Several studies have also suggested that Wnt
Several studies have also suggested that Wnt signalling can in turn stimulate the transcriptional activity of Yap/Taz (Figure 2, Figure 3). In the Azzolin model, Yap/Taz not only block β-catenin signalling, but their association with Axin1 and/or β-catenin also maintains them in an inactivate state (Figure 2b) [21•, 31], with Yap/Taz inhibition partly dependent on Pkcζ recruitment, which directly phosphorylates and destabilizes β-catenin and Yap . Wnt-dependent disruption of the β-catenin destruction complex is then proposed to activate Yap/Taz. In a second model, Cai et al. suggested that the tumour suppressor Apc interacts with Lats and Sav1 to promote Lats-mediated phosphorylation and inactivation of Yap (Figure 2c) . Moreover, they propose the Hippo scaf folding properties of Apc are disrupted by GSK-3β inhibition but not Wnt stimulation, as proposed by Azzolin et al. Finally, Park et al. suggested an alternative mechanism whereby Wnts trigger a non-canonical FZD/ROR-Gα12/13-RhoGTPases-Lats1/2 pathway that activates Yap/Taz (Figure 2d) .
By fully integrating Hippo into the Wnt pathway, these models certainly provide a simple explanation for how Yap/Taz might be activated within gut epithelium. In agreement, Wnt-driven tumours typically show high levels of nuclear Yap. However, in the homeostatic gut neither Yap or Taz are required for Wnt signalling [10••, 21•], suggesting the pathway is not a core Wnt pathway component. Furthermore, as many extrinsic cues regulate Hippo, one cannot exclude that Yap activation in tumours is due to other signals. Indeed, using validated Yap antibodies, it is clear that Yap localization is largely unaffected by acute 2870 of Apc in the gut epithelium [10••, 33]. Accordingly, Yap activation during adenoma initiation likely depends on other signals independent of Wnt activity (see below). Clearly further studies are warranted to resolve these questions.
Regulation of Hippo signalling in the gut Our understanding of downstream events triggered by Yap/Taz in intestinal regeneration and cancer initiation is expanding, but our understanding of what regulates Hippo, particularly in vivo, remains limited. This question is a challenge to answer, as Hippo is not regulated like a classical ligand-receptor pathways (e.g. Wnt, Tgfβ, Notch, etc.). Rather, the opposite is true, Hippo activity depends on a panoply of input signals including mechanotransduction, G protein-coupled receptors (GPCR), metabolites, receptor tyrosine kinases and more . Below we discuss how some of these may regulate Hippo during gut regeneration and tumorigenesis.
Immune signalling and the Hippo pathway Inflammation and tumourigenesis are linked and may provide a mechanism whereby cytokine signalling is coupled to Yap (Figure 3). To support this notion, Taniguchi et al. showed that overexpression of activated gp130, a common coreceptor of the Il-6 family of inflammatory cytokines, stimulated Yap-dependent crypt hyperproliferation . Furthermore, Il-6 triggered Yap activation via Src family kinases independently of Stat3 and a recent follow up study showed that Yap in turn stimulates expression of Il6ST, the gene encoding gp130 in Apc mutant cells [35••, 36]. Another key element as yet unexplored, is the impact of pathogens and their metabolites on inflammation and Yap-driven regeneneration and tumorigenesis. As a central driver of gut inflammation determining how the microbiome might intersect with Hippo signalling should prove interesting [37, 38]. Indeed, in Drosophila fat bodies stimulation of the innate immune receptor, Toll, activates Hippo signalling, resulting in inactivation of Yki, the fly Yap/Taz ortholog . In turn, activation of Yki stimulates Cactus (or IκB) and vulnerability to Gram-positive bacterial infection. If conserved in mammals, this study would clearly open the door to a whole new facet of Hippo regulation.