Wnt/β-catenin signaling is activated repeatedly during an animal’s lifespan, and it controls gene expression through its essential nuclear effector, β-catenin, to regulate embryogenesis, organogenesis, and adult homeostasis. Although the β-catenin transcriptional complex has the ability to induce the expression of many genes to exert its diverse roles, it chooses and transactivates a specific gene set from among its numerous target genes depending on the context. For example, the β-catenin transcriptional complex stimulates the expression of cell cycle-related genes and consequent cell proliferation in neural progenitor cells, while it promotes the expression of neural differentiation-related genes in differentiating neurons. Recent studies using animal and cell culture models have gradually improved our understanding of the molecular basis underlying such context-dependent actions of the β-catenin transcriptional complex. Here, we describe eight mechanisms that support β-catenin-mediated context-dependent gene regulation, and their spatio-temporal regulation during vertebrate development. In addition, we discuss their contribution to the diverse functions of Wnt/β-catenin signaling.
Diverse Functions of Wnt/β-Catenin Signaling
The core Wnt/β-catenin signaling system. In unstimulated conditions, levels of cytoplasmic β-catenin are kept low by a destruction complex. Binding of Wnt to the receptors promotes dissociation of the β-catenin destruction complex and consequently stabilizes cytoplasmic β-catenin. As a result, β-catenin accumulates and enters the nucleus, where it forms complexes with Tcf/Lef that activate gene expression. Ub: ubiquitin, P: phosphorylation. β-cat: β-catenin.
Multiple roles of Wnt/β-catenin signaling during vertebrate brain development. During vertebrate brain development, Wnt/β-catenin signaling plays multiple roles, including head organizer formation, patterning of anterior–posterior (A–P) and dorsal–ventral (D–V) axes, size expansion and morphogenesis of each brain region, promotion of neural progenitor cell (NPC) proliferation, and neuronal differentiation. A: anterior, P: posterior, D: dorsal, V: ventral.
Cell proliferation and differentiation are inversely correlated processes. Therefore, it is interesting that the “apparently simple” Wnt/β-catenin signaling system promotes both proliferation and differentiation. Furthermore, such context-dependent multiple functions of Wnt/β-catenin signaling are observed not only in brain but also in many other tissues, including the digestive and blood systems (21–23), and support their development and maintenance. However, the mechanisms that enable the Wnt/β-catenin signaling system to exert its multiple functions are not well-understood. Here, we discuss how Wnt/β-catenin signaling displays context-dependent diverse functions during animal development, with a particular focus on vertebrate development.
The Target Gene Set of Wnt/β-Catenin Signaling Differs Depending on Cell Context
Wnt/β-catenin signaling must have an ability to induce the expression of numerous genes to control the formation of a variety of tissues. This signaling appears to selectively activate specific gene sets depending on the situation. Consistent with this concept, Railo et al. reported that stimulation with a Wnt family ligand, Wnt-3a, induced significant changes in the expression levels of 355 genes in the mouse fibroblast cell line NIH3T3, while it induced 129 genes in the rat pheochromocytoma PC12 cell line. Only two Wnt-target genes were shared by both cell lines (24). Qu et al. showed that the genomic regions to which β-catenin was recruited differed between proliferating NPCs and differentiating neurons in adult mouse brain. In NPCs, β-catenin was bound to the promoter region of the cyclin D1 gene but not to that of neurogenin 2, and activated the expression of cyclin D1. By contrast, β-catenin was recruited to the neurogenin 2 promoter but not to the cyclin D1 promoter and activated neurogenin 2 expression (20), suggesting that the target gene of Wnt/β-catenin signaling is switched in response to a change from proliferation to differentiation. Although proliferation and differentiation would not occur simultaneously, such selective target gene activation and switching must enable the Wnt/β-catenin signaling system to promote both proliferation and differentiation. Therefore, to understand the molecular basis underlying the multi-functionality of Wnt/β-catenin signaling, it is important to clarify two questions. One is how Wnt/β-catenin signaling induces the expression of a context-specific gene set. The other is how the Wnt/β-catenin signaling target gene set is “switched” in response to a change of cell context.
