Abstract
Cancer acquires metastatic potential and evolves via co-opting gene regulatory networks (GRN) of embryonic development and tissue homeostasis. Such GRNs are encoded in the genome and frequently conserved among species. Considering that all metazoa have evolved from a common ancestor via major macroevolutionary events which shaped those GRNs and increased morphogenetic complexity, we sought to examine whether there are any key innovations that may be consistently and deterministically linked with metastatic potential across the metazoa clades. To address tumor evolution relative to organismal evolution, we revisited and retrospectively juxtaposed seminal laboratory and field cancer studies across taxa that lie on the evolutionary lineage from cnidaria to humans. We subsequently applied bioinformatics to integrate species-specific cancer phenotypes, multiomics data from up to 42 human cancer types, developmental phenotypes of knockout mice, and molecular phylogenetics. We found that the phenotypic manifestations of metastasis appear to coincide with agnatha-to-gnathostome transition. Genes indispensable for jaw development, a key innovation of gnathostomes, undergo mutations or methylation alterations, are aberrantly transcribed during tumor progression and are causatively associated with invasion and metastasis. There is a preference for deregulation of gnathostome-specific versus pre-gnathostome genes occupying hubs of the jaw development network. According to these data, we propose our systems-based model as an in silico tool the prediction of likely tumor evolutionary trajectories and therapeutic targets for metastasis prevention, on the rationale that the same genes which are essential for key innovations that catalyzed vertebrate evolution, such as jaws, are also important for tumor evolution.