Molecular Characterization and Novel Therapeutic Strategy for X-Linked Sideroblastic Anemia Associated with ALAS2 Missense Variants

Background: Congenital sideroblastic anemia (CSA) is an inherited anemia characterized by the presence of ring sideroblasts (RSs) in bone marrow. X-linked sideroblastic anemia (XLSA) is the most common form of CSA and is caused by germline mutations in the erythroid-specific 5-aminolevulinate synthase (ALAS2) gene. Each ALAS2 mutation has a variable effect on the protein's enzymatic function, and only half of XLSA patients are responsive to pyridoxal 5′-phosphate (PLP) treatment. Thus, novel therapeutic strategies should be explored. We have established an in vitro XLSA model in human induced pluripotent stem cell-derived erythroid progenitors (HiDEP-1) by disrupting GATA-1-binding intronic enhancer region of ALAS2, which accounts for a minor XLSA subtype with PLP refractoriness (Saito and Ono et al. MCB 2019; ASH 2018). However, evidence regarding molecular characterization of ALAS2 missense variants is lacking, presumably due to difficulties in introducing the desired missense mutations in primary erythroblasts. In this study, we optimized a protocol for introducing point mutations in human umbilical cord blood-derived erythroid progenitor (HUDEP)-2 cells and subsequently characterized XLSA models harboring ALAS2 missense variants.

Methods: Based on homology-directed CRISPR/Cas9 system, we introduced an ALAS2 mutation from arginine at amino acid residue 170, one of the XLSA hot spots, into leucine (R170L) or histidine (R170H). A twenty-nucleotide targeted genomic sequence, located upstream of protospacer adjacent motif (PAM) near ALAS2 R170, was cloned into plasmid vectors expressing Cas9 nuclease and green fluorescent protein (GFP). Each homologous recombinant template was designed as a 127 single-stranded oligodeoxynucleotide (ssODN) asymmetrically distributed from the Cas9-mediated cleavage site. PAM silent mutations were added to avoid continuous cleavage after successful mutation. HUDEP-2 cells were co-transduced with gRNA/Cas9/GFP-expressing vectors and ssODNs using Amaxa Nucleofector 2b (Lonza). After 48 hours, cells were isolated by GFP-positive expression; subsequently, single-cell dilution was performed, followed by clonal analysis.

Results: Both XLSA clones (ALAS2 R170L and R170H) appeared pink/pale colored, reflecting impaired hemoglobin biosynthesis. Consistently, quantitative RT-PCR analysis demonstrated decreased globin gene (HBB) expression. On the other hand, XLSA clones did not show RSs and were morphologically similar to wild-type controls. When XLSA clones were induced to undergo erythroid differentiation by co-culturing with OP9 cells in a 100 mM sodium ferrous citrate (SFC)-supplemented medium, increased numbers of RSs were observed in mutant clones, mimicking findings in XLSA patients. Electron microscopy confirmed an aberrant mitochondrial iron deposit in XLSA clones. Also, HBB expression and intracellular heme concentration were significantly lower than those in wild-type controls. Intriguingly, while OP9 co-culture and SFC addition promoted erythroid differentiation in both wild-type and XLSA clones, XLSA clones exhibited a more immature morphological phenotype than wild-type controls. Expression profiling revealed that 317 and 86 genes were commonly up- and downregulated >2-fold in XLSA clones compared with those in wild-type controls. Gene ontology analysis showed significant (p < 0.01) enrichment of genes associated with mitochondrial gene expression and organization, suggesting that mitochondrial alteration is involved in XLSA pathogenesis. On the other hand, 188 genes in total were differentially expressed between ALAS2 R170L and R170H. Genes associated with ferroptosis, such as GPX4 and GCLC, showed a significantly lower expression in ALAS2 R170H clones, possibly accounting for phenotypic differences among patients with ALAS2 R170L or R170H mutation. Finally, supplementation with 5-aminolevulinic acid (ALA) significantly improved the compromised heme biosynthesis in XLSA clones, suggesting ALA treatment to be a promising therapeutic option for XLSA.

Conclusion: The established models may act as useful tools for exploring the precise molecular mechanisms of XLSA harboring missense mutations and for drug testing. Also, our homology-directed CRISPR/Cas9-based protocol can be applied to establish a wide variety of CSA as well as congenital anemia models.