Sustained &bgr;1-Adrenergic Stimulation Modulates Cardiac Contractility by Ca2+/Calmodulin Kinase Signaling Pathway

A tenet of &bgr;1-adrenergic receptor (&bgr;1AR) signaling is that stimulation of the receptor activates the adenylate cyclase-cAMP-protein kinase A (PKA) pathway, resulting in positive inotropic and relaxant effects in the heart. However, recent studies have suggested the involvement of Ca2+/calmodulin-dependent protein kinase II (CaMKII) in &bgr;1AR-stimulated cardiac apoptosis. In this study, we determined roles of CaMKII and PKA in sustained versus short-term &bgr;1AR modulation of excitation-contraction (E-C) coupling in cardiac myocytes. Short-term (10-minute) and sustained (24-hour) &bgr;1AR stimulation with norepinephrine similarly enhanced cell contraction and Ca2+ transients, in contrast to anticipated receptor desensitization. More importantly, the sustained responses were largely PKA-independent, and were sensitive to specific CaMKII inhibitors or adenoviral expression of a dominant-negative CaMKII mutant. Biochemical assays revealed that a progressive and persistent CaMKII activation was associated with a rapid desensitization of the cAMP/PKA signaling. Concomitantly, phosphorylation of phospholamban, an SR Ca2+ cycling regulatory protein, was shifted from its PKA site (16Ser) to CaMKII site (17Thr). Thus, &bgr;1AR stimulation activates dual signaling pathways mediated by cAMP/PKA and CaMKII, the former undergoing desensitization and the latter exhibiting sensitization. This finding may bear important etiological and therapeutical ramifications in understanding &bgr;1AR signaling in chronic heart failure.

A s a prototypical member of G protein-coupled receptor (GPCR) superfamily, ␤-adrenergic receptor (␤AR) plays a central role in sympathetic regulation of cardiac function. 1,2 Stimulation of ␤AR by catecholamines induces robust chronotropic, inotropic, and relaxant effects via the Gs-adenylate cyclase-cAMP-protein kinase A (PKA) pathway. 3,4 This signaling pathway is also thought to be responsible for other functions of ␤AR, such as regulation of metabolism, gene expression, cell growth, and apoptosis. 2 However, sustained ␤AR activation under pathological conditions such as hypertension and congestive heart failure will result in downregulation and desensitization of ␤AR attributable to the negative feedback of this pathway. [5][6][7] Recent studies have revealed unanticipated complexity of ␤AR signal transduction. For ␤ 2 AR subtype stimulation in the heart, a parallel activation of G i protein counterbalances G s -mediated contractile response. Whereas ␤ 1 AR stimulated contractile response is thought to be mediated exclusively by the cAMP/PKA signaling pathway, 8,9 ␤AR regulation of cell growth and remodeling involves mitogen-activated protein kinase (MAPK) cascades and phosphoinositol 3 kinase (PI3K) pathway. 10,11 Moreover, we have recently shown that sustained (24-hour) ␤ 1 AR stimulation progressively activates Ca 2ϩ /calmodulin-dependent protein kinase II (CaMKII), which is obligatory to cardiac apoptosis. 12 Therefore, in addition to the classic cAMP/PKA pathway, chronic ␤AR stimulation under certain physiological and pathophysiological circumstances may evoke pathways other than cAMP/ PKA. 13,14 These lines of evidence raise the question whether ␤ 1 AR modulates cardiac contractility using different sets of signaling mechanisms in short-term versus prolonged receptor stimulation.
CaMKII is a widely expressed protein kinase that modulates various functions ranging from learning and memory of the nervous system, muscle contraction, cell secretion to gene expression. 15 In the heart, CaMKII␦ is the predominant isoform and plays a pivotal role in regulating cardiac performance and remodeling such as myocyte hypertrophy, 16 apoptosis, 12 and heart failure. 17 Furthermore, CaMKII modulates an array of key proteins involved in cardiac excitationcontraction (E-C) coupling and Ca 2ϩ handling, such as the sarcoplasmic/endoplasmic reticulum Ca 2ϩ -ATPase (SERCA) and its regulator, phospholamban (PLB), ryanodine receptor (RyR) Ca 2ϩ release channels, and sarcolemmal L-type Ca 2ϩ channels (LCC). 18 -22 However, the involvement of CaMKII in ␤ 1 AR modulation of myocardial contractility remains obscure.
The present study aimed at appraising roles of CaMKII and PKA in ␤ 1 AR modulation of cardiac E-C coupling, with an emphasis on the signaling mechanism for the sustained ␤ 1 AR stimulation. We found that both short-term and sustained ␤ 1 AR stimulation are efficacious in mediating positive inotropic and relaxant effects in cardiac myocytes. Unlike the short-term ␤ 1 AR stimulation, the sustained responses are mediated mainly by the CaMKII rather than the cAMP/PKA pathway. Furthermore, molecular integration of these two signaling pathways is mediated by dual site phosphorylation of key proteins involved in cardiac E-C coupling and its physiological regulation.

