Molecular mechanisms of neuronal nitric oxide synthase in cardiac function and pathophysiology

Abstract

Neuronal nitric oxide synthase (nNOS or NOS1) is the major endogenous source of myocardial nitric oxide (NO), which facilitates cardiac relaxation and modulates contraction. In the healthy heart it regulates intracellular Ca²⁺, signalling pathways and oxidative homeostasis and is upregulated from early phases upon pathogenic insult. nNOS plays pivotal roles in protecting the myocardium from increased oxidative stress, systolic/diastolic dysfunction, adverse structural remodelling and arrhythmias in the failing heart. Here, we show that the downstream target proteins of nNOS and underlying post-transcriptional modifications are shifted during disease progression from Ca²⁺-handling proteins [e.g. PKA-dependent phospholamban phosphorylation (PLN-Ser¹⁶)] in the healthy heart to cGMP/PKG-dependent PLN-Ser¹⁶ with acute angiotensin II (Ang II) treatment. In early hypertension, nNOS-derived NO is involved in increases of cGMP/PKG-dependent troponin I (TnI-Ser^23/24) and cardiac myosin binding protein C (cMBP-C-Ser²⁷³). However, nNOS-derived NO is shown to increase S-nitrosylation of various Ca²⁺-handling proteins in failing myocardium. The spatial compartmentation of nNOS and its translocation for diverse binding partners in the diseased heart or various nNOS splicing variants and regulation in response to pathological stress may be responsible for varied underlying mechanisms and functions. In this review, we endeavour to outline recent advances in knowledge of the molecular mechanisms mediating the functions of nNOS in the myocardium in both normal and diseased hearts. Insights into nNOS gene regulation in various tissues are discussed. Overall, nNOS is an important cardiac protector in the diseased heart. The dynamic localization and various mediating mechanisms of nNOS ensure that it is able to regulate functions effectively in the heart under stress.

Introduction

Nitric oxide (NO) has been identified as an important signalling molecule involved in a broad range of biological functions since the late 1970s (Arnold et al. 1977; Gruetter et al. 1979; Furchgott & Zawadzki, 1980), a discovery that led Robert F. Furchgott, Louis J. Ignarro and Ferid Murad to win the Nobel Prize in Physiology or Medicine in 1998. It is now clear that NO-producing enzymes, NO synthases (NOSs), are specified by three genes: neuronal NOS (nNOS or NOS1); inducible NOS (iNOS or NOS2), and endothelial NOS (eNOS or NOS3). Until recently, eNOS has been considered the only isoform of NOS in the myocardium and its regulation of a variety of functions in the heart is well documented (Shah & MacCarthy, 2000; Massion et al. 2003). In 1999, Xu et al. (1999) demonstrated that nNOS is constitutively expressed in the sarcoplasmic reticulum (SR) of cardiac myocytes and regulates sarcoplasmic Ca²⁺ ATPase (SERCA) reuptake of intracellular Ca²⁺. nNOS is also the only NOS isoform that is expressed in intrinsic cardiac neurons from autonomic nerves and ganglions (Mohan et al. 2000; Choate et al. 2001; Danson et al. 2005) and controls parasympathetic and sympathetic regulation of cardiac rhythm and contractility. Furthermore, nNOS is expressed in human coronary artery smooth muscle cells (Han et al. 2007) and maintains the basal blood flow (Seddon et al. 2009). Taken together, these findings confirm that nNOS is an important NOS in the heart that is expressed in all the vital fractions and plays important roles in regulating the rhythm, contractility and microcirculation of the heart (Fig. 1A).

Figure 1.

Distribution and structure of neuronal nitric oxide synthase (nNOS) protein in the heart

A, expression of nNOS protein in cardiac myocytes, coronary artery smooth muscle cells, cardiac ganglia, and sympathetic and parasympathetic nerves. B, structure of nNOS. Each monomer of nNOS contains an oxygenase domain (–COOH terminal) and a reductase domain (–NH2 terminal). Electrons from NADPH transfer from the reductase domain [via flavin adenine dinucleotide (FAD) and flavin mononucleotide (FMN)] to the oxygenase domain (heme iron), enable nNOS to catalyse the oxidation of l-arginine to l-citrulline and release NO. These two domains are linked and the enzyme is activated by Ca²⁺–CaM.

