Heterotrimeric G proteins play a crucial role in regulating signal processing to maintain normal cellular homeostasis, and subtle perturbations in its activity can potentially lead to the pathogenesis of renal disorders or diseases. Cell-surface receptors and accessory proteins, which normally modify and organize the coupling of individual G protein subunits, contribute to the regulation of heterotrimeric G protein activity and their convergence and/or divergence of downstream signaling initiated by effector systems. Activators of G protein signaling (AGS) are a family of accessory proteins that intervene at multiple distinct points during the activation–inactivation cycle of G proteins, even in the absence of receptor stimulation. Perturbations in the expression of individual AGS proteins have been reported to modulate signal transduction pathways in a wide array of diseases and disorders within the brain, heart, immune system, and more recently, the kidney. This review will provide an overview of the expression profile, localization, and putative biologic role of the AGS family in the context of normal and diseased states of the kidney.
The heterotrimeric G protein signaling system enables the kidney to readily adapt to an ever-changing environment by controlling cellular responses to various hormones and physical stimuli to maintain normal homeostasis of kidney function. The diversity of the heterotrimeric G proteins allows these distinct subunits to perform unique modes of action within the cell, including the mobilization of specific signaling molecules to intracellular organelles or microdomains, receptor phosphorylation and trafficking, and integration of signaling pathways to optimize specificity and/or intensity (Wettschureck and Offermanns, 2005). Historically, canonical G protein signaling involves the stimulation of cell surface G protein–coupled receptors (GPCRs), which function as guanine nucleotide exchange factors (GEFs) to facilitate Gα subunits to release GDP and promote GTP binding. As a consequence of the central role of cell-surface receptors in mediating G protein signaling, there has been enormous interest in the development of drugs to control the signaling output from this GPCR–G protein complex (Wettschureck and Offermanns, 2005; Brogi et al., 2014).
Over the past 20 years, however, the regulation of G protein signaling has become increasingly more diverse through the identification of an increasing number of accessory proteins. Initially, numerous groups discovered negative regulators of GPCR signaling, known as regulators of G protein signaling (RGS) (De Vries et al., 1995; Berman et al., 1996; Druey et al., 1996; Hunt et al., 1996; Koelle and Horvitz, 1996; Siderovski et al., 1996; Watson et al., 1996), which function as GTPase-activating proteins (GAPs) by accelerating the intrinsic GTP catalysis in G protein α-subunits. Subsequently, Cismowski et al. (1999) identified a novel family of distinct GPCR-independent regulators, known as activators of G protein signaling (AGS), using a pheromone receptor–deficient yeast screen. The AGS family members are a compilation of previously unrelated proteins to be grouped together based on protein structure and/or biologic function (i.e., modulate G protein signaling).
Unlike RGS proteins, AGS proteins regulate G proteins with a broader array of mechanisms by forming complexes with Gα or Gβγ independent of the typical heterotrimeric Gαβγ, and are categorized into four separate groups: GEFs, guanine nucleotide dissociation inhibitors (GDIs), Gβγ binding proteins, and Gα16 binding proteins. Although these modes of action by the AGS proteins have been extensively reviewed elsewhere (Sato et al., 2006a; Blumer et al., 2007; Blumer and Lanier, 2014), the biologic function in the kidney remains to be fully determined.
In the normal kidney, AGS proteins appear to be fairly quiescent and play a minimal role in the maintenance of renal function. However, there is emerging and compelling evidence that some of the AGS proteins can play a pivotal role during renal pathologies due to biologic or genetic stress (Nadella et al., 2010; Zheng et al., 2010; Regner et al., 2011; Kwon et al., 2012; Lenarczyk et al., 2014a,b; Potla et al., 2014; Wang et al., 2015). These functional changes in the renal phenotype are likely associated with the ability of AGS proteins to intervene at alternate sites during the typical activation/inactivation cycle of the heterotrimeric G protein complex, as well as the distinct distribution pattern of AGS proteins and their associated heterotrimeric G protein subunits within the nephron, glomerulus, and vasculature.
In this minireview, we will summarize the expression and spatial localization of G proteins and the AGS protein family in normal and pathologic diseased kidneys, and provide some biologic insight into their functional roles in the context of the kidney.
G Protein Subunit Localization in the Kidney
Currently, there are four distinct isoforms of Gα, specifically Gαs, Gαi/o, Gαq/11, and Gα12/13, as well as 5 Gβ and 12 Gγ subunits (Wettschureck and Offermanns, 2005), and their localization can be segment specific within the nephron and renal vasculature. A summary of this section on G protein subunit localization is shown in Table 1.
Biodistribution of Gα subunits in human tissues using Northern blot analysis demonstrated minimal to moderate mRNA expression of Gαs compared with Gαi (Brann et al., 1987), which was consistent with the expression of Gαs protein in the whole renal cortex and inner medulla from mice (Yu et al., 1998). Gαs mRNA is widely expressed throughout the renal cells in the nephron and blood vessels. Gαs mRNA was detected in isolated rat nephron segments, including medullary thick ascending limbs and collecting ducts using reverse-transcription polymerase chain reaction (RT-PCR) (Senkfor et al., 1993), and mouse glomeruli and proximal tubules using in situ hybridization (Williamson et al., 1996; Yu et al., 1998). At the protein level, Gαs was detected in the glomeruli (Brunskill et al., 1991; Yanagisawa et al., 1993, 1994; Nitta et al., 1994), proximal convoluted tubules (Brunskill et al., 1991; Stow et al., 1991), thick ascending limbs of Henle (Stow et al., 1991), collecting ducts (Stow et al., 1991), and preglomerular vessels from rats (Ruan et al., 1999).
