Physiological and pathological roles of protein kinase A in the heart

Abstract

Protein kinase A (PKA) is a central regulator of cardiac performance and morphology. Myocardial PKA activation is induced by a variety of hormones, neurotransmitters, and stress signals, most notably catecholamines secreted by the sympathetic nervous system. Catecholamines bind β-adrenergic receptors to stimulate cAMP-dependent PKA activation in cardiomyocytes. Elevated PKA activity enhances Ca²⁺ cycling and increases cardiac muscle contractility. Dynamic control of PKA is essential for cardiac homeostasis, as dysregulation of PKA signalling is associated with a broad range of heart diseases. Specifically, abnormal PKA activation or inactivation contributes to the pathogenesis of myocardial ischaemia, hypertrophy, heart failure, as well as diabetic, takotsubo, or anthracycline cardiomyopathies. PKA may also determine sex-dependent differences in contractile function and heart disease predisposition. Here, we describe the recent advances regarding the roles of PKA in cardiac physiology and pathology, highlighting previous study limitations and future research directions. Moreover, we discuss the therapeutic strategies and molecular mechanisms associated with cardiac PKA biology. In summary, PKA could serve as a promising drug target for cardioprotection. Depending on disease types and mechanisms, therapeutic intervention may require either inhibition or activation of PKA. Therefore, specific PKA inhibitors or activators may represent valuable drug candidates for the treatment of heart diseases.

1. Introduction of the PKA signalling system

Protein kinases play fundamental roles in nearly every aspect of cell biology and physiology, and thus their malfunctions are frequently associated with diseases. The 3',5'-cyclic adenosine monophosphate (cAMP)-dependent protein kinase [protein kinase A (PKA)], first discovered in 1968,¹ has been viewed as the prototype for all protein kinases.² PKA belongs to the AGC [named after the representative members PKA, 3′,5′-cyclic guanosine monophosphate (cGMP)-dependent protein kinase (PKG) and protein kinase C (PKC)] kinase family, which contains >60 serine/threonine protein kinases. PKA is evolutionarily conserved from fungi to humans and is ubiquitously expressed in all mammalian cell types.² The PKA holoenzyme, composed of two regulatory (R) and two catalytic (C) subunits, remains in its inactive state as a PKA tetramer (R₂C₂) in the absence of cAMP. The PKA regulatory (PKA-R) subunits have four isoforms, R1α, R1β, R2α, and R2β (encoded by PRKAR1A, PRKAR1B, PRKAR2A, and PRKAR2B, respectively), whereas three isoforms of the PKA catalytic (PKA-C) subunits have been identified, Cα, Cβ, and Cγ (encoded by PRKACA, PRKACB, and PRKACG, respectively).³ These isoforms differ in their expression patterns and levels and may exhibit alternative splicing. Depending on the R subunit, PKA holoenzyme is divided into type I and II PKA, which bind to R1 and R2, respectively. In general, R1α and Cα1 are the most abundant and ubiquitously expressed PKA-R and PKA-C subunits, respectively.³

The canonical PKA signalling pathway is essential for the cardiac actions of many hormones and neurotransmitters, particularly the catecholamines including norepinephrine secreted by cardiac sympathetic nerve terminals, and epinephrine released by the adrenal medulla (Figure 1).⁴ Catecholamines bind the transmembrane β-adrenergic receptors (β-ARs), major G-protein-coupled receptors in the heart, leading to release of the stimulatory G-protein α subunit (Gα_s) inside the target cells. Gα_s then binds and activates adenylyl cyclases (ACs), which convert ATP into cAMP, resulting in a rapid increase in intracellular cAMP levels. The cAMP molecules bind the PKA-R subunits to induce dissociation of the tetrameric PKA holoenzyme, leading to release of the free PKA-C subunits and subsequent PKA activation. PKA phosphorylates a plethora of substrates to regulate cellular activities. It is worth noting that holoenzyme dissociation may not be necessary for PKA activation under certain circumstances.⁵ PKA can also be activated through non-canonical pathways in a cAMP-independent manner (Figure 1), by a variety of stimuli including reactive oxygen species (ROS),^6–8 lipopolysaccharides (LPS)/interleukin-1 (IL-1),⁹ endothelin-1 (ET-1)/angiotensin II (Ang II),¹⁰ transforming growth factor-β (TGF-β),¹¹ sphingosine,¹² and peroxynitrite.¹³ For example, R1α can serve as a redox sensor and undergo oxidant-induced protein degradation, leading to PKA holoenzyme dissociation and kinase activation.^6–8 In addition, the PKA-C subunits are also sequestered by the inhibitor of κB (I-κB) proteins within the I-κB-PKA-C complex, and can be activated following I-κB degradation.^9,10

Figure 1.

PKA signalling pathway. The canonical cAMP-dependent PKA pathway is initiated by catecholamines, which stimulate β-adrenergic receptor (β-AR) to induce adenylyl cyclase (AC)-mediated cAMP synthesis. The second messenger cAMP then binds the PKA regulatory (PKA-R) subunits, leading to release of the PKA catalytic (PKA-C) subunits and subsequent kinase activation. The non-canonical pathways of PKA activation are cAMP-independent. For example, oxidative stress or TGF-β induces PKA activation through blocking PKA-R-mediated sequestration of PKA-C, whereas ET-1, Ang II, LPS, or IL-1 derepresses I-κB-mediated inhibition of PKA-C. On the other hand, cAMP also exerts its biological functions through direct binding with other effectors beyond PKA, including Epac, PRKX, PRKY, CNG ion channels, HCN ion channels, and POPDC proteins. Ang II, angiotensin II; cAMP, 3',5'-cyclic adenosine monophosphate; CNG, cyclic nucleotide-gated; Epac, exchange protein directly activated by cAMP; ET-1, endothelin-1; HCN, hyperpolarization-activated cyclic nucleotide-gated; I-κB, inhibitor of κB; IL-1, interleukin-1; LPS, lipopolysaccharides; PKA, cAMP-dependent protein kinase, or protein kinase A; PDE, cyclic nucleotide phosphodiesterase; PKI, a PKA inhibitor; POPDC, popeye domain containing; PRKX/Y, protein kinase X/Y-linked; TGF-β, transforming growth factor-β. Arrows indicate activation; bar-headed lines indicate inhibition.