Isolation and Culture of Cardiac Myocytes
Cardiac myocytes were isolated from male Sprague-Dawley rat hearts (Charles River Laboratories, Wilmington, Mass) using standard enzymatic technique as described previously. 23 Freshly isolated myocytes were plated at a density of 0.5 to 1ϫ10 4 /cm 2 in dishes precoated with 20 g/mL laminin (Upstate Biotechnology). The culture medium (M199, SIGMA) containing (in mmol/L) creatine 5, L-carnitine 2, taurine 5, insulin-transferrin-selenium-X 0.1%, HEPES 25, and penicillin plus streptomycin 1%, were adjusted to pH 7.4 with NaOH at 37°C. For sustained ␤ 1 AR stimulation, norepinephrine (100 nmol/L with vitamin C) was added along with ␣ 1 AR antagonist, prazosin (1 mol/L), for 24 hours. All the antagonists were added at least 10 minutes before norepinephrine.

Dominant-Negative CaMKII␦C Adenovirus Construction and Myocyte Infection
Dominant-negative CaMKII␦C (DN-CaMKII) was generated by replacing the residue lysine43 with alanine (K43A) using the transformer site directed mutagenesis kit (Clontech). Adenoviral expression of ␤-gal or the HA-tagged DN-CaMKII was performed at the multiplicity of infections (MOI) of 100. Twenty four hours after adenoviral infection, norepinephrine (100 nmol/L) was added.

Measurements of Cell Shortening and Ca 2؉ Transients
Measurements of cell contraction and Ca 2ϩ transients were performed 24 hours after norepinephrine exposure. Normal cultured myocytes (24 hours) were used for short-term ␤ 1 AR stimulation. Myocytes were field-stimulated at 0.5 Hz in perfusion solution containing (in mmol/L) NaCl 137, KCl 4.9, CaCl 2 1, MgSO 4 1.2, NaH 2 PO 4 1.2, glucose 15, and HEPES 20 (pH 7.4). Prazosin was added 10 minutes before short-term norepinephrine treatment. Protocols pertaining to specific experiments were given in the respective result figure In indicator-unloaded myocytes, cell length was monitored by an optical edge tracking method at a 3-ms time resolution. 23 In myocytes loaded with the Ca 2ϩ indicator fluo-4/AM (Molecular Probes, 20 mol/L for 30 minutes), Ca 2ϩ transients and cell shortening were measured with a confocal laser scanning microscope (LSM510, Carl Zeiss). Digital image analysis used customerdesigned programs coded in Interactive Data Language (IDL).

Receptor Radioligand Binding Assay
Cardiac myocytes were homogenized and crude membranes were prepared by centrifuging at 35 000g for 20 minutes at 4°C twice. ␤ 1 AR radioligand binding studies were performed in membranes (25 to 100 g/tube) using the nonselective ␤AR antagonist ligand 125 I-cyanopindolol ( 125 I-CYP, 1 to 300 pmol/L) as described previously. 24 Nonspecific binding was determined in the presence of 10 mol/L propranolol. The maximal numbers of binding sites (B max ) and equilibrium dissociation constants (Kd) for 125 I-CYP were determined by Scatchard analysis.

Immunostaining
Fixed cells were incubated with anti-HA antibody (1:500, Covance Research Products Inc) at 4°C overnight, followed by Cy5conjugated secondary antibody (1:1000, Jackson ImmunoResearch laboratories). Negative controls were obtained by incubating cells with only the secondary antibody.

