Altered Hematopoiesis in Trisomy 21 As Revealed Through In Vitro Differentiation of Isogenic Human Pluripotent Cells

Abstract 921

Down Syndrome (DS, trisomy 21) is associated with increased incidence of acute myeloid leukemia, specifically megakaryoblastic leukemia (DS-AMKL), and the pre-leukemic condition, transient myeloproliferative disorder (DS-TMD). Both AMKL and TMD are invariantly associated with amino-terminal truncating, somatic mutations in the transcription factor Gata1 (termed Gata1s). It has been proposed that Gata1s cooperates with an underlying disturbance of hematopoiesis in trisomy 21 to generate these conditions. Initial support for this model came from reports of enhanced myeloid and erythroid progenitors in trisomy 21 fetal livers in the absence of Gata1 mutations. Since mouse models of trisomy 21 do not exhibit hematopoietic abnormalities in fetal liver or develop TMD, we have sought to establish a human platform for assessment of hematopoiesis in trisomy 21. To this end, we have performed in vitro differentiation of human pluripotent cells that have been carefully controlled to minimize effects of inherent individual or clonal variation that often confound studies with such cells.

We have used two different types of pluripotent human cell lines in these studies. We generated trisomy 21 induced pluripotent cells (iPS) by expressing reprogramming factors in human fibroblasts. Upon progressive passage of trisomy 21 iPS cells, we observed the segregation of disomic cells at low frequency. Taking advantage of this phenomenon, we isolated isogenic passage-matched disomic and trisomic iPS clones in order to mitigate clone-to-clone variation and performed in vitro differentiation experiments. To validate observations from 4 disomic and 4 trisomic isogenic clones, we also assayed 2 independent trisomic iPS lines, and 2 trisomic human ES lines, along with appropriate disomic controls.

Human pluripotent cells were subjected to differentiation as embryoid bodies with successive addition of growth factors. Initially, cells were treated with BMP4, Activin A and FGF to promote mesoderm formation. After 4 days, VEGF, IL-6, IL-3, IL-11 and SCF were added, and after 4 more days, EPO and TPO were added for the duration of differentiation. As results were consistent for iPS and hES cells, we summarize the differences between normal and trisomic hematopoiesis as follows. During in vitro differentiation, we observed that trisomy 21 cells gave rise to elevated numbers of multilineage hematopoietic progenitor cells. Colony forming assays performed with cells isolated from Day 10 (D10) of differentiation revealed that trisomic clones exhibited greater colony-forming potential for CFU-E, CFU-G, and CFU-M colonies. Furthermore, enhanced CFU-meg forming potential was observed, and trisomic cells also formed megakaryocyte colonies that were larger than disomic controls. FACS analysis identified a CD235+ population of cells that is enhanced 4-fold in trisomic cells at D10; prior to D10, no differences in CD31/CD34+ progenitor populations were observed. From D10-20 of differentiation, we observed ∼3-fold increases in CD235+ and CD41+ cells in trisomic cells, representing cells presumably committed to either the erythroid or megakaryocytic lineage. Quantitative RT-PCR of sorted CD235+ populations indicated that γ-globin is the predominant β-like globin expressed in erythroid cells. Hence, under the conditions of in vitro differentiation we have employed, fetal liver-type hematopoiesis is highly favored. As this stage is likely to represent the target for effects of somatic mutations in Gata1 and cooperating genes, the experimental system is well suited for further dissection of the pathogenesis of DS-AMKL.

In conclusion, using hES and isogenic iPS cells, we have observed that trisomy 21 leads to a disturbance of hematopoietic development, characterized by expansion of a common myeloid progenitor population. These results are reminiscent of observations that have been reported for trisomy 21 fetal liver. Our studies further establish that trisomy 21 results in a cell autonomous defect in hematopoiesis. The robust in vitro assay we have established can now be used to further investigate i) molecular targets responsible for perturbed hematopoiesis in trisomy 21 and ii) the effect of Gata1s mutation in concert with trisomy 21. The use of isogenic disomic and trisomic cells provides a powerful approach to reducing potential artifacts due to inherent clonal variation of pluripotent cell lines.