Polyploid giant cancer cells (PGCC): short-term return to multicellularity

Abstract Background Polyploidization is associated with progression of cancer, making cancer cells more dangerous. The common polyploid cancer cells constitute a considerable part of tumors (up to 56% in metastases). The giant polyploid cancer cells (PGCC), which appear under severe stress caused by treatment when the majority of cells die, present an enigmatic phenomenon both in fundamental and practical sense because they develop treatment resistance. Results Using transcriptome meta-analysis, we studied different types of polyploid cancer cells and found that in common polyploid cancer cells, the genes of unicellular (UC) origin and stemness are upregulated (compared to diploid cancer cells). At that, the upregulated UC genes show a higher local and global protein interactome centrality than the upregulated stemness genes, suggesting that the UC interactome attractor is a driving force behind this backward movement along the evodevo axis. Surprisingly, PGCC show the opposite picture. There occurs the suppression of UC and stemness genes with the upregulation of multicellular genes (especially those involved in intercellular communication), suggesting a reversal towards multicellular (MC) state. This effect is enhanced in PGCC's early progeny but diminished in the late progeny, indicating its transient nature. PGCC of different origin (breast, ovarian, prostate cancers), induced by different stresses (radiation or drugs with various mechanisms of action), show a similar behavior. The first principal component of transcriptome profiles, which is common for all cell types (initial cancer cells, PGCC, early and late progeny) and contains the major part of expression variance, is also directed along the gene evolutionary age axis. Conclusions While the common polyploid cancer cells comply with the 'serial atavism' model of oncogenesis, PGCC present a unique phenomenon of the short-term return to multicellularity probably associated with collective acquisition of resistance to treatment. Our analysis revealed also the evolutionary origin of the main differences in gene expression, emphasizing the importance of gene age axis in transcriptome analyses. The deep evolutionary basis of variation in gene expression across and within cell types might become a general framework for interrelated problems of cell and cancer biology and regenerative medicine.