RGS4 is a member of the R4 subfamily of RGS proteins, which contain only short N- and C-terminal protein sequences in addition to their conserved RGS domain. These proteins act as non-discriminatory GAPs for both Gα_i/o and Gα_q/11 (Table 1). RGS4 mRNA is enriched in the heart, but there is little to no detectable RGS4 protein expression (Posokhova et al., 2010; Yang et al., 2010). The lack of correlation between RGS4 mRNA and protein levels observed (Xie et al., 2009) may be due to a destabilizing N-terminal cysteine residue whose arginylation targets RGS4 for degradation through the ubiquitin-dependent N-end rule pathway (Davydov and Varshavsky, 2000; Lee et al., 2005). Using an RGS4 knockout mouse expressing LacZ behind the endogenous RGS4 promoter, selective enrichment of RGS4 mRNA in SAN was observed. This is an indirect means to measure RGS4 protein localization and may not reflect endogenous RGS4 levels due to differences in protein regulation (e.g., by the N-end rule pathway) between RGS4 and the exogenously expressed LacZ.

Though RGS4-null animals exhibited no difference in basal HR compared to their wild-type counterparts, these mice show enhanced bradycardia in response to systemic administration of the M2R agonist carbachol (CCh). In anesthetized animals, where parasympathetic tone is increased, RGS4-null mice did show reduced basal HR. These results are consistent with the fact that sympathetic drive dominates in conscious rodents as opposed to the greater parasympathetic stimulation of heart under basal conditions observed in humans. Knockout animals also displayed a larger change in HR upon administration of atropine, a M2R antagonist, suggesting these mice exhibit increased vagal tone. These results were replicated in isolated perfused hearts indicting that the enhanced bradycardia resulting from RGS4 loss is due to intrinsic alterations in cardiac automaticity (Cifelli et al., 2008).

SAN cells isolated from RGS4-deficient animals also showed increased sensitivity to CCh-induced inhibition of AP firing and changes in maximum diastolic potential (MDP). The correlation between inhibition of AP firing and MDP implicated enhanced GIRK current as a possible determinant of the enhanced M2R signaling RGS4-null cells. Sinoatrial myocytes isolated from RGS4 knockout mice also exhibited a loss of rapid M2R-activated GIRK current activation and deactivation kinetics as well as a loss of current desensitization with no difference in the magnitude or dose-dependency of the response (Cifelli et al., 2008). The increased time course of receptor deactivation and desensitization are consistent with a model whereby RGS4, acting as a GAP for Gα_i/o-coupled M2Rs, functions to inactivate Gβγ-induced I_KACh. Indeed, it was previously established that endogenous RGS proteins are required for proper activation and deactivation kinetics of agonist-induced GIRK currents (Lambert et al., 2010). RGS4 was shown to mediate these effects in the SAN, but RGS4 mRNA was not detected in atrial and ventricular cardiomyocytes indicating lack of RGS4 involvement in M2R signaling in these cell types (Cifelli et al., 2008). RGS4 appears to also mediate agonist and voltage-dependent GIRK channel relaxation (Fujita et al., 2000; Inanobe et al., 2001; Ishii et al., 2001), a process involving relief of channel pore blockade by intracellular cations and polyamines at hyperpolarizing potentials and increasing agonist concentration. As these experiments were performed in heterologous expression systems or isolated cells, the in vivo significance of this latter effect remains unclear. Nevertheless, the phenotype of enhanced parasympathetic tone in mice expressing RGS-insensitive Gα mutants may be due, in part, to loss of RGS4-mediated regulation of M2Rs in the SAN.

RGS6 is a member of the R7 subfamily of RGS proteins characterized by a distinct three domain structure. The conserved RGS domain confers their GAP activity directed specifically toward Gα_i/o subunits while the Disheveled-EGL-10-Pleckstrin (DEP) homology domain and Gγ subunit-like (GGL) domain facilitate binding of R7 family members to two accessory proteins: R7 family binding protein (R7BP), required for membrane targeting, and Gβ₅, an atypical G protein required for R7 family protein stability, respectively (Posner et al., 1999; Anderson et al., 2009). Unlike RGS4, RGS6 expression is detectable at appreciable levels in the SAN, AVN, atria, and ventricles of mouse hearts via PCR, western blot, and immunohistochemical staining (Posokhova et al., 2010; Yang et al., 2010). RGS6 is also highly enriched in the SAN and AVN along with cardiac GIRK channel subunit GIRK1 (Posokhova et al., 2010; Yang et al., 2010). Similar to the phenotype observed in RGS4-deficient mice, loss of RGS6 was associated with exaggerated bradycardia in response to CCh in conscious and anesthetized mice and isolated perfused hearts (Yang et al., 2010). Isolated perfused hearts from RGS6^−/− mice also exhibited significant CCh-induced AV block as evidenced by dramatic prolongation of the PR interval on electrocardiogram tracings.

