DDX41 Dissolves G-Quadruplexes to Maintain Erythroid Genomic Integrity and Prevent Genome DNA Leakage-Induced Cgas Activation

DDX41 is one of the most frequently mutated genes that are associated with familial myeloid neoplasms. To understand the pathogenesis of DDX41 mutations, which often lead to loss of function, we developed various mouse models with lineage-specific deletion of Ddx41. We found embryonical lethality with the hematopoietic specific knockout (KO) of Ddx41 ( VavCre:Ddx41 ^fl/fl). Examination of E13.5 fetuses showed that Ddx41 deficient fetal livers are pale with marked apoptosis, indicating defects in erythropoiesis. Indeed, depletion of Ddx41 in the early stage of erythropoiesis ( EpoRCre:Ddx41 ^fl/fl) also led to embryonic death. However, mice with Ddx41 KO in the late-stage erythropoiesis ( HBBCre:Ddx41 ^fl/fl) survived with normal complete blood count, indicating a stage-specific requirement of Ddx41 in erythropoiesis. We further revealed that mice with Ddx41 KO in monocytes ( LysMCre:Ddx41 ^fl/fl), granulocytes ( MRP8Cre:Ddx41 ^fl/fl), and dendritic cells ( CD11cCre:Ddx41 ^fl/fl) also showed no obvious phenotypes, suggesting that Ddx41 is dispensable in other myeloid cells in mice.

We further discovered that Ddx41 KO in erythroid cells induces the accumulation of G-quadruplexes (G4), a noncanonical DNA structure that is often associated with active replication. We found that erythroid cells contain significantly higher G4 signals. G4 accumulation reaches its peak level in the CFU-Es and proerythroblasts, which is likely due to replication stress at this stage of erythropoiesis and consistent with the lethality of EpoRCre:Ddx41 ^fl/flmice. A CUT&Tag analysis revealed genome-wide colocalization of Ddx41 and G4 in hematopoietic stem and progenitor cells, which became further enhanced in erythroid cells. We confirmed the direct binding of Ddx41 to G4, which is sequence independent. A fluorescence resonance energy transfer (FRET) assay demonstrated that Ddx41 binds to and dose-dependently dissolves G4. These findings indicate a critical role of DDX41 in erythropoiesis by maintaining the level of G4 and genomic integrity.

Consistent with the knowledge that increased level of G4 induces genomic instability, we found an increased level of γ-H2AX in Ddx41 KO erythroid cells or cells treated with pyridostatin (PDS), a well-known G4 stabilizer. One unique feature of developing erythroblasts is the dynamic nuclear opening with the release of genomic materials, which is critical for nuclear condensation. However, this process also poses potential risks of triggering the cGAS-STING pathway with the increased damage-associated molecular patterns (DAMPs), particularly under situations with increased genomic instability. Indeed, we found an increased level of cGAS and STING in Ddx41 KO and PDS-treated erythroid cells. Consistently, erythroblasts from cGAS KO mice are resistant to PDS treatment mediated cell death. The significance of the cGAS pathway was genetically confirmed in that cGAS KO completely rescued the lethality of VavCre:Ddx41 ^fl/fl mice. In addition, we demonstrated that Ddx41 KO and PDS treatment-mediated activation of the cGAS-STING pathway induces an increased level of interferon, which results in ribosome biogenesis defect, p53 upregulation, and cell apoptosis. Notably, cGAS KO in VavCre:Ddx41 ^fl/fl mice counteracted these effects and rescued erythroid ribosomopathy, although the level of G4 remains upregulated.

Overall, our study reveals a critical role of DDX41 in dissolving G4 and maintaining the genomic integrity of the erythroid cells. Our data also indicate that defect in the erythroid lineage through ribosomopathy is one of the major pathologic features in DDX41 mutated myeloid neoplasms, which is important for the development of novel therapies.