G protein-coupled receptor signalling in the cardiac nuclear membrane: evidence and possible roles in physiological and pathophysiological function

Abstract

G protein-coupled receptors (GPCRs) play key physiological roles in numerous tissues, including the heart, and their dysfunction influences a wide range of cardiovascular diseases. Recently, the notion of nuclear localization and action of GPCRs has become more widely accepted. Nuclear-localized receptors may regulate distinct signalling pathways, suggesting that the biological responses mediated by GPCRs are not solely initiated at the cell surface but may result from the integration of extracellular and intracellular signalling pathways. Many of the observed nuclear effects are not prevented by classical inhibitors that exclusively target cell surface receptors, presumably because of their structures, lipophilic properties, or affinity for nuclear receptors. In this topical review, we discuss specifically how angiotensin-II, endothelin, β-adrenergic and opioid receptors located on the nuclear envelope activate signalling pathways, which convert intracrine stimuli into acute responses such as generation of second messengers and direct genomic effects, and thereby participate in the development of cardiovascular disorders.

Stanley Nattel, Terry Hébert and Bruce Allen (shown from left to right) work at the Montreal Heart Institute, University of Montreal and McGill University. The cellular signalling and pathophysiological roles of nuclear-membrane GPCRs constitute one of their major research interests. Their backgrounds are in basic and clinical electrophysiology, molecular pharmacology, and signal-transduction. Artak Tadevosyan and George Vaniotis are doctoral students at the University of Montreal working at the Montreal Heart Institute. Using primary cells, subcellular fractionation, fluorescence imaging, and biochemical assays, their work has contributed to the elucidation of a role for nuclear-membrane localized G protein-coupled receptors in the regulation of gene-expression in normal and diseased hearts.

Introduction

The cardiovascular system is regulated by a diverse array of hormones, neurotransmitters and growth factors, many of which exert their biological effects through the stimulation of G protein-coupled receptors (GPCRs; reviewed in Tang & Insel, 2004; Regard et al. 2008). GPCRs comprise a diverse family of heptahelical transmembrane proteins involved in cellular signalling, and represent the largest class of gene products targeted by therapeutic agents (reviewed in Devi, 2001; Lefkowitz, 2004). Activated by a variety of ligands, including light, neuropeptides, nucleosides/nucleotides, hormones, calcium ions, chemokines, biogenic amines, proteases, lipids and fatty acid mediators, GPCRs transduce these signals into intracellular biochemical responses (see Hermans, 2003). The cellular/physiological responses mediated by the GPCRs result from interactions of the receptor with different heterotrimeric guanine (G) nucleotide-binding proteins composed of α, β and γ subunits. The principal Gα subunits are: G_q, which activates phospholipase C; G_s, which activates adenylyl cyclase (AC); G_i, which inhibits AC; and G₁₂ and G₁₃, which regulate RhoGEF (reviewed in Siehler, 2009). In their inactive state, Gα subunits are bound to GDP. Following receptor activation, Gα subunits undergo conformational changes leading to reduced affinity for, and hence dissociation of, GDP followed by the binding of GTP (Bourne et al. 1991). GTP binding is generally thought to result in dissociation of Gα-GTP from the receptor and the Gβγ complex, at least in vitro (Lambright et al. 1996). However, evidence indicates that this might not always be the case: GPCRs may also simply undergo a series of conformational changes that allow for effector activation (reviewed in Hebert et al. 2006). Both the Gα-GTP and Gβγ complexes can stimulate downstream effector molecules associated with several cellular signalling pathways, thereby modulating the production or release of second messengers (e.g. Ca²⁺, cGMP, cAMP, diacylglycerol (DAG), nitric oxide (NO), inositol-1,4,5-trisphosphate (IP₃)) which ultimately lead to cellular responses (reviewed in Hamm, 1998). Sequestration or metabolism of second messengers, intrinsic GTPase activity of Gα subunits, desensitization, internalization, and down-regulation of receptor responses by second messenger-activated or G protein-coupled receptor kinases (GRKs), β-arrestins, or regulators of G protein signalling (RGS) proteins, negatively regulate the initial wave of GPCR signalling (reviewed in Barren & Artemyev, 2007; Ribas et al. 2007; Kovacs et al. 2009).

In conventional GPCR signalling, receptors localized to the plasma membrane modulate the activity of membrane-localized enzymes and ion channels and/or the abundance or localization of second messengers within the target cell. Research conducted in recent years has revealed that endocytosed GPCRs may regulate distinct signalling pathways, suggesting that the biological responses mediated by GPCRs do not arise solely at the cell surface, but may result from the integration of extracellular and intracellular signalling events (Terrillon & Bouvier, 2004; Hanyaloglu & von Zastrow, 2008; Calebiro et al. 2009; Sorkin & von Zastrow, 2009). Recent experimental evidence suggests both nuclear localization and action of GPCRs. This distinct pool of GPCRs may play central roles in many cellular processes including regulation of gene transcription, ionic homeostasis, cellular proliferation and remodelling (reviewed in Boivin et al. 2008). Intracellular GPCRs may be constitutively active, or may be activated by ligands internalized from the extracellular space or synthesized within the cell. They can regulate signalling pathways distinct from those activated by the same receptor at the cell surface (reviewed in Re, 1999). This new paradigm for cellular signalling adds a further layer of complexity to studies of the function and physiological roles of GPCRs.

Nuclear GPCRs and their intracellular ligands

GPCRs localized to the nuclear membrane

Several GPCRs have now been shown to localize to the nuclear membrane, in a wide variety of different cell systems and tissues (Table 1). There are five major families of GPCRs, and members of the three main classes have been detected on the nuclear membrane. The remaining two classes, which we will not discuss further here, are the adhesion and the frizzled/taste receptor families, respectively (Millar & Newton, 2010).

Table 1.

