YAP (Yes-associated protein) oncogene has been found to form a stable complex with members of the Angiomotin (Amot) family of proteins, which bind WW domains of YAP and sequester the protein in the cytoplasm and junctional complexes. The Amot-mediated retention of YAP in the cytoplasm results in the inhibition of its proliferative function. Using apoptotic ‘read-out’ of YAP in HEK293 cells, we confirmed the molecular mode by which Amot regulates YAP. We showed that a representative member of the Amot family, AmotL1 (Angiomotin-like-1), uses its PPxY motifs to bind WW domains of YAP and inhibit YAP's nuclear translocation and pro-apoptotic function. Recently we also showed that YAP uses its PDZ-binding motif to interact with zona occludens-2 (ZO-2) protein, which promotes YAP's translocation to the nucleus. We also asked if AmotL1, YAP and ZO-2 signal together. We report here that AmotL1 and ZO-2 form a tripartite complex with YAP and regulate its function in HEK293 cells in opposite directions. AmotL1 inhibits pro-apoptotic function of YAP, whereas ZO-2 enhances it. As YAP is a potent oncogene, the identification and characterization of its regulators is important. AmotL1 and ZO-2 are two candidates that could be harnessed to control the oncogenic function of YAP.
Currently, YAP (Yes-associated protein) and its close paralog TAZ (transcriptional co-activator with PDZ domain) are intensely studied genes because they act as the main effectors of the newly described tumor suppressor pathway known as Hippo (Zhao et al., 2011). YAP gene is frequently overexpressed or amplified in several human cancers, and it is considered a potent oncogene (Pan, 2010). However, there are several cancers where YAP has a role of tumor suppressor (Yuan et al., 2008; Ehsanian et al., 2010). To better understand the molecular function of YAP gene, we elected to identify those YAP-binding proteins that assemble on YAP via its specialized regions, which evolved to mediate protein-to-protein binding, namely the WW domains and PDZ-binding motif. Several YAP-interacting proteins that were identified in protein-binding assays, turned out to be bona fide regulators of YAP signaling (reviewed in Sudol, 2010; Sudol and Harvey, 2010). Most recently, we showed that zona occludens-2 (ZO-2) protein binds YAP via a PDZ domain complex and regulates the nuclear localization and apoptotic activity of YAP (Oka and Sudol, 2009; Oka et al., 2010). In most cell culture and animal models used to study YAP, its overexpression results in the accumulation of YAP in the nucleus and this event correlates with growth-promoting and pro-oncogenic activity (Overholtzer et al., 2006). However, under conditions of stress such as UV light or low serum media, the nuclear YAP associates with a pro-apoptotic factor, p73, and causes cell death (Strano et al., 2001; Basu et al., 2003; and Matallanas et al., 2007). In our study, we employ HEK293 cells that are maintained in low serum (1%) media, in which overexpression of YAP causes apoptosis (Oka et al., 2008). The HEK293 cells provide a sensitive assay of YAP activity, which allows one to test a large number of candidates as potential regulators of YAP function. The easy detection of endogenous PARP as a marker of apoptosis, the detachment of cells (measured by cell counting) and a high efficiency of DNA transfection of HEK293 cells are behind the power of this model (Oka et al., 2008).
Angiomotin (Amot) was originally identified as a binding partner of angiostatin, and was found to regulate endothelial cell migration through angiostatin binding (Troyanovsky et al., 2001; Bratt et al., 2002). In addition, Amot was shown to be a part of the actin cytoskeleton and cell-to-cell junction complexes. Amot is a member of a family of proteins composed of Amot and two paralogs, angiomotin-like-1 (AmotL1) and angiomotin-like-2 (AmotL2) (Nishimura et al., 2002; Bratt et al., 2005).
Amot was first detected as a partner of YAP in a large screen of multi-protein complexes that assemble on de-ubiquitinating enzymes (Sowa et al., 2009). The complexes were detected by an affinity capture-mass spectrometry method. Because in that article no functional studies were reported for YAP and Amot, we decided to examine the role of YAP-Amot complex, especially as Amot contains the PPxY sequence motif that is required for ligands of WW domains.