Cellular cAMP levels are determined by the balance between the enzymatic activities of ACs and cyclic nucleotide phosphodiesterases (PDEs). In contrast to ACs, PDEs hydrolyze cAMP to 5′-AMP, thereby reducing cAMP levels and resulting in a decrease in PKA activity (Figure 1). In addition to cAMP, some PDEs also hydrolyze cGMP.¹⁴ PKA activity is inhibited by the endogenous PKI protein, a pseudosubstrate of PKA that is capable of sequestering and inactivating the PKA-C subunits.² Unlike PKA-R, the interaction between PKI and PKA-C is not modulated by the levels of cAMP. A-kinase anchoring proteins (AKAPs) tether PKA holoenzyme to certain subcellular locations, allowing for rapid and precise control of its substrates within the specific compartments.¹⁵ For example, short AKAP7 (also known as AKAP15) anchors PKA at the plasma membrane to regulate L-type Ca²⁺ channel (LTCC, Ca_v1.2) activity, whereas muscle AKAP (mAKAP, AKAP6) and long AKAP7 (AKAP18) tether PKA at the sarcoplasmic reticulum (SR) to modulate the functions of ryanodine receptors (RyRs) and sarco/endoplasmic reticulum Ca²⁺-ATPase (SERCA), respectively.¹⁵

PKA is a key regulator of cardiac function, structure, and remodelling. Recent studies have uncovered novel roles of PKA in normal and diseased heart, and have provided new insights on the underlying mechanisms. Here, we review our current knowledge of PKA in cardiac physiology and pathophysiology.

2. PKA and cardiac contraction/relaxation

Cardiac muscle contraction and relaxation determine systolic and diastolic functions of the heart. Contraction is initiated by electrical excitation (i.e. membrane depolarization through the action potential), a process termed excitation–contraction coupling.¹⁶ Depolarization of the sarcolemma induces cellular Ca²⁺ entry through LTCC. Elevated cytosolic Ca²⁺ level triggers Ca²⁺ release from SR through RyR2 channel, further raising intracellular Ca²⁺ concentration. Ca²⁺ binds to cardiac troponin C (cTnC), relieves cardiac troponin I (cTnI)-mediated inhibition of actin-myosin filament interaction, leading to contraction. During cardiac relaxation, Ca²⁺ is dissociated from cTnC, transported back into the SR via SERCA, and to the extracellular space via the Na⁺/Ca²⁺ exchanger (NCX) as well as the sarcolemmal Ca²⁺-ATPase.

The force (inotropy), rate (chronotropy) of cardiac contraction, and the ability of relaxation (lusitropy) are modulated by the sympathetic (or adrenergic) nervous system through secretion of catecholamines. In the fight-or-flight response, rapid release of catecholamines markedly increases heart rate and contractility through β-AR-dependent activation of the cAMP/PKA signalling pathway. Particularly, catecholamines may bind β-ARs at the sarcolemma and the SR to induce local PKA activation.¹⁷ PKA, in turn, phosphorylates target proteins involved in the regulation of contraction/relaxation, including Ca_v1.2, RyR2, phospholamban (PLN), cTnI, and cardiac myosin-binding protein C (cMyBP-C).¹⁶

PKA positively regulates inotropy through phosphorylation of Ca_v1.2, PLN, and cMyBP-C. Phosphorylation of Ca_v1.2 or PLN increases Ca²⁺ availability, whereas phosphorylation of cMyBP-C improves responsiveness of the contractile apparatus to Ca²⁺. In response to β-adrenergic stimulation, PKA phosphorylates Ca_v1.2 at S1700, resulting in increased LTCC activity and Ca²⁺ influx, thereby enhancing contraction.¹⁸ The above model is supported by studies in the knock-in mice expressing a non-phosphorylatable mutant of endogenous Ca_v1.2 (S1700A), which show decreased basal and β-adrenergic-stimulated calcium currents, as well as reduced cardiomyocyte contractility.¹⁸ Surprisingly, overexpression of the same Ca_v1.2 mutant (S1700A) using a transgenic approach does not interfere with Ca²⁺ current, arguing against a role for S1700 phosphorylation.¹⁹ A most recent study further reveals that PKA augments Ca_v1.2 channel activity indirectly, through phosphorylation and depletion of the Ca_v1.2 inhibitor Rad.²⁰ Despite the ongoing debate about the mechanisms, there is consensus that LTCC is essential for PKA-dependent Ca²⁺ influx and cardiac contraction (Table 1). PKA-mediated phosphorylation of PLN at S16 relieves PLN-dependent inhibition of SERCA, resulting in increased SR Ca²⁺ uptake from the cytosol and consequently increased SR Ca²⁺ load.²⁵ Therefore, higher levels of Ca²⁺ can be released from SR for binding to the contractile proteins, leading to more forceful contraction. The physiological significance of PKA-dependent RyR2 phosphorylation has been controversial. Although there is evidence that phosphorylation of RyR2 at S2808 by PKA is necessary for β-adrenergic-induced SR Ca²⁺ release and cardiac contraction, contradictory findings have been reported.^30,31 The discrepancy may be due to the fact that either maximal or minimal RyR2 phosphorylation by PKA favours SR Ca²⁺ leak.³² Phosphorylation of cMyBP-C at S273, S282, or S302 by multiple kinases including PKA increases myocardial contractility through weakening cMyBP-C-mediated inhibition of myosin,³⁴ increasing force-producing myosin heads,³⁵ or accelerating cross-bridge cycling,³⁶ respectively. Thus PKA-mediated cMyBP-C phosphorylation is also a key regulator of cardiac inotropy.

Table 1.

Myocardial PKA substrates and their biological functions

Substrate	Phosphorylation site(s)	Molecular/cellular function	Physiological/pathological function
Cav1.2	Ser1700	Cav1.2 activation, Ca2+ influx, Ca2+ overload, necrosis	Inotropy,18,19 chronotropy,21,22 heart failure23,24
Rad	Ser25, Ser38, Ser272, Ser300	Cav1.2 activation, Ca2+ influx	Inotropy20
PLN	Ser16	SERCA activation, SR Ca2+ uptake	Inotropy,25 chronotropy,21,22,26,27 lusitropy,25 heart failure28,29
RyR2	Ser2808	SR Ca2+ release, Ca2+ overload, necrosis	Inotropy,30–32 heart failure28,29,33
cMyBP-C	Ser273, Ser282, Ser302	Myosin activation, actin-myosin cross-bridge detachment	Inotropy,34–36 lusitropy37,38
cTnI	Ser23, Ser24	Ca2+-cTnC dissociation	Lusitropy39–41
CcO		CcO activity↓, ROS generation, necrosis	I/R injury42
HSP20	Ser16	Autophagy↑, apoptosis↓, necrosis↓	I/R injury↓43
eNOS	Ser633, Ser1177	eNOS activation	I/R injury↓44,45
CREB	Ser133	CREB-mediated transcription of genes involved in hypertrophy (foetal genes) and apoptosis (ICER,46,47 Bim,48 Bcl-249)	Hypertrophy,49,50 I/R injury→51
GSK-3β	Ser9	NFAT-mediated transcription of hypertrophic genes↑	Hypertrophy52,53
NFAT	Ser245, Ser269, Ser294	NFAT-mediated transcription of hypertrophic genes↓	Hypertrophy↓54–56
Drp1	Ser637	Mitochondrial fission↓	Hypertrophy↓57
HDAC5	Ser279	MEF2-mediated transcription of hypertrophic genes↓	Hypertrophy↓58,59
HDAC4		MEF2-mediated transcription of hypertrophic genes↓	Hypertrophy↓60
Titin, N2B element	Ser4185, Ser4010	Titin compliance↑, cardiomyocyte stiffness↓	Diastolic function↑, HFpEF↓61–64
RPN6	Ser14	Proteasome activation, clearance of misfolded proteins↑	Proteinopathy-related HFpEF↓65

