Myelodysplastic syndromes (MDS) are clonal hematologic disorders characterized by peripheral blood cytopenias and a dysplastic bone marrow (BM). Despite their relatively high incidence, these syndromes remain poorly understood and poorly studied, largely due to the unavailability of good animal models and the challenges of the ex vivo culture of primary MDS BM cells: their scarcity, poor proliferative potential and cellular heterogeneity. MDS BM cells exhibit poor growth and clonogenic capacity in culture, suggestive of a cell-intrinsic defect, but the cellular processes that are abnormal (e.g. proliferation, differentiation, cell death) remain elusive.
We set to establish an in vitro system of pure clonal MDS hematopoiesis as a new platform to investigate the pathophysiology of MDS. We used reprogramming technology to derive induced pluripotent stem cells (iPSCs) from BM mononuclear cells of 3 MDS patients (RAEB by FAB) using our excisable polycistronic lentiviral vector (Papapetrou et al. Nat Biotech, 2009) or Sendai viruses. We derived 4 iPSC lines from a del(20q)-MDS patient (MDS-0), one line from a del(7q)-MDS patient (MDS-206), as well as 10 normal (wt-) iPSC lines derived in parallel in one reprogramming experiment from the same starting BM sample (MDS-206). We also derived 9 iPSC lines with chromosome 7 uniparental disomy (UPD) from a third patient (MDS-L1). Karyotyping and aCGH analyses confirmed that the MDS-iPSC lines harbored typical chromosomal deletions (20q12-q13.2 and 7q21.3-qter, respectively), identical to the starting cells. The wt- iPSCs had a normal karyotype and were confirmed to be isogenic to the del(7q) MDS-206.13 line by DNA fingerprinting. All wt- and MDS-iPSC lines display characteristic morphology and pluripotency marker expression. 6 selected lines were shown to fulfill all criteria of pluripotency, including teratoma formation.
One del(7q)- and two del(20q)- iPSC lines so far studied show a 2- to 6- fold reduced proliferation rate (quantified by CFSE dilution and growth curves) compared to that of isogenic and non-isogenic wt-iPSCs, a phenotype much more pronounced in the del(7q) MDS-206.13 line, but absent from all 3 MDS-L1 UPD lines. Cell cycle analysis showed a relative accumulation in G0-G1 phase (40% in MDS-206.13 vs 23–25% in controls). Annexin V staining showed no differences in the percentage of apoptotic cells. Microarray analysis revealed 675 and 780 significantly differentially expressed genes in del(7q) MDS-206.13 and del(20q) MDS-0.12 iPSCs, respectively, compared to the wt MDS-206.12 line. In both cases, these were most enriched in the Gene Ontology categories of cellular growth and proliferation, cellular development and cell death. Ingenuity pathway analysis identified activation of p53 and FOS-JUN (AP1 transcription factor) among predominant potential regulators. Out of ∼1150 protein-coding genes residing in chromosome 7, 102 genes in 7q had reduced expression by at least 1.5-fold (23 of which by 2-fold) in the del(7q) iPSC line MDS-206.13 compared to its isogenic diploid line MDS-206.12.
The hematopoietic potential of the MDS-206.13 line and its normal isogenic control MDS-206.12 was assessed in embryoid body differentiation culture with cytokine supplementation. Strikingly, after mesoderm specification for 3 days followed by 10 days of hematopoietic differentiation, less than 1% of MDS-206.13 vs 48% of MDS-206.12 cells became committed to the hematopoietic lineage (CD34+/CD45+co-expression). Consistent with this, hematopoietic colony formation in methylcellulose and further differentiation in erythroid culture of del(7q)-iPSCs was altogether absent, in contrast to the robust clonogenic and erythroid differentiation potential of the isogenic control line.
Our data suggest that impaired cell proliferation may be integral to the pathophysiology of del(7q)-MDS. Since this phenotype is predominant in del(7q)-iPSCs, but absent from UPD7-iPSCs, it may be caused by reduced dosage of one or more genes on chromosome 7 (haploinsufficiency). Further studies with additional iPSC lines patient-derived and genetically engineered to harbor artificial 7/7q deletions are underway.
In summary, we have developed a novel MDS model of patient-derived and isogenic normal iPSCs. This model should prove useful to study the cellular, molecular and genetic pathogenesis of MDS, identify critical genes and test therapeutic compounds.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.