Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2014 Mar 13.
Published in final edited form as: Cell. 2013 Sep 5;154(6):10.1016/j.cell.2013.08.025. doi: 10.1016/j.cell.2013.08.025

Spatial organization of Hippo signaling at the plasma membrane mediated by the tumor suppressor Merlin/NF2

Feng Yin 1, Jianzhong Yu 1, Yonggang Zheng 1, Qian Chen 1, Nailing Zhang 1, Duojia Pan 1,2
PMCID: PMC3835333  NIHMSID: NIHMS526293  PMID: 24012335

Abstract

Although Merlin/NF2 was discovered two decades ago as a tumor suppressor underlying Neurofibromatosis type II, its precise molecular mechanism remains poorly understood. Recent studies in Drosophila revealed a potential link between Merlin and the Hippo pathway by placing Merlin genetically upstream of the kinase Hpo/Mst. In contrast to the commonly depicted linear model of Merlin functioning through Hpo/Mst, here we show that in both Drosophila and mammals, Merlin promotes downstream Hippo signaling without activating the intrinsic kinase activity of Hpo/Mst. Instead, Merlin directly binds and recruits the effector kinase Wts/Lats to plasma membrane. Membrane recruitment, in turn, promotes Wts phosphorylation by the Hpo-Sav kinase complex. We further show that disruption of the actin cytoskeleton promotes Merlin-Wts interactions, which implicates Merlin in actin-mediated regulation of Hippo signaling. Our findings elucidate an important molecular function of Merlin and highlight the plasma membrane as a critical subcellular compartment for Hippo signal transduction.

Introduction

The tumor suppressor Merlin encodes a member of the ezrin, radixin, moesin (ERM) family of membrane-cytoskeleton adapter proteins (Rouleau et al., 1993; Trofatter et al., 1993). It is inactivated not only in the familial cancer syndrome Neurofibromatosis type II (NF2), but also at varying frequency in sporadic tumors outside the nervous system such as mesotheliomas, bladder, thyroid and skin cancer (http://www.sanger.ac.uk/perl/genetics/CGP/cosmic?action=gene&ln=NF2). Although it is generally believed that Merlin impacts cell signaling as a membrane-cytoskeleton scaffold, the precise mechanisms by which Merlin functions as a tumor suppressor are poorly understood and remain an active area of research (Li et al., 2012). In this study, we use a combination of Drosophila and mammalian models to investigate the molecular function of Merlin. This combinatorial approach is rooted in the remarkable conservation of Merlin function between Drosophila and mammals, as illustrated by the ability of the human Merlin gene to rescue Drosophila mer mutations (LaJeunesse et al., 1998). For purposes of clarity, we shall use “Merlin” as a non-species-specific reference to this tumor suppressor protein; “mer/Mer” and “Nf2/NF2” will be used to refer to the corresponding Drosophila and mammalian gene/protein, respectively.

The Hippo signaling pathway regulates organ size in Drosophila and mammals, and its dysfunction contributes to human cancers (Halder and Johnson, 2011; Harvey and Tapon, 2007; Pan, 2010). Central to the Hippo pathway is a kinase cascade whereby the Ste20-like kinase Hippo (Hpo) (Mst1/2 in mammals) activates the Nuclear Dbf2-related (NDR) family kinase Warts (Wts) (Lats1/2 in mammals) by phosphorylating the latter on its hydrophobic motif; activated Wts/Lats, in turn, phosphorylates and inactivates the transcriptional coactivator Yorkie (Yki) (YAP/TAZ in mammals). Compared to the core kinase cascade leading from Hpo to Yki, signaling events upstream of the Hippo kinase cascade are less well understood. In searching for upstream components of the Hippo pathway, Hamaratoglu et al. analyzed mutant cells lacking Mer and the related FERM protein Expanded (Ex) and provided the first evidence placing mer genetically upstream of hpo (Hamaratoglu et al., 2006). A Merlin-Hippo connection is also supported by mouse genetics, wherein heterozygosity of Yap greatly suppresses the Nf2-deficient phenotypes in multiple tissues (Zhang et al., 2010). Consistent with these genetic findings, Mer overexpression promotes Yki phosphorylation and Wts hydrophobic motif phosphorylation in Drosophila cells (Hamaratoglu et al., 2006; Yu et al., 2010), and conversely, loss of NF2 leads to reduced YAP and Lats phosphorylation (Zhang et al., 2010). In principle, these genetic and biochemical data are compatible with at least two possibilities. First, Merlin may directly or indirectly activate Hpo/Mst, which in turn phosphorylates Wts/Lats at its hydrophobic motif. Alternatively, Merlin may directly or indirectly translocate Wts/Lats to a subcellular compartment where Wts/Lats undergoes hydrophobic motif phosphorylation by Hpo/Mst. For simplicity, these models will be referred to as linear and parallel model, respectively. Interestingly, although the existing data cannot formally distinguish between the two models, the linear model is widely assumed in the field (Halder and Johnson, 2011; Harvey and Tapon, 2007; Pan, 2010).

Here we show that in both Drosophila and mammals, Merlin does not promote Hippo signaling through Hpo/Mst in a linear pathway. Instead, Merlin directly binds and helps recruit Wts/Lats to plasma membrane, where it is activated by the Hpo-Sav kinase complex. Our studies therefore uncover a critical function of Merlin in spatial organization of Hippo signaling at the plasma membrane.

Results

Merlin promotes Wts/Lats phosphorylation without stimulating the intrinsic kinase activity of Hpo/Mst in Drosophila and mammalian cells

To distinguish between the linear and parallel models discussed above, we first examined the effect of Mer on Hpo activation in Drosophila S2R+ cells. To facilitate this analysis, we took advantage of Mer1-600, an activated form of Mer lacking 35-a.a. at its C-terminus (LaJeunesse et al., 1998). Mer1-600 was previously identified from an unbiased structure-function analysis, but the underlying molecular mechanism for its gain-of-function activity, including its relevance to Hippo signaling, was unknown (LaJeunesse et al., 1998). As reported before, overexpression of Mer1-600, but not the wildtype Mer, by the wing-specific nub-Gal4 driver resulted in a significant reduction in wing size (Figure S1A). In S2R+ cells, Mer1-600 showed stronger activities than wildtype Mer in inducing Wts T1077 (hydrophobic motif) phosphorylation (Figures 1A and S1B) and Yki S168 phosphorylation (Figure 1B), suggesting that the gain-of-function of Mer1-600 may be attributed to its enhanced activity in promoting Hippo signaling. Interestingly, under these conditions, neither Mer nor Mer1-600 enhanced Hpo kinase activity, as revealed by a phospho-specific antibody against Hpo’s autophosphorylation site (T195) (Figure 1C). This was further confirmed by in vitro kinase assays using immunoprecipitated Hpo against bacterially purified GST-Wts as substrate (Figure 1D). Even though Mer or Mer1-600 did not enhance Hpo’s kinase activity per se, Mer- or Mer1-600-mediated Wts T1077 phosphorylation requires Hpo (Figure 1E), consistent with Hpo being a hydrophobic motif kinase of Wts. Conversely, RNAi knockdown of Mer resulted in decreased Yki and Wts phosphorylation without a concomitant decrease in Hpo activity – to the contrary, we consistently observed a slight increase in Hpo activity upon Mer RNAi (Figure 1F) (see later for further discussion). Thus, while Mer promotes Wts phosphorylation at its Hpo-responsive hydrophobic motif, it does so without activating the intrinsic kinase activity of Hpo. These results therefore contradicts a simple linear model placing Mer biochemically upstream of Hpo activation.

Figure 1. Merlin promotes Hippo signaling without stimulating the intrinsic kinase activity of Hpo/Mst. See also Figure S1.

Figure 1

Open in a new tab

(A-C) Mer/Mer1-600 promotes Wts and Yki phosphorylation without affecting Hpo phosphorylation. S2R+ cell lysates expressing the indicated constructs were probed for P-Wts (A), P-Yki (B) or P-Hpo (C). Relative P-Wts and Wts levels were quantified in Figure S1B.

(D) In vitro kinase assay. Myc-Hpo was expressed alone or together with HA-Mer1-600 in S2R+ cells, immunoprecipitated, and incubated with GST-Wts, and the reaction products were probed for P-Wts.

