SEARCH

SEARCH BY CITATION

Keywords:

  • actin;
  • AP2;
  • arrestin;
  • dynamin;
  • epsin;
  • GPCR;
  • internalization;
  • numb;
  • trafficking;
  • ubiquitin

Abstract

  1. Top of page
  2. Abstract
  3. G protein-coupled receptor endocytosis
  4. Clathrin-mediated endocytosis
  5. Arrestins and endocytosis
  6. Adaptor protein complex-2 and GPCR endocytosis
  7. Ubiquitination and clathrin adaptors
  8. Dab2/ARH/Numb
  9. Actin and GPCR endocytosis
  10. Conclusions and perspectives
  11. Acknowledgments
  12. References

The heptahelical G protein-coupled receptor (GPCR) family includes ∼900 members and is the largest family of signaling receptors encoded in the mammalian genome. G protein-coupled receptors elicit cellular responses to diverse extracellular stimuli at the plasma membrane and some internalized receptors continue to signal from intracellular compartments. In addition to rapid desensitization, receptor trafficking is critical for regulation of the temporal and spatial aspects of GPCR signaling. Indeed, GPCR internalization functions to control signal termination and propagation as well as receptor resensitization. Our knowledge of the mechanisms that regulate mammalian GPCR endocytosis is based predominantly on arrestin regulation of receptors through a clathrin- and dynamin-dependent pathway. However, multiple clathrin adaptors, which recognize distinct endocytic signals, are now known to function in clathrin-mediated endocytosis of diverse cargo. Given the vast number and diversity of GPCRs, the complexity of clathrin-mediated endocytosis and the discovery of multiple clathrin adaptors, a single universal mechanism controlling endocytosis of all mammalian GPCRs is unlikely. Indeed, several recent studies now suggest that endocytosis of different GPCRs is regulated by distinct mechanisms and clathrin adaptors. In this review, we discuss the diverse mechanisms that regulate clathrin-dependent GPCR endocytosis.

Seven transmembrane G protein-coupled receptors (GPCRs) belong to the largest and most diverse family of signaling receptors encoded in the mammalian genome. The superfamily of human GPCRs is comprised of ∼900 members, including ∼375 receptors with known ligands and functional responses, several hundred odorant receptors, as well as ∼160 non-olfactory orphan receptors that have no identified native ligand (1). The GPCR family appears to have expanded enormously during evolution because only three GPCRs exist in the genome of the single-cell yeast eukaryote Saccharomyces cerevisiae. G protein-coupled receptors transduce signals of diverse extracellular stimuli such as photons, ions, biogenic amines, lipids, bile acids, peptides, proteases and many others. The diversity of these receptors is further exemplified by the role of GPCR signaling in vast physiological responses including vision, olfaction, gustation, cardiac function, neurotransmission and others. In addition, nearly half the drugs currently in use target GPCRs either directly or indirectly, indicating the tremendous success in using the GPCR signaling pathway for therapeutic development to treat a wide range of human diseases.

Most GPCRs couple to heterotrimeric G proteins comprised of α, β and γ subunits at the plasma membrane. Upon agonist binding, GPCRs undergo conformational changes exposing cytoplasmic sequences that interact with heterotrimeric G proteins. Activated GPCRs promote exchange of GDP for GTP on the α subunit causing dissociation of Gα and Gβγ subunits which signal to a variety of effectors including phospholipases, ion channels, guanine nucleotide exchange factors (GEFs) for small GTPases, mitogen-activated protein kinases (MAPKs) and other enzymes. G protein-coupled receptors can also signal to certain downstream MAPKs, including extracellular signal-regulated kinases (ERK1 and ERK2) independent of G protein activation (2). In some cases, the multifunctional adaptor protein arrestin cointernalizes with activated GPCRs and functions as a scaffold that binds and sustains ERK1 and ERK2 signaling in the cytoplasm. Thus, diverse mechanisms must exist to precisely regulate the temporal and spatial aspects of GPCR signaling.

Desensitization and receptor trafficking are the predominant mechanisms that control GPCR signaling. Activated GPCRs are rapidly desensitized by phosphorylation and arrestin binding that promotes receptor uncoupling from G proteins within seconds. G protein-coupled receptor internalization ensues within minutes and removes activated receptor from G proteins and signaling effectors at the plasma membrane. Once internalized, some GPCRs may continue to signal from endosomes, until agonist eventually dissociates from receptors or receptor signaling is shut off. G protein-coupled receptors are then dephosphorylated and recycled back to the cell surface in a resensitized state competent to signal again. Thus, endocytosis serves multiple functions in regulation of GPCR signaling including signal termination and propagation and receptor resensitization. Trafficking of internalized GPCRs from endosomes to lysosomes and consequent receptor degradation is also an important process that terminates receptor signaling (3). Our knowledge of the mechanisms that regulate GPCR endocytosis is limited and based on a few well-studied receptors. To date the best characterized pathway for GPCR endocytosis occurs through an arrestin-, clathrin- and dynamin-dependent process. Given the vast number and diversity of GPCRs, the increasing complexity of clathrin-mediated endocytosis and the identification of alternate clathrin adaptors, our basic understanding of mechanisms that control GPCR internalization is now challenged. The focus of this review is the emerging concept that mammalian GPCR endocytosis is controlled by diverse and complex clathrin-dependent mechanisms.

