背景:尽管经历上皮-间质转化(EMT)的肿瘤细胞在实验模型中通常表现出纺锤体形态,EMT的这种组织形态学证据主要在罕见的原发性梭形癌中观察到。遗传不受干扰的非纺锤癌中自发EMT的特征和转录调节因子仍未得到充分研究。
方法:我们使用原代培养结合RNA测序(RNA-seq),单细胞RNA-seq(scRNA-seq),和原位RNA-seq探讨非纺锤状乳腺癌中潜在自发性EMT的特征和转录因子(TFs)。
结果:我们的原代培养显示癌细胞表达不同的上皮间质性状,符合上皮-间质可塑性。重要的是,经历自发性EMT的癌细胞不一定表现出纺锤体形态,即使在接受完整的EMT时。EMT是一个受欢迎的过程,而间充质-上皮转化似乎对继发性肿瘤的生长至关重要。通过scRNA-seq,我们确定了随着癌细胞通过EMT过程进展而顺序和显着上调的TFs,与VIM表达增加相关。一旦上调,TFs在整个EMT过程中保持活跃。ZEB1是EMT的关键发起者和维持者,正如它在EMT过程中最早的显著上调所表明的那样,它与VIM表达的确切相关性,以及ZEB1敲低后EMT的逆转和EMT上调的TFs的下调。ZEB1和波形蛋白在三阴性乳腺癌和化生性乳腺癌肿瘤队列中的表达之间的相关性进一步强调了其作用。ZEB2在ZEB1之后立即上调,同时观察到ZEB1或ZEB2的敲低与EMT的逆转同时下调ZEB1和ZEB2,建议他们在EMT中进行功能合作。这个发现,以及SNAI1,SNAI2和TWIST1表达与间充质表型缺乏相关性,表明EMT-TF在EMT中具有上下文依赖的作用。EMT期间EMT相关基因特征的上调与不良患者预后相关,强调模型的生物学重要性。与波形蛋白阴性癌细胞相比,波形蛋白阳性的EMT基因特征升高,ZEB1和ZEB2表达增加,证实ZEB1和ZEB2是内在的,而不是微环境诱导的,EMT调节器,和波形蛋白作为EMT的体内指标。
结论:我们的发现为原发性非纺锤形癌自发性EMT的特征和转录调节因子提供了见解。
BACKGROUND: Although tumor cells undergoing epithelial-mesenchymal transition (EMT) typically exhibit spindle morphology in experimental models, such histomorphological evidence of EMT has predominantly been observed in rare primary spindle carcinomas. The characteristics and transcriptional regulators of spontaneous EMT in genetically unperturbed non-spindled carcinomas remain underexplored.
METHODS: We used primary culture combined with RNA sequencing (RNA-seq), single-cell RNA-seq (scRNA-seq), and in situ RNA-seq to explore the characteristics and transcription factors (TFs) associated with potential spontaneous EMT in non-spindled breast carcinoma.
RESULTS: Our primary culture revealed carcinoma cells expressing diverse epithelial-mesenchymal traits, consistent with epithelial-mesenchymal plasticity. Importantly, carcinoma cells undergoing spontaneous EMT did not necessarily exhibit spindle morphology, even when undergoing complete EMT. EMT was a favored process, whereas mesenchymal-epithelial transition appeared to be crucial for secondary tumor growth. Through scRNA-seq, we identified TFs that were sequentially and significantly upregulated as carcinoma cells progressed through the EMT process, which correlated with increasing VIM expression. Once upregulated, the TFs remained active throughout the EMT process. ZEB1 was a key initiator and sustainer of EMT, as indicated by its earliest significant upregulation in the EMT process, its exact correlation with VIM expression, and the reversal of EMT and downregulation of EMT-upregulated TFs upon ZEB1 knockdown. The correlation between ZEB1 and
vimentin expression in triple-negative breast cancer and metaplastic breast carcinoma tumor cohorts further highlighted its role. The immediate upregulation of ZEB2 following that of ZEB1, along with the observation that the knockdown of ZEB1 or ZEB2 downregulates both ZEB1 and ZEB2 concomitant with the reversal of EMT, suggests their functional cooperation in EMT. This finding, together with that of a lack of correlation of SNAI1, SNAI2, and TWIST1 expression with the mesenchymal phenotype, indicated EMT-TFs have a context-dependent role in EMT. Upregulation of EMT-related gene signatures during EMT correlated with poor patient outcomes, highlighting the biological importance of the model. Elevated EMT gene signatures and increased ZEB1 and ZEB2 expression in
vimentin-positive compared to
vimentin-negative carcinoma cells within the corresponding primary tumor tissue confirmed ZEB1 and ZEB2 as intrinsic, instead of microenvironmentally-induced, EMT regulators, and
vimentin as an in vivo indicator of EMT.
CONCLUSIONS: Our findings provide insights into the characteristics and transcriptional regulators of spontaneous EMT in primary non-spindled carcinoma.