Nuclear GPCRs in various organ systems

GPCR subfamilies	Cell/tissue	Species	Method of detection	Function	Reference
Family 1
Muscarinic acetylcholine receptor (mAChR)	Cornea, corneal epithelial & endothelial cells	Rabbit	Autoradiographic ligand binding, ligand binding assay, functional assays	↑ DNA and RNA Pol II activity, ↑ nuclear cGMP	Cavanagh & Colley (1989); Lind & Cavanagh (1993, 1995))
Apelin receptor (APJ)	Cerebellum, hypothalamus, transfected cerebellum D283 Med cells	Human	IHC, GFP-CM, Western blot*	Unknown	Lee et al. (2004)
Alpha- adrenergic receptor (α-AR)	Neonatal ventricular cardiomyocytes, adult ventricular α1ARKO cardiomyocytes	Rat, murine	Ligand binding assay, GFP-CM, subcellular ligand binding	α-AR; ERK activation, proliferation, survival	Buu et al. (1993); Huang et al. (2007)
β-Adrenergic receptor (β-AR)	Adult & neonatal ventricular cardiomyocytes	Rat, murine	Ligand binding, Western blot†, ICC†, functional assays	↑ Nuclear cAMP, G protein coupling, transcription, PKB activation	Buu et al. (1993); Boivin et al. (2006); Vaniotis et al. (2011b)
Angiotensin receptor (ATR)	Vascular endothelial cells (ECs), VSMCs,adult ventricular cardiomyocytes, human adrenal tissue, neuronal cells, hepatocytes,HEK-293T cells, cerebellar cortex	Rat, hamster	Autoradiographic ligand binding, ligand binding assay, GFP-CM, ICC†, Western blot†, IHC†, functional assays	↑ Nuclear µCa2+½, ERK & p38 activation, tyrosine phosphorylation, G protein coupling, transcription, proliferation	Lee et al. (2004); Re et al. (1984); Robertson & Khairallah (1971); Haller et al. (1999); Cook et al. (2006); Tadevosyan et al. (2010)
Cysteinyl leukotriene receptor 1 (CysLT1)	Colon tissue, colorectal carcinomas, Int 407 cells, CaCo-2 cells	Human	IHC†, ICC†, GFP-CM, Western blot†, functional assays	↑ Nuclear µCa2+½, ERK activation, DNA replication, proliferation	Nielsen et al. (2005)
C-X-C chemokine receptor type 4 (CXCR4)	Hepatoma, colorectal cancer HT-29 cells, adenocarcinoma, (NSCLS), HeLa	Human	IHC†, ICC†, flow cytometry	Tumorigenesis	Shibuta et al. (2002); Gobeil et al. (2006a); Spano et al. (2004)
Prostaglandin receptor (EPR)	Brain, myometrium, liver, transfected HEK293 & Swiss 3T3, cerebralmicrovessel ECs	Murine, human, porcine, rat	Radioligand binding, ICC†, IHC†, Western blot†, functional assays	↑ Nuclear µCa2+½, ERK & PKB activation, ↑ nuclear cAMP & IP3, ↑ eNOS mRNA,G-protein coupling, transcription, pro-inflammatory response	Bhattacharya et al. (1998); Bhattacharya et al. (1999); Gobeil et al. (2002); Provost et al. (2010)
Endothelin receptor (ETR)	Liver, VSMCs, adult ventricular myocytes, heart, brain, endocardial ECs	Rat, human, canine, sheep	Ligand binding assay, ICC†, Western blot†, radioligand-binding-SDS-PAGE, functional assays	↑ Nuclear µCa2+½, ↑ phosphorylation, transcription	Hocher et al. (1992); Boivin et al. (2006); Jacques et al. (2005); Vaniotis et al. (2011b)
Gonadotropin-releasing hormone Receptor (GnRHR)	Germline, intestine, HEK293HTR-8/SVneo	C. elegans, human, murine	IHC†	Acetylation & phosphorylation of Histone H3	Vadakkadath Meethal et al. (2006); Re et al. (2010)
Lysophosphatidic acid-1-receptor (LPA1R)	Hepatocytes, cerebral microvessel ECs, transfected HTC4 cells	Rat, porcine	Western blot†, ligand binding assay, IHC, ICC, functional assays	↑ Nuclear µCa2+½, ERK & PKB activation, iNOS activation, transcription	Gobeil et al. (2003, (2006b); Waters et al. (2006)
Melanocortin type 2 receptor (MC2R)	Adrenal Y6 cells	Human	Functional assays	Unknown	Doufexis et al. (2007)
Melatonin receptor (MT2)	Placental choriocarcinoma cells	Human	ICC*	Unknown	Lanoix et al. (2006)
Natural killer receptor (NKR)	Dorsal root ganglia neural cells, HEK 293, Ventral tegmental area (VTA) neurons	Rat	ICC†, IHC†, electron microscopy*	Unknown	Boer & Gontijo (2006); Lessard et al. (2009)
Opioid peptide receptor (OPR 1/2)	NG 108–15 neurohybrid cells (OPR1), cardiomyopathic adult ventricular myocytes (OPR2)	Human, hamster	IHC†, ICC†, ligand binding assay, functional assays, transcription	G protein coupling, PKC activation	Belcheva et al. (1993); Ventura et al. (1998)
Platelet activating factor receptor (PAFR)	Liver, cerebral microvascular ECs, brain, transfected CHO, HEK293 & COS1cells, brain ECs	Rat, porcine	Flow cytometry, ICC†, IHC†, radio ligand binding assay, GFP-CM, Western blot†, functional assays	↑ Nuclear DAG, ↑ nuclear µCa2+½,↓ nuclear cAMP, ERK phosphorylation,G protein coupling, transcription, PLC activation. Pro-inflammatory response	Miguel et al. (2001); Marrache et al. (2002); Zhu et al. (2006)
Prolactin receptor (PRLR)	Liver, splenocytes	Rat	Functional assays	Mitogenesis	Buckley et al. (1988)
Vasoactive intestinal peptide receptor(VPAC)	HT29 cells, breast cancer cells	Human	Ligand binding assay, radioligand-binding-SDS-PAGE	cAMP production	Omary & Kagnoff (1987); Valdehita et al. (2010)
Neuropeptide Y1 receptor (Y1R)	Pituitary gland, pituitary cells, endocardial EC	Rat, human	ICC†	Unknown	Chabot et al. (1988); Jacques et al. (2003)
Family 2
Parathyroid hormone 1 receptor (PTH1R)	Kidney, liver, gut, uterus, ovary, MC3T3-E1, ROS17/2.8, UMR106 and SaOS-2 cells, osteoclast-like multinucleated cells, LLC-PK1 cells	Rat, murine, human, deer	IHC†, ICC†, Western blot†, far Western blot GFP-CM	DNA synthesis, mitosis	Watson et al. (2000a, b); Faucheux et al. (2002); Pickard et al. (2006)
Family 3
Metabotropic glutamate receptor (mGluR)	Neuronal cells, transfected HEK293cells, neonatal striatal neurons	Murine	IHC†, ICC†, Western blot†, functional assays	↑ Nuclear µCa2+½, transcription, CREB phosphorylation	O’Malley et al. (2003); Jong et al. (2005, 2007)

