PerspectiveCell Biology

The Hippo Size Control Pathway—Ever Expanding

See allHide authors and affiliations

Science Signaling  22 Jan 2013:
Vol. 6, Issue 259, pp. pe4
DOI: 10.1126/scisignal.2003813


An important regulator of organ size and tumorigenesis is the Hippo pathway. Recent studies have unveiled increasing complexity in regulation of Hippo pathway activity at the level of the oncoprotein Yes-associated protein (YAP). The protein tyrosine phosphatase 14 (PTPN14, known as Pez in Drosophila) was identified as a protein that antagonizes the function of the key Hippo pathway protein YAP by promoting its cytoplasmic localization under high cell density conditions. In Drosophila, Pez was identified as a repressor of epithelial proliferation in vivo. Studies in mammalian cells showed that a family of G protein–coupled receptors, the protease-activated receptors, functioned as activators of YAP. These studies shed light on the intricate regulation of the Hippo pathway and also highlight the importance of investigating these newly discovered regulatory links in physiological and pathological settings to fully appreciate their importance.

Despite being recognized as a fundamental principle of biology, the mechanism by which the size of organs is specified is still poorly understood. An important inroad into tackling this challenging problem was the discovery, in Drosophila 10 years ago, of the Hippo pathway (1), an evolutionarily conserved regulator of organ size (24). A substantial increase in knowledge of Hippo pathway signal transduction has been enabled primarily by genetic and proteomic screens, with at least 35 Hippo pathway proteins identified in both the Drosophila and mammalian Hippo pathways (5). Hippo pathway proteins fall into three main classes: (i) the core kinase cassette; (ii) the downstream transcriptional regulators; and (iii) the upstream branches, many of which also control cellular processes, such as apicobasal cell polarity, planar cell polarity, and cell-cell adhesion (5). These upstream branches converge on the transcriptional regulators Yes-associated protein (YAP) and TAZ, which are oncoproteins of the pathway. The activity of YAP and TAZ is inhibited by the kinase cassette, consisting of MST1 and -2 (MST1/2) and the downstream kinases LATS1 and -2 (LATS1/2). LATS1/2 phosphorylation of YAP and TAZ limits their nuclear localization (6). The identification of additional regulators of YAP localization, protein tyrosine phosphatase 14 (PTPN14, known as Pez in Drosophila) (710) and a family of G protein–coupled receptors (GPCRs), the protease-activated receptors (PARs) (11), has uncovered further complexity in the Hippo pathway.

PTPN14: Adding Complexity Through WW Domain–Mediated Interactions

WW domains play key roles in Hippo pathway signal transduction, and this domain and proteins containing the sequence motif (PPxY or PY motif) that it recognizes are enriched in this signaling network (12). Four independent groups reported that PTPN14, a protein with two PY motifs, participates in regulation of Hippo signaling through a WW domain–mediated interaction. Using proteomic approaches with different rationales, the WW domain–containing proteins Kibra and YAP were both found to physically interact with PTPN14 in a manner dependent on a WW domain–PY motif interaction (710). In cultured human cells, the Hippo pathway represses YAP activity in response to increased cell density (13). Consistent with this, the ability of PTPN14 to repress YAP correlated with cell density: In confluent, high-density cell culture, the abundance of PTPN14 was increased and PTPN14 was required to inhibit YAP (810). The relevance of this phenomenon in vivo in mammals is unclear.

Currently, support for an in vivo role for PTPN14 as a Hippo pathway regulator comes from studies of Pez in Drosophila (7). Poernbacher et al. found that Pez represses Yorkie (Yki, the fly homolog of YAP and TAZ) activity, primarily in the adult gut epithelium. Pez mutant flies displayed excessive Yki activity in the adult midgut, which resulted in stem cell expansion and hyperplasia. yki hemizygosity completely suppressed the gut defects in Pez mutant flies, indicating that repression of Yki is a major function for Pez with respect to organism viability and size. Loss of Pez in larval epithelial tissues did not cause obvious overgrowth, suggesting functional redundancy—possibly with one or more of the several upstream Hippo pathway proteins—at the larval stage of development. However, Pez must play some essential functions during Drosophila development because, in addition to the gut overgrowth phenotype in the adults, these flies exhibited a developmental delay and growth retardation after the larval stage, which may been due to starvation caused by the gut hyperplasia (7).