Brann et al. (1987) demonstrated Gαi2 mRNA expression, but Gαi1 mRNA was undetectable in rat kidneys using Northern blot analysis. However, Hansen et al. (2003) detected all three isoforms of Gαi mRNA in the renal cortex using RT-PCR. To obtain more discrete information on the localization of Gαi subunits within the kidney, RT-PCR analysis was performed on isolated nephron and blood vessel segments from rats (Senkfor et al., 1993; Hansen et al., 2003). Gαi2 was highly identified as the predominant subunit in the medullary thick ascending limb and collecting ducts (Senkfor et al., 1993). Gαi1, Gαi3, and Gαo were undetectable in the nephron segments (Senkfor et al., 1993), but Gαi1 was robustly expressed in preglomerular vessels (Hansen et al., 2003).
Localization of Gαi subunits at the protein level was fairly consistent with spatial expression detected with mRNA assays. Gαi2/i3 was detected in the rat glomeruli using immunohistochemistry (Brunskill et al., 1991; Yanagisawa et al., 1993, 1994) and all Gi1/i2/i3 isoforms were detected in isolated glomerular membranes (Nitta et al., 1994) and preglomerular blood vessels from rats using immunoblot analysis (Ruan et al., 1999). Using immunohistochemistry with rat kidneys, Gαi subunits were segment specific within the nephron, where Gαi1 was localized to the thick ascending limb, including the macula densa, and papillary epithelial cells; Gαi2 was found in the collecting duct cells; and Gαi3 was in the S1 segment proximal tubule cells and macula densa cells in the thick ascending limb of Henle (Stow et al., 1991). Any differences between mRNA and protein localization of the Gαi subunits may be attributed to the relative nonselectivity of the antibodies compared with the nucleotide probes/primers used in the previous studies. In most cases, however, the mRNA and protein results provided confirmatory evidence that one, if not all, of the isoforms exist within specific cell types in the kidney.
Gαo mRNA expression was undetectable in rodent kidneys using Northern blot analysis (Brann et al., 1987; Strathmann et al., 1990) and RT-PCR analyses (Hansen et al., 2003), although RT-PCR analysis showed that a specific GαoA mRNA isoform could be minimally detected in the mouse kidney (Strathmann et al., 1990). Gαo protein expression was minimally expressed in isolated glomerular membranes from humans (Nitta et al., 1994).
The Gq class of α-subunits consists of Gαq, Gα11, Gα14, and Gα15 (or Gα16 in humans). Gαq, Gα11, and Gα14 mRNA were highly expressed (Strathmann et al., 1989; Wilkie et al., 1991), but Gα15 mRNA was minimally expressed using Northern blot analysis in mouse kidney tissue (Wilkie et al., 1991). Antibodies targeted to Gαq and Gα11 subunits confirmed their expression in isolated whole rat kidney membranes (Gutowski et al., 1991). In human tissue, Gαq presented variable renal mRNA levels from fetus to adult, whereas Gα14 mRNA was only detected in fetal human kidney tissue (Rubio et al., 1999).
Using RT-PCR in isolated rat nephron segments, Gα11 and Gα14 were minimally detected in the outer medullary and upper portion of the inner medullary collecting ducts, respectively (Senkfor et al., 1993). Immunoblot analyses demonstrated Gαq/11 protein in preglomerular vessels (Ruan et al., 1999) and glomeruli from rats (Brunskill et al., 1991; Yanagisawa et al., 1993, 1994).
Gα12 was robustly detected in the brush border membrane of the proximal tubule, with moderate staining in the thick ascending limb of Henle, cortical collecting ducts, and renal blood vessels in rats (Zheng et al., 2003). In mice, Gα12 was detected uniformly throughout the nephron and the podocytes of the glomeruli (Boucher et al., 2012). Gα13 was similarly expressed in high abundance in the proximal tubules, juxtaglomerular cells, and renal artery with lesser staining in the distal convoluted tubules, medullary collecting ducts, and renal veins (Zheng et al., 2003).
The molecular expression of all five Gβ mRNA isoforms was detected in either mouse, chicken, cow, or human kidneys using RNA detection methods. Bovine Gβ1 mRNA (Fong et al., 1987), chicken Gβ3 mRNA (Tummala et al., 2011), and human Gβ5 mRNA (Snow et al., 1999) were detected in the kidneys using Northern blot methods. Mouse Gβ4 (von Weizsäcker et al., 1992) or bovine Gβ2 mRNA (Fong et al., 1987) was not detected in the kidney using Northern blot analysis, but a more sensitive RT-PCR assay followed by Southern blotting of the PCR bands enabled the detection of mouse kidney Gβ1, Gβ2, and Gβ4 mRNA (von Weizsäcker et al., 1992). These contrasting results are likely attributed to the sensitivity of the RNA detection methods, and the lack of bovine Gβ2 mRNA expression in the kidney may be attributed to differential species-specific gene expression regulation.
Of the 12 Gγ isoforms, only 6 were detected in the kidney. Northern blot analysis was used to detect the mRNA for human Gγ4 (Ray et al., 1995), bovine Gγ5 (Cali et al., 1992), bovine Gγ7 (Cali et al., 1992), human Gγ10 (Ray et al., 1995), and human Gγ11 (Ray et al., 1995). Gγ12 protein was detected in the rat kidneys using Western blot analysis (Asano et al., 1998).
The spatial localization of each distinct β or γ isoforms within distinct cell types in the kidney remains to be determined. However, the diversity of each distinct Gα subunit within specific cell types of the kidney may be more expansive than described using molecular techniques, particularly for Gαq. Many published studies investigating GPCR activity have described Gα function to specific vascular, tubular, or interstitial cells within the kidney, without actual evidence for the presence of those G protein subunits being expressed in those segments (Zhang et al., 2011; Zhuo and Li, 2011; Horita et al., 2013). As an example, angiotensin II is well known to regulate proximal tubular sodium reabsorption through activation of the AT1A receptor and its association with Gαq, but there are no definitive data showing the presence of this subunit in that particular segment of the kidney other than interaction studies unrelated to the kidney (Sano et al., 1997; Li et al., 2011b).
