The Dominant Negative Acting VPS4AE228Q Mutation Causes Dyserythropoiesis in Human iPSC-Derived Erythroblast Cultures That Phenocopies Cimdag Syndrome-Associated Anemia

Vacuolar protein sorting 4A (VPS4A) is a member of the AAA protein family (ATPases Associated with diverse cellular Activities) and a necessary component of the Endosomal Sorting Complex Required for Transport (ESCRT)-III membrane scission machinery which performs many essential cellular processes including the abscission step of cytokinesis, vesicle budding during endo-lysosomal transport, and nuclear envelope repair. Dominant-negative (DN) mutations affecting the ATPase domain of VPS4A cause the multi-system syndrome CIMDAG (Cerebellar hypoplasia and cataracts, Impaired intellectual development, congenital Microcephaly, Dystonia, Anemia, and Growth retardation). The penetrance of the hematologic phenotype is incomplete; some patients with CIMDAG present with severe, transfusion-dependent hemolytic and dyserythropoietic anemia with iron overload indicating ineffective erythropoiesis, although occasionally with brisk reticulocytosis, while others have mild or subclinical anemia. Despite its ubiquitous role in cellular functions, the precise mechanism by which DN VPS4A mutations affect erythropoiesis is not known, nor is the cause of reduced penetrance of the anemia phenotype which may involve compensatory mechanisms.

To determine if DN VPS4A is sufficient to produce the erythropoiesis defects observed in patients, we introduced a rationally designed DN mutation, VPS4A^E228Q, into healthy donor derived iPSCs using CRISPR gene editing and evaluated their differentiation and maturation into erythroblasts alongside patient-derived and healthy donor control iPSCs. VPS4A^E228Q mutant protein lacks ATP hydrolysis activity and exerts a dominant-negative effect on wild-type VPS4 proteins. The VPS4A^E228Q mutant iPSCs largely phenocopied the patient-derived lines, exhibiting slower growth and producing fewer GPA+ erythroid progenitors and precursors. Similar to patient erythroblasts, VPS4A^E228Q erythroblasts differentiated more rapidly compared to controls, and had increased frequency of binucleated erythroblasts, dysplastic nuclei, and cells joined by cytoplasmic bridges. Erythroblasts from VPS4A^E228Q mutant iPSCs also had increased abundance of some integral membrane proteins such as glycophorin A (GPA/CD235a) and CD71, which was also observed in patient-derived iPSC cultures and primary patient blood samples. To identify the pathways perturbed in VPS4A mutant erythroblasts, we performed single cell RNA sequencing on erythroblasts cultured from VPS4A^E228Q mutant, patient, and control iPSCs. Preliminary analysis yielded at least 8 distinct cell clusters, 5 of which represent different erythroblast states. Genes involved in heme metabolism, ribosome synthesis, and endosomal transport were among the differentially expressed genes in both VPS4A^E228Q and patient-derived erythroblasts compared to the controls, consistent with previous RNA sequencing data from CD71+ reticulocytes isolated from patient blood samples. Taken together, these data support the conclusion that the DN VPS4A^E228Q mutation is sufficient to cause dyserythropoiesis and phenocopies the mutations found in patients with DN-acting VPS4A variants. The patient and CRISPR iPSC models developed provide a valuable resource for further investigating the role of VPS4A and ESCRT-III in hematopoiesis and possible compensatory molecular mechanisms.