No abstract

We have previously identified a novel intra-tumoral dichotomy in triple-negative breast cancer (TNBC) based on the differential responsiveness to a reporter containing the Sox2 regulatory region-2 (SRR2), with reporter responsive (RR) cells being more stem-like than reporter unresponsive (RU) cells. Using bioinformatics, we profiled the protein-DNA binding motifs of SRR2 and identified Myc as one of the potential transcription factors driving SRR2 activity. In support of its role, Myc was found to be highly expressed in RR cells as compared to RU cells. Enforced expression of MYC in RU cells resulted in a significant increase in SRR2 activity, Myc-DNA binding, proportion of cellsexpressing CD44+/CD24–, chemoresistance and mammosphere formation. Knockdown of Myc using siRNA in RR cells led to the opposite effects. We also found evidence that the relatively high ERK activation in RR cells contributes to their high expression of Myc and stem-like features. Using confocal microscopy and patient samples, we found a co-localization between Myc and CD44 in the same cell population. Lastly, a high proportion of Myc-positive cells in tumors significantly correlated with a short patient survival. In conclusion, inhibition of the MAPK/ERK/Myc axis may be an effective approach in eliminating stem-like cells in TNBC.

We have previously identified a novel intra-tumoral dichotomy in breast cancer based on the differential responsiveness to a Sox2 reporter (SRR2), with cells responsive to SRR2 (RR) being more stem-like than unresponsive cells (RU). Here, we report that RR cells derived from MCF7 and ZR751 displayed a higher tolerance to oxidative stress than their RU counterparts, supporting the concept that the RR phenotype correlates with cancer stemness. Sox2 is directly implicated in this differential H2O2 tolerance, since siRNA knockdown of Sox2 in RR cells leveled this difference. Interestingly, H2O2 converted a proportion of RU cells into RR cells, as evidenced by their expression of luciferase and GFP, markers of SRR2 activity. Compared to RU cells, converted RR cells showed a significant increase in mammosphere formation and tolerance to H2O2. Converted RR cells also adopted the biochemical features of RR cells, as evidenced by their substantial increase in Sox2-SRR2 binding and the expression of 3 signature genes of RR cells (CD133, GPR49 and MUC15). Lastly, the H2O2-induced RU/RR conversion was detectable in a SCID mouse xenograft model and primary tumor cells. To conclude, the H2O2-induced RU/RR conversion has provided a novel model to study the acquisition of cancer stemness and plasticity.

No abstract