α-1 Subunit

If the main effects of β-AR stimulation on voltage dependence and amplitude of the I_Ca,L and their consequences for the fight-or-flight response are well documented in the literature, the molecular events that mediate the increase of Ca_V1.2 activity during a sympathetic stimulation remain elusive; this is due to the difficulty of reconstituting such modulation in heterologous overexpression systems. The α_1C-subunit of Ca_V1.2 channels exhibits multiple potential PKA phosphorylation sites in the N- and the C-terminal regions (Figure 37-3).² Despite the fact that α_1C is a substrate for phosphorylation by PKA in vitro,^16,17 several attempts to mimic the adrenergic stimulation of Ca_V1.2 channels in expression systems have failed.^18–20 This failure led to the hypothesis that at least one missing link in heterologous systems would preclude the reconstitution of such modulation of overexpressed Ca_V1.2 channels. One of these links could be an A-kinase-anchoring protein (AKAP) that allows the cAMP modulation and the PKA phosphorylation of serine 1928 in the C-terminus of Ca_V1.2 channels in HEK293 cells.²¹ A conserved leucine zipper motif in the C-terminus of Ca_V1.2 identified in native cardiac cells directly anchors a low molecular weight AKAP15-PKA complex to ensure a fast and efficient β-adrenergic modulation of I_Ca,L (see Figure 37-3),²² allowing its phosphorylation at serine 1928 when β-ARs are stimulated.²³ Surprisingly, the C-terminal part of rabbit Ca_V1.2 undergoes proteolytic processing by calpain at residue 1821, leaving two size forms of the α_1C-subunit of these channels expressed in cardiac cells,¹⁷ whereas this distal 37 to 50-kDa peptide contains the serine 1928 phosphorylated by PKA and the binding site for AKAP15. In fact, this fragment constitutes a potent autoinhibitory domain that covalently associates with the proximal C-terminal part of α_1C, reducing its open probability and shifting the voltage dependence of activation to depolarized potentials.²⁴ The role of serine 1928 in I_Ca,L upregulation by β-AR has since been challenged. A first study showed that a DHP-resistant Ca_V1.2 channels mutated at position 1928 (S1928A) displays an unaltered response to the β-AR agonist isoproterenol when overexpressed in isolated ventricular myocytes.²⁵ Moreover, the generation of a S1928A knock-in mouse model confirmed that the phosphorylation of this residue by PKA is not required for the functional effects of β-AR stimulation on Ca_V1.2 in cardiac cells, because these mice exhibit an essentially conserved response of I_Ca,L to isoproterenol and typical chronotropic and inotropic responses to β-AR stimulation.²⁶ Therefore, although phosphorylation of serine 1928 within the C-terminus of Ca_V1.2 definitely occurs when β-ARs are stimulated,^17,21,23 it does not correlate with the functional effects observed on I_Ca,L. As a result, PKA would phosphorylate other PKA sites within the channel or another binding partner to mediate the increase of its activity.

β-Subunit

Three serines (S459, S478, and S479) of the cardiac the Ca_Vβ_2a subunit were identified as putative PKA phosphorylation sites.²⁷ Coexpression of this subunit with a Ca_V1.2 lacking the C-terminal part including the serine 1928 in TsA-201 cells allowed an upregulation of the current by cAMP/PKA pathway, whereas expression of mutated Ca_Vβ_2a at serine 478 and 479, but not at position 459, was unable to do so.²⁸ The authors concluded that the Ca_Vβ_2a ancillary subunit was the main target for PKA to mediate the β–AR stimulation of I_Ca,L; however, this conclusion has been questioned by experiments realized in a more physiologic context. Overexpression of a Ca_Vβ_2a construct mutated for its PKA phosphorylation sites in ventricular cardiac cells did not prevent the cAMP/PKA modulation of I_Ca,L.¹⁴ Nonetheless, the Ca_Vβ₂ subunit can associate with the giant cytoskeletal protein of 700 kDa named ahnak, which emerged as an important player in the β-AR regulation of cardiac I_Ca,L (see Figure 37-3). Through its interaction with Ca_Vβ₂, ahnak would serve as a brake on I_Ca,L that would be relieved by PKA phosphorylation of the ancillary subunit and the cytoskeletal protein.²⁹ Interestingly, ahnak polymorphism occurs, and the genetic variant generated interferes with the β-AR stimulation of I_Ca,L by reducing the Ca_Vβ₂ interaction with ahnak.²⁹ However, if the involvement of the Ca_Vβ or other binding partners such as ahnak cannot be ruled out completely, recent studies have reaffirmed the importance of the proteolytically cleaved distal C-terminus of Ca_V1.2 for its upregulation upon β-AR stimulation. In addition, a new model for the molecular basis of β-AR stimulation of Ca_V1.2 has been proposed.³⁰ Overexpression in TsA-201 cells of a truncated Ca_V1.2 channel at position A1800 (the site of in vivo proteolytic cleavage previously determined for skeletal muscle Ca_V1.1 channel³¹) with α₂δ₁ and Ca_Vβ_1b and the distal C-terminal part of the channel (peptide from 1821-2171) produced a functional channel that is autoinhibited. Two presumed sites for PKA phosphorylation were identified upstream from the proteolytic site—a serine at position 1700 and a threonine 1704 at the interface of the distal and the proximal C-terminal parts. Although a mutation of S1928 confirmed that its phosphorylation is not required for the increase of the current, substitution of T1704 and S1700 for alanines revealed that phosphorylation of T1704 is required for basal Ca_V1.2 channel activity, whereas S1700 is crucial for its cAMP/PKA-induced upregulation. In this scheme, the AKAP15 is still required to coordinate the PKA phosphorylation of S1700 that disrupts the interaction of the noncovalently distal C-terminus, thus relieving the inhibition of the channel.³⁰ This assumption is reinforced by the fact that mice expressing a Ca_V1.2 channel deleted for the distal C-terminus display altered cAMP/PKA regulation accompanied with reduced expression of AKAP15.³²

