Perspective

β-Arrestin and Mdm2, Unsuspected Partners in Signaling from the Cell Surface

See allHide authors and affiliations

Science's STKE  27 Nov 2001:
Vol. 2001, Issue 110, pp. pe41
DOI: 10.1126/stke.2001.110.pe41

Abstract

Mdm2 is a ubiquitin-protein ligase known to ubiquitinate p53, promoting its degradation by the ubiquitin-proteasome system. Shenoy and co-workers showed that Mdm2 can act as a key factor in the sequestration of the cell surface β2-adrenergic receptor (β-AR) through interactions with β-arrestin. Strous and Schantl discuss how Mdm2 may be a switch connecting extracellular signals mediated through G protein-coupled receptors (GPCRs) to p53 and its functions in apoptosis and cell cycle progression.

Mdm2 is a RING-type ubiquitin-protein ligase known to ubiquitinate p53, promoting its degradation by the ubiquitin-proteasome system (1). Shenoy and co-workers showed that Mdm2 can also act as a key factor in the sequestration of the cell surface β2-adrenergic receptor (β-AR) (2). Its direct target in this signaling pathway is β-arrestin, the protein adaptor responsible for terminating signaling from the activated receptor and promoting receptor internalization. This finding reveals a novel connection between the role of Mdm2 in the nucleus in the regulation of apoptosis and cell proliferation, and at the cell surface in the down-regulation of signaling receptors.

Mdm2 and the Regulation of p53

Mdm2 is well known as a suppressor of p53 activity (3). The p53 protein, which inhibits cellular proliferation and induces cell death, lies at the heart of stress response pathways that prevent growth and survival of potentially malignant cells. Mdm2 is the major regulator that can both inhibit p53 transcriptional activity and target p53 for degradation (4). Because Mdm2 is also a transcriptional target of p53, an autoregulatory feedback loop exists in which increased activity of p53 leads to increased expression of its own negative regulator. Loss of Mdm2 function (such as by mutations or knockout technology) results in p53-driven apoptosis early in embryogenesis (5), whereas overexpression leads to p53 loss and tumor development.

Activation of p53 requires phosphorylation and acetylation, which dramatically increases its half-life (6). Mdm2 regulation of p53 in the absence of stress occurs through p53 ubiquitination, export from the nucleus, and degradation by the proteasome. With stress or DNA damage, p53 and Mdm2 are phosphorylated and acetylated, and they dissociate (7). Subsequently, p53 forms a tetramer that acts as transcriptional activator leading either to cell cycle arrest through the increased expression of the tumor suppressor protein ARF (8), or apoptosis through the increased expression of the proapoptotic Bcl-2 family member of Bax (9) (Fig. 1). After growth factor stimulation, Mdm2 is phosphorylated by AKT (also known as protein kinase B) and enters the nucleus. This leads to reduction of both p53 levels and transactivation activity. These results establish a novel mitogen-regulated pathway linking phosphatidyl inositol 3-kinase (PI3K) and AKT to the regulation of the Mdm2-p53 complex (10).

Fig. 1.

A schematic view of the regulatory functions of Mdm2. In the absence of stimuli, p53 is ubiquitinated (dark ovals) by Mdm2 and translocated to the cytosol along with Mdm2. Cytosolic Mdm2, presumably stabilized by sumoylation (SUMO) (16), can be degraded by the proteasome, phosphorylated and translocated back into the nucleus, or available for other cytosolic functions, such as membrane receptor transport. Upon agonist binding, GPCR activates G proteins, becomes phosphorylated, and recruits β-arrestin. Mdm2 binds and ubquitinates β-arrestin (ARR), and the trimeric complex interacts with the clathrin-mediated endocytosis machinery. DNA damage and oncogenes activate p53 by acetylation and phosphorylation, causing dissociation of the Mdm2-p53 complex, or by translocating Mdm2 to the nucleolus.

Mdm2 interacts with several proteins in addition to p53, some of which are involved in regulating the p53-Mdm2 interaction. Other Mdm2 partners include the ubiquitin ligase-like tumor susceptibility gene product Tsg101, the p53 coactivator p300, the p53 co-regulator Mdmx, the tumor suppressor pRb, the transcription factor E2F, the ribosomal protein L5, the p53-related protein p73, and ARF (11, 12). Mdm2 is also capable of autoubiquitination and might regulate its own stability (11). The stability of Mdm2 is also regulated by sumoylation (13, 14) and by interaction with other targets, such as Tsg101, Mdmx, p53, and p73 (15). Sumoylation stabilizes Mdm2, and nonsumoylated Mdm2 is ubiquitinated and degraded (16). Tsg101 plays an important role in the regulation of the cellular protein concentration of both Mdm2 and p53: Mdm2 ubiquitination and degradation is inhibited by Tsg101, and Mdm2 accelerates degradation of Tsg101 (15). Deregulation of oncogenes can result in sequestration of Mdm2 by ARF in the nucleolus and stabilization of p53 (17).

