Discrimination of fibroblast subtypes by multivariate analysis of gene expression

摘要: Fibroblasts and myofibroblasts from normal, fibrotic or tumoral breast tissues present multiple quantitative differences in gene expression even when grown isolation. We were therefore prompted to investigate whether one could recognize various subtypes by their constitutive-gene profile. Quantitative autoradiographic data for 34 constitutively expressed transcripts submitted multivariate analysis of variance, followed discriminant single linkage cluster analysis. Models assuming up 8 putative fibroblast (among fibroblasts skin, normal mammary stroma, tumor-adjacent "normal" post-radiation fibrosis lesions benign malignant tumors) an epithelial-cell group used as internal control resulted 100% correct classification. Myofibroblasts origins clustered close to, although distinctly apart from, corresponding alpha-smooth-muscle-actin-negative counterparts. Malignant tumor phenotypically more distant cells compared with other pathological types. Our results support the hypothesis co-adaptive transformation stromal epithelial during development suggest that different types give rise myofibroblasts. Discriminant molecular variation may be considered a powerful artificial-intelligence method cell typing should particularly useful no reliable discrete markers are available.