CaMKII Activity and cAMP Accumulation
Cell lysate (500 g protein) was first immunoprecipitated with anti-CaMKII antibody (1:100, Santa Cruz Biotechnology) in a Ca 2ϩ -free medium. The precipitated proteins were evaluated with CaMKII assay kits (Upstate Biotechnology Inc) using a specific peptide substrate (KKALRRQETVDAL) as previously described. 12 cAMP level was measured with the cAMP ( 3 H) assay system (Amersham Biosciences) as described previously. 24

Western Blotting Analysis of PLB Phosphorylation
PLB phosphorylation was detected with the phosphorylation sitespecific antibodies recognizing P- 16 Ser PLB or P- 17 Thr PLB (1:10000, PhosphoProtein Research). Total PLB was detected with monoclonal antibody against PLB (Upstate Biotechnology).

Statistics
Data were reported as meanϮSEM. Student t test and ANOVA with repeated measurement were applied, when appropriate, to determine statistical significance of the differences. PϽ0.05 was considered statistically significant.

Sustained ␤ 1 AR Stimulation Induced Positive Inotropic and Relaxant Effects
To investigate possible modulatory effects of sustained ␤ 1 AR stimulation on cardiac E-C coupling, enzymatically isolated rat ventricular myocytes were cultured in the presence of ␤ 1 AR stimulation (norepinephrine 100 nmol/L plus the ␣ 1 AR blocker, prazosin 1 mol/L) for 24 hours. Cell shortening measured in the continued presence of ␤ 1 AR agonist was used as the end-point readout. Figure 1 shows typical examples and the average data of contractile response to short-term and sustained ␤ 1 AR stimulation. Sustained ␤ 1 AR stimulation (24 hours) elicited a 3.3-fold increase of the contraction amplitude ( Figure 1C), which was accompanied by a 47% decrease of the 90% relaxation time ( Figure 1D). Notably, the effects of sustained ␤ 1 AR stimulation on cell contraction and relaxation were fully reversible on washout of the agonist, as was the case with short-term ␤ 1 AR stimulation ( Figure 1). That sustained contractile responses are fully and rapidly reversible excludes the possibility that they were due simply to a long-term "memory" formed during prolonged receptor stimulation. 25 As is the case with short-term ␤ 1 AR stimulation, sustained contractile response of ␤ 1 AR was sensitive to a nonselective ␤AR blocker, propranolol (10 mol/L), or a specific ␤ 1 AR blocker, CGP 20712A (0.5 mol/L). Pretreatment with propranolol or CGP 20712A prevents the increase of contractility by subsequent 24-hour norepinephrine stimulation (Table  1); ICI 118,551 (0.5 mol/L), a ␤ 2 AR specific antagonist, did not affect the 24-hour norepinephrine-induced response (data not shown). Radioligand binding assay of receptor density in crude membrane fraction revealed a trend of decrease (though statistically insignificant) in ␤ 1 AR density after 24-hour norepinephrine stimulation (Table 2), consistent with previous in vitro reports. 26 When compared with the response to short-term ␤ 1 AR stimulation (10 minutes), we found that sustained ␤ 1 AR stimulation was equally efficacious in mediating the positive inotropic and relaxant effects ( Figure 1). This observation is somewhat unexpected because it has been well established that cAMP/PKA signaling desensitizes within hours during prolonged ␤ 1 AR stimulation. 5,27