G protein-coupled receptor endocytosis

  1. Top of page
  2. Abstract
  3. G protein-coupled receptor endocytosis
  4. Clathrin-mediated endocytosis
  5. Arrestins and endocytosis
  6. Adaptor protein complex-2 and GPCR endocytosis
  7. Ubiquitination and clathrin adaptors
  8. Dab2/ARH/Numb
  9. Actin and GPCR endocytosis
  10. Conclusions and perspectives
  11. Acknowledgments
  12. References

The original reports of GPCR internalization (also termed sequestration) were from studies of the endogenous β2-adrenergic receptor (β2AR) in frog erythrocytes and human astrocytoma cells. Chuang and Costa (4) observed that incubation of frog erythrocytes with isoproterenol caused an increase in the number of ‘soluble’ β2ARs detected in a cytosolic fraction by radioligand binding. Using human astrocytoma cells, Harden et al. (5) showed that agonist induced the redistribution of a β2AR population from a heavy-vesicle fraction (plasma membrane) to a light-vesicle fraction that exhibited properties of a ‘desensitized’ receptor, that is loss of high-affinity ligand binding and insensitivity to guanine nucleotides. Subsequent studies confirmed the movement of activated β2AR to clathrin-coated pits and internalization from the plasma membrane in intact cells (6).

Several reports indicate that GPCRs of distinct conformations internalize through clathrin-coated pits by different mechanisms. G protein-coupled receptors isomerize between distinct conformational states, including the active and inactive states stabilized by agonists and antagonists, respectively. In many cases, most notably the β2AR, activated GPCR phosphorylation mediated by G protein-coupled receptor kinases (GRKs) and arrestin binding is required for endocytosis. In contrast, some unactivated GPCRs constitutively internalize through clathrin-coated pits independent of phosphorylation and arrestin binding. This has been well documented for protease-activated receptor-1 (PAR1), a GPCR for thrombin, and the thromboxane-A2β receptor (TPβ) (7,8). However, phosphorylation is required for agonist-induced internalization of PAR1 and TPβ, suggesting that constitutive and agonist-induced internalization are specified by distinct endocytic machinery. In addition, antagonist binding to the 5-hydroxytryptamine 2A (5-HT2A) serotonin receptor promotes internalization through an arrestin-independent but dynamin-dependent pathway without affecting receptor activity (9), indicating that for some GPCRs stabilization of an inactive conformation is sufficient to trigger endocytosis. Protein kinase C (PKC)-mediated phosphorylation of unactivated CXCR4 also induces receptor endocytosis through a dynamin-dependent mechanism that appears distinct from ligand-induced internalization (10,11). Internalization of several other unactivated GPCRs by direct activation of PKC has also been reported (12). Thus, multiple distinct GPCR conformations appear to require different endocytic machinery for internalization, suggesting that diverse mechanisms control clathrin-dependent endocytosis of GPCRs in mammalian cells.

The internalization of cargo through clathrin-coated pits is a highly regulated and dynamic process occurring within the 20- to 90-second lifetime of a clathrin bud. Endocytosis of different types of cargo can be saturated but is noncompetitive, indicating that diverse cargo can proceed efficiently through the same clathrin-coated pit, perhaps by using different endocytic machinery (13,14). A substantial literature supports the notion that clathrin-coated pits are high-capacity carriers that efficiently internalize diverse cargo probably through the use of distinct clathrin adaptors (15). Classic electron microscopy studies as well as live-cell imaging of fluorescently tagged clathrin adaptors also strongly suggest that clathrin-coated pits are uniform and homogenous in nature (16). Recent studies from von Zastrow and coworkers suggest that internalization of certain GPCRs occurs through subpopulations of clathrin-coated pits (17). These findings raise the possibility that distinct endocytic machinery is segregated in subsets of clathrin-coated pits. Moreover, these subsets of clathrin-coated pits appear to have the capacity to regulate GPCRs containing post synaptic density protein (PSD95), Drosophila disc large tumor suppressor (DlgA), and zo-1 protein (PDZ) ligands through an actin-dependent mechanism. This results in a difference in the surface residence times, but not in the early recruitment of diverse GPCRs to clathrin-coated pits, and suggests a new role for actin in directly regulating mammalian GPCR endocytosis. Interestingly, Mundell et al. (18) recently observed that the P2Y1 and P2Y12 purinergic GPCRs, which both contain type I PDZ ligands in their cytoplasmic tails, segregate from each other and internalize through distinct populations of clathrin-coated pits. Clearly, this newly proposed regulated form of GPCR endocytosis deserves further research. Although clathrin-coated pits appear to form at preferential sites on the plasma membrane perhaps through interactions with the actin cytoskeleton and are unlikely to be induced by overexpressed cargo proteins (14,19), it will be important to determine whether endogenous GPCRs, like ectopically expressed GPCRs, are sequestered and/or retained in subsets of clathrin-coated pits in native systems.

