The Role of Canonical Glycolytic Endpoints in the Development and Function of Hematopoietic Stem and Progenitor Cells

Glucose catabolism via glycolysis is thought to be an essential metabolic pathway. After glycolysis converts glucose to pyruvate, it terminates by either the reduction of pyruvate to lactate by lactate dehydrogenase (LDH) or by the oxidative decarboxylation of pyruvate to acetyl-CoA by pyruvate dehydrogenase (PDH), followed by further oxidation in the tricarboxylic acid (TCA) cycle. The textbook description of glycolysis is that cells must use either one or the other terminal step to complete the pathway. However, LDH and PDH have never been ablated simultaneously in any tissue to test if this assumption is true in vivo. Specifically in the hematopoietic system it has not been systematically defined which cell types particularly depend on glucose oxidation via PDH or glucose fermentation via LDH or both catabolic arms for survival and function. To test the essentiality of glucose catabolism in HSCs and other hematopoietic lineages, we generated single, double and triple genetic knockouts (KO) of Ldha, Ldhb, and Pdha1. The main objective of this study is to understand how glycolysis operates in hematopoietic stem cells (HSCs) and other hematopoietic cells and to identify possible alternative routes of glucose catabolism in the absence of LDH and PDH.

Ldha, Ldhb, and Pdha1 were conditionally deleted in the adult hematopoietic system using Mx1Cre. Ldhb^Δ/Δ mice did not develop anemia indicating that LDHB is dispensable in erythroid development at steady state. Ldha deletion alone or in combination with Ldhb and/or PDH caused anemia. PDH deletion in Ldha^Δ/Δ;Pdha1^Δ mice exacerbated the anemia suggesting that LDH and PDH redundantly support erythropoiesis. Analysis of erythropoiesis revealed an increase of CD71⁺Ter119⁺ erythroid progenitors in the bone marrow BM of Ldha^Δ/Δ mice. This is dependent on PDH activity, because this increase was abolished in Ldha^Δ/Δ;Pdha1^Δ bone marrow. Ldha^Δ/Δ;Pdha1^Δ mice instead showed an increase in upstream CD71^-Ter119^int and CD71⁺Ter119^int erythroid progenitors. Analysis of more immature erythroid progenitors revealed an accumulation of pre-CFU-E and a more pronounced decrease in CFU-E in the bone marrow of Ldha^Δ/Δ;Pdha1^Δ mice as compared to Ldha^Δ/Δ mice. These results suggested that in the bone marrow, some erythroid progenitors do not depend on either LDHA or PDH while others rely on either LDHA or PDH. In contrast to bone marrow erythropoiesis arrest, spleen size, cellularity, and erythropoiesis expanded in all these genotypes. Therefore, there is a differentiation-stage and organ-specific requirement for glycolytic lactate production or glucose oxidation in erythroid development.

Parallel analysis in T cell development also showed differentiation-stage specific requirements for LDH and PDH. Ldha^Δ/Δ or Ldha^Δ/Δ;Ldhb^Δ/Δ mice did not show any changes in thymocyte number or composition, but Ldha^Δ/Δ;Pdha1^Δ or Ldha^Δ/Δ;Ldhb^Δ/Δ;Pdha1^Δ mice showed a significant decrease in thymocytes from double negative (DN1) stage to all subsequent stages of T cell development. We previously showed that PDH is required to support double positive (DP) T cell development. Therefore, our results suggest that LDH and PDH are required either alone or redundantly at different stages of T cell development. Further and ongoing experiments have defined the requirements for glycolytic lactate production or glucose oxidation in HSCs, multipotent and restricted progenitors, and other hematopoietic lineages. Our results suggest multiple cell-type specific dependencies on different arms of glycolysis in vivo across hematopoietic differentiation.

To understand how glycolysis operates in hematopoietic cells, we performed rare cell metabolomics and stable isotope tracing in vivo in HSCs, restricted progenitors, B cells, neutrophils, and T cells in these single or combined conditional knockouts. We observed multiple cell-type specific metabolic consequences of blocking glycolysis or glucose oxidation. As an example, pyruvate and alanine were significantly increased in Ldha^Δ/Δ;Pdha1^Δ T cells and lactate, and acetylcarnitine and TCA cycle metabolites were significantly decreased suggesting a shift to alanine metabolism. Our work comprehensively map the cell-type specific operation and functional requirements of terminal glycolytic endpoints in a tissue for the first time and uncover cell-type specific requirements in many hematopoietic cell lineages.