Adrenergic and calcium modulation of the heart in stress: From molecular biology to function

Abstract

There is strong evidence about the importance of catecholamines and calcium signaling in heart function. Also, interaction of these two systems is well documented. Catecholamines signal through adrenergic receptors, and further activate calcium transport either from the extracellular space, or from the intracellular calcium stores. This review summarizes current knowledge on catecholamine production in the heart, with special focus on the final enzyme in the catecholamine synthesizing pathway, phenylethanolamine N-methyltransferase (PNMT), in different cell types in the heart. Further, signaling through different types of adrenergic receptors in physiological conditions and after exposure to different stressors is discussed. Also, part of this review considers activation of an intracellular calcium transport system via inositol 1,4,5-trisphosphate receptor and to possible functional consequences in control and stress conditions.

• Stress, catecholamines
• phenylethanolamine N-methyltransferase
• adrenergic
• receptors
• calcium transport
• inositol 1,4,5-trisphosphate receptors

Catecholamines and the heart

The catecholamines epinephrine and norepinephrine have profound influence on cardiac function, beginning almost as soon as the heart starts beating and continuing through the remaining stages of life. The cardiovascular augmentation during the classical stress “fight or flight” response is thought to be primarily due to catecholamine release from sympathetic nerve endings and from the adrenal medulla. Furthermore, the obligatory role of catecholamines in fetal cardiac development has been underscored by the finding that mice with knock-out of the genes of enzymes for catecholamine synthesis die in utero due to cardiac failure at an early developmental stage, before cardiac sympathetic innervation occurs (Thomas et al. 1995; Zhou et al. 1995). Under physiological conditions the principal catecholamine that regulates cardiac function is norepinephrine, since it is the main mediator of the sympathetic neural influence on the heart. Nevertheless, epinephrine is also important in heart function, since it increases heart rate and force of contraction, alters blood flow and uptake of glucose, oxygen, potassium and phosphate (Elayan et al. 1990a).

It has been generally assumed that epinephrine affects the heart by passage through the bloodstream from the adrenal medulla (Elayan et al. 1990a; Esler et al. 1990). Nevertheless, there is some evidence that epinephrine can be synthesized directly in the cardiac tissue. Kvetnansky et al. (1979) found a slight but significant elevation of plasma epinephrine levels in adrenal medullectomized rats exposed to stress, suggesting an extra-adrenal source of epinephrine. Esler et al. (1991) showed that epinephrine is released by the human heart in vivo, and it is also present in human plasma and urine after bilateral adrenalectomy (Shah et al. 1984). Moreover, patients with heart transplants maintain adequate cardiac function even in the absence of sympathetic reinnervation, although epinephrine from the adrenal medulla is still available (Rowan and Billingham 1988; Beckers et al. 2004).

Phenylethanolamine N-methyltransferase in the heart

The final enzyme in the catecholamine synthesizing cascade that converts norepinephrine to epinephrine is phenylethanolamine N-methyltransferase (PNMT, EC 2.1.1.28). Although PNMT activity is largely restricted to the adrenal medulla, where the majority of epinephrine is synthesized, findings of low PNMT activity in the rat heart have been reported by several authors (Axelrod 1962; Culman et al. 1987; Torda et al. 1987; Elayan et al. 1990b; Kennedy and Ziegler 1991, 2000; Kennedy et al. 1995; Ebert et al. 1996). PNMT mRNA was detected for the first time in embryonic rat hearts (Ebert et al. 1996) and subsequently in adult rat hearts (Krizanova et al. 2001; Slavikova et al. 2003). Moreover, recently we have shown that PNMT mRNA is expressed also in transplanted human hearts (Goncalvesova et al. 2004). Interestingly, distribution of PNMT mRNA levels within the heart seems not to be uniform. The left cardiac atrium has a much greater abundance of the PNMT mRNA compared to the right atrium, whereas levels in the ventricles are about 10-times lower (Krizanova et al. 2001). These data are in good agreement with previously published reports concerning localization of the PNMT activity in the cardiac atria and ventricles (Culman et al. 1987; Torda et al. 1987; Elayan et al. 1990b; Kennedy et al. 1995). PNMT activity in the adult rat heart was found to be greatest in the left atrium and many times lower in the ventricles (Culman et al. 1987; Torda et al. 1987). Despite the presence of the PNMT mRNA in the heart being confirmed by many authors (Ebert et al. 1996; Krizanova et al. 2001; Slavikova et al. 2003; Micutkova et al. 2004), the cell types that are able to express the PNMT in the heart are still a matter of investigation and discussion. Huang et al. (1996, 2005) identified special cardiac cells capable of adrenergic paracrine signaling in mammalian hearts. These cells, which do not have neuronal or chromaffin cell ultrastructural morphology, were described as intrinsic cardiac adrenergic (ICA) cells. ICA cells containing mRNA and protein of enzymes involved in catecholamine biosynthesis could be identified in fetal human hearts at a developmental stage before sympathetic innervation occurs (Huang et al. 1996, 2005), but also in the adult heart. These findings provide evidence for the existence of an endogenous system responsible for catecholamine production in the fetal heart. Ebert and Thompson (2001) have shown that cardiac cells expressing catecholamine biosynthetic enzymes are present in the rat embryonic heart, where they seem to be progressively associated with regions of the cardiac conduction system. Slavikova et al. (2003) identified two types of catecholamine-handling intrinsic ganglion neurons: small intensely fluorescent (SIF) cells and large-diameter neurons. SIF cells exhibited tyrosine hydroxylase (TH)-immunoreactivity, but they were not positive for dopamine-β-hydroxylase (DBH). In contrast, large-diameter intrinsic TH-positive neurons displayed also DBH-immunoreactivity, thus indicating the capacity for the synthesis of norepinephrine. Presence of the DBH mRNA in rat atrial samples was also verified using reverse transcription and subsequent polymerase chain reaction (Slavikova et al. 2003). This finding of DBH expression in rat heart was confirmed by several authors (Kvetnansky et al. 2004; Micutkova et al. 2004). On the other hand, the presence of TH mRNA in the heart is still under investigation.