Abbreviations for detection methods: IHC, Immunohistochemistry; GFP-CM, green fluorescent protein-confocal microscopy; ICC, immunocytochemistry. Antibodies used for verification of nuclear localisation by Western blotting or immunochemistry were targeted to either ^*epitope tagged versions of the receptor, or (more commonly) †endogenous receptors.

Class A, or rhodopsin-like receptors, contains 672 family members and are divided into 19 separate subfamilies which account for roughly 85% of all GPCRs (including olfactory receptors), representing the most diverse of the three classes (reviewed in Millar & Newton, 2010). Receptors from this class respond to a variety of different ligands that range from hormones and neurotransmitters to light itself, and all signal through interactions with heterotrimeric G proteins. Class A GPCRs also includes a relatively large number of receptors with no known physiological ligand, referred to as orphan receptors. Receptors in this family have been localized to the nuclear membrane, including members from almost every subfamily such as chemokine receptors (CXCR4), lysophosphatidic acid (LPA) receptors and opioid-like peptide receptors 1/2 (Ventura et al. 1998; Shibuta et al. 2002; Gobeil et al. 2003). Multiple GPCRs have been localized on the nuclear membrane of cardiomyocytes including angiotensin II type 1 receptor (AT₁R) (Tadevosyan et al. 2010), angiotensin II type 2 receptor (AT₂R, Robertson & Khairallah, 1971; Lee et al. 2004; Tadevosyan et al. 2010), α₁AR (Huang et al. 2007), β₁-AR, β₃-AR, and endothelin B receptor (ET_BR, Boivin et al. 2003; Boivin et al. 2006) to name a few (Table 1).

Class B GPCRs are primarily hormone receptors, including receptors for secretin and glucagon (Gkountelias et al. 2010). This family consists of three main subfamilies plus several as yet unclassified receptors. Members of this group have begun to be demonstrated at the level of the nuclear membrane of osteoclast-like cells, as in the case of the parathyroid hormone receptor (Faucheux et al. 2002).

Finally Class C, including metabotropic glutamate (mGlu) and γ-aminobutyric acid (GABA)-B receptors are primarily involved in the central and peripheral nervous systems (Jong et al. 2007; Ribeiro et al. 2010). Interestingly, of the three distinct groups of mGluRs with diverse roles, function and localization, only group I (e.g. mGluR₁ and mGluR₅) have been demonstrated on nuclear membranes of neuronal cells (Jong et al. 2005; Jong et al. 2007).

GPCR effectors in the nuclear membrane

Most of the downstream molecules normally associated with GPCR signalling at the plasma membrane have also been found in the nucleus (Fig. 1). This includes various G protein subunits: all three major forms of the α subunit as well as βγ complexes (Zhang et al. 2001; Gobeil et al. 2002; Boivin et al. 2005). Furthermore, a number of effector molecules have also been shown to be located at the level of the nucleus, including AC (Yamamoto et al. 1998), PLA₂, PLCβ₁ and PLCδ₁ (Schievella et al. 1995; Freyberg et al. 2001) as well as a variety of ion channels, including Ca²⁺ (reviewed in Bootman et al. 2009), K⁺ (Mazzanti et al. 1990; Gobeil et al. 2002; Quesada et al. 2002) and Cl⁻ channels (Tabares et al. 1991; Rousseau et al. 1996; Franco-Obregon et al. 2000). Moreover, the second messengers associated with downstream signalling events, including cAMP (Sastri et al. 2005; Boivin et al. 2006), IP₃ and DAG (Kumar et al. 2008), are also present at the level of the nucleus or generated in isolated nuclei. As such, all of the components necessary for GPCR signalling are present at the nuclear membrane. Additionally, several studies have shown that GPCRs and their effectors interact at the nuclear membrane. Nuclear bradykinin type 2, endothelin B, lysophosphatidic acid receptor-1, prostaglandin (EP₃), mGlu5 and AT1 receptors have been shown to modulate nucleoplasmic µCa²⁺½ (Devi, 2001; Gobeil et al. 2002, 2003; Bkaily et al. 2003a,b; Boivin et al. 2003; O’Malley et al. 2003; Savard et al. 2008; Tadevosyan et al. 2010). Consistent with a role in regulation of nucleoplasmic calcium transients, nuclear mGlu5Rs and AT1R also generate nuclear IP₃ (Kumar et al. 2008; Tadevosyan et al. 2010). In contrast, nuclear β₁-ARs stimulate adenylyl cyclase and increase nuclear µcAMP½ (Boivin et al. 2006). In addition to effectors, many of the systems involved in regulation of GPCR signalling are also be found in the nucleus, including RGS proteins (Burchett, 2003), β-arrestin-1 (Wang et al. 2003), as well as GRKs (Yi et al. 2002; Johnson et al. 2004). This indicates that nuclear GPCR signalling is tightly regulated, in a manner analogous to their cell surface counterparts, and hence reinforces the idea that these receptors are functionally relevant. Additionally, effectors associated with post G protein or G protein-independent GPCR signalling, including mitogen-activated protein kinase (MAPK) and PI3K/PKB pathways, are also present in the nucleus and GPCRs in the nuclear membranes are able to activate these various pathways (Marrache et al. 2002; Gobeil et al. 2006b; Vaniotis et al. 2011a). Furthermore, certain GPCRs, as is the case with the GABA_B receptors as well as G proteins themselves, have been shown to interact directly with transcription factors (Nehring et al. 2000; Robitaille et al. 2010; Sato et al. 2011), while nuclear bradykinin B2 and type I gonadotropin releasing hormone receptors appear to be able to regulate histone acetylation (Savard et al. 2008; Re et al. 2010), adding yet another potential method by which nuclear localized GPCR signalling might modulate gene transcription.

