AMOT (angiomotin) is a membrane-associated protein that is expressed in ECs (endothelial cells) and controls migration, TJ (tight junction) formation, cell polarity and angiogenesis. Recent studies have revealed that AMOT and two AMOT-like proteins, AMOTL1 and AMOTL2, play critical roles in the Hippo pathway by regulating the subcellular localization of the co-activators YAP (Yes-associated protein) and TAZ (transcriptional co-activator with PDZ-binding motif). However, it has been unclear how AMOT is regulated. In the present study, we report that AMOT undergoes proteasomal degradation. We identify three members of Nedd4 (neural-precursor-cell-expressed developmentally down-regulated)-like ubiquitin E3 ligases, Nedd4, Nedd4-2 and Itch, as the ubiquitin E3 ligases for the long isoform of AMOT, AMOT/p130. We demonstrate that Nedd4, Nedd4-2 and Itch mediate poly-ubiquitination of AMOT/p130 in vivo. Overexpression of Nedd4, Nedd4-2 or Itch leads to AMOT/p130 proteasomal degradation. Knockdown of Nedd4, Nedd4-2 and Itch causes an accumulation of steady-state level of AMOT/p130. We also show that three L/P-PXY motifs of AMOT/p130 and the WW domains of Nedd4 mediate their interaction. Furthermore, Nedd4-like ubiquitin E3 ligases might compete with YAP for the binding to AMOT/p130, and subsequently targeting AMOT/p130 for ubiquitin-dependent degradation. Together, these observations reveal a novel post-translational regulatory mechanism of AMOT/p130.
Protein ubiquitination has emerged as a fundamental mechanism for regulating the half lives and activity of many cellular proteins. The specificity of ubiquitination reactions is achieved by the E3 ubiquitin ligases, which mediate the transfer of ubiquitin from E2 ubiquitin-conjugating enzymes to substrates . Ubiquitination controls turnover and abundance of proteins by targeting them for proteasomal degradation. The Nedd4 (neural-precursor-cell-expressed developmentally down-regulated 4)-like family contains nine members in humans including Nedd4, Nedd4-2, Itch, Smurf (SMAD-specific E3 ubiquitin protein ligase) 1, Smurf2, WWP (WW domain-containing E3 ubiquitin protein ligase) 1, WWP2, NedL (Nedd4-like E3 ubiquitin-protein ligase) 1 and NedL2, and they are characterized by a distinct modular domain architecture, with each member consisting of a Ca2+/lipid-binding domain (C2 domain) involved in membrane targeting, 2–4 WW domains conferring substrate specificity, and a HECT (homologous with E6-associated protein C-terminus)-type ligase domain required for catalytic activities [2,3]. This family of ubiquitin E3 ligases has been shown to regulate diverse biological processes through targeted degradation of proteins that generally have one or more L/P-PXY motifs for WW domain recognition . Since the WW domains of Nedd4-like family of E3 ubiquitin ligases have similar binding affinities, albeit to different degrees, for the L/P-PXY motif-containing substrates, these substrates can be degraded by one or more Nedd4-like E3 ubiquitin ligases at different cellular contexts. For example, the oncoprotein ErbB4 (v-erb-a erythroblastic leukaemia viral oncogene homologue 4) was found to be negatively regulated by WWP1, Itch and NedL1 , whereas the TGF-β (transforming growth factor β) receptor can be degraded by Smurf1/2, WWP1 and NEDD4-2 . To date, a large number of P/L-PXY motif-containing proteins have been identified as substrates of Nedd4-like E3 ubiquitin ligases by various methods. But given the prevalence of this motif in the human proteome , additional substrates for this family of ubiquitin E3 ligases remain to be identified and characterized.
AMOT (angiomotin) was identified in a yeast two-hybrid screen as an interacting protein of angiostatin, an inhibitor of angiogenesis . In vivo and in vitro data indicate an essential role of AMOT in endothelial cell motility . AMOT-knockout mice exhibit severe vascular insufficiency in the intersomitic region as well as dilated vessels in the brain . The chemotactic response to VEGF (vascular endothelial growth factor) was abolished in AMOT-deficient cells, suggesting a critical role for AMOT during vascular patterning and endothelial polarization . Mechanically, AMOT binds to Rich1 [RhoGAP interacting with Cdc42 (cell division cycle 42)-interacting protein 4 homologues protein 1] and is thereby targeted to a protein complex at TJs (tight junctions) containing the PDZ-domain proteins Pals1 (protein associated with Lin Seven 1), Patj (Pals1-associated TJ protein) and Par-3 (proteinase-activated receptor 3) . AMOT and Rich1 maintain TJ integrity by the co-ordinatated regulation of Cdc42 and by linking specific components of the TJ to intracellular protein trafficking . Furthermore, AMOT can form a ternary complex with Patj [or its paralogue Mupp1 (multiple PDZ-domain-containing protein 1)] and the RhoGEF (guanine-nucleotide-exchange factor) protein Syx . This complex controls RhoA activity at the leading edge of migrating cells, and knockdown of either AMOT or Syx results in inhibition of migration of intersegmental vessels during zebrafish angiogenesis . AMOT belongs to a protein family composed three members, AMOT and two paralogues AMOTL (AMOT-like)-1 and AMOTL2. AMOTL1 and AMTOL2 are relatively less studied than AMOT. It was reported that AMOTL1 localizes to endothelial lamellipodia and TJs, regulating sprouting angiogenesis by affecting tip-cell migration as well as controlling cell–cell adhesions in vivo . AMOTL2 was found to regulate cell migration by binding to, and promoting peripheral membrane translocation of, the non-receptor tyrosine kinase c-Src in zebrafish models .