(E) S2R+ cells transfected with the indicated plasmids along with control or hpo dsRNA were probed with the indicated antibodies. Note reduction of P-Wts levels by hpo RNAi.

(F) RNAi of mer leads to decreased P-Yki and P-Wts but increased P-Hpo levels in S2R+ cells. Also note comparable levels of total Yki, Wts or Hpo in control and RNAimer cells.

(G) NF2 promotes Lats1 phosphorylation without affecting Mst1 phosphorylation. HEK293 cells expressing the indicated constructs were probed for P-Lats and P-Mst.

(H) Control (Nf2flox/flox) and Nf2 (Alb-Cre; Nf2flox/flox) livers from 2-month old littermates were probed with the indicated antibodies. Note the comparable levels of Lats1 and YAP but diminished P-Lats and P-YAP levels in Nf2 mutant livers. Also note the increased P-Mst levels in Nf2 mutants.

After demonstrating that Mer promotes Wts and Yki phosphorylation without activating the Hpo kinase activity in Drosophila, we examined whether a similar mechanism also operates in mammals. Indeed, overexpression of NF2 in HEK293 cells promoted Lats1/2 phosphorylation without enhancing Mst1/2 kinase activity (Figures 1G and S1C). These results were further corroborated with loss-of-function studies. We recently reported that Nf2-deficient livers (Alb-Cre;Nf2flox2/flox2) display reduced Lats1/2 and YAP phosphorylation, supporting the view that NF2 is normally required for Hippo signaling in mammals (Zhang et al., 2010). If NF2 activates downstream Hippo signaling by stimulating the intrinsic kinase activity of Mst1/2, one would expect to see a corresponding decrease in Mst1/2 activity in Nf2-deficient livers. Contrary to this prediction, we found that concurrent with a decrease in Lats1/2 and YAP phosphorylation, the Nf2-deficient livers instead showed a significant increase in Mst1/2 activity (Figure 1H). The opposite changes in Lats/YAP and Mst1/2 phosphorylation upon loss of NF2 therefore argue against the generally assumed linear model in which NF2 signals through activation of Mst1/2.

The increase in Hpo/Mst activity upon loss of Merlin in S2R+ cells or mouse livers is likely due to negative feedback in Hippo signaling, whereby increased Yki/YAP activity leads to elevated transcription of multiple upstream components such as Expanded and Kibra (Hamaratoglu et al., 2006; Genevet et al., 2010; Xiao et al., 2011), which in turn, may promote Hpo/Mst activation. While the molecular details of this feedback regulation warrant further investigation, our results clearly demonstrate that in the absence of Merlin, even enhanced Hpo/Mst kinase activity cannot stimulate its downstream kinase Wts/Lats. Taken together, these results indicate that in both Drosophila and mammals, Merlin activates Wts/Lats phosphorylation without stimulating the intrinsic kinase activity of Hpo/Mst.

Merlin promotes plasma membrane association of Wts/Lats in Drosophila and mammalian cells

Having excluded the linear model, we next explored the possibility that Mer, a predominantly membrane-associated protein (LaJeunesse et al., 1998), may play a role in targeting Wts to the plasma membrane. Strikingly, co-expression of Mer1-600, but not the wildtype Mer or the related protein Ex, resulted in a dramatic translocation of Wts to the plasma membrane, as revealed by subcellular fractionation (Figure 2A-B) and immunofluorescent staining (Figure 4G). This translocation is independent of hydrophobic motif phosphorylation or Mats binding, since Wts mutants lacking its hydrophobic motif phosphorylation site (WtsT1077A) (Yu et al., 2010) or the Mats-binding site (WtsR708A) (Ho et al., 2010) were translocated by Mer1-600 to the membrane fraction as efficiently as wildtype Wts (Figures 2B and S2A-C). Consistent with these overexpression studies, RNAi knockdown of Mer resulted in a significant increase of endogenous Wts in the cytosol relative to the membrane fraction (Figure 2C).

Figure 2. Merlin promotes plasma membrane association of Wts/Lats in Drosophila and mammalian cells. See also Figure S2.

Figure 2

Open in a new tab

(A) Mer1-600, but not Mer or Ex, induces Wts membrane translocation. S2R+ cells expressing the indicated constructs were fractioned into membrane (M) and cytosolic (C) fraction and probed with the indicated antibodies. T: total cell lysates.

(B) Mer1-600 induces membrane association of Wts and WtsT1077A in S2R+ cells. Subcellular fractions were analyzed as in (A).

(C) RNAi of mer results in a significant increase of endogenous Wts in the cytoplasmic fraction relative to the membrane fraction. Subcellular fractions were probed with α–Wts antibody. Nervana, a plasma membrane-localized Na+/K+ ATPase, serves as a control for fractionation.

(D) NF2 promotes membrane association of Lats1 and Lats1T1079A in HEK293 cells, as revealed by subcellular fractionation.

(E) Loss of NF2 results in increased cytosolic accumulation of endogenous Lats1 and Lats2. Subcellular fractions of confluent FH-912 and FC-912 cells were probed.

(F) FC-912 cells were transfected with empty vector (pcDNA), NF2 or NF2 mutants along with Myc-Lats1. Subcellular fractions were probed for Myc-Lats1. FC-912 cells expressing empty vector or NF2 mutants had higher Myc-Lats1 levels in cytosolic than membrane fraction, but FC-912 cells expressing wildtype NF2 showed similar cytosolic and membrane Myc-Lats1 levels. The cell lysates were also probed for the expression of NF2 mutants (bottom gel).

(G) Liver sections from 8-day-old control and Nf2-deficient littermates, stained with Lats2 (red) and the plasma membrane marker Pan-cadherin (green). Lats2 is concentrated at the cell cortex immediately on the inner face of plasma membrane in wildtype hepatocytes, but shows a uniform cytoplasmic localization in Nf2-deficient hepatocytes.

Figure 4. Identification of a conserved N-terminal motif in Wts/Lats that mediates Merlin-Wts/Lats interactions. See also Figure S4.

Figure 4

Open in a new tab

(A) HEK293 cells expressing HA-NF2 and various truncations of Lats1 were subjected to immunoprecipitation as indicated.

(B) Alignment of the conserved N-terminal motif (CNM) in Wts/Lats protein from different species.

(C) Myc-Lats1, but not Myc-Lats13m, was co-immunoprecipitated with HA-NF2 in HEK293 cells.

(D-E) The CNM is required for NF2-induced membrane association of Lats1 in HEK293 cells. Subcellular fractionation (D) and immunofluorescent staining (E) of Myc-Lats1 or Myc-Lats13m in the presence or absence of NF2.

(F-G) The CNM is required for Mer1-600-mediated membrane association of Wts in Drosophila S2R+ cells. Subcellular fractionation (F) and immunofluorescent staining (G) of V5-Wts or V5-Wts3m in the presence or absence of Mer1-600.

(H) UAS-Wts and UAS-Wts3m transgenes inserted at identical chromosomal locus were crossed to nub-Gal4 driver. Note the small-wing phenotype induced by Wts, but not Wts3m.

Next, we examined whether NF2 plays a similar role in mammals. Indeed, co-expression of NF2 in HEK293 cells resulted in a dramatic translocation of Lats1/2 to plasma membrane, as revealed by subcellular fractionation (Figures 2D and S2D) and immunofluorescence (Figure 4E). To complement these gain-of-function studies, we examined the effect of NF2-deficiency on endogenous Lats1/2 localization in mouse Schwann cells, a cell type that is directly involved in the pathogenesis of the NF2 disease. FC-912 and FH-912 are genetically matched Schwann cells derived from primary Nf2flox/flox mouse Schwann cells with and without in vitro adenoviral Cre infection, respectively (Li et al., 2010). As expected, the Nf2-/- FC-912 cells were defective in Lats1/2 phosphorylation (Figure S2E-F). Subcellular fractionation revealed a significant increase of endogenous Lats1/2 in the cytoplasmic relative to the membrane fraction in the Nf2-/- FC-912 Schwann cells compared to the Nf2+/+ FH-912 Schwann cells (Figure 2E). This localization defect was rescued by re-expression of NF2 (Figure 2F). Consistent with these findings, examination of Nf2-deficient livers revealed a loss of membrane-associated Lats2 localization in hepatocytes (Figure 2G). Taken together, these results indicate that Merlin is required for Wts/Lats membrane association in both Drosophila and mammalian cells.