Consistent with this phenotype of enhanced parasympathetic stimulation of heart, CCh-induced inhibition of spontaneous AP firing was exaggerated in SAN pacemaker cells isolated from RGS6^−/− animals (Yang et al., 2010). Like RGS4, RGS6 also appears to be required for CCh-induced GIRK channel gating kinetics. In atrial myocytes from wild-type mice, application of CCh elicited rapid I_KACh that showed significant desensitization over time, followed by rapid deactivation upon removal of agonist. Conversely, cells isolated from RGS6^−/− mice exhibited a significant reduction in the time course of activation and deactivation, as well as the extent of I_KACh desensitization (Posokhova et al., 2010; Yang et al., 2010). The observation that RGS6 is required for desensitization and rapid activation and deactivation of I_KACh in atrial myocytes and SAN cells is consistent with its role as a GAP for G_i/o. Indeed, such defects in GIRK desensitization and deactivation in RGS6-deficient atrial cells likely underlie the exaggerated bradycardia response to CCh in RGS6^−/− mice and isolated hearts. Therefore, like RGS4, RGS6 is also essential for modulation of M2R signaling in heart and loss of either protein recapitulates the phenotype of enhanced parasympathetic tone observed in mice expressing RGS-insensitive Gα_i2. The phenotypes of the various mouse models used to evaluate the role of RGS proteins in modulating autonomic control of heart are summarized in Table 2.

That fact that both RGS4 and RGS6 knockout animals exhibit similar phenotypes of enhanced M2R signaling imply that, while each of these proteins may function in a similar signaling cascade, they fail to completely compensate for each other’s loss. A number of hypotheses can be posed to explain these observations. It is possible that RGS4 and RGS6 regulate distinct receptor sub-populations, act as GAPs for different G proteins (e.g., Gα_i vs. Gα_o), or function to regulate M2Rs in separate cell types. Indeed, RGS6 is known to bind Gα_o with higher affinity than Gα_i isoforms in vitro (Hooks et al., 2003) implying it may selectively regulate M2Rs coupled to Gα_o. The observed lack of RGS4 expression outside of the SAN implies that RGS6 may be the primary negative regulator of M2R-mediated GIRK channel activation in AVN and extra-nodal tissues. In transfected cells expressing exogenous proteins, RGS4 appears to form a complex with the M2R and GIRK channel subunit GIRK1, though this interaction has not been confirmed in vivo (Jaen and Doupnik, 2006). Conversely, RGS6 fails to directly bind GIRK1 in vitro or in vivo (Posokhova et al., 2010; Yang et al., 2010), but does interact with GIRK4 in a heterologous expression system (Posokhova et al., 2010). These results suggest RGS4 and RGS6 may selectively regulate distinct GIRK channel subunits in heart, but the physiological significance of these observations remains unclear. Clearly, the generation of double knockout mice is required to facilitate investigations into the redundant, additive, or synergistic function of RGS4 and RGS6 in regulating parasympathetic stimulation of heart.