Clathrin-mediated endocytosis

  1. Top of page
  2. Abstract
  3. G protein-coupled receptor endocytosis
  4. Clathrin-mediated endocytosis
  5. Arrestins and endocytosis
  6. Adaptor protein complex-2 and GPCR endocytosis
  7. Ubiquitination and clathrin adaptors
  8. Dab2/ARH/Numb
  9. Actin and GPCR endocytosis
  10. Conclusions and perspectives
  11. Acknowledgments
  12. References

The majority of cell-surface receptors and integral membrane proteins are internalized by clathrin-mediated endocytosis. Clathrin-coated pits are composed of clathrin heavy- and light-chain molecules that form a polymeric lattice as well as dozens of regulatory proteins. Release of the clathrin-coated pit from the plasma membrane is mediated by the large GTPase dynamin. Clathrin, adaptor proteins and numerous accessory proteins are recruited to plasma membrane regions enriched in phosphatidylinositol (4,5)-bisphosphate (PIP2) and together support coated pit assembly and invagination (20). The function of clathrin adaptors is to enrich select cargo within a forming vesicle. The adaptor protein complex-2 (AP2) is the most abundant adaptor protein found in clathrin-coated pits at the plasma membrane and recognizes short linear tyrosine- and dileucine-based cytoplasmic sequences of cargo proteins (Table 1) (21). However, endocytic clathrin-coated pits also accommodate cargo proteins that do not harbor these motifs, suggesting that AP2 alone is unlikely to have a universal function in cargo recruitment.

Table 1.  Clathrin adaptors and cargo recognition
AdaptorSignal typeRecognition domainCargo examples
  1. LAMP-1, lysosomal associated membrane protein-1; CI-MPR, cation-independent mannose 6-phosphate receptor; NHE5, Na+/H+ exchanger 5; IGF-1, insulin-like growth factor-1; ENaC, epithelial sodium channel; AMPA, α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid-type glutamate receptors; ?, unknown.

AP2YXXφμ2 subunitTransferrin receptor, LAMP-1, CI-MPR, PAR1 (GPCR)
[DE]XXXL[LI]α/σ2 subunitsTyrosinase, GPCRs-
ArrestinsPhosphorylationN-domain/polar coreGPCRs, VE-cadherin, NHE5, IGF-1, TGF-β
Dab2/ARHFXNPXYPTB domainLDL receptors
NumbFXNPXYPTB domainEGFR, Notch
Epsin/Eps15UbiquitinationUBDsEGFR, ENaC
Ent1 and Ent2/Ede1UbiquitinationUBDsSte2 (yeast GPCR)
HIP1/HIP1R??AMPA receptors
CALM/AP180???

Increasing evidence suggests that monomeric clathrin adaptors that bind clathrin and PIP2 co-operate with AP2 to recruit cargo and promote clathrin-dependent endocytosis (Figure 1). Recent work has also identified an α-helical motif in several monomeric adaptor proteins including arrestins, autosomal recessive hypercholesterolemia (ARH) and epsins that bind the β2-appendage of AP2 with high affinity (22). Clathrin adaptors interact with short linear sequences, in addition to recognizing phosphorylated and ubiquitinated cargo (Table 1). Clathrin adaptors themselves are also modified by phosphorylation and ubiquitination, suggesting that clathrin adaptor activity is tightly regulated (Figure 2). The first monomeric adaptor proteins shown to function in clathrin-mediated endocytosis were the nonvisual arrestins. Nonvisual arrestins interact with clathrin and AP2 and facilitate internalization of activated and phosphorylated GPCRs. A second group of monomeric clathrin adaptor proteins include epsins, clathrin assembly lymphoid myeloid leukemia protein (CALM)/AP180 (AP180 is a neuronal ortholog of CALM) and huntington-interacting protein-1 (HIP1) and HIP1 related (HIP1R). These proteins contain an N-terminal PIP2-binding domain [termed epsin N-terminal homology (ENTH)/AP180 N-terminal homology (ANTH) domain] and interact with clathrin and AP2. In addition, epsin harbors three ubiquitin-binding domains (UBDs), binds eps15 and eps15 related (eps15R), which also contain UBDs, and recruits ubiquitinated cargo to clathrin-coated pits. Both CALM/AP180 and HIP/HIP1R lack UBDs and how they function in cargo recruitment is unclear. HIP1 and HIP1R dimerize via their coiled-coil regions and HIP1R binds directly to F-actin linking the actin cytoskeleton to the endocytic machinery (Figure 1). A third group of monomeric clathrin adaptors include disabled-2 (Dab2), ARH and numb (Figure 1). Members of this group bind AP2, some interact with clathrin and all contain a phosphotyrosine-binding (PTB) domain that recognizes FXNPXY motifs, although tyrosine phosphorylation is not a prerequisite for binding. Besides arrestins, the function of other clathrin adaptors in mammalian GPCR endocytosis is relatively unexplored and has been described for only a few GPCRs as discussed below. Distinct clathrin adaptors might function to ensure noncompetitive endocytosis of diverse GPCRs through the same clathrin-coated pit. Moreover, distinct clathrin adaptors might function to control critical aspects of GPCR signaling, similar to arrestins.