Recently, we showed that the complex between YAP and ZO-2 could be visualized for the endogenous proteins by immunostaining and co-immunoprecipitation (Oka and Sudol, 2009; Oka et al., 2010). The PDZ-binding motif in YAP and the 1st PDZ domain in ZO-2 were required for the formation of the complex and overexpression of ZO-2 affected the subcellular localization of YAP. Moreover ZO-2 cooperated with YAP to enhance detachment of HEK293 cells, but in MDCK cells the induction of ZO-2 inhibited YAP-induced cell proliferation (Oka et al., 2010).
Results and discussion
Several recent reports suggest that members of Amot family interact with YAP (and its paralog, TAZ) via WW domain and PPxY-mediated complexes (Wang et al., 2011; Zhao et al., 2011; Chan et al., 2011). The WW domains of YAP and TAZ, and PPxY motifs of Amots were shown to be critical for the binding. Using different approaches, these three studies documented that the Amot/YAP(TAZ) complexes have inhibitory function on YAP's (and TAZ) ability to promote growth and oncogenic transformation. The main mechanism of this regulation is the sequestration of YAP and TAZ proteins by Amots in the cytoplasm and in cell junction complexes, thereby preventing YAP and TAZ from functioning as transcriptional co-activators in the nucleus. We independently confirmed that the YAP-Amot complexes are mediated between YAP WW domains and Amots PPxY motifs (Supplementary Figure 1). More specifically, we showed that Amots bind to YAP2 isoform (that contains two WW domains) but not to YAP1 isoform (that contains a single WW domain). The binding was independent of the carboxy-terminal PDZ-binding motif of motins and the first PPxY motif of AmotL1 was important for the binding (Supplementary Figure 1). Most importantly, we documented that the complex between endogenous YAP and endogenous Amot and AmotL1 could be observed in HEK293 cells (for unknown reasons, we could not detect the endogenous complex between YAP and AmotL2). It did not escape our attention that HEK293 cells are known to express a relatively high level of Amots, compared with other commonly used cells lines, and therefore were helpful in our study (Chan et al., 2011). Because the YAP1 isoform did not interact with Amots, all the remaining experiments were performed with YAP2 isoform only, which we refer to simply as YAP.
We then asked if ZO-2, YAP and Amot form either a tripartite or binary and mutually exclusive complexes. We also tested if ZO-2 could enhance or oppose the inhibitory function of Amot on YAP, using HEK293 cell readout. As Amots contain PDZ-binding motifs and ZO proteins contain several PDZ domains (Figure 1a), we first tested if ZO-2 protein could interact with AmotL1 via PDZ complexes. Surprisingly, ZO-2 protein could be co-precipitated with AmotL1 even if the PDZ-binding motif of AmotL1 was deleted (Figure 1b). Therefore, we speculated that YAP could mediate the complex between ZO-2 and AmotL1. Indeed, when the level of YAP was increased by overexpression, a high level of ZO-2 could be co-precipitated with Amot and AmotL1 (Figure 1c). To analyze this binding in detail, we employed various mutants of AmotL1 and YAP (Figures 1d and 2a). AmotL1 bound to YAP (WW domains) mainly via its 1st PPEY motif, although the 2nd PPEY motif played a minor role in the binding (Figure 1d, top panel, lanes 2–5). In the presence of YAP, ZO-2 was co-precipitated with both AmotL1 WT and AmotL1N construct that encodes the amino terminal 70% of AmotL1 (Figure 1d, middle panel, lanes 2 and 6). The amount of co-precipitated ZO-2 decreased when either of the two PPEY motifs in AmotL1 was mutated, and the co-precipitated ZO-2 was barely detectable when both PPEY motifs were mutated (Figure 1d, middle panel, lane 3, 4 and 5). ZO-2 was not co-precipitated with AmotL1C construct that contains carboxyterminal 30% of AmotL1 (Figure 1d, middle panel, lane 7). These results suggest that YAP mediates an association between Amot L1 and ZO-2. A binding assay between ZO-2 and YAP mutants further supported our assumption (Figure 2a). The WT-YAP was able to bind AmotL1 (Figure 2a, top panel, lanes 3 and 5). In the presence of WT-YAP, ZO-2 was co-precipitated with AmotL1 WT (Figure 2a, middle panel, lane 3). However, when YAP delta C mutant (Oka and Sudol, 2009) was used instead of YAP WT, ZO-2 could no longer be co-precipitated with AmotL1 WT (Figure 2a, middle panel, lane 5). The simplest explanation is that YAP delta C binds to AmotL1, but it fails to bind to ZO-2 because it lacks the C-terminal PDZ domain-binding motif and therefore can no longer act as a scaffold protein. This result suggests that YAP-mediated association between ZO-2 and AmotL1 requires both the WW domains and the C-terminal PDZ domain-binding motif of YAP2. To confirm our hypothesis further, we reduced the level of the endogenous YAP by RNA interference (RNAi) and then tested if ZO-2 and AmotL1 form a complex. The endogenous YAP in HEK293 cells mediates AmotL1 and ZO-2 association, even though the detected signal is relatively modest (Figure 2b, middle panel, lane 2). However, when the endogenous YAP was removed, ZO-2 failed to form a complex with AmotL1 (Figure 2b, middle panel, lane 4). In sum, these results suggest that YAP mediates the association between AmotL1 and ZO-2. The two PPEY motifs in AmotL1, two WW domains and one PDZ-binding motif in YAP, plus the 1st PDZ domain in ZO-2 are necessary for the formation of the tri-partite complex.
