Overexpression of HOX11 leads to the immortalization of embryonic precursors with both primitive and definitive hematopoietic potential

Primitive and definitive erythropoiesis represent distinct hematopoietic programs that differ with respect to stage of development, transcriptional control, and growth regulation. Although these differences have been recognized for some time, the relationship of the two erythroid lineages to each other is not well established. We have used a model system based on the hematopoietic development of embryonic stem (ES) cells in culture to investigate the origins of the earliest hematopoietic populations. Using ES cells transduced with a retrovirus that overexpresses the HOX11 gene, we have established factor-dependent hematopoietic cell lines that represent novel stages of embryonic hematopoiesis. Analysis of three of these cell lines indicates that they differ with respect to cytokine responsiveness, cell surface markers, and developmental potential. Two of the cell lines, EBHX1 and EBHX11, display the unique capacity to generate both primitive and definitive erythroid progeny as defined by morphology and expression of betaH1 and betamajor globin. The third line, EBHX14, has definitive erythroid and myeloid potential, but is unable to generate cells of the primitive erythroid lineage. Analysis of the cytokine responsiveness of the two lines with primitive erythroid potential has indicated that exposure to leukemia inhibitory factor (LIF) results in the upregulation of betaH1 and a change in cellular morphology to that of primitive erythrocytes. These findings are the first demonstration of a clonal cell line with primitive and definitive hematopoietic potential and support the interpretation that these lineages may arise from a common precursor in embryonic life. In addition, they suggest that LIF could play a role in the regulation of primitive erythropoiesis.