The Ubiquitin-Proteasome System

The selective degradation of many short-lived proteins in eukaryotic cells is carried out by the ubiquitin system. In this pathway, proteins are targeted for degradation by covalent ligation to ubiquitin, a highly conserved small protein. Ubiquitin-mediated degradation of regulatory proteins plays important roles in the control of numerous processes, including cell cycle progression, signal transduction, transcriptional regulation, receptor down-regulation, and endocytosis. The specificity of ubiquitination depends largely on the enzymes that recognize the substrates, the class of ubiquitin ligases called E3s (18). E3s are the final components in the multienzyme process that eventually leads to the covalent modification of proteins with ubiquitin. In E3s of the HECT family (for homologous to E6-AP COOH-terminus), a thioester intermediate between ubiquitin and the E3 ligase is formed. In RING-containing E3 enzymes, the E3 ligase binds to the ubiquitin conjugase (E2) and mediates the direct transfer of ubiquitin from E2 to substrates. Mdm2 is a member of the latter family of RING-containing E3s.

The ubiquitin-proteasome system is involved in endosomal trafficking of membrane receptors, transporters, and channels (19). In yeast, the availability of several permeases at the cell surface is regulated by the ubiquitin system by the HECT-domain E3, Rsp5 (20). In only two cases of mammalian membrane proteins have the E3 enzymes been identified: The RING domain-containing adaptor protein c-Cbl is the E3 ubiquitin ligase mediating epidermal growth factor (EGF) receptor ubiquitination and sorting into multivesicular bodies (21), and the HECT domain-containing E3 Nedd4 serves as E3 for the regulated internalization of the epithelial sodium channel ENaC (22). The work of Shenoy et al. demonstrates a role for a third E3 of the RING domain family in regulating another class of membrane proteins, the G protein-coupled receptors (GPCRs) (2).

β-Adrenergic Receptor Signaling and Termination

The β-AR belongs to the GPCR family. GPCRs lack intrinsic kinase activity, but agonist binding stimulates the exchange of guanosine diphosphate (GDP) for guanosine triphosphate (GTP) at the bound heterotrimeric GTP-binding proteins (G proteins), which in turn dissociate. The α and βγ subunits of these G proteins modulate effectors such as adenylyl cyclases, phospholipases, ion channels, and protein kinase cascades (23). The GPCR family includes receptors for many different signaling molecules, such as hormones, chemokines, taste and odor molecules, neurotransmitters, photons, and calcium ions. Once activated, the Gβγ subunits can initiate the Ras-dependent extracellular signal-regulated kinase (ERK) pathway (24).

Agonist-bound β-ARs are phosphorylated by G protein-coupled receptor kinase (GRK), which induces β-arrestin protein binding and ultimately can stimulate endocytosis (25). The affinity of the G protein for the GPCR-arrestin complex is low and the G protein dissociates (26). Thus, arrestin binding causes termination of signaling and renders the complex endocytosis-competent. Almost all GPCRs interact with β-arrestins, although not all of these receptors use β-arrestins for their internalization (23, 25). With a specific cluster of serine and threonine residues in a receptor's COOH-terminal tail, the GPCR-arrestin complex remains firmly associated inside the cell, rendering the receptor inactive for a longer time than those receptors that dissociate rapidly (23). For receptors that lack this cluster, such as β-AR, the GPCR-arrestin complex dissociates shortly after endocytosis (27), allowing the resensitized receptor to recycle back to the cell surface.

β-arrestin is a soluble cytosolic phosphoprotein. It is dephosphorylated when it binds to the β-AR (28). The GPCR-arrestin complex binds AP2 (a clathrin adaptor) and clathrin to initiate endocytosis through clathrin-coated pits and vesicles (29). Moreover, β-arrestin binding provides a signaling platform for many downstream factors. Interaction of the GPCR-arrestin complex with Src promotes activation of the ERK pathway and causes phosphorylation of other factors of the endocytic machinery such as dynamin (30, 31), whereas recruitment of apoptosis-stimulating kinase-1 (ASK1), together with the c-Jun-NH2-terminal kinase 3 (JNK3) module and mitogen-activated protein kinase kinase kinase 4 (MKK4), induces activation of JNK (32). For ERK activation, both Src recruitment and clathrin-coated pit internalization are required (31). Thus, β-arrestin not only terminates receptor-G protein coupling, but also initiates another and distinct wave of signal transduction in which it functions as a scaffold linking the desensitized GPCR to mitogen-activated protein kinase (MAPK) signaling modules.

GPCRs Get Ubiqutinated

The finding by Shenoy and co-workers links the endocytosis of GPCRs to the activity of the ubiquitin system (2). Until now, the ubiquitin system was known to regulate the uptake of some nutrients and ions by controlling the endocytosis of their permeases and transporters (22). Generally, the initial event is a phosphorylation step followed by recruitment of the ubiquitin ligase and ubiquitination. For example, Nedd4 is the ubiquitin ligase for phosphorylated ENaC (20, 22). In yeast, the regulation of the pheromone receptor Ste2p depends on mono-ubiquitination (33). In these cases, the membrane protein is routed to the lysosomes for degradation. It is important to realize that proteins can only get to the lysosomes after a second controlling step, which occurs in the multivesicular endosomes. This step is also controlled by the ubiquitin system in a more general way through a protein complex known as endosomal sorting complex required for transport (ESCRT-I). In this process, Tsg101 (also known as Vps23) recognizes ubiquitinated proteins (34), so only proteins that are recognized by a specific ubiquitin ligase are selected for ESCRT-I handling. Examples are the EGF receptor after ubiquitination by c-Cbl, and the growth hormone receptor (GHR) after ubiquitination by a still-unknown E3 (35, 36).