CcO, cytochrome c oxidase; cMyBP-C, cardiac myosin-binding protein C; CREB, cAMP-response element binding protein; cTnC, cardiac troponin C; cTnI, cardiac troponin I; eNOS, endothelial nitric oxide synthase; GSK-3β, glycogen synthase kinase-3β; HDAC4/5, histone deacetylase 4/5; HFpEF, heart failure with preserved ejection fraction; HSP20, heat shock protein 20; ICER, inducible cAMP early repressor; I/R, ischaemia/reperfusion; MEF2, myocyte enhancer factor 2; NFAT, nuclear factor of activated T cells; PLN, phospholamban; ROS, reactive oxygen species; RyR2, ryanodine receptor 2; SERCA, sarco/endoplasmic reticulum Ca²⁺-ATPase; SR, sarcoplasmic reticulum; ↑, increase; ↓, decrease; →, no change.

PKA drives cardiac chronotropy primarily through phosphorylation of SR Ca²⁺ cycling proteins in cardiac pacemaker cells (i.e. sinoatrial nodal cells, a group of specialized cardiac myocytes).²¹ Pacemaker cells are unique in their high basal cAMP level and PKA activity, which are necessary and sufficient for the generation of rhythmic internal Ca²⁺ store oscillations and spontaneous beating, even in the absence of β-adrenergic stimulation. In response to β-AR stimulation, however, PKA-mediated phosphorylation of PLN at S16 speeds up SR Ca²⁺ uptake and contributes to the acceleration of heart rate.²¹ PKA-dependent generation of rhythmic action potentials also requires coupling between spontaneous local Ca²⁺ release and the membrane ion channels including NCX and hyperpolarization-activated cyclic nucleotide-gated (HCN) channels.⁶⁶ Cardiac-specific overexpression of AC8 increases AC activity in the sinoatrial node tissue, which is accompanied by a marked increase in heart rate.⁶⁷ Pharmacological inhibition of PDE3A and PDE4B, the major PDE subtypes expressed in pacemaker cells, synergistically increase the spontaneous beating rate.²⁶ PDE2 inhibition with Bay 60–7550 also accelerates heart rate.²² Conversely, cardiac-specific PDE2 overexpression results in a decrease in phospho-PLN(S16) level, which is associated with reduced resting and maximal heart rate.²² Cardiac-specific PKI transgenic mice show lower PKA activity, reduced PLN phosphorylation at S16, and slower heart rate following β-AR stimulation.²⁷ Intriguingly, basal contractility is enhanced in the transgenic heart overexpressing either PDE2 or PKI, likely due to reduced SR Ca²⁺ leak, increased myocyte Ca²⁺ transient and enhanced myofilament Ca²⁺ sensitivity.^22,27 Aberrant PKA activation can cause Ca²⁺ cycling protein dysfunction and lead to cardiac arrhythmias. For example, overexpression of PDE2 decreases arrhythmia occurrence following myocardial infarction,²² but PDE4D gene deficiency induces PKA-dependent RyR2 hyperphosphorylation, which enhances SR Ca²⁺ leak and promotes arrhythmias.⁶⁸

PKA enhances cardiac lusitropy through phosphorylation of PLN, cTnI, and cMyBP-C. Myocardial relaxation depends on efficient cytosolic Ca²⁺ removal. Phosphorylation of PLN by PKA accelerates SERCA-mediated transport of cytosolic Ca²⁺ into SR, resulting in faster relaxation.²⁵ PKA-dependent phosphorylation of cTnI and cMyBP-C augments lusitropy primarily through reducing Ca²⁺ sensitivity. The N-terminal region of cTnI binds cTnC to increase its Ca²⁺ affinity, which determines myofibrillar Ca²⁺ sensitivity and contractile activation.³⁹ Phosphorylation of cTnI at the N-terminal residues S23/24 by PKA weakens cTnI–cTnC interaction, thereby promoting dissociation of Ca²⁺ from cTnC and accelerating myofibril relaxation.⁴⁰ Importantly, monophosphorylation at S24 is sufficient to reduce Ca²⁺ sensitivity.⁴¹ Phosphorylation of cMyBP-C also represents a major lusitropic mechanism during β-adrenergic stimulation. Concurrent non-phosphorylatable mutations of all PKA target sites on cMyBP-C (S273A, S282A, and S302A) impair diastolic function without altering intracellular Ca²⁺ handling, suggesting that cMyBP-C phosphorylation enhances relaxation likely through increasing the rates of actin-myosin cross-bridge detachment.^37,38 Conversely, phosphomimetic mutations of all three serine residues of cMyBP-C (S273D, S282D, and S302D) increase peak myocardial relaxation velocity, indicating enhanced lusitropy.³⁷ However, the specific role of each individual site remains to be determined.

Although required by the positive inotropic, chronotropic, and lusitropic effects of β-adrenergic stimulation, PKA activity is not necessary for the maintenance of basal heart function.²⁷ Thus, pharmacological PKA inhibition might be a safe and viable strategy for clinical use. Since PKA inhibition with PKI preferentially reduces PKA activity in the SR and myofilament but not the sarcolemma and nuclei, it is possible that basal heart function is maintained by sarcolemmal or nuclear PKA. Nonetheless, long-term PKA inhibition reduces cardiac reserve and impairs exercise capability.²⁷

In summary, PKA plays an essential role in neurohormonal regulation of cardiac function, through phosphorylation of its substrates including Ca_v1.2, RyR2, PLN, cTnI, and cMyBP-C. Although loss of PKA activity does not diminish cardiac function at resting conditions, it can limit the heart’s ability to cope with stress.

3. PKA and ischaemic heart disease

Ischaemic heart disease, also known as coronary artery disease, occurs when blood flow to the heart muscle is reduced or blocked, most commonly due to atherosclerosis. The biological functions of PKA in atherosclerosis pathogenesis have been described elsewhere.⁶⁹ This review will summarize our current knowledge of PKA in myocardial ischaemia, with a particular focus on cardiac myocytes. Myocardial ischaemia is characterized by tissue hypoxia (i.e. oxygen deprivation due to insufficient blood supply). Restoration of blood supply by reperfusion results in tissue re-oxygenation but paradoxically also causes ischaemia/reperfusion (I/R) injury. During myocardial ischaemia, interstitial catecholamine concentration dramatically increases by 100–1000 fold (Figure 2),^70,71 due to hyperactivation of the sympathetic nervous system.⁷² Accordingly, both cAMP level^73,74 and PKA activity^75,76 are increased in the ischaemic myocardium. Following reperfusion, myocardial catecholamine level rapidly declines and returns to baseline within 120 min.^70,71 However, cardiac PKA activity remains elevated after reperfusion,^76,77 due at least in part to R1α loss as a result of oxidation and dimerization.^8,77

Figure 2.

