1 μM zint-β₂-AR stimulation elicited a typical increase in I_Ca,L (124±12%, N = 3). In another group of +PKA myocytes from the same hearts (A), prior exposure to GW5074 for 30 min had no significant effects on basal I_Ca,L amplitude (open bar) or zint-β₂-AR stimulation of I_Ca,L (126±11%, N = 3). In contrast, as shown in panel B, control zint-β₂-AR stimulation of I_Ca,L in–PKA myocytes (B; 207±23%, N = 3) was enhanced compared to +PKA myocytes (A; 124±12%, N = 3) due to activation of cPLA₂ signaling, as previously reported [4]. In–PKA myocytes, GW5074 had no effects on basal I_Ca,L amplitudes (open bars) but in contrast to +PKA myocytes, GW5074 significantly inhibited zint-β₂-AR stimulation of I_Ca,L (B; 40±5%, N = 3). Additional experiments showed that zint-β₂-AR stimulation of I_Ca,L in +LMN myocytes produced results similar to those found in–PKA myocytes (control; 192±13%, N = 3) and (GW5074; 33±9%, N = 3) (see S1 Fig for data).

Fig 1.

Effects of Raf-1 inhibition (10 μM GW5074) on zint-β₂-AR stimulation of I_Ca,L in +PKA (A), -PKA (B) atrial myocytes. A; In +PKA myocytes, GW5074 (30 min) had no significant effect on zint-β₂-AR stimulation of I_Ca,L. B; In -PKA myocytes, zint-β₂-AR stimulation of I_Ca,L was enhanced compared to control responses (A) and GW5074 significantly inhibited zint-β₂-AR stimulation of I_Ca,L. Numbers in parentheses indicate the number of myocytes studied. * = P<0.05.

Comparison of zint-β₂-AR induced I_Ca,L currents in the Adv-βgal and dn-Raf-1 mutant infected atrial myocytes

To further establish the role of Raf-1, we infected atrial myocytes with an Adv that expresses a dn-Raf-1 mutant (generously provided by Dr. Kuppuswamy [9, 18]. Control cells were infected with Adv-βgal. Infected cells were cultured on LMN overnight and therefore represent +LMN myocytes. As shown in Fig 2A, in control +LMN myocytes expressing βgal, zint-β₂-AR stimulation elicited a typically enhanced increase in I_Ca,L (202±19%). In +LMN myocytes expressing the dn-Raf-1 mutant, zint-β₂-AR stimulation of I_Ca,L was significantly inhibited (105±14%; P<0.02) compared to control. Together, these findings indicate that Raf-1 signaling plays no role in β₂-AR stimulation of I_Ca,L in freshly isolated atrial myocytes not attached to LMN (Fig 1A). However, when PKA is inhibited by either PKA inhibitor or cell attachment to LMN, β₂-AR stimulation acts via Raf-1 to activate cPLA₂ and stimulate I_Ca,L.

Fig 2. Effects of dn-Raf-1-Adv on zint-β₂-AR stimulation of I_Ca,L in +LMN myocytes.

In control +LMN myocytes (βgal) zint-β₂-AR stimulation elicited a typically enhanced increase in I_Ca,L. In +LMN myocytes infected with dn-Raf-1-Adv zint-β₂-AR stimulation of I_Ca,L was significantly inhibited compared with controls. Numbers in parentheses indicate that number of cells studied. * = P<0.05.

Inhibition of PKA, increases ERK1/2 phosphorylation in atrial myocytes

It is well established that Raf-1 activates downstream MEK/ERK1/2 signaling [7, 8]. Moreover, our previous results indicated that inhibition of MEK/ERK1/2 signaling by U0126 inhibited zint-β₂-AR stimulation cPLA₂ signaling [3]. Therefore, inhibition of PKA should activate β₂-AR stimulation of ERK1/2 phosphorylation. As shown in Fig 3, we performed Western blots and probed for zint-β₂-AR-mediated ERK1/2 and p38^MAPK phosphorylation in the absence (+PKA) and presence of PKA inhibitor (–PKA). Atrial myocytes were plated on poly-L-lysine, a biologically inactive substrate and therefore represent +PKA myocytes. As shown in Fig 3in control +PKA myocytes zint-β₂-AR stimulation modestly increased ERK1/2 phosphorylation (solid bars). Exposure of +PKA myocytes to the PKA inhibitor (1 μM) KT5720, (–PKA myocytes) further increased ERK/1/2 phosphorylation, suggesting that inhibition of basal PKA activity removes PKA-induced inhibition of basal ERK1/2 phosphorylation. In–PKA myocytes, zint-β₂-AR stimulation prominently stimulated ERK1/2 phosphorylation. Moreover, zint-β₂-AR stimulation only modestly stimulated p38^MAPK phosphorylation (open bars), and was not significantly affected by KT5720. The last lane on the Western blot is a positive control showing ERK1/2 and p38^MAPK phosphorylation in A7r5 cells stimulated by 1μM angiotensin II (5 min). In conjunction with our previous findings [3], the present results indicate that inhibition of PKA preferentially activates β₂-AR stimulation of ERK1/2 signaling.

