G Protein–Coupled Receptor Kinase
G protein–coupled receptor kinases (GRKs) represent a family of seven serine/threonine kinases that, based on their sequence similarities, can further be broken down into three subfamilies. These subfamilies include: GRK1, composed of GRK1 (rhodopsin kinase) and GRK7 (cone opsin kinase); GRK2, including GRK2 and GRK3; and GRK4, made up of GRK4, 5, and 6. These kinases were initially identified for their ability to phosphorylate G protein–coupled receptors (GPCRs). Phosphorylation of the receptor by GRKs leads to the recruitment of β-arrestins and consequently desensitization and internalization of the receptor. This internalization can then lead to additional signaling cascades. Furthermore, it has recently become evident that individual GRKs can interact in a kinase dependent or independent manner with nonreceptor substrates and influence a variety of physiological functions and pathologies.
Evolutionarily, GRKs are present in most vertebrate and invertebrates species as well as identified in nonmetazoan species, with the kinase domain being the most evolutionarily conserved between the families (reviewed in Gurevich and Gurevich 2016). All seven GRK members are ~500–700 amino acids long, multidomain proteins containing a family specific N-terminal domain, followed by an RGS homology domain (RH domain), an AGC protein kinase domain, and a C-terminal domain (reviewed in Ribas et al. 2007). The C-terminal is variable among the different subfamilies and contains structural elements that are partly responsible for the diversity of GRK function, regulation, and interaction. Specifically, members of the GRK1 family have C-terminal prenylation sites while GRK2 subfamily members have a pleckstrin homology domain in their C-termini, that can interact with Gβγ subunits (reviewed in Gurevich et al. 2012), and the GRK4 family members have palmitoylation sites with the exception of GRK5 that contains a positively charged lipid-binding amphipathic helix in its C-terminus (Premont et al. 1999).
In the 1970s rhodopsin phosphorylation was identified and its role in the rapid desensitization of rhodopsin signaling was attributed to “opsin kinase” now commonly known as GRK1 (Bownds et al. 1972). It was through the cloning of the β2-adrenergic receptor kinase (now named GRK2) that demonstrated that GRK1 was, in fact, a member of a family of related kinases that have a common mechanism that can recognize and regulate the active state of a variety of GPCRs (Benovic et al. 1989). Furthermore, GRKs have been shown to have critical regulatory roles outside of GPCR phosphorylation both in enzymatic and nonenzymatic ways – a wide variety of intracellular proteins have been reported to phosphorylate or be phosphorylated by various GRKs. Some of these substrates include: membrane receptors not belonging to the GPCR family, transcription factors, adaptor proteins, cytoskeletal components, and signal transduction proteins (reviewed in Gurevich et al. 2012). Additionally, these GRKs can also be regulated by the presence and/or absence of other GRKs, in part via phosphorylation, making an intricate and complex system of regulation (reviewed in Ribas et al. 2007).
GRKs play a vital role in the regulation and behavior of many biological processes via one of its multiple regulatory mechanisms. Thus, it is impossible to define their role in the body with a simple definition. It is perhaps due to this variety of functions and interactions that they are so heavily conserved evolutionarily. To highlight some of the intricacies and variation between the subfamilies, specific functions and regulations of the individual GRKs are addressed below.
The GRK1 subfamily is composed of the retinal GRKs, GRK1 (rhodopsin kinase) and GRK7 (cone opsin kinase). Both of these GRKs are expressed in the retina of the vertebrates but expression pattern within the retinal cells is species-dependent. For example, rod cells in all vertebrates express GRK1 but the cones express either GRK1 or GRK7 or both depending on the species (reviewed in Orban and Palczewski 2016). GRK7, however, was originally designated the cone-specific receptor kinase due to its expression and activity in cones. The differences in expression begin to play a major role when we look at the effects of mutations and deficiencies of these GRKs in the retina. Oguchi disease, or stationary night blindness, results in prolonged insensitivity of rod-controlled vision after light exposure. This disease has been linked to homozygous or heterozygous deletions or mutations in the GRK1 or arrestin genes in mice. Mice express only GRK1 in both rods and cone photoreceptors whereas in humans GRK7 is also expressed in cones. This leads to a partial amelioration of the light insensitivity in humans, highlighting the variability between species and demonstrating a compensatory or protective role for GRK7 in humans (reviewed in Orban and Palczewski 2016).
While GRK1 and GRK7 share major domains and sequence homology, much of the research done on this subfamily has been done on GRK1. This is partly due to the early discovery of GRK1 (1970s) that was decades earlier than GRK7, but also because an X-ray structure has been determined for GRK1 providing a fundamental basis for understanding enzymatic regulation at molecular level. Functionally, following the activation of rhodopsin by light, the visual desensitization cascade is in part facilitated by GRK1 phosphorylating rhodopsin at serine and threonine residues on the C-terminal domain leading to arrestin recruitment and binding, resulting in inhibition of receptor interaction with Gt (transducin) and subsequent internalization (Krupnick et al. 1997). This interaction between GRK1 and rhodopsin is suggested to be regulated by the N-terminal domain of GRK1 and facilitated by conformational changes induced by interacting with either activated rhodopsin or ATP•Mg2+ (reviewed in Orban and Palczewski 2016).
