The different types of colony morphologies allow us to predict some stem cell characteristics43. and in vivo. Furthermore, MAP17 increased the exosomes in tumor cells, where MAP17 was released as cargo, and this horizontal propagation also increased the EMT in the recipient cells. Importantly, an antibody against MAP17 in the media reduces the EMT and stemness alterations promoted by the conditioned media from MAP17-expressing cells. Therefore, MAP17 expression promotes the horizontal propagation of EMT and metastasis by transferring the MAP17 protein between subsets of neoplastic cells. Thus, MAP17 can be used to describe a new mechanism for cell malignity at distance, without the involvement of genetic or epigenetic modifications. MAP17 can also be taken in consideration as new target for metastatic high-grade breast tumors. levels and tumor progression, we used Finak data set (“type”:”entrez-geo”,”attrs”:”text”:”GSE9014″,”term_id”:”9014″GSE9014) from Oncomine (https://www.oncomine.org). In addition, we used R2 webpage resource to compare MAP17 expression across data sets, using 219630_at as probe for MAP17 and the algorithm MAS5.0 for data normalization (see Supplementary Table 1). KaplanCMeier method was used for survival analysis, according to R2 webpage adjustments. TCGA Wanderer resource (data sets for Breast Invasive Carcinoma, Colon Adenocarcinoma and Lung Adenocarcinoma) was used to analyze the methylation state of in human samples33, considering CG probes cg15187606 and cg26523175, both upstream of MAP17 gene. To find genes correlated with expression, we selected 31 breast cancer databases (see Supplementary Table 1), all freely accessible through R2 webpage (http://r2.amc.nl). We used two different gene filters: Oncogenesis (GeneCategory) and Pathways in Cancer (KEGG Pathway); both options included in R2. We searched for correlations using the probes TIC10 listed in Supplementary Table 1, establishing a value ?0.05 to identify significant differences. From the list of correlated genes, we separated genes positively from genes negatively correlated with expression, generating two gene lists for each TIC10 database. To look for altered biological processes connected to changes in expression, we used enrichment analysis from Gene Ontology consortium webpage (http://geneontology.org/page/go-enrichment-analysis). The obtained GO terms, from genes that were either positively or negatively correlated with MAP17 expression, were compared using Venny tool34. In addition, we used Panther (http://www.pantherdb.org/) to group the list of genes according to protein class. TransmiR v2.0 software (http://www.cuilab.cn/transmir) was used to find miRNAs regulated by NOTCH1, HES1, or HES5. Data sets “type”:”entrez-geo”,”attrs”:”text”:”GSE20685″,”term_id”:”20685″GSE20685 and “type”:”entrez-geo”,”attrs”:”text”:”GSE7390″,”term_id”:”7390″GSE7390 were used to separate patients according to tumor type (primary vs metastasis) and levels (low vs high), using GEO2R (https://www.ncbi.nlm.nih.gov/geo/geo2r/) to obtain the expression values of each individual gene. Cell lines and cellular assays T-47D, MDA-MB-231, MDA-MB-468, and MCF10A cells were obtained from the European Collection of Authenticated Cell Cultures (ECACC) commercial repository at the beginning of this study. No further authentication was performed in these KSR2 antibody cell lines. AA, AW, AX, BC, and CE cell lines, derived from sarcoma patients, were described previously35. T-47D, MDA-MB-231, and MDA-MB468 cells were maintained in DMEM (Gibco), whereas sarcoma cells were TIC10 maintained in F10 (Gibco), all supplemented with 10% fetal bovine serum (FBS; Life Technologies), penicillin, streptomycin, and fungizone. All cell lines were regularly tested for mycoplasma. MAP17 expression was induced through transfection with plasmid pBabe-MAP17, previously described12,15. All transfected cells were selected with 1?g?mL?1 of puromycin. Clonogenicity assays, holo- and paraclone analysis and tumorspheres analysis were performed as previously described36. miRNAs analysis We extracted total RNA from T-47D cells, overexpressing MAP17 or control, using Qiazol and miRNAeasy kit (Qiagen, USA). To find miRNAs with significant differences between both conditions, we used the Cancer Pathway Finder miScript miRNA PCR Array (Qiagen, USA), following manufacturers instructions. All miRNAs detected with significant differences were analyzed using miRTarBase resource (miRTarBase.mbc.nctu.edu.tw/), focusing only in changes in gene.