image

Figure 1. A graphical representation of the endocytic clathrin adaptor protein–protein interaction network. Adaptor protein complex-2 and clathrin, which serve as hubs, bind to several monomeric clathrin adaptor proteins and recruit diverse cargo to clathrin-coated pits. Clathrin adaptors including AP2 bind to the plasma membrane lipid PIP2, and clathrin binds to adaptor proteins but not PIP2. The dotted line indicates that HIP1 and HIP1R can dimerize and link the endocytic machinery to the actin cytoskeleton. Epsin and eps15 (and eps15R) also interact and function cooperatively to recruit ubiquitinated cargo to endocytic clathrin-coated pits. Stonin2 interacts with AP2 and eps15 to promote endocytosis of cargo proteins. Note that not all proteins involved in clathrin-mediated endocytosis are shown. Adapted from Traub (63).

image

Figure 2. Clathrin adaptor endocytic activity is regulated by phosphorylation and ubiquitination. Arrestins are basally phosphorylated and dephosphorylated upon recruitment to activated and phosphorylated GPCRs. Arrestins are also ubiquitinated by the E3 ligase MDM2 that is critical for its endocytic function. The μ2 subunit of AP2 is phosphorylated by the serine/threonine protein kinase AAK1, which enhances its affinity for tyrosine-based motifs. In contrast, epsin dephosphorylation by the phosphatase calcineurin and deubiquitination by FAM/USP9X, a deubiquitinating enzyme (DUB), is important for enhancing its endocytic activity.

Arrestins and endocytosis

  1. Top of page
  2. Abstract
  3. G protein-coupled receptor endocytosis
  4. Clathrin-mediated endocytosis
  5. Arrestins and endocytosis
  6. Adaptor protein complex-2 and GPCR endocytosis
  7. Ubiquitination and clathrin adaptors
  8. Dab2/ARH/Numb
  9. Actin and GPCR endocytosis
  10. Conclusions and perspectives
  11. Acknowledgments
  12. References

The ubiquitously expressed arrestin2 and arrestin3, also known as β-arrestin1 and β-arrestin2, are best known to function in regulation of GPCR desensitization and internalization. Arrestins bind activated and phosphorylated GPCRs, which promotes receptor uncoupling from G proteins and, in some cases, internalization through clathrin-coated pits. Structural studies of arrestin have revealed distinct N- and C-domains composed of antiparallel β-sheets that are linked by an unusual 12-residue polar core (23). The basal conformation of arrestin is maintained by a network of intramolecular interactions between charged residues buried within the polar core (24). The N-domain of arrestin binds to diverse GPCR intracytosolic sequences, whereas the positively charged polar core engages receptor-associated phosphates. Upon binding to activated and phosphorylated GPCRs, arrestins undergo a conformational change induced by engagement of receptor-associated phosphates with the polar core. This confers high-affinity binding of arrestin to cytoplasmic sequences of some, but not all GPCRs. Binding is followed by exposure of the arrestin C-terminal domain, which interacts with components of the endocytic machinery including the clathrin heavy-chain and the β2-adaptin subunit of AP2, and thereby facilitates GPCR internalization. In Drosophila, visual arrestins have recently been shown to function in endocytosis of rhodopson, a light-activated GPCR, although these arrestin variants lack a clathrin-binding domain (25,26).