Fig 3. Effect of PKA inhibition on β₂-AR-medited phosphorylation of ERK1/2.

Western blots of atrial cell homogenates show that in control -PKA myocytes zinterol modestly increased ERK1/2 phosphorylation (solid bars). Inhibition of PKA by 1 μM KT5720 (KT) increased basal ERK/1/2 phosphorylation. In the presence of KT5720, zinterol prominently stimulates ERK1/2 phosphorylation. Zinterol only modestly stimulated p38MAPK phosphorylation (open bars), and without significant effect by KT5720. The last lane on the Western blot is a positive control showing ERK1/2 and p38MAPK phosphorylation in A7r5 cells stimulated by 1 μM angiotensin II (5 min). Similar results were obtained in a total of 3 experiments. * = P<0.05.

Role of PLC in the enhanced zint-β₂-AR induced I_Ca,L currents in –PKA atrial myocytes

It is well established that Raf-1 activates downstream MEK/ERK1/2 signaling [7, 8]. Moreover, our previous results indicated that inhibition of MEK/ERK1/2 signaling by U0126 inhibited zint-β₂-AR stimulation of cPLA₂ signaling [3]. This raises the question of how β₂-AR stimulation activates Raf-1. In a variety of cell systems [10], including cardiac muscle [9], protein kinase C (PKC) activates Raf-1/MEK/ERK1/2. Receptor-mediated activation of phospholipase C (PLC) and the hydrolysis of phospholipids results in the production of diacylglycerol and stimulation of PKC isoenzymes. Therefore, as shown in Fig 4we determined the effects of 0.5 μM edelfosine, a specific inhibitor of phospholipase C [19], on zint-β₂-AR stimulation of I_Ca,L in +PKA and–PKA myocytes. In control cells (A), edelfosine had little effect on basal I_Ca,L amplitude (open bars) or zint-β₂-AR stimulation of I_Ca,L (zint, 125±3% vs edelfosine, 114±7%). In–PKA myocytes (B) zint-β₂-AR stimulation elicited a typically enhanced increase in I_Ca,L (215±5%) compared to control responses (A) due to activation of cPLA₂ [3]. In contrast to +PKA myocytes, edelfosine now significantly inhibited zint-β₂-AR stimulation of I_Ca,L (78±11%; P<0.0005). Additional experiments in +LMN myocytes showed results similar to those found in–PKA myocytes, i.e. zint, 204±2% vs edelfosine, 84±8%; P<0.03 (see S2 Fig for data). In other words, inhibition of PKA allows β₂-AR stimulation to act via PLC to stimulate cPLA₂ signaling.

Fig 4.

Effects of 0.5 μM edelfosine on zint-β₂-AR stimulation of I_Ca,L in +PKA (A), -PKA(B) and atrial myocytes. A; in control +PKA myocytes edelfosine had no significant effects on zint-β₂-AR stimulation of I_Ca,L. B; in -PKA myocytes zint-β₂-AR stimulation elicited a typically enhanced increase in I_Ca,L compared to controls (A) and edelfosine significantly inhibited zint-β₂-AR stimulation of I_Ca,L. Numbers in parentheses indicate the number of myocytes studied. * = P<0.05.

Role of PKC in the enhanced zint-β₂-AR induced I_Ca,L currents in –PKA atrial myocytes

Because activation of PLC produces diacylglycerol and activation of PKC we next determined the effects of 4 μM chelerythrine, an inhibitor of Ca²⁺-dependent and Ca²⁺ -independent PKCs [20] on zint-β₂-AR stimulation of I_Ca,L in +PKA and–PKA atrial myocytes. As shown in Fig 5, in +PKA myocytes (panel A), chelerythrine had no significant effects on basal I_Ca,L amplitude (open bars) or zint-β₂-AR stimulation of I_Ca,L (control,106±8% vs cheler, 116±3%). In–PKA myocytes (panel B) zint-β₂-AR stimulation elicited a typically enhanced increase in I_Ca,L (278±13%) compared to control (panel A) due to activation of cPLA₂ [3]. In contrast to +PKA myocytes, chelerythrine now significantly inhibited zint-β₂-AR stimulation of I_Ca,L (71±13%; P<0.002). These findings indicate PKC signaling plays no role in β₂-AR stimulation of I_Ca,L in freshly isolated atrial myocytes. However when PKA is inhibited β₂-AR stimulation acts via PKC to stimulate cPLA₂, consistent with β₂-AR stimulation of PLC.

