Regulators of G protein signaling in cardiovascular function during pregnancy

Abstract

G protein-coupled receptor signaling mechanisms are implicated in many aspects of cardiovascular control, and dysfunction of such signaling mechanisms is commonly associated with disease states. Investigators have identified a large number of regulator of G protein signaling (RGS) proteins that variously contribute to the modulation of intracellular second-messenger signaling kinetics. These many RGS proteins each interact with a specific set of second-messenger cascades and receptor types and exhibit tissue-specific expression patterns. Increasing evidence supports the contribution of RGS proteins, or their loss, in the pathogenesis of cardiovascular dysfunctions. This review summarizes the current understanding of the functional contributions of RGS proteins, particularly within the B/R4 family, in cardiovascular disorders of pregnancy including gestational hypertension, uterine artery dysfunction, and preeclampsia.

INTRODUCTION

Pregnancy induces profound physiological changes in nearly every organ system (30, 33). In particular, the cardiovascular system undergoes significant adaptations. Most of the changes, such as increases in heart mass (ventricular wall mass), occur within the first trimester. Myocardial contractility and cardiac compliance, as well as heart rate and stroke volume all increase significantly. Overall, these changes increase cardiac output by 30–50% and preferentially increase uterine blood flow by 10-fold compared with prepregnancy. Total blood volume increases similarly by 40–50% during pregnancy (33). In early to midgestation of normal pregnancies, vascular resistance decreases, accompanied by blood pressure reductions (75) in response to nitric oxide and prostaglandins (29, 167). Failure to reduce blood pressure in midgestation carries increased risk for stillbirth, proteinuria, hypertension, intrauterine growth restriction, and preeclampsia (PreE) (132). Thus, it follows that mechanisms that regulate vascular function in early and midgestation may represent causal mechanisms for, and potential therapeutic targets to prevent, such complications of pregnancy.

Women who develop complications during pregnancy, such as hypertension, often receive suboptimal treatment because of the acute changes in pharmacokinetics and pharmacodynamics of drugs during pregnancy (30, 33). Cardiovascular adaptations and organ-specific changes that influence pharmacokinetics and pharmacodynamics such as increased cardiac output during pregnancy result in a 50% increase in renal plasma flow, glomerular filtration rate, and creatinine clearance. Increases in maternal plasma volume cause a progressive decrease in plasma albumin, which significantly alters binding, volume, and distribution of drugs administered during pregnancy (42, 50). In addition, expression of drug-metabolizing enzymes such as cytochrome P450 are increased during pregnancy, and overall metabolism of drugs in the liver is increased (30, 42, 50).

Hypertension during pregnancy is associated with perinatal death, placental abruption, indicated preterm birth, small fetuses, and miscarriage. Treatment with antihypertensive drugs can improve these risks and decrease hypertensive morbidities like stroke. However, in more severe disorders of pregnancy that include hypertension, such as PreE, treatment with antihypertensive drugs does not stop progression of the disorder (1, 106). PreE affects 5–7% of all pregnancies and requires more acute treatments to avoid maternal and fetal death as well as preventing intracerebral hemorrhage, placental abruption, or intrauterine growth restriction (30). However, current treatments for disorders such as PreE do not stop the progression of the disorder. Furthermore, preventative strategies for PreE are modest.

Daily low-dose aspirin (81 mg per day) has been endorsed as a potential preventative intervention for women who are at increased risk of developing PreE. The risk reduction for development of PreE at this dose, however, is only ~10–15% (8, 112). Recently, the ASPRE trial (The Combined Multimarker Screening and Randomized Patient Treatment with Aspirin for Evidence-Based Preeclampsia Prevention) has further examined higher-dose aspirin as a preventative for PreE. The study identified women who were at high risk for developing preterm PreE by measuring uterine artery function via Doppler ultrasound. Women identified at high risk for PreE in the treatment group received 150 mg aspirin per day, and this treatment resulted in significantly reduced risk of developing PreE (149). Identifying women at highest risk of PreE by uterine artery Doppler measurements remains difficult, however, and is not feasible for many or most clinical settings. Therefore, continued identification of additional novel risk factors and therapeutic targets is still desperately needed for disorders such as PreE.

Increasing evidence implicates arginine vasopressin (AVP) (158, 160, 205), endothelin-1 (ET-1) (43, 186), and angiotensin II (ANG) (43, 78, 96) in the pathogenesis of pregnancy-associated cardiovascular disorders such as PreE. The major receptor subtypes that transduce signaling in response to these vascular hormones, including the V_1A (2, 165, 171, 182), V₂ (145), ET_A/ET_B (91, 111, 184), and AT₁ receptors (74, 173), respectively, activate G proteins and their second-messenger cascades in target tissues. Downstream mediators or modulators of G protein signaling, such as the regulator of G protein signaling (RGS) family (Fig. 1), may represent a common mechanism contributing to cardiovascular disorders of pregnancy and may offer new targets for more effective treatment of pregnancy-associated cardiovascular disorders. The purpose of the current review is to summarize the current understanding of the status and contribution of various RGS family members in cardiovascular function during pregnancy. We propose the general hypothesis that dysregulation or dysfunction of RGS family members may mechanistically contribute to the pathogenesis of cardiovascular disorders of pregnancy, and thus RGS family members may represent novel screening or therapeutic targets for such disorders.

Fig. 1.

Canonical function of regulator of G protein signaling (RGS) proteins. RGS proteins function to accelerate hydrolysis of GTP bound to the G_α subunit, thereby promoting reassembly of the G_α and G_βγ heterotrimer, and termination of second-messenger signaling. GPCR, G protein-coupled receptor.