Recently, several reports demonstrated that members of the AMOT family (AMOT/p130, AMOTL1 and AMOTL2) are important regulators of the downstream effectors of the Hippo pathway YAP (Yes-associated protein) and its paralogue, TAZ (transcriptional co-activator with PDZ-binding motif) [13–16]. The Hippo signalling pathway is a highly conserved pathway that controls organ size by phosphorylating and inhibiting the transcription co-activators YAP/TAZ in mammals and Yki (Yorkie) in Drosophila . The Hippo pathway also plays a critical role in the self-renewal and expansion of stem cells and tissue-specific progenitor cells . The AMOT family of proteins was found to specifically interact with YAP/TAZ through the WW domain–PPXY interaction. The transcriptional activities of YAP and TAZ were inhibited by AMOT-mediated TJ localization [13–16], and down-regulation of AMOTL2 in cells promotes EMT (epithelial–mesenchymal transition) . Although the AMOT protein family plays important roles in cell migration, TJ formation, angiogenesis and growth control [10,18–20], whether these proteins are regulated at transcriptional, translational and post-translation level remains largely unknown.
In the present study, we report that AMOT/p130 is negatively regulated by three members of Nedd4-like ubiquitin E3 ligases. We demonstrate that Nedd4, Nedd4-2 and Itch directly interact with and act as robust E3 ubiquitin ligases for AMOT/p130. Overexpression of these three ubiquitin E3 ligases promote cytoplasmic translocation, ubiquitination and proteasomal degradation of AMOT/p130. Conversely, knockdown of these three ubiquitin ligases by siRNAs (small interfering RNAs) increases AMOT/p130 abundance.
Cell culture and transfection
HEK (human embryonic kidney)-293T and H1299 cells were obtained from the A.T.C.C. HEK-293T cells were maintained in DMEM (Dulbecco's modified Eagle's medium) with 10% FBS (fetal bovine serum). H1299 cells were maintained in RPMI 1640 medium with 10% FBS. Cells were transiently transfected using Lipofectamine™ (Invitrogen) according to the manufacturer's instructions.
Human AMOT/p130 and AMOT/p80 constructs were provided by Professor Lars Holmgren (Cancer Centre Karolinska Institute, Stockholm, Sweden). HA (haemagglutinin)–AMOTL1 (JEAP) and AMOTL2 (MASCOT) were provided by Dr Makoto Adachi (Kyoto University, Kyoto, Japan). WT (wild-type) and a CA (catalytically inactive) mutant of Itch were provided by Dr Gerry Melino (Leicester University, Leicester, U.K.). Nedd4-2 construct was provided by Dr Christie P. Thomas (University of Iowa, Iowa City, U.S.A.). Myc–Smurf1/2 and Myc–WWP1 were provided by Dr Kohei Miyazono (University of Tokyo, Tokyo, Japan). Myc–WWP2 was provided by Dr Ying Jin (Chinese Academy of Sciences, Shanghai, China). All other AMOT- or Nedd4-deletion mutants were generated using the KOD-Plus Mutagenesis Kit (Toyobo).
RNAi (RNA interference)
The RNAi oligonucleotides were purchased from Genepharma. The RNAi oligonucleotide sequences for Nedd4 were: RNAi #1, 5′-UGGCGAUUUGUAAACCGAAdTdT-3′ (where dT is deoxyribothymidine) and RNAi #2, 5′-GAUG-AAGCCACCAUGUAUAdTdT-3′. The RNAi oligonucleotide sequences for Nedd4-2 were: RNAi #1, 5′-AACCACAACAC-AAAGUCAdTdT-3′ and RNAi #2, 5′-AAGUGGACAAUUUA- GGCCGAAdTdT-3′. The RNAi oligonucleotide sequences for Itch were: RNAi #1, 5′-CCAGUUGGACUCAAGGAU- UUAdTdT-3′ and RNAi #2, 5′-GGUGACAAAGAGCCA- ACAGAGdTdT-3′. The sequence of negative control was: Control RNAi, 5′-ACAGACUUCGGAGUACCUGdTdT-3′.
