(Journal of Leukocyte Biology. 2002;71:753-763.)
© 2002
by Society for Leukocyte Biology
The Cbl family of ubiquitin ligases: critical negative regulators of tyrosine kinase signaling in the immune system
Navin Rao*,
Ingrid Dodge* and
Hamid Band
* Division of Medical Sciences and
Lymphocyte Biology Section, Division of Rheumatology, Immunology and Allergy, Department of Medicine, Brigham and Women’s Hospital and Harvard Medical School, Boston, Massachusetts
Correspondence: Hamid Band, M.D., Ph.D., Brigham and Women’s Hospital, Smith Building 538C, One Jimmy Fund Way, Boston, MA 02115. E-mail: hband{at}rics.bwh.harvard.edu
 |
ABSTRACT
|
|---|
The Cbl family of proteins are evolutionarily conserved negative regulators of activated tyrosine kinase-coupled receptors. Antigen receptors are prominent targets of negative regulation by the Cbl family members, Cbl and Cbl-b, which proteins function as ubiquitin ligases. Cbl and Cbl-b contain substrate recognition domains that interact specifically with activated protein tyrosine kinases of the Src and Syk/ZAP-70 families. Cbl-mediated ubiquitination of these kinases leads to their degradation, resulting in attenuation of receptor signals. Cbl may also control activation-induced monoubiquitination of antigen receptors, thus facilitating their delivery to lysosomes for subsequent degradation. Finally, the interactions of Cbl proteins with downstream targets of tyrosine kinases, such as PI-3-kinase and Vav, could provide an additional mechanism to attenuate receptor signaling. By targeting multiple components of antigen receptor signaling for degradation, the Cbl protein family provides a critical mechanism to ensure an appropriate immune response. The hyperresponsiveness of Cbl-/- and Cbl-b-/- lymphocytes and the autoimmune phenotype of Cbl-b-/- mice lend strong support for this proposal. The ability to control early receptor signals through regulated protein degradation provides a novel paradigm of immunoregulation.
Key Words: PI-3K • immunoregulation • antigen receptors • SFK • Syk/ZAP-70
|
INTRODUCTION
|
|---|
Protein tyrosine phosphorylation provides a fundamental switch to initiate leukocyte responses to diverse extracellular stimuli, ranging from cytokines to cell-cell interactions. In addition to responding to these generalized external cues, leukocytes are uniquely endowed with the ability to respond to foreign antigens. Foreign antigen-dependent leukocyte activation is mediated by the T- and B-cell antigen receptors (TCR and BCR, respectively) and the immunoglobulin (Ig) Fc receptors (FcRs). It is now firmly established that leukocyte development and responses to antigen are absolutely dependent on tyrosine kinase activation [1
2
3
]. Furthermore, although antigen receptor signaling involves some immune system-specific adaptations, the basic signaling mechanisms are shared with other tyrosine kinase-coupled receptors, thus establishing antigen receptors as valid models to study the regulation of tyrosine kinase signaling pathways.
The antigen receptors (TCR, BCR, and FcRs) are Ig superfamily transmembrane glycoproteins that do not interact directly with their signaling machinery. Instead, the antigen-binding receptor components associate noncovalently with invariant signaling subunits (CD3
, 
, 
, and 
chains for the TCR; Ig
and ß chains for the BCR; and or chains for the FcRs) [1
2
3
]. These invariant subunits contain a conserved amino acid sequence (D/E)XXYXX(L/I)X6–8YXX(L/I), referred to as the immune receptor tyrosine-based activation motif (ITAM; chain has three ITAMs) within their cytoplasmic tails. The phosphorylation of paired ITAM tyrosine residues is a crucial event in cellular activation through antigen receptors. ITAM phosphorylation is mediated by members of the Src-family kinases (SFKs), a group of protein tyrosine kinases (PTKs) anchored to the inner leaflet of the plasma membrane through N-terminal myristoylation [1
2
3
]. Specific SFKs are associated weakly with the TCR (Fyn, Lck), BCR (Fyn, Lyn, and Blk), and FcRs (Fyn and Lyn). In addition, the TCR coreceptors CD4 and CD8 associate strongly with Lck, and this pool of Lck is crucial for coreceptor-dependent cellular activation via the TCR [1
2
3
]. Phosphorylated ITAMs function as high-affinity docking sites for the tandem SH2 domains of the cytoplasmic PTKs, Syk (associated with BCRs and FcRs) and ZAP-70 (associated with TCRs). This promotes their recruitment to activated antigen receptors and facilitates their activation by SFKs. Naturally occurring or engineered PTK mutations have demonstrated the essential role of sequential SFK and Syk/ZAP-70 activation in the induction of all cellular responses triggered by the antigen receptor [1
2
3
].
To ensure an appropriate immune response to antigenic challenge without eliciting an ongoing inflammatory or autoimmune response, it is critical that the level of antigen receptor-induced cellular activation be regulated precisely. Tyrosine phosphatases provide an essential mechanism of negative regulation for tyrosine kinase-coupled receptors [4
]. Down-regulation of surface levels of activated antigen receptors and/or their costimulatory receptors is thought to provide another mechanism of signal attenuation [5
], similar to ligand-induced lysosomal targeting of growth factor receptors via the endocytic pathway [6
]. However, the precise role of this mechanism in regulating antigen receptor signaling has not been delineated. Given the consequences of dysregulated PTK activation, it is likely that additional mechanisms exist to regulate tyrosine kinase-coupled receptor activation, independently or in concert with the mechanisms mentioned above. Recent advances have identified a novel mechanism of negatively regulating antigen receptor and other tyrosine kinase-coupled receptor signals, which is mediated by Cbl family proteins [7
8
9
10
]. This involves ubiquitination of activated PTKs and other crucial early signaling intermediates. This review will focus on the role of this novel Cbl-mediated pathway, with specific emphasis on attenuation of antigen receptor signaling.
|
UBIQUITIN-DEPENDENT PROTEIN DEGRADATION AS A REGULATORY MECHANISM
|
|---|
Ubiquitination is an evolutionarily conserved covalent modification of cellular proteins in which ubiquitin is attached via an isopeptide bond between its C-terminus and a lysine residue in the target protein [11
]. Target proteins can be modified by mono- or polyubiqitination. Monoubiquitination involves the addition of one or two ubiquitin moieties to a specific lysine residue(s) in the target protein (Fig. 1
). Polyubiquitination involves chains of four or more ubiqitin moities attached to one or more target protein lysine residues. The type of ubiquitin modification, mono or poly, determines the functional consequences for the modified protein. Chains of four or more ubiquitins, typically seen on cytoplasmic proteins, are recognized efficiently by the proteasome and lead to target protein degradation. In contrast, monoubiquitination does not serve as a proteasome targeting signal, but plays other distinct roles (Fig. 1
; and discussion below).