Sustained ␤ 1 AR Contractile Response Is Mediated by CaMKII Signaling Pathway
Whereas short-term and sustained ␤ 1 AR stimulations elicited indistinguishable responses in terms of cell contraction, there is no a priori reason that the same PKA-dependent mechanism is responsible for the contractile responses in both cases. To appraise role of PKA in mediating the sustained ␤ 1 AR responses, we used two inhibitors of the cAMP/PKA signaling: PKI, a membrane permeable peptide inhibitor of PKA, and Rp-cpt-cAMPS, an inhibitory cAMP analogue. We found that PKI treatment (10 mol/L for 30 minutes) largely blocked the effect of short-term ␤ 1 AR stimulation (Figure 2A and 2B) as expected. By contrast, PKI was unable to reverse the increase in contraction amplitude in cells exposed to ␤ 1 AR stimulation for 24 hours (Figure 2A and 2B). Likewise, the cAMP antagonist Rp-cpt-cAMPS (100 mol/L for 30 minutes) significantly inhibited the effects of short-term, but not sustained, ␤ 1 AR stimulation ( Figure 2B). These results indicate that sustained contractile response of ␤ 1 AR stimulation is largely PKA-independent.
Recently, it has been shown that sustained ␤ 1 ARstimulated cardiomyocyte apoptosis requires activation of CaMKII signaling independently of PKA. 12 Previous studies have also established an important role for CaMKII modulation of phosphorylation and function of key proteins involved in cardiac E-C coupling, including LCC, 21,22 SERCA, 18 and its regulator PLB, 19 and RyR. 20,28 Next, we examined the involvement of CaMKII signaling in the contractile response to sustained ␤ 1 AR stimulation. As shown in Figure 2C and 2D, the inotropic effect of sustained ␤ 1 AR stimulation was reversed by KN93 (10 mol/L for 30 minutes), a synthetic CaMKII inhibitor, whereas the effect of short-term ␤ 1 AR stimulation was unaffected by KN93. KN92 (10 mol/L), an  inactive analogue of KN93, exerted no significant effects on either short-term or sustained ␤ 1 AR stimulation (data not shown). Similar to KN93, a cell-permeable peptide inhibitor of CaMKII, myristoylated autocamtide-2 related inhibitory peptide (AIP, 10 mol/L for 30 minutes), completely abolished the positive inotropic effect of sustained ␤ 1 AR stimulation, without affecting those of short-term ␤ 1 AR stimulation ( Figure 2D). Thus, inhibition of cAMP/PKA and CaMKII preferentially blocked contractile modulation by short-term and sustained ␤ 1 AR stimulation, respectively, indicating that the initial responses depend mainly on PKA, whereas the sustained responses require CaMKII signaling.

CaMKII-Dependent Increase of Ca 2؉ Transients in Response to Sustained ␤ 1 AR Stimulation
To further investigate cellular mechanisms underlying the contractile responses in sustained versus short-term ␤ 1 AR stimulation, we measured Ca 2ϩ transients and the corresponding cell contraction in Ca 2ϩ indicator-loaded myocytes, using confocal microscopy. Figure 3A shows typical micrographs of cellular Ca 2ϩ transients and shortenings from cells that underwent short-term and sustained ␤ 1 AR stimulation in the absence or presence of PKA or CaMKII inhibitor. Both short-term and sustained ␤ 1 AR stimulation increased Ca 2ϩ transient amplitude (⌬F/F 0 from 2.7Ϯ0.1 to 5.5Ϯ0.2 or 6.0Ϯ0.2 for cells with short-term or sustained ␤ 1 AR stimulation, respectively, nϭ37 to 55 cells) and contractile amplitude (from 2.5Ϯ0.3% to 12.2Ϯ0.6% or 11.1Ϯ0.6% for cells with short-term or sustained ␤ 1 AR stimulation, respectively, nϭ33 to 55; Figure 3). The short-term ␤ 1 AR stimulationinduced Ca 2ϩ and contractile responses were blocked by Rp-cpt-cAMPS but not by AIP (Figure 3). In contrast, the sustained ␤ 1 AR stimulation-induced increases in Ca 2ϩ transients and contraction were resistant to Rp-cpt-cAMPS, but were reversed by AIP (Figure 3). These data indicate that both cAMP/PKA and CaMKII signaling pathways of ␤ 1 AR stimulation augment cell contraction by enhancing intracellular Ca 2ϩ transients.

Effects of Dominant-Negative CaMKII on Short-Term and Sustained ␤ 1 AR Stimulation
To confirm that the sustained ␤ 1 AR contractile effect is CaMKII-dependent, we resorted to molecular and genetic manipulation of the CaMKII signaling system. An adenovirus carrying the HA-tagged CaMKII gene with a dominantnegative mutation (DN-CaMKII, mutation K43A) that abrogates the kinase activity was constructed and infected cardiac myocytes in culture. Twenty-four hours after adenoviral infection at MOI of 100, nearly 100% cells showed intense immunostaining of DN-CaMKII as visualized by an anti-HA antibody ( Figure 4A). Western Blotting analysis revealed that CaMKII-specific phosphorylation of PLB at 17 Thr was also markedly reduced in the DN-CaMKII expressing myocytes ( Figure 4B). Figure 4C and 4D show that sustained ␤ 1 AR stimulation in DN-CaMKII-expressing cells failed to elicit significant increases of Ca 2ϩ transient amplitude or contraction amplitude. However, no significant difference between ␤-gal and DN-CaMKII expression groups was found in contractile and Ca 2ϩ responses to short-term ␤ 1 AR stimulation ( Figure 4C and 4D). These data corroborate that sustained ␤ 1 AR stimulation enhances Ca 2ϩ transients and cell contraction through a CaMKII pathway, whereas the initial ␤ 1 AR response is largely CaMKII-independent.