The mechanism by which arrestins promote GPCR endocytosis is best characterized for the β2AR and has been recently reviewed (27). Activation of β2AR causes translocation of arrestins to the plasma membrane where the receptor preferentially binds arrestin3 rather than arrestin2. The activated β2AR–arrestin complex redistributes to clathrin-coated pits and arrestin dissociates upon receptor internalization. Arrestin interaction with clathrin and AP2 is essential for agonist-induced β2AR internalization. The endocytic activity of arrestin is controlled by both phosphorylation and ubiquitination (Figure 2). Arrestins are basally phosphorylated and dephosphorylated upon recruitment to the plasma membrane. Dephosphorylation of arrestins is required for agonist-promoted β2AR internalization. In addition to phosphorylation, the ubiquitination status of arrestin is important for its endocytic activity. Arrestins interact with the E3 ubiquitin ligase mouse double minute-2, Mdm2, which specifically ubiquitinates arrestins bound to activated β2AR (28). Ubiquitination of arrestin is essential for β2AR internalization (Figure 3). Other regulators of clathrin-dependent β2AR endocytosis include ARF6 and its GEF ARF nucleotide-binding site opener (ARNO) that bind arrestins, and the GTPase activating protein GIT1, PI3-kinase and erzrin, all of which interact with GRK2. The β2AR illustrates the complexity by which clathrin-dependent GPCR endocytosis is controlled and suggests that endocytosis of other GPCRs is equally complex. In addition to GPCRs, arrestins regulate endocytosis of other cell-surface receptors and integral membrane proteins including the insulin-like growth factor-1 receptor, type III transforming growth factor-β receptor, low-density lipoprotein (LDL) receptor, the Na+/H+ exchanger 5 transporter and most recently vascular/endothelial (VE)-cadherin (Table 1) (29). In some cases, arrestins facilitate endocytosis by binding to phosphorylated serine/threonine residues in the cytoplasmic regions of these cargo proteins. Thus, arrestins are capable of regulating endocytosis of diverse cargo like other clathrin adaptors.

image

Figure 3. G protein-coupled receptors endocytosis is regulated by distinct clathrin adaptors. Mammalian GPCRs are also known as seven transmembrane receptors (7TMRs). The best characterized pathway for mammalian GPCR endocytosis is exemplified by the β2AR that involves arrestin recruitment to activated and phosphorylated GPCRs. PAR1, a GPCR for thrombin, contains a tyrosine-based motif in its cytoplasmic tail that is recognized by the μ2 subunit of AP2. Both AP2 and the tyrosine-based motif are essential for PAR1 constitutive internalization. The mechanisms that regulate activated PAR1 internalization have yet to be defined. Ste2, an yeast GPCR, is phosphorylated and ubiquitinated and internalization is facilitated by Ent1,2, epsin homologs, and Ede1, an eps15 homolog.

Adaptor protein complex-2 and GPCR endocytosis

  1. Top of page
  2. Abstract
  3. G protein-coupled receptor endocytosis
  4. Clathrin-mediated endocytosis
  5. Arrestins and endocytosis
  6. Adaptor protein complex-2 and GPCR endocytosis
  7. Ubiquitination and clathrin adaptors
  8. Dab2/ARH/Numb
  9. Actin and GPCR endocytosis
  10. Conclusions and perspectives
  11. Acknowledgments
  12. References

The finding that several GPCRs internalize independent of arrestins suggests that other clathrin adaptors function as critical regulators of mammalian GPCR endocytosis. The first reports of arrestin-independent GPCR internalization employed dominant-negative arrestin and dynamin mutants. Using these mutants, the m1, m2, m3 and m4 muscarinic acetylcholine receptors (30,31), prostacyclin receptor (32) and 5-HT2A serotonin receptor (9) were shown to internalize through an arrestin-independent, but dynamin-dependent pathway. More definitive evidence for an arrestin-independent mechanism for GPCR endocytosis came from the use of mouse embryo fibroblasts (MEFs) derived from arrestin2 and arrestin3 knockout mice (33). Both constitutive and agonist-induced internalization of PAR1 were shown to proceed through an arrestin-independent but clathrin- and dynamin-dependent pathway in MEFs lacking endogenous arrestins (7). Subsequent studies used arrestin-null MEFs to demonstrate arrestin-independent internalization of the N-formyl peptide receptor (34), the human viral chemokine receptor US28 (35) and the urotensin receptor (36). These studies provide strong evidence that alternate clathrin adaptors function in GPCR endocytosis.