View larger version (14K):
[in this window]
[in a new window]
|
Figure 1. The different functional roles of ubiquitin. Target proteins are monoubiquitinated on one or more lysine residues, which serve as a signal for intracellular sorting as well as endocytosis of cell-surface proteins. Polyubiquitination, which involves chains of usually four or more ubiquitin (Ub) moieties, serves as an efficient signal for recognition by the proteasome and subsequent protein degradation.
|
|
While most immunologists are familiar with the role of proteasomal degradation in generating peptides for loading onto major histocompatibility complex class I molecules [12
], the role of the ubiquitin proteasome system in other pathways of immunological interest has only been recognized recently. Proteolytic degradation of signaling molecules can, in principle, provide a powerful mechanism to attenuate an intracellular signal by removing activated proteins. Degradation of signaling proteins may also generate a refractory period in a cell, as new signaling proteins may need to be synthesized before the signal can be fully propagated again. Proteolytic degradation can also serve as a mechanism for removal of the inhibitory components of signaling protein complexes, as is the case for nuclear factor-
B activation (NF-B) [13
]. Thus, polyubiquitination-dependent targeting of proteins to the proteasome can provide a mechanism for regulating activated signaling proteins specifically, thereby controlling the level of cellular activation.
While proteasomal degradation of cytosolic and nuclear signaling proteins is well established, ubiquitin modification of cell-surface transmembrane receptors is thought not to target them for proteosomes, as they are generally monoubiquitinated. Studies carried out initially in yeast, and extended recently to mammalian cells, have demonstrated that monoubiquitin modification of cell-surface receptors serves as a signal to target them to the lysosome, where they are degraded by lysosomal proteases [14
]. Lysosomal degradation provides an efficient method to remove activated cell-surface receptors, thus attenuating the activation signal. A large array of cell-surface receptors coupled to tyrosine kinase activation have now been shown to undergo activation-dependent ubiquitin modification, indicating that ubiquitin-dependent lysosomal targeting is a general mechanism of signal attenuation [6
, 14
].
|
THE UBIQUITIN MODIFICATION MACHINERY
|
|---|
Given its dramatic role in dictating protein fates, it is not surprising that the attachment of ubiquitin to target proteins is tightly regulated. Ubiquitin is synthesized in a polymeric form and then posttranslationally cleaved into monomeric subunits [15
]. Monomeric ubiquitin is activated by a ubiquitin-activating enzme, or E1, and involves the attachment of ubiquitin to E1 via a thioester bond. There is only one described mammalian E1 enzyme. The E1 then transfers activated ubiquitin to one of approximately 25 ubiquitin-conjugating enzymes, or E2s, also via a thioester linkage. The E2s transfer ubiquitin to lysine residues of target proteins bound to ubiquitin ligases, or E3s. Attachment of ubiquitin to target protein-lysine residues involves an isopeptide bond. The large number of target-specific E3 ligases provides exquisite specificity for the ubiquitin system (Fig. 2
). In some cases (as with HECT domain-containing E3s), ubiquitin is first transferred to E3 via a thioester linkage and is subsequently transferred by E3 to a target protein [15
]. In all cases, physical association of E2 with E3 appears to be essential for target-specific ubiquitination. Over the past several years many laboratories have demonstrated that Cbl and related family members constitute a novel group of E3 ubiquitin ligases directed at components of tyrosine kinase signaling pathways linked to membrane receptors, as discussed below.
View larger version (34K):
[in this window]
[in a new window]
|
Figure 2. Comparison of Cbl ubiquitin ligase with a SCF ubiquitin ligase. In both models, a ubiquitin-activating enzyme (E1, orange) activates Ub and transfers it to a ubiquitin-conjugating enzyme (E2, yellow), which interacts with a ubiquitin ligase (E3) and transfers Ub to the target protein (blue). The SCF E3s are multiprotein complexes (left panel), containing an F-box protein (Cdc4, green) that recognizes the target protein. Cullin and a RING finger-containing protein (Rbx1, purple) bind the E2. A linker protein (Skp1, pink) serves as an adaptor. The Cbl family of E3s recruits E2 proteins (RING finger, purple) and substrate binding through different domains of a single polypeptide. Cbl contains several target protein-binding domains such as the TKB domain, the prolin-rich region, as well as induced phosphorylation sites.
|
|
The biochemical basis for Cbl ubiquitin-ligase function strongly parallels other recently described RING finger domain-containing E3 ligases, such as the Skp1/Cullin/F-box (SCF) family of E3s [7
]. In the SCF family of E3s (Fig. 2)
, an F-box-containing protein, such as CDC4, mediates target-protein recognition [15
]. The F-box of CDC4 interacts with Skp1, which links the CDC4-target protein complex to the rest of the ubiquitin-ligase machinery. The ubiquitin-ligase machinery is composed of a cullin and a Ring finger-containing protein, Rbx1. The RING finger of Rbx1 is critical for recruitment of E2 and ubiquitin ligation, as mutations in this domain abrogate E3 activity. A prominent example of immunologically relevant SCF-type E3s is ßTrCP, which regulates IB degradation in the process of NF-B activation [13
]. Cbl family proteins can function as E3s in an analogous manner [16
17
18
], but with the important distinction that all of the necessary protein-protein interactions are carried out by a single polypeptide (Fig. 2)
. The specific target recognition domains in Cbl direct this E3 almost exclusively to components of tyrosine kinase signaling pathways (see below).
|
Cbl PROTEIN FAMILY
|
|---|
The Cbl protein family contains three related but distinct mammalian gene products: Cbl, the highly related Cbl-b, and the more distantly related Cbl-c (Fig. 3
). Cbl and Cbl-b share overall domain structure and exhibit high sequence identity, particularly within the N-terminal half (84% identity; Table 1
). Cbl and Cbl-b are composed of an N-terminal tyrosine kinase-binding (TKB) domain, which mediates binding to specific phosphotyrosine motifs on activated PTKs and possibly other signaling proteins; a Zn-coordinating RING finger domain that interacts with E2s; a proline-rich region that contains functional SH3 domain-binding sites; and a C-terminal leucine zipper-like region with high homology to ubiquitin-associated (UBA) domains [8
]. Cbl-c retains the TKB and RING finger domains but possesses only a short proline-rich region and no UBA domain [7
, 8
, 10
, 19
].
