AUTHORS
Kazuhiro Noguchi, MD.,PhD., Carlo Castruccio Castracani, PharmD, PhD2, Jean Ann Maguire3, Alyssa Gagne4, Wei Tong, PhD5, Giulia Pavani, PhD6, Stefano Rivella, PhD7
AFFILIATIONS
1 Children's Hospital of Philadelphia, Philadelphia, Pennsylvania, United States
2 Children's Hospital of Philadelphia, Philadelphia, Pennsylvania, United States
3 Children's Hospital of Philadelphia, Philadelphia, Pennsylvania, United States
4 Children's Hospital of Philadelphia, Philadelphia, Pennsylvania, United States
5 Children's Hospital of Philadelphia, Philadelphia, Pennsylvania, United States
6 Children's Hospital of Philadelphia, Philadelphia, Pennsylvania, United States
7 Children's Hospital of Philadelphia, Philadelphia, Pennsylvania, United States
ABSTRACT
Background
X-linked sideroblastic anemia (XLSA), the most common inherited sideroblastic anemia, results from mutations in the erythroid-specific enzyme 5-aminolevulinate synthase 2 (ALAS2), which catalyzes the rate-limiting step in heme biosynthesis. Reduced ALAS2 activity impairs heme production, causing hypochromic anemia and mitochondrial iron accumulation in erythroblasts. Although XLSA is monogenic, how impaired heme synthesis drives mitochondrial dysfunction, altered metabolism, and ineffective erythropoiesis (IE) remains incompletely understood. The R452C mutation, the second most common cause of XLSA (~5% of cases), was used to (1) generate and characterize novel experimental models of severe XLSA to define its mitochondrial and metabolic alterations, and (2) develop an in vivo adenine base editing (ABE) therapy delivered via lipid nanoparticles (LNPs). Here we report the generation of these models, an integrated metabolomics characterization of ALAS2 deficiency, and successful in vitro base-editing correction of the ALAS2 R452C mutation.
Methods
We generated erythroid (K562) and induced pluripotent stem cell (iPSC) lines carrying the ALAS2 R452C mutation via CRISPR-mediated recombination, along with a novel mouse model carrying a humanized ALAS2 R452C allele to enable evaluation of human-specific base editing. Mutant K562 and iPSC lines were differentiated into erythroid cells and assessed for heme content, mitochondrial function (Seahorse analysis), mitochondrial iron accumulation, reactive oxygen species, membrane peroxidation (flow cytometry), and Fe-S cluster biogenesis (aconitase 2/citrate synthase [ACO2/CS] and complex I/citrate synthase [CI/CS] ratios). Cellular bioenergetics (oxidative phosphorylation and glycolysis) were measured by Seahorse metabolic flux analysis. Untargeted metabolomic profiling of early erythroid cells from Alas2-knockout mice was performed by CE-MS (OmegaScan platform), optimized for polar metabolites. To evaluate base editing efficiency, K562 ALAS2 R452C cells were transfected with plasmids encoding a single-guide RNA (sgRNA) and ABE, differentiated into erythroid cells, and assessed for editing efficiency and phenotypic rescue.
Results
CRISPR-mediated genome editing successfully generated ALAS2-R452C K562 cells, iPSCs, and mice. Conditional Alas2 deletion in murine bone marrow produced ring sideroblasts and a severe erythroid maturation block at the polychromatic erythroblast stage, coinciding with the onset of hemoglobin synthesis. Across cellular and in vivo models, ALAS2 deficiency reduced oxidative phosphorylation and electron transport chain activity, with compensatory increases in glycolysis, alongside mitochondrial iron accumulation and reduced ACO2/CS and CI/CS ratios. Metabolomic profiling of Alas2-knockout erythroid cells revealed reduced medium-chain fatty acid import and lower organic acid/carboxylate levels (succinate and citrate), consistent with reduced DMT1-mediated heavy-metal chaperoning, alongside Kennedy pathway activation (decreased choline/ethanolamine with increased phosphocholine/phosphoethanolamine), indicating phospholipid membrane remodeling. 2,3-diphosphoglycerate was markedly elevated, consistent with increased mitophagy and/or intracellular iron pooling, while antioxidant (glutathione, taurine) and osmolyte levels were unchanged, arguing against oxidative or osmotic stress and instead implicating disrupted mitochondrial membrane integrity as the primary stress mechanism, consistent with aberrant mitochondrial morphology on electron microscopy. The optimized ABE/sgRNA combination corrected the R452C mutation in up to 94% of K562 cells in vitro, restoring heme synthesis, mitochondrial function, iron homeostasis, and Fe-S cluster biogenesis.
Conclusion
These novel models and integrated metabolomic profiling establish a robust platform for studying severe XLSA, showing that ALAS2 deficiency disrupts mitochondrial bioenergetics, membrane phospholipid composition, and metabolic homeostasis during erythropoiesis, and identifying disrupted mitochondrial membrane integrity as a likely driver of IE. Building on these findings, we established a base editing therapy targeting ALAS2 R452C that restores the XLSA phenotype in vitro. We are now generating LNP containing gRNA and ABE to develop an in vivo base-editing approach in the humanized XLSA mouse model.
DISCLOSURES
Kazuhiro Noguchi, MD.,PhD.: No relevant conflicts of interest to declare.