PKA signalling in myocardial ischaemia/reperfusion injury. Both ischaemia and reperfusion induce PKA activation, which can contribute to oxidative stress and Ca²⁺ overload, leading to cardiac cell death and myocardial damage. However, cAMP-induced activation of Epac, eNOS, proteasome, and HSP20 appear to be cardioprotective. CcO, cytochrome c oxidase; ECs, endothelial cells; eNOS, endothelial nitric oxide synthase; Epac, exchange protein directly activated by cAMP; HSP20, heat shock protein 20; MnSOD, manganese superoxide dismutase; ROS, reactive oxygen species. Arrows indicate activation; bar-headed lines indicate inhibition.

At physiological levels, catecholamines mediate the fight-or-flight response to rapidly increase contractility and heart rate. At extremely high concentrations, however, catecholamines can cause myocardial tissue damage. Administration of a single, high dose of the synthetic catecholamine isoproterenol in rats induces PKA-dependent RyR2 hyperphosphorylation, leading to SR Ca²⁺ leakage and subsequent myocyte death via apoptosis or necrosis (Figure 2).³³ Moreover, isoproterenol induces PKA-mediated phosphorylation of cAMP-response element binding protein (CREB) at S133, leading to enhanced binding of CREB with the cAMP-response element in the promoter region, and subsequent transcription of pro-apoptotic genes including inducible cAMP early repressor (ICER)^46,47 and Bim.⁴⁸ It is noteworthy that CREB deficiency has no impact on apoptosis following I/R injury,⁵¹ likely because CREB also mediates transcription of the anti-apoptotic gene Bcl-2 in certain context.⁴⁹ In addition, acute β-adrenergic stimulation with isoproterenol in mice transiently increases myocyte membrane permeability, an important marker of necrosis.⁷⁸ Intriguingly, acute catecholamine injury appears to be fully reversible.^33,78 By contrast, sustained activation of the PKA substrate LTCC in mice results in cardiac dysfunction and premature death, which is exacerbated by isoproterenol infusion.²³ Mechanistically, isoproterenol exposure augments LTCC activity and enhances Ca²⁺ influx, leading to Ca²⁺ overload and myocyte necrosis.²³

Activation of β-AR/cAMP signalling during the ischaemic phase contributes to myocardial I/R injury. Overexpression of PDE3A1 reduces myocardial cAMP levels and attenuates I/R-induced myocyte apoptosis, possibly due to decreased ICER and increased Bcl-2 expression.⁷⁹ Moreover, treatment with the β₁-AR inhibitor CGP-20712A⁸⁰ or the PKA inhibitor H89/PKI ^42,81 diminishes I/R-induced myocardial necrosis. In the ischaemic heart, excessive cAMP induces PKA-dependent phosphorylation and inhibition of cytochrome c oxidase (CcO), resulting in augmented ROS production.⁴² In turn, oxidative stress prompts R1α loss to activate PKA,^8,77 leading to further CcO inhibition and ROS generation.⁷ Therefore, a rapid rise in cAMP level may trigger a vicious cycle of oxidative stress through activation of PKA (Figure 2).

Reduced cAMP level during ischaemia, caused by ischaemic preconditioning, is associated with marked protection against myocardial I/R injury.^73,74 Interestingly, ischaemic preconditioning initially increases cAMP level during intermittent ischaemia (i.e. the preconditioning phase), but later reduces cAMP accumulation during sustained ischaemia (i.e. the main ischaemia phase) possibly due to β-AR desensitization or compensatory activation of PDEs.^73,74 Myocardial I/R injury is attenuated by pre-ischaemic cAMP up-regulation using isoproterenol (a β-AR agonist), forskolin (an AC agonist),^74,81–83 milrinone/olprinone/amrinone (PDE3 inhibitors),^84,85 trapidil (a non-selective PDE inhibitor),⁸⁶ or dibutyryl-cAMP (a cell-permeable cAMP analogue).^75,84 Germline ablation of PDE3B increases cardiac cAMP level at baseline, reduces infarct size, and improves cardiac function following I/R.⁸⁷ The cardioprotective effect of PDE3B ablation is blocked by treatment with KT5720, a PKA inhibitor.⁸⁷ Intramyocardial injection of PKA-Cα siRNAs before ischaemia increases infarct size and impairs systolic function after I/R, possibly due to NF-κB activation and superoxide production.⁸⁸ Together these studies suggest that pre-ischaemic cAMP/PKA activation represses PKA activity during ischaemia, thereby leading to cardioprotection.

Emerging evidence suggests that PKA could be a better drug target than β-AR, because PKI is more effective than the β₁-AR blocker metoprolol in alleviating myocyte apoptosis and cardiac dysfunction following myocardial infarction.⁸⁹ Mechanistically, β-AR and cAMP induce PKA activation to cause cardiotoxicity but stimulate exchange protein directly activated by cAMP (Epac) to mediate cardioprotection (Figure 2).⁸⁹ In this regard, cardiac cell death following simulated I/R is exaggerated by inhibition of type 10 soluble AC (sAC) with gene silencing, KH7 treatment, or bicarbonate withdrawal, but mitigated by overexpression of sAC or treatment with the PDE2 inhibitor Bay 60–7550.⁹⁰ Moreover, incubation with the cAMP inhibitor Rp-cAMP abolishes glucagon-like peptide 1 (GLP-1)-induced protection against I/R Injury.⁹¹ Interestingly, the beneficial effects of GLP-1 are mediated by sAC-dependent cAMP/PKA activation, and subsequent endothelial nitric oxide synthase (eNOS) phosphorylation.^44,45 PKA has also been shown to protect against I/R injury through proteasome activation,⁹² as well as phosphorylation of heat shock protein 20.⁴³ Therefore, although believed to be detrimental during myocardial I/R injury, cAMP and PKA may also activate certain protective molecules in some contexts.

In summary, the majority of studies suggest that activation of PKA during ischaemia contributes to I/R injury through aggravating cardiomyocyte apoptosis and/or necrosis. Ischaemic preconditioning reduces PKA activity during ischaemia and thus attenuates I/R injury. Notably, elevated cAMP levels during ischaemia may also activate protective mechanisms in a PKA-dependent or independent manner.