Based on data confirming much higher abundance of the PNMT mRNA in the cardiac atria compared to the ventricles, it seems most likely that the majority of the PNMT is expressed predominantly in neuronal cardiac ganglia, which are localized in the atria (Krizanova et al. 2001; Slavikova et al. 2003). On the other hand, a small amount of the PNMT mRNA was observed also in the cardiac ventricles, so the possibility that also cardiomyocytes express PNMT could not be excluded. Moreover, dual origin of the cardiac PNMT mRNA was supported by the finding that immobilization provoked different effects on the PNMT mRNA in the ganglionic (the part of the atria containing cardiac neuronal ganglia) and nonganglionic (the part of the atria without cardiac neuronal ganglia) parts of the atrium (Kvetnansky et al. 2004). Recently, we observed different effects of hypoxia and also cold stress on the PNMT mRNA in the cardiac atria and ventricles (Tillinger et al. 2006a). Finally, we have clearly confirmed PNMT gene expression in isolated cardiomyocytes and also in primary culture of isolated cardiomyocytes (Tillinger et al. 2006b). These results were confirmed by demonstrating PNMT-immunoreactivity in isolated cardiomyocytes (Tillinger et al. 2006a). Taking together, it becomes clear that PNMT in the heart is expressed in several different types of cells in the heart.

Phenylethanolamine N-methyltransferase and stress

The significance of PNMT gene expression in the heart was studied also under stress conditions. Single and also repeated immobilization stress significantly increases the PNMT mRNA levels in both, cardiac atria and ventricles (Krizanova et al. 2001; Kvetnansky et al. 2004; Micutkova et al. 2004). Since immobilization stress markedly activates many cardiovascular, metabolic, neuroendocrine and other processes, the highly stress-induced increase in cardiac PNMT mRNA levels suggests participation of the cardiac adrenergic system in physiological and/or pathophysiological processes (Krizanova et al. 2001). Subsequently we have found different modulation of PNMT gene expression in the heart by various stressors (cold, immobilization and hypoxia). Cold exposure for seven days significantly decreased PNMT mRNA levels in the left and right atrium of the rat compared to controls. Another stress stimulus, hypoxia for 6 h, significantly increased PNMT mRNA levels in cardiac atria compared to normoxic controls (Tillinger et al. 2006a). In contrast, PNMT gene expression in the ventricles was not affected by cold or the hypoxic stimulus. Thus, PNMT expression in the cardiac atria seems to be more vulnerable to different stressors compared to the ventricles. Moreover, these data indicated different regulatory mechanisms by which PNMT gene expression is modulated in different parts of the heart.