Another unanswered question highlighted by these studies is the mechanism by which PTPN14 inhibits YAP activity. On the basis of experiments using phosphatase-deficient versions of PTPN14 and Pez, the consensus view is that PTPN14 limits YAP activity and Pez limits Yki activity in phosphatase-independent fashions (710), suggesting that the WW domain–mediated interactions are key to the regulation. Liu et al. and Wang et al. showed that at least in cultured cells, PTPN14 limits access of YAP to the nucleus, where it regulates transcription (8, 9). What is unclear is whether PTPN14 blocks nuclear entry, enhances nuclear exit, or acts as a cytoplasmic “sink” for YAP. PTPN14 might also repress YAP by promoting Kibra activity, which promotes Hippo pathway signaling by activating the core kinase cassette (1416) (Fig. 1). Experiments with flies (7) and mammals (8) indicate that Pez and PTPN14 interact with Kibra and that PTPN14 enhances phosphorylation of YAP at a site that is targeted by LATS1/2 (9). PTPN14 could promote the assembly of a complex containing the Hippo pathway core kinase cassette and Kibra and thus facilitate LATS1/2 activity. PTPN14 could also inhibit YAP by competing with positive regulators of YAP, such as WBP2, that interact with the YAP WW domains (17, 18). The studies of PTPN14 and Pez add further complexity to WW domain–dependent regulation of YAP because in addition to positive regulators, the YAP WW domains also mediate interactions with negative regulators, such as LATS1/2 and the Angiomotin protein family (19).

Fig. 1

PTPN14 and PARs regulate YAP localization and activity. (A) Cells at low density have active YAP and TAZ. (B) Cells at high density have active Hippo signaling that inhibits YAP and TAZ transcriptional activity. Under conditions of high cell density, the abundance of PTPN14 increases, resulting in direct inhibitory binding with and maintenance of YAP in the cytosol. PTPN14 itself is targeted for degradation by an E3 ubiquitin ligase, LRR1, and this also occurs in a cell density–dependent fashion. (C) In response to ligands, PAR promotes YAP nuclear localization and activity through a pathway involving the actin cytoskeleton. PAR, protease-activated receptor; TFs, transcription factors.


Another interesting question arising from the studies of Wang et al. (8) pertains to the regulation of PTPN14. In cultured cells, the abundance of PTPN14 was regulated by the ubiquitin ligase LRR1; PTPN14 abundance increased in cultures with a high cell density, implying that LRR1 activity was suppressed under these conditions (8). What then regulates LRR1 activity or its ability to target PTPN14, and does LRR1 regulate PTPN14 abundance in vivo? Understanding the regulation of the Hippo pathway is important because this pathway is a key regulator of tissue growth during development and homeostasis, and loss of appropriate regulation can contribute to cancer (2).

PARs: Adding Complexity Through GPCR Signaling

Various studies in Drosophila (20, 21) and mammalian cells (2224) have shown that changes in the actin cytoskeleton regulate YAP phosphorylation and nuclear localization, but whether this pathway involves the core kinases of the Hippo pathway—MST1/2 and LATS1/2—has been unclear. Two reports have connected GPCRs, acting through the actin cytoskeleton, to the regulation of YAP (11, 25). Mo et al. (11) defined a link between the Hippo pathway kinases LATS1/2 and thrombin, an extracellular serine protease that activates PARs (11). In cell culture studies, PAR1 activated YAP and TAZ through the G12/13 family of G proteins and the Rho guanosine triphosphatase (GTPase), which inhibited LATS1/2, reducing YAP and TAZ phosphorylation and enabling their accumulation in the nucleus. The reduction in YAP phosphorylation relied on Rho GTPase–mediated changes in the actin cytoskeleton to inhibit LATS1 (11). Rho-mediated coordination of YAP phosphorylation and localization did not require its downstream effector the kinase ROCK [also known as ROK] (11), a finding that contrasts with the requirement for ROCK in regulation of YAP downstream of changes in cell density and mechanical forces (22, 24). YAP localization is predominantly nuclear in situations in which cells spread and stress fibers (composed of F actin) become abundant, such as when cells are plated in low density (22) or when cells are in high-tensile states (24). In both of these scenarios, inhibition of ROCK resulted in decreased YAP nuclear localization, most likely because of a reduction in stress fiber formation.

