Abstract
There is currently no standardized treatment regimen established for Congenital dyserythropoietic anemia (CDA), supportive care and HSC transplantation remain the mainstream strategies. Here, we demonstrated that luspatercept, an ActRIIB ligand trap, rescues the disease phenotype in one CDAII patient and its effect verified in an in vitro model. Our work suggests a potential treatment method for CDAII with a clinically approved product.
A CDAII patient confirmed by whole-exome sequencing to have a SEC23B gene mutation received luspatercept at a dose of 1 mg/kg once every 3 weeks. After 3 cycles of treatment, the hemoglobin level increased from 76 g/L to 102 g/L, achieving hematological improvement-erythroid (HI-E).
To further assess the effect of luspatercept on erythroid development, we constructed a patient-derived iPSC line for in vitro erythroid culture using STEMdiff™ Erythroid Kit. High-purity CD34+CD43 progenitor cells (CON 59.9% vs CDAII 56.9%) were harvested. The patient cells exhibited high proliferative tendency (CON 0.94±0.15×10^6 vs CDAII 2.35±0.23×10^6, p=0.0040). Next, cells were seeded at 4×10^4 cells/well (CON, CDAII, CDAII+lus, 4 wells/group, data reported in such order on forth), followed by a fourteen-day erythroid induction culture period (day 7 replating at 1.6×10^5/well) with a luspatercept concentration of 20μg/ml.
At day 7, flow cytometry showed rapid erythroid lineage expansion of the CDAII samples, (CD71+GPA+): 11.6% vs 57.0% vs 54.3%, indicating the characteristic ineffective erythropoiesis of CDAII. We further analyzed erythroid progenitor markers for BFU-E and CFU-E forming cells, (CD45+GPA-CD123-CD34+CD36-CD71low) and (CD45+GPA-CD123-CD34-CD36+CD71high), respectively. The results indicated an unbalanced progenitor cell pattern compared to CON, which was partially adjusted with luspatercept (BFU-E forming: 1.9% vs 1.6% vs 1.3%; CFU-E forming: 2.5% vs 1.2% vs 1.4%; Ratios CFU-E/BFU-E: 1.31 vs. 0.75 vs. 1.07). At day 14, CON showed increased (CD71+GPA+) subsets while the CDAII groups' subsets deceased: 45.1% vs 49.1% vs 46.3%. The cell counts at day 7 and day 14 are (0.75 vs 1.25 vs 1.05)×10^6 and (0.8 vs 0.45 vs 0.6)×10^6 cells/well, respectively.
We next performed scRNA-seq at each time point. At day 7, both CDAII and CDAII+lus samples presented huge clusters of immature early erythroblasts, while their CON counterpart was only a small fraction (44.1% vs 47.3% vs 3.2%). Downstream, the intermediate erythroblast subsets were both missing from the CDAII and CDAII+lus groups (CON 9.0%) but the late erythroblast cluster remained (3.0%, 3.1% vs CON 1.8%). At day 14, the intermediate and late erythroblast clusters continued to expand in CON cells (15.0%, 16.3%). As a result of ineffective erythropoiesis, the huge cluster of early erythroblasts in CDA cells almost disappeared (<1%). In contrast, the CDAII+lus early erythroblast cluster remained (25.3%).
Further, gene differential expression analysis revealed significantly downregulated GATA1 expression, a key erythroid lineage regulating transcription factor, in erythroid progenitors (avg_lg2FC CDAII vs CON -1.20, p=1.26e-13; CDA+LUS vs CON -0.82, p=9.1e-7). Interestingly, with luspatercept, both GATA1 expression level and GATA1-expressing-cell percentage increased (avg_lg2FC CDAII vs CDAII+lus 0.37, p=0.003; ratio 59% vs 36% vs CON 77%). Moreover, HBB expression maintained higher levels in CDAII+lus cells at d7 in late erythroblast clusters (avg_lg2FC vs CDAII 0.48, p=0.004; vs CON 3.31, p=1.5e-33) and similarly at d14 (avg_lg2FC vs CDAII 0.54, p=0.4; vs CON 0.39, p=1.7 p=0.84).
In addition, lncRNA SNHG31, SLC351, and SLC26A3 showed consistently higher (Always top 5 upregulated genes) expression in CDAII and CDAII+lus cells vs CON specifically across erythroid clusters at all time points. These were previously reported not to be detected in normal in vitro erythroid differentiation (An et al. Blood. 2014), indicating another potential effect of SEC23B mutation on erythroid development.
Taken together, our findings demonstrated luspatercept's rescue effect on the patient, offering a potential treatment strategy for CDAII patients. The in vitro analysis model suggests that such effect may be achieved by the increased GATA1 expression level in erythroid progenitors. However, the downstream effect on HBB and other erythroid lineage related genes requires more refined exploration.