It is well known that glucocorticoids are critical regulators of PNMT gene expression (Bohn et al. 1981; Teitelman et al. 1982; Schmid et al. 1995; Tai et al. 2002) and also induction of PNMT in response to acute and chronic stress (Sabban et al. 1995; 1998). The final enzyme in epinephrine synthesis is regulated by glucocorticoids in two ways: postranscriptionally via regulation of PNMT co-substrate S-adenosylmethionine levels and transcriptionally through stimulation of glucocorticoid responsive elements (GREs), which occur in the proximal 5′- flanking sequences of the PNMT gene promoter (Jeong et al. 2000; Tai et al. 2002; Wong et al. 2002). Besides GREs, a synergistic activation of the PNMT promoter by EGR-1, AP-2 and the glucocorticoid receptor has been observed (Wong et al. 1998). Considering the important regulatory role of glucocorticoids, Kvetnansky et al. (2004, 2006), used corticotropin-releasing hormone (CRH) knockout mice (CRH KO) as an experimental model. These mice, unable to synthesize a functional CRH, can help to define mechanisms that are involved in regulation of the PNMT during stress. CRH is the major driver of the hypothalamic pituitary adrenocortical (HPA) system and it triggers both basal and stress-induced release of ACTH from the anterior pituitary, subsequently leading to release of glucocorticoid hormones from the adrenal cortex (Bruhn et al. 1984; Venihaki and Majzoub 2002). CRH KO mice exhibit impaired endogenous glucocorticoid synthesis and secretion (Muglia et al. 2001). Kvetnansky et al. (2004, 2006), have shown that single immobilization and also repeated stress exposure induce a significant increase in PNMT mRNA levels in the heart atria and ventricles of wild-type (WT) mice. In contrast to these findings, neither single immobilization, nor seven daily repeated immobilizations were able to induce a significant elevation of PNMT gene expression in the cardiac atria of CRH KO mice. A different picture was observed in the cardiac ventricles. Surprisingly, PNMT mRNA levels were significantly increased in both ventricles of CRH KO mice exposed to a single immobilization and a decrease was found only after repeated stress. These findings are in good agreement with a previously published report about different responses of PNMT gene expression in the atria and ventricles of adrenalectomised or hypophysectomised rats after stress exposure (Krizanova et al. 2001). In conclusion, all these data strongly indicate that the HPA axis plays a crucial role in regulation of the PNMT gene expression in the cardiac atria under stress conditions; however, PNMT gene expression in the cardiac ventricles is probably regulated by a different mechanism(s).

Adrenergic receptors and their occurrence in the heart

Both epinephrine and norepinephrine bind to adrenergic receptors (adrenoceptors). Adrenergic receptors are members of the G-protein-coupled receptor (GPCR) family. In the heart, several subtypes are expressed both at the gene and protein levels (Brodde and Michel 1999; Gauthier et al. 2000). The β₁ subtype is considered to be the main subtype, since about 80% of cardiac β-adrenoceptors are β₁ subtype. β₂- and β₃-adrenoceptor subtypes are also present in the cardiac tissue (Gauthier et al. 1998; Brodde and Michel 1999). Although human β₃-adrenoceptors share 51% identity with β₁-adrenoceptors and 46% identity with β₂-adrenoceptor amino acid sequences, and there is 54% homology between β₁- and β₂-adrenoceptor amino acid sequences, the receptor subtypes show closer relationship between β₁ and β₃ subtypes than between β₁ and β₂ subtypes, what can be important in receptor regulation (see below).

In addition to β-adrenoceptors, α₁-adrenoceptors and α₂-adrenoceptors are also present in cardiac tissue. The α₁-adrenoceptor and β-adrenoceptor subtypes are effector receptors, i.e. when activated lead to functional changes in the functioning of the myocardium/cardiomyocytes (Brodde and Michel 1999). In contrast, α₂-adrenoceptors modulate catecholamine release from the nerve terminals, i.e. they are located pre-junctionally (Brodde and Michel 1999). Also, these receptors are as much as 30-times less abundant (at the mRNA level) than α₁-adrenoceptors (α_2C vs. α_1A; Berkowitz et al. 1994). Moreover, both α₁-adrenoceptors and β₁-adrenoceptors have specific low-affinity binding sites (α_1L, so called “β₄”, respectively) that can affect cardiac physiology in other ways than via conventional binding sites (Ford et al. 1997; Takahashi et al. 2000; Kaumann et al. 2001; Santos et al. 2005). The questions arise, why are targets for catecholamine action in the heart so abundant, and why are there several adrenoceptor subtypes and what is their role in cardiac function?

Signaling pathways of adrenergic receptors

Firstly, it is necessary to mention the function/coupling/signalling pathways of the individual adrenoceptor subtypes. Although α₁-adrenoreceptors in humans are less highly expressed in the myocardium than β-adrenoceptors, in rodents (especially in rat) the amount of α₁-adrenoceptor binding sites is greater than the amount of β-adrenoceptor binding sites. Despite the lower density of α₁-adrenoreceptors in the human heart, their physiological relevance is undoubted. These receptors are responsible for enhanced inotropy under most conditions, but opposite effects are also possible. In addition, stimulation of α₁-adrenoceptors increases the Ca^{2 +} sensitivity of the myofilaments. Apart from these short-term effects, extended stimulation of cardiac α₁-adrenoceptors may also cause the development of a hypertrophic cardiac phenotype in rats (Zimmer 1997). Nevertheless, a hypertrophic effect due to extended stimulation of cardiac α₁-adrenoceptors was shown in rodents and it is difficult to extrapolate these results into human physiology. α₁-adrenoceptors couple to Gq_/11 proteins (Table I). Subsequent steps in the intracellular signalling cascade include activation of phospholipase C (PLC) and increase in intracellular diacylgylcerol (DAG) and inositol 1,4,5-trisphosphate (IP₃) concentration (Figure 1). Cardiac α₁-adrenoceptors couple not only to PLC; phospholipase D can be also activated and coupling to various ion currents, i.e. L-type Ca^{2 +} channels, the transient outward current Ito, the delayed rectifier K⁺ current, and the acetylcholine-activated K⁺ channel has been demonstrated (Endoh et al. 1991). Moreover, the Na⁺/H⁺ exchanger and Na⁺/K⁺-ATPase can be activated (Endoh et al. 1991).

