Solid tumors consist of a heterogeneous mixture of cells that display large variations in their capacity to drive tumor growth. Breast cancer stem cells (referred here as breast cancer-initiating cells, or BCICs) are a subpopulation of cells identified within primary tumors that have an enhanced ability to form tumors in immunocompromised mice. These tumorigenic cells have been prospectively identified and isolated based on the CD44+/CD24−/lo/ESA+/lo surface marker profile. An implication of the presence of a highly tumorigenic subpopulation is that total elimination of this cell fraction is imperative for tumor control. Radiotherapy (RT) is an integral component of breast cancer treatment; therefore, understanding the response of BCICs to radiation is of particular importance. This current study sought to further characterize the effect of radiation on BCICs. To measure BCIC survival following radiation, SUM149, SUM159 and SUM1315 breast cancer cell lines were irradiated as adherent cultures and seeded on ultra-low adherence plates for propagation of surviving BCICs as tumorspheres. Likewise, the tumorigenic capacity of these cells was tested directly by assaying xenograft tumor formation in NOD/SCID mice. In addition, the percent of BCICs was analyzed 24 hours following radiation by flow cytometry. Finally, BCICs were isolated from the SUM159 cell line using fluorescently-labeled antibodies against the BCICs surface markers and sorted into BCICs and non-BCIC subpopulations. The radiosensitivity of these two populations was compared by assaying clonogenic survival. Irradiation of all three cell lines resulted in a significant decrease in tumorsphere formation (mean tumorsphere formation {95% CI}; SUM149: 0 Gy = 84.5 {72.3–97.4}, 9 Gy = 23.6 {11.1–36.3}, 5 × 3 Gy = 13.4 {9.7–21.0}; SUM159: 0 Gy = 517.8 {460.3–575.4}, 9 Gy = 181.8 {124.3–239.4}, 5 × 3 Gy = 198.8 {165.4–232.3}; SUM1315: 0 Gy = 86.0 {64.6–107.4}, 9 Gy 1.8 {0–23.3}), 5 × 3 Gy = no tumorspheres {0–21.4} and p < 0.001 for unirradiated vs. irradiated samples in all three cell lines. Xenograft tumor formation with 1000 cells per injection was also found to be decreased by pre-treatment of these cell lines with radiation (0 Gy = 33.3% tumor formation, n = 6 vs. 9 Gy = no tumor formation, n = 6, 80 days post-injection). There was no significant difference in radiosensitivity of BCICs vs non-BCICs in the clonogenic survival assay. However, cells with the CD44+/CD24−/lo/ESA+/lo cell surface phenotype, which have been used to identify a subpopulation enriched for BCICs, were found to be enriched 24 hours following irradiation as measured by flow cytometric analysis (mean %CD44+/CD24−/lo/ESA+/lo {95% CI}; SUM149: 0 Gy = 0.50% {0.30–0.70%}, 5 × 3 Gy = 1.45% {1.23–1.64%}, p = 001; SUM159: 0 Gy = 0.92% {0.66–1.19%}, 5 × 3 Gy = 2.57% {2.30–2.83%}, p < 0.001; SUM1315: 0 Gy = 0.83% {0.42–1.24%}), 5 × 3 Gy = 2.23% {1.82–2.64%}, p < 0.001. The tumorigenic capacity of breast cancer-initiating cells can be effectively eradicated by radiation. This effect was seen with both single fraction and fractionated delivery of the radiation. Finally, although BCICs may show an initial increase in their proportion of the cell fraction following radiation, this did not translate into an increase in clonogenic survival as adherent colonies, tumorspheres or xenograft tumors.
Read full abstract