Fig 5.

Effects of 4 μM chelerythrine (cheler) on zint-β₂-AR stimulation of I_Ca,L in +PKA(A) and -PKA (B) myocytes and effects of dn-PKCα-Adv (C) on zint-β₂-AR stimulation of I_Ca,L in +LMN myocytes. A; in control +PKA myocytes, chelerythrine had no significant effects on zint-β₂-AR stimulation of I_Ca,L. B; in—PKA myocytes, zint-β₂-AR stimulation elicited a typically enhanced in increase in I_Ca,L compared to controls (A) and chelerythrine significantly inhibited zint-β₂-AR stimulation of I_Ca,L. C; in control +LMN myocytes (βgal) zint-β₂-AR stimulation of I_Ca,L was typically enhanced compared to controls (A). In +LMN myocytes infected with dn-PKCα-Adv, zint-β₂-AR stimulation of I_Ca,L was significantly inhibited. Numbers in parentheses indicate that number of cells studied. * = P<0.05.

In another approach we infected atrial myocyte with an Adv expressing a dominant-negative mutant of Ca²⁺-dependent PKCα (dn-PKCα). Control cells were infected with Adv-βgal. Infected cells were cultured overnight on LMN and therefore represent +LMN myocytes. As shown in Fig 5C, in control +LMN myocytes (Adv-βgal) zint-β₂-AR stimulation elicited a typically enhanced increase in I_Ca,L (208±8%) compared to +PKA myocytes (panel A). In +LMN myocytes expressing the dn-PKCα mutant, zint-β₂-AR stimulation of I_Ca,L was significantly inhibited (control, 208±8% vs dn-PKCα, 57±5%; P<0.002). Similar experiments performed on +LMN myocytes expressing a dn-PKCε mutant showed no differences with control +LMN (β-gal) myocytes (control, 158±21% vs dn-PKCε, 165±31%, N = 4). These findings provide additional support for the idea that β₂-AR stimulation acts via Ca²⁺-dependent PKCα to activate cPLA₂.

Role of cPLA2 in the enhanced zint-β₂-AR induced I_Ca,L currents in LMN plated atrial myocytes

Activation of cPLA₂ requires phosphorylation at serine sites S505 and S515 [21–23]. We therefore determined the effect of β₂-AR stimulation in +LMN myocytes infected with Adv that either expressed a non-phosphorylatable dominant-negative cPLA₂ (dn-cPLA₂^S505A/S515A) mutant, wild-type cPLA₂ (wt-cPLA₂) or βgal as control. Infected atrial myocytes were cultured on LMN overnight and therefore represent +LMN myocytes. Fig 6shows that in control +LMN myocytes expressing βgal, β₂-AR stimulation elicited a typical increase in I_Ca,L (166±11%). In +LMN myocytes expressing the double dn-cPLA₂^S505A/S515A mutant β₂-AR stimulation of I_Ca,L was significantly attenuated (84±8%) compared to control. In +LMN myocytes overexpressing wt-cPLA₂, β₂-AR stimulation of I_Ca,L was significantly enhanced (307±8%) compared to either control or dn-cPLA₂^S505A/S515A. These results confirm that cell attachment to LMN activates β₂-AR/cPLA₂ signaling and further indicates that β₂-AR stimulation of cPLA₂ requires serine phosphorylation at one or both sites.

Fig 6. Effects of dn-cPLA₂^S515A/S505A and wt-cPLA₂ on zint-β₂-AR stimulation of I_Ca,L in +LMN myocytes.

Compared to control +LMN myocytes (βgal), zint-β₂-AR stimulation of I_Ca,L was significantly inhibited and enhanced in myocytes overexpressing dn-cPLA₂^S515A/S505A and expressing wt-cPLA₂, respectively. Numbers in parentheses indicate that number of cells studied. * = P<0.05.

