A model that cancer cells’ metastatic competence is determined by their states

A polyclonal mouse model of breast tumor heterogeneity using model cell line 4T1 was developed previously. It was shown that 4T1 cells were composed of mixed subpopulations with specializations, including dominating the primary tumor, contributing to metastatic populations, etc. Here in a different angle, we reveal that each 4T1 cells can switch between 2 distinct states, which could be easily misclassified into 2 subpopulations, as the 2 states can be distinguished by distinct patterns of gene expression, dramatic difference of metastatic competence, stem-cell like feature or not, etc. The event for each 4T1 cell to exhibit one state or another is stochastic and the metastatic competence of clones generated from a single 4T1 cell is variable rather than stable. Further, given that metastatic competence of 4T1 cells is not stable, classification of clones based on metastatic competence is highly biased. Similarly, the classification of 4T1 clones based on gene expression profiles is also highly biased. Taken together, our data support that the exhibition of metastatic competence depends on the state of 4T1 cells and do not support the existence of subpopulations with specialization. To the best of our knowledge, no similar work has ever been reported. 4T1 cells were cultivated in 96-well plate at the concentration of one cell per well with 200 μL complete RPMI-1640. After 14 days of culture, 10 SC clones and 10 NSC clones were randomly collected. We collect three batches of SC and NSC from 3 independent experiments. Total RNA was extracted using the RNeasy kit (Qiagen, Germany). The quality of the total RNA was verified with the BioAnalyzer (Agilent, Santa Clara, CA, USA).