View larger version (30K):
[in this window]
[in a new window]
|
Figure 3. Conservation of structure in the Cbl family of proteins. The comparison includes mammalian Cbl family members (Cbl, Cbl-b, and Cbl-c), chick (Gallus gallus), Cbl homologue (gCbl), the long and short Drosophila melanogaster homologues (dCbl-L and dCbl-S), and the C. elegans Cbl homologue (SLI-1). The Cbl domains are TKB domain (green), Ring finger domain (purple), proline-rich region (yellow), and leucine-zipper/UBA domain (blue).
|
|
Related gene products have been identified in Caenorhabditis elegans (SLI-1), Drosophila (dCbl), Xenopus, and chick [10
]. Comparison of the primary sequences reveals strong evolutionary conservation of the TKB domain and an even stronger conservation of the RING finger domain (Table 1)
. The Cbl homologue SLI-1 exhibits a domain structure similar to that of Cbl-c, although it shows comparable sequence identity within the TKB and RING finger domains to all mammalian Cbl proteins. dCbl is expressed as a long version with domain structure identical to that of Cbl and Cbl-b, and an alternative transcript encodes a short form composed solely of the TKB and RING finger domains [20
]. The high degree of conservation of the TKB and RING finger domains underscores their importance in Cbl function and suggests that the C-terminal motifs may play roles unique to particular Cbl family members (Table 1)
.
Notably, RNA expression data indicate that mammalian Cbl family members are expressed differentially. For example, although Cbl and Cbl-b are expressed broadly, the highest levels of Cbl mRNA are found in thymus and testis [21
, 22
]. Conversely, Cbl-c is expressed at low levels in hematopoietic tissues but is strongly expressed in organs with a high epithelial component [19
]. dCbl forms appear to be regulated based on cellular activation state [23
]. At present, there is no information on the relative expression of Cbl proteins in various tissues and at different stages of development or if there are any alterations in Cbl protein levels during immune responses or disease states.
|
Cbl IS AN E3 LIGASE
|
|---|
Genetic evidence from studies of C. elegans vulval development demonstrated that the SLI-1 is a negative regulator of LET-23, the mammalian epidermal growth factor receptor (EGFR) homologue [24
]. Subsequently, dCbl proteins were also shown to negatively regulate Drosophila EGFR-mediated cell fate decisions [25
, 26
]. Complementary studies in mammalian cells showed that Cbl functions as a negative regulator of the EGFR and other receptor tyrosine kinases [27
28
29
30
]. Cbl regulates EGFR signaling negatively via enhanced ubiquitination of the receptor. The RING finger domain of Cbl was found to be critical for ubiquitination of receptor tyrosine kinases and, together with the TKB domain, was sufficient for ubiquitination activity in vitro and in vivo [6
]. Biochemical mutational analyses and crystal structure data demonstrate that Cbl’s RING finger domain mediates binding to E2s, indicating that Cbl is an E3 ubiquitin ligase [16
, 18
, 31
]. Several in vitro and in vivo analyses have established this conclusively [16
, 17
, 28
, 32
33
34
35
].
|
Cbl-MEDIATED NEGATIVE REGULATION OF Syk/ZAP-70 PTKs
|
|---|
The first potential targets of Cbl-mediated negative regulation in the immune system were Syk/ZAP-70 PTKs. Cbl forms stimulation-dependent complexes with Syk and ZAP-70, via the TKB domain of Cbl binding directly to autophosphorylated Syk and ZAP-70 [36
, 37
]. Phosphopeptide library screening and mutational approaches have defined N/DXpYXXX
(where is any hydrophobic residue, usually proline) as the binding motif for the TKB domain [28
, 38
, 39
]. Definition of Cbl’s TKB domain binding site on Syk/ZAP-70 as the negative regulatory linker autophosphorylation site (Y292 in human ZAP-70 and Y323 in human Syk) suggested that Cbl was a negative regulator of these PTKs [30
, 37
, 38
, 40
]. Previous analyses had demonstrated that Syk/ZAP-70 mutants with Y 
F substitution of these phosphorylation sites led to lymphocyte hyperresponsiveness upon antigen receptor stimulation [41
42
43
]. Furthermore, ZAP-70-Y292F knockin mice exhibit a reduced TCR down-regulation rate, enhanced proximal TCR signaling events, and increased numbers of T cells producing interleukin (IL)-2 and interferon- in response to antigen [44
]. As the specific activity of the ZAP-70-Y292F mutant was not elevated, the hyperactive phenotype in vivo suggested that the linker phosphorylation site recruits Cbl as a negative regulatory protein [42
].
Indeed, studies in lymphoid and nonlymphoid cells demonstrate directly that Cbl functions as a negative regulator of Syk and ZAP-70 by interacting with the linker phosphorylation site upon antigen receptor stimulation. For example, reconstitution of the ZAP-70-deficient Jurkat T-cell line p116 with wild-type and mutant Syk or ZAP-70 clearly demonstrated that Cbl functions as a negative regulator of Syk and ZAP-70 [33
, 45
]. This functional effect, as measured by nuclear factor of activated T cells (NFAT)-luciferase activity, requires an intact Cbl-TKB domain as well as a Syk/ZAP-70 linker autophosphorylation site. In contrast, one study suggests that Cbl-b, which interacts with ZAP-70 through the same mechanism as Cbl, enhances ZAP-70-dependent NFAT-luciferase activity in a Jurkat T-cell transfection system [46
]. This result is inconsistent with the studies showing that Cbl-b functions as a negative regulator of EGFR [47
], as well as the phenotype of Cbl-b knockout mice (see below). Further studies are warranted to clarify these discrepant results.
