Abstract
Glucose, the most abundant in vivo circulating nutrient, is essential for cell survival. Biochemistry textbooks posit glucose catabolism must terminate either via the glycolytic conversion of glucose to lactate by lactate dehydrogenase (LDH) or via the oxidative decarboxylation of pyruvate to acetyl-CoA by pyruvate dehydrogenase (PDH), followed by further oxidation in the tricarboxylic acid (TCA) cycle. Most previous studies have focused on the role of each terminal glycolytic step in isolation. However, the two enzymes LDH and PDH have never been simultaneously ablated in any tissue to test if cells in vivo must terminate glycolysis in the way established in textbooks. Here, we investigated whether glucose catabolism terminating in the canonical endpoints of LDH and PDH is required across hematopoietic cell types. Hematopoietic stem cells (HSCs) have been suggested to rely on glycolytic production of lactate or on glucose oxidation in various contexts. We generated hematopoietic-specific deletions of Ldha, Ldhb, and Pdha1 using Mx1Cre and validated them by western blot, single HSC-derived colony genotyping, and enzymatic assays. Single deletions of Ldha or Ldhb did not affect HSC frequency. Competitive transplantation revealed that Ldha or Ldhb loss had no impact on peripheral blood reconstitution or bone marrow chimerism in primary recipients, and secondary transplants showed stable or slightly increased HSC function. Combined Ldha;Ldhb deletion completely ablated LDH activity but caused no functional deficits in blood reconstitution or bone marrow chimerism after serial competitive transplantation. Therefore, contrary to current thinking in stem cell metabolism, HSCs do not require LDH and glycolytic lactate production in vivo.
To assess if HSCs can survive by switching between glycolysis and glucose oxidation, we generated Ldha/Pdha1 double- and Ldha/Ldhb/Pdha1 triple conditional knockout mice. These mice developed progressive anemia and reduced bone marrow cellularity, with increased HSC frequency. Serial competitive transplantation revealed intact myeloid and B cell reconstitution but selective loss of T cells. HSPC bone marrow chimerism was unaffected. Thus, HSCs do not require LDH or PDH for survival or multilineage reconstitution. To identify metabolic adaptations enabling HSC survival without LDH and PDH, we used methods we previously developed to perform in vivo rare-cell metabolomics and in vivo U-¹³C-glucose labeling of sorted HSC/MPP (Lin-Sca-1+Kit+CD48-), progenitors (HPC1, Lin-Sca-1+Kit+CD48+CD150-), and myeloid progenitors (MP, Lin-Sca-1-Kit+). We observed accumulation of upper glycolytic intermediates (glucose-phosphate, fructose-6-phosphate) and increased alanine, a transamination product of pyruvate, across all populations. Notably, pyruvate oxidation was minimal, with no significant changes in acetylcarnitine or TCA cycle metabolites (citrate, fumarate, malate). To assess whether pyruvate was exported, we analyzed bone marrow interstitial fluid and detected markedly elevated pyruvate in double and triple knockout mice. U-¹³C-glucose tracing confirmed that extracellular pyruvate originated from glucose in vivo. These findings indicate that glycolysis can terminate with pyruvate export. This adaptation rendered HSCs vulnerable to inhibition of monocarboxylate transporter 1 (MCT1). Transplant recipients of Ldha;Ldhb and Ldha/Ldhb/Pdha1 bone marrow cells treated with AZD3965, an MCT1 inhibitor, exhibited impaired HSC function. Given the role of LDH and PDH in NAD⁺ regeneration, we hypothesized that mitochondrial complex I would compensate for redox imbalance in the absence of LDH and PDH. Indeed, complex I inhibition in double and triple knockout mice reduced HSC function, highlighting a reliance on mitochondrial NAD⁺ regeneration.In conclusion, our findings demonstrate that in the absence of LDH and PDH, glycolysis terminates at pyruvate, which is exported, and HSCs maintain redox balance via mitochondrial complex I. These results: a) Suggest that glycolytic lactate production is not essential in HSCs in contrast to the idea that HSCs are glycolytic; b) Reveal metabolic plasticity that sustains HSC survival and function independent of canonical glycolytic endpoints; and c) Most fundamentally, they redefine how glucose can be catabolized in vivo by showing that the canonical endpoints of glycolysis are dispensable for metabolism and cell function.
