Structural and Functional Insights into Dishevelled-Mediated Wnt Signaling

Abstract

Dishevelled (DVL) proteins precisely control Wnt signaling pathways with many effectors. While substantial research has advanced our understanding of DVL's role in Wnt pathways, key questions regarding its regulatory mechanisms and interactions remain unresolved. Herein, we present the recent advances and perspectives on how DVL regulates signaling. The experimentally determined conserved domain structures of DVL in conjunction with AlphaFold-predicted structures are used to understand the DVL's role in Wnt signaling regulation. We also summarize the role of DVL in various diseases and provide insights into further directions for research on the DVL-mediated signaling mechanisms. These findings underscore the importance of DVL as a pharmaceutical target or biological marker in diseases, offering exciting potential for future biomedical applications.

Keywords: AlphaFold; Dishevelled; Wnt; cancer; mechanism; post-translation.

Figure 1

Plausible mechanism of DVL-mediated Wnt signaling pathways. (A) The destruction complex degrades the β-catenin protein. DVL is inactivated through autoinhibition. (B) DVL regulates both canonical and non-canonical Wnt signaling through its interaction with Frizzled (FZD) and other co-receptors. Upon ligand binding, DVL inhibits the β-catenin degradation complex in canonical signaling or activates downstream effectors in non-canonical signaling. (C) Conformational changes of DVL may regulate Wnt signaling.

Figure 2

Human DVL proteins have three paralogs, DVL1/2/3. (A) Sequence alignment of human DVL proteins using ClustalW [42]. The conserved DIX (blue), PDZ (green), and DEP (yellow) domains are highlighted. The intrinsically disordered regions (IDRs) are indicated. The red box represents the IDR1-basic 2 region. (B) AlphaFold-predicted DVL3 structure shows the autoinhibition conformation. In the intrinsically disordered regions, AF3 predicts two putative helical structures. IDR2_H1 is located between the PDZ and DEP domains. IDR3_H1 is near the C-terminal region of DVL. (C) Predicted aligned error (PAE) maps for AF3-predicted human DVL3 protein structure.

Figure 3

The structures of the highly conserved domains and IDR1 region of hDVL3. (A) Structure diagram of hDVL3 protein (UniProt ID: Q92997). (B–D) AlphaFold-predicted structure of the highly conserved domains: DIX, PDZ, and DEP domain. (B) The DVL DIX domain can interact with other DIX domains, such as Axin, through head and tail interfaces. (C) PDZ domain is a protein-protein interaction module. The red circle represents the PDZ-binding site. More than 30 binding partners for the PDZ domain were reported [3,24,37]. (D) The DEP domain can interact with the plasma membrane through its positively charged surface with the negatively charged lipids, such as PIP₂ of the inner layer of the plasma membrane. The red circle represents the DEP-finger. (E) AF3-predicted the complex structure of the hFZD3 transmembrane domain (TMD) with the hDVL3 DEP domain. The binding site of hFZD3 TMD and hDVL3 DEP was expanded. The residues in the binding interfaces, hydrogen bonds, and salt bridges of the FZD3-DVL3 DEP complex predicted by AF3 are summarized in Table S1 and Figure S2. The DEP-finger region may bind to the intracellular region of hFZD3. (F) AF3-predicted structure of IDR1_basic2 region of hDVL3, showing the putative α-helical structure. ReSMAP prediction shows that IDR1_basic2 of DVLs may bind to the plasma membrane (Figure S3 shows the IDR1_basic region of hDVL1).