G PROTEIN SIGNALING

G protein-coupled receptors (GPCRs) remain core targets of pharmaceutical drug development with over 40% of currently marketed drugs targeting GPCRs (45). These plasma membrane receptors regulate virtually all physiological processes including vision, taste, smell, neurotransmission, metabolism, blood clotting, cardiac output, blood pressure, and cell growth and development (208). Disrupted GPCR function or expression can lead to aberrant signaling, inactivation of the receptor, desensitization, and altered ligand binding, which can result in cardiovascular disorders, hypertension, diabetes, neurodegenerative disorders, anxiety, and cancer (166). All GPCRs contain a seven-transmembrane domain (7TM) consisting of seven transmembrane helices. Ligand binding to a GPCR induces a conformational change, and the receptor acts as a GTP exchange factor for G proteins by exchanging GDP for GTP on the G_α subunit of the G protein (166, 180). This GDP-to-GTP exchange on the G_α subunit stimulates dissociation of the G_α and G_βγ subunits, which then activate or inhibit various downstream effectors and result in a number of cellular responses. There are several subfamilies of G_α subunits including G_αq, G_αs, G_αi (including G_αz and G_αo), and G_α12/13. Each G_α subunit can differ in specific effector activation/inactivation (180).

Duration of activated GPCR signaling can be regulated many ways including posttranslational modification of the receptor by G protein-coupled receptor kinases (189). Phosphorylation of GPCRs results in recruitment of β-arrestins and internalization of the activated receptor (189). G proteins can be modified, as well, to regulate second messenger activation. Protein kinase C phosphorylates G_αz preferentially in the GDP bound state and prevents association of G_αz with the G_βγ subunits, thereby potentially leading to aberrant signaling through G_αz (201). In addition, blocking myristoylation of G_α subunits alters their affinity for the cell membrane, altering their activation and function (23, 140). Hydrolysis of GTP to GDP on the G_α subunit controls the duration of second messenger signaling, as this allows for their reassociation with the G_βγ subunits and termination of signaling (180).

RGS Proteins

In some tissues, termination of GPCR signaling activation is 100-fold faster than the predicted rate of GTP hydrolysis based on rates calculated from in vitro experiments (9). Additionally, rates of deactivation for G protein activated calcium and potassium channels can be much faster in vivo than in vitro (210). This indicates there are additional mechanisms by which G protein signaling can be regulated. Other such regulators of G protein signaling were discovered in yeast. Mutants of the gene SST2 were found to be hypersensitive to G protein coupled stimulation to induce growth arrest. Up to 200 times less ligand could be used to elicit an exaggerated growth arrest response compared with wild type, and SSST2P was subsequently defined as a negative regulator of G protein signaling. Other negative regulators were also discovered in Caenorhabditis elegans (EGL-10), and monocytes (GOS8) (66, 176, 210). Additionally, the nucleotide sequences of 15 mammalian genes from a cDNA library revealed a conserved 130-amino acid core domain within SSST2P, EGL-10, and GOS8. These proteins were defined as “regulators of G protein signaling,” and the core domain was, therefore, deemed the “RGS” domain.

Approximately 37 RGS proteins have been identified, and an additional 14 contain a nonfunctional RGS homology domain. Proteins with a functional RGS domain are classified into subfamilies (180). There are two types of classification systems. The first classification system organizes subfamilies by letter (A–F) (66, 214). The second classification system derives the name of a subfamily from the representative RGS member. For example, the RZ family is named for the representative RGSZ protein (66, 150). Each family grouping is based on amino acid sequence identity, structure, and function. For instance, members of the C/R7 family have a unique disheveled, EGL-10, pleckstrin (DEP) domain, known as the R7H domain. In contrast, the A/RZ and B/R4 RGS proteins have little more than a functional RGS domain. The main function of RGS proteins is to stabilize the G_α subunit in its GTP hydrolysis transition state and lower the free energy needed to complete the reaction. Hydrolysis of GTP allows for the G_α protein, bound to GDP, to reassociate with the G_βγ subunits and terminate the second messenger signaling cascade (Fig. 1). Though RGS proteins bind to the transition state, but not active conformation of G proteins, some RGS proteins (such as RGS4) interfere with interaction of activated G_α subunits with downstream effectors, functioning as effector antagonists (64).

Expression of RGS protein mRNA, due to transcription factor regulation, can be ubiquitous or cell-type specific. Rgs1, Rgs2, and Rgs3 mRNA is expressed in both cardiac myocytes and fibroblasts of the heart. RGS2 is even more ubiquitously expressed and can be found in smooth muscle and endothelial cells of the vasculature (127). However, some RGS protein expression is restricted, such as in the brain, where several RGS proteins are limited to specific regions of the brain (51, 211). Within the placenta, most of the 37 RGS genes are expressed at detectable levels throughout gestation and across various layers (Fig. 2, A–B). RGS genes are also expressed in circulating human CD4+ T cells in both the nonpregnant and pregnant states and in cord blood (Fig. 2C). Although the placenta and T cell functions (and dysfunctions) are clearly implicated in cardiovascular complications of pregnancy, the functions of RGS proteins during pregnancy have been largely unstudied to date, with the recent exceptions of RGS2 and RGS5 (72, 88, 96). Future studies to elucidate the role of RGS proteins in these tissues in pathogenic complications of pregnancy are clearly required.

Fig. 2.

Expression of mRNA for various RGS family members in placenta and immune cells during human pregnancy. In silico reanalysis of published processed data sets GSE9984 (108), GSE93839 (53), and GSE31976 (161) reveals differential expression of the various RGS family members in whole-thickness human placenta across gestation (A), in different layers of term human placenta (B), and in subsets of human CD4⁺ T cells from nonpregnant women, and maternal and cord blood samples at term (C). While many genes were either not detected or reported in circulating CD4⁺ cells in the GSE31976 data set, the concordance of relative expression levels of individual genes tested, across cell types, is notable.