Table I.  Pharmacological characteristics of individual adrenoreceptor subtypes and coupling to G proteins with subsequent activation of second messengers.

Receptor	Adrenaline	Noradrenaline	G protein	Second messenger
α1A	0.3	1.2–2.8	Gq/G11	IP3/Ca^2 + ; DAG
α1B	6	10	Gq/G11	IP3/Ca^2 + ; DAG
α1D	0.016	0.012	Gq/G11	IP3/Ca^2 + ; DAG
α2A	8.3–5.6*	8.4–5.6*	Gi/Go	↓ Cyclic AMP
α2B	6.2–5.2*	9.1–5.6*	Gi/Go	↓ Cyclic AMP
α2C	6.2–5.8*	8.7–5.9*	Gi/Go	↓ Cyclic AMP
β1	1.0 2.7#	0.5 0.8#	Gs	↑ Cyclic AMP
β2	0.9–0.7 2.2#	10–6 (9.7) 36#	Gs	↑ Cyclic AMP
β3	4.7–3.9* 49#	3.8–5.3* 6.3#	Gi Gs	↓ Cyclic AMP ↑ Cyclic AMP

Pharmacological characteristics of adrenoceptor subtypes are expressed as EC50 (expressed as μmol l − 1) or pKi (*) or K_AC (activation constant of AC stimulation; expressed as nmol.l − 1)#. Data are given as in Bylund et al. (1994), Chen et al. (1994) (EC50), IUPHAR receptor database (http://www.iuphar-db.org/GPCR/; Ki), and Skeberdis (2004) (K_AC). Coupling to G proteins and second messenger system(s) are also shown. Please note that the coupling of α₁-adrenoceptors with PLD (α_1A>α_1B>α_1D) was not included in the table as the precise mechanism of activation is not known yet; all three subtypes are able to activate PLD, but the mechanisms may differ in different cell types.

Figure 1 Signaling cascades of individual types of adrenergic receptors. AC, adenylate cyclase; DAG, diacylglycerol; DHPR, dihydropyridine receptor, G_s; G_i, Gq_/11, G proteins type s, i and q; IP₃, inositol 1,4,5-trisphosphate; IP₃R, IP₃ receptor; NCX, sodium calcium exchanger; PKA, protein kinase A; PKB, protein kinase B; PKC, protein kinase C; PLC, phospholipase C; PIP2, phosphatidylinositol 4,5-diphosphate; RyR, ryanodine receptor; PLB, phospholamban; SERCA, sarcoplasmic reticulum calcium ATPase; CaMKII, calmodulin-dependent kinase II; PI3K, phosphatidyl inositol 3-kinase.

As mentioned above, the role of α₂-adrenoceptors is modulatory rather than affecting cardiac function directly. α₂-adrenoceptors mediate prejunctional inhibition of norepinephrine release. All α₂-adrenoceptor subtypes inhibit adenylyl cyclase (AC), which results in decreased intracellular cAMP content (Table I).

Important differences occur between the two “classical” heart β-adrenoceptors and the β₃ subtype. The human β₃-adrenoceptor gene contains introns, whereas the β₁- and β₂-adrenoceptor genes do not (Brodde and Michel 1999). Moreover, the β₃-adrenoceptors have no phosphorylation sites for protein kinase A (PKA) and therefore they are considered as relatively insensitive to receptor desensitization in vitro and in vivo following activation with agonists (Gauthier et al. 2000).

The β₁-adrenoceptors are considered as the main heart adrenoceptor subtype. Their role in cardiac physiology is stimulatory (the effects are positively inotropic, chronotropic, dromotropic and bathmotropic). Thus, stimulation of these receptors increases the strength of heart muscle contraction, the heart rate, increases the atrio–ventricular conduction velocity and increases the heart muscle excitability (the last two effects are sometimes considered as secondary and only inotropic and chronotropic effects are described). This receptor subtype couples to Gs proteins (Figure 1), stimulates AC, which subsequently leads to increase in cAMP intracellular levels (Table I), activation of PKA, and further to phosphorylation of several sarcolemmal proteins, including L-type Ca^{2 +} channels and phospholamban, which regulates reticular Ca^{2 +} -ATPase (Kaumann and Molenaar 1997). Phosphorylation of L-type Ca^{2 +} channels promotes calcium influx and thus enhances contraction; phosphorylation of phospholamban may be involved in enhanced diastolic relaxation by increasing calcium uptake into the sarcoplasmic reticulum. Gs can also directly activate L-type calcium channels (it is not clear yet, whether this applies also for human myocardium, see Brodde and Michel 1999 for discussion).