Switch of cAMP and CaMKII Signaling During Sustained ␤ 1 AR Stimulation
To better characterize the aforementioned signaling shift during ␤ 1 AR stimulation, we directly measured cellular cAMP accumulation and CaMKII activation over a wide period of time (from 10 minutes up to 24 hours). ␤ 1 AR stimulation elicited a rapid increase of cellular cAMP production that reached its peak in 10 minutes. In the continued presence of the ␤ 1 AR agonist, however, cAMP production was reduced by 66% at 3 hours after stimulation ( Figure 5A), and returned to basal level at the steady state (12 hours). The gradual decay of cAMP is consistent with the notion that cAMP/PKA signaling undergoes substantial desensitization during prolonged receptor stimulation. [5][6][7] Parallel measurement of CaMKII activity revealed a distinctly different temporal pattern for CaMKII response during ␤ 1 AR stimulation. CaMKII activity rose exponentially (time constant ϭ15 minutes) without an initial overshoot ( Figure  5B). The plateau was reached after Ϸ1 hour stimulation and was stable for at least 24 hours. Overall, the gradual sensitization of CaMKII signaling roughly mirrored the desensitization of cAMP/PKA signaling, indicating a shift from a cAMP/PKA-dominant signaling to a CaMKII-dominant signaling. The slow and nondecremental activation of CaMKII provides the basis for the sustained contractile and Ca 2ϩ responses to ␤ 1 AR stimulation.

Phospholamban (PLB) as a Molecular Integrator of ␤ 1 AR-Stimulated PKA and CaMKII Signals
The SR protein PLB, in its unphosphorylated form, serves as a constitutive inhibitor of the SR Ca 2ϩ ATPase. PKA and CaMKII can independently phosphorylate PLB at 16 Ser and 17 Thr, respectively, and either site phosphorylation is sufficient to reverse its inhibition on SERCA activity and subsequently elicit positive inotropic and relaxant responses. 29 Thus, PLB with its dual-site phosphorylation might operate as a molecular integrator of both short-term and sustained ␤ 1 AR signaling.
To explore this possibility, we examined the site-specific phosphorylation of PLB in response to sustained and shortterm ␤ 1 AR stimulation. The phosphorylation at PKAdependent site (P-16 Ser) was increased by 7.2Ϯ0.9-fold (nϭ4, PϽ0.001 versus control) at 10 minutes after exposure to norepinephrine, but was then diminished toward the basal level during sustained ␤ 1 AR stimulation ( Figure 5C). Conversely, phosphorylation at the CaMKII-dependent site (P-17 Thr) was significantly increased in response to sustained, but not short-term, ␤ 1 AR stimulation ( Figure 5D). These data  support the idea that dual site phosphorylation by PKA and CaMKII in effector proteins (eg, PLB) serves to integrate the dual signaling pathways of ␤ 1 AR stimulation.

Effect of cAMP/PKA on CaMKII Mediated ␤ 1 AR Contractile Response
Because ␤ 1 AR-stimualted cAMP/PKA signaling precedes the CaMKII signaling, it might be argued that activation of CaMKII pathway in sustained ␤ 1 AR stimulation is still dependent on the initial PKA activation. However, preinhibition of cAMP/PKA with Rp-cpt-cAMPS (100 mol/L) did not influence the sustained ␤ 1 AR-stimulated contractile and relaxant responses or the blockade effect of AIP ( Figure 6A). Conversely, direct activation of cAMP/PKA pathway by forskolin (1 mol/L), an adenylate cyclase activator, or cpt-cAMP (100 mol/L), an active cAMP analogue, elicited no CaMKII-dependent component in the sustained inotropic response ( Figure 6B and 6C). Thus, cAMP/PKA signaling appears to be neither sufficient nor necessary for ␤ 1 AR activation of CaMKII.