Consistent with Cbl’s ubiquitin-ligase function, overexpression of Cbl in lymphoid and nonlymphoid cells led to increased degradation of Syk/ZAP-70 as a result of Cbl-mediated polyubiquitination. This required the TKB and RING finger domains of Cbl and could be blocked by proteasome inhibitors [33
]. Furthermore, the TKB and RING finger domains were sufficient to induce Syk degradation. Importantly, endogenous Cbl was shown to mediate ubiquitination of Syk upon BCR stimulation, as introduction of a dominant-negative RING-finger mutant-abrogated Syk ubiquitination [33
]. Due to the activation-dependent association of Cbl and Syk/ZAP-70, Cbl targets the activated pools of these kinases specifically for proteosomal degradation. Cbl-induced degradation of Syk/ZAP-70 appears to be the sole mechanism for Cbl regulation of these kinases, as the specific activity of Syk was unaltered by its association with Cbl [33
]. Similar to Syk/ZAP-70, Cbl-dependent ubiquitination of receptor tyrosine kinases also requires the TKB and RING finger domains, and these domains are sufficient for Cbl’s negative regulatory effects [32
]. Therefore, the TKB and RING finger domains define a core structural unit whereby Cbl negatively regulates signaling molecules in the immune system and elsewhere. Recent studies of this unit are therefore discussed first, as these provide insights of general significance.
|
STRUCTURAL BASIS FOR THE FUNCTION OF Cbl’s TKB AND RING FINGER DOMAINS
|
|---|
The structural basis for Cbl binding to activated tyrosine kinases was not apparent immediately from its primary sequence. Crystal structural analysis revealed the Cbl TKB domain to be composed of three interacting domains: a four-helix bundle (4H), a calcium-binding EF hand (EF), and a divergent SH2 domain [39
]. The EF hand wedges between the 4H and SH2 domains and determines their orientation. The 4H domain helps complete the phosphotyrosine-binding pocket of the SH2 domain, which lacks the ß strands D, E, and F, as well as the BG loop; the latter, together with the EF loop, is known to help form the specificity pocket of the SH2 domain. Point mutations that inactivated the 4H, EF, or SH2 regions of Cbl’s TKB domain abrogated binding to ZAP-70, demonstrating that these domains form an integrated phosphotyrosine-binding platform [39
].
A more recent structure of the N-terminal region, containing the TKB and RING finger domains, revealed that the RING finger domain is anchored to the TKB domain by interacting with the 4H bundle [31
]. A total of seven highly conserved RING finger domain residues interact with the 4H bundle of the TKB domain, suggesting that the precise arrangement of these two domains is important for Cbl function. The TKB-RING finger linker region also packs onto the TKB domain forming a loop and an -helix. The TKB–linker interactions prominently feature the conserved Y368 and Y371 residues of the Cbl linker region. Deletion of either tyrosine residue causes Cbl to become oncogenic [48
], indicating that the integrity of this TKB–linker interface is also necessary for Cbl function. Furthermore, the structure revealed that Cbl binds the E2 UbcH7, using its RING finger domain and the linker helix.
|
Cbl INTERACTION WITH Src FAMILY KINASES
|
|---|
In addition to TKB and RING finger domains, most Cbl family members also possess a proline-rich region that can interact with the SH3 domain of SFKs [10
]. SFKs, such as Lck, Fyn, and Lyn, are crucial initiators of antigen receptor signaling, and inappropriate activation of SFKs can result in uncontrolled leukocyte proliferation. For example, the Herpesvirus saimiri TIP protein binds to and activates Lck, which is thought to promote leukemogenesis [49
]. Importantly, SFKs also play vital roles during leukocyte development as shown by gene ablation in mice [50
]. Given their pivotal role in leukocyte signaling, regulation of SFKs is of substantial interest.
The crystal structures of several SFKs, together with a large body of mutational data, have established a general model for basal repression of SFKs (Fig. 4
) [51
, 52
]. Intramolecular SH3 domain binding to a type II polyproline-like helix in the SH2-kinase linker region, together with SH2 domain binding to the phosphotyrosine residue near the C-terminus, force the kinase domain into an inactive conformation [51
, 52
]. Activation signals are hypothesized to displace the SH2 and SH3 domains from their intramolecular ligands, allowing the SH2 and SH3 domains to assemble signaling complexes. Indeed, mutations within the SH2-kinase linker that abolish binding to the SH3 domain or overexpression of high affinity SH3 domain-binding ligands result in increased SFK kinase activity [53
]. Although the above paradigm elegantly accounts for basal repression and provides a plausible scheme for activation of SFKs, it is not clear if and how activated SFKs are returned to their basal repressed conformation. Given recent evidence that SFKs require cellular chaperones for proper folding [54
], it is likely that cells use additional mechanisms for deactivation of SFKs, such as Cbl-mediated degradation (Fig. 4)
.
View larger version (25K):
[in this window]
[in a new window]
|
Figure 4. Cbl-mediated regulation of activated Src-family PTKs. In their repressed form, SFKs are bound intramolecularly, forcing the kinase domain into an inactive conformation. Activation signals release the SH3 and SH2 domains from intramolecular ligands and allow the kinase domain to become active. Furthermore, the SH2 and SH3 domains can now interact with signaling proteins, as well as with Cbl, which targets activated SFKs for degradation. The presumed return of active SFKs to their repressed forms would require dephosphorylation of the autophosphorylation sites, phosphorylation of the C-terminal, negative regulatory tyrosine by CSK, and chaperone-mediated refolding. At present, it is not clear if fully activated SFKs do indeed return to their native, repressed state in cells. Thus, Cbl-mediated degradation may represent an important negative regulatory mechanism.
|
|
Defining the role of Cbl-mediated regulation of SFKs is important not only because of its intrinsic biological significance, but also because these PTKs interact with Cbl in a highly complex manner [55
]. In contrast to Syk/ZAP-70, which interacts with Cbl exclusively via the TKB domain, Cbl-SFK association involves binding between the SFK-SH3 domain and the proline-rich sequences in the C-terminal half of Cbl [9
]. The SH2 domains of SFKs are capable of interacting with phosphotyrosine motifs also found in the C-terminal half of Cbl [56
]. Furthermore, a phosphorylated tyrosine motif in SFKs, located in the activation loop, can interact with the Cbl-TKB domain [55
, 57
]. Thus, Cbl-SFK interactions likely involve multiple and possibly concurrent intramolecular interactions.