One clathrin adaptor shown to function directly in mammalian GPCR endocytosis is AP2. The heterotetrameric AP2 complex is composed of α, β2, μ2 and σ2 subunits and recognizes short linear sequences that reside in the intracytosolic regions of transmembrane proteins (37). The α subunit of AP2 binds to PIP2, localizing AP2 on the plasma membrane, whereas the β2 subunit binds to the clathrin heavy chain. Adaptor protein complex-2 is able to recruit cargo through μ2 subunit interactions with intracytosolic tyrosine-based motifs (YXXφ). Additionally, α/σ2 subunits, and possibly β2 subunit, recognize acidic dileucine motifs ([DE]XXXL[LI]) in the cytosolic regions of cargo proteins (Table 1). The endocytic activity of AP2 is regulated by μ2 subunit phosphorylation mediated by the AP2-associated kinase-1, AAK1, which enhances μ2 affinity for YXXφ motifs (Figure 2) (38). Adaptor protein complex-2 is known to mediate endocytosis of a variety of cell-surface receptors and integral membrane proteins and a recent study indicates that AP2 functions directly in GPCR endocytosis. Paing et al. (39) found that rather than arrestins, AP2 is required for PAR1 constitutive internalization through a clathrin- and dynamin-dependent pathway (Figure 3). PAR1 constitutive endocytosis generates an intracellular protected pool of naïve receptor that replenishes the cell surface after thrombin exposure and is required for rapid cellular resensitization independent of de novoreceptor synthesis. The μ2 subunit of AP2 was shown to directly bind to a tyrosine-based motif (YXXL) in the cytoplasmic tail of PAR1 using surface plasmon resonance. Moreover, the expression of a PAR1 tyrosine mutant or depletion of AP2 by siRNA led to significant inhibition of PAR1 constitutive internalization, loss of intracellular uncleaved PAR1 and failure of endothelial cells and other cell types to resensitize to thrombin signaling. However, internalization of activated PAR1 is neither dependent on AP2 nor arrestins, suggesting that constitutive and agonist-induced internalization require distinct endocytic machinery.

Several other GPCRs also contain tyrosine- and dileucine-based motifs conforming to canonical and noncanonical AP2 recognition motifs in their cytoplasmic tails (39), suggesting that AP2 might function in trafficking of other GPCRs. A related YXXXφ motif exists in the TPβ receptor cytoplasmic tail and functions in arrestin-independent constitutive internalization, but the role of AP2 has not been tested (8). Interestingly, YXXGφ motifs bind μ2 at the same site as canonical YXXφ motifs, where the Y and φ residues are accommodated in the same hydrophobic pockets (40), further suggesting that AP2 might function in TPβ constitutive endocytosis. The chemokine CXCR2 receptor has two cytoplasmic tail dileucine motifs and a dileucine mutant receptor-bound arrestin but failed to interact with AP2 and to undergo agonist-induced internalization (41). The β2AR and CXCR4 also contain a cytoplasmic tail dileucine motif that appears to function in receptor endocytosis. Mutation of the β2AR cytoplasmic tail dileucine motif impairs agonist-induced endocytosis (42), but the affects on arrestin and AP2 binding are not known. Interestingly, agonist-induced CXCR4 internalization is dependent on serine residues and an adjacent dileucine motif in certain cell types (10), whereas in other cell types the serine and dileucine endocytic signals are required for PKC-mediated internalization but not for agonist-induced internalization (11,43). Other GPCRs harbor noncanonical sequences that bind AP2 and function in endocytosis. The μ2 subunit of AP2 was shown to interact with an unusual stretch of eight arginine residues in the cytoplasmic tail of the α1b-adrenergic receptor, and a receptor mutant, in which the arginine stretch was deleted, failed to bind μ2 and displayed defects in agonist-induced endocytosis (44). The constitutively active human viral chemokine receptor US28 has also been reported to internalize through a clathrin-dependent pathway involving AP2, and not arrestins, but the AP2 recognition motif has not been identified (35). Although the clathrin adaptor AP2 can directly regulate GPCR endocytosis independent of arrestins as shown for PAR1, in some cases, such as the β2AR, AP2 is likely to co-operate with arrestins to promote endocytosis.

Ubiquitination and clathrin adaptors

  1. Top of page
  2. Abstract
  3. G protein-coupled receptor endocytosis
  4. Clathrin-mediated endocytosis
  5. Arrestins and endocytosis
  6. Adaptor protein complex-2 and GPCR endocytosis
  7. Ubiquitination and clathrin adaptors
  8. Dab2/ARH/Numb
  9. Actin and GPCR endocytosis
  10. Conclusions and perspectives
  11. Acknowledgments
  12. References

Arrestins play a major role in desensitization and internalization of several mammalian GPCRs; however, their functions are not evolutionarily conserved. The yeast S. cerevisiaedo not express arrestins, but the protein Vps26 has a structural fold similar to arrestins (45). Both biochemical and genetic studies indicate that Vps26 exists as part of the retromer complex that functions in retrograde trafficking between endosomes and the transGolgi network, and Vps26 is unlikely to function in endocytosis. Instead of arrestins, internalization of yeast GPCRs requires ubiquitin, a small 76-amino acid polypeptide, that functions as an internalization and endocytic-sorting signal. Hicke et al. (46) have demonstrated an essential role for ubiquitination in endocytosis of Ste2, a GPCR for the mating α-factor. Studies using yeast strains that lack specific ubiquitin-modifying enzymes and Ste2 ubiquitin mutants or chimeras indicate that monoubiquitination is necessary and sufficient for both constitutive and agonist-induced internalization (47). Although monoubiquitination is sufficient for endocytosis, attachment of short ubiquitin chains to lysine-63 of an adjacent ubiquitin facilitates endocytosis of many transmembrane proteins (48). In addition, phosphorylation of activated Ste2 facilitates ubiquitination and promotes internalization (49). A role for ubiquitination in endocytosis of Ste3, a GPCR for the mating a-factor, has also been established (50). In contrast, several studies suggested that mammalian GPCR ubiquitination is essential for lysosomal sorting but not for receptor internalization (28,51). However, our recent work indicates that ubiquitination differentially regulates clathrin-dependent internalization of PAR1 (52). The clathrin adaptor that mediates agonist-induced internalization of ubiquitinated PAR1 has yet to be identified.