Figure 1. Known nuclear angiotensin II, β-adrenergic, endothelin and opioid receptor-activated signalling cascades in the heart.

Different receptor systems activate distinct Gα-GTP and Gβγ complexes which stimulate downstream effector molecules modulating the production or release of second messengers (e.g. Ca²⁺, cAMP, NO, IP₃) which may ultimately lead to changes in gene transcription. Considering the potential physiological functions of nuclear GPCRs in various cardiovascular diseases, future studies may provide a foundation for selective pharmacological interventions targeting these systems.

Ligands acting intracellularly

The term ‘intracrine’ ligand was coined to describe intracellular molecules binding to and activating intracellular receptors, including those on the nucleus. A list of known intracrine ligands in cardiovascular physiology is found in Table 2. These ligands are either synthesized within the target cell or are internalized from the extracellular space (Prochiantz & Theodore, 1995; Re, 1999). Enzymes involved in the generation and metabolism of phosphoinositides (Barlow et al. 2010) or processing peptide ligands (e.g. angiotensin converting enzyme) have also been localized to the nuclei of different cell types (Lucero et al. 2010). Further, these intracrine signalling loops are not restricted to GPCRs and may include a number of other classes of receptors including ALK4/ALK5, which are TGF-β superfamily receptors responsive to activin A (Gressner et al. 2008), as well as VEGF receptors (Lee et al. 2007). The potential actions of intracellular ligands include: (1) binding to intracellular membrane receptors and generation of second messengers (e.g. adrenergic, angiotensin, endothelin, insulin: Goldfine et al. 1982; Boivin et al. 2003, 2006; Tadevosyan et al. 2010), (2) binding to nuclei and nucleolus (e.g. angiogenin, fibroblast growth factor (FGF), parathyroid hormone-related protein (PTHrP): Bouche et al. 1987; Moroianu & Riordan, 1994b; Nguyen & Karaplis, 1998; Nguyen et al. 2001), (3) binding to chromatin (e.g. platelet-derived growth factor (PDGF), epidermal growth factor (EGF), nerve growth factor (NGF): Rakowicz-Szulczynska et al. 1986), (4) binding to receptors on the endoplasmic reticulum (ER) immediately following their synthesis (e.g. PDGF-B/v-sis: Fleming et al. 1989; Lee & Donoghue, 1992), and (5) binding to intravesicular receptors with consequent generation of second messengers (e.g. EGF: Wiley et al. 1998).

Table 2.

Intracellular ligands involved in cardiac physiology

Intracrines	References
Angiogenin	Moroianu & Riordan (1994b)
Angiotensin II	Re et al. (1984); Erdmann et al. (1996)
Angiotensin (l-7)	Camargo de Andrade et al. (2006)
Angiotensin converting Enzyme (ACE)	Cristovam et al. (2008)
Angiotensinogen	Sherrod et al. (2005)
Atrial natriuretic peptide (ANP)	Morel et al. (1988)
Brain-derived neurotrophic factor	Marmigere et al. (2001)
Dynorphin	Ventura et al. (1998)
Endothelial cell growth factor, platelet-derived (PD-ECGF)	Fox et al. (1995); Matsukawa et al. (1996)
Endothelin Converting Enzyme	Russell & Davenport (1999)
Epidermal growth factor (EGF)	Rakowicz-Szulczynska et al. (1986); Wiley et al. (1998)
Erythropoietin	Mtjavila et al. (1991)
Fibroblast growth factor (FGF; 1, 2, 3, 10)	Bouche et al. (1987)
Gonadotropin	Millar et al. (2004)
Growth hormone	Ardail et al. (2010)
Hepatoma-derived growth factor	Kishima et al. (2002)
Hepatopoietin	Lu et al. (2002)
Insulin	Goldfine et al. (1982)
Insulin-like growth factor binding Protein (IGFBP; 3, 5, 6)	Schedlich et al. (2000); Iosef et al. (2008)
Interferon-y (IFN-y)	Nagy et al. (2000)
Interleukins	Clevenger et al. (1990); Curtis et al. (1990)
Nerve growth factor (NGF)	Rakowicz-Szulczynska et al. (1986)
Neuropeptide Y	Wirth et al. (2005)
Oxytocin	Kinsey et al. (2007)
Parathyroid hormonerelated protein (PHTrP)	Nguyen & Karaplis (1998)
Phospholipase A2-I	Fayard et al. (1998)
Platelet-derived growth factor (PDGF)	Rakowicz-Szulczynska et al. (1986)
Renin/prorenin	Lee-Kirsch et al. (1999)
Thrombospondin-1	Ashton et al. (2004)
Thyroid-releasing hormone (TRH)	Garcia et al. (2000)
Transforming growth factor α (TGF-α)	Grasl-Kraupp et al. (2002)
Vascular endothelial growth factor (VEGF)	Li & Keller, (2000)
WNT13	Struewing et al. (2006)

