a, b, Leukemic mice established as described in Extended Data. 2 were treated with 5 daily injections of free DXR at 0.5 or 0.15 µg/g with and without chemotherapy. Alternatively, a single injection on day 1 of 0.8 or 2.5 µg/g of DXR-loaded nanoparticles (NanoDXR) was given with and without chemotherapy. At 10 days post-treatment, BM was analyzed by flow cytometry to determine frequency of LSCs (a) and HSPCs (b). n=6,10,3,6,10,6,3 (a) and n=6,10,8,6,10,6,3 (b) (left to right, respectively) biologically independent replicates for each condition; mean ± SEM is indicated. Note that 5 doses of 0.15 µg/g DXR is ineffective; however, a single NanoDXR injection with a similar cumulative dose (0.8 µg/g) is most effective at reducing LSCs while allowing for HSPC recovery. c-d, Cohorts of leukemic mice were prepared and treated as in Figure S2 but with low-dose NanoDXR. At 12 days post-treatment, BM was harvested from treated mice and transplanted into sub-lethally irradiated NSG recipients. c, Treatment schematic and Kaplan-Meier curves of recipient mice. The free low-dose DXR treatment group (solid line) from Fig. 5f is shown for comparison (n = 30 per group). d, Recipients of BM from low-dose DXR and low-dose NanoDXR treated leukemic mice were analyzed by flow cytometry at 6 months post-transplant for Blast cells, HSPCs and LSCs. n=27 (free_ and 24 (nano) biologically independent replicates; mean ± SD is indicated. Statistics determined by unpaired two-tailed t-test.