How Does Wnt/β-Catenin Signaling Induce the Expression of a Specific Gene Set?
The control of Wnt/β-catenin signaling-mediated context-specific gene expression must be supported by mechanisms that recruit the β-catenin transcriptional complex to the specific gene promoters where it is activated. There are eight factors that can affect such β-catenin recruitment and activity: (i) distinct activities between Tcf/Lef members; (ii) distinct activities between Tcf/Lef splicing variants; (iii) methylation of the target gene promoter; (iv) histone modifiers; (v) cooperation with other transcription factors; (vi) Tcf/Lef-independent β-catenin-recruitment; (vii) long noncoding RNA (lncRNA)-mediated recruitment; and (viii) strength of Wnt/β-catenin signaling. Each of these factors is discussed in detail below.
Distinct activities between Tcf/Lef members
Eight factors that affect activation and recruitment of the β-catenin transcription complex to specific gene promoters. (A) Distinct activities between Tcf/Lef members. (B) Distinct activities between Tcf/Lef splicing variants. (C) Methylation of the target gene promoter. (D) Histone modifiers. (E) Cooperation with other transcription factors. (F) Tcf/Lef-independent β-catenin-recruitment. (G) lncRNA-mediated recruitment. (H) Strength of Wnt/β-catenin signaling. Ac: acetylation, MHB: midbrain–hindbrain border.
Distinct activities between Tcf/Lef splicing variants
Numerous splicing variants of Tcf/Lef exist and display differential binding affinity to DNA, β-catenin, and other co-factors. For example, Tcf7E and Tcf7L2E, which are long variants of Tcf7 and Tcf7L2, respectively, own a second DNA binding domain called the C clamp that binds to GC-rich sequence helper sites (5, 28). Tcf7E and other Tcf7 isoforms lacking the C clamp bind to distinct genomic regions (Fig. 3B) (27). On the other hand, the amino terminal β-catenin binding region-lacking variants of Lef1 and Tcf7 (ΔN-Lef1 and ΔN-Tcf7) can prevent β-catenin-mediated gene expression (Fig. 3B) (29–31). The Tcf/Lef variant ratio can be also involved in context-dependent Wnt/β-catenin-mediated gene expression.
Methylation of the target gene promoter
Methylation of the target gene promoter may prevent Wnt/β-catenin-mediated gene expression. The promoter of a Wnt target gene ZBP1 is methylated in some breast cancer cells. This methylation prevents the binding of β-catenin to the promoter and, thus, the Wnt/β-catenin signaling-mediated activation of ZBP1 transcription (Fig. 3C) (32). However, the importance of DNA methylation in regulating Wnt target genes in physiological conditions remains unclear.
Histone modifiers
Enzymes catalysing acetylation, deacetylation, and methylation play important roles in context-specific β-catenin–Tcf/Lef-mediated gene expression. CREB-binding protein (CBP) and p300 are histone acetyltransferases. Both bind to β-catenin to activate β-catenin-mediated transcription, although the binding of each seems to result in distinct effects (33). For example, treatment with ICG-001, a chemical inhibitor that blocks β-catenin binding to CBP but not to p300, promotes differentiation of rat PC12-related cells (34) and C2C12 myoblasts (35). In contrast, treatment with the small molecule IQ-1, which promotes β-catenin binding to CBP and inhibits its binding to p300, inhibits the differentiation of C2C12 myoblasts, alveolar epithelial cells (AECs), and tracheal epithelial cells (35). Based on these findings, it is thought that the CBP- and p300–β-catenin complexes positively regulate the expression of genes promoting cell proliferation and differentiation, respectively (Fig. 3D, left) (33). However, histone deacetylase 1 (HDAC1) binds to Tcf/Lef and deacetylates histones proximal to the nucleosome to suppress β-catenin–Tcf/Lef-mediated transcription (Fig. 3D, right) (36). We, and others, reported that inhibition of HDAC1 enhanced β-catenin–Tcf/Lef-mediated transcription of cell proliferation-promoting genes in zebrafish midbrain and retinal NPCs (37, 38). This suggests that HDAC1 negatively regulates NPC proliferation through the inhibition of β-catenin–Tcf/Lef.