Biosynthesis and actions of intracrine ligands

Hormones and growth factors fated to act within their target cells can be produced in various ways (Fig. 2). Intracrine ligands can be synthesized and targeted to the Golgi apparatus for secretion; however, they may also act intracellularly either before or after secretion (following reuptake; Re, 1999). In the case of peptide agonists, the intracrine gene product might arise from an alternative transcription initiation site, differences in mRNA maturation or translation leading to a gene product lacking secretory signals and consequently active only in the intracellular space (Kiefer et al. 1994; Lee-Kirsch et al. 1999; Xu et al. 2009). Alternative splicing can result in the generation of peptides that lack ER signal sequences or contain an altered hydrophobic sequence. For example, the mitochondrial genome does not contain the gene for renin, and hence the mitochondrial isoform of this peptide must be derived from the nuclear genome and must be synthesized in the cytosolic compartment and imported post-translationally into this subcellular organelle. In fact, the amino-terminal deletion of the ER signal sequence and 10 amino acids of preprorenin leads to trafficking to and retention within the mitochondria (Clausmeyer et al. 1999). A similar process might be envisaged for the nuclear translocation of intracellular ligands. Intramolecular events such as alternative translation initiation sites, the use of internal ribosomal entry sites and protein splicing through inteins that do not require additional enzymes for protein cleavage could also introduce protein variability and be a relevant aspect of intracrine biology and signalling (Amitai et al. 2004). This is particularly true for the Wnt family, where through the use of alternative promoters, alternative translation initiation sites and RNA splicing, three isoforms of Wnt 13 (Wnt 13 A, B, C) are synthesized. One of the isoforms, non-glycosylated Wnt 13B, traffics to the nucleus, increases the β-catenin/T-cell factor activity in HEK293 cells and promotes apoptosis in aortic endothelial cells (Struewing et al. 2006).

Figure 2. Sources of intracrine ligands.

Intracrine systems appear in various forms; hormones, cardiac growth factors, DNA-binding proteins and enzymes can all display intracrine functionality. There are many ways to generate intracrine ligands: (1) ligands can be synthesized within the cell and remain in the cytoplasm where they can be trafficked to organelles such as the nucleus (e.g. ET-1); (2) ligands may pass through channels or pores, or active transport across the plasma membrane (e.g. noradrenaline, NA); (3) ligands sufficiently small may cross the plasma membrane directly, while hydrophobic ligands may freely cross the plasma membrane at will (e.g. leukotrienes, LTs); (4) internalization of endogeneously or exogenously formed ligands following binding to cell-surface receptors and subsequent release from internalized endosomes within the cell (e.g. opioids).

Intracrine ligands have many functional similarities. Many of these ligands regulate the generation of second messengers, translocate and bind to the nucleus, and can be pro-angiogenic or anti-angiogenic (Moroianu & Riordan, 1994a,b; Wakasugi et al. 2002; Tsuji et al. 2005). Intracrine ligands might function in similar or different ways depending on the cell type in question and can serve to amplify or attenuate cellular signals coming from cell membrane events. Infusion of angiotensin-II (Ang-II) increases hepatic and renal angiotensinogen mRNA levels (Kobori et al. 2001a,b) and it has been proposed an amplification system exists whereby increases in intracellular Ang-II results in increased Ang-II expression. Consistent with this, Ang-II up-regulates angiotensinogen and renin expression when administered to isolated hepatic nuclei (Eggena et al. 1993, 1996). However, in isolated rat renal cortical nuclei, Ang-II has no effect on angiotensinogen mRNA, indicating that this system for amplifying the intracellular effects of Ang-II is cell-type specific (Li & Zhuo, 2008). In contrast, extracellular ET-1 reduces ET-1 mRNA levels in endothelial cells, suggesting that extracellular ET-1 may suppress intracellular ET-1 signalling (Farhat et al. 2008). Nuclear binding of PTHrP stimulates proliferation of vascular smooth muscle cells and is implicated in cell cycle progression, whereas its binding to receptors at the plasma membrane suppresses proliferation and mitogenesis (Massfelder et al. 1997; Re, 1999; Tovar Sepulveda et al. 2002). Although still poorly understood, the diverse functions exerted by agonists and hormones acting at intracellular GPCRs suggests that intracrine signalling may play important roles distinct from those of the same receptors activated at the cell surface.

Nuclear GPCRs in the heart

Among the GPCRs currently known to be present on nuclear membranes, few receptors have been studied in cells specifically derived from components of the cardiovascular system. Here, we will review four nuclear GPCRs where function and mode of action has begun to be elucidated.

Renin–angiotensin system (RAS)

