Extended Data Fig. 5. Venetoclax and prednisolone synergize in primary ALL with low CELSR2 expression and CELSR2 knockdown in cell lines disregulation of Bim/Bcl2 axis.

(a.) Response surface model plot of cytotoxicity from prednisolone plus venetoclax at concentrations indicated for the 697 leukemia cell line transduced with non-targeting control vector. (b.) Response surface model plot for the 697 leukemia cell line transduced with CELSR2 shRNA knockdown vector (for a and b individual points represent n= 3 independent experiments performed in technical duplicate; response surface model two-tailed t-test p-value). The alpha (α) value indicates antagonism < 0 or synergy > 0 with greater synergy from higher value. P-value describes overall model fit. Individual plots of prednisolone effect (mean ± S.D.; n= 3 independent experiments) (c.) NALM-6 and (d.) 697 leukemia cell lines at one concentration of venetoclax (mean ± S.D.; n= 3 independent experiments). Black lines are non-targeting control cells and red lines are CELSR2 knockdown cells, dashed lines indicate predicted additivity curve fit based on single drug treatments; data left of the dashed lines represent additivity/synergy. Solid lines represent fit of measured values. (e.) Venetoclax sensitivity of independent cohort of patients (n=96 ALL patients) grouped based on prednisolone sensitivity (LC₅₀) (f.) Bcl2 expression associated with sensitivity to venetoclax (n= 81 ALL patients) (g.) Primary ALL cells from patients (n=6 patient samples) and human leukemia cell lines assessed for additivity/synergy with prednisolone and venetoclax (for all box plots horizontal bars depict medians and boxes represent 25^th and 75^th percentiles, whiskers represent ±1.5x IQR; linear model p-values).(h.) mRNA expression (n=1 experiment run in technical triplicate) of CELSR2 in patient samples assessed for synergy.