4. PKA and cardiac hypertrophy

It is well established that chronic stimulation of β₁-AR induces cardiac hypertrophy.⁴ However, whether PKA contributes to β₁-AR-mediated hypertrophy remains elusive. Earlier studies reveal that cardiac-specific overexpression of β₁-AR,⁹³ Gα_s,⁹⁴ or PKA-Cα,²⁸ markedly increases cardiomyocyte size, but only moderately increases heart weight due to concurrent apoptosis or necrosis. Notably, β-adrenergic stimulation is sufficient to cause apoptosis in adult cardiomyocytes, in a PKA-dependent manner.⁹⁵ Overexpression of type 5 AC (AC5), a major cardiac isoform, exacerbates apoptosis and only marginally increases cardiomyocyte size following chronic infusion of isoproterenol.⁹⁶ Mechanistically, AC5-mediated cAMP/PKA activation augments oxidative stress through repression of manganese superoxide dismutase (MnSOD) expression (Figure 2).⁹⁶ Cardiac-specific transgenic expression of AC6, another major cardiac isoform, does not alter heart weight in mice with ischaemic cardiomyopathy.⁹⁷ Moreover, overexpression of AC8 fails to increase heart weight, despite four-fold higher PKA activity.⁹⁸ These studies suggest that PKA activation may primarily provoke myocyte loss, which can eventually lead to compensatory hypertrophy. Given the essential role of PKA in regulating cardiac contractility, it is not surprising that interfering with PKA-dependent phosphorylation can disrupt normal heart function, which may also result in hypertrophic remodelling to compensate for reduced cardiac output. Indeed, mice with non-phosphorylatable mutations of Ca_v1.2,^18,24 PLN,⁹⁹ cTnI,¹⁰⁰ or cMyBP-C^37,38 at their PKA sites exhibit cardiac dysfunction and hypertrophy. However, myocardial PKA inactivation by overexpression of PKI does not induce cardiac dysfunction or hypertrophy in mice up to 12 months of age.²⁷ Germline deletion of AC5 in mice does not blunt physiological hypertrophy during postnatal heart development (up to 3–6 months of age),¹⁰¹ or attenuate pathological hypertrophy induced by pressure overload.¹⁰² Again, disruption of AC5 primarily suppresses apoptosis, due in part to up-regulation of the anti-oxidant protein MnSOD and the anti-apoptotic molecule Bcl-2.^101,102 Collectively, the above in vivo studies do not support a major, direct role of the AC/cAMP/PKA signalling in hypertrophy.

Myocardial hypertrophic growth is not associated with a change in total PKA activity.¹⁰³ However, in vitro real-time imaging reveals that hypertrophy is accompanied by relocation of PKA and PDE activity within myocytes, resulting in altered cAMP/PKA compartmentalization.¹⁰⁴ Upon induction of hypertrophy, expression, and activity of several cAMP-hydrolyzing PDEs, including the pro-hypertrophic PDE1C,¹⁰⁵ PDE2,⁵⁴ and PDE10A¹⁰⁶ are increased, whereas PDE3 and PDE4 are down-regulated.^107,108 Interestingly, PDE3 or PDE4 appears to be anti-hypertrophic as PDE3/4 inhibition causes spontaneous hypertrophy,⁵⁴ rendering myocytes insensitive to catecholamine stimulation in vitro⁵⁴ or pressure overload in vivo.¹⁰⁹ Mechanistically, PDE4 suppresses cardiomyocyte hypertrophy through inhibition of nuclear PKA activity.^47,50 Since PDEs differ in their subcellular distribution,¹⁴ their opposing roles in hypertrophy could be due to differential modulation of PKA activity within specific subcellular compartments.

Indeed, nuclear PKA activation is associated with augmented cardiac hypertrophy (Figure 3). Overexpression of nuclear-targeted PKA-C increases cardiomyocyte size, whereas overexpression of cytosolic-targeted PKA-C enhances Ca²⁺ transients and cardiac contractile force without inducing hypertrophy.¹¹⁰ β-Adrenergic stimulation or pressure overload-induced hypertrophy requires sAC, which is known to mediate nuclear cAMP synthesis.¹¹¹ Interestingly, activation of β₁-AR increases nuclear PKA activity,⁴⁷ and results in hypertrophy.⁹³ By contrast, stimulation of β₂-AR does not induce nuclear PKA activation,⁴⁷ and fails to cause hypertrophy.¹¹² Mechanistically, PKA phosphorylates CREB at S133 in the nuclei of cardiomyocytes, thereby initiating CREB-mediated transcription of hypertrophy-related genes.^49,50 Since β₁-AR-mediated nuclear PKA activation is abolished by PKI,⁴⁷ cardiac-specific PKI transgenic mice are resistant to isoproterenol-induced CREB phosphorylation at S133²⁷ and subsequent pathological hypertrophy.⁸⁹ Moreover, overexpression of a non-phosphorylatable CREB mutant (S133A) attenuates isoproterenol-induced hypertrophy.⁴⁶ Collectively, these studies suggest that CREB-mediated transcription is a critical pro-hypertrophic mechanism downstream of nuclear PKA. β-AR-mediated nuclear PKA activation requires the scaffold protein mAKAP.⁴⁷ Interestingly, mAKAP facilitates calcineurin-dependent activation of the pro-hypertrophic transcription factor nuclear factor of activated T cells (NFAT), which binds GATA4 or myocyte enhancer factor 2 (MEF2) to stimulate expression of hypertrophy-related genes.¹¹³ Glycogen synthase kinase-3β (GSK-3β), which phosphorylates NFAT to trigger its nuclear export and inactivation,⁵² is repressed by PKA via PKA-dependent phosphorylation at S9.⁵³ Therefore, nuclear PKA may promote hypertrophy in a synergistic manner with calcineurin, through relieving GSK-3β-mediated inhibition of NFAT (Figure 3). In line with these findings, global PKA-Cβ null mice are resistant to Ang II-induced cardiac hypertrophy.¹¹⁴ Intriguingly, overexpression of PKA-Cα in cardiac fibroblasts also induces ventricular dilation and cardiomyocyte hypertrophy, possibly via a paracrine mechanism.¹¹⁵ Taken together, PKA mediates cardiac hypertrophy while localized in the nuclei of cardiomyocytes or in cardiac fibroblasts.

Figure 3.

PKA signalling in cardiac hypertrophy. Key molecular events contributing to cardiac hypertrophy include MEF2-, CREB-, NFAT-mediated gene transcription, or Drp1-mediated mitochondrial fission. Gα_s stimulates synthesis of the second messenger cAMP, which activates nuclear PKA to induce hypertrophy and cytoplasmic PKA to inhibit hypertrophy. Nuclear PKA promotes hypertrophy through: (i) activating CREB; and (ii) alleviating GSK-3β-mediated NFAT repression. Cytoplasmic PKA inhibits hypertrophy through: (i) invoking HDAC4/5-mediated MEF2 repression; (ii) inactivating NFAT; and (iii) suppressing Drp1. By contrast, Gα_q acts through the second messenger Ca²⁺ to provoke Ca²⁺/CaMKII- and calcineurin-mediated hypertrophy. CaMKII activates MEF2 and CREB, whereas calcineurin activates NFAT and Drp1. Nuclear PKA may act cooperatively with CaMKII and calcineurin to induce hypertrophy, but cytoplasmic PKA antagonizes CaMKII- and calcineurin-dependent hypertrophy. AC, adenylyl cyclase; CaMKII, Ca²⁺/calmodulin-dependent protein kinase type II; CREB, cAMP-response element binding protein; GSK-3β, glycogen synthase kinase-3β; HDAC4/5, histone deacetylase 4/5; MEF2, myocyte enhancer factor 2; NFAT, nuclear factor of activated T cells; PLC, phospholipase C. Arrows indicate activation; bar-headed lines indicate inhibition.