Neuronal NOS shares similarities with eNOS in structure and the logistics of NO production. Active nNOS is a homodimer in which each monomer is composed of an N-terminal oxygenase domain and a C-terminal reductase domain (Fig. 1B). NO production takes place in the oxygenase domain by catalysing l-arginine to l-citruline [endorsed by O₂ as the other substrate in the presence of cofactor tetrahydrobiopterine (BH4) and electron transfer from nicotinamide adenine dinucleotide phosphate (NADPH) via the flavin adenine dinucleotide (FAD)–flavin mononucleotide (FMN) axis to the heme iron in the oxygenase domain of the other monomer]. Ca²⁺-calmodulin is essential in linking FMN and heme iron to ensure efficient electron transfer by facilitating FMN donation of electrons and the alignment of the FMN and heme domains (Masters et al. 1996; Feng et al. 2014) (Fig. 1B). In cardiac myocytes, nNOS differs from eNOS in localization and translocation upon stimulation, post-transcriptional modification and the mechanisms mediating physiological and pathological functions in the cardiovascular system (Alderton et al. 2001; Massion et al. 2003; Belge et al. 2005; Zhang & Casadei, 2012). For example, eNOS is primarily located in caveolae (Feron et al. 1996) and mediates mechanical stress stimulation of intracellular Ca²⁺ release from ryanodine receptors (RyR) (Petroff et al. 2001). nNOS is located in the SR (Xu et al. 1999; Bendall et al. 2004) and in plasma membrane (Ueda et al. 2008) and regulates Na⁺ and Ca²⁺ homeostasis in cardiac myocytes. nNOS-derived NO modulates myocyte contraction and facilitates relaxation by targeting key elements of excitation–contraction coupling [e.g. phospholamban (PLN) or L-type Ca²⁺ channels (LTCC)]. This is understandable because NO has a short half-life (∼10 s) and limited diffusion area (due to abundant NO scavenging myoglobin and NO-interacting biochemicals and proteins); thus the compartmentation of nNOS or the translocation of nNOS to the vicinity of target proteins seem essential to its competent effects. In essence, the spatial confinement of nNOS and its dynamic translocation in response to stimuli ensure that it is able to regulate diverse intracellular signalling pathways and myocardial functions in normal and diseased hearts.

The function and mechanisms of nNOS regulation in the heart

nNOS in the healthy heart

Using gene knockout of nNOS (Barouch et al. 2002; Sears et al. 2003; Dawson et al. 2005; Saraiva et al. 2005) and cardiomyocyte-specific nNOS overexpression models (Burkard et al. 2007; Loyer et al. 2008; Burkard et al. 2010) in conjunction with nNOS inhibitors [e.g. S-methyl-l-thiocitrulline (SMTC), N5-(1-imino-3-butenyl)-l-ornithine (l-VNIO) or 7-nitroindazole (7-NI)] (Xu et al. 1999; Sears et al. 2003; Bendall et al. 2004; Seddon et al. 2009), it has been established that NO produced from constitutive nNOS provides an intrinsic regulatory mechanism of myocardial contraction and relaxation in both the healthy and the diseased heart. In the healthy heart, nNOS-derived NO attenuates basal cardiac inotropy by modulating the activities of LTCC in the plasma membrane to reduce the amplitude of intracellular Ca²⁺ transients (Sears et al. 2003) through S-nitrosylation- or cGMP-dependent mechanisms. By contrast, the activity of the cardiac Na⁺–Ca²⁺ exchanger is unaffected (Sears et al. 2003). In the SR, nNOS-derived NO facilitates myocyte relaxation by promoting SERCA reuptake of intracellular Ca²⁺ by increasing PKA-dependent (but cGMP/PKG-independent) PLN phosphorylation at serine16 (PLN-Ser¹⁶) and Ca²⁺-calmodulin-dependent kinase II (CaMKII)-dependent PLN phosphorylation at threonine17 (PLN-Thr¹⁷), subsequent to the inhibition of cytosolic protein phosphatase 2A (PP2A) and protein phosphatase 1 (PP1) activities by nNOS-derived NO (Zhang et al. 2008). nNOS-derived NO may activate SERCA either directly through S-nitrosylation (Burger et al. 2009) or indirectly through peroxynitrite-dependent S-glutathionylation (Adachi et al. 2004). The role of nNOS-derived NO in RyR activity is controversial because nNOS gene deletion has been associated with both increased RyR leak (Gonzalez et al. 2007) and decreased open probability of RyR (Wang et al. 2010) (Fig. 2). In addition, nNOS forms a macrocomplex with the plasma membrane Ca²⁺ pump (PMCA4b) and cardiac Na⁺ channel (SCN5A) linked by α-syntrophin (SNTA1) and is tonically inhibited by PMCA4b. Dissociation of nNOS with PMCA4b in an SNTA1 mutation liberates nNOS from inactivation and subsequently increases S-nitrosylation of SCN5A and late Na⁺ current. Under these conditions, nNOS is associated with SCN5A-dependent long QT syndrome (Ueda et al. 2008) (Fig. 2).