The polycomb group (PcG) consists of PRC1 and PRC2 complexes that possess histone H2A-ubiquitinating and histone H3 Lys-27(H3K27)-methylating activities, respectively. The PcG negatively regulates gene expression via its histone modification activity and modifies Wnt/β-catenin-mediated regulation of the cell fate in brain (39, 40). NPCs in the mouse embryonic cerebral cortex have the potential to differentiate into both neurons and astrocytes. When NPCs differentiate to neurons, Wnt/β-catenin signaling induces the expression of neurogenin 1 to promote their differentiation (41). However, when NPCs differentiate to astrocytes, the PcG blocks Wnt/β-catenin signaling-mediated neurogenin 1 expression through methylation of H3K27 on the neurogenin 1 promoter (42).
Cooperation with other transcription factors
The β-catenin–Tcf/Lef complex can be recruited to specific genomic regions by other transcription factors. For example, the transcription factor Cdx1 is essential for normal anterior–posterior vertebral patterning. Cdx1 physically interacts with Lef1 to recruit the β-catenin–Lef1 complex to the Cdx1 promoter. The β-catenin–Lef1–Cdx1 complex then activates the expression of Cdx1 to control vertebral patterning (Fig. 3E) (43). Sp1-like transcription factors (Sp5 and Sp8) also directly interact with Tcf/Lef to recruit the β-catenin–Lef1 complex to the promoter/enhancer regions of selective Wnt target genes, including Sp5 and brachyury (44). Sp5 and Sp8 are expressed only in restricted regions of the mouse embryo (45, 46), suggesting that they may support context-specific Wnt target gene selection.
Tcf/Lef-independent β-catenin recruitment
Using a systems biology approach, Schuijers et al. (47) showed that Wnt/β-catenin signaling-induced transcriptional activation is mediated exclusively by Tcf/Lef in mouse small intestinal epithelium and human embryonic kidney (HEK) 293 cells. However, rarely, β-catenin regulates gene expression through other transcription factors, but not Tcf/Lef. For example, β-catenin binds to and activates the transcription factor Oct4 in a Tcf/Lef-independent manner to maintain the pluripotent state of mouse embryonic stem cells (Fig. 3F) (48). In addition, the transcription factors Sox17, FOXO, HIF-1, and SATB1 also interact with and recruit β-catenin to their genomic binding sites to promote expression of their target genes (49–52).
lncRNA-mediated recruitment
Recently, Giakountis et al. reported that a lncRNA, WiNTRLINC1, can recruit the β-catenin–Tcf/Lef complex to a specific genomic region. WiNTRLINC1 is a direct target of the β-catenin–Tcf7L2 complex, and it interacts with and recruits β-catenin-Tcf7L2 to the enhancer regions of the Ascl2 gene that codes for a transcription factor controlling the fate of intestinal stem cells. This recruitment enforces β-catenin-Tcf7L2-mediated Ascl2 expression (Fig. 3G) (53). Although WiNTRLINC1 is the only β-catenin-recruiting lncRNA that has been reported, similar lncRNA-mediated β-catenin recruitment might occur in other contexts.
