The Usp8 deubiquitination enzyme is post-translationally modified by tyrosine and serine phosphorylation.

The ERBB1-ERBB4 receptors belong to a family of receptor tyrosine kinases that trigger a network of signaling pathways after ligand binding, thereby regulating cellular growth, differentiation and development. Ligand-induced signaling through ERBB1, also known as EGFR, is attenuated by the clathrin-dependent receptor-mediated endocytosis and RING E3-ligase Cbl-mediated receptor ubiquitination, which is followed by incorporation into multi-vesicular bodies (MVBs) and subsequent degradation in lysosomes. Before incorporation into MVBs, the EGFR is deubiquitinated by Usp8. We previously demonstrated that Usp8 is tyrosine phosphorylated in an EGFR- and SRC-kinase dependent manner. In the present study we show that overexpression of constitutively active SRC enhances constitutive and ligand-induced Usp8 tyrosine phosphorylation. We also show that enhanced endosomal recycling of the EGFR induced by TGFα stimulation is associated with decreased Usp8 tyrosine phosphorylation. We therefore hypothesize that tyrosine phosphorylation of Usp8 could regulate the function of Usp8. To identify Usp8 tyrosine phosphorylation site(s), we used Usp8 deletion constructs, site-directed mutagenesis of nine individual Usp8 tyrosine residues and mass spectrometry (MS) analysis. Our results demonstrate that the MIT-domain is necessary for ligand-induced tyrosine phosphorylation of Usp8 1-504. However, mutation of three MIT domain tyrosine residues did not abolish Usp8 tyrosine phosphorylation. Similar results were obtained upon mutation of six exposed tyrosine residues in the Rhod domain and linker region. Repeated MS analysis of both Usp8 WT and C748A mutants readily detected serine phosphorylation, including the S680 14-3-3 binding site, but did not reveal any phospho-tyrosine residues. Notably, mutation of the tyrosine residue in the Usp8 14-3-3 binding motif (Y679) did not abolish phosphoserine-dependent binding of 14-3-3 to Usp8. Our findings are most consistent with the model that MIT domain-dependent recruitment of Usp8 to endosomal membranes is important for low stoichiometry SRC-mediated tyrosine phosphorylation of multiple Usp8 tyrosines. Our findings demonstrate that Usp8 is a target for the post-translational serine and tyrosine phosphorylation, most likely characterized by low abundant tyrosine phosphorylation on multiple residues, and high abundant serine phosphorylation on several residues.

Cbl and Itch binding sites in ERBB4 CYT-1 and CYT-2 mediate K48- and K63-polyubiquitination, respectively.

ERBB receptors have an important function in mammalian development and normal physiology, but overexpression and poor downregulation of ERBB receptors have been associated with malignant growth. Ligand-induced ERBB receptor signaling is terminated by clathrin-dependent receptor endocytosis, followed by incorporation of activated receptor complexes into multi-vesicular bodies and subsequent degradation in lysosomes. In the case of ERBB1, also known as the EGF receptor, it has been shown that ubiquitination serves as a signal to facilitate internalization and subsequent endosomal sorting, but little is known about the role of ubiquitination of other ERBB receptors. In the present study we investigated the regulation of ubiquitination and deubiquitination of the ERBB4 CYT-1 and CYT-2 isoforms in the context of chimeric EGFR-ERBB4 receptors. We demonstrate that EGFR-ERBB4 CYT-2 chimera shows decreased ligand-induced downregulation and EGF-degradation, as well as enhanced EGF recycling, when compared to EGFR-ERBB4 CYT-1. Moreover we show that the mutation Y1103F in the binding site for Cbl which is present in both CYT-1 and CYT-2, does not influence ERBB4 endosomal trafficking. We further demonstrate that total ligand-induced ubiquitination of CYT-1 is higher than that of CYT-2, whereby CYT-1 ubiquitination is mainly dependent on the PPXY(1056) Itch binding site for the E3-ligase Itch which is only present in CYT-1, while that of CYT-2 is dependent on the Y1103 Cbl binding site. The E3-ligase c-Cbl is more efficiently phosphorylated upon EGF stimulation of the CYT-2 than the CYT-1 isoform. Moreover our data show that the pY1103 Cbl binding site is required for K48-polyubiquitination of both CYT-1 and CYT-2, whereas the PPXY(1056) Itch binding site is required for K63-polyubiquitination of CYT-1. We further demonstrate that EGF stimulation of EGFR-ERBB4 CYT-1 and CYT-2 does not result in efficient binding to and tyrosine phosphorylation of the ESCRT-0 subunit Hrs. Finally, even though CYT-1 shows ligand-induced K63-polyubiquitination, it is not subjected to deubiquitination by the K63 polyubiquitin-specific AMSH deubiquitinating enzyme, while CYT-1 is slightly deubiquitinated by USP8. We conclude that Cbl and Itch binding sites in ERBB4 CYT-1 and CYT-2 mediate K48- and K63-polyubiquitination, respectively.

