Erythroid-specific, inducible CDAN1∆/∆ mouse model phenocopies congenital dyserythropoietic anemia type-la

Abstract Congenital dyserythropoietic anemias (CDAs) are a heterogenous group of inherited blood diseases, characterized by ineffective erythropoiesis, hemolysis, secondary iron overload, and the presence of bi- or multinucleated erythroblasts (EBs) in the bone marrow (BM). Several CDA types exist which are classified based on their causative gene mutations, when known. CDA type-Ia is caused by biallelic pathogenic variants in the CDAN1 gene, which encodes Codanin-1, a highly conserved protein. There are no known patients with complete loss of CDAN1, suggesting that total Codanin-1 deficiency is incompatible with life. Patients with CDA-Ia frequently require transfusions in utero to survive and present with varying degrees of hemolytic anemia with relative reticulocytopenia, suggesting ineffective erythropoiesis. BM pathology reveals erythroid hyperplasia with 3-10% binucleated EBs, the rare but characteristic finding of internuclear chromatin bridges between nearly-completely separated EBs, and, under transmission electron microscopy (TEM), a “Swiss cheese” appearance in the nuclear heterochromatin of erythroid precursors. Many patients with CDA-Ia require blood transfusions and iron chelation, while treatment with IFN-α is also an option. Most patients treated with IFN-α achieve transfusion independence and appear to have alleviated iron overload, but the mechanism by which IFN-α improves the Codanin-1 deficient erythropoiesis is still unknown. Previous efforts to make an erythroid-specific deletion of CDAN1 in mice utilized the constitutive erythroid-specific Epor-cre recombinase; this model was embryonal lethal by mid-gestation (E12.5–E13.5) due to severe anemia, allowing limited in vivo studies to investigate the role of CDAN1 in erythropoiesis (Noy-Lotan S. et al, 2021 Front Physiol). To circumvent the embryonal lethality caused by Codanin-1 deficiency during fetal development and its associated stress erythropoiesis, we used the erythroid-restricted, inducible Gata1creERT2 mouse model (Yu L. et al, 2021 Blood) to drive deletion of CDAN1 in adult mice, with the goal to investigate the pathological mechanisms of Codanin-1 deficiency in erythropoiesis and the molecular effects of IFN-α as a therapeutic intervention. We developed CDAN1flox/flox mice utilizing the CDAN1 gene trapped (tm1a) mouse, obtained by the Mutant Mouse Resource & Research Centers (MMRRC), and bred them, as well as CDAN1WT/WT mice, with Gata1creERT2 mice. We induced CDAN1 deletion via weaning 4-week-old Gata1creERT2; CDAN1flox/flox and Gata1creERT2; CDAN1WT/WT mice onto tamoxifen chow; mice were studied after 2 to 3 months on this diet. Whole blood, hind limbs, and spleens were collected for downstream analyses such as CBCs, flow cytometry, and cell sorting . We found that CDAN1∆/∆ mice had a 75 ± 16% reduction of CDAN1 mRNA expression compared to controls from magnetically separated TER119+ cells. CDAN1∆/∆ mice developed a macrocytic anemia and splenomegaly. Flow cytometry of CDAN1∆/∆ EBs revealed a reduction in late-stage EBs, compared to controls, not only in the BM, but also in the spleen, maintaining the dyserythropoietic phenotype even in extramedullary erythropoiesis, which failed to compensate for the anemia. In CDAN1∆/∆ mice, we found binucleated EBs and chromatin bridges, characteristic pathological findings of CDA-Ia. Under TEM, we observed the “Swiss cheese” phenotype in the heterochromatin of BM and splenic EBs, the hallmark finding of CDA-Ia. In TER119+ cells from the BM and spleen, we observed increased erythroferrone mRNA expression. We assayed liver nonheme iron content via a colorimetric assay (Torrance J.D. & Bothwell T.H., 1980, Churchill Livingstone) and observed an increase in liver nonheme iron stores, suggesting the onset of secondary iron loading. After flow sorting for the different populations of BM erythroblasts, we performed assay for transposase accessible chromatin (ATAC) with high-throughput sequencing per each erythroblast population in CDAN1∆/∆ versus CDAN1WT/WT mice while on long-term tamoxifen induction; our preliminary analysis suggests differential gene accessibility between control and CDAN1∆/∆EBs. In conclusion, we have generated a reliable model of CDA-Ia which we use as a platform to investigate the role of Codanin-1 in chromatin organization and condensation during erythropoiesis and how IFN-α may exert a therapeutic effect on patients with CDA-Ia.