Study on normal and tumoral cell subpopulations and their interactions in the mammary gland cancer of humans and animals

Human breast cancer (HBC), canine (CMT), and feline mammary tumors (FMT) are extremely common and are characterized by a remarkable both inter- and intra-tumor heterogeneity. Intra-tumor heterogeneity is due to the coexistence of cancer cells that differ between each other in terms of phenotypic, genetic, behavioral characteristics, and metastatic potential. Cancer stem cells (CSCs) are thought to be responsible for such heterogeneity, resistance to therapy, and metastasis development. Several pathways are altered in CSCs, such as the oncogenic Wnt/-catenin and Hippo pathways, and CSCs are associated to the epithelial-to-mesenchymal transition (EMT) process. The aims of this study were to i) isolate and characterized mammary CSCs; ii) investigate EMT process and Wnt/-catenin and Hippo pathways in mammary cancer of the three species; iii) establish a metastatic mouse model of breast cancer seeking for genes responsible of metastatic dissemination; iv) isolate and characterize extracellular vesicles (EVs), which is one of the main forms of intercellular communication, from canine and feline mammary tumors as well as study the role that EVs play during tumor development. CSC-like cells were isolated from established canine and feline mammary tumor cell lines (CYPp and FMCp, respectively) and phenotypically and molecularly characterized for common CSC markers: CD44, CD24, CD133, SOX2, OCT4. Moreover, gene (qPCR) and protein (IHC and WB) expression of Wnt/-catenin and Hippo pathways-related molecules (-catenin, CCND1, YAP, TAZ, CTGF, ANKRD1) as well as protein expression (IHC) of EMT-related molecules (E-cadherin, SNAIL, TWIST, ZEB) were evaluated in a subset of human, canine, and feline mammary cancer tissues, that were also phenotypically characterized for the following markers: CK8/18, CK5/6, CK14, CD44, and vimentin. Additionally, triple negative breast cancer (TNBC) cell line MDA-MB-231 was used to establish a clinically relevant in vivo metastatic model. Finally, EVs were isolated and characterized from CYPp and FMCp and human glioblastoma-derived EVs were used to study tumor angiogenesis. We found that CD44, CD133, SOX2, and OCT4 expression increase in CSC-like cells (mammospheres) compared to parental adherent cells, therefore they could be used as useful markers in CMTs and FMTs. Wnt/-catenin and Hippo pathways seem to be deregulated at a post-transcriptional level in HBCs, CMTs, and FMTs. Interesting similarities were confirmed between TNBCs and FMTs, as well as between ER+ HBC and CMTs. In our metastatic model, mice developed distant metastases and we found a few genes that might play a role during metastatic dissemination. Among these, caspase 3 seems to be involved in brain metastases. Additionally, EVs were isolated from CYPp and FMCp, visualized by transmissible electron microscopy, counted using nanoparticle tracking analysis, and characterized by immunogold and WB (Alix, CD63, TSG101). Finally, using a human glioblastoma cell line (GBM8) we demonstrated that EVs are directly involved in tumor angiogenesis. Overall, this study confirms the use of dogs and cats as spontaneous models of mammary cancer development due to highly interesting biological similarities among the three species. Also, identification of EVs in CMTs and FMTs opens an interesting unexplored field in veterinary medicine.