Figure 2.

Neuronal nitric oxide synthase (nNOS) protein expression, compartmentation and function in healthy cardiac myocytes

The diagram illustrates the localization of nNOS and target proteins and regulation. nNOS is located in the sarcoplasmic reticulum (SR), mitochondria, plasma membrane and possibly nucleus. nNOS interacts with ryanodine receptor (RyR), xanthine oxidoreductase (XOR), α-syntrophin [forming a macrocomplex with Na⁺ channel (SCN5a) and plasma membrane Ca²⁺-ATPase (PMCA4b)]. nNOS modifies the activities of various proteins either through S-nitrosylation (-SNO) or through PKA-dependent, cGMP/PKG-dependent phosphorylations of downstream proteins.

Accumulating evidence shows that nNOS-derived NO regulates cardiac function by targeting proteins other than Ca²⁺-handling elements. For example, results reported by our group and others show that nNOS tonically controls the activities of constitutive cardiac oxidases such as xanthine oxidoreductase (Kinugawa et al. 2005; Idigo et al. 2012) and NADPH oxidase (Zhang et al. 2009; Jin et al. 2012) and moderates the levels of intracellular superoxide and reactive oxygen species (ROS). As ROS (e.g. hydrogen peroxide) and reactive nitrogen species (RNS) (e.g. peroxynitrite) target protein kinases/phosphatases (Brennan et al. 2006; Burgoyne et al. 2007; Kohr et al. 2009), nNOS regulation of ion channel activity and intracellular Ca²⁺-handling proteins may be mediated by an array of post-transcriptional modifications such as NO-dependent S-nitrosylation, ROS-dependent oxidation and kinase/phosphatase-dependent phosphorylation. In essence, the genres of the target proteins of nNOS-derived NO determine the downstream post-transcriptional modifications. Furthermore, nNOS may affect myocardial function by regulating mitochondrial proteins. Indeed, nNOS-derived NO was reported to inhibit the mitochondria respiration chain, including complexes I, III and IV (Torres et al. 1995; Welter et al. 1996; Chouchani et al. 2013) and reduces mitochondrial oxygen consumption and affects cardiac metabolism (Fig. 2). Whether such regulation is detrimental or beneficial in myocardial function in the healthy or diseased heart remains to be determined. Recently, nNOS was shown to be recruited by α-syntrophin to the nucleus membrane of murine skeletal muscle C2C12 cells and human HeLa cervix carcinoma cells and to be involved in mitochondrial biosynthesis by increasing the respiratory chain complexes through the promotion of cAMP response element-binding (CREB) protein S-nitrosylation-dependent activation of peroxisome proliferator-activated receptor γ co-activator 1α (PGC-1α) and its downstream oxidative phosphorylation genes PGC-1α, NRF-1 and mtTFA (Aquilano et al. 2013). Whether similar mechanisms apply to the myocardium remains unknown. Overall, current evidence shows that nNOS-derived NO exerts cardiac functions by targeting essential proteins in various organelles in the myocardium.