Active Merlin recruits Wts/Lats to plasma membrane through direct protein-protein interaction

To understand the molecular mechanism by which Mer1-600 promotes membrane association and phosphorylation of Wts, we examined protein interactions between Mer1-600 and Wts. Consistent with its gain-of-function nature in stimulating Wts/Yki phosphorylation and Wts membrane association, Mer1-600, but not the wildtype Mer or the related protein Ex, coimmunoprecipitated with Wts in Drosophila S2R+ cells (Figure 3A). Thus, the ability of Mer1-600 to promote Hippo signaling and Wts membrane localization correlates with its ability to physically associate with Wts. The functional significance of Mer1-600-Wts interaction was also confirmed by genetic interactions in vivo, whereby co-expression of Wts with Mer1-600, but not with wildtype Mer, resulted in a synergistic reduction in eye size (Figure S3A).

Figure 3. Active Merlin binds to Wts/Lats through its FERM domain. See also Figure S3.

Figure 3

Open in a new tab

(A) Wts associates with Mer1-600, but not Mer or Ex, as revealed by immunoprecipitation in S2R+ cells as indicated.

(B) Mer1-600 exhibits an open conformation. S2R+ cells expressing Mer-N fragment (1-375) and wildtype (375-635) or truncated (375-600) Mer-C fragment were subjected to immunoprecipitation.

(C) Lats1 associates with N-terminal half of NF2, as revealed by immunoprecipitation in HEK293 cells.

(D) Co-immunoprecipitation between endogenous NF2 and Lats1/2 in HEK293 cells. Endogenous Lats1 or Lats2 was detected in α-NF2, but not control, immunoprecipitates.

(E) NF2-Lats1 binding in vitro. HEK293 cell lysates containing Myc-Lats1 or Myc-Lats13m were incubated with glutathione Sepharose beads containing GST (as a control) or GST-NF2. GST-NF2 bound to Myc-Lats1 but not Myc-Lats13m.

(F-H) C-terminal truncation of NF2 results in a more open conformation and enhances Lats1 binding and phosphorylation. HEK293 cells expressing the indicated constructs were probed for association between NF2-N and NF2-C fragments (F), NF2-Lats1 association (G) and Lats1 hydrophobic motif phosphorylation (H).

(I) Wildtype NF2, but not disease-associated missense mutants, associates with Lats1, as revealed by immunoprecipitation in HEK293 cells.

(J) Wildtype NF2, but not disease-associated missense mutants, promotes membrane association of Lats1 in HEK293 cells, as revealed by subcellular fractionation.

Next, we investigated, at a quasi-structural level, why Mer1-600 confers gain-of-function activity. Merlin, like the related ezrin-radixin-moesin (ERM) proteins, can exist in two functional/conformational states (Bretscher et al., 2002). In the open and more-active conformation, the N-terminal FERM domain of Merlin is free to bind to its effectors. In the more-closed conformation, the C-terminal tail (CTT) domain associates with the N-terminal FERM domain, blocks the accessibility of the FERM domain to its effectors and therefore results in a less-active state (Sher et al., 2012). Conversion between the two states can be regulated by phosphorylation on a unique CTT residue (S518) in mammalian NF2 (Jin et al., 2006; Shaw et al., 2001; Surace et al., 2004), although this residue is not conserved in Drosophila Mer. The ability of Mer1-600, but not Mer, to bind Wts suggests that the former, but not the latter, may represent an open conformation. To test this hypothesis, we used co-immunoprecipitation to examine any physical interactions between the C-terminal half of Mer (Mer-C), derived from either wildtype or truncated Mer, and the N-terminal half of Mer (Mer-N). Of note, a similar assay was used recently to assess the conformational states of mammalian NF2 (Sher et al., 2012). As shown in Figure 3B, the wildtype Mer-C, but not the truncated Mer-C, strongly interacted with Mer-N. These results suggest that Mer1-600 displays a more open conformation than the wildtype Mer, which provide a molecular explanation for the gain-of-function activity of Mer1-600 in growth suppression, activation of Hippo signaling, Wts-binding and membrane recruitment.

As in Drosophila, NF2 and Lats1/2 co-immunoprecipitated with each other in HEK293 cells (Figures 3C and S3B). This interaction was confirmed with endogenous NF2 and Lats1/2 proteins (Figure 3D), and by in vitro GST pull-down assays (Figure 3E). NF2 appeared to be intrinsically more active than Mer, since the full-length NF2 (but not Mer) sufficed to bind and to recruit Lats1/2 to plasma membrane (Figures 3C and 2D; compare to Figures 3A and 2A). When we used a similar co-immunoprecipitation assay to detect interaction between the N- and C-terminal domains of NF2, this interaction was significantly weaker than that detected between their Drosophila counterparts at comparable protein input levels (Figure 3F vs. Figure 3B). Thus, the stronger activity of NF2 in binding Lats1/2 correlates with an intrinsically more open conformation of NF2 than Mer. Despite such differences, a C-terminal truncation of NF2 (NF21-559, which is analogous to Mer1-600) resulted in enhanced binding and activation of Lats1 (Figure 3G-H), as well as a more open conformation (Figure 3F). In addition, the phosphodeficient S518A mutant of NF2, which is known to display a more open/active conformation compared to the phosphomimetic S518D mutant (Sher et al., 2012), also showed greater ability to bind Lats1 and to promote Lats1 phosphorylation (Figure S3C-D). Thus, despite differences in basal activity and “openness” of the proteins, in both Drosophila and mammalian cells, a conformational change in Merlin to a more open form leads to enhanced Wts/Lats-binding and Hippo signaling.

The FERM domain of Merlin and an evolutionarily conserved N-terminal motif in Wts/Lats mediates Wts/Lats-Merlin binding

Using co-immunoprecipitation assays, we mapped protein domains involved in Mer/NF2-Wts/Lats interactions. The N-terminal half of Mer and NF2, which corresponds to the FERM domain, is required for interaction with Wts (Figure S3E-F) and Lats1 (Figure 3C), respectively. Further truncations into the FERM domain abolished this interaction (Figure S3G). Interestingly, most disease-associated missense mutations in NF2 map to its FERM domain (Ahronowitz et al., 2007). We therefore examined a collection of human disease-associated NF2 mutations in the FERM domain (L46R, F62S, L64P and L141P; mapped to the 3D structure of NF2 FERM domain (Kang et al., 2002) in Figure S3H) for their ability to interact with Lats1 and to recruit Lats1 to plasma membrane. Unlike wildtype NF2, these pathological mutants failed to interact with Lats1 (Figure 3I) or to promote membrane association of Lats1 in HEK293 cells (Figure 3J). We also examined these missense mutations for their ability to rescue the Lats1 membrane association defect in the Nf2-/- FC-912 Schwann cells. Unlike wildtype NF2, none of the pathological mutations could rescue the Lats1 localization defect in FC-912 cells (Figure 2F). The impaired ability of pathological NF2 mutations to interact with and to recruit Lats1 to plasma membrane further support the physiological relevance of NF2-Lats interactions.