For Western blot analysis, the following antibodies were used: anti-AMOT (sc-82491, Santa Cruz Biotechnology), anti-Nedd4 (sc-25508; Santa Cruz Biotechnology), anti-Nedd4-2 (4013S, Cell Signaling Technology), anti-Itch (sc-28367, Santa Cruz Biotechnology), anti-Myc (9E10, Sigma), anti-FLAG (M2, Sigma), anti-HA (MM5-101R, Convance), anti-GFP (green fluorescent protein) (sc-8334, Santa Cruz Biotechnology) and anti-actin (AC-74; Sigma).
Cells were lysed with cell lysis buffer [20 mM Tris/HCl (pH 7.5), 150 mM NaCl, 1% Triton X-100, 1 mM EGTA, 1 mM Na2EDTA, 2.5 mM sodium pyrophosphate, 1 mM 2-glycerophosphate, 1 mM Na3VO4 and 1 μg/ml leupeptin) and the lysate was centrifuged (12000 g for 20 min). The supernatant was pre-cleared with Protein A/G beads (Sigma), followed by incubation with 2 μl of antibody for 2 h and thereafter with Protein A/G beads for 2 h, all at 4°C. Pellets were washed four times with lysis buffer, resuspended in sample buffer and analysed by SDS/PAGE (10% gel).
Western blot analysis
Cell lysates and immunoprecipitates were subjected to SDS/PAGE (10% gel) and proteins were transferred on to nitrocellulose membranes (GE Healthcare). The membrane was blocked in PBS containing 5% non-fat milk and 0.1% Tween-20, washed twice in PBS and incubated with primary antibody at room temperature (24°C) for 2 h, followed by incubation with secondary antibody at room temperature for 45 min. Afterward, the proteins of interest were visualized using ECL (enhanced chemiluminescence) (Santa Cruz Biotechnology).
GST (glutathione transferase) pull-down assay
HEK-293T cells were lysed 36 h after transfection with cell lysis buffer for 30 min at 4°C. GST fusion proteins were immobilized on glutathione–Sepharose beads (Amersham Biosciences). After washing with pull-down buffer [20 mM Tris/HCl (pH 7.5), 150 mM NaCl, 0.1% Nonidet P40, 1 mM DTT (dithiothreitol), 10% glycerol, 1mM EDTA, 2.5 mM MgCl2 and 1 μg/ml leupeptin], the beads were incubated with lysates of transfected H1299 cells for 4 h. The beads were then washed four times with binding buffer and resuspended in sample buffer. The bound proteins were subjected to SDS/PAGE (10% gel).
H1299 cells cultured on coverslips were fixed in 4% paraformaldehyde for 10 min and permeabilized in 0.2% Triton X-100 for 5 min at room temperature. The coverslips were blocked with 5% normal goat serum plus 2% BSA for 1 h and then incubated with mouse anti-HA and anti-ZO-1 (zonula occludens 1) antibodies (1:300 dilution) for 1 h at room temperature, which was followed by sequential incubation with a Texas Red-conjugated goat anti-mouse secondary antibody at 1:300 dilution, and with DAPI (4′,6-diamidino-2-phenylindole; 1:500 dilution) for 10 min. Epifluorescence images were captured using Olympus Inverted System microscope.
qRT–PCR (quantitative reverse transcription–PCR)
Total RNA was isolated from HEK-293T cells using the TRIzol® reagent (Tiangen), and cDNA was reversed-transcribed using the Superscript® Reverse Transcription kit (TOYOBO), according to the manufacturer's instructions. PCR primer sequences for human AMOT/p130 were selected as follows: forward, 5′-CTTGAGCGAGCAGAGAGTC-3′ and reverse, 5′-CAGAGGTCAGAGAAGAAGTAGG-3′. The amplified sequence was located to the N-terminus of AMOT/p130 and it was designed to detect the mRNA level of AMOT/p130, but not the AMOT/p80 isoform. Human Nedd4 primer sequences were: forward, 5′-CCTCCTCCTCCTCCACAG-3′ and reverse, 5′-CGGCTATAACTCTTACTCTCAC-3′. Human Nedd4-2 primer sequences were: forward, 5′-GTGATGTGGATGTGAAT-GACTG-3′ and reverse, 5′-GCTGAGGACCATTGGAACC-3′. Human Itch primer sequences were: forward, 5′-CAAGACC-TTCACGACCACCAC-3′ and reverse, 5′-TCCAGATGTTGC-TCCTTCAGATG-3′. PCR amplification was performed using the SYBR Green PCR master mix kit (TOYOBO). All quantifications were normalized to the level of endogenous control GAPDH (glyceraldehyde-3-phosphate dehydrogenase).
The results of the present study are expressed as means±S.D. for three independent experiments. Statistical analysis was performed using one-way ANOVA with a Newman–Keuls post-hoc test. P<0.05 was considered statistically significant.
Identification of AMOT/p130 as potential novel substrates of the Nedd4-like ubiquitin E3 ligases family
To identify additional substrates of Nedd4-like ubiquitin E3 ligases, we searched the SwissProt database using ExPASy Scansite with the L/P-P-X-Y consensus motif against human proteome sequences. We focused on those proteins that contain multiple L/P-P-X-Y motifs, because these proteins were more likely to be substrates of Nedd4-like ubiquitin E3 ligases. The long isoform of AMOT (AMOT/p130) was identified in this screen.