From studies with cloned β₂-adrenoceptors, it is known that the functions, coupling and signalling pathways resemble those of the β₁ subtype (Table I). Differences from β₁-adrenoceptors that should be mentioned when considering the importance of β₂-adrenoceptors in the heart include: β₂-adrenoceptor stimulation (in contrast to β₁-adrenoceptor stimulation) does not correlate well with increases in intracellular cAMP and does not lead to phosphorylation of phospholamban in these cells. Moreover, there are kinetic differences between β₁- and β₂-adrenoceptor-mediated activation of L-type calcium channels. As another difference, more effective coupling of β₂-adrenoceptors to AC was observed, which might explain why adrenaline causes nearly identical increases in force of contraction via β₁- or β₂-adrenoceptor stimulation (Brodde and Michel 1999). Moreover, chronic stimulation of β₁-adrenoceptors and β₂-adrenoceptors elicits opposing effects on the fate of cardiomyocytes: β₁-adrenoceptors induce hypertrophy and apoptosis; but β₂-adrenoceptors promote cell survival (Zheng et al. 2004). The other difference between β₁-adrenoceptors and β₂-adrenoceptors lies in their differential role in regulating autonomic signalling: it has been demonstrated that heart rate variability in β₁- and β₂- KO mice shows significant differences (Ecker et al. 2006). In other words, this study demonstrated the specific roles of individual adrenoceptor subtypes in regulating both heart rate and heart rate variability. Finally, there is evidence that β₂-adrenoceptors couple, in addition to the “classical” cascade, also to G_i proteins (Kilts et al. 2000). Thus, in addition to cAMP and calcium signalling pathways, β₂-adrenoceptors activate guanylyl cyclase via G_i/G₀ family proteins (Li et al. 2004). Stimulation of guanylyl cyclase causes an increase in cGMP levels, and subsequent activation of protein kinase G.

One of the intriguing questions in cardiac physiology is about the presence of another β-adrenoceptor subtype(s). The presence of β₃-adrenoceptor in the heart is supported by these findings: β₃-adrenoceptor mRNA has been found repeatedly in the heart (Gauthier et al. 2000; Myslivecek et al. 2006), the β₃-adrenoceptor has been also found using immunohistochemical detection (De Matteis et al. 2002), and negative inotropic effects have also been repeatedly recorded when using β₃-adrenoceptor agonists (Gauthier et al. 1996, 1998, 1999). Moreover, the importance of β₃-adrenoceptors was demonstrated using mice over-expressing this receptor (Tavernier et al. 2003). On the other hand, Heubach et al. (2002) did not find negative inotropic effects of adrenaline as mediated via β₃-adrenoceptors. These receptors would couple to G_i proteins in the heart (Table I), which have inhibitory effects on AC and decreases the amount of intracellular cAMP, while in fat tissue β₃-adrenoceptors are coupled to G_s. It is supposed that effects of β₃-adrenoceptors on heart rate are mediated via increased nitric oxide synthase activity (Rozec and Gauthier 2006). Nevertheless, physiological relevance of these receptors in the heart remains to be elucidated.

During the 1990s it was believed that another, fourth, putative β₄-adrenoceptor is present in the heart of several species (rat, mouse, ferret, human–Molenaar et al. 1997). Nevertheless, at the beginning of the new millennium it has been recognized that these binding sites are a rather low affinity state of β₁-adrenoceptors rather than another β-adrenoceptor subtype (Granneman 2001). On the other hand, some results (Santos and Spadari-Bratfisch 2001; Santos et al. 2005) and those from our laboratory (Myslivecek et al. 2003) indicated that the situation with regard to low-affinity binding sites could be more complicated as in some conditions (stress), these sites show behavior distinct from β₁-adrenoceptors.

Catecholamine action through receptors depends on several factors, according to general receptor occupational theory (Kenakin 2004). Specifically, action depends on (1) relative receptor density in the target tissue; (2) intrinsic efficacy (the ability to activate the intracellular signalling pathways); (3) receptor affinity, expressed as equilibrium dissociation constant of the drug (catecholamine)-receptor complex. Thus, when intrinsic activity and receptor density is higher and equilibrium dissociation constant is lower, drug (catecholamine) action is stronger. All these factors are considered when analyzing effects of receptor activation in target tissue, but the above is a simplification, which may lead to misinterpretation of results (Kenakin 2004). Considering the effects of catecholamines on heart adrenoceptors:

1. The relative amount of receptor in the tissue, and region-specific differences in receptor densities. β₁-adrenoceptors are more abundant than β₂-adrenoceptors and much more abundant than α₁-adrenoceptors and β₃-adrenoceptors, but there are some regions in which the relative density of β₂-adrenoceptors is higher than that of β₁-adrenoceptors (Brodde and Michel 1999; Myslivecek et al. 2006).