In the reverse direction, the N-terminal half of Lats1 (Lats11-500), but not its C-terminal half containing the kinase domain, is required for interaction with NF2 (Figure 4A). The NF2-binding region was further mapped to Lats11-152 (Figure 4A) and a similar region in Lats2 (Figure S4A). The overall sequence similarity between Drosophila and human Wts/Lats is very low in this region. However, a careful examination of Wts/Lats orthologues from multiple insects and vertebrates revealed a conserved sequence motif (Figure 4B). In fact, this sequence, hereafter referred to as conserved N-terminal motif (CNM), represents the only evolutionarily conserved sequence in the N terminal half of Wts/Lats. Interestingly, mutation of the three most conserved residues in the CNM (Lats1-A77P-I81T-L85P, abbreviated as Lats13m) abolished NF2-Lats1 interactions in co-immunoprecipitation (Figure 4C) and GST pull-down assays (Figure 3E), and abolished NF2-induced membrane association of Lats1 (Figure 4D-E). Introduction of the same mutation into Wts (Wts-A20P-I24T-L28P, abbreviated as Wts3m) greatly diminished Mer1-600-induced Wts membrane association (Figure 4F-G), and abolished the growth-suppressive activity of Wts in Drosophila (Figure 4H). When fused to GFP, a minimal fragment containing the CNM (a.a. 1-50) was sufficient to recruit GFP to plasma membrane in a Mer1-600-dependent manner (Figure S4B). Interestingly, although the CNM and the Mer-C domain bear no sequence similarities, interaction between the CNM and the Mer-N domain could be competed away by the Mer-C domain (Figure S4C). Taken together, these results uncover a previously unrecognized but evolutionarily conserved sequence motif in Wts/Lats that is crucial for Merlin-binding and growth-suppressive activity.

Constitutive membrane targeting promotes Hpo-mediated hydrophobic motif phosphorylation and tumor suppressor activity of Wts

The data presented so far support a working model in which active Mer functions by recruiting Wts to plasma membrane, where Wts undergoes Hpo-mediated hydrophobic motif phosphorylation. We conducted additional experiments to further corroborate this model.

One prediction of this model is that constitutive targeting of Wts to plasma membrane may result in enhanced hydrophobic motif phosphorylation and gain-of-function activity for Wts. This was tested by fusing Wts with a well-characterized plasma-membrane targeting sequence, the myristylation signal (Myr) of Drosophila Src (Myr-Wts) (Struhl and Adachi, 1998). As a control, we fused Myr with the hydrophobic motif-defective WtsT1077A. As expected, Myr-Wts and Myr-WtsT1077A showed exclusive plasma membrane association (Figure S5A-B). In agreement with our prediction, Myr-Wts stimulated Wts and Yki phosphorylation more potently than Wts (Figure 5A). We also compared the growth-inhibitory activity of Myr-Wts and Wts in Drosophila. Overexpression of Myr-Wts by the wing-specific nub-Gal4 driver led to a complete loss of wing tissue, whereas overexpression of Wts led to a less severe reduction in wing size (Figure 5B), demonstrating that membrane targeting of Wts promotes its tumor suppressor activity in vivo.

Figure 5. Constitutive membrane targeting promotes Hpo-mediated hydrophobic motif phosphorylation and tumor suppressor activity of Wts. See also Figure S5.

Figure 5

Open in a new tab

(A) S2R+ cells expressing the indicated constructs were probed for P-Yki and P-Wts.

(B) Enhanced growth-suppressive activity of membrane-targeted Wts and dominant negative activity of membrane-targeted WtsT1077A. All UAS Wts transgenes were inserted at the same locus. Note the complete loss of adult wings in nub: Myr-Wts flies, and the suppression of small-wing phenotype of nub:Hpo in flies that co-expressed Myr-WtsT1077A.

(C-E) Partial rescue of hpo mutant clones by Myr-Wts. Eye discs containing clones of the indicated genotype (GFP+, arrowheads) were stained for Diap1 (red). Note elevated Diap1 levels in hpo clones without or with Myr-Wts expression. Also note the smaller clone size and less rounded shape of the latter genotype. Clones size is quantified in (E).

Another prediction of this model is that hydrophobic motif phosphorylation is required for the gain-of-function activity of Myr-Wts. Indeed, in contrast to Myr-Wts, Myr-WtsT1077A showed no activity in vivo (Figure 5B), demonstrating that membrane-targeted Wts requires hydrophobic motif phosphorylation for its activation. In fact, Myr-WtsT1077A behaves as a dominant negative form -- it greatly suppressed Hpo-induced undergrowth of eyes and wings, and enhanced Yki-induced eye overgrowth (Figures 5B and S5C). The ability of a membrane-targeted phospho-defective Wts mutant to interfere with Hippo signaling further implicates the plasma membrane as a critical subcellular compartment for Wts phosphorylation and activation.

A third prediction of this model is that despite its gain-of-function activity in wildtype background, Myr-Wts should not rescue hpo mutant clones, as membrane-targeted Wts should require Hpo for its activation. To test this prediction, we analyzed hpo mutant clones with concurrent Myr-Wts overexpression. Indeed, Myr-Wts cannot rescue hpo mutant phenotypes, as hpo mutant clones with Myr-Wts overexpression still exhibited elevated diap1 expression, a hallmark of defective Hippo signaling (Figure 5C-D). Interestingly, hpo mutant clones expressing Myr-Wts were smaller in size than hpo mutant clones (Figure 5C-E), suggesting that Myr-Wts may partially bypass Hpo. Given that T1077 phosphorylation is absolutely required for Myr-Wts activity in vivo (Figure 5B), the simplest possibility is that in the absence of Hpo, an unknown Hpo-like kinase may phosphorylate the hydrophobic motif of Wts. Although the identity of such Hpo-like kinases requires further investigation, this view is consistent with the stronger overgrowth of wts compared to hpo null mutant tissues in Drosophila (Wu et al., 2003), as well as multiple studies in mammalian cells that unequivocally demonstrated Mst1/2-independent phosphorylation of Lats1/2 hydrophobic motif (Yu et al., 2012; Zhou et al., 2009).

Sav and Mer play distinct roles in spatial organization of the Hippo kinase cassette at the plasma membrane

The gain-of-function activity of Myr-Wts and the dominant negative activity of Myr-WtsT1077A suggest that the plasma membrane is an important subcellular compartment for Hippo signal transduction. A comprehensive survey of Hippo pathway components showed that like Merlin, Sav/Sav1, a scaffold protein that has been reported to bind Wts/Lats and Hpo/Mst (Harvey and Tapon, 2007), is localized almost exclusively in the membrane fraction in both Drosophila and mammalian cells (Figure S6A-B). Given its near exclusive membrane association, we investigated the possibility that Sav may also play a role in scaffolding the Hippo kinase cassette at the plasma membrane.

Using subcellular fractionation, we examined whether Sav affects the membrane association of Hpo and Wts. Interestingly, co-expression of Sav induced translocation of Hpo but not Wts to the membrane fraction (Figure 6A-B). Conversely, RNAi knockdown of Sav led to a significant decrease of endogenous Hpo in the membrane fraction relative to the cytoplasmic fraction (Figure 6C). Sav-induced membrane translocation of Hpo is mediated by direct binding, since it was abolished by mutation of Sav’s SARAH domain (Figure 6D), which is known to mediate Sav-Hpo binding (Scheel and Hofmann, 2003). Notably, despite its potent activity in recruiting Wts to plasma membrane, Mer1-600 had no effect on membrane association of Hpo (Figure 6A-B). Unlike Mer and Sav, other components of the Hippo pathway failed to promote Hpo or Wts membrane association in our assays (Figure S6C-D). Thus, it appears that Mer and Sav play separable roles in membrane organization of the Hippo kinase cassette by targeting Wts and Hpo to plasma membrane, respectively. Similar observations were made in mammalian cells, as Sav1 promoted membrane association of Mst1 but not Lats1, and conversely, NF2 promoted membrane association of Lats1 but not Mst1 (Figure S6E-F). Consistent with fractionation, immunofluorescence confirmed that co-expression of Merlin1-600 and Sav induced translocation of both Wts and Hpo to plasma membrane in Drosophila S2R+ cells (Figure S6G). Likewise, co-expression of NF2 and Sav promoted membrane translocation of Lats1 and Mst1 in HEK293 cells (Figure S6H).

Figure 6. Sav-mediated membrane association of Hpo and regulation of Merlin-Wts binding by actin cytoskeleton. See also Figure S6.

Figure 6

Open in a new tab

(A) S2R+ cells expressing Myc-Hpo together with HA-Mer1-600 or FLAG-Sav were fractioned into membrane (M) and cytosolic (C) fraction. T: total cell lysates. Sav, but not Mer1-600, promoted Hpo membrane association.

(B) S2R+ cells expressing V5-Wts together with HA-Mer1-600 or FLAG-Sav were analyzed by subcellular fractionation. Mer1-600, but not Sav, promoted Wts membrane association.