Strength of wnt/β-catenin signaling
Wnt/β-catenin signaling is thought to induce different cell fates through Wnt concentration gradients during formation of the embryonic axis (12, 54). For example, Kiecker and Niehrs used Xenopus animal cap assays to show that strong and mild activation of Wnt/β-catenin signaling induced the hindbrain marker Krox20 and the midbrain–hindbrain border maker En1, respectively, while inhibition of Wnt signaling promoted expression of the fore- and midbrain marker, Otx2 (Fig. 3H) (12). In addition, signaling strength affected blood cell fate decisions in mice. Specifically, mild (low) levels of Wnt/β-catenin signaling activation promoted, while intermediate level activation inhibited, hematopoietic stem cell self-renewal and stimulated myeloid development. Intermediate-high level activation enhanced early T cell differentiation, whereas very high level activation resulted in impaired T cell differentiation (23, 55). Thus, signaling strength varied the outcomes of Wnt/β-catenin signaling. However, the mechanisms by which differences in signaling strength modify the functions of Wnt/β-catenin signaling remain unclear.
How Is the Switching of a Wnt/β-Catenin Signal-Inducing Gene Set in Response to a Change of Context Regulated?
In several progenitor cells, Wnt/β-catenin signaling changes its roles depending on the situation. For example, Wnt/β-catenin signaling initially stimulates the proliferation of NPCs and subsequently promotes their differentiation (Fig. 2). In intestine, Wnt/β-catenin signaling maintains stem cells while it promotes Paneth cell differentiation (22). Therefore, Wnt/β-catenin signaling appears to switch its target gene set in response to changes in the cell context, and such switching may be essential for tissue morphogenesis. However, the molecular mechanisms underlying Wnt/β-catenin target gene set switching have yet to be well-characterized.
Based on the above discussion, we consider that selection of the Wnt/β-catenin target gene set can be influenced by the expression and activity levels of each Tcf/Lef member and their variants (Fig. 3A and B), β-catenin–Tcf/Lef-cooperating transcription factors (Fig. 3E), β-catenin-binding non-Tcf/Lef transcription factors (Fig. 3F), β-catenin–Tcf/Lef-recruiting noncoding RNA (Fig. 3G), methylation levels of the target gene promoter (Fig. 3C), the extent of interaction between β-catenin–Tcf/Lef and histone modifiers (Fig. 3D), and the strength of Wnt/β-catenin signaling (Fig. 3H). Context-specific and spatio-temporal regulation of these pathways would enable appropriate Wnt/β-catenin target gene set switching.
Context-specific Wnt/β-catenin signaling modulators that regulate histone modifiers and their possible roles in Wnt/β-catenin target gene set switching. (A) Histone acetylation (Ac) and histone H3 Lys-4 tri-methylation (Me) are regulated on the β-catenin–Tcf/Lef binding genomic region during the transition from NPC proliferation to differentiation. CycD1: Cyclin D1, Ngn2: neurogenin 2. (B) In proliferating NPCs, NLK phosphorylates Lef1 thereby preventing HDAC1-mediated Lef1 inhibition and allowing NPCs to proliferate. In contrast, HDAC1 appears to positively regulate neuronal differentiation and inhibit cell proliferation in differentiating neurons. (C) PKC-mediated p300 phosphorylation promotes the association of p300 and the dissociation of CBP with β-catenin during AEC differentiation. (D) Pygo2 binds to the β-catenin–Tcf/Lef complex and recruits histone H3K4 methyltransferase (HMT) to facilitate H3K4 methylation and consequent expression of cell proliferation-related genes in mammary progenitor cells.
Recently, we and other researchers showed that regulation of HDAC1 recruitment is involved in zebrafish brain neurogenesis. HDAC1 negatively regulates the β-catenin-Lef1-mediated transcription of cell proliferation-promoting genes in zebrafish (37, 38). A serine/threonine kinase, Nemo-like kinase (NLK), blocks the association of HDAC1 with Lef1 through direct phosphorylation of Lef1 to accelerate Lef1-mediated gene expression and consequent NPC proliferation (Fig. 4B) (38). This indicates that HDAC1 is inactivated in proliferating NPCs. However, interestingly, Harrison et al. (57) showed that HDAC1 activity was required for neuronal differentiation in zebrafish brain, suggesting that HDAC1 was active during neuronal differentiation. Possibly, HDAC1 might be activated in differentiating neurons (in this situation, NLK might be inactivated) and inhibit the proliferation-related Wnt-target genes specifically. Such inhibition might shift the Wnt/β-catenin target gene set from a proliferation- to a differentiation-promoting gene set (Fig. 4B). This idea is consistent with the above-mentioned previous findings that acetylated histone H3 is not detected in the promoter of cyclin D1 in differentiating adult mouse brain neurons (20). Thus, control of HDAC1 recruitment may play important roles in Wnt/β-catenin target gene switching during neurogenesis.