AMOT was originally identified as a binding protein of angiostatin, and was subsequently found to regulate endothelial cell migration through interaction with angiostatin. AMOT has two main isoforms, AMOT/p80 and AMOT/p130 . We found three L/P-PXY motifs (106 aa~109 aa, LPTY); (239 aa~242 aa, PPEY) and (284 aa~287 aa, PPEY) located in the N-terminus of AMOT/p130. These motifs were not present in the short isoform AMOT/p80 (Figure 1A). Notably, the three L/P-P-X-Y motifs in AMOT/p130 were highly conserved across a wide range of metazoans (Figure 1B). Moreover, the three L/P-P-X-Y motifs were also present in AMOTL1 and AMOTL2, except for the third L/P-P-X-Y motif that was not conserved in AMOTL2 (Figures 1A and 1B).
Identification of AMOT/p130 as potential novel substrates of the Nedd4-like ubiquitin E3 ligases family
To determine whether Nedd4-mediated degradation is mediated by the proteasome, we treated HEK-293T cells with the proteasome inhibitor MG132 for 0, 0.5, 1 and 2 h and determined the protein levels of endogenous AMOT by Western blot. The commercial antibody used in the present study recognized both isoforms of AMOT. As shown in Figure 1(C, left-hand panel), MG132 treatment led to a rapid increase in AMOT/p130 as well as AMOT/p80 protein levels in HEK-293T cells. Since MG132 also inhibits non-proteasomal enzymes, we repeated the experiment using a second mechanistically different proteasome inhibitor, lactacystin, whose only known target is the proteasome . As shown in Figure 1(C, right-hand panel), lactacystin treatment also led to an increase in AMOT/p130 and AMOT/p80 protein levels. We excluded the possibility that elevated protein level might have resulted from up-regulation of transcription by performing qRT–PCR to measure the mRNA level of AMOT/p130 in HEK-293T cells upon proteasome inhibitor treatment (results not shown). Furthermore we measured AMOT/p130 turnover in HEK-293T cells treated with CHX (cycloheximide) to block protein synthesis, as shown in Figure 1(D), the half-life of AMOT/p130 was about 2 h and the protein turnover can be blocked by the MG132. Together, these results suggested that AMOT/p130 stability is regulated by the proteasomal pathway.
We determined if one or more member of Nedd4-like ubiquitin E3 ligases might mediate its degradation. HA–AMOT/p130 or AMOT/p80 was co-expressed with a panel of Nedd4 family of E3 ubiquitin ligases, including Itch, Smurf1, Smurf2, WWP1, WWP2, Nedd4 and Nedd4-2, in H1299 cells. As shown in Figure 1(E), of all E3 ubiquitin ligases tested, Nedd4, Nedd4-2 and Itch efficiently promoted AMOT/p130 degradation in a dose-dependent manner. However, none of these tested E3 ligases has any effect on the level of AMOT/p80 protein (Figure 1F). These results suggested the N-terminal part, which harbours three L/P-P-X-Y motifs, was essential for Nedd4-, Nedd4-2- and Itch-mediated degradation. Since two paralogues of AMOT, AMOTL1 and AMOTL2, have a similar L/P-P-X-Y motif arrangement, we wanted to determine if these two proteins are also subjected to Nedd4-like ubiquitin E3 ligases-mediated degradation. HA–AMOTL1 or AMOTL2 was co-expressed with Nedd4. We found Nedd4 efficiently promoted AMOTL1 protein degradation, but Nedd4, Nedd4-2 or Itch had no effect on the level of AMOTL2 protein (Figure 1G). AMOTL2 has a tyrosine-to-phenylalanine substitution in the third L/P-P-X-Y motif compared with AMOT/p130 and AMOTL1, which was predicted to lose WW domain-binding capacity. One possibility was that the third L/P-P-X-Y motif might be important for Nedd4-mediated degradation. Lastly, we found that AMOT/p130 degradation induced by Nedd4 co-expression could be significantly rescued by the proteasome inhibitor MG132 or lactacystin (Figure 1H). The above results suggested that AMOT/p130 and probably AMOTL1 were novel proteasome-dependent degradation substrates of Nedd4-like ubiquitin E3 ligases.
Nedd4-like E3 ubiquitin ligases control AMOT/p130 stability in vivo
To verify the roles of Nedd4, Nedd4-2 and Itch in AMOT/p130 degradation, we tested E3 ligase mutants in which the highly conserved cysteine residue in its HECT domain was mutated to alanine. This cysteine-to-alanine mutation has been shown to abrogate ubiquitin ligase activities. As shown in Figure 2(A), WT, but not the CA mutant, promoted AMOT/p130 degradation in a dose-dependent manner, indicating that the HECT domain and its ubiquitin ligase activities of Nedd4, Nedd4-2 and Itch are required for promoting AMOT/p130 degradation.