Regardless, these molecular studies describing the expression of Gα, Gβ, and Gγ subunits in the kidney provides some potential insight into the types of interactions that may occur with accessory proteins in each respective renal cell type, but further studies are needed to confirm the identity of particular G protein subunits within specific segments of the kidney.
Since the initial findings by Cismowski et al. (1999) using the genetically modified yeast system, there are a total of 19 accessory proteins divided into 4 distinct groups within the AGS family. Thirteen AGS proteins were identified using cDNA libraries from multiple organ systems, and an additional 6 accessory proteins were added based on their mode of action (Sato et al., 2006a; Blumer et al., 2007). The renal localization of the 13 AGS proteins identified using the yeast screen were summarized in Table 2.
Group I AGS Proteins
There are currently four AGS proteins that are categorized in this group: 1) Activator of G protein Signaling 1 (AGS1)/RasD1/Dexras1; 2) RasD2/Ras homolog enriched in striatum (Rhes)/tumor endothelial marker 2 (TEM2); 3) GIV/Girders of actin filaments (Girdin)/Akt phosphorylation enhancer (APE); and 4) resistance to inhibitors of cholinesterase 8 (Ric-8). The latter three accessory proteins were not identified from the yeast screen, but have similar protein structure and/or mechanism of action so they have been included as part of the group I AGS proteins.
AGS1 was the initial AGS protein that was characterized by Cismowski et al. (1999), and demonstrated homology with a dexamethasone-inducible Ras protein, known as Dexras1 or RasD1 (Cismowski et al., 2000). Ras proteins are a family of small (typically 20–25 kDa) monomeric G proteins, which can alternate the binding of guanine nucleotides (GDP and GTP) to control the signaling output within the cell. AGS1 demonstrates selective interaction with Gαi, and not Gαs, Gα16, or Gβγ subunits (Cismowski et al., 1999, 2000, 2001). In yeast, overexpression of AGS1 with either mutant Gαi2 (Cismowski et al., 1999, 2000) or RGS4 (Cismowski et al., 1999), a GAP for Gi/Go proteins, prevented the functional signaling output mediated by AGS1. These studies demonstrate that the biologic function of AGS1/Dexras1 requires the GTP-bound forms of Gαi subunits to activate signaling pathways (Tu and Wu, 1999).
The renal expression of AGS1/Dexras1 mRNA was developmentally regulated where higher levels were detected in the adult versus fetus (Kemppainen et al., 2003). In terms of tissue biodistribution, AGS1/Dexras1 mRNA was minimally expressed in adult mice (Kemppainen and Behrend, 1998) or human kidneys (Tu and Wu, 1999; Kemppainen et al., 2003) compared with other sites, including the liver and brain. AGS1/Dexras1 mRNA expression was induced by dexamethasone administration (Kemppainen and Behrend, 1998; Tu and Wu, 1999; Brogan et al., 2001; Vaidyanathan et al., 2004), including in the kidney (Kemppainen and Behrend, 1998), or by biologic stress applied to the kidney (Lenarczyk et al., 2014a,b). Coincidentally, dexamethasone has been shown to exhibit a protective role in the kidney after renal ischemia-reperfusion injury (Rusai et al., 2013). Considering that AGS1/Dexras1 protein was predominantly localized to the cortical and outer medullary proximal tubular epithelial cells in normal mouse kidneys (Lenarczyk et al., 2014a,b), which are nephron sites that are highly sensitive to ischemia (Devarajan, 2006; Bonventre and Yang, 2011), there may be a role for AGS1/Dexras1 to modulate the extent of renal tubular epithelial cell injury through its ability to activate downstream signaling pathways (Tu and Wu, 1999), which are involved in cell growth, differentiation, and transformation (Kemppainen and Behrend, 1998).
There may be additional roles for AGS1/Dexras1 in the kidney to control cardiovascular function. AGS1/Dexras1 can control renin transcription (Tan et al., 2011) or nitric oxide synthase activity (Fang et al., 2000), which are well established to play a crucial role in long-term sodium reabsorption and blood pressure regulation by controlling regional blood flow distribution and fluid and electrolyte balance in the kidney (O'Connor and Cowley, 2010; Aksu et al., 2011; Toda and Okamura, 2011). In cell culture, AGS1/Dexras1 modulated cAMP production by adenylyl cyclase II through the actions of the dopamine D2 receptor (Nguyen and Watts, 2006). This may implicate another alternate role for AGS1/Dexras1 in the kidney by regulating Na+-K+ ATPase activity in a cAMP-dependent mechanism through the dopamine D2 receptor (Bertorello and Aperia, 1990). The functional roles of AGS1/Dexras1 and its effect on renal function require further investigation.
RasD2/Rhes/TEM2 has moderate sequence homology (approximately 60%) to AGS1/RasD1. Rhes/RasD2 was highly expressed in the striatum with minimal detection in the kidney using RNA blot analysis (Spano et al., 2004). Unlike AGS1/Dexras1, Rhes/RasD2 has yet to be confirmed as a GEF, so further evaluation is necessary to confirm whether it fulfills the criteria to be categorized as a legitimate AGS protein (Falk et al., 1999; Vargiu et al., 2004; Thapliyal et al., 2008; Harrison and He, 2011).