The B/R4 Family of RGS Proteins

There are a total of 10 RGS proteins that make up the B/R4 family, RGS-1 to -5, -8, -13, -16, -18, and -21, as summarized in Table 1 (13, 180). These are the smallest RGS proteins in terms of size (with the exception of RGS3), and they primarily regulate G_αi or G_αq signaling pathways (13, 180). RGS1 functions mainly in the immune system, with high concentrations in germinal centers of B lymphocyte maturation (13, 70). RGS1 functions to regulate B lymphocyte maturation induced by chemokines and can also be found in T lymphocytes, natural killer (NK) cells, dendritic cells, and monocytes (4, 38, 94, 113, 169). The next RGS protein, Rgs2, is ubiquitously expressed, with expression documented in the CNS, heart, vasculature, kidney, immune system, lungs, bone and ovaries (13, 89, 98, 125, 163). The main function of RGS2 is to regulate G_αq and G_αi signaling (64, 89, 92, 180). Loss of RGS2 effects T lymphocyte proliferation and reduces inflammation responses to viral challenge. In addition, Rgs2-deficient mice exhibit hypertension with increased vascular resistance, hypertrophy of renal vasculature, and extended calcium signaling in the presence of vasoconstrictors (13, 124). As for RGS3, of the various isoforms, both short isoforms contain barely more than an RGS domain, while the longer isoforms contain other protein interaction domains including a PDZ domain (13, 90). Similarly to Rgs1, Rgs3 is mainly found in immune cells and regulates immune cell function (143, 144, 162). Alternatively, RGS4 expression is highest in the brain and heart (13, 44, 212). RGS4 has been implicated in regulating downstream signaling of opioid, cholinergic, and serotoninergic receptors and has been linked to schizophrenia by microarray and genotype single nucleotide polymorphism (SNP) analysis (109, 117). RGS5 expression appears to be highest in the vasculature as well, with high levels in pericytes and vascular smooth muscle cells. This RGS protein can be found in microvasculature as well, such as capillaries and arterioles, the aorta, and carotid artery (13, 16, 101). Like most other B/R4 members, Rgs8 is highly enriched in the brain. Various isoforms of Rgs8 have been implicated in receptor type-specific inhibition of G_αq signaling in which the shorter form of RGS8 yielded a less potent inhibition response compared with the other isoforms (156). RGS13 is found in T and B lymphocytes as well as mast cells (13, 170). Like most other B/R4 members, Rgs16 has a fairly ubiquitous expression and can be found in the heart, brain, liver, NK cells, and T lymphocytes (13). Conversely, RGS18 is mainly found in bone marrow-derived cells such as hematopoietic stem cells, megakaryocytes, platelets, granulocytes, as well as osteoclasts (13, 47, 119, 134). RGS18, like most RGS B/R4 members, regulates both G_αi- and G_αq-mediated signaling (13, 180). The last B/R4 member, RGS21, is found in sensory taste cells, while others have reported expression in various tissues such as brain, heart, placenta, lung, liver, and others (13, 102).

Table 1.

RGS B/R4 family members

Member	Interactions	Vascular	Receptor Association	Refs.
RGS1	Gαi/Gαq	no	ANG, sphingosine-1	(13, 89, 122)
RGS2	Gαi/Gαq (and AC)	yes	ANG, ET-1, AVP, α1/β2AR, muscarinic, CCK2	(15, 23, 64, 89, 107, 122, 128, 151, 152, 188, 191, 193, 198, 199)
RGS3	Gαi/Gαq/Gαz/Gαo	yes	ET-1, sphingosine-1, GnRH, M1/M3 muscarinic, βAR	(21, 89, 90, 120, 121, 139, 164)
RGS4	Gαi/Gαq/Gαz	yes	Kir3, ANG, ET-1, muscarinic	(81, 133, 151, 153, 185, 192, 209)
RGS5	Gαi/Gαq/Gαo	yes	ANG, ET-1, muscarinic	(10, 26, 54, 68, 141, 215)
RGS8	Gαi/Gαq/Gαo	yes	ANG, GIRK, M1 muscarinic	(79, 154, 179, 187)
RGS13	Gαi/Gαq/Gαo	yes	sphingosine-1, muscarinic, CXCR	(12, 85, 170)
RGS16	Gαi/Gαq/Gαo	yes	PAF, rhodopsin, GIRK	(14, 22, 86, 213)
RGS18	Gαi/Gαq	no	ANG, M1 muscarinic, OGR1, CCR2	(80, 119, 134, 207)
RGS21	Gαi/Gαq/Gαo	no	T2R	(20, 31, 194)

ANG, angiotensin II; AC, adenylyl cyclase; ET-1, endothelin-1; AVP, arginine vasopressin; AR, adrenergic receptor; CCK, cholecystokinin; GnRH, gonadotropin-releasing hormone; Kir, potassium inwardly-rectifying channel; GIRK, G protein-coupled inwardly-rectifying potassium channel; CXCR, C-X-C motif chemokine receptor; PAF, platelet-activating factor receptor; OGR1, Ovarian cancer G-protein coupled receptor 1; CCR2, C-C chemokine receptor type 2; T2R, (TAS2R) taste receptor type 2.

ROLE FOR B/R4 FAMILY MEMBERS IN THE CARDIOVASCULAR SYSTEM

The RGS B/R4 family members are highly expressed in the cardiovascular system in contrast to other RGS protein families (48). All of the RGS B/R4 members have been reported to be expressed at the mRNA level in cardiac tissue, with the exception of RGS13. RGS2, RGS3, RGS5, and RGS16 are the most highly expressed in the heart and vasculature. However, RGS protein expression varies in that many of the RGS B/R4 family members have not been detected (181). RGS8 has been detected in cardiac tissue, albeit at very low levels (181, 211). Rgs8 mRNA is primarily expressed in the brain, with the most densely expressed region being the cerebellum. Specifically, RGS8 seems to be localized to Purkinje cells where it presumably regulates G_αo and G_αi signaling (155, 157). RGS21 is also not expressed in cardiovascular tissue. In most recent studies it appears to be selectively expressed in circumvallate taste papillae or lung airway epithelial cells (31, 194). RGS21 was found to interact with G_αq, G_αo, G_αi, and G_αz subunits but not with G_αs subunits (194). However, since RGS21 is a newly recognized B/R4 family member and as few studies have been published on its function, a cardiovascular function of RGS21 cannot be ruled out.