nNOS in the diseased heart

Importantly, nNOS protein expression and activity are increased in the myocardium in the diseased heart, such as in ischaemia–reperfusion injury (Sun et al. 2006; Aragon et al. 2011), infarct (Takimoto et al. 2002; Bendall et al. 2004; Dawson et al. 2005; Burger et al. 2009), hypertrophy and heart failure (Damy et al. 2003, 2004; Niu et al. 2012), and exert manifold beneficial effects: nNOS-derived NO prevents diastolic dysfunction and increases β-adrenergic reserve, reduces left ventricular hypertrophy/dilatation/infarct size and protects the myocardium from arrhythmogenesis. In fact, recent findings by our group demonstrate that nNOS upregulation is an early event following pathogenic insult and during disease progression. We have shown that acute angiotensin II (Ang II, 1 μm) treatment to isolated left ventricular (LV) myocytes in vitro significantly increases mRNA/protein expressions and the activity of nNOS (Jin et al. 2012). In turn, nNOS-derived NO reduced NADPH oxidase production of superoxide and facilitated LV myocyte relaxation through cGMP/PKG-dependent (and PKA-independent) phosphorylation of PLN-Ser¹⁶ (Jin et al. 2012) (Fig. 3). Similarly, nNOS protein expression and activity were enhanced in LV myocytes from Ang II-induced early hypertensive rats (this is because Ang II infusion in vivo for 4 weeks did not develop hypertrophy and systolic/diastolic dysfunction detected using echocardiography). nNOS did not change contractility or LTCC activity, but did facilitate LV myocyte relaxation via myofilament Ca²⁺ desensitization mediated by cGMP/PKG-dependent phosphorylations of TnI-Ser^23/24 and cMyBP-C-Ser²⁷³ (Jin et al. 2013) (Fig. 3). Interestingly, PLN-Ser¹⁶ was increased and the kinetics of Ca²⁺ transient decay (Ca²⁺ reuptake via SERCA) were faster in LV myocytes from hypertensive rats; however, these changes were independent of nNOS or cGMP/PKG-dependent signalling (Jin et al. 2013). In line with these findings, nNOS protein expression was increased in the myocardium after 3 days and 6 days of Ang II infusion in rats (Tambascia et al. 2001; Moreno et al. 2002). These results demonstrate that nNOS upregulation is an early event during disease progression and nNOS plays protective roles in the heart under stress. In addition, the downstream targets of nNOS-derived NO are dynamic in order to facilitate effective function.

Figure 3.

Cardiac neuronal nitric oxide synthase (nNOS) during disease progression

nNOS is functionally expressed in healthy myocardium. From the early stages of disease progression (pre-disease) to hypertension, cardiac nNOS is upregulated and facilitates myocyte relaxation. The mechanisms mediating nNOS shift from PKA-dependent PLN-Ser¹⁶ to PKG-dependent PLN-Ser¹⁶ at pre-disease and PKG-dependent cMBP-C Ser²⁷³ and cTnI Ser^23/24 at hypertension. Myocyte contraction is not affected by nNOS.

In stellate ganglia neurons, the protein expression and activity of nNOS are reduced and intracellular [Ca²⁺]_i are increased in pre-hypertensive (young spontaneously hypertensive) rats. Conversely, targeted nNOS transfer into sympathetic neurons using a novel noradrenergic cell-specific vector reduced [Ca²⁺]_i (Li et al. 2013). As nNOS is important in modulating sympathetic neural transmission (Paton et al. 2002), these results suggest that nNOS in the autonomic nervous system and contracting myocardium may coordinate to facilitate cardiac protective effects (e.g. the positive lusitropic effect) in early hypertension.