Many of the other RGS proteins expressed at the mRNA level in heart are capable GAPs for Gα_i/o (Table 1) and their role in regulating cardiac automaticity remains unknown. A few RGS proteins are emerging as potential additional players in the regulation of cardiac automaticity. For example, the enhanced susceptibility to atrial fibrillation (AF) in RGS2^−/− mice suggests involvement of additional RGS players in heart, though RGS2 is a selective GAP for Gα_q/11 (Heximer et al., 1999) and likely mediates these effects by regulating Gα_q-coupled muscarinic M3 receptors (Tuomi et al., 2010). There is also evidence for crosstalk between β₁AR signaling and M2R signaling in heart as isoproterenol, a β-adrenoreceptor agonist, induces a slowing of I_KACh deactivation kinetics, a process abolished by knockdown of RGS10 in isolated atrial myocytes (Bender et al., 2008). These results implicate RGS10 as a downstream target of the β₁AR in heart and suggest RGS10 might also function to regulate M2R signaling, though this effect has yet to be confirmed in vivo. Delineation of the distinct functions of each of the RGS proteins in controlling cardiac automaticity is essential to the design and utilization of targeted therapeutics aimed at modulating RGS protein function in heart. GPCR signaling pathways involved in cardiac automaticity and known to be regulated by endogenous RGS proteins in the SAN and atrial myocytes are summarized in Figures 2 and 3, respectively.

Thus far the specific RGS proteins (RGS4 and RGS6) involved in regulating Gα_i/o-coupled M2Rs have been identified. The enhanced bradycardia in mice lacking either of these RGS proteins appears to be determined, at least in part, by enhanced M2R-mediated I_KACh. Ventricular myocytes isolated from mice expressing RGS-insensitive Gα_i2 also exhibit sensitization to CCh-induced inhibition of contractility implicating endogenous RGS proteins as key regulators of this process (Waterson et al., 2011). Due to its expression in atria and ventricles, RGS6 might also regulate this inotropic response to M2R activation, though this effect has not been directly tested. Studies in cells expressing RGS-insensitive knock-in Gα alleles also implicated endogenous RGS proteins as key negative regulators of both the Gα_o-coupled A₁R and β₂AR signaling through Gα_i/o. To date, the consequences of loss of RGS-mediated regulation of cardiac GPCRs on cAMP levels and PKA-dependent modulation of I_f and I_Ca,L remains unexplored. As the cardioprotective effects of β₂AR stimulation occur via Gα_i/o-dependent mechanisms (Xiao, 2001; Patterson et al., 2004), identification of the specific RGS protein(s) responsible for negative regulation of this signaling cascade might allow for the development of targeted therapeutics aimed at specifically enhancing the protective, Gα_i/o-mediated effect without simultaneous activation of Gα_s.

Both conscious RGS6^−/− and anesthetized RGS4-null mice display mild resting bradycardia and a greater positive chronotropic response to administration of the muscarinic antagonist atropine implying these mice exhibit increased vagal tone (Cifelli et al., 2008; Posokhova et al., 2010). Thus far, enhanced vagal tone in these knockout mice has only been assessed using muscarinic cholinergic receptor blockade. Information gleaned from this method is limited and fails to assess the full physiological significance of enhanced vagal tone in these animal models. Fluctuations in parasympathetic nerve activity under resting conditions are the primary determinant of HR variability and such measures provide an additional index of the magnitude of parasympathetic modulation of HR (Chapleau and Sabharwal, 2011). ECG recordings in mice under basal conditions compared to recording upon administration of autonomic blockers can be used to assess sympathovagal balance, which would be expected to be skewed toward the parasympathetic division in mice lacking RGS6 or RGS4. The vagus is activated by multiple reflex pathways including the arterial baroreceptors and peripheral chemoreceptors to slow HR in response to increases in blood pressure or decreased blood oxygen, respectively (Chapleau and Sabharwal, 2011). RGS null mice may exhibit exaggerated bradycardia upon administration of a potent vasoconstrictor or under conditions of hypoxia, hypercapnia, or acidosis. Such measurements would give a more accurate approximation of the physiologically relevant regulation of cardiac M2Rs by RGS proteins without the administration of exogenous pharmacological agents.

In humans, RGS6 may indeed be essential for proper control of parasympathetic stimulation of heart as a single nucleotide polymorphism in the RGS6 gene is associated with deficits in HR recovery after exercise (Vasan et al., 2007). During periods of intense physical exercise, the sympathetic nervous system is activated with a corresponding inhibition of the parasympathetic division. Upon exercise termination, reactivation of vagal efferents is required for the rapid, exponential decline in HR (Chapleau and Sabharwal, 2011). Unsurprisingly, muscarinic receptor blockade compromises HR recovery after exercise (Coote and Bothams, 2001). Post-exercise HR recovery is used as a diagnostic and prognostic tool clinically as it has been shown to be predictive of morbidity and mortality in chronic heart failure and other cardiovascular diseases with faster recovery, indicative of increased vagal tone, associated with better health and lower mortality risk (Thayer and Lane, 2007; Tang et al., 2009). Indeed, shorter HR recovery is predictive of overall mortality independent of cardiovascular risk factors (Cole et al., 1999). Based on its essential role in regulating M2R signaling in heart, loss of RGS6 and the consequent increased vagal tone may promote faster HR recovery in mice. Such a result would implicate RGS6 as a novel target for therapeutic intervention in human patients with failing cardiovascular health.