Recent analyses have demonstrated that antigen receptor-associated SFKs are direct targets of Cbl-mediated negative regulation. Overexpression analyses indicated that Cbl targets activated Fyn and Lck for degradation, reducing SFK-mediated activation of the serum response element used as a reporter [34
, 55
]. Moreover, the level of Fyn protein was increased in Cbl-/- fibroblasts and T cells [55
]. Recent results from our laboratory have extended these findings to the lymphoid-restricted SFK, Lck (unpublished results). Notably, use of a ZAP-70-negative Jurkat T-cell line p116 demonstrated that Cbl overexpression reduced the TCR-induced, Lck-mediated activation of the ERK pathway (unpublished results). Finally, Cbl has also been shown to inhibit Src- and Fyn-mediated enhancement of mitogenic signals in fibroblasts [58
]. Thus, it has become clear that Cbl can negatively regulate multiple members of the Src family of PTKs.
Recent studies have also revealed SFKs to be targets ofubiquitination. For example, c-Src, expressed in c-terminal src kinase (CSK)-/- fibroblasts and oncogenic v-Src were shown to be ubiquitinated. Furthermore, treatment with proteasome inhibitors led to increased protein levels [59
, 60
]. It is interesting that several groups have demonstrated recently that Cbl polyubiquitinates activated pools of Fyn, Lck, and Src leading to their proteasomal degradation ([34
, 35
] and unpublished results). Cbl-mediated ubiquitination was dependent on an intact RING finger domain. Interestingly, Cbl-/- cells showed reduced levels of Fyn and Lck ubiquitination, indicating an important role for endogenous Cbl in regulating SFK turnover ([34
] and unpublished results). A critical role for Cbl-dependent ubiquitination in regulating SFK protein levels was provided by analyses of Chinese hamster ovary (CHO)-TS20 cells, which express a thermolabile ubiquitin-activating (E1) enzyme [34
]. In this system, Cbl-mediated degradation of Fyn was blocked when E1 was rendered nonfunctional by growth at nonpermissive temperature.
While recent studies have clearly established Cbl as a negative regulator of SFK-mediated cellular activation, analyses of Src-dependent cell spreading and migration in macrophages and bone resorption in osteoclasts have suggested a positive role of Cbl in these responses downstream of Src [61
, 62
]. It is unclear whether this reflects a distinct role for Cbl, possibly as an adaptor, or the complexity of the readout system that requires Cbl-mediated, negative regulation to achieve a cellular response.
Recent findings have also implicated SFK ubiquitination in viral pathogenesis. For example, Winberg et al. [63
] demonstrated that the latent membrane protein 2A of Epstein-Barr virus enhances the ubiquitination of the SFK Lyn in B cells. Furthermore, the human papilloma virus E6 oncogene was shown to interact directly with the B-cell-specific SFK Blk and induce its degradation through the E3 ligase E6AP [64
]. At present, the role of Cbl in viral inactivation of SFKs remains unknown. Whether E6AP is a physiological ubiquitin ligase for Blk or other SFKs and whether Cbl and E6AP might work in concert are obvious questions that will require further examination.
Thus, in addition to Syk/ZAP-70, Cbl also functions as a negative regulator of antigen receptor signals by targeting SFKs, the proximal antigen receptor PTKs (Fig. 5
). Given the requirement of SFKs for Syk/ZAP-70 docking and activation, Cbl-mediated degradation of activated SFKs is likely to be central to Cbl’s role as a negative regulator of antigen receptor signaling.
View larger version (45K):
[in this window]
[in a new window]
|
Figure 5. Cbl-mediated ubiquitination and degradation of activated Syk/ZAP-70 and Src-family PTKs as a mechanism to regulate activation signals downstream of antigen receptors. This figure depicts the TCR, a prototype of antigen receptors. Cbl binds to activated PTKs of the Syk/ZAP-70 and SFK family and targets them for polyubiquitination and proteasome-mediated degradation. Degradation of active SFKs and Syk/ZAP-70 can reduce the phosphorylation of effector proteins, resulting in the reduction of the intensity of downstream signals. Furthermore, degradation of SFKs such as Lck can reduce the phosphorylation of ITAM chains (such as ) and activation of Syk/ZAP-70 PTKs.
|
|
|
INSIGHTS INTO Cbl PRTOEIN FUNCTION FROM Cbl-DEFICIENT MICE
|
|---|
The role of mammalian Cbl ubiquitin ligases has been examined genetically using Cbl-/- and Cbl-b-/- mice. The most prominent defect in both strains of mice is lymphocyte hyperresposiveness to antigen receptor triggering, supporting the role of Cbl as a negative regulator. Cbl-/- mice exhibit increased thymic cellularity, splenomegaly, and lymphadenopathy, as well as increased ductal branching and density of mammary fat pads [65
]. Cbl deficiency also enhances the positive selection of CD4+ thymocytes on a TCR transgenic background [66
]. Positive selection and thymocyte maturation are known to be Lck-dependent events [67
], suggesting that these observed phenotypes may reflect elevated Lck activity as a result of Cbl deficiency. Consistent with this idea, our in vitro studies have demonstrated that Cbl-/- T-cell lines have enhanced levels of kinase-active Lck and Fyn; moreover, these cell lines show a profound defect in SFK ubiquitination and subsequent degradation ([34
, 55
] and unpublished results). Cbl-/- thymocytes also have elevated levels of kinase-active ZAP-70, which could be due to increased SFK activity or a lack of direct Cbl-mediated negative regulation of ZAP-70 [68
]. Given that Lck kinase activity influences CD4+ fate decision, more studies are required to examine if Cbl regulates this process. Together, these data reinforce the role of Cbl as a key regulator of antigen receptor signaling and suggest Cbl ubiquitin-ligase activity plays a prominent role in thymocyte positive selection.
In contrast to Cbl-/- mice, no aberrant thymic development has been shown in Cbl-b-/- mice. However, mature T and B cells are hyperproliferative in these animals. In one study, Cbl-b-/- mice developed spontaneous autoimmunity [69
]; another study demonstrated that Cbl-b-/- mice exhibited increased susceptibility to experimental autoimmune encephalomyelitis [70
]. Moreover, mature T cells from Cbl-b-/- mice do not require CD28 costimulation for optimal T-cell proliferation and other activation events, such as lipid raft aggregation and IL-2 production [69
70
71
]. The distinct phenotypes of Cbl-/- and Cbl-b-/- mice may reflect differences in Cbl and Cbl-b expression between thymocytes and peripheral T cells. In Drosophila, it is known that the two forms of D-Cbl are regulated dynamically depending on cellular activation states [23
]. Further studies detailing the expression patterns of various Cbl family members during thymic development, maturation, and T-cell activation should help elucidate the redundant versus specific roles of Cbl family proteins. Whether the differences in severity of autoimmunity in Cbl-b-/- mice demonstrated by the two groups relate to distinct environmental factors, different levels of compensation by Cbl family members, or other factors remains to be investigated.