Role of IP3 receptor in the enhanced zint-β₂-AR induced I_Ca,L currents in -PKA atrial myocytes

Our previous work showed that β₂-AR stimulation of cPLA₂ signaling is dependent on intracellular Ca²⁺ [3]. The fact that β₂-AR stimulation acts via PLC to activate cPLA₂ (Fig 3) suggests the potential involvement of IP₃-mediated Ca²⁺ signaling. We therefore determined the effects of 2 μM 2-APB, a putative IP₃ receptor (IP₃R) blocking agent [24], on β₂-AR stimulation of I_Ca,L in +PKA and–PKA myocytes. As shown in Fig 7A, in +PKA myocytes 2-APB had no significant effect on basal I_Ca,L (open bars) or zint-β₂-AR stimulation of I_Ca,L (control, 121±4% vs 2-APB, 113±5%). In–PKA myocytes (B) zint-β₂-AR stimulation elicited a typically enhanced increase in I_Ca,L compared to control responses (panel A). In contrast to +PKA myocytes, 2-APB significantly inhibited zint-β₂-AR stimulation of I_Ca,L (control, 245±25% vs 2-APB, 89±19%; P<0.02). Similar results were obtained in +LMN myocytes where zint-β₂-AR stimulation of I_Ca,L also was significantly blocked by 2-APB (control, 186±26%; N = 4 vs 87±7%; N = 4, P<0.01) (see S3 Fig online data).

Fig 7.

Effects of 2 μM 2-APB on zint-β₂-AR stimulation of I_Ca,L in +PKA (A), -PKA atrial myocytes. A; in control +PKA myocytes, 2-APB had no significant effects on zint-β₂-AR stimulation of I_Ca,L. B; in -PKA myocytes, zint-β₂-AR stimulation elicited a typically enhanced increase in I_Ca,L compared to controls (A) and 2-APB significantly inhibited zint-β₂-AR stimulation of I_Ca,L. Numbers in parentheses indicate the number of myocytes studied. * = P<0.05.

In another approach, we infected atrial myocytes with an adenovirus that expresses an IP₃ affinity trap which binds to IP₃ in the cytosol and thereby inhibits IP₃-mediated Ca²⁺ signaling by preventing IP₃ from reaching and activating the IP₃ receptor (IP₃R) [17]. Cells were cultured overnight on LMN and therefore represent +LMN myocytes. As shown in Fig 8, in control +LMN myocytes expressing βgal, zint-β₂-AR stimulation elicited a typical increase in I_Ca,L (182±4%). However, in +LMN myocytes expressing the IP₃ trap zint-β₂-AR stimulation of I_Ca,L (111±17%) was significantly inhibited compared to controls (P<0.002). Together, these findings suggest that IP₃R signaling plays no role in zint-β₂-AR stimulation of I_Ca,L in freshly isolated atrial myocytes. However, when PKA is inhibited by either PKA inhibitor (Fig 7) or cell attachment to LMN (Fig 8), β₂-AR stimulation of cPLA₂ is dependent on IP₃-mediated Ca²⁺ signaling.

Fig 8. Effect of adenovirus IP₃ affinity trap on zint-β₂-AR stimulation of I_Ca,L in +LMN myocytes.

Compared with control +LMN myocytes (βgal), zint-β₂-AR stimulation of I_Ca,L was significantly inhibited in +LMN myocytes infected with adenovirus expressing the IP₃ affinity trap. Numbers in parentheses indicate the number of myocytes studied. * = P<0.05.

Role of IP3 receptors located on SR or nuclear envelope in the enhanced zint-β₂-AR induced I_Ca,L currents in -PKA atrial myocytes

IP₃Rs are thought to be located primarily on the sarcoplasmic reticulum (SR) and nuclear envelope membranes. Because the SR and nuclear envelope membranes are highly interconnected, inhibition of SR Ca²⁺ uptake by thapsigargin depletes intracellular Ca²⁺ stores from both sites [25]. Therefore, in Fig 9we determined whether depletion of SR and nuclear Ca²⁺ by 5 μM thapsigargin (10 min; Thaps) inhibits β₂-AR stimulation of I_Ca,L in +PKA (A) and–PKA (B) myocytes. In +PKA myocytes (A) thapsigargin had no effect on basal I_Ca,L amplitude (open bars). Compared to control responses (107±8%), thapsigargin slightly enhanced zint-β₂-AR stimulation of I_Ca,L (143±4%), although the change was not statistically significant. This modest increase in β₂-AR stimulation of I_Ca,L is consistent with the effects of thapsigargin to inhibit SR Ca²⁺ release and thereby inhibit Ca²⁺-mediated inactivation of I_Ca,L. In fact, separate experiments showed that thapsigargin abolished SR Ca²⁺ transients (data not shown). In–PKA myocytes (B), zint-β₂-AR stimulation of I_Ca,L (207±17%) was typically enhanced compared to control responses (107±8%) obtained in +PKA myocytes (A). Interestingly, in–PKA myocytes treated with thapsigargin β₂-AR stimulation of I_Ca,L (238±12%) was still enhanced. In other words, depletion of Ca²⁺ from SR and nuclear membranes failed to prevent PKA inhibition from enhancing β₂-AR signaling. Moreover, the last column in Fig 9B shows that in–PKA myocytes treated with thapsigargin, AACOCF₃ (cPLA2 inhibitor) significantly inhibited zint-β₂-AR stimulation of I_Ca,L (62±12%), indicating that the enhanced response to zint-β₂-AR stimulation was in fact due to activation of cPLA₂ signaling and that depletion of SR and nuclear Ca²⁺ stores failed to prevent β₂-AR/cPLA₂ signaling. Similar results were obtained when Ca²⁺ stores were depleted by treatment (10 min) with 10 μM ryanodine (data not shown). These results indicate that IP₃-dependent Ca²⁺ signaling is not mediated via IP₃Rs located on SR or nuclear envelope membranes.