We previously reported that YAP WT promoted apoptosis in HEK293 cells grown in low serum containing media (Oka et al., 2008). We used this read-out to determine if AmotL1 has any influence on the pro-apoptotic function of YAP. AmotL1 was transfected into HEK293 cells before the induction of YAP expression. After incubation for 96 h in Dulbecco's-modified Eagle's medium-containing 1% serum, the attached cells were counted. In the absence of AmotL1 transfection, YAP expression reduced the number of cells that remained attached (the ratio of attached cells with induced YAP WT to attached cells without induced YAP WT was 0.478; Figure 3a). However, in the context of AmotL1 WT transfection, the ability of YAP to reduce the number of attached cells was impaired (ratio increased to 0.709; Figure 3a). This change correlated with the diminished signal of PARP cleavage, which is the marker of cell death, suggesting that YAP-induced apoptosis was inhibited. Similar results were obtained when AmotL1 2nd PPEY* or AmotL1 N mutant were overexpressed (Figure 3a). These three AmotL1 recombinant proteins showed relatively strong binding to YAP (Supplementary Figure 1E). However, AmotL1 1st PPEY*, 1 & 2 PPEY*, or C, which showed relatively weak binding to YAP (Supplementary Figure 1E), did not significantly affect the pro-apoptotic function of YAP (Figure 3a). These results suggest that AmotL1-YAP complex formation impairs the pro-apoptotic function of YAP in HEK293 cells. Other members of the Amot family: Amot and AmotL2 showed a similar effect on pro-apoptotic function of YAP (Figure 3b). We recently reported that ZO-2 enhances pro-apoptotic function of YAP in HEK293 cells (Oka et al., 2010), which is opposite to the role of AmotL1. Accordingly, we have shown that ZO-2 and AmotL1 tend to neutralize each other in terms of the pro-apoptotic function of YAP (Figure 3c), suggesting that their relative balance determines the function of YAP.
We also downregulated the level of AmotL1 by RNAi and monitored changes in cell apoptosis induced by YAP. After the expression of YAP WT was induced, either AmotL1-specific RNAi oligonucleotides or control RNAi oligonucleotides were transfected into the cells. The efficiency of RNAi inhibition of AmotL1 expression was confirmed by immunoblotting (Figure 3d). The removal of Amot L1 alone slightly enhanced YAP-induced apoptosis, but the individual removal of Amot or Amot L2 did not (Figure 3d). This result could be explained by the functional redundancy among the members of the family, which share a high degree of sequence similarity. Therefore, we investigated if the removal of all three Amots by RNAi affects the function of YAP. As expected, YAP-induced apoptosis was enhanced considerably when all the members of Amot family were suppressed (Figure 3d). Perhaps a further study of Amots and ZO-2 will help to decipher the duality of YAP function as an oncogene and tumor suppressor.