GIV/Girdin/APE was initially shown to regulate actin organization and cell motility by acting as a substrate for Akt phosphorylation. Subsequently, studies have demonstrated that the C terminus of GIV/Girdin can directly interact with receptor tyrosine kinases, G protein subunits, and Akt to control processes, including cell migration (Ghosh et al., 2008, 2010), autophagy (Garcia-Marcos et al., 2011), and metastasis (Ghosh et al., 2010). Initial biodistribution studies by Anai et al. (2005) were unable to reliably detect the expression of GIV/Girdin/APE mRNA and protein in the kidney using Northern blot and immunoblot analysis after immunoprecipitation of tissue lysates, respectively. More recently, the expression of GIV/Girdin was highly induced in glomerular podocytes after administration of puromycin aminonucleoside in rats (Wang et al., 2015), which has been previously shown to be an animal model of reversible podocyte injury (Pippin et al., 2009). In this study, GIV/Girdin promoted the formation of a signaling complex with Gαi and vascular endothelial growth factor receptor 2, leading to elevated phosphoinositide-3 kinase/Akt survival signaling and protection from damage by puromycin aminonucleoside (Wang et al., 2015).
In mammals, there are two isoforms, Ric-8A and Ric-8B, which function as GEFs for the Gα subunits (Tall et al., 2003; Chan et al., 2011). Ric-8A promoted the intrinsic nucleotide exchange rate for Gαi, Gαq, and Gα12/13, but Ric-8B interacts only with Gαs subunits. Ric-8A functions as a chaperone protein to direct Gα subunit association to the plasma membrane. In the absence of Ric-8A/Ric-8B, the steady state levels of Gα are appreciably reduced suggesting a role for Ric-8 to stabilize the Gα subunit from degradation during the biosynthetic process (Gabay et al., 2011). To date, the role of Ric-8 in the kidney remains largely unexplored since Ric-8A was exclusively detected in the nervous system of lower invertebrates and mice (Miller et al., 2000; Tõnissoo et al., 2003, 2006; Maldonado-Agurto et al., 2011). Using a more sensitive RT-PCR assay, however, adult frogs presented a more widespread expression pattern for Ric-8 mRNA, including modest expression in the kidney, but these levels may not be sufficient to alter renal function.
Group II AGS Proteins
Group II AGS proteins are categorized by virtue of one or more protein sequences known as GoLoco (Siderovski et al., 1999) or G protein regulatory (GPR) domains (Peterson et al., 2000; Bernard et al., 2001). Depending on the protein structure, three distinct types of group II AGS proteins have been identified.
AGS3 is also known as G protein signaling modulator 1 (GPSM1), and was first described by Takesono et al. (1999) as the initial member of the GDIs. AGS3/GPSM1 has a protein structure containing N-terminal tetratricopeptide repeats, a linker region, and four C-terminal GPR/GoLoco motifs, which can bind one or more Gαi/o subunits and regulate heterotrimeric G protein activity (Takesono et al., 1999; Peterson et al., 2000; Bernard et al., 2001).
Under basal conditions, full-length and truncated forms of AGS3/GPSM1 mRNA (De Vries et al., 2000a; Pizzinat et al., 2001) and protein (De Vries et al., 2000a; Pizzinat et al., 2001; Blumer et al., 2002; Nadella et al., 2010; Regner et al., 2011; Kwon et al., 2012) were expressed at minimal to undetectable levels in mouse or rat kidneys. This was likely attributed to the selective localization of AGS3/GPSM1 to the distal portion of the nephron (Nadella et al., 2010; Regner et al., 2011). No expression was detected in other nephron segments of the renal cortex or medulla, and no expression was detected in the vasculature or glomerulus.
Under conditions of acute or chronic kidney injury to the renal tubular epithelia, the extent of AGS3/GPSM1 protein induction was dependent upon the site of epithelial cell injury or disruption (Nadella et al., 2010; Regner et al., 2011; Kwon et al., 2012). Ischemia-reperfusion injury, a common form of acute kidney injury, resulted in a dramatic temporal induction of AGS3/GPSM1 expression during the reparative phase of recovery in the outer medullary proximal tubular epithelial cells (Regner et al., 2011; Lenarczyk et al., 2014a,b), which was a nephron segment normally absent in AGS3/GPSM1 expression. In multiple genetic models of polycystic kidney disease from mice, rats, and humans (Nadella et al., 2010; Kwon et al., 2012), AGS3/GPSM1 expression was abnormally elevated in the cystic tubular epithelial cells derived from the collecting ducts (Nadella et al., 2010; Kwon et al., 2012).
However, AGS3/GPSM1 does not appear to play a role during compensatory hypertrophy, which is an alternate mechanism to produce increased cell size (Lenarczyk et al., 2014b). During hypertrophy, AGS3/GPSM1 protein expression was not induced (Lenarczyk et al., 2014b). Moreover, unilateral nephrectomy in the Gpsm1-deficient mice did not affect the ability of the remaining kidney to increase its size compared with wild-type mice (Lenarczyk et al., 2014b).
From these studies, AGS3/GPSM1 appeared to act as a stress-response protein to promote tubular epithelial cell repair in an attempt to restore normal kidney function after renal injury or in genetic diseases involving cellular hyperplasia.