Although some of the RGS B/R4 members cannot be detected in cardiovascular tissue directly, many of the B/R4 members are expressed in various immune cells (12, 137, 143). Both the innate and adaptive immune systems have been implicated in cardiovascular disorders such as hypertension and PreE (123). For example, T cells are implicated in kidneys of hypertensive rodents and humans (123, 146). In addition, deletion of T and B cells by RAG1 knockout in mice blunts hypertension induced by challenge with ANG II infusion or deoxycorticosterone (DOCA)-salt, and such animals are protected from renal damage (55). Transfer of T cells back into this model restores the hypertensive phenotype (55). In addition, B cells are implicated in hypertension. Specifically, autoantibodies to the AT_1A receptor have been detected in humans with PreE, and experimental models of PreE, and studies have demonstrated CD4⁺ T cells are required for production of these autoantibodies (36, 37, 97, 195). Similar antibodies can also be detected in humans with secondary malignant hypertension (46). A more comprehensive review of immune cell function/dysfunction in hypertension (123) highlights numerous studies where immunosuppression blunts hypertensive responses in humans and in various animal models. As such, although RGS B/R4 members may not be detected in cardiovascular tissue per se, their roles in immune cells may contribute to cardiovascular function and dysfunction.

Regulator of G Protein Signaling-1

Regulator of G protein signaling 1 (RGS1), as mentioned previously, is primarily expressed in lymphoid-derived immune cells such as B lymphocytes, T cells, NK cells, and dendritic cells. Expression of RGS1 (mRNA and protein) in these cells has been detected in peripheral blood, germinal centers, and in bone marrow (4, 94, 99, 103, 114, 142, 148, 169, 190). Due to high expression of RGS1 in lymphoid derived cells, RGS1 is implicated in many immune-related diseases such as celiac disease, multiple sclerosis, and multiple myeloma with mutations within the RGS1 gene being associated with development of these immune related diseases (73, 77, 148). In addition, a SNP in the 5′-region of the RGS1 gene (rs2816316) was recently associated with development of celiac disease and Type 1 diabetes. Biopsies of intestinal regions indicate RGS1 is expressed in intraepithelial lymphocyte compartments (73, 178). Interestingly, another SNP located within the 5′-region of RGS1 gene (rs2760524) is significantly associated with risk for developing multiple sclerosis. The multiple sclerosis SNP (rs2760524) has also been found to exist in strong linkage disequilibrium with the celiac disease SNP (rs2816316) (77). Within B lymphocytes, RGS1 mainly mediates G_αi-induced signaling and is linked to regulating migration of immune cells (58).

RGS1 mRNA expression can be observed in the heart at low levels, and it is more highly expressed in the aorta than heart tissue (25). However, this expression has been shown not to be in cardiovascular cells (smooth muscle or endothelial cells), but in resident immune cells of aorta and heart (137). Despite little expression in cardiovascular cells, RGS1 is still linked to cardiovascular function by its role in immune cells. RGS1 expression is significantly reduced in monocytes from patients with early-onset coronary artery disease, and increased RGS1 expression has been implicated in diabetes mellitus (19, 49, 175). In a study performed to analyze genetically dysregulated signaling pathways in obesity and atherosclerosis, microarray analysis of white adipose tissue from obese patients and atherosclerotic aortae demonstrated RGS1 was one of the highest ranked genes involved in inflammatory pathways of these disorders (115). High expression of RGS1 can be found in atherosclerotic plaques and aortic aneurysms (137). RGS1 was shown to regulate leukocyte recruitment, and mice lacking RGS1 had significantly reduced trafficking to the aortic wall after ANG infusion than mice with intact RGS1 and ANG infusion. In addition, loss of trafficking reduced the occurrence of ANG induced aortic aneurysm rupture in RGS1 knockout mice (137). Furthermore, RGS1 can functionally regulate ANG-II induced ERK activation, although RGS1 has little effect on ET-1-induced signaling (25). Although RGS1 is linked to cardiovascular-related disorders, its role appears to be primarily mediated through effects in immune cells.

Regulator of G Protein Signaling-2

Regulator of G protein signaling-2 (RGS2) is highly expressed in brain, kidneys, heart, and vascular smooth muscle cells (84, 89, 126). Within the brain, Rgs2 expression can be seen in the hippocampus, striatum, and cortex. Expression of RGS2 can be significantly induced by activation of synaptic activity and is implicated in regulation of both G_αq- and G_αi-induced M2 muscarinic acetylcholine receptor (AChR) activation (76). Both heterozygous and homozygous RGS2 knockout mice exhibit a hypertensive phenotype. This leads to a hypertensive phenotype that is indistinguishable between the heterozygous and homozygous knockouts (62). In addition, loss of RGS2 results in an increase in medial thickness of renal arterial walls, which is characteristic of humans that have chronic hypertension (62). After challenge with vasodilator agonists, the rate of blood pressure decrease was significantly lower in RGS2 knockout mice than wild-type mice. Furthermore, vasoconstrictor challenge of RGS2 knockout aortic vascular smooth muscle cells results in increased peak calcium responses and prolonged decline of intracellular calcium levels postchallenge (62). Hypertension mediated by loss of RGS2 is due to increased vascular tone as demonstrated by increased contraction of mesenteric resistance arteries, renal interlobar arteries, and increased hypertension responses to ANG (62, 63, 127). Although RGS2 mainly functions to inhibit G_αq signaling in vascular smooth muscle, RGS2 was shown to regulate G_αi/o signaling in the vascular endothelium as well. G_αi/o signaling inhibits vasodilation, and loss of RGS2 results in enhanced G_αi/o signaling and impaired vasodilation in response to acetylcholine (127). Alternatively, RGS2 was shown to alter G_αs signaling as well. However, it does not interact with G_αs itself. RGS2 has been demonstrated to inhibit the production of cAMP in olfactory membranes independently of G_αs, and so is thought to regulate adenylyl cyclase itself. Of the various adenylyl cyclases, RGS2 has been shown to regulate adenylyl cyclase III, V, and VI (174). Additionally, RGS2 has been implicated in regulating AVP-induced V₂ receptor (G_αs)-mediated signaling. In the kidneys, RGS2 expression is localized to parts of the nephron that express V₂ receptors, and kidneys from RGS2 knockout mice exhibit increased levels of cAMP, the main signaling molecule of V₂ receptors. In the same study, RGS2 knockout mice that were water restricted followed by acute water loading exhibited significantly lower urine volume output after water restriction and more urine volume output after acute water loading compared with wild-type mice. This effect was lost when acute water loading injections contained a V₂ antagonist, suggesting RGS2 functionally regulates vasopressin-induced G_αs signaling through the V₂ receptor in the kidneys (216).

