Inducible Knockout of Codanin-1: An Adult Mouse Model of Congenital Dyserythropoietic Anemia Type-I

Congenital dyserythropoietic anemias (CDAs) are a heterogenous group of congenital blood diseases, characterized by ineffective erythropoiesis, hemolysis, and the presence of bi- or multi-nucleated erythroblasts in the bone marrow (BM). Several CDA types exist, originally classified by bone marrow pathology and now more accurately by their causative mutations, when known. Mutations in codanin-1 ( Cdan1) are the major contributor (≈80%) to CDA type-I (CDA-I). Patients with CDA-I frequently require transfusions in utero to survive and present with varying degrees of hemolytic anemia with relative reticulocytopenia suggesting ineffective erythropoiesis, binucleated erythroblasts, and, under transmission electron microscopy (TEM), a “swiss cheese” appearance in the nuclear heterochromatin of erythroblasts. Treatment for patients with CDA-I includes blood transfusions and iron chelation, or IFN-α, which allows most treated patients to achieve transfusion independence. However, the molecular mechanisms of dyserythropoiesis in CDA-I as well as those of the therapeutic effect of IFN-α are unknown. To study the pathogenesis of the disease and the mechanisms of IFN-α action in vivo, we generated an adult mouse model of CDA-I. Previous efforts to produce a CDA-I mouse model were achieved via a constitutional, EpoR ^Cre+/--driven erythroid-specific deletion of Cdan1; however, this mouse suffered embryonic lethality by E12.5 [Noy-Lotan et al, Frontiers in Physiology, 2021]. We developed a transgenic mouse model in which the most common CDAN1 missense mutation (p.R1042W in humans corresponding to mouse p.R1047W) was introduced using CRISPR-Cas9 editing, and the mice were bred to homozygosity. These mice survived to term but exhibited no significant anemia, altered erythropoiesis, or the characteristic “swiss cheese” appearance in the heterochromatin of the erythroblasts under TEM. They demonstrated no differences from control mice in recovery after induction of stress erythropoiesis by phlebotomy (in agreement to the findings reported in parallel by Noy-Lotan et al, 2021), suggesting that this model is inadequate for studying the CDA-I phenotype. Therefore, we proceeded to generate an inducible Cdan1 knockout mouse model to test the hypothesis that deletion of Cdan1 in the erythroid lineage,mediated by Gata1 ^creERT2 [Yu et al. Blood. 2021], will recapitulate the CDA-I phenotype in adult mice. This maneuver avoids embryonal lethality during the extreme stress erythropoiesis phase of fetal development. We bred mice with the wildtype (WT) or floxed Cdan1 gene ( Cdan1^WT/WT and Cdan1^Fl/Fl, respectively) with mice carrying Gata1 ^creERT2. Cdan1^WT/WT; Gata1^creERT2+/- mice served as WT controls and Cdan1^Fl/Fl; Gata1^creERT2+/- mice served as the experimental mice with erythroid-specific Cdan1 deficiency after 10 injections of tamoxifen every other day to activate Cre expression. We evaluated the phenotype of these mice using complete blood counts (CBCs), light microscopy, TEM, and flow cytometry of erythroid progenitors and precursors. The Cdan1-deficient mice became anemic with reduction in RBC count, hemoglobin ( Fig. 1), and hematocrit. Light microscopy of the BM and spleen revealed erythroblasts joined by a thin-interchromatin bridge and binucleated erythroblasts in the Cdan1-deficientcompared to WT mice (z-test of proportions, P < 0.001), appearing at a frequency of ≈13% of all erythroblasts (see Fig. 2). TEM revealed a “swiss cheese” appearance in the heterochromatin of erythroblasts, the pathognomonic finding of CDA-I. Flow cytometry of the BM and spleen revealed decreased late-stage erythroblasts, demonstrating ineffective erythropoiesis. To determine a possible stage in which IFN-α may be directly acting on erythropoiesis, we examined the expression of IFNαR1 in erythroblasts. In all mice, basophilic, polychromatic, and orthochromatic erythroblasts had relatively low expression of IFNαR1. Interestingly, proerythroblasts had relatively high expression of IFNαR1 but only in the spleen. These data provide a starting point for localizing the effect of IFNα in the CDA-I phenotype. In conclusion, erythroid-specific deletion of Cdan1 via Gata1 ^creERT2 in adult mice recapitulates the CDA-I phenotype and provides a viable mouse model to investigate the disease mechanism and the therapeutic effect of IFN-α on CDA-I.