In contrast, accumulating evidence suggests that cytoplasmic PKA may inhibit hypertrophy. In response to PDE2 inhibition, an increased local pool of cAMP induces PKA-dependent phosphorylation of NFAT at S245/S269/S294, resulting in cytoplasmic retention of NFAT and inhibition of hypertrophy.⁵⁴ Cytoplasmic NFAT retention requires mitochondrial localization of PKA, as disruption of AKAP1 (also known as AKAP121, which targets PKA to mitochondria) reduces mitochondrial PKA activity, leading to augmented nuclear NFAT accumulation and aggravated hypertrophy both in vitro and in vivo.^55,56 These results suggest that cytoplasmic PKA may directly repress NFAT (Figure 3). Since cytoplasmic PKA activity is restricted by R1α,¹¹⁶ we recently delete the R1α-encoding gene PRKAR1A in mouse heart to assess the impact of cytoplasmic PKA activation on physiological cardiac hypertrophy during development.⁵⁷ Because homozygous PRKAR1A ablation inhibits cardiomyocyte proliferation and results in embryonic lethality,¹¹⁷ we use a heterozygous deletion approach and show that PRKAR1A deficiency suppresses hypertrophic heart growth, likely through inhibition of mitochondrial fission via PKA‐dependent Drp1 phosphorylation at S637 (Figure 3). Importantly, PRKAR1A mutations/deletions in humans are also associated with reduced left ventricular mass.⁵⁷ In addition, germline PRKAR2B ablation attenuates age-related cardiac hypertrophy, possibly through a non-cardiomyocyte-autonomous mechanism because R2β is predominantly expressed in non-cardiac tissues.¹¹⁸

Although stimulation of either Gα_s-coupled receptors (e.g. β₁-AR) or Gα_q-coupled receptors (e.g. α₁-AR, Ang II type 1 receptor, ET A receptor) can lead to cardiomyocyte hypertrophy, Gα_s signalling appears to antagonize Gα_q-mediated hypertrophy. For example, treatment with forskolin (an AC agonist),⁵⁷ or cAMP⁵⁸ attenuates α₁-AR-mediated hypertrophy in vitro. Moreover, Gα_q-mediated cardiac foetal gene expression, a common hypertrophy marker, is blocked by isoproterenol, forskolin, or a selective PKA activator 8-CPT-6-Phe-cAMP, but augmented by a PKA inhibitor H89.¹¹⁹ Mechanistically, Gα_s-dependent PKA activation induces histone deacetylase 5 (HDAC5) phosphorylation at S279 ^58,59 or dephosphorylation at S259/S498,¹²⁰ resulting in translocation of HDAC5 from the cytoplasm to the nucleus, where it represses MEF2-mediated transcription of the hypertrophic gene program (Figure 3). PKA also induces nuclear import of HDAC4 and subsequent MEF2 repression.⁶⁰ By contrast, Gα_q stimulates phospholipase C to increase intracellular Ca²⁺ level, leading to activation of Ca²⁺/calmodulin-dependent protein kinase type II (CaMKII) and calcineurin (i.e. protein phosphatase 2B). Intriguingly, CaMKII induces HDAC4 nuclear export to provoke MEF2-mediated pathological cardiac hypertrophy.¹²¹ On the other hand, calcineurin dephosphorylates NFAT at S245/S269/S294 and Drp1 at S637, the same sites phosphorylated by cytoplasmic PKA.^122,123 Therefore, cytoplasmic PKA may counteract CaMKII- and calcineurin-mediated hypertrophy (Figure 3). It is worth noting that CaMKII also phosphorylates CREB at S133,¹²⁴ suggesting that CaMKII and nuclear PKA may promote CREB-dependent hypertrophy in a redundant manner.

In summary, PKA differentially modulates cardiac hypertrophy depending on its subcellular localization. Nuclear PKA induces cardiomyocyte hypertrophy through activation of CREB and NFAT-mediated transcription of hypertrophic genes, whereas cytoplasmic PKA inhibits hypertrophy via repression of MEF2 and NFAT-mediated transcription, as well as inhibition of Drp1-mediated mitochondrial fission. In cardiomyocytes, Ca²⁺ and cAMP are major second messengers downstream of Gα_q and Gα_s, respectively. Ca²⁺-dependent activation of CaMKII or calcineurin provokes hypertrophy, which is likely augmented by cAMP-dependent activation of nuclear PKA, but attenuated by cytoplasmic PKA. These points need to be further investigated with approaches that specifically target PKA in various intracellular compartments.

5. PKA and heart failure

Heart failure occurs when the heart is unable to supply adequate blood to meet the metabolic needs of the body. Decreased cardiac output induces compensatory activation of the sympathetic nervous system, which increases circulating catecholamines in an attempt to restore heart function. Although acute neurohormonal activation maintains cardiac output, chronic β-AR stimulation impairs myocardial β-AR responsiveness through receptor down-regulation (via β-arrestin-mediated internalization) and desensitization (via uncoupling from G-proteins).⁴ Decreased β-AR responsiveness blunts catecholamine-induced PKA activation, thereby reducing myocardial contractility. Thus β-AR antagonism with β-blockers has been a mainstay of heart failure therapy.

Human heart failure is associated with diminished phosphorylation of PLN,¹²⁵ cTnI,^126–128 and cMyBP-C.^128,129 Although the cause of discrepancy remains an unsettled debate,¹³⁰ animal models of heart failure exhibit increased cTnI phosphorylation.^131,132 In failing rabbit cardiomyocytes, β-AR-induced PKA activation is diminished at the SR and sarcolemma, but enhanced at the myofilaments, indicating intracellular PKA activity redistribution.¹³² Accordingly, PKA-mediated phosphorylation of the SR protein PLN is decreased, but phosphorylation of the myofilament protein cTnI is increased in failing myocytes. Myofilament PKA activation in heart failure is due to reduced local PDE activity and enhanced β₂-AR signalling. Mechanistically, heart failure is associated with a redistribution of β₂-AR and PKA from t-tubules to other sarcolemmal areas owing to reduced expression of caveolin-3, a major structural protein of caveolae.^132–134 It is worth noting that total PKA activity following cAMP stimulation is comparable between failing and non-failing human hearts, suggesting that heart failure alters PKA signalling primarily at the level of β-AR.¹³⁵ Nonetheless, failing human hearts express lower level of R1α,^126,136 but opposite findings have also been reported.¹²⁸

