Modeling autosomal dominant erythrocytosis due HIF-2α E398K mutation reveals novel mechanism of erythropoiesis control Free

We report family with autosomal dominant erythrocytosis, when the proband (woman, age 67 when first studied had hemoglobin 15.8 g/dL and erythropoietin (EPO) level 9 mU/mL and history of deep vein thrombosis and pulmonary emboli. Her now 47 years old son had a hemoglobin level of 20.3 g/dL and coronary thrombosis. Both have early erythroid progenitors (BFU-Es) hypersensitive to EPO,no splenomegaly, normal leukocyte and platelet numbers and normal p50 indicating normal hemoglobin oxygen dissociation. Their targeted sequencing of EPO, EPOR, EGLN1, SH2B3 and JAK2 were normal. Whole exome sequencing identified a heterozygous missense variant in exon 9 of EPAS1 gene encoding HIF-2α c.1192G>A, p.E398K; its computer modeling by multiple analytical tools were compatible with the gain-of-function EPAS1 mutation. The proband's other son (now 51 years old), also affected by erythrocytosis has not yet been studied by us; both sons have one child with normal hemoglobin levels. All three affected individuals with erythrocytosis are being treated by their providers by frequent phlebotomies.

By immunoprecipitation assay, the HIF-2α E398K variant, together with two known variants F374Y (exon 9) and M535V (exon 12), were found less ubiquitinated, suggesting a less effective protein degradation. The signal of ubiquitinated HIF-2α was 38.31%, 42.27% and 54.05% for variants E398K, F374Y and M535V when normalized to HIF-2α wild-type. The protein stability measured by a cycloheximide chase assay showed that the wild-type HIF-2α was rapidly degraded whereas E398K variant was more stable with their protein half-lives 7.44 min (WT) and 155.50 min (E398K). The E398K variant is located in the proximity of secondary hydroxylation site of HIF-2α (P405), and our results suggest that it disrupts the prolyl hydroxylation driven degradation of the HIF-2α protein and therefore leads to constitutive activation of hypoxia signaling. Upregulated expression of HIF-2 targets genes TFRC, VEGF, PDK1 and BNIP3 were detected in proposita granulocytes and BFU-Es.

Although HIF2 regulates EPO transcription (PMID:17404621), the proband and her son exhibited inappropriately normal EPO levels and had BFU-Es hypersensitive to EPO. This finding implies that gain-of-function HIF-2α may contribute to erythrocytosis through EPO-independent mechanism. Perrotta et al proposed the existence of additional, yet unidentified, molecular pathways mediating another mutated HIF-2α–driven erythrocytosis (PMID:23716564). Because of limited number of available erythroid progenitors from the propositi, we created iPSCs utilizing the CRISPR/Cas9 technology. The EPAS1^E398Kclones and control lines were differentiated to hematopoietic stem cells, characterized by flow cytometry, qPCR and colony forming assays. The purified CD³⁴⁺cells were further differentiated into the CD⁷¹⁺/CD^235a+erythroid precursors. The iPSC clones carrying the EPAS1^E398Kmutation recapitulated the erythrocytosis phenotype exhibiting increased proliferation and accelerated differentiation into erythroid progenitors compared to wild-type controls. Moreover, purified CD³⁴⁺ cells harboring the EPAS1^E398Kvariant had an enhanced capacity to generate BFU-E colonies. Transcriptomic and proteomic analysis of differentiated CD³⁴⁺ cells is currently underway to elucidate the molecular cause underlying HIF-2α E398K–driven erythropoiesis.

In summary, we report a novel EPAS1^E398Kmutation and demonstrate that gain-of-function EPAS1 mutations not only augment erythropoiesis by previously reported increased EPO production (secondary erythrocytosis) but also, as we show here, causing additional intrinsic hyperproliferative defect of erythroid progenitors (primary erythrocytosis). Whether this hyperactive HIF-2 also augments transcription of prothrombotic, inflammatory and other HIFs-regulated genes, as we found in polycythemia vera (PMID:32203583), remains to be demonstrated. We plan to determine this by studying the effect of FDA-approved HIF-2α inhibitor belzutifan in vivo and in vitro in this family, in another EPAS1-mutated family with thromboses (PMID:31289208), and in Chuvash erythrocytosis (PMID:14726398). We plan to determine whether belzutifan not only, as we anticipate, corrects erythrocytosis and its phlebotomies induced iron deficiency, but also normalizes expression of prothrombotic genes.

Grant support: Czech Science Foundation, 24-11730S and GAUK 238121 (VZ, LL) and VA Merit grants JTP.