Receptor Specificity

Although only three types of β-adrenergic receptors (β-ARs) have been cloned (β₁-, β₂-, and β₃-ARs), the effect of catecholamines in the human heart is generally attributed to β₁– and β₂-ARs. β₁– and β₂-ARs are highly homologous receptors and are both positively coupled to AC/cAMP/PKA cascade, I_Ca,L and cardiac performance³³; however, they exert opposite effects on hypertrophy^38,39 and apoptosis,^40,41 and their respective contribution varies significantly depending on the cardiac tissue, pathophysiologic state, age, or developmental stage.⁴² Part of the difference is because β₂-ARs couple not only to G_αs but also to G_αi proteins, and this confines the cAMP-dependent signal to the membrane compartment and to activation of the LTCCs.⁴³ Another difference between the two β-AR subtypes is their location at the cell surface, with β₂-ARs present in the caveolae/lipid rafts^44,45 of the transverse-tubular structure⁴⁶ and β₁-ARs distributed throughout both caveolae–lipid rafts and nonlipid raft membrane domains⁴⁴ and in both plasma and T-tubular membranes.⁴⁶ Accordingly, the β₂-AR downstream activation of I_Ca,L is sensitive to disruption of caveolae by cholesterol depletion, whereas the β₁-AR stimulatory effect is not.⁴⁷ Therefore, β₂-AR stimulation exerts a local activation of LTCCs, whereas β₁-AR stimulation leads to activation of LTCCs in the distance.^43,48,49

The β₃-AR differs from β₁-and β₂-AR subtypes in its molecular structure and pharmacologic functions.⁵⁰ Expression of β₃-AR was demonstrated in human myocardium at the mRNA^51,52 and protein levels.^52–55 Interestingly, β₃-AR activation produces a negative inotropic effect in human endomyocardial biopsies from transplanted hearts^50,51 and in left ventricular samples from failing and nonfailing explanted hearts^50,54 that is due to activation of the NO/cGMP pathway, but increases I_Ca,L and contractility in human atrial tissue via the cAMP/PKA pathway (Figure 37-4).⁵⁶ This effect is reminiscent of the contractile effects of the serotonin 5-HT₄ receptors,⁵⁷ which are also coupled to increases in force of contraction⁵⁸ or I_Ca,L⁵⁹ in atria but not in healthy ventricles.

Role of A-Kinase-Anchoring Proteins

Cyclic AMP signaling components are organized into multiprotein complexes, an arrangement that increases both efficiency and specificity of the transduction cascade. AKAPs have an essential role in these arrangements. AKAPs form a large family of proteins comprising more than 50 members whose primary function is to anchor PKA in the vicinity of its substrates, thus ensuring the preferential phosphorylation of a limited number of targets.60,61 As discussed earlier, AKAP15 (also called AKAP 7 or AKAP15/18) is the main AKAP controlling the PKA phosphorylation of LTCCs. However, in a recent study, Ca_V1.2 phosphorylation and β-AR stimulation of I_Ca,L was found to be unchanged in mice in which the gene encoding this protein was inactivated, suggesting that PKA is anchored by a different protein to the channel.⁶² Furthermore, AKAP5 was found to target ACs, PKA and phosphatases within caveolae to allow specific PKA phosphorylation of the subpopulation of channels present in this compartment upon β-AR stimulation.⁶³ Although AKAPs share in common their ability to bind PKA, they are remarkably diverse scaffold proteins. Within each signalosome, AKAPs couple PKA to different substrates, enhancing the rate and fidelity of their phosphorylation by the kinase. Importantly, AKAPs not only bind PKA but act as scaffold proteins for other signaling components, including phosphatases 1 and 2,^64,65 Epac,⁶⁶ adenylyl cyclases,^67,68 and PDEs.⁶⁹ Recently, it has been demonstrated that the phosphoinositide 3-kinase p110γ, that was shown recently tether PKA,⁷⁰ orchestrates multiprotein complexes including different PDEs to control cardiac Ca_V1.2 phosphorylation during β-AR stimulation.⁷¹ The combination of PDEs and phosphatases present in individual AKAP complexes will affect the duration, amplitude, and spatial extent of cAMP/PKA signaling. Thus, by bringing together different combinations of upstream and downstream signaling molecules, AKAPs provide the architectural infrastructure for specialization of the cAMP signaling network.^61,66,72