2. Receptor affinity. As can be deduced from Table I, for some receptor subtypes expressed in the heart, there are substantial differences among receptor subtypes affinities to epinephrine or norepinephrine. Moreover, differences can be observed also in affinity to catecholamines for specific subtype (for example, β₂-adrenoceptors have higher affinity to epinephrine than to norepinephrine).

3. Receptor function, or post-receptor coupling. Different adrenoceptors have different functions, e.g. some are cardiostimulatory (β₁-adrenoceptor) and others are cardioinhibitory (β₃-adrenoceptors). The subtype with complex function (α₁-adrenoceptors) is also expressed in the heart. It has been mentioned above that adrenoceptors couple to specific G proteins and signalling systems. Despite that, there is evidence that β₂-adrenoceptor couple in addition to the “classical” cascade also to G_i proteins (Kilts et al. 2000). This finding can extend our understanding of peculiarities of cardiac signaling, since disruption of coupling of these receptors to G_i enhances their ability to stimulate AC. Also, it should be mentioned that coupling of β₃-adrenoceptors is “atypical” as they normally couple to G_s and not to G_i.

Adrenoceptors, similarly to other GPCRs undergo changes when stimulated more or less. Typically, with increased receptor stimulation receptors undergo desensitization-internalization and down-regulation (Morris and Malbon 1999; Fukunaga et al. 2006). However, there are some differences in the receptor regulation processes among adrenoceptor subtypes in the heart. As mentioned above, β₃-adrenoceptors are relatively insensitive to receptor desensitization (Gauthier et al. 2000). There is also a difference in the regulatory processes between β₁-adrenoceptors and β₂-adrenoceptors. Although with stimulation β₁-adrenoceptors are internalized like β₂-adrenoceptors, there is no reduction in receptor number, but β₁-adrenoceptors accumulate inside the cells. Thus, while β₂-adrenoceptors undergo lysosomal degradation, the β₁-adrenoceptors do not (Liang et al. 2003).

Besides other systems, stress affects also cardiac adrenoceptors, and alter signalling (Santos and Spadari-Bratfisch 2006). Moreover, balance between the sympathetic and parasympathetic systems can also be disturbed as a consequence of stress (Farah et al. 2006). Finally, stress changes receptor densities (Myslivecek et al. 2004).

An intriguing finding is that just two catecholamines (epinephrine, norepinephrine) have many “target” -receptors that differ greatly in the functional consequences of receptor activation: while β₁- and β₂-adrenoceptors cause cardiostimulation (Kaumann and Molenaar 1997), β₃-adrenoceptors are responsible for cardioinhibition (Gauthier et al. 2000) and α₁-adrenoceptors contribute to enhanced inotropy (Xiao et al. 2006), while the role of α₂-adrenoceptors is in the regulation of catecholamine release (Docherty 2002). Moreover, the heart is equipped with a very efficient system of receptors that binds transmitters and activates intracellular signalling not in parallel, but as a consequence of differences in affinity, coupling, signalling and regulatory events. Conceivably, catecholamines first activate β₁-adrenoceptors, than β₂-adrenoceptors and α₁-adrenoceptors (which are less expressed) and if the presence of catecholamines persists, a mechanism of cardioinhibition through β₃-adrenoceptors may then be activated. It could be hypothesized that β₃-adrenoceptors are “back-up” receptors activated during extreme or stressful conditions (Pelat et al. 2003). This hypothesis awaits further investigation.

Catecholamines and calcium release through the inositol 1,4,5-trisphosphate receptors

The catecholamines norepinephrine and epinephrine can modify calcium homeostasis, either by increasing cytoplasmic calcium by stimulating entry from the extracellular space through L-type calcium channels, or from intracellular calcium stores (sarcoplasmic reticulum) through inositol 1,4,5-trisphosphate (IP₃) receptors (IP₃R; Figure 1). Calcium is a key molecule controlling several cellular processes, from fertilization to cell death, and in all cell types. In excitable and contracting cells, such as cardiomyocytes, Ca^{2 +} controls muscle contractility. The spatial and temporal segregation of Ca^{2 +} concentrations are central to maintain its concentration gradients across the cells and the cellular compartments for normal function.