(C) RNAi of sav results in decreased membrane localization of endogenous Hpo. Subcellular fractions were probed with α–Hpo antibody.

(D) Savshrp6 mimics a sav allele that deletes just the SARAH domain (Kango-Singh et al., 2002). Unlike wildtype Sav, Savshrp6 failed to induce membrane recruitment of Hpo.

(E) S2R+ cells expressing the indicated constructs were analyzed for P-Wts. Wildtype Sav, but not Savshrp6 (Kango-Singh et al., 2002) or SavΔFBM (Yu et al., 2010), greatly potentiated Mer1-600-induced Wts phosphorylation.

(F-G) S2R+ cells expressing the indicated constructs were treated with DMSO (0.1%, lane 1), LatB (1 μg/ml, lane 2) or Fosk (30 μM, lane 3) for 1 hr, followed by immunoprecipitation. LatB, but not Fosk, promoted Mer-Wts interaction (F). Neither drug affected Sav-Hpo interaction (G).

(H) S2R+ cells expressing the indicated constructed were treated with DMSO (0.1%, lane 1), LatB (1 μg/ml, lane 2) or C3 (2 μg/ml, lane 3), followed by immunoprecipitation.

(I) S2R+ cells were treated with DMSO (0.1%, lane 1), LatB (1 μg/ml, lane 2) or C3 (2 μg/ml, lane 3), followed by immunoprecipitation with anti-Mer antibody and detection with anti-Wts antibody.

(J) S2R+ cells expressing HA-Mer1-600 and V5-Wts were treated with DMSO, LatB or C3 as in (H-I) followed by immunoprecipitation. Neither treatment affected interactions between Mer1-600 and Wts.

(K) Schematic models of Merlin function in Hippo signaling. For simplicity, only proteins relevant to the current study are illustrated in the schematic diagrams. Left: the prevailing linear model placing Merlin upstream of Hpo activation. Right: a new model based on the current study highlighting a dual mechanism in which Merlin and Sav promotes the membrane association of Wts and Hpo, respectively.

To investigate the functional role of Sav-induced membrane recruitment of Hpo, we analyzed the effect of Sav on Hippo signaling output in S2R+ cells. Although overexpression of Sav alone had no visible effect on Wts phosphorylation, co-expression of Sav with Mer1-600 led to synergistic activation of Wts phosphorylation. Mutation of the SARAH domain of Sav abolished this synergistic effect (Figure 6E), supporting the importance of Sav-Hpo binding in Hpo membrane recruitment and signal transduction. Sav was recently reported to bind to Mer through an evolutionarily conserved FERM-binding motif (FBM) (Yu et al., 2010). Interestingly, mutation of the FBM in Sav also abolished this synergistic effect (Figure 6E), suggesting that a Mer-Sav interaction may be involved in bringing Hpo and Wts together at the plasma membrane to further facilitate the Hippo kinase cascade.

Regulation of Merlin-Wts interaction by the actin cytoskeleton

Although wildtype NF2 suffices to bind Lats1/2 in mammalian cells, our analysis of Mer-Wts interactions in Drosophila has so far relied on the use of a constitutive active mutant, Mer1-600. Thus it remains an open question if Wts could interact with wildtype Merlin under any circumstances. We therefore used the S2R+ cells to investigate whether physical interactions between Wts and wildtype Merlin can be detected under conditions that potentially stimulate Hippo signaling. Disruption of actin cytoskeleton has been reported to activate Hippo signaling in both Drosophila and mammalian cells (Dupont et al., 2011; Fernandez et al., 2011; Sansores-Garcia et al., 2011; Wada et al., 2011; Zhao et al., 2012). In addition, certain GPCR ligands such as epinephrine and lysophosphatidic acid (LPA) have been reported to activate or inhibit Hippo signaling in mammalian cells (Yu et al., 2012). Since Drosophila lacks LPA receptors (Metpally and Sowdhamini, 2005) and does not produce epinephrine (Roeder, 2005), these particular ligands are unlikely to regulate Hippo signaling in Drosophila. Nevertheless, we tested Protein Kinase A (PKA), which was reported to lie downstream of GPCRs in GPCR-Hippo signaling in mammalian cells (Yu et al., 2013). Strikingly, we detected robust interactions between wildtype Merlin and Wts in S2R+ cells treated with the actin polymerization inhibitor Latrunculin B (LatB), but not in cells treated with DMSO (control) or the PKA agonist Forskolin (Fosk) (Figure 6F). In contrast, interactions between Sav and Hpo were not modulated by LatB or Fosk (Figure 6G). Thus, it appears that actin cytoskeleton disruption specifically enhanced physical interactions between Merlin and Wts.

To further corroborate a link between the actin cytoskeleton and Merlin, we treated S2R+ cells with the Rho GTPase inhibitor C3. Like LatB, treatment with C3 also promoted Merlin-Wts interaction (Figure 6H), in agreement with findings in mammalian cells implicating the Rho GTPase as a critical regulator of actin cytoskeleton in modulating Hippo signaling (Dupont et al., 2011; Wada et al., 2011; Zhao et al., 2012). In fact, only when S2R+ cells were treated with these actin inhibitors could we detect interactions between endogenous Merlin and Wts (Figure 6I). These results suggest that Merlin may be a critical mediator of actin-regulated Hippo signaling. Consistent with this hypothesis, RNAi knockdown of Mer attenuated LatB- or C3-induced Yki phosphorylation in S2R+ cells (Figure S6I). Significantly, while LatB and C3 promoted interactions between Wts and wildtype Merlin, neither inhibitor affected physical interactions between Wts and Mer1-600, the constitutively active mutant form of Merlin (Figure 6J). Such epistasis further supports a functional link between the actin cytoskeleton and Merlin function.

A dual mechanism of Wts/Lats activation is required for physiological Hippo signaling in Drosophila and mammalian tissues in vivo

The results presented so far lead to the following model of Merlin-mediated Hippo signaling: Merlin binds and recruits Wts to plasma membrane, which in effect increases the concentration of Wts at the plasma membrane; furthermore, Sav acts as a membrane-associated scaffold that increases the concentration of Hpo at the plasma membrane. Thus, it appears that full activation of Wts requires a dual mechanism that involves increasing the concentration of both the substrate (Wts) and its activating kinase (Hpo) at the membrane (Figure 6K).

A genetic prediction of this dual mechanism is that double mutants disabling both arms of the regulation may result in stronger phenotypes than disabling each single arm. Indeed, examination of supernumerary interommatidial cells in Drosophila pupal retina, a hallmark of defective Hippo signaling, revealed a massive increase in the number of supernumerary interommatidial cells in mer; sav double mutant eyes compared to mer or sav single mutant eyes (Figure 7A). To determine if this is also true in mammals, we compared mice with liver-specific knockout of Nf2, Sav1, or both genes (Figure 7B-F). As shown previously, Nf2-deficient livers appeared macroscopically of normal size at 8 days of age (~5% liver/body ratio) and microscopically showed mild focal overproliferation of biliary epithelial cells (BECs). Sav1-deficient livers appeared macroscopically of normal size at 8 days of age (~5% liver/body ratio) and microscopically showed normal BECs. Strikingly, age-matched Nf2; Sav1 double mutant livers showed massive liver overgrowth (~18% liver/body ratio) (Figure 7E). Microscopically, the double mutant livers showed massive overproliferation of BECs resembling bile duct hamartoma, in which BECs had overtaken the hepatocytes to become the predominant cell types in the liver (Figure 7B-D). The double mutant livers also showed a more dramatic decrease of YAP and Lats1/2 phosphorylation compared to the single mutant tissues (Figure 7F). Furthermore, in contrast to the life expectancy of >1 yr for each single mutant (Lu et al., 2010; Zhang et al., 2010), all the double mutant mice died before reaching 3 weeks of age (N=7). Thus, the dual mechanism of Wts/Lats activation is required for physiological Hippo signaling in both Drosophila and mice.

Figure 7. Loss of Mer/NF2 and Sav/Sav1 leads to synergistic defects in Hippo signaling in Drosophila and mice.