Given their essential role in regulating parasympathetic activation in heart, alterations in RGS4 or RGS6 expression might provide symptomatic relief in patients with cardiac diseases involving impaired or excessive vagal activity. Amongst these disorders are sick sinus syndrome and AV block (Fu et al., 2007). Excessive vagal stimulation leads directly to AV block, an effect that is primarily dependent on increased GIRK current in mice (Drici et al., 2000; Hardouin et al., 2002) but can also result from inhibition of I_Ca,L by maternal auto-antibodies in cases of congenital heart block in humans (Garcia et al., 1994). Activation of RGS6 or RGS4 activity in heart would be expected to dampen parasympathetic stimulation and relieve these symptoms.

Atrial fibrillation is also known to lead to sinoatrial nodal dysfunction in humans and rodents (Sparks et al., 1999; Hadian et al., 2002; Hocini et al., 2003). Often GIRK currents are constitutively active in human patients with chronic AF (Dobrev et al., 2005) and a polymorphism in the Gβ₃ subunit of the heterotrimeric G protein complex resulting in reduced GIRK current is linked to decreased risk of AF in humans (Schreieck et al., 2004). Loss of M2R-mediated GIRK channel activation in mouse models, either through genetic ablation of cardiac specific GIRK channel GIRK4 or reduced membrane targeting of Gβγ, also resulted in reduced HR variability and risk for AF (Kovoor et al., 2001; Gehrmann et al., 2002). Clearly the extent of M2R-mediated GIRK current is a critical determinant of AF risk. Thus, activation of RGS protein-mediated negative regulation of cardiac I_KACh might be a viable means of treating AF in the clinic.

Heart failure can lead to diseases of the cardiac conduction system such as sick sinus syndrome and AVN dysfunction, which may result from both cardiac damage as well as negatively chronotropic pharmacological interventions including the use of β-blockers and digoxin (Monfredi et al., 2010). Impairment of the baroreceptor reflex is also associated with increased mortality in heart failure (Mortara et al., 1997; Schwartz and De Ferrari, 2011). Emerging evidence suggests that increased parasympathetic tone may be protective in heart failure. Donepezil, an acetylcholinesterase inhibitor used to enhance cholinergic signaling in patients with Alzheimer’s disease, can produce adverse cardiovascular events involving bradycardia and AV block (Hernandez et al., 2009) presumably due to enhanced ACh-mediated M2R signaling in heart. Nevertheless, Donepezil treatment regimens also reduce the risk for heart failure related mortality in human patients (Sato et al., 2010), suggesting that under certain pathological conditions, enhanced parasympathetic stimulation of the heart may be cardioprotective. These results have been recapitulated in rodent models of heart failure following aortocaval shunt placement (Handa et al., 2009) and during the acute and chronic phases post-myocardial infarction (Okazaki et al., 2010; Arikawa et al., 2011). The effects are likely due to increased parasympathetic modulation of heart as an increase in HR variability, a common measure of parasympathetic tone, was observed in Donepezil-treated animals (Okazaki et al., 2010). Similar observations have also been found using vagal nerve stimulation protocols in humans (Schwartz and De Ferrari, 2009) and rodents (Li et al., 2004).