In order to phosphorylate and regulate rhodopsin, GRK1 and 7 must be properly localized to the membrane to facilitate interaction with the receptor. These GRKs are posttranslationally modified to ensure their proper subcellular localization within the photoreceptor cells. GRK1 differs slightly from GRK7 in this regard in that GRK1 is posttranslationally modified via farnesylation and blocking this farnesylation markedly limits its ability to phosphorylate activated rhodopsin (Inglese et al. 1992). In addition to the prenylation site, GRK1 has been shown to autophosphorylate and that this can affect the affinity of GRK1 toward its substrates including activated rhodopsin (Buczylko et al. 1991). In contrast, based on the sequence of GRK7, it is predicated to have a geranylgeranyl modification site on its C-terminus rather than its N-terminus like GRK1. It is possible that these posttranslational modifications serve as hydrophobic anchors that secure GRK1 and GRK7 to the membranes of these photoreceptor cells (Inglese et al. 1992). Although much has been accomplished with the X-ray structure of GRK1, there is still work to be done deciphering the differences and similarities between GRK1 and GRK7 on a molecular and functional scale.
The GRK2 subfamily includes GRK2 and GRK3. This subfamily is ubiquitously found throughout the body but its localization and expression vary from tissue to tissue. One primary difference between this subfamily and the others is that the ability of GRK2 and 3 to phosphorylate GPCRs is strictly dependent on ligand activation (Li et al. 2015). In addition, RGS homology domain (RH) in the N-terminus of the GRK2 family members is able to bind activated Gαq and inhibit its downstream signaling (reviewed in Gurevich et al. 2012). Another unique feature to this subfamily is the presence of a pleckstrin homology domain (PH domain) which is present in the C-terminus and can facilitate the binding of phosphatidylinositol 4,5-bisphosphate (PIP2) as well as the Gβγ subunit aiding in membrane localization. This is particularly important for receptor phosphorylation since both GRK2 and 3 are primarily found in the cytoplasm and the shuttling by Gβγ is a potential mechanism for active receptor recognition by GRK2 and 3 (Benovic et al. 1986). In addition to localizing GRK2 and 3 to the membrane, a side effect of this high affinity binding to the Gβγ subunit can sequester and interfere with its downstream GPCR signaling very similar to that of Gαq. The presence of the PH domain and the cytoplasmic localization of these GRKs may also provide a mechanism for this subfamily to influence nonreceptor substrates in the presence and absence of ligand activity.
GRK2 (β-adrenergic receptor kinase-1) is expressed ubiquitously throughout the body and is localized primarily in the cytoplasm. Although it is ubiquitously expressed GRK2 has been shown to mainly phosphorylate β-adrenergic and angiotensin II type 1 receptors (reviewed in Gurevich et al. 2012). In addition to phosphorylating and regulating GPCR function, GRK2 is able to phosphorylate and/or interact with a large interactome of cellular proteins in a tissue- and context-specific manner. This combination of canonical and noncanonical roles of GRK2 is now attributed to a multitude of vital physiological functions including: cell migration, proliferation, metabolism, angiogenesis, and insulin resistance. Role of GRK2 in development was demonstrated using knockout mouse models wherein homozygous deletion of GRK2 led to embryonic lethality due to heart defects (Jaber et al. 1996). This along with a number of other studies including cardiac-targeted knockout and overexpression models confirmed key role for GRK2 in cardiovascular diseases, especially heart failure (reviewed in Sato et al. 2015). GRK2 has since been implicated in the pathogenesis of neurodegenerative diseases including Alzheimer’s, multiple sclerosis, and Parkinson’s disease. Finally, GRK2 has been shown to have a significant role in the pathogenesis of polymicrobial sepsis, as well as endotoxemia, primarily through the regulation of the NFκB pathway (reviewed in Packiriswamy and Parameswaran 2015). This vast array of influence makes GRK2 a popular target of study for both diagnostic opportunities as well as therapeutic interventions.
