Eltrombopag Improves Erythroid Differentiation in a Human Induced Pluripotent Stem Cell Model of Diamond Blackfan Anemia

Abstract

Diamond Blackfan Anemia (DBA) is a congenital macrocytic anemia associated with ribosomal protein haploinsufficiency. Ribosomal dysfunction delays globin synthesis, resulting in excess toxic free heme in erythroid progenitors, early differentiation arrest, and pure red cell aplasia. In this study, DBA induced pluripotent stem cell (iPSC) lines were generated from blood mononuclear cells of DBA patients with inactivating mutations in RPS19 and subjected to hematopoietic differentiation to model disease phenotypes. In vitro differentiated hematopoietic cells were used to investigate whether eltrombopag, an FDA-approved mimetic of thrombopoietin with robust intracellular iron chelating properties, could rescue erythropoiesis in DBA by restricting the labile iron pool (LIP) derived from excessive free heme. DBA iPSCs exhibited RPS19 haploinsufficiency, reduction in the 40S/60S ribosomal subunit ratio and early erythroid differentiation arrest in the absence of eltrombopag, compared to control isogenic iPSCs established by CRISPR/Cas9-mediated correction of the RPS19 point mutation. Notably, differentiation of DBA iPSCs in the presence of eltrombopag markedly improved erythroid maturation. Consistent with a molecular mechanism based on intracellular iron chelation, we observed that deferasirox, a clinically licensed iron chelator able to permeate into cells, also enhanced erythropoiesis in our DBA iPSC model. In contrast, erythroid maturation did not improve substantially in DBA iPSC differentiation cultures supplemented with deferoxamine, a clinically available iron chelator that poorly accesses LIP within cellular compartments. These findings identify eltrombopag as a promising new therapeutic to improve anemia in DBA.

Trial registration: ClinicalTrials.gov NCT00027274.

Keywords: diamond blackfan anemia; disease modeling; drug testing; eltrombopag; genome editing; induced pluripotent stem cells.

Figure 1

Generation and characterization of isogenic and DBA iPSCs. (A) Genomic RPS19 sequence trace chromatograms for isogenic (top panel) and DBA863 (bottom panel) iPSCs to confirm CRISPR/Cas9-mediated correction (grey highlight) of heterozygous c.185G>A (pR62Q) nonsense mutation (green highlight). In isogenic iPSCs, the overlapping G/C bases represent a silent mutation introduced in the template DNA during genome editing to complete an AscI restriction site (GGCGCGCC) for rapid RFLP screening of RPS19 gene corrected iPSC clones. (B) Western blot for RPS19 and β-tubulin loading control for wild-type (WT), DBA and isogenic iPSCs. (C) Sucrose gradient (10–50%) polysome profiling analyses for isogenic and DBA iPSCs. Peaks represent ribosomal subunits (40S and 60S), 80S monosomes, and ribosome clusters (polysomes). (D) Calculated 40S/60S ratios for isogenic and DBA iPSCs. Data are presented as mean ± standard error of the mean (SEM). Unpaired t-test, * p ≤ 0.05 (n = 3).

Figure 2

Hematopoietic differentiation of DBA iPSCs phenocopies in vitro the erythroid maturation defect found in DBA. (A) Stages of erythroid maturation defined by cell populations expressing variable levels of CD71 (transferrin receptor), CD235a (glycophorin A) and CD45. (B) Representative flow cytometry contour plots depicting percentages of late-stage CD45⁻ erythroblasts co-expressing CD71 and CD235a at culture day 21 of isogenic and DBA iPSC differentiation. (C) Percentages of CD71⁺CD235a⁺ late-stage CD45⁻ erythroblasts at culture day 19 or 21 of isogenic and DBA iPSC differentiation (n = 9). (D) Representative Giemsa stains of hematopoietic cells at culture day 21 of isogenic and DBA iPSC differentiation. (E) Number of colony forming units (CFUs) per 3000 hematopoietic cells harvested at culture day 19 or 21 of isogenic and DBA iPSC differentiation (n = 16). (F) Representative CFU plates for hematopoietic cells harvested at culture day 21 of isogenic and DBA iPSC differentiation. White arrows point to erythroid colonies. In panels C and E, data are presented as mean ± SEM. In panel E, statistical analysis is presented for each colony type relative to the “DBA iPSCs” group. Unpaired t-test, **** p ≤ 0.0001, ns: not significant.