As described above, CBP and p300 promote β-catenin-mediated cell proliferation and differentiation, respectively (33). Therefore, the ratio of CBP to p300 in the β-catenin transcriptional complex would affect the β-catenin-mediated cell fate decision. Recently, Rieger et al. (35) reported that the interaction of CBP and p300 with β-catenin is changed dynamically during AEC culture. In undifferentiated AECs, β-catenin binds to both CBP and p300. However, in differentiating AECs, protein kinase C (PKC) is activated and phosphorylates p300 to promote formation of the p300–β-catenin complex and consequent dissociation of the CBP–β-catenin complex. This β-catenin cofactor switching from CBP to p300 is required for AEC differentiation (Fig. 4C) (35). Thus, PKC-mediated p300 phosphorylation appears to contribute to Wnt/β-catenin-mediated cell differentiation in AEC cultures. However, it remains unclear whether PKC is also involved in Wnt/β-catenin target gene switching in vivo and in other type cells.
Not only histone acetylation, but also H3K4 tri-methylation is involved in determining the fate of progenitor cells. In mouse mammary progenitor cells, the Pygopus 2 (Pygo2) protein binds to the β-catenin–Tcf/Lef complex and recruits histone H3K4 methyltransferase to facilitate histone H3K4 methylation and consequent expression of cell proliferation-related genes. This results in an expansive self-renewal of mammary progenitor cells (Fig. 4D) (58, 59). Interestingly, Pygo2 knockout, as well as the double knockout of Pygo2 and the related Pygo1 gene, reduce the activity of the β-catenin–Tcf/Lef reporter BAT-gal and cell proliferation in several tissues, including mammary tissue. These mice survive to birth (60), suggesting that Pygo2 is not a core component of Wnt/β-catenin signaling and participates in context-specific regulation. Thus, context-specific modifiers that affect the recruitment of histone modifiers to β-catenin–Tcf/Lef play crucial roles in the switching of the Wnt/β-catenin target gene set.
As mentioned above, the strength of Wnt/β-catenin signaling can affect selection of the Wnt/β-catenin target gene set (Fig. 3H). We and other researchers have discovered several proteins that modulate the Wnt/β-catenin signaling strength, including dimerization-partner 1 (DP1), homeodomain-interacting protein kinase 2 (Hipk2), and Notch-regulated ankyrin-repeat protein (Nrarp). These modulators are non-universal regulators that alter the Wnt/β-catenin signaling strength in specific cell-types/tissues. For example, DP1 enhances the nuclear translocation and DNA-binding of β-catenin, and consequent β-catenin–Tcf/Lef-mediated transcription of posterior neural markers, such as Krox20, in the Xenopus posterior neural plate (61). Hipk2 potentiates Wnt/β-catenin signaling activity through multiple mechanisms. Hikasa et al. (62) reported that Hipk2 phosphorylates Tcf7L1 and relieves Tcf7L1-mediated repression of vent2 and cdx4 genes, which are required for posterior body formation in Xenopus embryos. We recently reported that Hipk2 promotes protein phosphatase 1c-mediated dephosphorylation and consequent stabilization of Dishevelled protein, which acts upstream of β-catenin, and boosts the Wnt/β-catenin signaling activity and β-catenin-mediated posterior marker gene expression in zebrafish embryos (63). Nrarp reinforces Lef1-mediated transcription by blocking Lef1 ubiquitination and degradation in neural crest and endothelial cells, and contributes to neural crest differentiation and angiogenesis in zebrafish and mouse (64, 65). Such modulator-mediated spatio-temporal control of the signaling strength may contribute to the cell-type- or tissue-specific target gene expression, and consequently support diverse roles of Wnt/β-catenin signaling (66, 67).