Figure 7

Open in a new tab

(A) Synergistic effect of mer and sav mutations in Drosophila. Mid-pupal retina of the indicated genotypes was stained for Discs-Large to highlight cell outlines. Note the significantly increased number of interommatidial cells in mer; sav double mutant eyes.

(B-E) Synergistic effect of Nf2 and Sav1 mutations in mice. Livers from 8 day-old pups of the following genotype were analyzed: control (Alb-Cre), Nf2 mutant (Alb-Cre; Nf2flox2/flox2), Sav1 mutant (Alb-Cre; Sav1flox/flox), and double mutant (Alb-Cre; Nf2flox2/flox2; Sav1flox/flox). Whole amount liver images (B), H&E staining (C) and Pan-CK staining (D) were shown, along with quantification of liver weight (E). While Nf2 and Sav1 livers displayed mild or no overproliferation of the CK-positive BECs, Nf2; Sav1 double mutant livers were predominantly occupied by the BECs. Also note the unique external appearance and overgrowth of Nf2; Sav1 double mutant livers.

(F) Western blot analysis of liver lysates from mouse livers analyzed in (B-E). Liver lysates from three mice of each genotype were probed. Note the further decrease in YAP and Lats phosphorylation in Nf2; Sav1 double mutant livers compared to the single mutant. Also note the comparable level of Lats1 in control and Nf2 mutant livers.

Discussion

The molecular function of Merlin in Hippo signaling

Since its initial discovery as a human disease gene underlying Neurofibromatosis 2, the tumor suppressor Merlin has been the subject of intense investigation. Besides the Hippo pathway, Merlin has been linked to a variety of mechanisms such as transmembrane receptor endocytosis/localization (EGFR and CD44) and signaling by Ras, Rac/PAK and PI3K pathways (Li et al., 2012). Paradoxically, as a membrane-associated tumor suppressor, Merlin was also reported to suppress tumorigenesis in mammalian cells by translocating to the nucleus to inhibit a specific E3 ubiquitin ligase (Li et al., 2012). Among these proposed targets, the linkage between Merlin and Hippo signaling has attracted much attention given the similarity of the respective mutant phenotypes in Drosophila, and the dosage sensitive genetic suppression of Merlin mutant phenotypes by heterozygosity of the Hippo effector YAP in multiple mouse tissues.

Despite the genetic evidence implicating Merlin in Hippo signaling, the molecular basis of this functional link was unknown. Our current study addresses this outstanding issue in two important ways. First, we provide molecular evidence showing that Merlin promotes downstream Hippo signaling without activating the intrinsic kinase activity of Hpo/Mst. Our studies therefore disprove the prevailing assumption that Merlin functions biochemically upstream of Hpo activation (Figure 6K). Along this line, we note that current models of Hippo signaling are actually a composite of true molecular relationships (such as Hpo acting upstream of Wts or Wts acting upstream Yki) and genetic epistasis relationships (such as Mer acting upstream of Hpo). In light of the current study, we caution that mix-and-match of biochemical and epistasis relationships in signaling diagrams can be misleading and should be clearly distinguished. Second, we elucidate direct physical interactions between Merlin and Wts/Lats, and show that such interactions promote Hippo signaling by recruiting Wts/Lats to plasma membrane. The discovery of physical interactions between Merlin and a key component of the Hippo pathway therefore provides molecular support for a Merlin-Hippo connection that has so far been based largely on genetics and indirect evidence. Interestingly, interactions between Merlin and Wts are regulated by the actin cytoskeleton, underscoring Merlin as a potential mediator of actin-regulated Hippo signaling.

Besides identifying a conserved molecular function for Merlin, our studies also revealed quantitative differences between Drosophila and mammalian Merlin. Wildtype Mer does not associate with Wts in Drosophila cells, yet wildtype NF2 suffices to bind Lats1/2 in human cells. Such differences correlate with an intrinsically more open conformation of NF2 compared to Mer. These findings agree with previous reports that the intramolecular interaction in NF2 is relatively weak and dynamic (Nguyen et al., 2001). We note that the intrinsically more active/open state of NF2 is consistent with the role of S518 phosphorylation in antagonizing NF2 activity, and the absence of this negative regulatory site in Drosophila Mer. Obviously, such negative regulation would be of more functional relevance in the context of an intrinsically more active Merlin protein as in mammals.

Plasma membrane as the key subcellular compartment for Hippo signal transduction

The plasma membrane is the entry point of diverse environmental stimuli and is intimately involved in spatial organization of signaling proteins (Groves and Kuriyan, 2010). Although many reported upstream regulators of the Hippo pathway in Drosophila are transmembrane proteins (e.g., Fat and Crumbs) or localized in apical membrane domains (e.g., Mer, Ex and Kibra), how these membrane-associated inputs spatially organize the Hippo kinase cassette was poorly understood. This question is further complicated by the apparent evolutionary divergence of upstream inputs into the pathway between Drosophila and mammals (Bossuyt et al., 2013). Notably, among these upstream inputs, Merlin is the only protein whose contribution to Hippo signaling has been genetically validated in both flies and mammals (Hamaratoglu et al., 2006; Zhang et al., 2010).

Our study demonstrates that an important and evolutionarily conserved molecular function of Merlin is to promote the membrane association of Wts/Lats. We also implicate Sav as a membrane-associated scaffold that promotes the membrane association of Hpo/Mst, the upstream kinase of Wts/Lats. Thus, two predominantly membrane-associated proteins, Merlin and Sav, are involved in targeting the two essential kinases of the Hippo kinase cassette to plasma membrane. It is tempting to speculate that at least some of the other upstream regulators of Hippo signaling may function in a similar manner by promoting the membrane association of the Hippo kinase cassette. We note that a functional role for Sav in membrane association of Hpo does not preclude the other previously described roles for the Sav scaffold in Hippo signaling, such as tethering Hpo and Wts. It is possible that Sav potentiates Hippo signaling both by tethering multiple signaling components and by localizing signaling activity to specific subcellular compartment, as shown in other well studied scaffold signaling proteins such as Ste5 and KSR (Good et al., 2011). Nevertheless, our study has uncovered a previously unrecognized role for Sav in spatial organization of the Hippo pathway.

Wts/Lats is known to be subjected to two modes of regulation, including phosphorylation and protein stability. Our study extends previous studies by showing that its membrane association represents an additional mode of regulation. In addition, our study suggests that Wts may be activated by alternate upstream kinase(s) besides Hpo. Identifying the kinases that mediate Hpo-independent activation and understanding the regulation of such kinases should greatly expand our knowledge about the physiological regulation of Hippo signaling. With its activity subjected to multiple modes of regulation, it is becoming increasingly clear that Wts/Lats represents as a critical node in the Hippo signaling network. These different modes of regulation are not exclusive of each other and indeed functionally intertwined, as membrane association of Wts/Lats also enhances its phosphorylation. Understanding how the multiple regulatory inputs into Wts/Lats are coordinated will shed light on the physiological regulation of Hippo signaling in normal development and offer new strategies for therapeutic intervention in pathological conditions such as Neurofibromatosis type II.

Experimental Procedures

Subcellular fractionation and cell staining

To separate membrane and cytoplasm-associated proteins, subcellular fractionation was carried out as described (Hergovich et al., 2005). Briefly, cells were incubated for 20 minutes on ice in S100/P100 buffer (20 mM Tris, 150 mM NaCl, 2.5 mM EDTA, 1 mM EGTA) supplemented with 1 mM PMSF and protease inhibitors (Roche). Cell lysate was then passed through a 26 G needle 20 times and nuclei were removed by centrifugation at 1,000 × g for 2 min. The supernatant was then ultracentrifuged for 60 min at 100,000 × g at 4°C to separate cytoplasmic fraction (supernatant) and membranous fraction (pellet).

For immunofluorescence staining, cells were grown on Lab-Tek® II Chamber Slide, washed with PBS, and fixed in 3% paraformaldehyde for 10 min at room temperature. Immunofluorescent staining was performed according to the protocols provided by the manufacturers of the commercial antibodies.