The RAS was originally thought to be either a systemic or local loop, where Ang-II is synthesized within the circulatory system or at the tissue level. However, there is now experimental evidence in support of an intracellular RAS. Specific Ang-II binding sites, including the AT1R and AT2R, have been demonstrated on the nuclear membranes in cardiomyocytes (Tadevosyan et al. 2010), hepatocytes, and vascular smooth muscle cell lines (Re et al. 1984; Booz et al. 1992; Cook et al. 2006). Adult cardiac ventricular myocytes express angiotensinogen, angiotensin converting enzyme (ACE) and renin (Malhotra et al. 1999; Tsai et al. 2008). Renin is synthesized as a preproprotein, although two splice variants of renin have been identified that lack exon 1 (Clausmeyer et al. 1999, 2000; Lee-Kirsch et al. 1999; Sinn & Sigmund, 2000). These variants do not possess the precursor domain, which contains a motif that directs preprorenin to the ER and thus the secretory pathway, but instead encode prorenin which remains within the cell. Interestingly, rat heart and lungs express only a prorenin variant, termed 1A renin (Clausmeyer et al. 2000), suggesting an important role for intracellular renin activity in these tissues. In addition, cells are able to internalize and activate circulating prorenin (Peters et al. 2002). Moreover, ACE, which converts angiotensin-I to angiotensin-II, was detected by immunocytochemistry inside nuclei of cultured rat cardiomyocytes and fibroblasts as well as mesangial cells cultured from spontaneously hypertensive rats (Dostal et al. 1992; Camargo de Andrade et al. 2006). Thus, there are means whereby Ang-II can be generated intracellularly. In vascular and cardiac tissue, chymase can replace ACE in the biosynthesis of Ang-II (Miyazaki & Takai, 2006) and chymase immunoreactivity has been detected in cytosolic granules of mast cells, endothelial cells, and mesenchymal interstitial cells of the cardiac interstitium (Urata et al. 1993).

Role of nuclear ATRs in cardiac hypertrophy and remodelling

The role of extracellular Ang-II has been extensively studied in the cardiovascular system, where it is crucial in the progression of cardiovascular remodelling, during the development of heart failure (Xu et al. 2010). Experimental evidence now suggests that intracellular Ang-II (iAng-II) might also be involved in the development of cardiovascular disease. Intracellular renin isoforms are implicated in blood pressure and fluid homeostasis through their potential generation of Ang-II (Lavoie et al. 2006; Schefe et al. 2006). Transduction of neonatal rat ventricular myocytes with an adenoviral construct expressing Ang-II led to the rapid induction of cellular hypertrophy (Baker et al. 2004). Transgenic mice engineered for cardiomyocyte-specific overexpression of rat angiotensinogen, as driven by the mouse α-myosin heavy chain (MHC) promoter, show increased levels of iAng-II and also develop cardiac hypertrophy (Mazzolai et al. 1998). Similarly, mice injected with a plasmid containing an α-MHC-Ang-II construct developed left and right ventricular hypertrophy within 48 h with no change in blood pressure (Baker et al. 2004). Whereas ventricular Ang-II levels were increased in this model, plasma Ang-II concentrations were unchanged and administration of the AT1 antagonist losartan did not prevent hypertrophy, suggesting the effects were a result of intracrine actions of Ang-II (Baker et al. 2004).

There is also accumulating evidence that intracellular Ang II (iAng-II) levels change in various forms of cardiomyopathy, with diabetic cardiomyopathy being the form with the greatest evidence. Ang-II immunoreactivity is increased 3-fold in cardiomyocytes and cardiac endothelial cells from patients with type 2 diabetes and a further 2-fold in diabetic patients with hypertension (Frustaci et al. 2000). High glucose concentrations increase Ang-II synthesis in rat neonatal cardiac ventricular myocytes (Singh et al. 2007), adult rat cardiac ventricular myocytes (Fiordaliso et al. 2001), fibroblasts (Singh et al. 2008a), vascular smooth muscle cells (VSMCs) (Lavrentyev et al. 2007), and mesangial cells (Vidotti et al. 2004). In cardiomyocytes, VSMCs and mesangial cells, the glucose-induced increase in iAng-II production is mediated by chymase rather than ACE. Notably, whereas high glucose increased both extracellular and iAng-II production in cardiac fibroblasts and VSMCs (Lavrentyev et al. 2007; Singh et al. 2008a), only iAng-II production was increased in mesangial cells and cardiomyocytes. In contrast, whereas isoproterenol increased both extracellular and iAng-II production in fibroblasts, only production of extracellular Ang-II was increased in cardiomyocytes. One week following induction of diabetes in rats following streptozotocin treatment, iAng-II levels were elevated in cardiomyocytes and both iAng-II synthesis and cardiac fibrosis were blocked by inhibition of renin, but not ACE or angiotensin receptor blockers (ARBs; Singh et al. 2008b). Many of the observed iAng-II effects are not prevented by ARBs, presumably because of their structures, lipophilic properties, or affinity for intracellular Ang-II receptors. In addition, transgenic mice engineered such that cardiomyocytes chronically secrete Ang-II failed to develop hypertrophy (although fibrosis was evident). In contrast, injection of mice with a plasmid containing a construct directing cardiomyocyte-specific overexpression of Ang II (which was retained within cardiomyocytes) did develop hypertrophy, suggesting that elevated extracellular Ang-II alone is insufficient to induce hypertrophy (van Kats et al. 2001; Baker et al. 2004).

Role in cardiac electrical function

In cardiomyocytes, electrical synchronization and the flow of chemical messages between neighbouring cells is achieved through gap junctions. A decrease in gap junction conductance and alterations in the distribution of connexin 43 are partially implicated in the impairment of impulse propagation and generation of abnormal heart rhythms (Severs, 1994). Dialysis with renin or Ang-II in the presence of enalapril (an ACE inhibitor) decreases gap junction conductance in cardiomyocytes isolated from rat or cardiomyopathic hamster ventricles (De Mello, 1995, 1996). Cell uncoupling and abnormal changes in conduction velocity enhance the generation of re-entrant heart rhythms. Moreover, intracellular injection of Ang-II in single isolated ventricular fibres from failing hearts of cardiomyopathic hamsters shows a significant alteration in resting potential, action potential duration and refractoriness (De Mello, 2011). Intracellular administration of Ang-II to cardiomyocytes significantly reduced inward calcium current (I_Ca) and increased peak I_Ca density in the failing hearts of 4-month-old cardiomyopathic hamsters, suggesting that iAng-II may be involved in the regulation of cardiac contractility (De Mello, 1998; De Mello & Monterrubio, 2004). Similarly, dialysis of cardiomyocytes with renin plus angiotensinogen increased L-type Ca²⁺ currents in the cardiomyopathic hamster, an effect which was abolished by intracellular, but not extracellular, application of losartan (De Mello, 2006).