Time-Dependent Shift of ␤ 1 AR Dual Signaling Pathways
We have systematically examined the signaling mechanisms underlying cardiac contractile modulation by short-term and sustained ␤ 1 AR stimulation (24 hours), using myocyte culture combined with genetic manipulation, confocal imaging, and biochemical measurements. In contrast to the anticipated receptor desensitization, [5][6][7]27 we demonstrated that shortterm and sustained ␤ 1 AR stimulation similarly enhance Ca 2ϩ transients and contraction and accelerate relaxation (Figures 1  and 3). Despite phenomenological similarities, inhibition of CaMKII by specific inhibitors or adenoviral expression of DN-CaMKII exerts profound inhibitory effects on the sus-tained, but not short-term, ␤ 1 AR responses, whereas inhibition of the cAMP/PKA pathway preferentially blocks the responses to short-term ␤ 1 AR stimulation (Figures 2 to 4). By tracking cAMP production and CaMKII activation over an extended time course (Figure 5), we have uncovered that, CaMKII activity rose to a plateau that does not show any  noticeable decay, whereas the cAMP/PKA signaling subsides in the continued presence of ␤ 1 AR agonist. These results indicate that ␤ 1 AR signaling undergoes a time-dependent switch from the PKA-dominant pathway to the CaMKIIdominant pathway after receptor stimulation. Because inhibition of the cAMP/PKA pathway did not alter the responses to sustained receptor stimulation and receptor-independent cAMP/PKA signal failed to elicit CaMKII-dependent response (Figure 6), activation of the CaMKII pathway may not be consequential to the transient cAMP/PKA activation. In other words, different pathways initiated from the same GPCR may manifest desensitization or sensitization independently. 6,12,30 As compared with ␤ 1 ARstimulated CaMKII activation in ␤ 1 ␤ 2 AR double knockout mouse cardiac myocytes overexpressing ␤ 1 AR (Ϸ3-fold, Ϸ60 minutes), the elevation of CaMKII activity by native ␤ 1 AR in heart cells is modest (35% over baseline) yet exhibits faster kinetics ( Ϸ15 minutes) ( Figure 5B). These quantitative discrepancies might reflect differences in the receptor density or the coupling efficiency of ␤ 1 AR to downstream signaling pathways in these two systems.

Molecular Integration of ␤ 1 AR-Stimulated PKA and CaMKII Signals
That short-term and sustained ␤ 1 AR contractile and Ca 2ϩ responses are virtually indistinguishable indicates a seamless integration of ␤ 1 AR-stimulated PKA and CaMKII signals at the molecular and cellular levels. Indeed, we detected that a shift of PLB phosphorylation from the PKA-dependent site ( 16 Ser) in short-term ␤ 1 AR stimulation to the CaMKIIdependent site ( 17 Thr), which correlates well with the timedependent changes in cAMP production and CaMKII activation in sustained ␤ 1 AR stimulation. Because either site phosphorylation of PLB is sufficient to release its inhibition on the SERCA, 28,31,32 PLB serves as a key molecular integrator of the dual signaling pathways of ␤ 1 AR stimulation. Disinhibition of SERCA activity by PLB phosphorylation will enhance SR Ca 2ϩ recycling, accelerate relaxation of Ca 2ϩ transients and cell contraction, and subsequently increase the SR Ca 2ϩ load, 12 contributing to both the positive inotropic and relaxant effects of ␤ 1 AR stimulation. It is noteworthy that the relaxant effect of sustained norepinephrine exposure was not significantly influenced by CaMKII inhibitors (Figure 2C), suggesting the possibility for a differential regulation of peak contraction and relaxation by sustained ␤ 1 AR stimulation.
In addition to PLB, previous studies have also shown that LCC Ca 2ϩ currents are regulated by both PKA and CaMKII. 21,33 Our preliminary data suggested the enhancement of LCC Ca 2ϩ currents during initial (10 to 30 minutes) and sustained (3 to 6 hours) ␤ 1 AR stimulation is mediated by PKA-and CaMKII-dominant mechanisms, respectively, suggesting LCC serves as another molecular integrator of the PKA and CaMKII signals of ␤ 1 AR stimulation. In addition, it has been documented that either PKA or CaMKII can phosphorylate RyRs, 20,34 although functional consequence of such phosphorylation remains controversial. 34 -36 Functional Significance of ␤ 1 AR Signaling Switch The finding of time-dependent switch between two signaling pathways of ␤ 1 AR stimulation bears important implications in understanding physiological modulation of cardiac function as well as the etiology and pathophysiology of cardiac diseases associated with chronically exaggerated ␤AR signaling. First, the present result reassures that short-term ␤ 1 AR stimulation, as in fight-or-flight response or during exercise, is largely mediated by the cAMP/PKA pathway, rapid activation of which enables a beat-to-beat regulation of cardiac performance. However, the cAMP/PKA signaling desensitizes nearly completely ( Figure 5A) and is unlikely to play any major role for more enduring responses. In contrast, the CaMKII signaling does not undergo any appreciable desensitization over the period of observation (up to 24 hours). If this can be extrapolated to chronic ␤ 1 AR stimulation in vivo, a corollary is that long-term ␤ 1 AR effects, beneficial or toxic, must be due primarily to CaMKII-dependent signal transduction. This concept resonates with several lines of evidence found in previous studies. First, ␤AR-stimulated cardiac cell hypertrophy is largely PKA-independent but requires CaMKII activation. 37,38 Second, CaMKII, but not PKA activation is obligatory to ␤ 1 AR-mediated cardiac apoptosis. 12 Third, whole-animal and clinic data hint on that cardiac toxic effects of chronically enhanced ␤AR stimulation are likely PKA-independent. 39 Emerging evidence also suggests that increased CaMKII activity in human heart failure might play a compensatory role for the decreased cardiac contractility in failing hearts. 40,41