Drosophila Genetics

The myristylation signal from Drosophila c-Src (amino acids 1–10) was inserted into the N- terminus to generate Myr-V5-Wts, Myr-V5-WtsT1077A. These fragments were then inserted into the pUAST-attB vector. Transgenic flies were created by phiC31-mediated site-specific transformation, using the attP2 site at 51C (Bischof et al., 2007). GMR-Gal4 and nub-Gal4 were used for overexpression analysis. All crosses were done at 25°C. The following genotypes were used:

  • GFP+hpo− clones:

    • UAS-GFP hs-Flp; tub-Gal80 FRT42D/FRT42hpo42-47; tub-Gal4

  • GFP+hpo− clones overexpressing Myr-Wts:

    • UAS-GFP hs-Flp; tub-Gal80 FRT42D/FRT42hpo42-47; tub-Gal4/UAS- Myr-Wts

  • Control pupal eyes:

    • y w ey-flp Ubi-GFP FRT19A/FRT19A

  • mer mutant pupal eyes:

    • y w ey-flp Ubi-GFP FRT19A/mer4FRT19A

  • sav mutant pupal eyes:

    • y w ey-flp; FRT 82B sav3/FRT 82B Ubi-GFP

  • mer sav double mutant pupal eyes:

    • y w ey-flp, Ubi-GFP FRT19A/mer4FRT19A; FRT82B sav3/FRT82B Ubi-GFP

Mouse genetics

To achieve liver-specific gene deletion, floxed alleles of Nf2 (Giovannini et al., 2000) and Sav1 (Cai et al., 2010) were bred to Albumin-Cre transgenic mice (003574) from the Jackson Laboratory (Bar Harbor, ME). To avoid potential variation related to gender, all experiments were performed in male mice with paternal inheritance of Alb-Cre. Mouse liver tissues were collected and fixed in 10% neutral buffered formalin solution (Sigma), paraffin-embedded, sectioned (at 5 μm), and stained with hematoxylin and eosin. For immunofluorescence, rabbit anti-Lats2 (Bethyl laboratories, 1:100) and mouse anti-pancadherin (Sigma, 1:500) were used, together with Cy3-conjugated goat anti-rabbit (1:250) and Alexa488-conjugated goat anti-mouse secondary antibodies (1:250) (Molecular Probes). Immunohistochemistry staining was performed on paraffin sections using a Vectastain ABC kit according to manufacturer’s instructions (Vector Laboratories, Burlingame, CA), using pan-CK antibody (DAKO, 1:500).

Supplementary Material

01

Highlights.

Merlin promotes Hippo signaling without stimulating the kinase activity of Hpo/Mst

Merlin promotes plasma membrane association of Wts/Lats through direct binding

Membrane targeting promotes Wts phosphorylation and tumor suppressor activity

Distinct roles of Sav and Mer in membrane association of the Hippo kinase cassette