Regulator of G Protein Signaling-3

As mentioned previously, the regulator of G protein signaling-3 (RGS3) gene generates multiple protein isoforms, and it is the largest B/R4 family member (13, 180). RGS3 mRNA is expressed in various different cells and tissues and functionally regulates G_αi/o and G_αq activation. Similar to RGS1, RGS3 can be found in immune-derived cells such as bone marrow derived and peripheral blood B and T lymphocytes as well as macrophages (143, 144, 162). Within these cells, RGS3 has been implicated in regulating B and T lymphocyte chemotaxis by lymphoid-derived chemokines (143). In addition, Rgs3 is expressed in various regions of the brain, as detected by in situ hybridization, and has been reported in several regions of the brain stem, cerebellum, dorsal thalamus, and hypothalamus (51). In addition, RGS3 expression can be found in several neuroblastoma cell lines and has been implicated in negative regulation of chemokine signaling in some of these cell lines (5).

RGS3 can be found in cardiovascular tissues as well and exhibits differential expression among the various isoforms. The PDZ-containing isoform of RGS3 can be found in the atria of the heart, whereas the long and short isoforms of RGS3 are more highly expressed in the ventricles of the heart (25). RGS3 expression is more highly localized within heart endothelial cells but has also been documented in smooth muscle cells, cardiomyocytes, and cardiac fibroblasts (25, 211). In smooth muscle cells, RGS3 regulates signaling induced through sphingosine 1-phosphate receptors and also attenuates signaling induced by ET-1 and ANG (25). In the presence of cardiovascular dysfunction RGS3 expression changes, RGS3 can be found in cardiomyocytes of adult rats and expression is significantly elevated in rat models of cardiac hypertrophy (104). In spontaneously hypertensive heart failure (SHHF) rats, RGS3 is significantly elevated in the failing myocardium (212). In contrast, the SHHF rat model that also develops congestive heart failure exhibits loss of Rgs3 mRNA and protein in failing hearts (212). In line with these findings, RGS3 mRNA and protein expression are significantly increased in myocardial samples from human end-stage failing hearts, suggesting a role in regulating cellular responses to hypertrophy in humans as well (131). Overall these studies suggest RGS3 may hold an important role in mediating cardiac hypertrophy and regulating cardiac function and failure.

Regulator of G Protein Signaling-4

Regulator of G protein Signaling-4 (RGS4) is expressed in various tissues, similar to RGS3. Expression of RGS4 mRNA and protein has been documented in immune cells such as T lymphocytes and B cell-derived cell lines where it is implicated in regulating chemotaxis signaling and differential regulation of G protein activation by varying localization within cells (40, 162). RGS4 expression by in situ hybridization reveals dense mRNA within specific areas of the brain as well. The densest regions include the neocortex, piriform cortex, caudoputamen and ventrobasal thalamus, and hippocampus (28, 51). In the cardiovascular system, Rgs4 is most densely located in the heart near the sinoatrial and atrioventricular nodal regions, the endogenous pacemakers for the heart (28, 181). RGS4 knockout mice exhibit reduced renal blood flow that can be reversed with ET-1 antagonists, implicating RGS4 function in regulating ET-1 activation as well (172). Additionally, exogenous expression of RGS4 in cardiomyocytes resulted in a dampening of ET-1 activation of PLC signaling, reducing myocyte contractility (181). In addition, similar to RGS3, RGS4 expression is increased in rat models of hypertrophy including cultured primary cardiomyocytes treated with growth factor and pulmonary artery banded mice. Increases in RGS4 can be seen in myocardium of SHHF rats as well (212). In humans, upregulation of RGS4 mRNA expression can be observed in patients with dilated and ischemic cardiomyopathy or end-stage heart failure (110, 131). RGS4 has been implicated in heart failure and hypertrophy, as well as in vitro cardiomyocyte contractility; however, more studies could be done to elucidate a role for RGS4 in control of cardiovascular function.

Regulator of G Protein Signaling-5

The highest expression of regulator of G protein signaling-5 (RGS5) appears to be within the brain and cardiovascular tissues. In situ hybridization reveals Rgs5 mRNA is densely expressed within three neuronal populations located in the paraventricular and supraoptic hypothalamic nuclei, as well as in the basolateral amygdala. Furthermore, Rgs5 has been detected within glial cell rich regions, suggesting RGS5 may also be expressed within glial cells of the brain (28). RGS5 has been implicated in the function of arteries, small renal arteries, and resistance arteries of the kidneys (48). RGS5 is mainly a G_αi/o and G_αq regulator and does not regulate or bind G_αs subunits (59, 215).

