Acute leukemia cells catabolize BCAA to maintain mitochondrial OXPHOS and non-EAA synthesis, sustaining cellular α-KG levels. (A) The schema for the BCAA metabolism pathway. (B) Comparison of BCAT1 expression levels in normal CD34⁺ HSPCs (n = 6), CD34⁺ AML cells (n = 9), and CD34⁺ ALL cells (n = 11). The MFI ratio of the target and isotype control for each sample is shown. (C) Cellular contents of α-KG in CD34⁺ HSPCs- (n = 14), AML cells (n = 45), and ALL cells (n = 20) in our metabolome analysis. (D) The alteration of cellular α-KG levels in primary AML cells (n = 5) and ALL cells (n = 9) by the addition of gabapentin (20 mM) and L-asparaginase (1.0 IU/mL) for 24 hours are shown. The α-KG levels under control conditions are set to 1.0. (E) The alteration of cellular α-KG levels in primary AML cells (n = 7) and ALL cells (n = 8) after 24-hour culture in control and BCAA- or Thr/Phe/Lys-free DMEM/F12 medium supplemented with 10% fetal bovine serum (FBS) are shown. The α-KG levels in control conditions are set to 1.0. (F) The schema and results for the isotope-tracing experiments using primary AML (n = 2) and ALL (n = 2) samples cultured in HPLM containing 160 uM of [¹³C₆,¹⁵N₁] Leu are shown. Percentages of [¹³C₆,¹⁵N₁] Leu–derived citrate (M+1 and M+2) and non-EAA (M+1) including Glu, Asp, Ala, and Ser among the total amount of corresponding metabolites quantified at the indicated time points are shown. Results from at least 3 independent experiments are shown. (G) The schema and results for isotope-tracing experiments using primary AML (n = 2) and ALL (n = 1) samples cultured in HPLM containing 30 uM of [¹³C₆] α-KIC are shown. Results from at least 3 independent experiments are shown. ∗P < .05, ∗∗P < .01, ∗∗∗P < .001, panels B, C, D, E, F, and G, mean ± SEM. α-KIC, α-ketoisocaproate; EAA, essential amino acids; HPLM, human plasma-like medium; OXPHOS, oxidative phosphorylation; TCA, trichloroacetic acid.