The biochemical phenotypes of Cbl-/- and Cbl-b-/- mice are also distinct. While Cbl-/- mice have enhanced phosphorylation of Lck, Zap-70 and downstream substrates such as LAT and SLP-76, Cbl-b-/- mice exhibit relatively selective elevation of Vav phosphorylation [68
, 69
]. These analyses again suggest that although Cbl family members possess a conserved domain structure, they are functionally distinct. However, a substantial level of functional redundancy does exist, as mice deficient in Cbl and Cbl-b are early embryonic lethal (H. Gu, personal communication). This vital role of Cbl proteins during development likely reflects their critical role in controlling receptor tyrosine kinase function, and mirrors developmental studies in Drosophila that have demonstrated that Cbl-mediated regulation of EGFR is essential during eye development and embryonic pattern formation [25
, 26
], and the role of SLI-1 in C. elegans vulval development [24
].
|
Cbl-DEPENDENT DOWN-REGULATION OF CELL-SURFACE RECEPTORS
|
|---|
As discussed above, Cbl-mediated regulation of EGFR, a model for ligand-induced receptor down-regulation, has been highly conserved throughout evolution. In contrast to the polyubiquitination of cytoplasmic PTKs, Cbl induces monoubiquitination of receptor tyrosine kinases such as EGFR, platelet-derived growth factor receptor, colony-stimulating receptor-1R, and the c-Met receptor (Fig. 6
). The Cbl-mediated monoubiquitination of receptor tyrosine kinases plays an important role in targeting these receptors to the lysosome, rather than the proteasome, for degradation [6
]. It is interesting that one phenotypic abnormality of Cbl-/- mice is increased surface TCR levels, which could reflect a failure to down-regulate the TCR [65
, 66
]. Previous studies have demonstrated that T-cell activation can be regulated by removal of the TCR from the cell surface [5
]. Upon activation, antigen receptor components are targeted for lysosomal degradation. Based on Cbl’s function as an E3 ubiquitin ligase and the observed phenotype of Cbl-/- mice, it seems likely that Cbl may also regulate cell-surface TCR levels by facilitating ubiquitination of receptor components and their lysosomal targeting. Although this hypothesis has not been tested directly, some lines of evidence hint that this might occur in vivo.
View larger version (20K):
[in this window]
[in a new window]
|
Figure 6. Cbl-mediated down-regulation of antigen receptor signaling. Signaling subunits of the antigen receptor, such as the chain (red), associate with proximal signaling proteins such as SFK or Syk/ZAP-70 PTKs (green). Cbl binds to these PTKs after activation, mediates their polyubiquitination, and targets them for proteasome-mediated degradation. Alternatively, Cbl could facilitate monoubiquitination of receptor subunits and target them for delivery to the lysosome for degradation. Receptors that escape ubiquitination (or are deubiquitinated) are recycled to the cell surface. Therefore, the E3 ubiquitin-ligase Cbl could tilt the balance in favor of lysosomal degradation.
|
|
Several studies have illustrated that ITAM-containing subunits of the TCR and FcRI are ubiquitinated upon antigen receptor stimulation [5
]. A recent study using an overexpression system demonstrated that the chain can be ubiquitinated in a Cbl-dependent manner [72
]. However, the significance of this finding is unclear, as the investigators demonstrated polyubiquitination of rather than the monoubiquitination observed under physiologic activation conditions [73
]. Notably, ZAP-70 was shown to serve as an adaptor to recruit Cbl for chain ubiquitination, and it remains possible that Lck will function in a similar manner. Use of Lck or ZAP-70 as an adaptor to link Cbl to the chain would provide an elegant mechanism to ensure Cbl-mediated ubiquitination of only the activated pool of antigen receptors. Ubiquitin ligases other than Cbl may also participate in antigen receptor ubiquitination, as Cbl and Nedd4, an unrelated ubiquitin ligase, were found recently to move into lipid rafts upon FcRI stimulation [74
].
Thus, the hyperresponsive phenotype of Cbl-deficient leukocytes may relate in part to the inability to down-regulate antigen receptors (or other tyrosine kinase-coupled receptors) efficiently, consistent with the increased surface TCR levels in Cbl-/- mice [65
, 66
]. Thus, Cbl may regulate leukocyte activation through polyubiquitination of activated PTKs (Syk/ZAP-70, SFKs) as well as through monoubiquitination of activated cell-surface receptors, facilitating their targeting to lysosomes. These two mechanisms are unlikely to be mutually exclusive and could allow adequate attenuation of activation signals and prevent uncontrolled immune responses observed in Cbl-deficient mice.
|
INTERACTION OF Cbl WITH NON-PTK SIGNALING PROTEINS
|
|---|
Many studies over the last seven years have defined associations of Cbl with signaling proteins other than PTKs [7
, 8
, 10
]. Although these protein-protein interactions have been interpreted in the context of a positive adaptor role for Cbl, there is only meager evidence for such a role. In fact, such a role is incompatible with the knockout mouse phenotypes described above. As discussed below, these protein-protein interactions may serve to target Cbl ubiquitin ligases to additional non-PTK targets and provide a mechanism to down-regulate PTK signaling that works in concert with direct negative regulation of PTKs.
Grb2-Cbl complexes
The proline-rich region of Cbl associates prominently with the N-terminal SH3 domain of Grb2 in most cell types examined [7
, 8
, 10
]. This association is constitutive, and it remains controversial, whether receptor engagement enhances or decreases this association. Given studies that the SH2 domain of Grb2, by itself or through Shc, may interact with phosphorylated antigen receptor components such as the chain, this interaction could provide an alternate mechanism to recruit Cbl ubiquitin ligases to activated antigen receptors. Notably, Grb2 is known to function as an adaptor to recruit Cbl to receptor tyrosine kinases [10
]. Because Cbl and the Ras exchange factor SOS interact with the N-terminal SH3 domain of Grb2, Cbl may sequester Grb2 from SOS and thus inhibit Ras activation. This would provide an independent mechanism to down-regulate antigen receptor signals [8
].