In the adult human, RGS5 is strongly expressed in peripheral arterial smooth muscle cells and regulates G_αi/G_αq activation in response to cardiovascular hormones, such as ANG (48, 197). RGS5 can also be found in mural cell populations, a mix of pericytes and smooth muscle cells. However, RGS5 expression is highest in medial smooth muscle cells of arteries (48). RGS5 has been implicated in regulating responses to ANG and ET-1, controlling calcium responses to these hormones as well as regulating activation of TGF-β and MEK/ERK pathways. Loss of RGS5 leads to enhanced activation of these pathways and results in cardiac hypertrophy in mice and increased renal hypertrophy and cardiovascular damage in the two-kidney one-clip model (48, 60, 100). In addition, SNPs in the RGS5 gene have been correlated with hypertensive populations. However, knockout models of RGS5 can display hypotension, hypertension, or normotensive phenotypes. In hypertensive models, loss of RGS5 leads to arterial hypercontractility, vessel stiffening, and increased vascular resistance. RGS5 has also been implicated in arteriogenesis. When femoral arteries were ligated to cause arterial growth, RGS5 expression in vascular smooth muscle cell significantly increased. In addition to being involved in cardiac tissue repair, RGS5 has been implicated in angiogenesis pathways as well. RGS5 is highly expressed in angiogenic vessels of solid tumors, in a model of pancreatic cancer, and is also highly expressed in angiogenic neovessels in brain tumors. It has even been suggested that RGS5 could be used as a pericyte marker for pathological neovascularization of tumors (48). Overall, RGS5 is a critical regulator of vascular function.

Regulator of G Protein Signaling-13

Regulator of G protein signaling-13 (RGS13) mRNA expression is mainly documented in lymphoid tissues with highest expression in B cells and monocytes (12). RGS13 activity has been characterized mainly within immune-derived cells. RGS13 was found to negatively regulate G_αi and G_αo signaling, as well as associate with G_αq subunits (85). In addition, RGS13 attenuates MAPK activation by G_αi and G_αq stimulation (85). Interestingly, RGS13, along with RGS2, typically displays nuclear localization. It was recently demonstrated upon activation of adenylyl cyclase and cAMP protein-dependent kinase (PKA), RGS13 translocates to the nucleus where it complexes with phosphorylated CREB to repress gene transcription (202). In addition, PKA activity phosphorylates RGS13, which results in stabilization of RGS13 and inhibits its degradation (204). In terms of vascular expression, there are only low levels of RGS13 found in the heart, and mostly found in cardiac fibroblasts (211).

Regulator of G Protein Signaling-16

Regulator of G protein signaling-16 (RGS16) mRNA and protein are expressed in the heart in both cardiac myocytes and fibroblasts, but at much lower levels than RGS2, -3, and -5 (87, 138, 211). RGS16 mRNA and protein are strongly expressed in the liver as well, and this expression can be controlled diurnally by feeding and fasting activity. In the liver, RGS16 regulates fatty acid oxidation metabolism (71, 136). Within cardiovascular tissue, RGS16 is implicated in regulation of G_aq activation. Bacterial endotoxin (LPS) is associated with impairment of myocardial contractility and acute heart failure. Treatment of cultured rat cardiomyocytes with LPS results in an increase of Rgs16 mRNA expression. Induction of Rgs16 expression by LPS resulted in blunted activation of PLC activity by ET-1. Additionally, LPS treatment of vascular smooth muscle cells and rat aortic rings can also induce Rgs16 mRNA expression (61). LPS treatment results in dampened contractile responses to vasoconstrictors such as ANG, suggesting RGS16 regulation of ANG receptors as well (61). Moreover, ET_B receptor has been implicated in regulating RGS16 expression, whereby activation of the receptor leads to increases in RGS16 expression (183). Overall, these studies implicate RGS16 in cardiac function.

Regulator of G Protein Signaling-18

Expression of regulator of G protein signaling-18 (RGS18) mRNA can be detected, but is very low, in the heart. RGS18 is not found in cardiomyocytes and exhibits only low expression in cardiac fibroblasts (47, 134, 211). Like RGS1, RGS18 is mostly expressed in immune cells. RGS18 mRNA and protein expression appears to be highest in bone marrow-derived cells, especially platelets. It can also be found in megakaryocytes and leukocytes (47). RGS18 functions to regulate G_αi and G_αq signaling in these cells but did not interact with G_αs, G_αz, or G_α12 (47). Additionally, Rgs18 is also expressed in osteoclasts where it functions as a negative regulator of osteoclast differentiation (80). However, to date, no studies have revealed a function for RGS18 within cardiovascular tissues.

GESTATIONAL DISORDERS AND THE RGS B/R4 FAMILY

Recently, various RGS B/R4 members have been implicated in vascular disorders of pregnancy. During normal pregnancy, the cardiovascular system changes dramatically. Cardiac output significantly increases, and by week 24 of gestation, the renin-angiotensin system is activated and ANG is secreted. AVP secretion is also activated, and both hormones act to increase water intake and retention. Retained fluids can be increased as much as 45%, resulting in a reduction in osmolality. Blood pressure also decreases up to 5–10 mm Hg during pregnancy, and most of these changes occur early during the 6–8 wk range of gestation (159). Despite activation of vasoconstrictor hormones, pregnancy displays an overall reduction in blood pressure thought to be mediated by relaxin, progesterone, or nitric oxide (27, 159). Furthermore, during pregnancy, there are attenuated vasoconstriction responses to ANG (32).

ANG receptors (AT₁) and AVP receptors (V_1A and V_1B, respectively) function by activating G_αq and G_αi/o pathways (39, 67), with the exception of AVP receptor V₂, which can activate G_αs (17, 93). Both hormones, in addition to other G protein-activating hormones such as ET-1, have been implicated in gestational cardiovascular conditions such as PreE (78, 82, 160). A disorder classified by elevated blood pressure and proteinuria, PreE affects ~10% of pregnancies each year and in some areas remains the leading cause of maternal and fetal death (41, 56). Yet, the cause of PreE remains unknown. Currently, there is a lack of adequate diagnostic tools to predict risk of PreE and the only cure is delivery of the fetus and placenta. RGS proteins function to regulate signaling by many of the cardiovascular hormones implicated in PreE, and therefore there is growing interest in the role of RGS proteins in the pathogenesis of cardiovascular complications of pregnancy such as PreE.