Possible Mechanisms Underlying CaMKII Activation
The exact mechanisms underlying CaMKII activation and the switch of signaling pathways remain unknown. Preliminary observations showed that the ␤ 1 AR CaMKII-dependent response is insensitive to G i /G o inhibition by pertussis toxin (1.5 g/mL, added 3 hours before norepinephrine). Results in Figure 6 further suggest that direct and sustained activation of the cAMP/PKA signaling failed to activate CaMKII, and that blockage of cAMP/PKA does not prevent CaMKII from activation after ␤ 1 AR stimulation; thus, the ␤ 1 AR-cAMP signaling is neither necessary nor sufficient for the activation of ␤ 1 AR-induced CaMKII signaling. Apart from the cAMP/ PKA pathway, some evidences support a direct coupling between LCC and GPCRs including ␤ 2 AR. 42,43 Overexpression of G s␣ has also been shown to activate LCC currents via PKA-independent mechanisms. 44 If LCC could be activated in a PKA-independent manner, LCC Ca 2ϩ entry might be responsible for the gradual CaMKII activation during ␤ 1 AR stimulation. In this scenario, the cAMP-dependent activation of LCC may differ from ␤ 1 AR-dependent, but cAMPindependent, LCC activation because sustained cAMPdependent activation of LCC by forskolin or active cAMP analogue does not express the CaMKII signaling even after 24 hours. Recent studies have also hinted on a G proteinindependent GPCR signaling. For instance, the carboxyl terminus of ␤ 1 AR can directly interact with PDZ-motif containing proteins such as PSD-95 and Ras exchanger regulatory factor. 45,46 By analogy, ␤ 1 AR might directly activate element(s) of the CaMKII pathway through G proteinindependent mechanisms. Future investigations are warranted to explore these possibilities.
In summary, ␤ 1 AR modulation of cardiac E-C coupling invokes dual signaling pathways mediated by cAMP/PKA and CaMKII, respectively. The cAMP/PKA signaling is biphasic with a prominent early peak, whereas the CaMKII signaling is slow but persistent. During the signaling switch, the inotropic and relaxant responses are maintained at the steady state because of the convergence of PKA and CaMKII signals onto common effector proteins involved in E-C coupling and intracellular Ca 2ϩ regulation. As a result, the effects of sustained ␤ 1 AR stimulation (eg, inotropy, cell growth, and cell death) are due primarily to CaMKII, rather than PKA signaling. These findings provide mechanistic insights into ␤AR modulation of cardiac function and GPCR signaling, and suggest new concepts for mechanistic understanding and therapeutic treatment of cardiac conditions such as hypertension and chronic heart failure that are associated with sustained elevation of endogenous catecholamines.