Nedd4-like E3 ubiquitin ligases control AMOT/p130 stability in vivo
Next, we determined the effects of Nedd4 and Nedd4 CA on AMOT/p130 protein turnover. H1299 cells were co-transfected with the HA–AMOT/p130 expression construct, empty vector and either Nedd4 or the Nedd4 CA mutant. After 24 h, the cells were treated with CHX. As shown in Figures 2(B) and 2(C), co-expression of WT Nedd4 and AMOT/p130 resulted in a decrease in AMOT/p130 protein level in a time-dependent manner. In contrast, AMOT/p130 was stabilized by the Nedd4 CA mutant, probably due to a dominant-negative effect of the CA mutant on the WT enzyme.
Given that Nedd4 is an ubiquitin ligase, and induced AMOT/p130 degradation upon co-expression, it is highly likely that AMOT/p130 is a ubiquitination substrate of Nedd4. To assess this possibility, we co-expressed FLAG–AMOT/p130, HA–ubiquitin and either Nedd4 or Nedd4 CA mutant in H1299 cells. The polyubiquitinated forms of AMOT/p130 were immunoprecipitated and then detected by Western blotting. As shown in Figure 2(D) (labelled Short exposure), AMOT/p130 was seen as a strong smear of bands when it was co-expressed with WT Nedd4. The polyubiquitinated forms of AMOT/p130 were also observed when co-expressed with empty vector or Nedd4-2 CA mutant in the long exposure, which might be mediated by endogenous ubiquitin E3 ligases. Similar results were obtained when AMOT/p130 was co-expressed with Nedd4-2 and Itch (results not shown).
Having demonstrated that WT Nedd4 and its paralogues were capable of inducing polyubiquitination and proteasomal degradation of transiently expressed AMOT/p130, we next determined their effects on the protein level of endogenous AMOT/p130. As shown in Figure 2(E), overexpression of the three WT Nedd4-like ubiquitin E3 ligases family members led to a significant reduction in endogenous AMOT/p130. In comparison, the corresponding CA mutants had no effect on the levels of endogenous AMOT/p130. In a complementary experiment, we knocked down the endogenous Nedd4, Nedd4-2 and Itch respectively, each by two specific siRNAs and determined the changes in the level of AMOT/p130 protein in HEK-293T cells. Knockdown of Nedd4 or Nedd4-2 in HEK-293T cells resulted in a marked increase in the level of endogenous AMOT/p130, but not AMOT/p80 (Figure 2F), whereas knockdown of Itch resulted in a smaller increase in the level of endogenous AMOT/p130. To exclude the possibility that AMOT/p130 protein elevation resulted from transcriptional up-regulation, we performed qRT–PCR to measure the mRNA levels of AMOT/p130 and ubiquitin E3 ligases in siRNA knockdown HEK-293T cells. In contrast with the significant decrease in mRNA transcripts of ubiquitin E3 ligases, the mRNA level of AMOT/p130 in ubiquitin E3 ligases-depleted HEK-293T cells stayed at a level similar to that of the control cells (Figure 2G). Furthermore, concomitant knockdown of three ubiquitin E3 ligases caused the most significant elevation of AMOT/p130 protein level in comparison with the knockdown of any individual ubiquitin E3 ligase (Figure 2H). It suggested that Nedd4, Nedd4-2 and Itch are functionally redundant in the regulation of AMOT/p130 protein stability. Taken together, these results further support the notion that Nedd4, Nedd4-2 and Itch collectively regulate AMOT/p130 ubiquitin-dependent proteome degradation in vivo.
AMOT/p130 forms a complex with Nedd4-like E3 ubiquitin ligases
Given that Nedd4-like ubiquitin E3 ligases contain multiple WW domains that are known to mediate its binding to substrates through the L/P-PXY motif, which is also present in AMOT/p130, we examined the interaction between these two proteins. We co-expressed HA–AMOT/p130 and the Nedd4 CA mutant (which was used to avoid degradation of the protein bound to Nedd4) in H1299 cells and immunoprecipitated Myc-tagged Nedd4. As shown in Figure 3(A), Nedd4 co-immunoprecipitated with AMOT/p130. Similar results were obtained in a reciprocal co-immunoprecipitation experiment using anti-HA antibody. In contrast, we did not detect any interaction between Nedd4 and the AMOT/p80 (Figure 3A), suggesting that the N-terminal part of AMOT/p130 was essential for Nedd4 binding. Similar results were obtained when AMOT/p130 was co-expressed with Nedd4-2 and Itch (results not shown). Next, we want to confirm if AMOT/p130 interacts with Nedd4, Nedd4-2 and Itch endogenously. When endogenous AMOT in HEK-293T cells was immunoprecipitated with an anti-AMOT antibody, endogenous Nedd4, Nedd4-2 and Itch were detectable in the immunoprecipitate by Western blotting (Figure 3B). Furthermore, a reciprocal immunoprecipitation experiment was performed in which endogenous Nedd4, Nedd4-2 or Itch were immunoprecipitated. As shown in Figure 3(C), AMOT/p130, but not AMOT/p80, was detectable in the immunoprecipitate by Western blotting. Although, a weak AMOT/p80 signal was shown in the immunoprecipitate using an anti-Itch antibody, the possible reason being because AMOT/p130 and AMOT/p80 could hetero-oligomerize , the AMOT/p130-bound AMOT/p80 was immunoprecipitated by Itch through indirect interaction. Furthermore, we did not detect Nedd4-2 and Itch in Nedd4 immunoprecipitate. Similar results were also observed in Nedd-2 and Itch immunoprecipitate. These results suggested that the three ubiquitn E3 ligases exist as distinct AMOT–HECT complexes and do not to form a large hetero-oligomer to carry out AMOT/p130 ubiquitination (Figure 3C). These results indicated AMOT/p130 strongly interacted with Nedd4, Nedd4-2 and Itch in vivo.