Mechanistically, AGS3/GPSM1 sequesters Gαi subunits to prevent the inactivation of Gβγ signaling (Nadella et al., 2010; Regner et al., 2011; Kwon et al., 2012). Gβγ activity has been implicated in mitotic spindle orientation (Sanada and Tsai, 2005), ion channel activity (Kwon et al., 2012), cAMP production (Fan et al., 2009), or modulation of GPCR-dependent (Sato et al., 2004; Oner et al., 2010, 2013) and GPCR-independent signaling pathways (Smrcka, 2008; Smrcka et al., 2008; Lin and Smrcka, 2011). Overexpression of AGS3/GPSM1 facilitated an increase of polycystin-1/-2 ion channel activity, which could be blocked by a scavenger of Gβγ dimers (Kwon et al., 2012). Alternatively, there was evidence to suggest that AGS3/GPSM1, through its interaction with Gαi subunits, could regulate the autophagic process (Pattingre et al., 2003, 2004; Ghosh et al., 2010; Groves et al., 2010), but this may be cell-type dependent (Vural et al., 2013). AGS3/GPSM1 may also influence signal transduction pathways through other protein–protein interactions, such as liver kinase B1 (Blumer et al., 2003), which is a serine/threonine kinase known to regulate AMP-activated protein kinase signaling during states of energetic stress (Alexander and Walker, 2011). These studies demonstrate the potentially diverse pathways that may be controlled by AGS3/GPSM1 during normal or pathologic states.
AGS5 is also known as LGN (Leu-Gly-Asn repeat-enriched protein) or G protein signaling modulator 2 (GPSM2), and functions as a GDI. AGS5/LGN exhibits a similar protein structure as AGS3/GPSM1, and is considered a mammalian homolog of AGS3/GPSM1. Unlike AGS3/GPSM1, AGS5/LGN was ubiquitously expressed in mammalian organs, including the kidney (Blumer et al., 2002). Our group has detected AGS5/LGN in mouse, rat, and human tubular epithelial cell lines in vitro and whole kidneys in vivo (Nadella et al., 2010; Regner et al., 2011; Kwon et al., 2012). In the normal rat kidneys, AGS5/LGN was detected in the distal tubules, including thick ascending limb and collecting ducts (Lenarczyk et al., 2014a,b). Unlike AGS3/GPSM1, there was no observable change in the AGS5/LGN protein levels or localization pattern after acute kidney injury (Regner et al., 2011; Lenarczyk et al., 2014b) or in genetically mutant cystic kidneys (Nadella et al., 2010).
AGS5/LGN and AGS3/GPSM1 are also considered as mammalian homologs of the Drosophila Partner of Inscuteable (Pins) protein. Functionally, AGS5/LGN has been well characterized as a regulator of mitotic spindle orientation, which is a process that participates in asymmetric cell division of epithelial cells (Fuja et al., 2004; Lechler and Fuchs, 2005; Culurgioni et al., 2011; Zhu et al., 2011). Consistent with these findings, AGS5/LGN was determined to play a key role in cyst formation by controlling mitotic spindle orientation in a renal epithelial cell system in vitro (Zheng et al., 2010; Xiao et al., 2012), but the LGN-Gαi complex may not be involved in asymmetric daughter cell formation (Xiao et al., 2012). These findings may suggest that AGS5/LGN could compensate for the loss of its homolog, AGS3/GPSM1, as observed in a recent study investigating acute or chronic kidney injury (Regner et al., 2011; Kwon et al., 2012). Alternatively, there may be other biologic roles for AGS5/LGN in tubular epithelial cells in the kidney by virtue of its interaction with other proteins, such as soluble guanylate cyclase (Chauhan et al., 2012), which is an important regulator of fluid and electrolyte balance through the actions of hormones, such as nitric oxide (O'Connor and Cowley, 2010; Aksu et al., 2011; Toda and Okamura, 2011). To date, the loss of AGS5/LGN in the kidney and its effects on renal function remains to be determined.
The second type of group II AGS proteins, including AGS6, RGS14, and Rap1GAP, were identified by the presence of a single GPR motif and other protein regulatory domains that function to accelerate Gα-GTP hydrolysis (Blumer and Lanier, 2014).
AGS6 has sequence identity with RGS12, which is the largest member of the RGS protein family. AGS6/RGS12 has a multidomain protein structure that contains the RGS domain, a single GoLoco/GPR motif, and other protein binding domains. AGS6/RGS12 inactivated G protein signaling by accelerating the intrinsic GTPase activity for Gαi/o subunits through the N-terminal RGS domain (De Vries et al., 2000b), whereas the C-terminal GPR/GoLoco motif interacted with all isoforms of the Gαi subunits (Kimple et al., 2001). The expression profile for AGS6/RGS12 in the kidney may be species or isoform dependent, since multiple RGS12 mRNA isoforms were expressed in the human kidneys (Snow et al., 1998), but not in mouse kidneys (Yang and Li, 2007). At the protein level, however, we have observed low uniform staining throughout the rat nephron and the endothelial cells of the blood vessels in the renal cortex (Lenarczyk et al., 2014b).
Functionally, there is a paucity of data regarding the role of AGS6/RGS12 in the kidney, but there are number of studies demonstrating that AGS6/RGS12 played a critical role in the regulation of calcium channels, including N-type (Schiff et al., 2000; Yang and Li, 2007) and CaV2.2 calcium channels (Anantharam and Diversé-Pierluissi, 2002; Richman and Diversé-Pierluissi, 2004; Richman et al., 2005). At this time, the role of the GPR domain in RGS12 and how it regulates G protein signaling compared with its RGS effects remains to be determined.
RGS14 is an accessory protein initially cloned by Snow et al. (1997), and contains a single GPR/GoLoco motif and a RGS domain. RGS14 interacted with Gαi/o subunits in the RGS box (Traver et al., 2000; Kimple et al., 2001), whereas the GPR/GoLoco motif selectively bound to Gαi1 and Gαi3 (Mittal and Linder, 2004) with weaker interaction with Gαi2 (Kimple et al., 2001). RGS14 mRNA distribution was relatively exclusive to the spleen, lung, and select regions in the rat brain (Snow et al., 1997; Reif and Cyster, 2000) with no apparent expression in the kidney (Snow et al., 1997).