Fig 9.

Effects of 5 μM thapsigargin (10 min; Thaps) on zint-β₂-AR stimulation of I_Ca,L in +PKA(A) and -PKA (B) myocytes. A; in +PKA myocytes, compared to control responses, thapsigargin slightly enhanced zint-β₂-AR stimulation of I_Ca,L. B; in -PKA myocytes, zint-β₂-AR stimulation elicited a typically enhanced increase in I_Ca,L compared to control (A) that was not prevented by treatment with thapsigargin. The addition of 10 μM AACOCF₃ (+ thaps) significantly inhibited zint-β₂-AR stimulation of I_Ca,L indicating that thapsigargin did not prevent zint-β₂-AR stimulation of cPLA₂. Numbers in parentheses indicate the number of myocytes studied. * = P<0.05.

Role of IP3 receptors located in caveolae in the enhanced zint-β₂-AR induced I_Ca,L currents in -PKA atrial myocytes

Alternatively, IP₃R protein is present in caveolae [26, 27], which are abundant in atrial myocytes [28]. We therefore determined the effects of 10 mM methyl-β-cyclodextrin (MCD; 30 min), an agent that disrupts caveolae formation [29], on zint-β₂-AR stimulation of I_Ca,L in +PKA and–PKA myocytes. As shown in Fig 10A, in +PKA myocytes MCD caused a modest but significant inhibition of zint-β₂-AR stimulation of I_Ca,L (control,122±5% vs MCD, 93±7%; P<0.05), representing a 24% decrease. This is consistent with the idea that β₂-ARs are normally localized to caveolae [30]. However, in–PKA myocytes (B) MCD elicited a significantly larger inhibition of zint-β₂-AR stimulation of I_Ca,L (control, 235±7% vs MCD, 54±12%; P<0.008), representing a 77% decrease. The fact that MCD elicited a significantly larger inhibition of β₂-AR signaling in–PKA compared to +PKA myocytes supports the idea that the IP₃Rs that are essential for β₂-AR/cPLA₂ signaling are localized to the caveolae.

Fig 10.

Effect of 2 mM methyl-β-cyclodextrin (MCD; 30 min) on zint-β₂-AR stimulation of I_Ca,L in +PKA (A) and -PKA (B) myocytes. A; in +PKA myocytes, MCD significantly decreased zint-β₂-AR stimulation of I_Ca,L compared to control responses (-24%). B; in—PKA myocytes, zint-β₂-AR stimulation elicited a typically enhanced increase in I_Ca,L compared to +PKA myocytes (A) and the effect of MCD to decrease zint-β₂-AR stimulation of I_Ca,L was enhanced (-77%) compared to +PKA myocytes (A). Numbers in parentheses indicate the number of myocytes studied. * = P<0.05.

Discussion

We previously reported that in atrial myocytes inhibition of adenylate cyclase/PKA by either cell attachment to LMN or inhibition of PKA in cells not attached to LMN activates β₂-AR/cPLA₂ signaling [2, 3]. The present study extends those findings by showing that when PKA is inhibited, β₂-AR stimulation acts via PKCα/Raf-1/MEK/ERK1/2 and IP₃-dependent Ca²⁺ signaling to activate cPLA₂.

