Modelling Congenital Dyserythropoietic Anemia Type II through Gene Editing in Hematopoietic Stem and Progenitor Cells

Congenital dyserythropoietic anemias (CDAs) are a group of inherited anemias that affect the development of the erythoid lineage. They are characterized by ineffective erythropoiesis with distinct morphologic abnormalities of erythroblasts, a degree of hemolysis, and secondary hemochromatosis. Patients usually present with congenital anemia, jaundice, splenomegaly, and an absolute reticulocyte count inadequate for the degree of anemia. CDA type II (CDAII) is the most frequent type. CDAII patients show anemia of variable degrees, and 20% are transfusion dependent. The most specific finding of CDAII marrow is the presence of more than 10% mature binucleated erythroblasts with two nuclei at the same erythroid maturation stage. Treatment of CDAII patients may involve blood transfusions, iron chelation therapy and splenectomy. The only described definitive therapy is allogeneic bone marrow transplantation, which implies additional side effects for these patients. Therefore, new therapeutic approaches are needed.

CDA II is caused by mutations in the SEC23B gene. SEC23B is part of coat protein complex II (COPII). COPII is involved in protein processing and Golgi-reticulum trafficking. However, how mutations in SEC23B cause CDA II is not known yet. Therefore, studying the role of SEC23B in the erythropoiesis will help to elucidate the underlying mechanism of CDA II and to develop new therapeutic approaches for the disease.

We have developed a CDA II model in human cells through the introduction of genomic mutations in the gene using the CRISPR/Cas9 gene editing system. Different single guides RNAs (sgRNA) targeting the start of the coding sequence of human SEC23B gene were designed and tested in human erythroleukemia K562 cell line and in healthy human cord blood hematopoietic stem and progenitors (hCB-CD34⁺). The gene editing outcome at SEC23B gene was assessed at: i) genomic level through Sanger sequencing, Inference of CRISPR Edits (ICE) and Next-Generation Sequencing (NGS). ii) Protein level through Western-blot analysis. iii) Functional level through morphological analysis and erythroid differentiation either in vitro or in vivo in human hematopoietic chimeras in NOD.Cg-Kit^W-41JTyr⁺Prkdc^scidIl2rg^tm1Wjl/ThomJ (NBSGW) mice.

K562 cells were nucleofected with three different sgRNAs, as ribonucleoprotein (RNP), independently or in combination. Afterwards, seventy five K562 clones were established from the cells nucleofected with the most efficient sgRNA or with the combination of the three sgRNAs. Forty per cent of them showed a high efficiency of knock-out (higher than 50% of alleles). Eight SEC23B^KO clones were selected for further analysis. All of those eight clones showed a reduction in SEC23B protein and six of them had a lower proliferation than control cells and morphological abnormalities, such as presence of bi/multinucleated cells. Moreover, when CB-CD34⁺ cells were nucleofected with the most efficient sgRNA or with the combination of the three sgRNAs, up to 80% of knock-out efficiency and close to 90% reduction of SEC23B protein were obtained. Interestingly, when those gene edited hematopoietic progenitors were differentiated in vitro to erythroid cells, their terminal differentiation was hampered, with a reduce percentage of enucleated cells and the presence of high number of bi/multinucleated cells. Similarly, the in vivo erythroid differentiation of these gene edited progenitors two months after HSPC transplant into NBSGW mice showed again an impairment of terminal erythroid differentiation with an increment in the percentage of erythroid bi/multinucleated cells without altering other human hematopoietic lineages.

In summary, CRISPR/Cas9 system has been used to model CDA II in a human cell line and in human hematopoietic progenitors through the knock-out of SEC23B gene. Our system reproduced the most relevant feature characteristic of CDA II pathology. This gene editing based CDA II model will allow the study of how mutations in SEC23B cause CDA II and the development of new therapeutic strategies to cure this disease.