AMOT/p130 forms a complex with Nedd4, Nedd4-2 and Itch
To verify the interaction between AMOT/p130 and Nedd4, we investigated whether these two proteins are localized to the same subcellular compartments. HA–AMOT/p130 and Myc–Nedd4 CA expression constructs were transfected into H1299 cells and their subcellular localizations monitored by immunofluorescence. AMOT/p130 was localized to the cell membrane and cytoplasm as reported previously  (Figure 4), Nedd4 was localized to the cytoplasm, and when AMOT/p130 and Nedd4 were co-expressed both of these proteins were co-localized in the cell membrane and cytoplasm (Figure 4).
Co-localization of AMOT/p130 and Nedd4 in H1299 cells
Binding of Nedd4 to AMOT/p130 is mediated through the WW domain of Nedd4 and the L/P-PXY motifs of AMOT/p130
AMOT/p130 contains three L/P-PXY motifs at the N-terminal part. Nedd4-like ubiquitin E3 ligases contains multiple WW domains in its central region (Figures 5A and 5B). To determine whether the AMOT/p130–Nedd4 interaction is mediated through these domains, we generated four truncation mutants of AMOT/p130 (D1–D4). We also generated a series of N-terminal part (D1) mutants in which one (ΔPY1, ΔPY1, ΔPY2 andΔPY3), two (ΔPY1/2, 1/3 and 2/3) or three (ΔPY1/2/3) L/P-PXY motifs were deleted. We also constructed a three-amino-acid substitution mutant (3PYF) by substituting the terminal tyrosine residue for phenylalanine. These mutations have been shown to abrogate binding of the L/P-PXY motif to the WW domain.
Nedd4 and AMOT/p130 interacts through WW domains and L/P-PXY motifs
A total of twelve mutants were fused to GST, expressed and purified from bacteria. We used a pull-down assay with GST fusion proteins of AMOT/p130 and Nedd4 overexpressed in H1299 cells. As expected, the N-terminal part (D1) mutant, but not the central or C-terminal part mutants (D2, D3, D4), specifically interacted with Nedd4 (Figure 5C). For the L/P-PXY motif deletion mutants of D1, we found that only a single L/P-PXY motif deletion causes no obvious binding affinity change compared with the D1 segment, whereas deletion of two L/P-PXY motifs caused a significant decrease in the binding affinity. Deletion of all three of the L/P-PXY motifs totally abolished the interaction between D1 and Nedd4 (Figure 5C). Additionally, the 3PYF mutation also causes a significant decrease in the binding affinity with Nedd4 (Figure 5C). Thus these results suggested that L/P-PXY motifs of AMOT/p130 were essential for Nedd4 binding, and each L/P-PXY motif partially contributes to the binding capacity.
Next, we generated three deletion mutants of Nedd4 and determined the minimal region that is sufficient to mediate its interaction with AMOT/P130. Thus the WT and three of the Nedd4-deletion mutants, containing the C2 domain (C2), the WW domain (WW) and the HECT domain (HECT) respectively, were fused to GST and expressed in and purified from bacteria. A GST pull-down assay for Nedd4 and HA–AMOT/p130 expressed in H1299 cells was carried out. As shown in Figure 5(D), AMOT/p130 interacted with the WW mutant harbouring four WW domains. Thus Nedd4 is capable of interacting with AMOT/p130 through multiple WW domains.
We also determined if the L/P-PXY motifs were also essential for Nedd4-mediated degradation. As shown in Figure 5(E), Nedd4 targeted WT AMOT/p130 for degradation efficiently, and deletion of either one or two L/P-PXY motifs had a negligible effect on Nedd4-mediated degradation. However, deletion of three L/P-PXY motifs provides total resistance against Nedd4-mediated degradation (Figure 5E). These results suggested that any one L/P-PXY motif was sufficient for Nedd4–AMOT/p130 interaction and subsequent degradation.