RGS5 has recently been linked to hypertensive disorders of pregnancy such as PreE (68). Myometrial arteries examined from hypertensive and PreE pregnancies display reduced expression of RGS5. Loss of RGS5 during pregnancy in mice induces gestational hypertension and reduces flow through uterine arteries, a key characteristic of PreE. In addition, RGS5 knockout pregnancies displayed an increased sensitivity to ANG challenge, and when treated with an AT₁ receptor antagonist, hypertensive phenotypes were ameliorated. Uterine arteries also displayed increased production of reactive oxygen species with ANG challenge (indicating oxidative stress) and an increased sensitivity to sodium nitroprusside (SNP, consistent with a nitric oxide signaling deficiency). Although there was no difference in fetal weight or placental weight with RGS5 deficiency, developmental effects were observed in the placenta. RGS5 knockout pregnancies displayed reduced labyrinth layer thickness and increased proliferation of spongiotrophoblast layers compared with normal pregnancies. Additionally, the gestational hypertension and vascular dysfunction observed in RGS5 knockout pregnancies could be attenuated by treatment with a peroxisome proliferating factor gamma (PPARγ) agonist. This supports links PPARγ, ANG, and RGS5 in the regulation of blood pressure during pregnancy (68).

RGS2 was also recently linked to PreE. A SNP (rs4606) in the 3′-untranslated region (UTR) of the RGS2 gene is associated with hypertension in selected human populations (168). This same SNP, rs4606, was correlated with increased risk of developing PreE, specifically in a population of overweight women (88, 96, 168). Women who carry rs4606 were at increased odds of developing PreE and cardiovascular disorders later in life (95). This SNP in the 3′-UTR of the RGS2 gene has been associated with decreased levels of RGS2 transcript expression (168). Interestingly, as noted previously, RGS2 heterozygous and knockout mice both display hypertensive phenotypes, suggesting that the RGS2 gene displays haplo-insufficiency. Moreover, both heterozygous and homozygous changes in the rs4606 C1114G SNP were linked to increased odds of developing human PreE.

Decreased uterine artery flow is a characteristic of PreE, and RGS2 function has recently been implicated in uterine artery flow control. Previously, it was demonstrated that reducing RGS2 increases blood pressure and decreases blood flow in kidneys (126). Most recently, Jie et al. (84) used Doppler ultrasonography to assess uterine artery blood flow in RGS2 wild-type, heterozygous, or knockout mice. In both heterozygous mice and homozygous nonpregnant female RGS2 mice, loss of RGS2 results in increased resistive index in uterine arteries. In addition, this group assessed myogenic responses (physiological response of arteries to differences in pressure) of uterine arteries. They determined that with increasing pressure, uterine artery myogenic responses in RGS2 knockout and heterozygous females was significantly increased as well as prolonged in the presence of calcium. In addition, RGS2 knockout and heterozygous mice had increased sensitivity to increases in intraluminal pressure. Lastly, they implicated loss of RGS2 in control of internal calcium release. Smooth muscle cells derived from wild-type RGS2 mice released reduced amounts of calcium compared with smooth muscle cells from RGS2 knockout mice (84). The same group also previously used RGS2 endothelial- and smooth muscle-specific knockout mice to delineate the mechanism by which RGS2 loss could cause cardiovascular dysfunction. Loss of RGS2 in smooth muscle of mesenteric arteries did not have much effect on vasodilation of the vessel. However, endothelial-specific reduction of RGS2 caused impaired vasodilatory responses to acetylcholine (ACh) in mesenteric arteries. In addition, loss of RGS2 in endothelial cells impaired vasodilation responses to endothelial-derived hyperpolarizing factor, and they could recover vasodilation responses by blocking G_αi/o activation. However, endothelial-specific and smooth muscle-specific loss of RGS2 had minor affects, overall, on blood pressure (127). Thus, although associated with gestational hypertensive disorders, the molecular and physiological mechanisms by which RGS proteins such as RGS2 and RGS5 contribute to these pathological states require further study.

CONTROL OF RGS PROTEIN FUNCTION

As highlighted throughout this review, RGS proteins are ubiquitously expressed. Yet recent studies have provided evidence for relatively specific interactions of various RGS proteins with individual GPCRs and G proteins. For example, RGS21 is known to interact with very few receptors, while RGS2 is known to interact with many GPCRs (Table 1). Furthermore, RGS1 regulates G_αi and G_αq signaling but not G_αo, whereas RGS4 and RGS16 preferentially interact with G_αq (Table 1).

How these RGS proteins mediate receptor-specific signaling is still under investigation. Selectivity of RGS proteins may be affected by subcellular localization. For example, if we focus on RGS2 as an exemplar member of the RGS family, this RGS protein can be found either in the nucleus or cytosol. Upon coexpression in HEK293T cells with β₂-adrenergic receptor or angiotensin AT_1A receptor, RGS2 is recruited to the plasma membrane, even in the absence of agonist treatment (151). Splice variants can also regulate RGS protein specificity. RGS3 has a splice variant containing a PDZ domain, which mediates interactions with the Ephrin-B receptor (105). NH₂-terminal structure can additionally alter RGS interactions with various GPCRs. RGS2 interacts with the M1 AChR through its NH₂ terminus (15). Furthermore, another study confirmed the NH₂-terminal domain of RGS2 was responsible for directly binding the α_1A-adrenergic receptor. However, this does not occur with the α_1B-adrenergic receptor (57). Instead of directly binding, a scaffolding protein known as spinophilin recruits RGS2 to the α_1B-adrenergic receptor/G_αq complex to regulate calcium signaling (199). Overall, subcellular localization, splice variants, NH₂-terminal structure, and other complexing proteins can mediate the specificity of RGS protein response to various GPCRs.