Rap1GAP has limited expression distribution in mammalian organs under normal conditions (Rubinfeld et al., 1991). Recent studies, however, have shown that the regulation of Rap1GAP expression was dependent upon the disease stimulus in the kidney. Rap1GAP expression was induced in renal glomerular podocytes from HIV-1 transgenic mice and in human kidney biopsies of focal and segmental glomerulosclerosis (Potla et al., 2014), which appeared to contribute to the podocyte dysfunction after injury by altering β1-integrin–mediated adhesion. In renal carcinoma, the Rap1GAP protein levels were reduced, leading to increased cellular invasion (Kim et al., 2012). The importance of these observations in terms of renal function needs further elucidation.
The third type of group II AGS proteins consists of three proteins: a truncated form of AGS3 (AGS3-SHORT) (Pizzinat et al., 2001), AGS4/GPSM3/G18.1b (Cao et al., 2004), and Pcp-2/L7/GPSM4 (Nordquist et al., 1988), which has multiple C-terminal GPR domains, but no other defined regulatory protein binding sites. Minimal-to-no expression of AGS3-SHORT (Pizzinat et al., 2001), AGS4/GPSM3/G18.1b (Cao et al., 2004; Zhao et al., 2010), or Pcp-2/L7/GPSM4 (Nordquist et al., 1988; Saito et al., 2005) has been detected in the kidneys.
Group III AGS Proteins
Group III AGS proteins include AGS2/TcTex1 (DiBella et al., 2001), AGS7/Trip13 (Li and Schimenti, 2007; Roig et al., 2010), AGS8/FNDC1 (Nielsen et al., 1993), AGS9/Rpn10/S5a/Psmd4 (Deveraux et al., 1995), and AGS10/GNAO (Neer et al., 1984; Sternweis and Robishaw, 1984). The biologic function of group III AGS proteins remains largely undefined in the kidney, but each of these proteins interacts directly with Gβγ subunits complexed with or without the Gα subunits.
AGS2 was initially characterized by Takesono et al. (1999), and was homologous to Tctex1. Tctex1 is a cytoplasmic dynein light chain essential for dynein assembly and participates in specific motor cargo interactions (DiBella et al., 2001). Previous studies have detected Tctex1 mRNA (DiBella et al., 2001) and protein (King et al., 1998) in fetal and adult mouse kidneys (King et al., 1998; DiBella et al., 2001). Human kidneys express higher levels of Tctex1 mRNA compared with the other organs analyzed (DiBella et al., 2001). Recent studies using neural progenitor cells have shown that AGS2/Tctex1 is recruited to the ciliary transition zone prior to the DNA synthesis phase of the cell cycle (Li et al., 2011a), and is modulated by interaction with AGS3/GPSM1 (Yeh et al., 2013). For this reason, AGS2/Tctex1 may have clinical relevance with renal pathologies associated with ciliary defects, such as polycystic kidney disease.
AGS7 was isolated from a yeast screen investigating receptor-independent G protein regulators using a prostate leiosarcoma cDNA library (Sato et al., 2006b). AGS7 demonstrated sequence identity with the C-terminal portion of the thyroid receptor–interacting protein 13 (Trip13) (Cismowski and Lanier, 2005), which was originally identified among a group of other proteins that interacted with the thyroid hormone receptor in a yeast two-hybrid screen (Lee et al., 1995). Neither the role of AGS7/Trip13 as a regulator of the thyroid hormone receptor nor G protein signaling functions have been elucidated in the kidney in vivo. Instead, most of the work involving AGS7/Trip13, which is also a homolog to mouse pachytene checkpoint 2 (mPCH2), has focused on its AAA-ATPase activity with emphasis on the mechanistic control of DNA damage repair (Bolcun-Filas et al., 2014), cell-cycle checkpoint, kinetochore control (Tipton et al., 2012; Wang et al., 2014), and meiotic recombination (Li and Schimenti, 2007; Roig et al., 2010). These initial studies led to recent high-impact findings regarding AGS7/Trip13 in reproductive cell biology (Li and Schimenti, 2007; Roig et al., 2010) and cancer (Banerjee et al., 2014).
In the kidney, the expression of Trip13 mRNA was minimally detected in the kidney using qualitative RT-PCR (Li and Schimenti, 2007). This was consistent with the Trip13 protein localization found only in the principal cells of the collecting ducts in the normal rat kidney (Lenarcyzk et al., 2014a). The site of expression in the nephron may implicate AGS7/Trip13 as a potential regulator of Na+, water, or urea reabsorption, but the biologic importance of AGS7/Trip13 during normal and pathologic conditions in the kidney needs further investigation.
AGS8 was identified using cDNA library from the ischemic injured mouse hearts (Nielsen et al., 1993), and was homologous to Fibronectin type III containing I (FNDC1) (Gao et al., 2003) and KIAA1866 (GenBank accession no. XM_217792.2). Transcript profiling by RT-PCR analysis showed that the basal expression of AGS8 mRNA was most abundant in the skeletal muscle and the kidney (Sato et al., 2006b). AGS8/FNDC1 localization was detected exclusively in distal tubular epithelial cells using immunohistochemistry of Sprague-Dawley rat kidneys (Lenarcyzk et al., 2014a).
Little is known about the function of AGS8/FNDC1 in the kidney. A recent genome-wide association study suggested a potential link between AGS8/FNDC1 and urinary albumin-to-creatinine ratio in African Americans from the CARe Renal Consortium (Liu et al., 2011). However, morpholino knockdown of AGS8/FNDC1 in zebrafish did not support a role for this protein during renal development or dysfunction (Liu et al., 2011).