GRK3 is ubiquitously expressed in the periphery as well as in the central nervous system, although its expression levels are generally lower than those of GRK2 (except within the brain). Despite the similarities in tissue expression, GRK3 has an independent role from GRK2 specifically in olfaction and neuronal functions. In the central nervous system, GRK3 is primarily located within olfactory neurons as well as dorsal root ganglions, where they are known to play a role in the desensitization of odorant receptors and α2-adrenergic receptors (reviewed in Pitcher et al. 1998). For example, when GRK3 is knocked down in mouse olfactory epithelia this leads to a significant reduction in odor-induced generation of secondary messengers and lack of receptor desensitization following odorant exposure. GRK3 has also been shown to play a critical role in opioid receptor signaling-Mu-opioid receptor and kappa-opioid desensitization is significantly slower in GRK3 knockout mice (reviewed in Premont and Gainetdinov 2007). In myocardial cells, knocking out GRK3 improves α-adrenergic receptor activity increasing systolic blood pressure. In addition, the lungs of GRK3 KO, but not GRK2 KO mice showed significant enhancement in airway response and sensitivity to cholinergic agonists (reviewed in Pitcher et al. 1998). Finally, based on its chromosomal location and sequence, GRK3 has been identified as a potential marker for bipolar disorder susceptibility through potential alterations in dopamine receptor desensitization (reviewed in Premont and Gainetdinov 2007).
The GRK4 subfamily is composed of GRK4, GRK5, and GRK6 and is found ubiquitously throughout the body with no specific expression pattern with the exception of GRK4, which is found primarily in the testes, kidney, and cerebellum (Inglese et al. 1993). Structurally, members of this subfamily share the AGC kinase domain as well as the N-terminus domain and also express the PIP2 specific binding site similar to the GRK2 subfamily. Additionally, this subfamily has a C-terminal domain aiding in its membrane localization that is regulated by palmitoylation sites on GRK4 and GRK6 and positively charged lipid-binding amphipathic helixes on GRK5 (Premont et al. 1999). These motifs keep these GRKs in equilibrium with the membrane and allows for significant phosphorylation of both activated and inactivated GPCRs (reviewed in Gurevich et al. 2012). Due to the importance of membrane targeting, these lipid-binding sites have emerged as key regulatory sites for this subfamily. Interestingly, GRK4 has relatively low sequence conservation with GRK5 and 6 in the C-terminal region that becomes apparent in the crystal forms of these molecules and has been shown to influence localization of GRK4 to the plasma membrane. (Xu et al. 2014). These changes in membrane targeting by the C-terminus may be influenced by the tissue-specific expression of GRK4 and may account for some of the functional differences amongst this subfamily.
GRK4 differs from GRK5 and GRK6 in its expression pattern throughout the body. Unlike GRK5 and 6, which are ubiquitous, GRK4 is most commonly expressed in the testes, cerebellum, and the kidneys. In these tissues, GRK4 can be alternatively spliced into four different variants that can have independent functions. For example, GRK4α constitutively phosphorylates the dopamine receptor reducing its responsiveness. The other three splice variants do not have this level of regulation but upon agonist stimulation, GRK4γ is also able to phosphorylate dopamine receptor, highlighting specificity but also some commonalities between the different variants (reviewed in Gurevich et al. 2012). This constitutive phosphorylation of the dopamine receptor is problematic in humans with hypertension and polymorphisms in GRK4 can lead to dysregulation of dopamine-stimulated salt and fluid excretion in the kidneys (reviewed in Sato et al. 2015). Expression of GRK4 in the cerebellum appears to be limited to the Purkinje cells and their presence here hints at regulation of metabotropic glutamate 1 receptors that may suggest a role for GRK4 in motor coordination as well as learning (reviewed in Premont and Gainetdinov 2007). When this was further investigated in mice, GRK4 knockout animals showed no differences in basal levels or in response to cocaine for locomotor activity or motor coordination (reviewed in Gainetdinov et al. 2004). In the same animal model, despite the high presence of GRK4 in the testes, there were no apparent changes in the fertility or sperm function and no obvious phenotypes were detected (reviewed in Gurevich et al. 2012). Thus there are still many unanswered questions about the role of GRK4 in physiological functions.