Several adaptor proteins including epsin and eps15 contain UBDs that recognize ubiquitinated cargo and facilitate internalization through clathrin-coated pits. In addition to UBDs, epsin contains NPF motifs that bind to epsin homology (EH) domains in eps15 (Figure 1). Eps15 also contains UBDs and an EH domain, as well as DPW motifs that bind to the α subunit of AP2, but does not bind clathrin directly. Epsin and eps15 interact with each other and are likely to function cooperatively to recruit ubiquitinated cargo to clathrin-coated pits. Similar to other clathrin adaptors, the endocytic function of epsin is regulated by phosphorylation and ubiquitination (Figure 2) (53). Epsin is basally phosphorylated and ubiquitinated and an influx of Ca2+ initiates rapid dephosphorylation (54). Dephosphorylated epsin is then deubiquitinated by the actions of a deubiquitinating enzyme FAM/USP9X. In yeast, the epsin homologs, Ent1 and Ent2, as well as the eps15 homolog, Ede1, have been shown to regulate Ste2 internalization (Figure 3) (55). In mammalian cells, epsin has been shown to regulate endocytosis of epidermal growth factor (EGFR) and the epithelial sodium channel (56–58), but its function in mammalian GPCR endocytosis is not known. Given that several GPCRs are ubiquitinated and internalize independent of arrestins, certain receptors will likely use an epsin- and/or eps15-dependent mechanism for internalization. In addition, many GPCRs couple to G proteins that activate phospholipase Cβ (PLCβ) resulting in PIP2 hydrolysis and Ca2+ mobilization raising the possibility that activated GPCRs may regulate epsin activity, thus allowing it to function as a clathrin adaptor for mammalian GPCRs.

Dab2/ARH/Numb

  1. Top of page
  2. Abstract
  3. G protein-coupled receptor endocytosis
  4. Clathrin-mediated endocytosis
  5. Arrestins and endocytosis
  6. Adaptor protein complex-2 and GPCR endocytosis
  7. Ubiquitination and clathrin adaptors
  8. Dab2/ARH/Numb
  9. Actin and GPCR endocytosis
  10. Conclusions and perspectives
  11. Acknowledgments
  12. References

Numb, Dab2 and ARH are clathrin adaptors that contain a PTB domain that specifically recognizes FXNPXY motifs in the cytosolic sequences of cargo proteins (Figure 1). Autosomal recessive hypercholesterolemia and Dab2 have well-established functions in LDL receptor internalization, whereas numb has been shown to regulate endocytosis of a variety of integral membrane proteins including EGFR and Notch (Table 1). A study recently demonstrated that numb is phosphorylated and redistributed to the plasma membrane following activation of the G protein-coupled neurokinin-1/substance P receptor which stimulates PLCβ activity and activates PKC (59); but numb is not essential for GPCR endocytosis. Although a critical role for numb, Dab2 and ARH function in GPCR endocytosis has not been demonstrated, a highly conserved N/DPX(n)Y motif is found at the end of the seventh transmembrane of most GPCRs. However, unlike FXNPXY motifs, the N/DPX(n)Y motif seems to be important for maintaining the structural integrity of the receptor protein and is unlikely to directly regulate GPCR trafficking (60), but whether FXNPXY motifs exist elsewhere in the receptor protein remains to be determined.