Proper functioning of the heart relies on efficient Ca²⁺ cycling, which is controlled by PKA-dependent phosphorylation of the Ca²⁺ handling proteins in a highly dynamic manner. Therefore, prolonged activation or inhibition of PKA as well as its substrates can cause heart failure. For example, either gain-^99,137 or loss-of-function PLN mutations¹³⁸ induce lethal dilated cardiomyopathy. Chronic LTCC activation provokes Ca²⁺ overload-mediated cardiomyocyte necrosis, eventually leading to heart failure.²³ On the other hand, permanent LTCC inactivation also results in heart failure and premature death.²⁴ Constitutive PKA activation induces PLN and RyR2 hyperphosphorylation, leading to reduced contractility, dilated cardiomyopathy and sudden death.²⁸ Although persistent PKA inhibition in mice does not cause functional or morphological abnormalities, depressed PKA activity impairs cardiac adaptation to stress, and may contribute to heart failure morbidity.²⁷

The role of PKA in heart failure is modulated by AKAPs. Failing human hearts exhibit decreased interaction between PKA and AKAP1.¹³⁶ Ablation of AKAP1 impairs mitochondrial PKA signalling and accelerates heart failure development after pressure overload.⁵⁵ In addition, heart failure progression is accompanied by myocardial AKAP5 (also known as AKAP150, or AKAP79) down-regulation.²⁹ Loss of AKAP5 abolishes PKA-mediated phosphorylation of PLN and RyR2, disrupts Ca²⁺ cycling, and predisposes to heart failure.²⁹

Heart failure differentially affects expression of the PDE family members. Human heart failure is associated with decreased levels of PDE3A and PDE4D, which are sufficient to provoke SR Ca²⁺ leak and cardiomyocyte apoptosis.^68,107 Interestingly, failing human hearts express higher levels of PDE1C,¹⁰⁵ PDE2,¹³⁹ and PDE10A,¹⁰⁶ which hydrolyze cAMP and may contribute to decreased contractile response. Inhibition of PDE1C augments adenosine A₂ receptor (A₂R)-dependent cAMP synthesis, thereby providing acute inotropic, lusitropic, and vasodilatory effects in tachypacing-induced heart failure.¹⁴⁰ Similarly, inhibition of PDE2¹⁴¹ or PDE10A¹⁰⁶ protects against pressure overload-induced heart failure. Since inhibition of PDE1C,¹⁰⁵ PDE2,¹⁴¹ or PDE10A¹⁰⁶ also increases cardiac cGMP level, their cardioprotective effects are likely mediated by both cAMP-dependent and cAMP-independent mechanisms.

Conventional animal models of heart failure, such as pressure/volume overload or ischaemic/toxic injury, typically develop heart failure with reduced ejection fraction (HFrEF).¹⁴² The most common form of heart failure, however, is heart failure with preserved ejection fraction (HFpEF), which is characterized by myocardial stiffness and diastolic dysfunction. At the molecular level, myofilament stiffness is associated with reduced phosphorylation of titin’s N2B element at the PKA/PKG sites S4185/S4010/S4099.^61,62 Stimulation with PKA-C reduces passive stiffness of human HFpEF cardiomyocytes in vitro.⁶³ In an HFpEF-like mouse model, administration of metformin augments N2B phosphorylation at the PKA-specific site S4010, leading to increased titin compliance and improved diastolic function.⁶⁴ In another HFpEF-like model, inhibition of PDE1 enhances PKA-mediated phosphorylation of RPN6 at S14, which augments proteasomal degradation of misfolded proteins and improves diastolic function.⁶⁵ A new animal model that recapitulates many features of human HFpEF has been developed recently,¹⁴³ and will likely advance our knowledge of PKA in HFpEF pathogenesis.

In summary, the dynamic, nearly instantaneous control of cardiac contractility by PKA is compromised in heart failure. Although acute PKA activation improves cardiac performance, chronic PKA activation or inhibition can lead to heart failure.

6. PKA and cardiomyopathies

6.1 Diabetic cardiomyopathy

Emerging evidence suggests that catecholamines antagonize insulin-induced cardiac glucose transport.¹⁴⁴ Sustained catecholamine stimulation not only contributes to heart failure progression, but also leads to insulin resistance in the heart.¹⁴⁵ Mechanistically, overstimulation of β₂-AR in cardiomyocytes inhibits insulin-induced translocation of glucose transporter 4 (GLUT4) to the plasma membrane, resulting in reduced glucose uptake in a PKA-dependent manner.¹⁴⁵ Treatment with the β-blockers propranolol or metoprolol prevents catecholamine-induced cardiac insulin resistance.¹⁴⁵ In adipocytes, catecholamines induce PKA-dependent lipolysis, which blocks insulin-induced, phosphoinositide 3-kinase/Akt/mammalian target of rapamycin-dependent GLUT4 translocation, resulting in decreased glucose uptake.¹⁴⁶

Insulin resistance is a common feature of type 2 diabetes mellitus, which can cause cardiac dysfunction through PDE4D-mediated decline in total PKA activity.¹⁴⁷ In type 2 diabetes, insulin excess induces PDE4D expression via G-protein-coupled receptor kinase 2 (GRK2)-dependent transactivation of β₂-AR/β-arrestin2/extracellular signal-regulated kinase pathway.¹⁴⁷ Therefore, treatment with the GRK2 inhibitor paroxetine or the β-blocker carvedilol diminishes PDE4D expression, normalizes cAMP/PKA activity, and prevents type 2 diabetes-associated cardiac dysfunction.¹⁴⁷ Interestingly, type 1 diabetes is also associated with reduced cardiac contractility due to decreased PKA activity.¹⁴⁸ Since type 1 diabetes is characterized by insulin deficiency, physiological level of insulin may be necessary for myocardial PKA activation. Together, these studies suggest that either excess or deficiency of insulin can diminish myocardial PKA activity and cause diabetic cardiomyopathy.

6.2 Takotsubo cardiomyopathy

Takotsubo cardiomyopathy (also known as stress cardiomyopathy, broken heart syndrome, or apical ballooning syndrome) is a form of acute, transient heart failure that typically occurs after physical or emotional stress. The incidence of takotsubo cardiomyopathy is markedly increased during the coronavirus disease 2019 pandemic, due to either coronavirus infection¹⁴⁹ or the pandemic-related psychological stress.¹⁵⁰ Although heart function usually recovers within a few weeks, takotsubo cardiomyopathy can cause long-term structural, metabolic changes and a heart failure phenotype.¹⁵¹ Takotsubo cardiomyopathy is associated with elevated circulating and myocardial catecholamine levels, which can be higher than after acute myocardial infarction.¹⁵² Excessive catecholamines directly contribute to myocardial contraction band necrosis, a pathological hallmark of takotsubo cardiomyopathy.¹⁵² Moreover, catecholamines induce severe lipid accumulation in the heart, resulting in lipotoxicity.¹⁵³ There is evidence that catecholamine hypersensitivity caused by genetic abnormalities may increase the susceptibility to takotsubo cardiomyopathy, owing to PKA-mediated hyperphosphorylation of RyR2, PLN, cTnI, and Ca_v1.2.¹⁵⁴ Therefore, a genetic predisposition for takotsubo cardiomyopathy is possible, but remains to be further characterized.