Furthermore, LATS and ROCK signaling also participate in cytoskeletal rearrangements related to cytokinesis. Overexpression of the ROCK downstream effector kinase LIMK results in polynucleate cells suggestive of defective cytokinesis, most likely because of enhanced actin polymerization; coexpression of LATS1 and LIMK results in a suppression of LIMK-mediated cytokinesis defects (26), suggesting that LATS limits LIMK in this process. Whether LIMK influences YAP function in response to PAR1 is undetermined; perhaps Rho, LATS1, and LIMK coordinate or balance signals from the actin cytoskeleton to regulate YAP localization and function. The components and their hierarchy in the signaling pathway between the actin cytoskeleton and core elements of the Hippo pathway in the regulation of YAP activity remain to be elucidated.

Although both MST1/2 and LATS1/2 are thought to function as a kinase cassette in the control of tissue growth, MST2 activity was not affected by the PAR1 agonist thrombin receptor activator peptide 6 (TRAP6) and was not involved in the PAR1-dependent changes in YAP phosphorylation. This suggests that PAR1-mediated regulation of YAP bypasses MST1/2 and instead inhibits LATS1/2 through a different mechanism. Also in studies of cultured cells, Yu et al. found regulation of YAP downstream of G12/13-, Gq/11-, Gs-, and Gi/o-coupled GPCRs through a mechanism that was independent of MST2 (25).

Although the MST1/2 part of the kinase cascade may not be relevant for GPCR-mediated regulation of YAP in response to changes in the actin cytoskeleton, MST1/2 may function in mediating other responses to changes in the actin cytoskeleton. MST2 stimulates the c-Jun N-terminal kinase (JNK) pathway in response to changes in actin cytoskeleton integrity, leading to a stabilization of the cell cycle inhibitor p21 (27). Whether this JNK regulatory pathway is mediated by the canonical MST1/2 to LATS1/2 to YAP and TAZ module is unknown. However, it appears that MST1/2 and LATS1/2 might distinguish distinct inputs from the actin cytoskeleton to promote different cellular responses.

Although only PARs and the G proteins G12/13, Gq/11, Gs, and Gi/o have been so far implicated in regulation of YAP, an important but complicated task will be to define which ligand-GPCR pairs activate YAP and TAZ and in what context. This is relevant for normal physiology as well as pathological settings such as cancer. Given that somatic mutations in genes that encode core components of the Hippo pathway are relatively rare, other genetic or epigenetic events must drive activation of YAP that is observed at a high frequency in different human cancers. On the basis of the study of Mo et al., one such mechanism for pathological YAP and TAZ activation is deregulated signaling through thrombin and PARs. For example, do hepatocellular carcinomas that have demonstrated PAR-mediated cell migration (28) also have increased YAP activity, and if so, are they dependent on YAP for growth? Do cancers that display increased YAP activity overlap (2931) with those that have increased thrombin or PAR abundance?

The featured studies have served to reveal further complexity in the wiring of the Hippo signaling network. Major challenges for the future include deciphering the precise in vivo roles of PARs and PTPN14 with respect to Hippo pathway regulation in mammals. Such studies have the potential to inform the mechanism by which the Hippo pathway controls organ size and cell fate, as well as its role in tumorigenesis.


View Abstract

Navigate This Article