Actin and GPCR endocytosis

  1. Top of page
  2. Abstract
  3. G protein-coupled receptor endocytosis
  4. Clathrin-mediated endocytosis
  5. Arrestins and endocytosis
  6. Adaptor protein complex-2 and GPCR endocytosis
  7. Ubiquitination and clathrin adaptors
  8. Dab2/ARH/Numb
  9. Actin and GPCR endocytosis
  10. Conclusions and perspectives
  11. Acknowledgments
  12. References

An essential role for a dynamic actin cytoskeleton in yeast S. cerevisiaeendocytosis has been clearly established, whereas its precise function in mammalian endocytosis is less clear. Yeast and mammalian endocytosis not only differ in their requirement for actin but also for dynamin. Although dynamin is essential for release of clathrin-coated pits in mammalian cells, it does not appear to have a direct role in yeast endocytosis. Recent studies from Drubin and others have pioneered the use of total internal reflection fluorescence (TIRF) microscopy in living cells to define the dynamics of actin, clathrin and cargo recruitment to endocytic sites in yeast (61). These studies used a combination of fluorescently tagged Ste2 or α-factor, pharmacological inhibitors and mutant yeast strains lacking critical actin components to clearly establish an essential role for the actin cytoskeleton in Ste2 endocytosis. In budding yeast, actin appears to function mainly in plasma membrane invagination, constriction and scission.

In contrast to yeast Ste2 receptor, the role of actin in mammalian GPCR endocytosis is less clear. In mammalian cells, actin appears at the plasma membrane in transient bursts just prior to clathrin-coated pit scission in living cells, but its precise function in endocytosis is not well understood (16). One issue contributing to the controversial role of actin in mammalian endocytosis is the apparent differences observed upon perturbation of the actin cytoskeleton with pharmacological inhibitors in different cell types and on different surfaces of polarized cells. Endocytosis in fibroblasts shows sensitivity to pharmacological inhibitors of actin compared to neuronal cells, which do not. Actin-depolymerizing agents inhibit clathrin-mediated endocytosis on the apical membrane of polarized epithelial cells, but not at the basolateral surface. Given that many GPCRs interact directly or indirectly with a variety of actin-binding proteins such as EBP50, spinophilin, filamin A/ABP-280, cofilin and spectrin, endocytosis of at least certain GPCRs will likely be regulated in an actin-dependent manner. Indeed, recent work showed that linking GPCRs to cortical actin by fusion of the actin-binding domain of ezrin slows endocytosis (17). Moreover, the cystic fibrosis transmembrane conductance regulator also transiently associates with the actin cytoskeleton through its C-terminal PDZ ligand-binding domain (62), but whether this affects the rate of endocytosis has not been determined. More research is clearly needed to define the function of actin in mammalian GPCR endocytosis.

Conclusions and perspectives

  1. Top of page
  2. Abstract
  3. G protein-coupled receptor endocytosis
  4. Clathrin-mediated endocytosis
  5. Arrestins and endocytosis
  6. Adaptor protein complex-2 and GPCR endocytosis
  7. Ubiquitination and clathrin adaptors
  8. Dab2/ARH/Numb
  9. Actin and GPCR endocytosis
  10. Conclusions and perspectives
  11. Acknowledgments
  12. References

The mechanisms responsible for mammalian GPCR endocytosis remain largely undefined. Because GPCR internalization controls the temporal and spatial aspects of GPCR signaling, it is critical to understand. Identifying the specific clathrin adaptors that regulate internalization of distinct GPCRs will enable us to develop new strategies to manipulate receptor signaling and will provide novel targets for the development of drugs that can be used in the prevention and treatment of a wide range of human ailments, including cardiovascular disease and cancer progression. Recent work has identified new mechanisms and clathrin adaptors that regulate endocytosis of certain GPCRs. However, the mechanisms that control endocytosis of other important GPCRs have yet to be fully elucidated and are vital to our understanding of a variety of human diseases. Although many studies have focused on arrestins in mammalian GPCR endocytosis, much more can be learned using techniques such as live-cell imaging, RNAi and bioinformatics tools to provide a systematic study of the function of all clathrin adaptors in GPCR endocytosis.

Acknowledgments

  1. Top of page
  2. Abstract
  3. G protein-coupled receptor endocytosis
  4. Clathrin-mediated endocytosis
  5. Arrestins and endocytosis
  6. Adaptor protein complex-2 and GPCR endocytosis
  7. Ubiquitination and clathrin adaptors
  8. Dab2/ARH/Numb
  9. Actin and GPCR endocytosis
  10. Conclusions and perspectives
  11. Acknowledgments
  12. References

We thank Dayle Houston, Adriano Marchese and members of the J. Trejo laboratory for comments and advice. This work is supported by NIH HL073328, American Heart Association Established Investigator Award and a Susan G. Komen Breast Cancer Foundation Award (to J. T.). An American Heart Association Predoctoral Fellowship supports B. L. W. We also acknowledge with apologies that because of space limitations a number of excellent studies in the field could not be cited.

References

  1. Top of page
  2. Abstract
  3. G protein-coupled receptor endocytosis
  4. Clathrin-mediated endocytosis
  5. Arrestins and endocytosis
  6. Adaptor protein complex-2 and GPCR endocytosis
  7. Ubiquitination and clathrin adaptors
  8. Dab2/ARH/Numb
  9. Actin and GPCR endocytosis
  10. Conclusions and perspectives
  11. Acknowledgments
  12. References