Recycling of EGFR and ErbB2 is associated with impaired Hrs tyrosine phosphorylation and decreased deubiquitination by AMSH.

ErbB receptors play an important role in normal cellular growth, differentiation and development, but overexpression or poor downregulation can result in enhanced signaling and cancerous growth. ErbB signaling is terminated by clathrin-dependent receptor-mediated endocytosis, followed by incorporation in multi-vesicular bodies and subsequent degradation in lysosomes. In contrast to EGFR, ErbB2 displays poor ligand-induced downregulation and enhanced recycling, but the molecular mechanisms underlying this difference are poorly understood. Given our previous observation that both EGFR and an EGFR-ErbB2 chimera undergo Cbl-mediated K63-polyubiquitination, we investigated in the present study whether activation of the EGFR and the EGFR-ErbB2 chimera is associated with tyrosine phosphorylation of the ESCRT-0 complex subunit Hrs and AMSH-mediated deubiquitination. EGF stimulation of the EGFR resulted in efficient Hrs tyrosine phosphorylation and deubiquitination by the K63-polyubiquitin chain-specific deubiquitinating enzyme AMSH. In contrast, EGF activation of EGFR-ErbB2 showed significantly decreased Hrs tyrosine phosphorylation and deubiquitination by AMSH. To test whether this phenotype is the result of endosomal recycling, we induced recycling of the EGFR by stimulation with TGFα. Indeed, even though TGFα-stimulation of EGFR is associated with efficient ligand-stimulated K63-polyubiquitination, we observed that Hrs tyrosine phosphorylation as well as AMSH-mediated deubiquitination is significantly reduced under these conditions. Using various EGFR-ErbB2 chimeras, we demonstrate that enhanced recycling, decreased Hrs tyrosine phosphorylation and decreased AMSH mediated deubiquitination of EGFR-ErbB2 chimeras is primarily due to the presence of ErbB2 sequences or the absence of EGFR sequences C-terminal to the Cbl binding site. We conclude that endosomal recycling of the EGFR and ErbB2 receptors is associated with significantly impaired tyrosine phosphorylation of the ESCRT-0 subunit Hrs as well as decreased deubiquitination by AMSH, which is consistent with the finding that recycling receptors are not efficiently incorporated in the MVB pathway.

ERBB2 is a target for USP8-mediated deubiquitination.

Overexpression and poor downregulation of ErbB receptor tyrosine kinases are associated with enhanced signaling and tumorigenesis. Attenuation of EGF-receptor (EGFR) signaling is mediated by endocytosis and ubiquitination by the E3-ligase Cbl. En route to lysosomes, but before incorporation of the EGFR into internal vesicles of MVBs, the EGFR undergoes Usp8-mediated deubiquitination. ErbB2 displays enhanced recycling back to the cell surface, and therefore we hypothesized that Usp8 is not part of the ErbB2 trafficking pathway. Here, we demonstrate, in the context of a chimeric EGFR-ErbB2 receptor, that (i) EGF induces pY1091 Cbl binding site-dependent K63-polyubiquitination of EGFR-ErbB2, (ii) Cbl is tyrosine phosphorylated upon stimulation of EGFR-ErbB2 wt and Y1091F mutant receptor, (iii) EGF-induced activation of EGFR-ErbB2 induces Usp8 tyrosine phosphorylation, and (iv) ubiquitination of the EGFR-ErbB2 wt and Y1091F mutant is enhanced upon coexpression of catalytically inactive Usp8-C748A in the presence and absence of EGF. We further show that Usp8 tyrosine phosphorylation upon stimulation of EGFR-ErbB2 is (a) independent of Y1091, (b) dependent on Src- and EGFR-ErbB2-kinase activity, (c) enhanced upon coexpression of Usp8-C748A, and (d) partly dependent on the Microtubule Interacting and Transport (MIT) domain of Usp8. Our findings demonstrate that Usp8 is part of the ErbB2 endosomal trafficking pathway.