Together with ryanodine receptors, IP₃ receptors are able to release calcium from the intracellular stores. While ryanodine receptors are high capacity channels, which are able to release huge amount of calcium and participate directly in the process of excitation–contraction coupling, IP₃ receptors can release only a moderate amount of calcium and the physiological relevance of calcium release through this mechanism is not yet clear. Within cardiomyocytes, IP₃ receptor mRNA levels are approximately 50-fold lower than levels of the cardiac ryanodine receptor mRNA (Moschella and Marks 1993). In the heart, all three types of IP₃ receptor are found, although with different predominance. We reported that there is an abundant expression of the type 1 IP₃ receptors in the cardiac atria in contrast to low expression in the ventricles (Lencesova et al. 2002). The type 1 IP₃ receptor is most abundant in the cardiac ganglia and neuronal cells (Slavikova et al. 2006), while the type 2 IP₃ receptor is most prevalent in cardiomyocytes (Perez et al. 1997).

The functional role of IP₃ in the heart is still unclear. Several hormones can activate IP₃ production in cardiomyocytes and some of their inotropic, chronotropic and arrhythmogenic effects may be due to the calcium release mediated by IP₃ receptors (Lipp et al. 2000). Compared to ryanodine receptors, responses to IP₃ are slow and weak, and the calcium so released does not contribute to calcium induced-calcium release. Thus, IP₃ is less likely to be important in regulating beat-to-beat changes in intracellular calcium under physiological conditions. Chronic mechanical overload of the atrial myocardium increased IP₃ receptor expression especially in patients with chronic atrial fibrillation (Yamada et al. 2002). Apart from the localization of IP₃ receptors in normal atrial tissue, several reports pointed out a characteristic up-regulation of IP₃ receptor mRNAs in hearts from hypertrophy and end-stage heart failure patients (Gutstein and Marks 1997), while mRNA for ryanodine receptors was down-regulated in the same tissue (Marks 2000). The IP₃ signaling cassette might be also a promising new target for understanding and managing atrial arrhythmia (Mackenzie et al. 2002). Thus, up-regulation of IP₃ receptors may be important in modulating intracellular calcium homeostasis and initiating and perpetuating atrial fibrillation. Furthermore, Xu et al. (2000) proposed that decreased levels of the IP₃ receptors might play a role in cardioprotection.

Inositol 1,4,5-trisphosphate receptors and stress

Stress is considered to be a major contributor to the development of cardiovascular disease and psychiatric illness. Therefore, it is crucial to clarify mechanisms involved in the conversion of brief beneficial responses to stressors into prolonged detrimental consequences. We have already shown different effects of single and repeated immobilization stress on gene expression of the IP₃ receptors. After a single immobilization stress, gene expression of the type 1 IP₃ receptor was increased (Lencesova et al. 2002; Krizanova et al. 2005), while after repeated immobilization stress (seven times), both gene expression and protein of the type 1 IP₃ receptor was decreased significantly in the left atrium of rats (Krepsova et al. 2004). Exposure to a relatively weak stressor, cold for 24 h, was not able to increase gene expression of IP₃ receptors (Krizanova et al. 2005). However, when rats were exposed to cold (4°C) for seven days, gene expression of the type 1 IP₃ receptor was significantly upregulated. Animals exposed to cold for 28 days became cold-adapted and did not show any difference in the type 1 IP₃ receptor mRNA level compared to non-stressed controls (Krizanova et al. 2005).

Several questions arise from these observations. Firstly, what elements are involved in the stress-induced increase in expression of IP₃R1 and also IP₃R2? Since the promoter region of the type 1 IP₃R contains a glucocorticoid responsive element, we proposed involvement of glucocorticoids in this process. We have shown that adrenalectomy completely abolished the immobilization-induced increase in both IP₃R1 and IP₃R2 mRNAs, but did not affect basal gene expression of the respective proteins in control conditions (Lencesova et al. 2002). We also tested the effect of catecholamines using the neurotoxin 6-hydroxydopamine. When 6-hydroxydopamine was given to animals, norepinephrine and subsequently epinephrine levels were markedly reduced. The partial depletion of catecholamines decreased significantly mRNA and protein levels of the type 1 IP₃ receptors in the heart, in both control conditions and after exposure to immobilization stress for 2 h (Jurkovicova et al. 2006). A similar effect was observed on gene expression for the type 2 and 3 IP₃ receptors. These findings provide a molecular basis for the hypothesis that the mechanism of adrenergic modulation in the heart involves IP₃ receptors. It was previously reported that changes in expression profiles for other genes in cardiomyocytes were observed after norepinephrine infusion (Ponicke et al. 2001; Li et al. 2003). By analogy to dihydropyridine receptors, for which norepinephrine treatment alters mRNA levels through the AP1 transcription factor (Deelman et al. 1998), it might be speculated that the effect of depleted norepinephrine is effected through AP1. The AP1 transcription factor is modulated by several proteins, products of immediate-early genes. We have shown that mice with knockout of the c-fos gene express significantly lower levels of the IP₃R1 mRNA compared to their heterozygous littermates (Figure 2A). Interestingly, this decrease was more profound in male atria than in female atria. Nevertheless, a single immobilization stress elicited a large increase in the IP₃R1 mRNA in male left atria. This increase was almost twice that induced by immobilization in control mice (Figure 2B). These results suggest that c-fos, and Fos protein, participates in regulating gene expression of the type 1 IP₃ receptor in physiological conditions, but it also might affect the increased expression of these receptors induced by immobilization stress.