Cbl and the Crk family of adapter proteins
The Crk adapter protein family is comprised of Crk-I and Crk-II, alternatively spliced products of a single gene, and Crk-L, a distinct Crk-like gene product [9
]. Stimulation through antigen receptors induces a prominent association of tyrosine-phosphorylated Cbl with the Crk adapter proteins [10
]. Notably, Cbl is the most dominant phosphoprotein associated with Crk in activated lymphocytes, and a relatively large pool of Cbl is recruited into this complex. In vitro binding analyses indicate that the Crk SH2 domain binds pY774 on Cbl [10
], suggesting that Cbl may form ternary complexes with Crk-SH3 domain-associated proteins such as C3G (Fig. 7
). Indeed, a ternary complex of Cbl, CrkL, and C3G was detected upon TCR ligation in Jurkat T cells [10
]. C3G is a guanine nucleotide exchange factor for Rap1. Thus, Cbl-Crk-C3G complexes generated upon antigen receptor activation provide a potential mechanism to regulate guanine nucleotide exchange on Rap1 or other small G-protein targets of C3G. We propose that Cbl proteins could negatively regulate C3G-mediated Rap activation by enhancing Crk or C3G polyubiquitination and subsequent degradation. However, this hypothesis remains to be tested.
View larger version (49K):
[in this window]
[in a new window]
|
Figure 7. Degradation of Cbl-associated effector proteins. Cbl is known to form activation-induced complexes with the Rap exchange factor C3G through Crk, PI-3K via p85, Vav, and Blnk. Ubiquitination of the adaptors or effector proteins can target them for degradation, resulting in Cbl-mediated negative regulation of antigen receptor signals.
|
|
Active [guanosine 5'-triphosphate-bound] Rap1 interacts with Ras targets (such as Raf and Ras-GAP) and has led to the suggestion that Rap1 functions in an antagonistic way toward Ras by sequestering Ras targets [75
]. It has therefore been argued that Cbl-Crk complexes could mediate the Ras-mitogen-activated protein kinase block observed in anergic T cells. Ras-dependent activation of the IL-2 promoter upon expression of an oncogenic Cbl mutant (70Z) in Jurkat cells is consistent with a role for Cbl in Ras regulation [10
]. However, the Ras antagonistic function of Rap has been challenged recently, casting doubt on this proposal [75
].
Cbl and the p85 subunit of phosphatidyl insositol 3 kinase (PI-3K)
PI-3K consists of an 85-kDa regulatory subunit and a 110-kDa catalytic subunit. PI-3K catalyzes the generation of phosphoinositide products that regulate diverse cellular functions, including mitogenesis, cell survival, migration, and vesicular membrane traffic. The Cbl phosphorylation-dependent (and hence, cellular activation-dependent) association between Cbl and p85 has been demonstrated in a number of cell types, including antigen receptor-stimulated lymphocytes [7
, 8
, 10
]. This interaction involves the p85-SH2 domain binding to pY731 in Cbl.
Although an adapter role of Cbl in recruiting PI-3K to antigen receptors cannot be excluded, recent evidence favors a likely negative regulatory role of Cbl proteins for the PI-3K pathway (Fig. 7)
. First, overexpression of Cbl was shown to enhance p85 polyubiquitination [76
], and p85 ubiquitination was found to be defective in Cbl-b-/- T lymphocytes [77
]. Second, the PI-3K activity was found to be important for hyperproliferation of Cbl-b-/- T cells. The investigators concluded that the Cbl-mediated ubiquitination did not lead to degradation of p85, but further investigations are needed to establish such a mechanism conclusively. Alternatively, Cbl may sequester PI-3K from its normal substrates. Nevertheless, these findings strongly support a role for Cbl ubiquitin ligases as negative regulators of PI-3K activation.
Cbl and Vav
Cbl also associates with Vav, a hematopoietic-restricted Rac/Rho guanine nucleotide-exchange factor, upon TCR stimulation. This interaction is thought to be mediated by the Vav SH2 domain binding to Cbl pY700 [9
], although Cbl-b association with Vav was shown to require Grb2-SH3 as well as SH2 domains [22
]. Moreover, Cbl-b was shown to inhibit the Vav-dependent increase in JNK activity in transfected fibroblasts [78
]. One prominent biochemical phenotype of Cbl-b-/- mice is elevated levels of phosphorylated Vav upon antigen receptor stimulation. A recent study of Cbl-b-/- T cells indicates that Cbl-b negatively regulates antigen receptor-mediated activation of Vav [71
]. Indeed, introducing Cbl-b deficiency into Vav-/- mice rescued the proliferation defect of peripheral T cells and led to spontaneous autoimmunity as the mice aged. Furthermore, Cbl-b deficiency restored receptor clustering and CDC42 activation in Vav-/- T cells [71
]. It will be of considerable interest to determine whether Cbl-mediated, negative regulation of Vav is mediated via polyubiquitination and subsequent proteasome-mediated degradation (Fig. 7)
. Notably, Vav has been shown to be a target of suppressor of cytokine signaling-mediated ubiquitination and degradation [79
].
Cbl and B-cell linker protein (BLNK)
Recently, Cbl was shown to interact with the B-cell adaptor protein BLNK in a TKB domain-dependent manner [80
]. This is the first example of the Cbl-TKB domain binding to a non-PTK target. BLNK is functionally analogous to LAT and the related adaptor protein Slp-76 in T cells and is a key mediator of Ca++ mobilization upon antigen receptor stimulation. It is notable that Cbl-/- DT-40 chicken B cells exhibited a prolonged BCR-induced calcium flux [80
]. The mechanism of Cbl-mediated negative regulation of BLNK remains to be determined. So far, all proteins that bind the Cbl-TKB domain are targets of Cbl’s E3 ligase activity. Therefore, it is possible that Cbl also regulates BLNK via polyubiquitination (Fig. 7)
.