Effects of the EGFR Inhibitor Erlotinib on Magnesium Handling.

A mutation in pro-EGF causes isolated hypomagnesemia, and monoclonal antibodies targeting the extracellular domain of the EGF receptor (EGFR) affect epithelial Mg(2+) transport. The effect of the EGFR tyrosine kinase inhibitor erlotinib on Mg(2+) homeostasis, however, remains unknown. Here, we injected C57BL/6 mice with erlotinib for 23 days and observed a small but significant decrease in serum Mg(2+) concentrations at days 16 and 23, but the fractional excretion of Mg(2+) remained unchanged after 23 days. Semiquantitative immunohistochemical evaluation did not reveal detectable changes in renal expression of transient receptor potential melastatin 6 (TRPM6) protein, the channel that mediates Mg(2+) reabsorption. Patch clamp analysis in TRPM6-expressing cells demonstrated that 30 muM erlotinib inhibited EGF-induced changes in TRPM6 current density and tyrosine phosphorylation of EGFR; 0.3 muM erlotinib did not have these effects. Furthermore, 30 muM erlotinib inhibited EGF-stimulated increases in the mobile fraction of endomembrane TRPM6 channels. In summary, erlotinib can influence Mg(2+) handling but its effect on the systemic Mg(2+) concentration seems less potent than that observed with antibody-based EGFR inhibitors. These data suggest that typical human dosages of erlotinib are unlikely to severely affect serum Mg(2+) concentrations.

Angiomotin-like protein 1 controls endothelial polarity and junction stability during sprouting angiogenesis.

RATIONALE: We have previously shown that angiomotin (Amot) is essential for endothelial cell migration during mouse embryogenesis. However, approximately 5% of Amot knockout mice survived without any detectable vascular defects. Angiomotin-like protein 1 (AmotL1) potentially compensates for the absence of Amot as it is 62% homologous to Amot and exhibits similar expression pattern in endothelial cells. OBJECTIVE: Here, we report the identification of a novel isoform of AmotL1 that controls endothelial cell polarization and directional migration. METHODS AND RESULTS: Small interfering RNA-mediated silencing of AmotL1 in mouse aortic endothelial cells caused a significant reduction in migration. In confluent mouse pancreatic islet endothelial cells (MS-1), AmotL1 colocalized with Amot to tight junctions. Small interfering RNA knockdown of both Amot and AmotL1 in MS-1 cells exhibited an additive effect on increasing paracellular permeability compared to that of knocking down either Amot or AmotL1, indicating both proteins were required for proper tight junction activity. Moreover, as visualized using high-resolution 2-photon microscopy, the morpholino-mediated knockdown of amotl1 during zebrafish embryogenesis resulted in vascular migratory defect of intersegmental vessels with strikingly decreased junction stability between the stalk cells and the aorta. However, the phenotype was quite distinct from that of amot knockdown which affected polarization of the tip cells of intersegmental vessels. Double knockdown resulted in an additive phenotype of depolarized tip cells with no or decreased connection of the stalk cells to the dorsal aorta. CONCLUSIONS: These results cumulatively validate that Amot and AmotL1 have similar effects on endothelial migration and tight junction formation in vitro. However, in vivo Amot appears to control the polarity of vascular tip cells whereas AmotL1 mainly affects the stability of cell-cell junctions of the stalk cells.

ErbB2 and ErbB4 Cbl binding sites can functionally replace the ErbB1 Cbl binding site.

Poor downregulation of ErbB receptors is associated with enhanced downstream signaling and tumorigenesis. It has been suggested that poor downregulation of ErbB-2, -3 and -4 receptors when compared to ErbB1 is due to decreased recruitment of Cbl E3 ligase proteins. However, a highly conserved Cbl binding site is not only present in ErbB1/EGFR (FLQRpY(1045)SSDP), but also in ErbB2 (PLQRpY(1091)SEDP) and ErbB4 (STQRpY(1103)SADP). We therefore replaced the ErbB1 Cbl binding site by that of ErbB2 and ErbB4. Whereas retrovirally infected NIH3T3 cells containing the EGFR Y1045F mutation showed dramatically impaired Cbl recruitment, EGFR ubiquitination and delayed EGFR degradation, replacement of the EGFR Cbl binding site by that of ErbB2 or ErbB4 did not affect Cbl recruitment, receptor-ubiquitination, -degradation, -downregulation or ligand degradation. We conclude that poor downregulation of ErbB2 and ErbB4 receptors is not due to sequence variations in the Cbl binding site of these receptors.