Recently, a new function of the AMOT protein family as regulators of the Hippo pathway was reported by several research groups [13–16]. AMOT/p130, AMTL1 and AMOTL2 all directly interact, and sequester the transcriptional co-activator YAP/TAZ into the cytoplasm or target them the TJs, thus inhibiting YAP/TAZ-induced transformation and causing loss of cell contact inhibition. AMOT/p130 or AMOTL1 was reported to bind to YAP/TAZ mainly via the first PPEY motif, and the second PPEY play a minor role in the binding. However, our results revealed another PPXY variant, LPTY, located in the N-terminal part of AMOT/p130 that may play a role in mediating AMOT/p130–Nedd4 interaction. This motif was well-conserved in AMOTL1 and AMOTL2. In recent reports, mutation of both PPEY motifs significantly decreased, but did not completely abolish, the interaction between AMOT/p130 and YAP/TAZ [13–16]. We hypothesized that the LPTY motif also participated in the interaction between AMOT/p130 and YAP. Unexpectedly, GST pull-down assay showed that deletion of a single LPTY motif abolished the interaction with YAP, whereas deletion of both PPEY motifs still retained some ability to bind to YAP (Figure 5F). This result suggested that the previously unidentified LPTY motif of AMOT/p130 (and probably the corresponding motif in AMOTL1 and AMOTL2) was critical for YAP interaction. The WW domain of YAP binds to AMOT/p130 mainly via the first and second L/P-PXY motifs, which is identical to the binding pattern of the WW domain of Nedd4, raising the possibility that YAP competes with Nedd4 for the binding to AMOT/p130. To test this possibility, we immunoprecipitated HA–AMOT/p130 and Nedd4 in the presence of YAP. In the absence of YAP, Nedd4 was co-immunoprecipitated with AMOT/p130. However, the level of AMOT/p130-associated Nedd4 was significantly decreased in the presence of YAP (Figure 5G). Furthermore, when endogenous AMOT/p130 was immunoprecipitated, endogenous Nedd4 and YAP were both detected in the AMOT immunoprecipitate. However, when Nedd4 was knocked down, more YAP1 protein was present in the immunoprecipitated AMOT than the control siRNA treatment, suggesting that ablation of Nedd4 may enhance the AMOT/p130–YAP interaction (Figure 5H). Taken together, these results suggested that that the binding of YAP and Nedd4 to AMOT/p130 is mutually exclusive and accumulation of YAP displaces the binding of Nedd4 and vice versa.
In the present study, we unravelled a new type of post-translational modification of AMOT/p130 by the Nedd4-like ubiquitin E3 ligases (Nedd4, Nedd4-2 and Itch). Among the multiple lines of experimental evidence are: (i) endogenous Nedd4-like ubiquitin E3 ligases interacted with AMOT/p130, and the interaction was mediated through the WW domain of Nedd4 and the L/P-PXY motifs of AMOT/p130; (ii) Nedd4-like ubiquitin E3 ligases can promote AMOT/p130 ubiquitin-dependent proteasomal degradation; and (iii) Nedd4 competed with YAP for the binding to AMOT/p130. The results of the present study suggest that the L/P-PXY motifs of AMOT/p130 were not only important for Hippo pathway suppression, but also for the control of the fate of its own protein.
We note that although the AMOT/p80 isoform lacks the L/P-PXY motifs, thus is not regulated by Nedd4-like E3 ubiquitin ligases as AMOT/p130, it is also stabilized by proteasome inhibitors in HEK-293T cells (Figure 1C and 1D). Thus AMOT/p80 might also be subjected to proteasome-dependent degradation controlled by other unidentified E3 ubiquitin ligases. It was reported that AMOT/p130 and AMOT/p80 are differently expressed and play distinct roles during angiogenesis. AMOT/p80 induces endothelial cell migration . In contrast, AMOT/p130 appears to play roles in the later phase of angiogenesis when stabilization and maturation are important . Thus the functional difference, the protein stabilities of AMOT/80 and AMOT/p130 might be regulated by different kinds E3 ubiquitin ligases and upstream signals.
In several recent reports, the two PPEY motifs of AMOT/p130 or AMOTL1 were suggested to be responsible for binding to the WW domain of YAP/TAZ [13–16]. We found that the LPTY motif was also critical for binding to the WW domain of YAP. Most of the Nedd4-like ubiquitin E3 ligases substrates contained one or more PPXY motifs; however, LPXY motifs were also present in a small subset of known substrates. Since deletion of this motif in AMOT/p130 nearly abolished the interaction with YAP, whereas deletion of two PPEY motifs in AMOT/p130 had a much less significant effect in its ability to binding to YAP, it is likely that all three L/P-PXY motifs might play partial roles in YAP binding. In a similar way, the LPTY motif was also involved the interaction between Nedd4 and AMOT/p130. However, deletion of either one or two L/P-PXY motifs had a negligible effect on Nedd4-mediated degradation. AMOT/p130 might interact with other WW domain-containing proteins through L/P-PXY motifs. It is possible that if these three sites are all occupied by other WW domain-containing proteins, AMOT/p130 may be stabilized. However, upon certain signal stimuli dissociation of these proteins from AMOT/p130 to expose only one L/P-PXY motif might be sufficient for recruitment of Nedd4-like E3 ubiquitin ligases and subsequent degradation.