In the present experiments, we showed that inhibition of PKA prominently stimulates β₂-AR-mediated phosphorylation of ERK1/2, consistent with β₂-AR stimulation of I_Ca,L via MEK/ERK1/2 signaling [3]. Moreover, inhibition of Raf-1 by either GW5074 or adenoviral expression of a dn-Raf-1 mutant significantly inhibited β₂-AR stimulation of I_Ca,L in–PKA or +LMN myocytes but failed to affect β₂-AR signaling in +PKA myocytes. These findings can be explained by the fact that PKA inhibits Raf-1 and Raf-1 activates MEK/ERK1/2 signaling [7, 31, 32]. Because atrial myocytes normally exhibit relatively high endogenous levels of adenylate cyclase/PKA activity [11], Raf-1 is normally inhibited and therefore β₂-AR stimulation is unable to activate Raf-1/MEK/ERK1/2 signaling in +PKA myocytes. However, when PKA is inhibited (–PKA or +LMN myocytes) β₂-AR stimulation acts via Raf-1/MEK/ERK1/2 to activate cPLA₂. The present experiments also show that adenoviral overexpression of wt-cPLA₂ or expression of a dn-cPLA₂^S505A/S515A mutant in +LMN myocytes significantly enhanced or inhibited, respectively, β₂-AR stimulation of I_Ca,L compared to control +LMN myocytes expressing βgal. In various cell systems including vascular smooth muscle and fibroblasts ERK1/2 phosphorylates cPLA₂ at S505 and S515 [15, 21–23]. In cardiac ventricular myocytes (cultured on LMN) β₂-AR stimulation acts via ERK1/2 and p38^MAPK to phosphorylate cPLA₂ at S505 [6].

In atrial myocytes, β₂-ARs are coupled to both Gs- and Gi-mediated signaling pathways [12]. We [3] and others [4] have reported that β₂-ARs act via pertussis toxin-sensitive G_i to activate cPLA₂. In various cell systems, including rat ventricular myocytes [33] activation of PKA phosphorylates β₂-ARs and thereby switches β₂-AR coupling from G_s-mediated adenylate cyclase to G_i-mediated ERK1/2 signaling [34, 35]. Moreover, inhibition of PKA by H-89 inhibited β₂-AR/G_i-mediated ERK1/2 signaling. These findings are not consistent with the present results, which indicate that in adult atrial myocytes inhibition of PKA by H-89 or KT5720 activates (rather than inhibits) β₂-AR/G_i-mediated ERK1/2 signaling [3]. However, our findings are consistent with reports that PKA inhibits Raf-1 and its downstream MEK/ERK1/2 signaling pathway [7, 31, 32]. Therefore, in atrial myocytes inhibition of PKA removes Raf-1 inhibition and thereby activates β₂-AR/G_i stimulation of Raf-1/MEK/ERK1/2 signaling.

The present results also indicate that inhibition of PLC (edelfosine) or PKC (chelerythrine or dn-PKCα) significantly inhibited β₂-AR stimulation of I_Ca,L in–PKA or +LMN myocytes but not in +PKA myocytes. Moreover, in +LMN myocytes expression of dn-PKCα significantly inhibited β₂-AR signaling while expression of dn-PKCε had no effect, indicating that β₂-ARs act via Ca²⁺-dependent PKCα to activate cPLA₂. These findings are consistent with reports that PKC activates Raf-1/MEK/ERK1/2 signaling in adult cardiac myocytes [9, 18] and that PKC activates cPLA₂ signaling [10].

Our previous findings showed that strong chelation of intracellular Ca²⁺ by BAPTA prevented β₂-AR stimulation via cPLA₂. This is consistent with the fact that several of the signaling molecules involved in the proposed β₂-AR/cPLA₂ signaling cascade are Ca²⁺-dependent, including PLC, PKCα and cPLA₂. In the present study, we investigated more specifically which source of intracellular Ca²⁺ is required for β₂-AR/cPLA₂ signaling. We found that 2-APB, an agent that inhibits IP₃-mediated Ca²⁺ release in cat atrial myocytes [24] significantly inhibited β₂-AR stimulation of I_Ca,L in both–PKA and +LMN myocytes but not in +PKA myocytes. Moreover, adenoviral expression of an IP₃ affinity trap which inhibits IP₃-mediated Ca²⁺ signaling [17] resulted in a similar inhibition of β₂-AR signaling in +LMN myocytes. The fact that 2-APB had no effect on β₂-AR signaling in control +PKA myocytes and that it exerted effects similar to those of the Adv-IP₃ affinity trap suggests that 2-APB acted specifically to inhibit IP₃-mediated Ca²⁺ signaling. Together, these results indicate that β₂-AR stimulation of cPLA₂ is dependent on IP₃-mediated Ca²⁺ signaling. Although IP₃Rs are typically located on SR and nuclear membranes, the present results showed that thapsigargin, an agent that depletes both SR and nuclear envelope Ca²⁺ stores [25] failed to prevent β₂-AR stimulation of I_Ca,L via cPLA₂. However, disruption of caveolae formation by methyl-β-cyclodextrin elicited a significantly larger inhibition of β₂-AR stimulation in–PKA myocytes than control +PKA myocytes. We therefore conclude that IP₃Rs on the SR and nuclear envelope membranes are not involved in the Ca²⁺ signaling required for β₂-AR stimulation of cPLA₂. Alternatively, β₂-AR stimulation of I_Ca,L via cPLA₂ is dependent on IP₃-mediated Ca²⁺ signaling in caveolae. In endothelial and smooth muscle cells plasmalemmal caveolae contain IP₃R protein that is speculated to mediate Ca²⁺ influx through the caveolar membrane [26, 27]. Moreover, in rat ventricular myocytes (cultured on laminin) β₂-ARs stimulate translocation of cPLA₂ to the low density caveolin-3 enriched membrane fraction suggesting that cPLA₂ translocates to caveolae [5]. In fact, lipid rafts i.e. caveolae contain a wide variety of signaling components including β₂-ARs, G_i, PKC, Raf-1, ERK1/2, IP₃Rs [36], involved in β₂-AR/cPLA₂ signaling.