Activation of the β-AR causes ubiquitination of both β-arrestin and the β-AR (the latter requiring the presence, but not the ubiquitination of, β-arrestin). Ubiquitination of β-arrestin is required for endocytosis by clathrin-coated pits. Furthermore, in cells lacking Mdm2, ubiquitination of β-arrestin (but not that of the receptor) was lost, resulting in decreased receptor internalization, but little effect on receptor degradation (2).

Thus, Mdm2 is not essential for receptor ubiquitination, suggesting the presence of still another E3 in the receptor-β-arrestin-Mdm2 complex. However, Mdm2 must be part of the protein complex that selects specific cargo for efficient endocytosis, suggesting that its role might be to stabilize the complex before it can interact with the AP2-clathrin complex. The presence of ubiquitin moieties on the receptor, β-arrestin, or Mdm2 itself, might serve as the handle by which the receptor complex attaches to the endocytosis machinery.

Making Connections

A most intriguing question is: Why is the ubiquitin system required for the regulation of GPCRs and other membrane proteins? The likely answer is that for selected proteins there are mechanisms in place to coordinate critical functions. For example, the ubiquitin ligase Rsp5 is an essential gene in oleic acid synthesis and in this way controls the membrane fluidity at the endoplasmic reticulum. In addition, Rsp5 takes part in controlling the number of permeases in cell surface membrane (22,37). Thus, the activity of the E3 Rsp5 links lipid metabolism and membrane fluidity with nutrient uptake. Along the same lines, it is likely that an essential role for Mdm2 in controlling p53 activities needs to be supported by controlling the signaling activity of key cell surface receptors that are active in regulating channel activity, sensing extracellular signals, and controlling metabolic processes. β-Arrestins bind to most, if not all GPCRs, and control their activities (23). The current data show that Mdm2 is at the controls that switch between the first wave of signal transduction through G proteins and the second through the scaffold function of β-arrestin to the MAPK cascades.

The relationship between Mdm2 in the nucleus and Mdm2 in the cytosol remains unclear, although the evidence is mounting for important functions in both locations (Fig. 1). Undoubtedly, a major task lies in the nucleus to tune the activities of p53. This function depends on strict protein-protein interactions, that is, on Mdm2 protein concentration. Upon detachment from p53, Mdm2 might shuttle back to the cytosol, where it can be degraded by the proteasome, or be available for other tasks. Distribution between the nucleus and the cytosol is seemingly under strict control, so the concentration in the cytosol is low compared to the nucleus. Confocal images of Mdm2 show most Mdm2 in the nucleus and scarce label associated with perinuclear membrane structures [Schantl and Strous, unpublished observations (38)]. The data from Shenoy et al. suggest that the β-arrestin-Mdm2 interaction must be stable. In addition, the mechanism emerging from these results also shows that endocytosis of the β-AR requires stoichiometric amounts of Mdm2 (2). This indicates that the efficiency of the endocytosis rate depends on the Mdm2 concentration in the cytosol. In addition to the rate of its synthesis, the concentration of Mdm2 in the cytosol depends on its interaction with p53 in the nucleus, the activity of the PI3K-AKT pathway, and its degradation rate. The balance among these processes determines how much Mdm2 is available for GPCRs. One can imagine that massive GPCR activation could deplete the cytosol of Mdm2, which would also deplete the nuclear pool, tipping the balance in the nucleus toward p53 activation. Another unanswered question is whether Mdm2 is consumed during recycling of GPCR. This would affect the duration of the change in p53 activity in response to massive GPCR activation and the ability of cells to adapt to persistent GPCR activation.

The findings of Shenoy et al. prompt another question: What determines the fate of the GPCR once it is internalized? Some receptors, such as the angiotensin-2 receptor or the vasopressin receptor (27), bind β-arrestin much more tightly than others, such as the β2-AR, which releases β-arrestin soon after endocytosis. Endocytosed growth factor receptors are transported to the lysosomes after selection by the ubiquitin system (34, 36, 39). Assuming that β-arrestin and Mdm2 travel with the receptor to the endosomes, they will encounter Tsg101 as part of the ESCRT-I complex. Tsg101 and Mdm2 are binding partners in the nucleus; therefore, they might also bind at the level of endosomes to target the GPCRs to lysosomes. A comparable scenario would apply to the EGF receptor and the GHR.

The intimate relationship between receptor trafficking and signaling is beginning to reveal its secrets. The versatile function of Mdm2 connects basic cell functions in the nucleus with extracellular signals mediated through GPCRs. How this circle closes in terms of molecular mechanisms remains a tantalizing question.

References

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
View Abstract

Stay Connected to Science Signaling

Navigate This Article