In terms of G protein regulation, epistasis experiments demonstrated that AGS8 interacted selectively with Gβγ, but not Gαi1/2/3/0 or Gαs. AGS8 has been shown to interact with Gβγ activates connexin-43 (CX-43), which can regulate gap junction function (Sato et al., 2009). Considering that CX-43 is abundantly expressed in the glomerulus and inner medullary collecting duct epithelial cells (Guo et al., 1998), which are a primary site of AGS8 expression, the role of AGS8 with connexins, including CX-43, may need further exploration in the kidney. Alternatively, AGS8/Fndc1 may play an important role in ischemia-reperfusion injury. In the heart, AGS8/Fndc1 gene expression could be robustly induced in cardiomyocytes during ischemia (Sato et al., 2006b, 2009, 2014). In the kidney, however, our laboratory detected a marked reduction in AGS8/Fndc1 mRNA after 24 hours after ischemia-reperfusion injury, which slowly returned toward normal levels after 168 hours (Lenarcyzk et al., 2014a). It is not presently known whether the reduction in renal AGS8/Fndc1 mRNA is attributed directly to negative transcriptional regulation or whether damaged nephron segments were unable to continue expressing AGS8/Fndc1 due to the biologic insult.
AGS9 has sequence identity to a protein known as S5a, which was encoded by the gene for the multiubiquitin-chain-binding protein (Mcb1) in rodents and humans (Ferrell et al., 1996; Pusch et al., 1998). S5a has multiple other nomenclatures, including Rpn10 and Psmd4. S5a/Rpn10 was ubiquitously expressed, but it has multiple splice variants, Rpn10a, Rpn10b, Rpn10c, Rpn10d, and Rpn10e. The expression pattern was dependent on the stage of development (Hamazaki et al., 2007), and the kidneys were one of the major organs that had the highest level of S5a/Rpn10 expression using Northern blot analysis (Pusch et al., 1998).
S5a is part of a large multimeric protein complex involved in protein degradation through the ubiquitin-proteosome system. In the ubiquitin-proteosome system, protein substrates that were destined for degradation were polyubiquitinated for subsequent catabolism through a nonlysosomal pathway (Ciechanover, 1993), which involved the 26S proteasome (Tanaka, 1995). The binding of the polyubiquitinated protein substrate to the 26S proteasome was facilitated by the S5a subunit, and was one of the major docking sites in the proteosome (Verma et al., 2004). Genetic loss of the S5a/Rpn10 expression in mice resulted in embryonic lethality (Hamazaki et al., 2007), and this would suggest that S5a/Rpn10-mediated degradation of ubiquitinated proteins was an indispensable cellular function essential to maintain mammalian life. To date, it remains to be determined which heterotrimeric G protein subunits, namely Gβγ, interact with S5a/Rpn10 to control renal function.
The more commonly known name for AGS10 is GNAO, which is a G protein α subunit that was initially isolated as a 39-kDa polypeptide from rat brain and liver, and bovine heart (Neer et al., 1984; Sternweis and Robishaw, 1984). AGS10/GNAO expression appears to be more broadly detected in many other cell types, including pancreatic cells (El-Mansoury and Morgan, 1998), but there is minimal, if any, expression within the kidney (Brann et al., 1987; Senkfor et al., 1993). Because of a lack of GNAO expression in the kidney as well as the lethality of GNAO knockout mice, there is minimal knowledge regarding any role for GNAO in the kidney.
Group IV AGS Proteins
The most recent AGS proteins were identified by their interaction with a specific subunit, Gα16, in the Gαq family (Sato et al., 2011). Sequence analysis of the Gα16-specific cDNAs were identified as the microphthalmia family (MiTF) of basic helix-loop-helix-leucine zipper transcription factors (Sato et al., 2011), specifically transcription factor E3 (TFE3), transcription factor EB (TFEB), and MiTF. Numerically, TFE3, TFEB and MiTF were designated as AGS11, AGS12, and AGS13, respectively (Sato et al., 2011).
In normal kidneys, the mRNA expression profile of AGS11/TFE3, AGS12/TFEB, and AGS13/MiTF is generally quite low (Kuiper et al., 2003, 2004), and the protein localization is either observed predominantly in the cytoplasm or weakly in the nucleus (Hong et al., 2010). There is some evidence that the basal role of TFE3 in proximal tubules is to control inorganic phosphate reabsorption (Miyamoto and Itho, 2001). However, MitF/TFE transcription factors appears to play a fundamentally important role during translocation renal cell carcinomas from children and young adults (Armah and Parwani, 2010; Hong et al., 2010), in which there is modest to robust nuclear localization of AGS11/TFE3 and to a lesser extent, AGS12/TFEB. Because of this shift in subcellular location to the nucleus, the MiTF/TFE transcription factors are being used as diagnostic biomarkers for specific subtypes of renal cell carcinoma. In these subtypes of renal cancer, the cellular transformation is believed to be attributed to abnormal chromosomal translocation or inversion of genomic DNA containing either TFE3 (Meloni et al., 1993; Weterman et al., 1996a,b) or TFEB (Kuiper et al., 2003) into other genes (Weterman et al., 2000, 2001; Mathur and Samuels, 2007). To date, the nephron segment responsible for the renal cell carcinoma remains to be determined, but there is evidence for TFEB translocation originating from distal tubules (Rao et al., 2012).
MiTF has also been linked as an oncogene for renal cell carcinoma, but its function has been described in greater detail for melanoma (Granter et al., 2002; Bertolotto et al., 2011). There is evidence that MiTF/TFE transcription factors have an inter-relationship between each other through redundant oncogenic function whereby TFE3 knockdown decreased viability of TFE3-translocated papillary renal carcinoma cells, but could be rescued by expression of MiTF (Davis et al., 2006).