Role in cardiac gene expression

Although systemic haemodynamic responses to Ang-II often occur in seconds or minutes, the genomic responses to Ang-II, including cellular growth and mitogenic and proliferative effects, occur at later periods ranging from hours and weeks to months. In vascular smooth muscle cells, microinjection of Ang-II rapidly increased nuclear µCa²⁺½ in both microinjected cells and also in adjacent cells (Haller et al. 1999). Similarly, iAng-II induced IP₃-mediated Ca²⁺ influx, rat aortic contraction, cellular proliferation and activation of cAMP-response element-binding protein (CREB) through phosphorylation of p38 mitogen-activated protein kinase (Haller et al. 1996; Brailoiu et al. 1999; Filipeanu et al. 2001; Cook et al. 2004). Although the molecular mechanisms of gene regulation by nuclear Ang-II receptors are poorly understood, nuclear Ang-II receptors are clearly able to regulate transcriptional events of target genes such as NF-κB, TGF-β1, MCP-1, and NHE-3 (Li & Zhuo, 2008). Moreover, stimulation of cardiac nuclear Ang-II receptors by their cognate ligands modulates NO and IP₃, which were shown to participate in fine-tuning the regulation of local calcium release and de novo genomic transcription (Tadevosyan et al. 2011). A sustained increase of nuclear Ca²⁺ may trigger chromatin activation leading to DNA fragmentation and apoptotic cell death, and hence play a determining role in a variety of pathological and toxicological processes (Orrenius & Nicotera, 1994). It appears that nuclear ionic homeostasis may directly affect cytosolic ionic balance and therefore changes in nuclear receptor expression/activation would directly affect excitation–contraction or excitation–secretion coupling and, in the long term, contribute to diseases such as hypertrophy, hypertension and atherosclerosis.

Endothelin

Endothelin receptors (ETRs) are class A GPCRs. There are two ETRs, ETA and ETB, both of which bind the ligand ET-1 with high affinity. Moreover, these receptors are primarily localized in endothelial cells and cardiomyocytes, where they produce both positive inotropic and chronotropic effects. The ETRs interact primarily with Gα_q and signal through the activation of PLC, PLD and PLA₂ and the regulation of Ca²⁺ homeostasis (van Heugten et al. 1994; Husain & Abdel-Latif, 1999). ETB localizes to the nuclear membrane in rat adult cardiac ventricular cardiomyocytes (Boivin et al. 2003) as well as cultured human vascular smooth muscle cells (Bkaily et al. 2003a), while ETA localizes primarily to cell surface membranes in these cells. The precise function of nuclear ETB has yet to be elucidated; however strides have been made into understanding what role they might play in the adult heart.

Role in modulation of nuclear Ca²⁺ buffering capacity

Nuclear ETB regulates nuclear Ca²⁺ levels independently of the plasma membrane receptors, potentially via both PTX- and CTX-sensitive G proteins, through a putative R-type Ca²⁺ channel in human aortic vascular smooth muscle cells (Bkaily et al. 2008). The regulation of nucleoplasmic Ca²⁺ implies a potential role in both gene transcription and apoptosis. Furthermore, these receptors have been shown to inhibit transcription initiation through an as yet unidentified pathway (Vaniotis et al. 2011b), further reinforcing the notion that nuclear ETB has a physiological and/or pathophysiological role in the cardiovascular system. ET-1 is believed to act in an autocrine/paracrine manner in the ventricular cardiomyocyte (Tirziu et al. 2010). ET-1 can be produced, stored and secreted by adult ventricular cardiomyocytes under basal conditions as well as in response to various stimuli (Thomas et al. 1996). The fact that adult cardiomyocytes have the ability to produce ET-1 internally suggests a possible intracrine mechanism by which these nuclear-localized ETRs may be able to bind ligand and thereby regulate calcium homeostasis and gene expression. In addition, it has been shown that extracellular ET-1 may reduce the intracellular levels of ET-1 mRNA (Farhat et al. 2008), indicating a potential regulatory mechanism wherein extracellular ET-1 actually suppresses intracrine endothelin signalling.

Role in nuclear reactive oxygen species generation

Reactive oxygen species (ROS) are important pathological mediators of cardiovascular diseases (Magder, 2006). ET-1 induces a sustained increase of nuclear ROS in endocardial endothelial cells and in human vascular smooth muscle cells (Provost et al. 2010). In resting and as well as stimulated states, ROS concentrations are higher in the nucleus than in the cytosol and the effect is reversed with glutathione (Go et al. 2007). Nuclear membrane ET-1 receptor generation of nuclear ROS could contribute to modulation of nuclear Ca²⁺ homeostasis as well as nucleoskeleton reorganization and nuclear hypertrophy. Indeed, intracellular ROS and Ca²⁺ signalling are intrinsically linked by nitric oxide produced from arginine by NOS (Mazars et al. 2010). Isolated rat liver nuclei treated with an NO donor (SNP) display nuclear Ca²⁺ mobilization through the sGC/cGMP/PKG pathway, which in turn induces NF-κB activation through protein kinases and subsequent changes in gene expression levels (Gobeil et al. 2006b).

β-Adrenergic receptors

There are three known β-AR subtypes: β₁-AR, β₂-AR and β₃-AR. All three β-ARs are found in the heart, but β₁-AR represents the predominant subtype, comprising roughly 70% of the total β-AR density (Tevaearai & Koch, 2004). The principal endogenous agonists for these receptors are noradrenaline and adrenaline; however, the receptors differ with respect to their affinity for these two ligands. β-ARs signal primarily through Gα_s/i, and modulate AC activity, and as a result cAMP levels (Xiao et al. 2006). In addition to effects on second messenger generation in isolated nuclei, nuclear β-AR have also been shown to activate additional effectors, including MAPK and PI3K/PKB signalling pathways (Vaniotis et al. 2011b). Recently, β₁-AR and β₃-AR were shown to localize to the nuclear membrane in adult ventricular myocytes (Boivin et al. 2006).

