Interconnections between DDX41-Dependent Post-Transcriptional and GATA Factor-Instigated Transcriptional Mechanisms That Govern Hematopoiesis

Heterozygous DDX41 germline mutations occur in familial myelodysplastic syndrome (MDS)/acute myeloid leukemia (AML) and acute erythroid leukemia (AEL). DDX41 encodes an RNA helicase that regulates RNA splicing, cGAS-Sting signaling in innate immunity, and maintains genome stability in erythropoiesis. We described the innovation of a facile genetic rescue system to compare RNA-regulatory activities of wild type and disease-associated DDX41 variants in engineered HoxB8-immortalized (hi) murine myeloid progenitor cells (hi-Ddx41+/+ and Ddx41+/- cells) (Kim et al., Leukemia 2023). We identified DDX41-regulated transcripts and membrane proteins as quantitative metrics of DDX41 activity that discriminate benign from pathogenic DDX41 alleles. As these transcripts and the cognate proteins were not known to be components of GATA2 genetic networks that govern hematopoietic stem/progenitor cell (HSPC) transitions, we hypothesized that DDX41 mechanisms in progenitors operate in parallel with or independent of GATA2 mechanisms. Considering the biological and pathological importance of DDX41 and many unanswered questions on how clinical variants impact DDX41 activity, we instituted studies to discover DDX41-regulated transcript isoforms. The datasets are being used to unveil how DDX41 controls cellular functions, how clinical variants derail mechanisms, and to establish interconnections or independence with GATA factor mechanisms.

To develop a broader perspective, we extended the myeloid data by innovating a rescue system to analyze DDX41 structure/function in G1E-ER-GATA1 erythroblasts that express a conditionally active ER-GATA1 allele. We used CRISPR/Cas9 to generate Ddx41^+/- clonal lines that expressed ~50% lower levels of DDX41 protein. A comparable system is being generated in human HUDEP2 erythroid precursor cells. We compared DDX41 with two disease-associated germline variants: G173R and K331deletion. G1E-ER-GATA1-Ddx41^+/-cells were infected with retrovirus expressing GFP, or GFP and DDX41 (WT, G173R, or K331del), GFP⁺ cells were sorted, and total RNA was analyzed by RNA-seq. Differential expression analysis revealed 138 Ddx41-regulated transcripts (P < 0.05), with 63 elevated and 75 decreased. Disease-associated variants failed to regulate these transcripts. To establish relationships between DDX41-regulated and GATA1-regulated transcripts in erythroblasts, we compared DDX41-regulated transcripts with our GATA1 transcriptomic data (Tanimura et al., EMBO Rep. 2016; Dev. Cell 2018; Liao et al., Cell Rep. 2020), which revealed 78 transcripts co-regulated by DDX41 and GATA1. The RNA analysis is being extended to the proteome.

To investigate similarities between DDX41-dependent post-transcriptional and GATA1-dependent transcriptional mechanisms, we analyzed the genes encoding the 78 co-regulated transcripts using gene ontology and pathway analyses. The gene signatures included downregulation of apoptosis (p=0.0006). We are testing the hypothesis that DDX41 and GATA1 operate additively or synergistically to control levels of transcripts encoding apoptotic regulators in developing erythroblasts, thus establishing a balance between pro-apoptotic and anti-apoptotic factors. We compared myeloid and erythroid DDX41-regulated transcript datasets to identify common and cell type-specific DDX41-regulated transcripts. This analysis unveiled 11 transcripts that are DDX41-regulated in both systems. One transcript encodes CLK3, a member of Cdc2-like kinase family that phosphorylates splicing factors. DDX41 increased intron 3 retention in Clk3 mRNA in both systems. By quantifying exon-intron junctions, we demonstrated that DDX41 promoted intron 3 retention in hi-Ddx41^+/- cells 1.7-fold (P = 0.0002) and in G1E-ER-GATA1- Ddx41^+/- cells 3.3-fold (P = 0.0002). We are testing models to explain the functional significance of DDX41-regulated Clk3 intron retention and discovered DDX41 regulates intron retention at other loci. We are assessing how DDX41 controls alternative splicing globally in the erythroid and myeloid systems and elucidating the biological and pathological consequences of corrupting these mechanisms. In aggregate, these studies established a system to discriminate pathogenic from benign DDX41 variants, and mechanistic insights that provide the basis for new myeloid and erythroid paradigms.