6.3 Anthracycline cardiomyopathy

The anthracycline family anticancer drugs, such as doxorubicin, are frequently used in the treatment of various cancers. Unfortunately, their uses are limited by irreversible and dose-dependent cardiotoxicity, which may eventually progress into heart failure. Doxorubicin-induced reduction in cardiac contractility is correlated with decreased Ca²⁺ transients, likely due to increased RyR2 phosphorylation by PKA.¹⁵⁵ Consistent with this, blockade of the β-AR/cAMP/PKA signalling attenuates anthracycline-related cardiotoxicity in cancer patients.¹⁵⁶ Interestingly, enhancing PKA-mediated PLN phosphorylation with the PDE3 inhibitor levosimendan maintains cardiac contractility following doxorubicin administration.¹⁵⁷ Levosimendan improves cardiomyocyte viability through activation of PKA, as the protective effect is abolished by the PKA inhibitor H89. In addition, PDE1 inhibition with IC86340 or via gene deletion attenuates doxorubicin-induced cardiomyocyte apoptosis, through augmenting A₂R-mediated cAMP synthesis.¹⁵⁸ Since β-AR-mediated cAMP production provokes apoptosis,⁹⁵ A₂R-derived cAMP likely resides in a different subcellular compartment to protect against apoptosis.¹⁵⁸

7. PKA and the sex differences in cardiac health or disease

Biological sex has a profound impact on the heart. Baseline heart function is better in premenopausal women, who are also at lower risk of heart disease compared with age-matched men. Better heart function in females is associated with higher basal PKA activity and distinct gene expression profiles in cardiomyocytes, indicating a fundamental difference between female and male cardiomyocytes.¹⁵⁹ Moreover, the female sex hormone oestrogen can activate cAMP/PKA and enhance cardiac contractility under basal conditions.^160,161 In response to increasing demand (e.g. exercise, stress, and disease), however, female hearts and myocytes exhibit smaller contractions than males. A possible explanation is that oestrogen antagonizes the function of catecholamines,¹⁶² thereby reducing intracellular cAMP levels and suppressing SR Ca²⁺ release.^163,164 Based on the above findings, we speculate that oestrogen counteracts catecholamine-mediated fight-or-flight response via up-regulating cAMP within different intracellular compartments. However, this possibility needs to be confirmed in future studies.

8. Study limitations

Our current knowledge of PKA has been largely based on studies of cAMP, PDEs, AKAPs, and pharmacological PKA inhibitors. Potential limitations of these approaches are summarized below.

8.1 cAMP ≠ PKA

In addition to PKA, cAMP activates other effectors including Epac,^165,166 protein kinase X-linked (PRKX)/protein kinase Y-linked (PRKY),¹⁶⁷ cyclic nucleotide-regulated cation channels [i.e. cyclic nucleotide-gated (CNG) channels and HCN channels],^168,169 and the popeye domain containing (POPDC) proteins (Figure 1).¹⁷⁰ Thus cAMP may produce PKA-independent responses.

On the other hand, PKA can also undergo cAMP-independent activation through diverse mechanisms including R1α oxidation,^6–8 I-κB degradation,^9,10 Smad3/4-PKA-R complex formation,¹¹ and sphingosine-PKA holoenzyme interaction (Figure 1).¹² Therefore, cAMP and PKA can be involved in completely different biological processes, and can have functions distinct from each other.

8.2 PDEs may modulate both cAMP and cGMP levels

The role of PKA in myocardial biology has also been investigated using genetic and pharmacological approaches to alter PDE activity. PDEs are divided into 11 super families (PDE1–11), comprising more than 80 different isoforms that modulate the levels of cAMP and/or cGMP.¹⁴ As a second messenger, cGMP regulates many aspects of cardiovascular homeostasis and pathophysiology through activation of PKG. Hence, PDEs may have cAMP/PKA-independent biological functions.

8.3 AKAPs may anchor other proteins not related with PKA

As scaffold proteins, AKAPs target PKA to subcellular locations and modulate local substrate function. However, AKAPs also anchor other kinases and regulatory proteins not related with PKA.¹⁷¹ Therefore, AKAP deficiency may cause phenotypes in a PKA-independent manner.

8.4 Specificity of PKA inhibitors

Many previous studies have relied on small molecule inhibitors to control PKA activity. Since pharmacological PKA inhibitors have off-target effects and interfere with a wide range of cellular activities,¹⁷² this approach may not accurately demonstrate the role of PKA.

9. Conclusions and perspectives

Preclinical studies over the past decades have greatly advanced our understanding of PKA in the heart. PKA regulates cardiac muscle contraction, relaxation and heart rate through modulating Ca²⁺ dynamics in cardiac myocytes. Myocardial PKA can be activated by the canonical and the non-canonical pathways. In the canonical PKA pathway, β-AR stimulation by catecholamines induces cAMP-dependent PKA activation. The non-canonical pathways prompt PKA activation in a cAMP-independent manner. Abnormal PKA activity has been observed in a wide range of cardiac pathologies. Therefore, PKA has the potential to serve as a drug target for the treatment of heart diseases.

Despite the great progress in recent years, the precise roles of PKA in heart disease pathogenesis remain not fully understood. Critical areas that warrant further investigation include:

• Spatiotemporal regulation of PKA activity in cardiac health and disease

• Proteins that determine PKA compartmentalization

• Control of cardiomyocyte morphology, fate, and function by PKA

• Signalling pathways that regulate apoptosis/necrosis downstream of PKA

• Connections between PKA and oxidative stress

• Additional PKA substrates in the heart

• Role of PKA in the sex differences of heart disease

• Heart disease types that require PKA-activating or inhibiting interventions

• Adverse effects of PKA activation or inhibition

• Development of potent and specific PKA activators and inhibitors.

Our current knowledge of cardiac PKA is mostly obtained using indirect approaches, which have limitations as described above. Strategies that directly alter the PKA holoenzyme should be considered in future research. Specific PKA inhibitors, such as PKI are also valuable tools for studying the biological functions of PKA in cardiac physiology and pathophysiology. Development of specific PKA activators or inhibitors with a satisfactory pharmacokinetic profile, good efficacy, and tolerability would greatly benefit future preclinical and clinical studies.

Funding

This work was supported by WSU College of Pharmacy and Pharmaceutical Sciences. Z.C. was supported by the National Heart, Lung, and Blood Institute (NHLBI, R00HL119605, R56HL145034), National Institutes of Health.