Figure 3

Defective erythropoiesis in DBA iPSCs is partially rescued by eltrombopag (EPAG). (A,B) Quantification of intracellular iron chelating properties of EPAG in erythroid cells differentiated from isogenic iPSCs. Representative flow cytometry histograms of intracellular calcein-AM fluorescence (A) and summary of the mean fluorescence intensity of calcein-AM (B) in isogenic iPSC-derived erythroid cells loaded with calcein-AM and then treated or not with EPAG. Peak height of histogram was normalized to mode (C) Representative flow cytometry contour plots depicting percentages of late-stage CD45⁻ erythroblasts co-expressing CD71 and CD235a at culture day 21 of isogenic iPSC differentiation in the presence or absence of EPAG. (D) Percentages of CD71⁺CD235a⁺ late-stage CD45⁻ erythroblasts at culture day 19 or 21 of isogenic iPSC differentiation in the presence or absence of EPAG (n = 7). (E) Representative Giemsa stains of hematopoietic cells at culture day 21 of isogenic iPSC differentiation in the presence of EPAG. (F) Number of colony forming units (CFUs) per 3000 hematopoietic cells harvested at culture day 19 or 21 of DBA iPSC differentiation in the absence or presence of EPAG (n = 8). (G) Representative CFU plates for hematopoietic cells harvested at culture day 21 of DBA iPSC differentiation in the presence or absence of EPAG. White arrows point to erythroid colonies. In panels B, D and F, data are presented as mean ± SEM. In panel F, statistical analysis is presented for each colony type relative to the “No EPAG” group. Unpaired t-test, **** p ≤ 0.0001, ns: not significant.

Figure 4

Intracellular iron restriction improves erythropoiesis in DBA iPSCs. (A–D) Quantification of intracellular iron chelating properties of DFX and DFO in erythroid cells differentiated from isogenic iPSCs. Representative flow cytometry histograms of intracellular calcein-AM fluorescence (A,C) and summary of the mean fluorescence intensity of calcein-AM (B,D) in isogenic iPSC-derived erythroid cells loaded with calcein-AM and then treated or not with DFX (A,B) or DFO (C,D). Peak height of each histogram was normalized to mode. (E) Representative flow cytometry contour plots depicting percentages of late-stage CD45- erythroblasts co-expressing CD71 and CD235a at culture day 21 of DBA iPSC differentiation in in the absence of an iron chelator (no treatment) or in the presence of deferasirox (DFX) or deferoxamine (DFO). (F) Percentages of CD71 + CD235a+ late-stage CD45⁻ erythroblasts at culture day 19 or 21 of DBA iPSC differentiation in the absence of an iron chelator (No Tx, no treatment) or in the presence of DFX or DFO (n = 3). (G) Number of CFUs per 3000 hematopoietic cells harvested at day 19 or 21 of DBA iPSC differentiation in the absence of an iron chelator (No Tx) or in the presence of DFX or DFO (n = 4). Insets at the bottom are representative CFU plates; white arrows point to erythroid colonies. In panels B, D, F and G, data are presented as mean ± SEM. In panel G, statistical analysis is presented for each colony type relative to the “No Tx” group. Unpaired t-test, * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001, **** p ≤ 0.0001, ns: not significant.

Figure 5

Summary. iPSC lines were generated from blood cells of DBA patients, and isogenic iPSC clones were produced by CRISPR-mediated correction of RPS19 mutations. Hematopoietic differentiation of DBA iPSCs phenocopied the erythroid maturation defect found in DBA. Defective erythropoiesis was partially rescued by eltrombopag and deferasirox, but did not improve substantially with deferoxamine, consistent with a mechanism based on intracellular iron chelation.