Acknowledgments

We thank Dr. Rick Fehon for the gift of anti-Mer antibody, Dr. Marco Giovannini for providing FC-912 and FH-912 Schwann cells, and Dr. Vijaya Ramesh for advice on GST-NF2 purification. This study was supported in part by grants from the National Institutes of Health (EY015708) and Department of Defense (NF093145). D.P. is an investigator of the Howard Hughes Medical Institute.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Ahronowitz I, Xin W, Kiely R, Sims K, MacCollin M, Nunes FP. Mutational spectrum of the NF2 gene: a meta-analysis of 12 years of research and diagnostic laboratory findings. Hum Mutat. 2007;28:1–12. doi: 10.1002/humu.20393. [DOI] [PubMed] [Google Scholar]
  2. Bischof J, Maeda RK, Hediger M, Karch F, Basler K. An optimized transgenesis system for Drosophila using germ-line-specific phiC31 integrases. Proc Natl Acad Sci U S A. 2007;104:3312–3317. doi: 10.1073/pnas.0611511104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bossuyt W, Chen CL, Chen Q, Sudol M, McNeill H, Pan D, Kopp A, Halder G. An evolutionary shift in the regulation of the Hippo pathway between mice and flies. Oncogene. 2013 doi: 10.1038/onc.2013.82. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bretscher A, Edwards K, Fehon RG. ERM proteins and merlin: integrators at the cell cortex. Nat Rev Mol Cell Biol. 2002;3:586–599. doi: 10.1038/nrm882. [DOI] [PubMed] [Google Scholar]
  5. Cai J, Zhang N, Zheng Y, de Wilde RF, Maitra A, Pan D. The Hippo signaling pathway restricts the oncogenic potential of an intestinal regeneration program. Genes Dev. 2010;24:2383–2388. doi: 10.1101/gad.1978810. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Dupont S, Morsut L, Aragona M, Enzo E, Giulitti S, Cordenonsi M, Zanconato F, Le DJ, Forcato M, Bicciato S, Elvassore N, Piccolo S. Role of YAP/TAZ in mechanotransduction. Nature. 2011;474:179–183. doi: 10.1038/nature10137. [DOI] [PubMed] [Google Scholar]
  7. Fernandez BG, Gaspar P, Bras-Pereira C, Jezowska B, Rebelo SR, Janody F. Actin-Capping Protein and the Hippo pathway regulate F-actin and tissue growth in Drosophila. Development. 2011;138:2337–2346. doi: 10.1242/dev.063545. [DOI] [PubMed] [Google Scholar]
  8. Genevet A, Wehr MC, Brain R, Thompson BJ, Tapon N. Kibra is a regulator of the salvador/warts/hippo signaling network. Dev Cell. 2010;18:300–308. doi: 10.1016/j.devcel.2009.12.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Giovannini M, Robanus-Maandag E, vand V, Niwa-Kawakita M, Abramowski V, Goutebroze L, Woodruff JM, Berns A, Thomas G. Conditional biallelic Nf2 mutation in the mouse promotes manifestations of human neurofibromatosis type 2. Genes Dev. 2000;14:1617–1630. [PMC free article] [PubMed] [Google Scholar]
  10. Good MC, Zalatan JG, Lim WA. Scaffold proteins: hubs for controlling the flow of cellular information. Science. 2011;332:680–686. doi: 10.1126/science.1198701. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Groves JT, Kuriyan J. Molecular mechanisms in signal transduction at the membrane. Nat Struct Mol Biol. 2010;17:659–665. doi: 10.1038/nsmb.1844. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Halder G, Johnson RL. Hippo signaling: growth control and beyond. Development. 2011;138:9–22. doi: 10.1242/dev.045500. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Hamaratoglu F, Willecke M, Kango-Singh M, Nolo R, Hyun E, Tao C, Jafar-Nejad H, Halder G. The tumour-suppressor genes NF2/Merlin and Expanded act through Hippo signalling to regulate cell proliferation and apoptosis. Nat Cell Biol. 2006;8:27–36. doi: 10.1038/ncb1339. [DOI] [PubMed] [Google Scholar]
  14. Harvey K, Tapon N. The Salvador-Warts-Hippo pathway - an emerging tumour-suppressor network. Nat Rev Cancer. 2007;7:182–191. doi: 10.1038/nrc2070. [DOI] [PubMed] [Google Scholar]
  15. Hergovich A, Bichsel SJ, Hemmings BA. Human NDR kinases are rapidly activated by MOB proteins through recruitment to the plasma membrane and phosphorylation. Mol Cell Biol. 2005;25:8259–8272. doi: 10.1128/MCB.25.18.8259-8272.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Ho LL, Wei X, Shimizu T, Lai ZC. Mob as tumor suppressor is activated at the cell membrane to control tissue growth and organ size in Drosophila. Dev Biol. 2010;337:274–283. doi: 10.1016/j.ydbio.2009.10.042. [DOI] [PubMed] [Google Scholar]
  17. Jin H, Sperka T, Herrlich P, Morrison H. Tumorigenic transformation by CPI-17 through inhibition of a merlin phosphatase. Nature. 2006;442:576–579. doi: 10.1038/nature04856. [DOI] [PubMed] [Google Scholar]
  18. Kang BS, Cooper DR, Devedjiev Y, Derewenda U, Derewenda ZS. The structure of the FERM domain of merlin, the neurofibromatosis type 2 gene product. Acta Crystallogr D Biol Crystallogr. 2002;58:381–391. doi: 10.1107/s0907444901021175. [DOI] [PubMed] [Google Scholar]
  19. Kango-Singh M, Nolo R, Tao C, Verstreken P, Hiesinger PR, Bellen HJ, Halder G. Shar-pei mediates cell proliferation arrest during imaginal disc growth in Drosophila. Development. 2002;129:5719–5730. doi: 10.1242/dev.00168. [DOI] [PubMed] [Google Scholar]
  20. LaJeunesse DR, McCartney BM, Fehon RG. Structural analysis of Drosophila merlin reveals functional domains important for growth control and subcellular localization. J Cell Biol. 1998;141:1589–1599. doi: 10.1083/jcb.141.7.1589. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Li W, Cooper J, Karajannis MA, Giancotti FG. Merlin: a tumour suppressor with functions at the cell cortex and in the nucleus. EMBO Rep. 2012;13:204–215. doi: 10.1038/embor.2012.11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Li W, You L, Cooper J, Schiavon G, Pepe-Caprio A, Zhou L, Ishii R, Giovannini M, Hanemann CO, Long SB, Erdjument-Bromage H, Zhou P, Tempst P, Giancotti FG. Merlin/NF2 suppresses tumorigenesis by inhibiting the E3 ubiquitin ligase CRL4(DCAF1) in the nucleus. Cell. 2010;140:477–490. doi: 10.1016/j.cell.2010.01.029. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Lu L, Li Y, Kim SM, Bossuyt W, Liu P, Qiu Q, Wang Y, Halder G, Finegold MJ, Lee JS, Johnson RL. Hippo signaling is a potent in vivo growth and tumor suppressor pathway in the mammalian liver. Proc Natl Acad Sci U S A. 2010;107:1437–1442. doi: 10.1073/pnas.0911427107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Metpally RP, Sowdhamini R. Cross genome phylogenetic analysis of human and Drosophila G protein-coupled receptors: application to functional annotation of orphan receptors. BMC Genomics. 2005;6:106. doi: 10.1186/1471-2164-6-106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Nguyen R, Reczek D, Bretscher A. Hierarchy of merlin and ezrin N- and C-terminal domain interactions in homo- and heterotypic associations and their relationship to binding of scaffolding proteins EBP50 and E3KARP. J Biol Chem. 2001;276:7621–7629. doi: 10.1074/jbc.M006708200. [DOI] [PubMed] [Google Scholar]
  26. Pan D. The hippo signaling pathway in development and cancer. Dev Cell. 2010;19:491–505. doi: 10.1016/j.devcel.2010.09.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Roeder T. Tyramine and octopamine: ruling behavior and metabolism. Annu Rev Entomol. 2005;50:447–477. doi: 10.1146/annurev.ento.50.071803.130404. [DOI] [PubMed] [Google Scholar]
  28. Rouleau GA, Merel P, Lutchman M, Sanson M, Zucman J, Marineau C, Hoang-Xuan K, Demczuk S, Desmaze C, Plougastel B, et al. Alteration in a new gene encoding a putative membrane-organizing protein causes neuro-fibromatosis type 2. Nature. 1993;363:515–521. doi: 10.1038/363515a0. [DOI] [PubMed] [Google Scholar]
  29. Sansores-Garcia L, Bossuyt W, Wada K, Yonemura S, Tao C, Sasaki H, Halder G. Modulating F-actin organization induces organ growth by affecting the Hippo pathway. EMBO J. 2011;30:2325–2335. doi: 10.1038/emboj.2011.157. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Scheel H, Hofmann K. A novel interaction motif, SARAH, connects three classes of tumor suppressor. Curr Biol. 2003;13:R899–R900. doi: 10.1016/j.cub.2003.11.007. [DOI] [PubMed] [Google Scholar]
  31. Shaw RJ, Paez JG, Curto M, Yaktine A, Pruitt WM, Saotome I, O’Bryan JP, Gupta V, Ratner N, Der CJ, Jacks T, McClatchey AI. The Nf2 tumor suppressor, merlin, functions in Rac-dependent signaling. Dev Cell. 2001;1:63–72. doi: 10.1016/s1534-5807(01)00009-0. [DOI] [PubMed] [Google Scholar]
  32. Sher I, Hanemann CO, Karplus PA, Bretscher A. The tumor suppressor merlin controls growth in its open state, and phosphorylation converts it to a less-active more-closed state. Dev Cell. 2012;22:703–705. doi: 10.1016/j.devcel.2012.03.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Struhl G, Adachi A. Nuclear access and action of notch in vivo. Cell. 1998;93:649–660. doi: 10.1016/s0092-8674(00)81193-9. [DOI] [PubMed] [Google Scholar]
  34. Surace EI, Haipek CA, Gutmann DH. Effect of merlin phosphorylation on neurofibromatosis 2 (NF2) gene function. Oncogene. 2004;23:580–587. doi: 10.1038/sj.onc.1207142. [DOI] [PubMed] [Google Scholar]
  35. Trofatter JA, MacCollin MM, Rutter JL, Murrell JR, Duyao MP, Parry DM, Eldridge R, Kley N, Menon AG, Pulaski K, et al. A novel moesin-, ezrin-, radixin-like gene is a candidate for the neurofibromatosis 2 tumor suppressor. Cell. 1993;72:791–800. doi: 10.1016/0092-8674(93)90406-g. [DOI] [PubMed] [Google Scholar]
  36. Wada K, Itoga K, Okano T, Yonemura S, Sasaki H. Hippo pathway regulation by cell morphology and stress fibers. Development. 2011;138:3907–3914. doi: 10.1242/dev.070987. [DOI] [PubMed] [Google Scholar]
  37. Wu S, Huang J, Dong J, Pan D. hippo encodes a Ste-20 family protein kinase that restricts cell proliferation and promotes apoptosis in conjunction with salvador and warts. Cell. 2003;114:445–456. doi: 10.1016/s0092-8674(03)00549-x. [DOI] [PubMed] [Google Scholar]
  38. Xiao L, Chen Y, Ji M, Dong J. KIBRA regulates Hippo signaling activity via interactions with large tumor suppressor kinases. J Biol Chem. 2011;286:7788–7796. doi: 10.1074/jbc.M110.173468. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Yu FX, Zhang Y, Park HW, Jewell JL, Chen Q, Deng Y, Pan D, Taylor SS, Lai ZC, Guan KL. Protein kinase A activates the Hippo pathway to modulate cell proliferation and differentiation. Genes Dev. 2013;27:1223–1232. doi: 10.1101/gad.219402.113. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Yu FX, Zhao B, Panupinthu N, Jewell JL, Lian I, Wang LH, Zhao J, Yuan H, Tumaneng K, Li H, Fu XD, Mills GB, Guan KL. Regulation of the Hippo-YAP Pathway by G-Protein-Coupled Receptor Signaling. Cell. 2012;150:780–791. doi: 10.1016/j.cell.2012.06.037. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Yu J, Zheng Y, Dong J, Klusza S, Deng WM, Pan D. Kibra functions as a tumor suppressor protein that regulates hippo signaling in conjunction with Merlin and expanded. Dev Cell. 2010;18:288–299. doi: 10.1016/j.devcel.2009.12.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Zhang N, Bai H, David KK, Dong J, Zheng Y, Cai J, Giovannini M, Liu P, Anders RA, Pan D. The Merlin/NF2 Tumor Suppressor Functions through the YAP Oncoprotein to Regulate Tissue Homeostasis in Mammals. Dev Cell. 2010;19:27–38. doi: 10.1016/j.devcel.2010.06.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Zhao B, Li L, Wang L, Wang CY, Yu J, Guan KL. Cell detachment activates the Hippo pathway via cytoskeleton reorganization to induce anoikis. Genes Dev. 2012;26:54–68. doi: 10.1101/gad.173435.111. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Zhou D, Conrad C, Xia F, Park JS, Payer B, Yin Y, Lauwers GY, Thasler W, Lee JT, Avruch J, Bardeesy N. Mst1 and Mst2 maintain hepatocyte quiescence and suppress hepatocellular carcinoma development through inactivation of the Yap1 oncogene. Cancer Cell. 2009;16:425–438. doi: 10.1016/j.ccr.2009.09.026. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

01
NIHMS526293-supplement-01.pdf (19.1MB, pdf)

RESOURCES