Abstract Knowledge gap: Fatal breast cancers are characterized by biological, genomic and extensive treatment heterogeneity. Although many breast cancers can now be cured by established therapies, treatment failure remains a major problem and is difficult to predict. In the current era of “personalized medicine”, a possible solution is to develop a large-scale system for quantifying responses to candidate treatments of individual malignant human mammary cells with in vivo clonogenic activity. Such cells can be detected by their ability to produce uniquely barcoded clones of progeny in xenografted immunodeficient mice and the clones obtained can be assessed for their size and number using next generation sequencing of tumor extracts. However, to pursue this approach it is first critical to establish how the clone content of a tumor may vary according to the number or type of competent tumorigenic and/or other cells that are present in the inoculum used to initiate tumor formation, and hence whether and how these parameters may influence assessment of the treatment responsiveness of these cells. Approach/methods: Here we describe the development and initial testing of a method to measure the treatment responsiveness of large numbers of tumorigenic cells using radiation as a prototypic treatment. Treatment sensitivity of in vitro colony-forming cells (CFCs) will then be compared with future measurements of in vivo clone-initiating tumorigenic cells obtained by sequencing the progeny of DNA-barcoded input cells. Results: In an initial series of experiments we showed that normal human luminal progenitor (LP) CFCs are ~1.5-fold more radioresistant than basal cell (BC) CFCs, and both are more sensitive than either type of mouse mammary CFCs. In vitro CFC assays of 2 human breast cancer cell lines (MDA MB231 and SUM149, with in vitro CFC frequencies of 70% and 40%, respectively) showed these to be 1.2- and 1.5-fold more radioresistant than normal LPs. Limiting dilution analysis showed the corresponding frequency of in vivo tumor-initiating cells in these 2 cell lines to be 1/6 and 1/47. Assessment of their response to radiation is complicated by the finding that the barcoded clone content of tumors initiated with >20,000 of these cells (untreated) is inversely related to the number injected and, at these input cell doses, very heterogeneous clone dynamics are also seen in successive passages. However, evidence of a positive linear cell dose-clone yield relationship is seen at input transplants of <1,000 cells from these 2 lines and this relationship would be predicted to extend to initial transplants of <2-5,000 cells. Recently we have found that normal human mammary LPs and BCs transduced with KRASG12D generate tumors efficiently in immunodeficient mice. Thus a next step is to determine if and how the radiosensitivity of the tumorigenic cells changes during the transformation process. Conclusion: These results highlight the complex clonal dynamics already operative in the growth of tumorigenic cells present in relatively homogeneous established human mammary cell lines and set the stage for future measurements of clone yields from irradiated cells derived from mammary tumors of different origins. Citation Format: Sneha Balani, Nagarajan Kannan, Long V. Nguyen, Sylvain Lefort, Davide Pellacani, Connie J. Eaves. Clonal analysis of normal and malignant human mammary epithelial cell responsiveness to radiation. [abstract]. In: Proceedings of the AACR Special Conference on Advances in Breast Cancer Research; Oct 17-20, 2015; Bellevue, WA. Philadelphia (PA): AACR; Mol Cancer Res 2016;14(2_Suppl):Abstract nr A63.