In the present study, we found that Nedd4 also promoted AMOTL1 degradation. It is not surprising, since AMOTL1 has three L/P-PXY motifs, whose sequences were identical to AMOT/p130. Nedd4-like E3 ubiquitin ligases might regulate AMOTL1 stability similar to AMOT/p130; however, AMOTL2 is not degraded by Nedd4. Since the third L/P-PXY motif was not conserved in AMOTL2, initially we speculated that the third L/P-PXY motif of the AMOT protein family might be essential for Nedd4 binding and degradation, but the results of the present study showed that deletion of the corresponding motif in AMOT/p130 had only a minor effect on Nedd4 interaction and degradation. These results suggested that, other than the three L/P-PXY motifs, other sequence constraints present in AMOT/p130 and AMOTL1, but not AMOTL2, might exist, which might also be important for Nedd4-like E3 ubiquitin ligases-mediated degradation. Furthermore, whether Nedd4-like ubiquitin E3 ligases have non-degradable regulation of AMOTL2 was not explored. These interesting issues need to be addressed in the future.
That AMOT/p130 can be regulated by three Nedd4-like E3 ubiquitin ligases suggested functional redundancy among the members of the family. The relative contribution might be dependent on the cellular context and difference in different cell types or tissues. Knockdown of Itch in HEK-293T cells causes only a small increase in AMOT/p130 protein level, whereas knockdown of Nedd4 or Nedd4-2 led to a significant increase in the level of AMOT/p130 protein. It is possible that the abundance of Itch was relatively low in HEK-293T cells compared with Nedd4 or Nedd4-2, so the knockdown effect might be easily compensated for by Nedd4 and Nedd4-2.
Epithelial and endothelial barrier function is maintained by intercellular TJs, multi-protein complexes that seal the space between adjacent cells. AMOT/p130 is localized to cell–cell junctions in primary endothelial cells, controlling the permeability of cell layers. Cytokines, such as INF-γ (interferon γ), TNF-α (tumour necrosis factor α) and VEGF perturbed the TJ function, resulting in enhanced paracellular permeability and increased exposure of tissues to luminal antigens in organ systems such as the gastrointestinal and respiratory tracts . The molecular mechanism that regulates these processes is less well known. Nedd4-like ubiquitin ligases were found to be involved in these processes by promoting ubiquitination and endocytosis of TJ proteins. For example, VEGF treatment induced TJ fragmentation and occludin trafficking from the cell border to the endosome, concomitant with increased occludin phosphorylation on Ser490 and ubiquitination mediated by Itch . Thus whether Nedd4-like ubiquitin E3 ligases-mediated AMOT/p130 ubiquitination also plays roles in cytokine-mediated TJ destabilization and enhanced paracellular permeability remain to be investigated.
cell division cycle 42
fetal bovine serum
green fluorescent protein
homologous with E6-associated protein C-terminus
human embryonic kidney
neural-precursor-cell-expressed developmentally down-regulated
Nedd4-like E3 ubiquitin-protein ligase
protein associated with Lin Seven 1
Pals1-associated tight junction protein
quantitative reverse transcription PCR
RhoGAP interacting with Cdc42-interacting protein 4 homologues protein 1
small interfering RNA
SMAD-specific E3 ubiquitin protein ligase
transcriptional co-activator with PDZ-binding motif
vascular endothelial growth factor
WW domain-containing E3 ubiquitin protein ligase
zonula occludens 1
Chenji Wang and Jian An performed most of the experiments; Pingzhao Zhang and Kun Gao performed the immunofluorescence analysis; Chen Xu, Di Wu and Dejie Wang performed the expression of the recombinant proteins. Hongxiu Yu, Jun Liu and Long Yu wrote the paper. Long Yu supervised the work.
We thank Dr Lars Holmgren (Cancer Centre Karolinska Institute, Stockholm, Sweden) for the AMOT/p130 and AMOT/p80 constructs, Dr Makoto Adachi (Kyoto University, Kyoto, Japan) for the HA–AMOTL1 and AMOTL2 constructs, Professor Gerry Melino (University of Leicester, Leicester, U.K.) for the Myc–Itch and Myc–Itch/C830A constructs, Professor Kohei Miyazono (University of Tokyo, Tokyo, Japan) for the Myc–Smurf1/2 and Myc–WWP1 constructs, Dr Xuejun Jiang (Memorial Sloan-Kettering Cancer Center, New York, NY, U.S.A.) for the Nedd4 construct, Dr Ying Jin (Shanghai Jiao Tong University School of Medicine, Shanghai, China) for the Myc–WWP2 construct, and Professor Christie P. Thomas (University of Iowa, Iowa City, IA, U.S.A) for the Nedd4-2 construct.
This work was supported by the National Key Sci-Tech Special Project of China [grant number 2008ZX10002-020] and the National Natural Science Foundation of China [grant numbers 30872947, 31071193 and 81171964].
These authors contributed equally to this work.