(Non)canonical Wnt signaling, cytoarchitecture and stemness: new insights from primary nonmetastatic, primary metastatic, regional and distant metastatic models of adrenocortical carcinoma

Tumor cells rearrange cytoskeletal networks, harmonize signaling pathways and modulate cell fate to initiate tumor growth and metastasis. Activating mutations in the canonical Wnt–β-catenin pathway are associated with aggressiveness and poor prognosis in adrenocortical carcinoma (ACC). However, the contribution of noncanonical Wnt pathways, which are known to regulate cell polarity and migration, is still widely unknown. Here we comprehensively investigated canonical and noncanonical Wnt signaling along with cytoskeleton networks, stemness programs and HOX gene expression in recently developed 2D and 3D cultures of primary, local and distant metastatic ACC models. We report that tumor spheroids derived from nonmetastatic primary tumors correlated strongly with the overexpression of canonical Wnt signaling components, while metastatic ACC models demonstrated shifts toward noncanonical Wnt pathways together with perturbed expression of matrix–cytoskeleton–nucleoskeleton markers and related changes in the cytoarchitecture of respective tumor spheroids. Consistent with a potential noncanonical Wnt activation in the metastatic models, an abundance and relocalization of proteins associated with tumor aggressiveness such as β-catenin, β-actin, vimentin and nucleolin as well as specific HOX-cluster activation was detected, suggesting strong potential for specific reconfigurations of metastatic tumor niches. Furthermore, all models demonstrated key properties of cancer stem cells, although with varying degrees of expression of progenitor and stem cell markers, which could be dynamically modulated by treatments and therapies. Overall, our results draw parallels between canonical and noncanonical Wnt signaling, differential cytoskeletal remodeling, stemness properties and various cellular plasticities in primary tumor and metastasis-derived models of ACC.