Conclusions and Future Challenges
Diverse context-dependent functions of Wnt/β-catenin signaling appear to be supported by context-specific selective gene regulation by the β-catenin transcriptional complex and its switching. Extensive studies using model animals and primary cell cultures have gradually revealed the mechanisms underlying the control of context-specific gene expression (Fig. 3). On the other hand, the mechanisms of switching the Wnt/β-catenin signaling target gene set in response to contextual change are still unclear. Recent studies by us, and others, showed that context-specific post-translational modifications in β-catenin–Tcf/Lef and the β-catenin–Tcf/Lef-binding proteins are involved in the switching process (Fig. 4). Interestingly, we recently discovered many new phosphorylation and ubiquitination sites in Tcf/Lef as well as new β-catenin–Tcf/Lef-binding histone modifiers, enzymes, and transcription factors using a proteomic approach (unpublished data). These findings indicate that β-catenin–Tcf/Lef can be regulated by a variety of post-translational modifications and binding proteins. Therefore, future investigations focusing on these factors may improve our understanding of the mechanisms involved in switching the Wnt/β-catenin signaling target gene set and the molecular systems supporting the multi-functionality of Wnt/β-catenin signaling.
Not only Wnt/β-catenin signaling, but also the other cell signaling pathways, including BMP and Shh, are used repeatedly during embryogenesis and organogenesis, and exert diverse context-dependent functions in different aspects. Hence, clarification of the molecular systems supporting the multi-functionality of Wnt/β-catenin signaling will contribute to the discovery of a new general principle underlying the cell signaling systems controlling embryogenesis and organogenesis. Furthermore, studies focusing on the context-dependent regulation of Wnt/β-catenin signaling will be important, not only in developmental biology but also in drug discovery. Because dysregulation of Wnt/β-catenin signaling causes various diseases, including cancer, mental disorders, osteoporosis, and obesity (2, 68), components of the Wnt/β-catenin signaling system are potential therapeutic targets. Several chemical inhibitors of Wnt/β-catenin signaling, such as XAV-939, IWR-1, and IWP-2 have been identified already. XAV-939 and IWR-1 reduce β-catenin protein levels by promoting Axin protein stabilization, and IWP-2 also blocks Wnt secretion (69, 70). These chemicals affect the activity of Wnt/β-catenin core signaling systems, which contribute to the homeostasis of various tissues. Hence, these inhibitors may not only affect abnormal tissues but may also damage healthy ones. In contrast, pharmacological inhibition of the context-specific modulators may enable cell type- or tissue-specific Wnt/β-catenin signaling regulation and contribute to disease treatment with few side effects.
Clarifying the mechanisms underlying the context-specific control of Wnt/β-catenin signaling target gene sets was difficult. However, recent progress in “omics” technology has somewhat simplified such studies. We can now compare the modifications, binding partners, and genomic distributions (genomic binding regions) of β-catenin–Tcf/Lef, and epigenetic modifications of the target gene promoters/enhancers between different contexts, simultaneously. Soon, mechanisms underlying the diverse context-dependent functions of Wnt/β-catenin signaling will be clarified.
Funding
This work was supported by the programs Grants-in-Aid for Scientific Research (B) (16H05141 to T.I.) and Grants-in-aid for Scientific Research on Innovative Areas (26114006 to T.I.) from the Japan Society for the Promotion of Science (JSPS), Takeda Foundation (T.I.), and Yasuda Medical Foundation (T.I.).
Conflict of Interest
None declared.