Role of Phosphodiesterases

Localized cAMP signals can be generated by the interplay between discrete cAMP production sites and restricted diffusion within the cytoplasm. Restricted diffusion of cAMP can be achieved by several means. A first possibility is that physical barriers are created by specialized membrane structures within the cytoplasm. This method was initially proposed to explain the differences in cAMP concentration elicited by PGE₁ at the plasma membrane and in the bulk cytosol of HEK293 cells, although an experimental proof that this actually occurs is still lacking.⁷³ Another important mechanism that limits cAMP diffusion is cAMP degradation by PDEs, which appears to be critical to the formation of dynamic microdomains that confer specificity of the response.^35,60

Cardiac cAMP PDEs degrading belong to five families (PDE1 to PDE4 and PDE8) that can be distinguished by distinct enzymatic properties and pharmacology.⁷⁴ Among these families various enzymes were shown to degrade cAMP to allow a fine tuning of I_Ca,L regulation by PKA in the heart. Although PDE2 is not highly expressed in cardiomyocytes, it controls LTCC activity in various species, including human atrial myocytes.^75–78 This enzyme is activated by cGMP, and stimulation of guanylyl cyclase strongly decreases local cAMP levels controlling I_Ca,L with only modest effects on its global concentration, suggesting the existence of a cAMP microdomain including β-AR and LTCCs under tight control of PDE2.⁷⁹ Contrary to PDE2, PDE3 is inhibited by cGMP; this explains in part why cGMP at low concentration can also increase basal I_Ca,L as shown in human atrial myocytes.^78,80 In rodents, PDE3 and PDE4 are the major contributors to the total cAMP-hydrolytic activity^34,81 and PDE4 is dominant to modulate β-AR regulation of cAMP levels.^49,82–84 Multiple PDE4 variants associate with β-ARs,^85–87 RyR2,⁸⁸ SERCA2,^89,90 I_Ca,L,⁹¹ and I_Ks⁷² to exert local control of ECC. In larger mammals, PDE3 activity is dominant in microsomal fractions,^92–94 and PDE3 inhibitors exert a potent positive inotropic effect.⁹⁵ Selective inhibition of PDE3 with milrinone has been shown to improve cardiac contractility in patients with congestive heart failure.⁹⁶ The role of PDE4 is less well defined, but evidence is emerging that PDE4 could also have an important role in these species. In the canine heart, a large PDE4 activity is found in the cytoplasm,⁹² but PDE4 is also present in microsomal fractions, where it accounts for approximately 20% of the activity.⁹³ Recent studies have indicated that PDE4 is expressed in the human ventricle where, similar to rodents, it associates with β-ARs, RyR2, and phospholamban.^81,88 Moreover, PDE4 is the main PDE modulating LTCC activity in rodent cardiomyocytes,^77,84 and PDE4 was recently shown to control I_Ca,L, ECC and arrhythmias in the human atrium.⁹⁷

Early evidence of the contribution of PDEs to intracellular cyclic nucleotide compartmentation was obtained by comparing the effects of the nonselective β-AR agonist Iso, or the nonselective PDE inhibitor, 3-isobutyl-1-methylxanthine (IBMX), or the PDE3 inhibitor milrinone in guinea pig perfused hearts. Whereas each of these treatments increased intracellular cAMP and produced positive inotropic and lusitropic effects, differences in the phosphorylation pattern of PLB, TnI, and MyBP-C by PKA were observed.⁹⁸ These results were attributed to a functional cellular compartmentation of cAMP and PKA substrates owing to a different expression of PDEs at the membrane and in the cytosol.⁹² In canine ventricular myocytes, an increase in particulate but not total cAMP correlated to an increase in Ca²⁺ transient amplitude and decay kinetics.⁹⁹ In response to β-AR stimulation, approximately 45% of the total cAMP was found in the particulate fraction, but this fraction declined to less than 20% when IBMX was added to Iso, although cAMP production was up to threefold to fourfold greater. These results show that cAMP-PDEs reside predominantly in the cytoplasm, where they prevent excessive cAMP accumulation upon β-AR activation. Thus, PDEs appear to be important to maintain the specificity of the β-AR response by limiting the amount of cAMP diffusing from membrane to cytoplasm. Similar results were obtained when studying the effect of PDE inhibition on I_Ca,L