"Cell dedifferentiation" versus "evolutionary reversal" theories of cancer: The direct contest of transcriptomic features.

Abstract

Cell dedifferentiation is considered an important hallmark of cancer. The atavistic reversal to a unicellular-like (UC) state is a less widely accepted concept, so far not included in the conventional hallmarks. The activated expression of ontogenetically earlier and evolutionary earlier genes in cancers supports both theories because ontogenesis partially recapitulates phylogenesis during cell differentiation (the cellular biogenetic law). We directly contested both types of gene signatures in stem vs. differentiated and cancer vs. normal cells, using meta-analysis of human single-cell transcriptomes (totally, 38 pairwise comparisons involving over 18,600 cells). Because compared cells can differ in proliferation rate, the correction for cell cycle activity was applied. Taken together as multiple variables in stem vs. differentiated cells analyses, the ontogenetic signature excluded the UC signature from predictive variables, usually even forcing it to change the sign of prediction (from plus to minus). In contrast, in cancer vs. normal cells, the UC signature excluded the ontogenetic signature. Thus, the direct contest decided in favor of the atavistic theory and placed a UC-like state as a central hallmark of cancer, which has a plausible evolutionarily formed mechanism (UC attractor) and can generate other hallmarks. These data suggest a paradigm shift in the understanding of oncogenesis and propose an integrative framework for cancer research. In a practical sense, the upregulation of UC signature over ontogenetic signature indicates potential tumorigenicity, which can be used in early diagnostics and regenerative medicine. For therapy, these results suggest the UC center of cellular networks as a universal target.