In human dilated cardiomyopathy and in the murine heart after myocardial infarction, nNOS upregulation is associated with reduced interaction with RyR and increased binding with caveolin-3 (Cav3) in plasma membrane of cardiac myocytes (Bendall et al. 2004; Sun et al. 2006). This location may potentiate its modulation of Ca²⁺ flux in plasma membrane or NADPH oxidase inhibition. Indeed, nNOS association with Cav3 is enhanced following ischaemia–reperfusion injury and nNOS attenuates β-adrenergic stimulation of I_Ca (by increasing S-nitrosylation of LTCC), and reduces the Ca²⁺ transient and SR Ca²⁺ contents, particularly in female mice (Sun et al. 2006). It has also been shown that nNOS protects the heart from ventricular arrhythmias after left coronary artery ligation by increasing S-nitrosylation and moderating the activities of key Ca²⁺-handling proteins (e.g. LTCCα1c, SERCA2 and RyR2) and reducing Ca²⁺ influx, and diastolic and systolic Ca²⁺ transient amplitudes (Burger et al. 2009) (Fig. 4). Furthermore, nNOS exerts anti-arrhythmic effects by potentiating the peripheral vagal-induced modulation of heart rate (Choate et al. 2001; Heaton et al. 2005). In fact, stimulation of the vagal nerve increased nNOS production of NO in the left ventricle and was anti-arrhythmic (Brack et al. 2009, 2011). Given that systemic sympathetic nervous activity is elevated in the diseased heart (Coote, 2005), vagal control of cardiac rhythm by nNOS appears to exhibit significant clinical relevance in the attenuation of fatal ventricular arrhythmias (Herring & Paterson, 2009).

Figure 4.

Neuronal nitric oxide synthase (nNOS) compartmentation, function and mechanism in failing cardiac myocyte

nNOS protein expression and activity are increased in failing myocardium. nNOS translocates to plasma membrane and interacts with caveolin-3 (Cav3) but dissociates from ryanodine receptors (RyR). nNOS-derived NO changes the activities of Ca²⁺ handling proteins via S-nitrosylation. Cardiac-specific overexpression of nNOS is associated with increased nNOS localization in mitochondria and moderates mitochondria activity.

The role of nNOS-dependent signalling in the mitochondria or myofilament in the diseased heart is not clear. In mice conditionally overexpressing nNOS, protein expression of nNOS is increased in mitochondria and myocardial availability of NO is increased. Consequently, nNOS reduced oxidative stress by attenuating mitochondria ROS and xanthine oxidoreductase (XOR) (Fig. 3) and attenuated infarct size following ischaemia–reperfusion (Burkard et al. 2010). Interestingly, depending on nNOS localization, downstream target proteins, mechanisms and functions, contrasting results occur in similar mouse models conditionally overexpressing nNOS (Burkard et al. 2007; Loyer et al. 2008). For example, overexpressed nNOS that is located predominantly in the SR increased LTCC activity and PLN-Ser¹⁶ and resulted in faster decay kinetics of [Ca²⁺]_i (Loyer et al. 2008). By contrast, overexpressed nNOS in the plasma membrane increased its interaction with LTCC, reduced LTCC activity and exerted little effect on PLN-Ser¹⁶ and, therefore, slowed the decay kinetics of [Ca²⁺]_i (Burkard et al. 2007). These results further strengthen the importance of spatial compartmentation of nNOS and its downstream target proteins in determining the phenotypes of diseased myocardium.