The exact mechanism underlying the cardioprotective effect of enhanced vagal tone remains unclear, though it is believed to involve the anti-apoptotic effects of PI3K/Akt/HIF-1α signaling through atropine-sensitive muscarinic ACh receptors (Kakinuma et al., 2005). Such receptors can couple either Gα_i/o or Gα_q, which activates or inhibits Akt signaling, respectively, suggesting the cardioprotective effects of vagal stimulation may occur through Gα_i/o-coupled M2Rs known to be regulated by RGS6 and RGS4 in heart. Thus, the RGS6 and RGS4 knockout mice, both of which exhibit a phenotype of enhanced parasympathetic tone, would be expected to be protected from heart failure induced cardiac damage and mortality. Identification of the specific receptor signaling pathways involved would facilitate investigations into the role of RGS proteins in modulating these processes and their potential as therapeutic targets in the treatment of heart failure. Loss of RGS-mediated regulation of Gα_i2 also protects isolated, perfused hearts against ischemic injury including reduction in infarct size and improved recovery of contractile function. Interestingly, inhibition of mitochondrial and sarcolemmal ATP-dependent potassium channels and not the PI3K/Akt/GSK-3β and MEK/ERK signaling cascades blocked the cardioprotective phenotype in these animals. This study points to an entirely novel, anti-apoptotic pathway regulated by RGS proteins in the heart (Waterson et al., 2011). Thus, inhibition of cardiac RGS proteins might represent a viable means to enhance the protective effects of endogenous GPCR agonists such as opioids, bradykinin, and adenosine released from the heart during periods of ischemia (Cohen et al., 2000).

RGS4 mRNA and protein levels are up-regulated in the left ventricular myocardium of human patients with dilated and ischemic cardiomyopathy (Mittmann et al., 2002) and end-stage heart failure (Owen et al., 2001) suggesting that, despite the absence of its expression in atria and ventricles under basal conditions, RGS4 might be an important regulator of GPCR signaling in heart under specific pathological conditions. Exogenous RGS4 expression in ventricular myocytes also suppressed endothelin-1 dependent activation of phospholipase C and the resulting increase in myocyte contractility. This function, if it occurs in vivo, is in direct opposition to the effects of RGS4 on M2R signaling implying that RGS4 may have opposing effects on cardiac inotropy and chronotropy depending on the physiological context (Mittmann et al., 2002). To complicate matters even more, transgenic overexpression of RGS4 in murine ventricular cardiomyocytes exacerbates heart damage and mortality due to heart failure after transverse aortic constriction but transiently improved cardiac hypertrophy (Rogers et al., 1999). RGS4-mediated protection from cardiac hypertrophy and contractile dysfunction was also observed in mice overexpressing Gα_q (Rogers et al., 2001). Further investigation into the effect of RGS4 knockout on heart failure progression in mouse models might shed some light on these seemingly paradoxical observations. Indeed, a majority of the studies in mice were performed in transgenic animals overexpressing exogenous RGS4. As RGS4 expression is not detectable in appreciable levels in the myocardium under basal conditions (Posokhova et al., 2010; Yang et al., 2010), the physiological relevance of these findings is questionable. Due to the promiscuous regulation of both Gα_i/o and Gα_q/11-dependent signaling pathways by RGS4 and the oft times opposing functions of these GPCR cascades, RGS6, which exclusively interacts with Gα_i/o, might be a more specific target to enhance parasympathetic signaling in heart. Pathological upregulation of RGS6 in failing myocardium of human patients and mouse models has yet to be explored.

Doxorubicin (Dox) is an anthracycline that is among the most effective and widely used chemotherapeutic drugs to treat human cancers including leukemia, lymphoma, multiple myeloma, and breast cancer (Carter, 1975). However, the major obstacle restricting clinical utility of anthracyclines is their dose-dependent, life-threatening cardiotoxicity (Gianni et al., 2008). Dox induces acute cardiotoxicity, occurring immediately or weeks after treatment cessation and commonly presenting as sinus tachycardia that suggests autonomic dysfunction (Bristow et al., 1978; Lipshultz et al., 2008); and chronic cardiotoxicity, which usually develops in the first year following treatment and presents as left ventricular dysfunction and chronic heart failure (Bristow et al., 1978; Praga et al., 1979; Singal and Iliskovic, 1998).

