a–c Mice with MLL-AF9 leukemia (n = 6/group) were administered 24 h high-dose cytarabine (Ara-C) chemotherapy, or vehicle control, prior to harvest of BM cells for flow cytometry analyses. a Dot plots show gating strategy and altered distribution of E-selectin-binding potential among BM AML blasts (GFP+ KIT+) before (top panel) and after chemotherapy (bottom panel). Representative dot plot from one mouse per group. b Fluorescence histogram overlay showing E-selectin-binding intensity of BM AML blasts before (black line) and after (pink line) chemotherapy. Grey fill is no binding negative control (n = 6 mice/group). Representative mouse shown. c Percentage of surviving phenotypic AML blasts after 24 h cytarabine administration compared to starting numbers per femur for each E-selectin-binding category (low, medium (med), high gates, as defined in dot plot panel in (5a) above. Mean ± SD of n = 6 mice/group. Statistical analysis, one-way ANOVA multiple comparison Bonferroni correction. d Experimental outline, (n = 6) mice were separated by FACS and 1500 sorted AML blasts (GFP+ KIT+) with highest or lowest E-selectin-binding (using the gating strategy shown in top panel above) were transplanted into 2.5 Gy conditioned recipient mice (n = 8/recipients group). In one cohort of donors GMI-1271 was administered 200 mg kg−1 BiD for last 48 h prior to harvest and sort. (Bottom) Kaplan Meier disease-free survival of recipient mice transplanted with 1500 sorted AML blasts. Solid lines represent survival of recipients of high E-selectin-binding whereas dashed lines are recipients of low E-selectin-binding AML blasts. Grey curves indicate recipients of AML blasts from vehicle treated donors, while red/orange indicates recipients of AML blasts from GMI-1271-treated donors. Statistics, Log-Rank (Mantel–Cox) survival curve comparison. e Percentage of AML blasts able to efflux Rhodamine-derivative TMRM dye by flow cytometry, within lowest or highest E-selectin binding groups (gating strategy shown in Supplementary Fig. 5). Each pair of dots are data from same individual mouse (n = 7 mice/group). Statistical significance calculated by two-tailed paired t test. Source data are provided as a Source Data file.