The nNOS gene, transcription and regulation

nNOS gene, multiple promoters/transcription and nNOS splice variants

The human nNOS or NOS1 gene is located on chromosome 12 (12q24.2) and the locus is dispersed over a region of 240 kb. The nucleotide sequence corresponding to the mRNA transcript is encoded by 29 exons and translates into a protein of 1434 amino acids with a predicted molecular weight of ≈160 kDa (Hall et al. 1994; Wang & Marsden, 1995). The nNOS gene has nine unique exon 1 variants for transcript initiation in different tissues and each transcript is expressed from a unique 5′-flanking region (Wang et al. 1999b); this makes the transcriptional/expressional regulation of nNOS extremely complex. In essence, diverse and structurally different nNOS mRNAs are initiated by a variety of transcriptional units via mechanisms including discrete promoters, alternative splicing, cassette exon deletions or insertions, and the usage of alternative polyadenylation signals (Wang et al. 1999a, b1999b). In neurons and in muscle, nNOS promoters are clustered in genomic regions upstream from exon 2. Heterogeneous mRNAs from these promoters encoding the same nNOS proteins differ in both enzymatic characteristics and structural features (Nakane et al. 1993; Xie et al. 1995; Young et al. 1995). The activation of promoters downstream of exon 2 results in the expression of transcripts that utilize alternative translation initiation sites and produce truncated forms of the protein (e.g. nNOSβ or nNOSγ); nNOSβ (but not nNOSγ) retains enzymatic activity but lacks a major PDZ protein–protein interaction domain, which is responsible for targeting nNOS to plasma or nucleus membranes (Eliasson et al. 1997). Different transcriptional initiation, processing and translational efficiency, stability and localization make nNOS transcripts susceptible to various stimuli and produce diverse isoforms. So far, five splice variants of nNOS are described (nNOSα, nNOSβ, nNOSγ, nNOSμ and nNOS2); however, little is known about the variants’ specificity, functional relevance and mechanisms in cardiac myocytes. nNOSα and nNOSμ are known to contain PDZ domain and localize in plasma membrane linked with α-syntrophin. With a 34 amino acid insertion, nNOSμ differs from nNOSα in electron transfer rates, modulation of electron flow by CaM, and heme–nitrosyl complex formation (Panda et al. 2013). nNOSβ and nNOSγ do not contain PDZ domains and may be located in the cytosol.

nNOS gene regulation in tissues other than the myocardium

The genomic regulation (transcriptional and translational mechanisms) of cardiac nNOS is sparse, despite increased expression of nNOS mRNA and protein in LV or in atria following various pathophysiological insults (Takimoto et al. 2002; Damy et al. 2004; Danson et al. 2004; Ganzinelli et al. 2007). Understanding the transcriptional control of nNOS is important in defining the transcription/translation of specific nNOS splice variants in cardiovascular normal physiology and diseases. In neurons, skeletal muscle or vascular or gastrointestinal smooth muscle, nNOS transcription is regulated by transcriptional factors such as CREB (Sasaki et al. 2000), SP/ZNF families (Saur et al. 2002) or NF-kB (Li et al. 2007). CREB regulates the transcription of sirtuin in neuronal cells by binding to Sirt-1 chromatin; Sirt-1, in turn, is recruited by CREB to DNA and promotes CREB-dependent expression of nNOS (Fusco et al. 2012). Sequence analysis of 5′-flanking regions of the human nNOS gene revealed potential binding sites for AP-2, TEF-1/MCBF, CREB/ATF/cFOS, Ets, NF-1 and NF-kB-like sequences (Hall et al. 1994). All of the transcription factors are present in cardiac myocytes and therefore this information provides useful clues to the mechanism(s) involved in affecting nNOS gene regulation in cardiac myocytes.

Similarly, evidence of nNOS in tissues other than the heart may enhance understanding of nNOS splice variants and the functions in the myocardium. For example, it is well known that nNOSμ is upregulated after exercise in skeletal muscle in both animal models (Vassilakopoulos et al. 2003) and in humans (McConell et al. 2007). Expressions of nNOSμ mRNA and protein levels are correlated to angiogenesis after exercise (Huber-Abel et al. 2012). Recently, nNOSβ has been found to target Golgi and to play a pivotal role in maintaining the structure of murine skeletal muscle and post-exercise muscle strength, as well as in increasing resistance to muscle fatigue (Percival et al. 2010). These results suggest that the functional significance of nNOS splice variants refers to the regulation of diverse skeletal muscle functions. However, whether nNOS variants (nNOSα, nNOSβ and nNOSμ) play distinctive roles in cardiac myocytes is not yet clear. In the rat kidney cortex during late pregnancy, nNOSβ protein expression was increased, whereas nNOSα was reduced; consequently, antioxidant capacity was increased so that the renal cortex was protected from oxidative stress (Cunningham et al. 2013). Additionally, a high salt diet (from 0.4% to 4%) increased expression of nNOSβ protein to a greater extent than that of nNOSα in the inner medullary collecting duct of rat kidney (Hyndman et al. 2013). Similar changes in nNOSβ occur after right kidney removal and left kidney two-thirds ablation/infarction, and in kidney transplant models (Tain et al. 2008, 2011). By the same token, relaxin was found to increase nNOSβ expression in mouse proximal colon neurons, but to decrease nNOSα expression in smooth muscle cells, resulting in reduced muscle tone and increasing the amplitude of spontaneous muscle contraction (Baccari et al. 2012). The rise in nNOSβ was selectively suppressed by the β-adrenergic antagonist, nebivolol, and the angiotensin II type 1 receptor antagonist, olmesartan (Sasser et al. 2012), further suggesting the distinctive regulation of nNOS splice variants in response to pathophysiological stimulus.