To elucidate the mechanism by which Amots inhibit apoptosis in our cell culture model, we examined whether Amots affect subcellular localization of YAP, by sequestering YAP in the cytoplasm and preventing its nuclear entry. We have shown previously that the nuclear localization of YAP is required for its pro-apoptotic function in HEK293 cells. green fluorescent protein (GFP)-YAP was co-transfected with Flag-fused constructs of AmotL1 WT, 1 & 2 PPEY* and N mutants into HEK293 cells and the cells were observed by fluorescence microscopy (Figure 4a). The Flag-tagged proteins were visualized by immuno-staining and the nuclei were decorated by 46-diamidino-2-phenyl indole staining. AmotL1-WT, AmotL1-1&2 PPEY* or/and AmotL1-N localized mostly in the cytosol of HEK293 cells. The signal of GFP-YAP in the nucleus was relatively weak when AmotL1-WT or AmotL1-N were co-transfected, compared with the GFP-YAP signal when AmotL1-1 & 2 PPEY* was co-transfected. These results suggest that GFP-YAP was eliminated from the nucleus by AmotL1-WT and/or AmotL1-N, but AmotL1-1 & 2 PPEY* failed to do so because it could not bind YAP. Unlike YAP WT, the GFP-YAP2 1 & 2 WW* remained in the nucleus when Flag-AmotL1 was expressed, because of the lack of binding (Figure 4b). The localization of YAP is preferentially cytoplasmic when AmotL1 is overexpressed.
In conclusion, ZO-2 and AmotL1 form a tripartite complex with YAP and regulate its pro-apoptotic function in HEK293 cells in opposing directions. AmotL1 inhibits pro-apoptotic function of YAP by sequestering YAP protein in the cytoplasm, whereas ZO-2 promotes YAP nuclear localization, its pro-apoptotic and growth inhibitory activity. Although ZO-2 and AmotL1 tend to drive YAP to different subcellular compartments, nucleus and cytoplasm, respectively, their regulatory roles on YAP oncogenic function, do not have to be opposite. For example, we anticipate that in MCF10A cells, where the overexpression of YAP causes transformation (Overholtzer et al., 2006), both Amots and ZO-2 would inhibit this process, Amots by cytoplasmic retention of YAP but ZO-2 by inhibiting cell growth while in complex with the nuclear YAP (Oka et al., 2010). As YAP is a potent oncogene, identification and characterization of its negative regulators is important. AmotL1 and ZO-2 are two candidates that could be harnessed to control oncogenic function of YAP.
Basu S, Totty NF, Irwin MS, Sudol M, Downward J . (2003). Akt phosphorylates the Yes-associated protein, YAP, to induce interaction with 14-3-3 and attenuation of p73-mediated apoptosis. Mol Cell 11: 11–23.
Bratt A, Birot O, Sinha I, Veitonmaki N, Aase K, Ernkvist M et al. (2005). Angiomotin regulates endothelial cell-cell junctions and cell motility. J Biol Chem 280: 34859–34869.
Bratt A, Wilson WJ, Troyanovsky B, Aase K, Kessler R, Van Meir EG et al. (2002). Angiomotin belongs to a novel protein family with conserved coiled-coil and PDZ binding domains. Gene 298: 69–77.
Chan SW, Lim CJ, Chong YF, Pobbati AV, Huang C, Hong W . (2011). Hippo pathway-independent restriction of TAZ and YAP by angiomotin. J Biol Chem 286: 7018–7026.
Ehsanian R, Brown M, Lu H, Yang XP, Pattatheyil A, Yan B et al. (2010). YAP dysregulation by phosphorylation or DeltaNp63-mediated gene repression promotes proliferation, survival and migration in head and neck cancer subsets. Oncogene 29: 6160–6171.
Matallanas D, Romano D, Yee K, Meissl K, Kucerova L, Piazzolla D et al. (2007). RASSF1A elicits apoptosis through an MST2 pathway directing proapoptotic transcription by the p73 tumor suppressor protein. Mol Cell 27: 962–975.
Matsuda T, Cepko CL . (2007). Controlled expression of transgenes introduced by in vivo electroporation. Proc Natl Acad Sci USA 104: 1027–1032.
Nishimura M, Kakizaki M, Ono Y, Morimoto K, Takeuchi M, Inoue Y et al. (2002). JEAP, a novel component of tight junctions in exocrine cells. J Biol Chem 277: 5583–5587.
Oka T, Mazack V, Sudol M . (2008). Mst2 and Lats kinases regulate apoptotic function of Yes kinase-associated protein (YAP). J Biol Chem 283: 27534–27546.