The mechanism of Dox-induced cardiotoxicity is not fully understood. However, Dox-induced production of reactive oxygen species (ROS) has been implicated as critical for subsequent cardiac myocyte apoptosis and heart damage (Li and Singal, 2000; Zhou et al., 2001; Pacher et al., 2003). The chemotherapeutic efficacy of Dox relies on its ability to kill cancer cells by activating the ATM–p53 apoptosis pathways (Evan, 1997; Canman et al., 1998; Khanna et al., 1998; Bakkenist and Kastan, 2003; Iliakis et al., 2003), which is also responsible for Dox-induced cardiotoxicity as inhibition of components of this pathway protects against Dox-induced heart injury in mice (Liu et al., 2004; Shizukuda et al., 2005; Yoshida et al., 2009; Zhu et al., 2009). RGS6 plays an essential role in mediating activation of ATM and p53 induced by Dox in a ROS-dependent manner (Huang et al., 2011), implicating RGS6, whose expression is enriched in heart (Yang et al., 2010), as a key mediator of Dox-induced cardiotoxicity. Consistently, loss of RGS6 reduces apoptosis of heart cells and consequent heart damage in Dox-treated mice (Huang et al., 2011).

RGS6 has been shown to be induced by Dox in MEFs and cancer cells (Huang et al., 2011). As described above, RGS6 plays a critical role in regulating parasympathetic vagal control in the heart. Thus, this early induction of RGS6 by Dox may be involved in Dox-induced acute AF by interfering with the normal gating kinetics of GIRK current. Evaluation of RGS6 protein levels and I_KACh properties in SAN cells isolated from WT and RGS6^−/− mice treated with Dox is required to further test this hypothesis.

RGS6 is also required for Dox-induced ROS production. Loss of RGS6 impairs Dox-induced ROS generation in MEFs (Huang et al., 2011). It has been recently reported that Dox induces a ROS- and CaMKII-dependent Ca²⁺ leakage from sarcoplasmic reticulum (SR; Sag et al., 2011). Normal heart rhythm is initiated and regulated by oscillatory SR Ca²⁺ release (Ca²⁺ clock) in SAN cells, which by rhythmically activating Na⁺/Ca²⁺ exchanger inward current, prompts membrane ion channels to fire rhythmic APs (Maltsev and Lakatta, 2007). Therefore, RGS6 may promote Dox-induced AF by mediating Dox-induced ROS generation and consequent SR Ca²⁺ leakage disturbing the normal rhythm of intracellular Ca²⁺ clock in SAN cells. Interestingly, the ability of RGS6 to mediate ROS generation is not dependent on its interaction with G proteins (Huang et al., 2011; Maity et al., 2011), suggesting a unique role of RGS6, but not other RGS proteins, in controlling Dox-induced autonomic dysfunction. The possibility that RGS6 induction by Dox might underlie cardiac pathologies resulting from anthracycline chemotherapy is intriguing, but remains an untested hypothesis requiring verification from multiple laboratories for final confirmation.

Due to their role in regulation of GPCR signaling pathways in the cardiovascular system in both physiological and pathophysiological contexts, RGS proteins have emerged as potential therapeutic targets in the treatment of various heart-related diseases. While GPCR agonists and antagonists have been used successfully in the clinic for years, additional therapeutic benefit could be derived from the additive or singular use of compounds designed to enhance or inhibit RGS protein function. In diseases involving loss of parasympathetic tone or over-activation of sympathetic nerves, inhibition, or activation of either RGS4 or RGS6 would be expected to enhance or block signaling resulting from endogenous release of vagal ACh and subsequent slowing of HR, respectively. In the case of pathological upregulation of RGS6 and RGS4 that occur in Dox-induced cardiotoxicity and heart failure, respectively, RGS protein inhibition might help to mitigate heart damage.

Small-molecule inhibitors of RGS4 have been developed. The first, CCG-4986 binds irreversibly to RGS4 and results in covalent modification and permanent, allosteric inhibition of the protein making it less than ideal for clinical use and in vivo investigation into RGS4 function (Roman et al., 2010). A more recent compound has been identified with similar but reversible effects (Blazer et al., 2010). It appears that some degree of selectivity can be achieved despite the similarity in protein sequence between members of the various RGS protein subfamilies. Targeting regulatory mechanisms might also represent viable means to manipulate the expression of specific RGS proteins. In addition to their clinical use, such compounds might also have utility in investigating the physiological and pathophysiological functions of RGS proteins, singularly, or for multiple proteins, for which there exists no genetic knockout model or those whose loss, alone or in combination, results in embryonic or post-natal lethality.