RGS proteins can also be modified or regulated directly, to alter GPCR signaling. NH₂-terminal posttranslational modifications can affect RGS protein interactions with GPCRs. Once again with RGS2 as an example, its protein half-life is regulated by the NH₂ end rule pathway. Some hypertensive populations carry mutations in the second amino acid of RGS2, which targets them for NH₂-terminal acetylation (Met-Gln-RGS2 to Met-Arg-RGS2) or NH₂-terminal Arginine degradation pathways. A mutation to leucine (Met-Leu-RGS2) targets RGS2 for both NH₂-end degradation pathways and significantly affects RGS2 protein half-life (135). In addition, RGS proteins can be phosphorylated to regulate activity. RGS5 is phosphorylated at a basal state in HEK293T cells, and with ET-1 activation, increases phosphorylation at Serine166. This phosphorylation interferes with RGS5 GAP activity but can be rescued with protein kinase C (PKC) inhibitors (116). In addition, RGS2 can be phosphorylated by PKC. Similar to RGS5, phosphorylation of RGS2 inhibits its GAP activity (34). Transcriptional regulation also controls RGS protein expression and function. A CRE element in the RGS2 promoter was found to regulate RGS2 expression in vascular smooth muscle cells, and treatment with forskolin increases RGS2 mRNA (203). Heat shock factor 1 has also been implicated in the transcription of RGS2. The RGS2 promoter contains a heat shock element. HeLa cells subjected to 1 h of heat shock had induced RGS2 mRNA, and chromatin immunoprecipitation revealed increased HSF1 at the RGS2 promoter (129). Similarly, in silico analyses reveal several other transcription factors that are also predicted to regulate RGS2 transcription (Fig. 3A).

Fig. 3.

Predicted regulators of RGS2 mRNA. A: transcription factor binding sites between −5 kb and +1 kb of the human RGS2 gene transcription start site (TSS), predicted using the MotifMap software (35), with false discovery rate (FDR) < 0.05. B: microRNA binding sites within the 681 base pair 3′-untranslated region (UTR) of the human RGS2 mRNA transcript, predicted using the TargetScan 7.1 software (3).

Epigenetic regulation of RGS proteins has also been reported. More specifically, DNA methylation of the RGS2 promoter, contributing to gene repression via a CpG island (Fig. 3A), has been demonstrated in various malignancies including ovarian cancer (18), bladder cancer (206), and prostate cancer (200), highlighting the importance of transcriptional fine-turning of these molecules. Alqinyah and Hooks (6) also recently published a review of the current state of knowledge regarding epigenetic regulation of RGS proteins.

Posttranscriptional mechanisms have also been demonstrated to regulate the function RGS proteins. MicroRNA (miRNA)-mediated mechanisms inhibit translation of proteins and reduce stability of mRNA transcripts by binding to the 3′-UTR of the RNA transcript (7). Of note, many of the RGS B/R4 members have been predicted, or evidenced, to be targets of various miRNAs (11, 24, 52, 147, 196). In silico analysis of the 3′-UTR of the RGS2 gene, again as a convenient example, identifies many conserved miRNA consensus binding sequences that are predicted to be targeted by miRNAs (Fig. 3B), with the most well characterized of these being miR-4717, miR-1271, miR-183, miR-96, and miR-22. MiR-1271 is highly expressed in the brain and shown to decrease Rgs2 expression (83). Furthermore, control of RGS2 by these miRNAs within the brain is associated with anxiety, panic disorders, and neurodegenerative disorders, such as Huntington’s disease. Studies have demonstrated links between RGS2 mutations, anxiety, and panic disorders (130, 177), and between miR-4717 or miR-22 and panic and anxiety disorders (69, 118), but direct mechanistic demonstrations through the miRNAs and RGS2 to mediate the disorders are yet to be demonstrated. Mutations in the RGS2 3′-UTR itself (such as rs4606) are located near these predicted miRNA binding sites. rs4606 is associated with anxiety disorders, hypertension, and PreE (96, 168, 177). It is possible that this SNP interferes with miRNA regulatory processes, but that effect also remains to be demonstrated experimentally. In addition to the 3′-UTR of the RGS2 gene, it is conceivable that mutations within miRNA themselves, or other regions of the RGS gene, may also play important roles in regulating expression or regulation of the RGS2 transcript in the disorders associated with this gene.

Thus, substantial work is needed to comprehensively understand developmental-, tissue-, cell-, and context-specific regulation RGS proteins, and more specifically, how dysregulation of these proteins may contribute to disease states such as pregnancy-associated cardiovascular diseases.

CONCLUSIONS AND FUTURE DIRECTIONS

RGS proteins are strongly expressed in cardiovascular and immune tissues, and its members have been variably implicated in cardiovascular diseases of the nonpregnant state. A few of the RGS B/R4 family members (such as RGS2 and RGS5) have been associated with the pathogenesis of hypertensive disorders of pregnancy. Substantial work is required to characterize the normal function(s) of RGS proteins in cardiovascular and immune adaptations to normal pregnancy and the ways in which dysfunctions in the expression, activity, or localization of RGS proteins may contribute to cardiovascular pathologies of pregnancy.

GRANTS

This work and the authors were supported by National Institutes of Health (NIH) Grants HL-134850, HL-084207, R21HD-091458, AA-025919, CA-161882; the American Heart Association (15SFRN23730000, 17PRE33660633, 18EIA33890055); NIH Reproductive Scientist Development Program; the University of Iowa Graduate College (2017 Post-Comprehensive Research Fellowship); and the University of Iowa Center for Hypertension Research.

DISCLOSURES

M.K.S. and J.L.G. have submitted patent applications describing potential use of the AVP system in the diagnosis and treatment of PreE.

AUTHOR CONTRIBUTIONS

K.J.P., G.D., and J.L.G. conceived and designed research; K.J.P. and G.D. analyzed data; K.J.P. and J.L.G. prepared figures; K.J.P. drafted manuscript; K.J.P., G.D., R.A.F., K.N.G.-C., M.K.S., and J.L.G. edited and revised manuscript; K.J.P., G.D., R.A.F., K.N.G.-C., M.K.S., and J.L.G. approved final version of manuscript.