Oka T, Remue E, Meerschaert K, Vanloo B, Boucherie C, Gfeller D et al. (2010). Functional complexes between YAP2 and ZO-2 are PDZ domain-dependent, and regulate YAP2 nuclear localization and signalling. Biochem J 432: 461–472.
Oka T, Sudol M . (2009). Nuclear localization and pro-apoptotic signaling of YAP2 require intact PDZ-binding motif. Genes Cells 14: 607–615.
Overholtzer M, Zhang J, Smolen GA, Muir B, Li W, Sgroi DC et al. (2006). Transforming properties of YAP, a candidate oncogene on the chromosome 11q22 amplicon. Proc Natl Acad Sci U S A 103: 12405–12410.
Pan D . (2010). The hippo signaling pathway in development and cancer. Dev Cell 19: 491–505.
Pei Z, Bai Y, Schmitt AP . (2010). PIV5M protein interaction with host protein angiomotin-like 1. Virology 397: 155–166.
Sowa ME, Bennett EJ, Gygi SP, Harper JW . (2009). Defining the human deubiquitinating enzyme interaction landscape. Cell 138: 389–403.
Strano S, Munarriz E, Rossi M, Castagnoli L, Shaul Y, Sacchi A et al. (2001). Physical interaction with Yes-associated protein enhances p73 transcriptional activity. J Biol Chem 276: 15164–15173.
Sudol M, Harvey KF . (2010). Modularity in the Hippo signaling pathway. Trends Biochem Sci 35: 627–633.
Sudol M . (2010). Newcomers to the WW domain-mediated network of the Hippo tumor suppressor pathway. Genes Cancer 1: 1115–1118.
Sugihara-Mizuno Y, Adachi M, Kobayashi Y, Hamazaki Y, Nishimura M, Imai T et al. (2007). Molecular characterization of angiomotin/JEAP family proteins: interaction with MUPP1/Patj and their endogenous properties. Genes Cells 12: 473–486.
Troyanovsky B, Levchenko T, Mansson G, Matvijenko O, Holmgren L . (2001). Angiomotin: an angiostatin binding protein that regulates endothelial cell migration and tube formation. J Cell Biol 152: 1247–1254.
Wang W, Huang J, Chen J . (2011). Angiomotin-like Proteins Associate with and Negatively Regulate YAP1. J Biol Chem 286: 4364–4370.
Yuan M, Tomlinson V, Lara R, Holliday D, Chelala C, Hanada T et al. (2008). Yes-associated protein (YAP) functions as a tumor suppressor in breast. Cell Death Differ 15: 1752–1759.
Zhao B, Li L, Lu Q, Wang LH, Liu CY, Lei Q et al. (2011). Angiomotin is a novel Hippo pathway component that inhibits YAP oncoprotein. Genes Dev 25: 51–63.
We thank Dr Makoto Adachi from Kyoto University in Japan for a kind gift of HA-tagged mouse Amot family plasmids and Priya Raghavan for experiments, which confirmed our initial observations. We also acknowledge members of the Sudol laboratory for helpful comments on the manuscript.
This research was supported by PA Breast Cancer Coalition grants (#60707 and #9200903) and by Geisinger Clinic.
The authors declare no conflict of interest.
Supplementary Information accompanies the paper on the Oncogene website
About this article
Cite this article
Oka, T., Schmitt, A. & Sudol, M. Opposing roles of angiomotin-like-1 and zona occludens-2 on pro-apoptotic function of YAP. Oncogene 31, 128–134 (2012). https://doi.org/10.1038/onc.2011.216
- angiomotin family
- WW domains
- PDZ domains
- YAP as scaffold
- cell detachment
Clostridium perfringens enterotoxin induces claudin-4 to activate YAP in oral squamous cell carcinomas
Modular mimicry and engagement of the Hippo pathway by Marburg virus VP40: Implications for filovirus biology and budding
PLOS Pathogens (2020)
Role of Clostridium perfringens Enterotoxin on YAP Activation in Colonic Sessile Serrated Adenoma/Polyps with Dysplasia
International Journal of Molecular Sciences (2020)
Biochimica et Biophysica Acta (BBA) - Biomembranes (2020)