We therefore propose (see Fig 11) that under normal conditions (A; +PKA myocytes), β₂-AR/G_s stimulation acts via adenylate cyclase (AC)/cAMP/PKA to stimulate I_Ca,L. The relatively high basal levels of endogenous PKA activity and β₂-AR-stimulated PKA activity both inhibit Raf-1 signaling, thereby preventing β₂-AR stimulation from normally activating cPLA₂. However, inhibition of PKA by either PKA inhibitor (B;–PKA myocytes) or by cell attachment to LMN (C; +LMN myocytes), removes inhibition of Raf-1. Under these conditions, β₂-AR/G_i stimulation of PLC activates PKCα and IP₃R-mediated Ca²⁺ signaling in caveolae. With PKA inhibited, PKCα is now able to stimulate Raf-1/MEK/ERK1/2 signaling. Therefore, β₂-AR/G_i-mediated activation of both ERK1/2 and IP₃-dependent Ca²⁺ signaling stimulates cPLA₂/AA. We previously reported that arachidonic acid (AA) stimulates I_Ca,L [3]. This scenario is consistent with the fact that cPLA₂ activation requires both elevation of submicromolar intracellular [Ca²⁺] and phosphorylation by various kinases (see also [21–23]).

Fig 11.

Schematic summary showing the proposed signaling mechanisms underlying zint-β₂-AR stimulation of I_Ca,L in +PKA (A), -PKA (B) atrial myocytes. A; in +PKA myocytes, zint-β₂-AR stimulation acts via G_s to activate adenylate cyclase (AC)/cAMP-dependent kinase (PKA) which in turn stimulates I_Ca,L. Both basal and stimulated PKA activity inhibits Raf-1 signaling, thereby preventing zint-β₂-AR stimulation of I_Ca,L via cPLA₂. B; in cells not attached to LMN, inhibition of PKA by KT5720 removes inhibition of Raf-1. C; cell attachment to LMN acts via β₁ integrins and FAK/PI-(3)K/Akt signaling to inhibit adenylate cyclase (AC)/cAMP-dependent kinase (PKA) activity, thereby removing inhibition of Raf-1. β₂-AR stimulation acts via G_i to stimulate PLC leading to activation of PKCα and IP₃R-mediated Ca²⁺ signaling within caveolae. With PKA inhibited, PKCα stimulates Raf-1/MEK/ERK1/2 signaling. Together, ERK1/2 and IP₃-mediated Ca²⁺ signaling activate cPLA/AA, resulting in stimulation of I_Ca,L.