nNOS compartmentation, binding partners and regulation in cardiac and non-cardiac tissues

Other than the essential components that assemble in stoichiometry to form the oxygenase and reductase domains (including co-factors and substrates), growing numbers of proteins are shown to represent ‘binding partners’ of nNOS. nNOSα and nNOSμ interact with proteins containing PDZ-domain through direct PDZ–PDZ binding or C-terminal–PDZ interactions. These interactions constrain nNOS localization to specialized cell compartments and to specific signal transduction pathways. Despite the importance of such interaction elsewhere, only a few proteins have been confirmed to interact with nNOS in cardiac myocytes and thus much of the current information is derived from findings in the brain or skeletal muscle. To date, nNOS binding proteins in cardiac myocytes are: (i) α-syntrophin; (ii) CAPON, and (iii) Cav3. The first of these, α-syntrophin, binds to nNOS via C-terminal–PDZ domains (in the β-hairpin finger structure region) and docks nNOS to the plasma membrane to form a macromolecular complex with PMCA4b and SCN5a (Ueda et al. 2008). CAPON is an adapter/regulator of nNOS in the brain. The C-terminal binding motif of CAPON interacts with the N-terminal PDZ domain of nNOS to activate nNOS (Jaffrey et al. 1998). In neurons, CAPON activation of nNOS is important in mediating neuronal excitotoxicity (Zhou & Zhu, 2009) and also facilitates iron uptake and neurotoxicity (Cheah et al. 2006). nNOS activation leads to S-nitrosylation of dexras1, which binds to benzodiazepine receptor-associated protein (PAP7) to bring dexras1 to divalent metal transporter (DMT1), an iron import channel, and induce iron influx (Cheah et al. 2006). Genome-wide analysis in patients with an abnormal QT interval has identified that the polymorphism of a common variant (rs10494366) of the NOS1 regulator, the NOS1AP gene (encoding CAPON) is strongly associated with variation in the cardiac QT interval (Arking et al. 2006); interestingly, the correlation is more significant in women than in men (Arking et al. 2006; Tobin et al. 2008). In fact, nNOS activated by CAPON has been implicated in the inhibition of LTCC, subsequently leading to shorter cardiac action potentials (Chang et al. 2008). nNOS increases its association with Cav3 in failing myocardium, but the functional relevance of this interaction remains unidentified. Cav3 may inhibit nNOS activity by preventing Ca-CaM binding to nNOS (Garcia-Cardena et al. 1997) and attenuate nNOS-derived cellular responses.

