Loss of Metabolic Control Beyond the Promyelocyte Stage Resolves Myeloid Maturation Arrest in Reticular Dysgenesis

Reticular Dysgenesis (RD) is a particularly grave syndromic form of severe combined immunodeficiency. The hematopoietic phenotype manifests in defective myeloid and lymphoid development with profound neutropenia and lymphopenia. RD is caused by biallelic mutations in the mitochondrial enzyme adenylate kinase 2 (AK2), which catalyzes the phosphorylation of adenosine monophosphate (AMP) to adenosine diphosphate (ADP) in the mitochondrial intermembrane space. ADP is then phosphorylated to adenosine triphosphate (ATP) during the OXPHOS process. While there are abundant other sources of APD, AK2 constitutes the only mitochondrial enzyme that contributes to AMP “recycling” by phosphorylation. Accordingly, AK2-deficient myeloid cells have been shown to exhibit increased AMP levels ¹. Beyond this observation, the cellular basis for the combined hematopoietic defect in RD remains elusive.

Using a CRISPR/Cas9 AK2 biallelic knock-out model in human hematopoietic stem and progenitor cells (HSPCs), we have previously shown that AK2^-/- HSPCs recapitulate RD myelopoiesis defects in vitro. AK2^-/- HSPCs exhibit a severe proliferative defect that becomes apparent at the myelocyte (MC) and neutrophil (NP) stages, while the promyelocyte (PM) stage appears relatively spared, consistent with observations in RD patients.

Our prior metabolomic analysis showed that AK2^-/- MCs not only have increased levels of AMP but up of 20-fold increased levels of IMP derived by converting excess AMP to IMP conversion via AMP deaminases (AMPDs). In addition, we have shown that beginning at the MC stage, AK2 ^-/- cells exhibit an increase in NADH:NAD+ ratio, consistent with reductive stress and decreased aspartate levels. Our new data revealed that expressing the aspartate transporter GLAST and/or the water-forming NADH oxidase LbNOX in AK2^-/- HSPCs partially rescues differentiation but not the proliferation defect. It raises the possibility that reductive stress and aspartate deficiency compromise protein synthesis necessary for maturation, but are separate from the proliferative defect related to high IMP and AMP levels. There were no metabolomic abnormalities noted at the PM stage.

PMs have a higher proliferative index than MCs and NPs. This raises the question why the most metabolically active stage is phenotypically spared by AK2 deficiency. To better understand why the metabolic abnormalities associated with RD selectively affect differentiation beyond the promyelocyte stage, we investigated key regulators of cellular metabolism and proliferation, i.e., AMPK, ACC1 and S6, a target of mTOR signaling, at PM, MC and NP stages. We found activation of AMPK, ACC1 and S6 only at the PM stage, while AMPK, ACC1 and S6 protein expression at MC and NP stages was virtually undetectable. Interestingly, AK2^-/- PMs exhibited notably lower ATP synthesis rates compared to controls, while ATP synthesis in AK2^-/- MCs and NPs was significantly higher than in controls of corresponding stages, paralleling the accumulation of AMP and IMP noted earlier.

These findings raise the possibility that tight metabolic control at the promyelocyte stage prevents AMP levels from rising by curtailing OXPHOS metabolism in AK2^-/- PMs. In contrast, lack of metabolic control beyond the PM stage permits unopposed OXPHOS activity, thereby allowing AMP and IMP levels to rise to toxic levels that interfere with proliferation.

In our ongoing work, we are testing whether pharmacologically redirecting metabolism from OXPHOS to glycolysis with UK5099, could rescue AK2^-/- myelopoiesis defects. UK5099 is a potent inhibitor of the mitochondrial pyruvate carrier, which facilitates pyruvate transport across the mitochondrial inner membrane to fuel the TCA cycle. Additionally, we developed a hematopoietic-specific and inducible Cre-mediated Ak2 knock-out mouse model, which recapitulates the profound neutropenia and lymphopenia observed in RD patients. This mouse model will allow us to corroborate our in vitro observations and to investigate if the same mechanistic underpinnings also apply to the lymphoid lineage.

• Rissone A, Weinacht KG, la Marca G, et al. Reticular dysgenesis-associated AK2 protects hematopoietic stem and progenitor cell development from oxidative stress. J Exp Med. Jul 27 2015;212(8):1185-202. doi:10.1084/jem.20141286