|
CONCLUSION
|
|---|
The Cbl family of proteins is an evolutionarily conserved group of negative regulators of PTK signaling. Cbl’s multidomain structure enables it to interact with a variety of proteins. The RING finger domain of Cbl interacts with E2 enzymes of the ubiquitination machinery, allowing Cbl to function as an E3 ubiquitin ligase. Recruitment of Cbl proteins to activated tyrosine kinase receptors promotes their monoubiquitination, resulting in sorting to the lysosome and subsequent degradation. Importantly, Cbl interacts directly with and promotes the polyubiquitination of antigen receptor-associated PTKs such as Syk/ZAP-70 and SFKs, targeting them for proteasomal degradation. Cbl interacts selectively with activated pools of PTKs, providing an elegant mechanism for postactivation down-regulation of antigen receptor signals. Down-regulation of the cell-surface receptors and degradation of tyrosine kinase target proteins known to complex with Cbl-ubiquitin ligases likely provide additional mechanisms of Cbl-mediated, negative regulation. Thus, the Cbl family of ubiquitin ligases represents a novel mode of regulating early antigen receptor signaling events through controlled degradation of proximal components of the signaling machinery. Biochemical as well as genetic studies provide strong evidence for a critical role of this powerful new mechanism in attenuating leukocyte activation. Future studies should focus on integrating Cbl-mediated ubiquitination into a broader understanding of antigen receptor trafficking and down-regulation.
|
ACKNOWLEDGEMENTS
|
|---|
This work was supported by National Institutes of Health (NIH) grants CA76118, CA87986, and CA75075 (H.B.) and the DOD Breast Cancer Research Program (DAMD17-001-0296). N. R. is an HHMI Pre-doctoral Fellow. I. D. is supported by NIH Training Grant (5T32AR07530).
Received December 18, 2001;
revised December 18, 2001;
accepted February 18, 2002.
|
REFERENCES
|
|---|
-
van Leeuwen, J. E., Samelson, L. E. (1999) T cell antigen-receptor signal transduction Curr. Opin. Immunol. 11,242-248[Medline]
-
Qian, D., Weiss, A. (1997) T cell antigen receptor signal transduction Curr. Opin. Cell Biol. 9,205-212[Medline]
-
Kane, L. P., Lin, J., Weiss, A. (2000) Signal transduction by the TCR for antigen Curr. Opin. Immunol. 12,242-249[Medline]
-
Tonks, N. K., Neel, B. G. (2001) Combinatorial control of the specificity of protein tyrosine phosphatases Curr. Opin. Cell Biol. 13,182-195[Medline]
-
Alcover, A., Alarcon, B. (2000) Internalization and intracellular fate of TCR-CD3 complexes Crit. Rev. Immunol. 20,325-346[Medline]
-
Waterman, H., Yarden, Y. (2001) Molecular mechanisms underlying endocytosis and sorting of ErbB receptor tyrosine kinases FEBS Lett. 490,142-152[Medline]
-
Thien, C. B., Langdon, W. Y. (2001) Cbl: many adaptations to regulate protein tyrosine kinases Nat. Rev. Mol. Cell Biol. 2,294-307[Medline]
-
Lupher, M. L., Jr, Rao, N., Eck, M. J., Band, H. (1999) The Cbl protooncoprotein: a negative regulator of immune receptor signal transduction Immunol. Today 20,375-382[Medline]
-
Miyake, S., Lupher, M. L., Jr, Andoniou, C. E., Lill, N. L., Ota, S., Douillard, P., Rao, N., Band, H. (1997) The Cbl protooncogene product: from an enigmatic oncogene to center stage of signal transduction Crit. Rev. Oncog. 8,189-218[Medline]
-
Tsygankov, A. Y., Teckchandani, A. M., Feshchenko, E. A., Swaminathan, G. (2001) Beyond the RING: CBL proteins as multivalent adapters Oncogene 20,6382-6402[Medline]
-
Weissman, A. M. (2001) Themes and variations on ubiquitylation Nat. Rev. Mol. Cell Biol. 2,169-178[Medline]
-
Stoltze, L., Nussbaum, A. K., Sijts, A., Emmerich, N. P., Kloetzel, P. M., Schild, H. (2000) The function of the proteasome system in MHC class I antigen processing Immunol. Today 21,317-319[Medline]
-
Orian, A., Gonen, H., Bercovich, B., Fajerman, I., Eytan, E., Israel, A., Mercurio, F., Iwai, K., Schwartz, A. L., Ciechanover, A. (2000) SCF(beta)(-TrCP) ubiquitin ligase-mediated processing of NF-kappaB p105 requires phosphorylation of its C-terminus by IkappaB kinase EMBO J 19,2580-2591[Medline]
-
Hicke, L. (2001) Protein regulation by monoubiquitin Nat. Rev. Mol. Cell Biol. 2,195-201[Medline]
-
Pickart, C. M. (2001) Mechanisms underlying ubiquitination Annu. Rev. Biochem. 70,503-533[Medline]
-
Joazeiro, C. A., Wing, S. S., Huang, H., Leverson, J. D., Hunter, T., Liu, Y. C. (1999) The tyrosine kinase negative regulator c-Cbl as a RING-type, E2-dependent ubiquitin-protein ligase Science 286,309-312[Abstract/Free Full Text]
-
Levkowitz, G., Waterman, H., Ettenberg, S. A., Katz, M., Tsygankov, A. Y., Alroy, I., Lavi, S., Iwai, K., Reiss, Y., Ciechanover, A., Lipkowitz, S., Yarden, Y. (1999) Ubiquitin ligase activity and tyrosine phosphorylation underlie suppression of growth factor signaling by c-Cbl/Sli-1 Mol. Cell 4,1029-1040[Medline]
-
Yokouchi, M., Kondo, T., Houghton, A., Bartkiewicz, M., Horne, W. C., Zhang, H., Yoshimura, A., Baron, R. (1999) Ligand-induced ubiquitination of the epidermal growth factor receptor involves the interaction of the c-Cbl RING finger and UbcH7 J. Biol. Chem. 274,31707-31712[Abstract/Free Full Text]
-
Keane, M. M., Ettenberg, S. A., Nau, M. M., Banerjee, P., Cuello, M., Penninger, J., Lipkowitz, S. (1999) cbl-3: A new mammalian cbl family protein Oncogene 18,3365-3375[Medline]
-
Robertson, H., Hime, G. R., Lada, H., Bowtell, D. D. (2000) A Drosophila analogue of v-Cbl is a dominant-negative oncoprotein in vivo Oncogene 19,3299-3308[Medline]
-
Langdon, W. Y., Hyland, C. D., Grumont, R. J., Morse, H. C., III (1989) The c-cbl proto-oncogene is preferentially expressed in thymus and testis tissue and encodes a nuclear protein J. Virol. 63,5420-5424
TOP
ABSTRACT
INTRODUCTION
