AUTHORS
Inci Aydemir, MD, Honghao Bi, MD, PhD2, Pan Wang, PhD3, Wenjie Cai4, Julia Galyayeva, BS5, Renee Ma6, Salina Dominguez7, Yijie Liu, PhD8, Madina Sukhanova9, Lucy Godley, MD, PhD10, Peng Ji, MD, PhD11
AFFILIATIONS
1 Northwestern University, Chicago, Illinois, United States
2 Northwestern University, Chicago, Illinois, United States
3 Northwestern University, Chicago, Illinois, United States
4 Northwestern University, Chicago, Illinois, United States
5 Northwestern University, Chicago, Illinois, United States
6 Northwestern University, Chicago, Illinois, United States
7 Northwestern University, Chicago, Illinois, United States
8 Northwestern University, Chicago, Illinois, United States
9 Northwestern University, Chicago, Illinois, United States
10 Northwestern University, Chicago, Illinois, United States
11 Northwestern University, Chicago, Illinois, United States
ABSTRACT
Background: Germline mutations in DDX41 and CHEK2 are strongly implicated in bone marrow failure syndromes, myelodysplastic neoplasms (MDS), and acute myeloid leukemia (AML). However, understanding how these mutations disrupt the complex 3D architecture of the human hematopoietic niche has remained elusive due to the limitations of conventional models. In this study, we utilized a patient-cell reprogrammed induced pluripotent stem cells (iPSC)-derived bone marrow organoid model to map the spatial, transcriptomic, and secretomic consequences of these mutations, uncovering their allele-specific regulatory role of and their genetic interaction.
Methods: Peripheral blood mononuclear cells (PBMCs) from patients with CHEK2 only, DDX41 only, and DDX41+CHEK2 (double mutant) germline mutations were reprogrammed into iPSCs and differentiated into bone marrow organoids. To achieve a mechanistic map, we integrated: 1) single-cell RNA sequencing (scRNA-seq) to quantify lineage commitment, 2) Luminex multiplex assays to profile the secretome, 3) 10x Xenium spatial transcriptomics to evaluate 3D niche architecture. CRISPR-Cas9 DDX41 knock-out models were generated to assess the impact of allelic loss on spatial compartmentalization.
Results: scRNA-seq revealed distinct lineage commitment and cellularity alterations in germline-mutated bone marrow organoids. In organoids derived from a normal individual, hematopoiesis was characterized by a dominance of late erythroid cells alongside normal myeloid differentiation. The CHEK2 only mutation showed a significant reduction in early erythroid cells, accompanied by an aberrant and prominent myeloid expansion. Conversely, the DDX41 only mutation maintained a cellular profile close to control. Germline DDX41+CHEK2 mutation reversed the CHEK2-driven myeloid expansion, but erythroid defects remained. We also generated DDX41-KO iPSCs via CRISPR. Organoids derived from these cells showed severe cellular depletion, as evidenced by profound hypoplasia at the single-cell level. Instead, heterozygous DDX41 organoids showed a milder phenotype, similar to that observed in DDX41-only germline patient-derived organoids.
Multiplex secretome profiling elucidated the cytokines and growth factors orchestrating these cellular phenotypes. CRISPR-engineered DDX41 knockouts drove a broad, robust activation across pro-inflammatory mediators and chemokines with no differential expression of growth factors, compared to control organoids. The organoids with DDX41-only germline mutation maintain a similar overall baseline to control organoids but drive targeted upregulation of a distinct stromal and chemokine cluster, led most prominently by FGF-2 and IL-1α. Strikingly, the DDX41+CHEK2 double mutant completely abolishes this DDX41-driven secretome signature, inducing deep repression of these growth factors while uniquely upregulating MDC.
High-resolution spatial transcriptomic profiling (Xenium 5K) revealed distinct architectural and lineage shifts across genotypes. While control organoids maintained structured, dense erythroid niches, homozygous DDX41 knockout led to near-total ablation of erythroblastic islands and marked stromal dominance. Heterozygous DDX41 knockout displayed spatial disorganization with fragmented lineage compartmentalization, whereas CHEK2-only organoids exhibited a unique spatial phenotype characterized by focal megakaryocytic expansion and altered erythroid distribution.
Conclusions: Integrated scRNA-seq, secretome, and spatial transcriptomics of patient iPSC-derived organoids reveal distinct niche remodeling by DDX41 and CHEK2 mutations. Homozygous DDX41 loss causes severe hypoplasia, stromal dominance, and pro-inflammatory signaling. In contrast, germline DDX41 mutations preserve baseline cellularity while selectively driving an FGF-2/IL-1α secretory module. CHEK2 mutations primarily trigger aberrant myeloid expansion and focal megakaryocytic clustering. Strikingly, the DDX41+CHEK2 double mutant abolishes both CHEK2-driven myeloid expansion and DDX41-induced growth factor upregulation, demonstrating clear genetic epistasis in the human 3D niche.
DISCLOSURES
Inci Aydemir, MD: No relevant conflicts of interest to declare.