Role in transcription

Similar to responses seen with plasma membrane receptor activation, β₁-AR activated adenylate cyclase and stimulated cAMP formation in isolated nuclei (Boivin et al. 2006). The activation of β₃-AR in isolated nuclei evoked a PTX-sensitive increase in transcription initiation (Boivin et al. 2006). Further studies showed that part of this increase is due to the up-regulation of 18S rRNA, whereas a concomitant decrease in mRNA for genes in the NFκB signalling pathway was also observed (Vaniotis et al. 2011b). Furthermore, nuclear β-ARs have been shown to activate the PI3K/PKB pathway (Vaniotis et al. 2011b). These results demonstrated a role for nuclear β-ARs in regulating gene expression and suggest potential roles in both physiological and pathological conditions. Unlike the ETRs, the endogenous ligand of the β-ARs is not produced or secreted by cells expressing β-ARs in the heart. In order for nuclear β-ARs to interact with either of their endogenous ligands, a mechanism for transporting ligand to the nucleus is required. While it remains unclear what this mechanism might be, noradrenaline is rapidly taken up into both neonatal and adult cardiomyocytes (Buu et al. 1993; Wright et al. 2008), and in neonatal rat ventricular cardiomyocytes norepinephrine is transported to the nucleus (Buu et al. 1993).

Nuclear opioid receptors

Cardiogenesis in the developing embryo involves cellular proliferation, differentiation and tissue organization. The role of intracellular opioid receptors and peptides in cellular growth and differentiation has been extensively studied. Dynorphin B, the end-product of the prodynorphin gene, is synthesised in and secreted from cardiomyocytes and acts as a κ-opioid receptor agonist (Ventura et al. 1994). Opioid receptors are present on myocyte nuclei, and binding of dynorphin to these receptors enhances prodynorphin gene expression, activates nuclear protein kinase C (PKC) and primes cardiogenesis in embryonic cells (Ventura et al. 1998; Ventura & Maioli, 2000; Ventura & Branzi, 2006). The nuclear binding of dynorphin B is elevated in cardiomyocytes from cardiomyopathic hamsters and promotes cellular growth and differentiation. The pluripotent embryonic P19 cell line mimics molecular and morphological events that occur during embryonic development, expresses prodynorphin and secretes dynorphin (McBurney et al. 1982). Administration of dynorphin to P19 cells increases expression of cardiac lineage-specific genes involved in normal heart development, including GATA4, Nkx-2.5, cardiac myosin heavy chain and cardiac myosin light chain, and promotes the development of beating colonies of cardiomyocytes (Ventura & Branzi, 2006). Studies also revealed that opioid receptors located on the nuclei of undifferentiated GTR1 cells also increase transcription of cardiac homeobox gene zinc finger containing GATA4 and Csx/Nkx-2.5 and increases prodynorphin gene expression, producing a positive feedback loop (Ventura et al. 2003). These findings suggest that opioid receptors on the nuclear membrane are involved in regulation of embryonic cardiac development.

Intracrine signalling in cardiac angiogenesis

Angiogenesis, the development of new vessels from existing vessels, is involved in the development and progression of ischaemic coronary artery lesions and restenosis (Tabibiazar & Rockson, 2001). The enhancement of angiogenesis through growth factors may represent a therapeutic strategy for cardiac revascularization in cardiovascular disease (Reilly et al. 2005). Proangiogenic factors include angiotensin, endothelin, angiogenin, PDGF, VEGF, bFGF and PTHrP. Many of these factors (e.g. angiogenin, bFGF) exert their angiogenic activity by trafficking to the nucleus and they may act via nuclear receptors, but direct evidence is lacking (Bonnet et al. 1996; Kishimoto et al. 2005). Neamine, which inhibits the angiogenic actions of VEGF and FGF2, has been also shown to block angiogenin trafficking to the nucleus and thus its angiogenic properties (Hu, 2001; Hirukawa et al. 2005; Yoshioka et al. 2006). Certain intracrine agonists traffic along the cytoskeleton, and consequently chemotherapeutic agents that interrupt the cytoskeleton are also antiangiogenic. Interruption of intracellular protein trafficking to the nucleus by interference with the cytoskeleton might be a strategy for the development of antiangiogenic compounds for the treatment of myocardial ischemia.

Conclusions

The discovery of GPCRs on the nuclear membrane and intracellular ligands opens new avenues in the physiology of GPCR signalling. Recent studies have begun to elucidate the roles played by nuclear GPCR signalling in both the healthy cardiovascular system and in response to pathological insults. Several GPCR ligands have now been shown to modulate nucleoplasmic second messenger levels and gene expression when acting on their cognate receptors located on the nuclear envelope. These receptors are well-positioned to play major roles in the effects of prolonged exposure to a specific ligand, while plasma membrane counterparts are likely to play a more pronounced role in acute responses to stimuli. This clearly isn't always the case, as many plasma membrane receptors have also been shown to regulate gene transcription. However, there is evidence suggesting that the profile of genes expressed in response to activation of receptors at the cell surface may differ from that induced by activating their nuclear membrane counterparts (Jong et al. 2009). Moreover, many GPCRs that have been identified at the nucleus are known to be involved in cardiovascular diseases, so it would not be a surprise to discover that their nuclear receptors might play a significant role in disease progression. Our improving understanding of the nature and role of cardiac nuclear GPCRs may have a major impact on our appreciation of physiological control mechanisms and thereby on therapeutic agent development in the future.

Glossary

Abbreviations

GPCR

G protein-coupled receptor

GRK

G protein-coupled receptor kinase

RAS

renin angiotensin system

RGS

regulator of G protein signalling