In addition to renal oncogenesis, there is increasing evidence that TFE3 plays a role in the pathogenesis of renal cystic disease. Increased chimeric TFE3 protein led to dedifferentiation of renal proximal tubule cells as determined by a loss of cilia formation and key functional transporter proteins (Mathur and Samuels, 2007). Other studies have shown that folliculin, a crucial negative regulator of TFE3 nuclear translocation, promoted renal cell carcinoma and produced the phenotypic appearance of polycystic kidney disease (Chen et al., 2008; Hong et al., 2010). In folliculin-deficient mice, renal tubular epithelial cells transformed into a cystic phenotype that increased nuclear localization of TFE3 compared with the noncystic kidneys (Chen et al., 2008).
There currently remains a paucity of information regarding the mode of action by which MitF/TFE regulates the process of oncogenesis and cystogenesis, particular its role in modulating G protein signaling in the kidney. Gα16 was normally localized to the cytoplasm and plasma membrane (Sato et al., 2011). Coexpression of Gα16 with TFE3 led to nuclear accumulation, which induced the expression of claudin 14, a cell junction protein, in cardiomyocytes (Sato et al., 2011). Claudins are widely expressed in the kidney and could implicate these transcription factors as regulators of paracellular ion transport to control fluid and electrolyte balance (Yu, 2015). On the other hand, MiTF has been shown to be a transcriptional regulator of hypoxia inducible factor 1α (Buscà et al., 2005), which was shown to be abundantly expressed in various renal cell types (Haase, 2006).
There is definitely a need to further elucidate the functional role of these transcription factors with respect to G protein regulation. Since the expression of Gα15/16 is much lower compared with other isoforms of the Gαq isoforms in the kidney (Wilkie et al., 1991), it remains to be determined whether this is a signaling mechanism by which AGS11–AGS13 promote carcinogenesis in the kidney.
Summary and Perspective
AGS proteins exhibit a distinct distribution pattern within the nephron, glomerulus, and vasculature, which helps in producing a diverse control mechanism on G protein signaling evoked by receptor-dependent or receptor-independent stimuli. Although none of the currently known AGS proteins were initially discovered using a normal or diseased kidney-specific cDNA library, several have shown biologic importance in the kidney, including AGS3 during acute and chronic kidney injury (Nadella et al., 2010; Regner et al., 2011; Kwon et al., 2012) and AGS11–AGS13 in different forms of renal cell carcinoma (Granter et al., 2002; Armah and Parwani, 2010; Hong et al., 2010; Bertolotto et al., 2011). Further progress regarding the importance of AGS proteins in the kidney is needed under various stressful conditions, and more importantly, to determine whether the mode of regulation is through the perturbation of G protein signaling. Some of the identified AGS proteins, such as AGS2, AGS7, AGS9, and AGS11–AGS13, were not previously known to interact with Gα or Gβγ subunits, and have only begun to be studied in greater detail.
Identification of additional AGS proteins, if they even exist, using kidney-specific cDNA libraries in the yeast screen may also offer new avenues of research in normal, and possibly diseased, kidneys. Even without the identification of new AGS family members, there will be continual accumulation of knowledge regarding the expression profile, spatial localization, and biologic function of the AGS proteins in various renal pathologies. This will warrant the development of therapeutic interventions targeting individual AGS proteins, particularly in disease states with an active state of growth or hyperplasia. As an example, group II AGS proteins contain one or more of the unique GPR motifs that could be targeted using peptide sequences to manipulate the binding of Gαi subunits in the renal epithelial cells (Blumer et al., 2003; Webb et al., 2005). Considering that the induction of the AGS proteins is generally cell-type specific, the pathologically defective or actively recovering renal cells may be more permissive for uptake of small molecular reagents (e.g., peptide or nucleic acid sequences) or chemical molecules to manipulate AGS function.
Modulating the protein function of intracellular versus cell-surface receptors in the kidney is obviously more challenging for a variety of factors, largely because of the complex architecture of the kidney with its numerous cell types, irregular blood blow distribution, solute gradients, and wide pH ranges. Each of these factors can complicate the development of an effective therapeutic to modulate AGS. However, the molecular and cellular technologies continue to evolve, leading to new discoveries focusing on effective drug delivery systems. Small molecules are being identified to target specific proteins (Jonasch et al., 2012) or cell types (Raj et al., 2011), which may be a viable approach to develop therapeutics targeting group IV AGS proteins to combat renal cell carcinoma. Alternatively, there are emerging new viral vector systems, such as replication-defective lentiviral vectors (Kim et al., 2010; Park et al., 2010) or adeno-associated virus type 9 (Bostick et al., 2007), which have shown promise for renal cell transduction. These vector systems could be designed to either overexpress small sequences to block Gαi binding or inhibit Gβγ activity, or possibly knockdown the expression levels of AGS genes to control their effect on G protein signaling. Although there has yet to be a successful method to correct renal diseases using these approaches to date, the continued study into the biology of AGS proteins will hopefully elucidate pathways that are crucial to G protein regulation during various renal pathologies.
In summary, the field of AGS protein biology is poised to make significant advances in the kidney over the next few years. With the inevitable convergence between the fields of medicinal chemistry, gene therapy, and G protein biology, new therapies targeting AGS proteins should be considered as potential clinical interventions for acute and chronic kidney diseases.
Wrote or contributed to the writing of the manuscript: Park.
- Received January 6, 2015.
- Accepted January 23, 2015.
This research was supported by the National Institutes of Health National Institute of Diabetes and Digestive and Kidney Diseases [Grant R01-DK09123] and institutional funds.
- activators of G protein signaling
- GTPase-activating protein
- guanine nucleotide dissociation inhibitor
- guanine nucleotide exchange factor
- G protein–coupled receptor
- G protein regulatory
- regulators of G protein signaling
- resistance to inhibitors of cholinesterase 8
- reverse-transcription polymerase chain reaction
- Copyright © 2015 by The American Society for Pharmacology and Experimental Therapeutics