Conclusions

These findings may have important implications with respect to the aging and/or failing heart, both of which exhibit decreases in adenylate cyclase activity. In both animal models [37] and in the human right atrium [38], increasing age is associated with a decrease in β₁-AR function that results from a decrease in adenylate cyclase activity. Likewise, adenylate cyclase activity is depressed in the failing human heart [39] and canine pacing-induced heart failure [40, 41] and yet β₂-AR signaling is preserved. Previous studies by Nalli et al suggest that PKA and PKG phosphorylates PLCβ3 in gastric smooth muscle cells and decreases PI hydrolysis [42]. The decrease in PI hydrolysis has been suggested to diminish IP₃-mediated Ca²⁺ release decreasing muscle contraction. Along these lines, the present studies also suggest that feline atrial cardiomyocytes exhibit a similar regulation as that seen in gastric muscle. Our research also suggests that increases in extracellular matrix proteins, i.e. fibrosis, may contribute to decreases in adenylate cyclase/PKA activity in the aging and/or failing heart, which in turn switches β₂-AR signaling from cAMP/PKA to cPLA₂/AA. This mechanism may be responsible for the preservation of β₂-AR signaling while β₁-AR signaling is depressed. Thus, the diminished PKA activity in cardiac fibrosis may decrease phosphorylation of PLC resulting in an increase in IP hydrolysis, Ca²⁺ release and greater contraction. Indeed, our present results indicate that inhibition of PLC significantly inhibited β₂-AR stimulation of I_Ca,L in–PKA or +LMN myocytes but not in +PKA myocytes. In addition, our results suggest that 2-APB, an agent that inhibits IP₃-mediated Ca²⁺ release in cat atrial myocytes [24] significantly inhibited β₂-AR stimulation of I_Ca,L in both–PKA and +LMN myocytes but not in +PKA myocytes. Together, these results indicate that β₂-AR stimulation of cPLA₂ in both–PKA and +LMN myocytes is dependent on PLC/IP₃-mediated Ca²⁺ signaling. Given that cPLA₂/AA signaling is a potentially pro-inflammatory mediator, the present findings suggest that fibrosis in the aging and/or failing heart may predispose inflammation and atrial dysfunction. In fact, AA is reported to slow atrial conduction and has been implicated in the development of postoperative atrial fibrillation [43]. On the other hand, activation of the Raf/MEK/ERK pathway may be cardioprotective [44]. In fact, inhibition of adenylate cyclase/PKA and subsequent activation of Raf-1/MEK/ERK1/2 signaling enhances cardiac resistance to oxidative stress, increases cell survival, and extends lifespan [45]. Therefore, inhibition of adenylate cyclase/PKA activity and the remodeling of β₂-AR signaling to activate Raf/MEK/ERK1/2 may help ameliorate the deterioration of function that occurs in the aging and/or failing atrium. Previous studies have indicated that, beside inhibition of COX, aspirin has COX-independent mechanisms that plays a role in tumor suppression [46, 47]. Aspirin reportedly inhibits PI3K/Akt kinase activity in epithelial ovarian cancer cells [48]. Aspirin down regulates the expression of PI3K, Akt and ERK in rodent model of acute pulmonary embolism [49] and in turn inhibit the release of inflammatory cytokines [50]. On the other hand, our studies indicate that laminin acts via FAK/PI-(3)K/Akt signaling to inhibit adenylate cyclase-mediated stimulation of I_Ca,L. Although, there is no direct evidence suggesting that aspirin or NSAIDs play a role in cardiomyocyte aging, we hypothesize that aspirin or NSAIDs may influence cPLA2/AA pathway and this requires further investigation.

Supporting Information

S1 Fig. Effects of Raf-1 inhibition (10 μM GW5074) on zint-β₂-AR stimulation of I_Ca,L in +LMN atrial myocytes.

A; In +LMN myocytes, zint-β₂-AR stimulation of I_Ca,L was enhanced and GW5074 significantly inhibited zint-β₂-AR stimulation of I_Ca,L. Numbers in parentheses indicate the number of myocytes studied. * = P<0.004.

(TIF)

S2 Fig. Effects of 0.5 μM edelfosine on zint-β₂-AR stimulation of I_Ca,L in +LMN atrial myocytes.

A; in +LMN myocytes zint-β₂-AR stimulation elicited a typically enhanced increase in I_Ca,L and edelfosine significantly inhibited zint-β₂-AR stimulation of I_Ca,L. Numbers in parentheses indicate the number of myocytes studied. * = P<0.03.

(TIF)

S3 Fig. Effects of 2 μM 2-APB on zint-β₂-AR stimulation of I_Ca,L in +LMN atrial myocytes.

A; +LMN myocytes, zint-β₂-AR stimulation elicited a typically enhanced increase in I_Ca,L and 2-APB significantly inhibited zint-β₂-AR stimulation of I_Ca,L. Numbers in parentheses indicate the number of myocytes studied. * = P<0.01.

(TIF)

Abbreviations

LMN

laminin

cPLA₂

cytosolic phospholipase A₂

AA

arachidonic acid

ERK

extracellular-signal regulated kinase

MEK

mitogen-activated protein kinase kinase

Raf-1 (also known as c-Raf)

MAP kinase kinase kinase

PLC

phospholipase C

FAK

focal adhesion kinase

PI-(3)K

phosphatidyinositol-3’ kinase

Akt

protein kinase B

zint-β₂-AR

zinterol β₂-adrenergic receptor

I_Ca,L

L-type Ca²⁺ current

Adv

adenovirus

dn

dominant negative

Data Availability

All relevant data are within the paper and its Supporting Information files.

Funding Statement

This work was supported by NIH grants HL079038 (S.L.L.), HL62426-Project 4 and AHA Scientist Development Grant (13SDG14000006) (M.R.P.) and a grant from the Dr. Ralph and Marian Falk Medical Research Trust (A.M.S.).