A number of nNOS binding proteins are reported in tissues other than myocardium. Representative proteins include: protein inhibitor of nNOS (PIN); a Ca²⁺-dependent protease, calpain; phosphofructokinase (PFK); various heat shock proteins (HSP90 and HSP70); nitric oxide synthase interacting protein (NOSIP); PSD95; NIDD, and carboxyl-terminal-binding protein (CtBP), etc. PIN physically interacts with nNOS and functions as an endogenous inhibitor of nNOS by destabilizing nNOS dimerization (Jaffrey et al. 1996); by binding to CBP, a fusion protein, CBP–PIN also reduces catalytic activity of nNOS without affecting its dimerization (Xia et al. 2006). Calpain binding leads to the acceleration of nNOS degradation, which determines the short lifetimes of nNOSα and nNOSμ (∼12 min and ∼50 min, respectively) (Laine & de Montellano, 1998). PSD95 links nNOS to N-methyl-d-aspartate receptor (NMDAR); NMDAR stimulation activates nNOS, which is critical to the postsynaptic activity of nNOS (Brenman et al. 1996; Doucet et al. 2012). Interestingly, CAPON is shown to compete with PSD-95 and PSD-93 for binding to nNOS, which results in the inactivation and cytosol localization of nNOS (Jaffrey et al. 1998). In nNOS-expressing synaptic vesicles, nNOS interacts with PFK and the product of PFK, fructose-1, 6-bisphosphate, may exert a neuroprotective effect (Firestein & Bredt, 1999). Furthermore, nNOS interacts with a transcription factor, CtBP, to target CtBP in cytosol and prevents its localization in the nucleus. By restricting CtBP from interacting with nuclear proteins (i.e. histone deacetylase and transcription factors), nNOS may maintain homeostasis of transcriptional repression and activation (Riefler et al. 2001). HSP90 binds to nNOS and enhances its Ca²⁺–CaM binding and activity (Song et al. 2001), and prevents nNOS aggregation and ubiquitination to uphold its stability and location (Corso-Diaz & Krukoff, 2010; Peng et al. 2012), all of which mediate protein–protein interaction and receptor–ligand interaction of nNOS to regulate cell function. By contrast, HSP70 promotes nNOS ubiquitination (Peng et al. 2012) and plays a contrasting role to HSP90. The identification of nNOS binding proteins and their functional regulations in special compartments are important in pinpointing the roles of nNOS in healthy and diseased myocardium.

Future perspectives: nNOS implications in the diseased heart

Cardiac nNOS regulates basal myocardial contraction and relaxation in the healthy heart. This is an important mechanism that protects against the progression and development of fatal heart diseases. Novel post-translational modifications of nNOS regulation are emerging and enhance our understandings of the important roles played by nNOS in cardiovascular disease. Aspects of nNOS gene regulation in response to pathophysiological stimuli, specific nNOS splice variants, and respective compartment/binding partners are known for various physiological systems, but cardiac-specific knowledge of nNOS and its downstream pathways in different heart diseases is required for the development of effective therapeutic strategies.

Glossary

eNOS

endothelial nitric oxide synthase

FAD

flavin adenine dinucleotide

FMN

flavin mononucleotide

iNOS

inducible nitric oxide synthase

LTCC

L-type Ca²⁺ channel

l-VNIO

N5-(1-imino-3-butenyl)-l-ornithine

7-NI

7-nitroindazole

NMDAR

N-methyl-d-aspartate receptor

nNOS

neuronal nitric oxide synthase

NO

nitric oxide

NOS

nitric oxide synthase

NOSIP

nitric oxide synthase interacting protein

PFK

phosphofructokinase

PIN

protein inhibitor of nNOS

PLN

phospholamban

RNS

reactive nitrogen species

ROS

reactive oxygen species

RYR

ryanodine receptors

SERCA

sarcoplasmic Ca²⁺ ATPase

SMTC

S-methyl-l-thiocitrulline

SR

sarcoplasmic reticulum

Biography

Dr. Yin Hua Zhang is Associate Professor at Seoul National University, College of Medicine, where she was trained as a Cardiac Electrophysiologist with Professor Yung E. Earm. She received further training in Bristol University with Professor Jules Hancox and in Oxford University with Professor Barbara Casadei. Her current research projects involve molecular mechanisms mediating cardiac protection from failing during disease progression (hypertension & metabolic syndrome). In addition, her team aims to reveal novel targets of nitric oxide and reactive oxygen species in regulating myocyte Ca²⁺ homeostasis, myofilament proteins, metabolism and contraction in the myocardium of cardiac arrhythmias and heart failure.

Additional information

Competing interests

None declared.

Author contributions

YHZ contributed to the writing of the paper and presented the research at the XXXVII Congress of the International Union of Physiological Sciences (IUPS) in Birmingham, UK. CZJ, JHJ and YW contributed to the work in YHZ's laboratory.

Funding

The present study was supported by a National Research Foundation of Korea (NRF) grant funded by the Ministry of Education, Science and Technology, South Korea (no. 2013R1A2A2A01068067), by the Brain Korea 21 Graduate Programme of the Korean Ministry of Education, Science and Technology, Seoul National University Hospital, the Korean Society of Hypertension (2013) and by the SK Telecom Research Fund (no. 3420130290).