Figure 2 Effect of c-fos knockout on the IP₃R1 mRNA in the cardiac left atria in control conditions (A) and after a single immobilization stress (B; IMO). Each column represents an average of three animals. Statistical significance *** represents p < 0.0001.

Physiological relevance of inositol 1,4,5-trisphosphate receptors in the heart

The second issue to be addressed is the physiological relevance of altered expression of the IP₃ receptor genes in the heart. It is generally accepted that IP₃ receptors do not play a major role in cardiac excitation–contraction coupling, although they could be involved in physiological modulation of cardiac contractility in response to pharmacological agents and hormones (Marks 1997). As mentioned above, IP₃ receptors are up-regulated in several cardiovascular diseases, and also during single immobilization stress exposure and seven-day cold exposure. In contrast, decreased levels of IP₃ receptors were observed after repeated immobilization stress (two hours daily for 7 days) in the heart (Krepsova et al. 2004), stellate ganglia (Micutkova et al. 2003) and kidney (Zacikova et al. 2000). Increased and decreased levels of IP₃ receptors have different physiological consequences.

Lack of type 1 IP₃ receptors was studied in IP₃R1 gene knockout mice (Matsumoto and Nagata 1999). These mice were rarely born alive, indicating that IP₃R1 has some important functions during embryonic development. Liveborn exhibited severe neurological symptoms, ataxia and epilepsy, and were shown to be deficient in cerebellar long-term depression. These results suggest that IP₃ receptors are crucial in neonates, while in adulthood they have regulatory role mainly in neuronal tissue.

Apoptosis is a “suicide” program highly conserved throughout evolution, and present in all metazoan cells (Evan et al. 1995). Apoptosis plays an important role in normal development and in pathological processes, such as heart failure and atherosclerosis. It has been proposed that elevation of cytoplasmic calcium concentration is a regulatory signal involved in apoptosis (Cohen and Duke 1984; Jones et al. 1989). IP₃ has been reported to be required for signaling pathways leading to apoptotic cell death in lymphoid cells, when apoptosis is induced by a number of different agents, including glucocorticoids (Jayaraman and Marks 1997). Recent studies have demonstrated that IP₃ receptors can also be activated during apoptosis independently of the production of IP₃. In the case of type 1 IP₃ receptors, this may occur following proteolysis by activated caspase 3.The type 1 IP₃ receptor has a single highly conserved DEVD cleavage site, which is caspase 3 specific. Truncation of IP₃ receptors by caspase 3 at this position removes a large portion of the protein, including the amino-terminal IP₃-binding region and other regulatory domains (Hanson et al. 2004). The part that remains in the endoplasmic reticulum membrane is a constitutively active channel that continuously leaks calcium (Nakayama et al. 2004).

In an alternative model for IP₃ receptor activation during apoptosis that is not reliant on caspase cleavage, cytochrome C released from mitochondria is suggested to control calcium movement through IP₃ receptors. Cytochrome C was shown to bind IP₃ receptors and reduce the calcium-dependent inhibition of channel opening (Boehning et al. 2003). These studies all point to a crucial role of IP₃ receptors in the generation of calcium signals that drive cells into apoptosis.

Conclusion

In summary, the catecholamine synthesizing pathway together with calcium is able to secure a wide variety of processes in the heart. Stress can affect not only production of catecholamines, but also adrenoceptors and IP₃ receptors. In this review we described in particular communication by the adrenergic receptor pathway through the release of calcium from intracellular stores, mainly through IP₃ receptors. Recently we have found that the amount of the type 1 IP₃ receptors is decreased when catecholamines are partially depleted. This finding provides a molecular basis for the hypothesis that a mechanism of adrenergic modulation in the heart involves the type 1 IP₃ receptors. Nevertheless, physiological relevance of this signaling remains to be elucidated. Since decreased expression and activity of IP₃ receptors inhibits apoptosis, this allows the speculation that an increase in IP₃ receptors due to increased and prolonged adrenergic modulation might be one of the stimuli involved in the process of apoptosis. Our further research will be oriented to clarify this issue.

Acknowledgements

This work was supported by grants Genomika SP 51/028 08 00/0280802, VEGA 2/6078, 2/5125 and 2/7123, GAUK11/06 and APVV 51-027-404.