GRK5 is unique from GRK4 and 6 through the presence of positively charged lipid-binding amphipathic helixes on the C-terminus. These lipid motifs, in combination with its polybasic regions in the N-terminus, predominantly localize GRK5 to the plasma membrane. GRK5 also has a nuclear localization signal and has the ability to accumulate in the nucleus where it can phosphorylate class II histone deacetylases and potentially other proteins and mediate gene transcription (reviewed in Sato et al. 2015). Throughout the body, GRK5 is expressed at high levels in the heart, lung, placenta, and kidney and expressed at lower levels in the brain, except within the limbic system. Levels of GRK5, however, have been shown to increase two-fold during neuronal differentiation (Inglese et al. 1993). Canonically, GRK5 can desensitize receptors from a wide variety of GPCR classes including: β-adrenergic, μ-opioid, muscarinic, and angiotensin II receptors. GRK5 can also influence nonreceptor substrates in a kinase-dependent or -independent manner. For example, GRK5 can phosphorylate α-synuclein and tubulin possibly regulating their polymerization that may be related to neuronal function and neurodegenerative disorders such as Alzheimer’s (reviewed in Sato et al. 2015). Furthering its role in Alzheimer’s, GRK5 deficiency is believed to mediate cholinergic neuronal dysfunction and degeneration via the impaired M2/M4 muscarinic receptor desensitization. GRK5 is also expressed in the immune cells suggesting regulation of inflammatory pathways. Role of GRK5 in inflammatory diseases was demonstrated in mouse models of sepsis. In addition, GRK5 was also shown to attenuate atherosclerosis by desensitizing CCR2 and inhibiting migration of monocytes. Finally, GRK5 was shown to play an important role in regulation of p53 thus underscoring a potential role in cancer biology. Using GRK5 knockout mice, GRK5 has been shown to have diverse and wide range of effects including distinct phenotypes in cholinergic responses and inflammation, enhanced hypothermia, hypoactivity, tremor and salivation by oxotremorine, and increased cardiac contractility and enhanced survival in heart failure models. Overexpression of GRK5 has been shown to inhibit contractility and heart rate in response to β-adrenergic receptor agonists, whereas in vascular smooth muscles GRK5 overexpression has been shown to increase blood pressure through β1-adrenergic and Angiotensin II receptors. GRK5 also has contrasting roles in the regulation of inflammation that is dependent on the tissue. While in some cell types GRK5 mediates NFκB signaling, in others, it inhibits NFκB pathways (reviewed in Packiriswamy and Parameswaran 2015). Ultimately these effects have differential consequences in pathophysiological outcomes.
GRK6 is found ubiquitously throughout the body and localizes primarily in the plasma membrane although GRK6α, one of the three splice variants, has also been found in the nucleus; though its function there remains unknown. GRK6 is one of the most prominent GRKs in the striatum and other dopaminergic brain areas, particularly in the GABAergic and cholinergic areas. In these areas GRK6 appears to localize primarily in the synaptic membranes, suggesting that there are neuron specific binding partners for GRK6 that help facilitate this localization (reviewed in Premont and Gainetdinov 2007). Studies have shown that D2 dopamine receptors are direct physiological targets for GRK6 and, unlike all the other GRKs, GRK6 knockout mice show significantly enhanced responsiveness to psychostimulants and other dopamine agonists. Because of this, GRK6 has been linked to Parkinson’s disorder and could potentially be involved in other diseases related to dopamine supersensitivity such as addiction, schizophrenia, and Tourette’s syndrome though these have not been investigated (Premont et al. 2007). Outside of the brain, GRK6 has been shown to have a physiological role in the apoptotic clearance in splenic red pulp, regulate integrin-mediated cell adhesion and migration of tumor cells, as well as modulate arthritis and colitis by regulating infiltration of immune cells. GRK6 is present at high levels in immune cells and can alter the progression of diseases by modifying immune cell chemotaxis through regulation of chemokine receptors CXCR2, CXCR4, and LTB4. This regulation of chemokine receptors is apparent in the GRK6 knockout mice where chemotaxis of T-cells is impaired and chemotaxis of neutrophils is enhanced toward various stimuli (reviewed in Packiriswamy and Parameswaran 2015).
GPCRs make up the largest class of receptors in the human genome with over 800 distinct types. Dysregulation of these receptors can lead to altered cell signaling that can oftentimes lead to pathology. With only four universally present GRKs and three others in specific subsets of tissues, understanding the regulation of these receptors is a critical area of study. This task has been greatly advanced due to the discovery of crystal structures for a number of these GRKs. Not only is there a tightly controlled regulation of GPCRs but also as more and more nonreceptor targets of GRKs become apparent, the degree of complexity for the function of an individual GRK multiplies. The use of genetic knockouts of individual GRKs has been beneficial in aiding in the discovery of pathologies due to lack of a specific GRK, but as more information is known we see overlapping functionalities and compensatory mechanisms from other GRKs within and outside of the specific subfamilies. The availability of CRISPR technology to generate novel mice with multiple knockouts can aid in unraveling some of these overlapping functions. As more novel GRK-interacting substrates are unraveled, our ability to target these kinases for therapeutic purposes becomes larger. Thus it is critical to keep expanding our knowledge and furthering our research on this family of kinases.
- Bownds D, Dawes J, Miller J, Stahlman M. Phosphorylation of frog photoreceptor membranes induced by light. Nature. 1972;237:125–7.Google Scholar
- Li L, Homan KT, Vishnivetskiy SA, Manglik A, Tesmer JJ, Gurevich VV, Gurevich EV. G protein-coupled receptor kinases of the GRK4 protein subfamily phosphorylate inactive G protein-coupled receptors (GPCRs). J Biol Chem. 2015;290(17):10775–90. doi: 10.1074/jbc.M115.644773.PubMedCrossRefPubMedCentralGoogle Scholar