Cdx4 is known to be of importance for specification of cell fate in embryonic hematopoiesis with defects leading to severe perturbation of blood formation. When overexpressed in a murine hematopoietic stem cell line, Cdx4 is capable to enhance progenitor formation in vitro and promote lymphoid reconstitution of lethally irradiated, transplanted mice in vivo.
In line with this important function of Cdx4 in early hematopoiesis, we analyzed expression of Cdx4 in highly purified subpopulations isolated from murine bone marrow (BM) cells by TaqMan qPCR. Cdx4 showed an expression profile known from other stem cell regulatory genes with high expression in early hematopoietic progenitors followed by decreasing expression towards the more differentiated stages of hematopoiesis, with a more than 1200-fold lower expression in total BM cells compared to progenitor enriched 5-FU BM cells (n=3).
To test the impact of Cdx4 on murine progenitors, we retrovirally transduced 5-FU BM cells with Cdx4. Overexpression of Cdx4 induced growth of BM cells in liquid expansion assay (Cdx4 5.7×108±2.2×108 SEM, EGFP 2.6×106±9×105 SEM, p=0.020; cell numbers after 14 days in cytokine supplemented medium, n=5). In addition, expression of Cdx4 conferred serial replating capacity to murine BM progenitors compared to empty vector control (CFU total after 3rd replating: 4.5×109±1.3×109 SEM/500 input cells in 1st CFC, n=5). This effect was significantly stronger compared to hematopoietic progenitors overexpressing the leukemogenic Cdx2 (p=0.008). Immunophenotyping of cells after 3rd replating showed expression of mainly myeloid antigens and cytospin preparation revealed a mature myeloid morphology. Interestingly, these colonies were able to engraft lethally irradiated mice and showed multilineage engraftment (lymphoid:myloid ratio week 16 after transplantation: 0.5:1, n=2), indicating the ability of Cdx4 expressing colonies to maintain stem cell properties in vitro.
In contrast to Cdx2-transplanted mice which showed a severe myeloid bias, regular peripheral blood analysis of mice transplanted with Cdx4 overexpressing BM cells showed multilineage engraftment confirmed by immunophenotyping and normal hematological parameters (RBC 6.7×109±4.2×108, WBC 5.8×106±5.19×105; lymphoid:myeloid ratio 1.4:1; week 8–28). Of note, with a median latency of 309 days after transplantation, nine out of ten mice transplanted with Cdx4-transduced BM cells died of transplantable leukemia. In six out of seven cases we found single retroviral integration sites, indicating a monoclonal origin of the disease. We could determine three different integration sites located between 200 and 700 bp upstream of coding sequences (n=4; Opa3, Akap1, Sema4d). The integration sites of two other mice were located intragenic (Zfyve2, Zfp407), indicating that insertional mutagenesis might be a necessary factor for Cdx4 induced leukemogenesis. Moreover, qRT-PCR revealed that Cdx4 in contrast to Cdx2 did not induce ectopic expression of the leukemogenic Hoxb8 and was associated with a significant lower (7.8-fold) expression of the leukemogenic Hoxb6 in transduced murine BM cells.
Taken together, these data indicate that Cdx4 plays a major role in the regulation of early hematopoiesis. Its expression profile and its hematopoietic activity in different hematopoietic assays clearly differs from Cdx2, which was shown to be highly leukemogenic in mice and to be ectopically expressed in human AML. Murine models analyzing the impact of Cdx4 and Cdx2 expression on hematopoietic development will help to delineate critical differences between the two related genes.
Disclosures: No relevant conflicts of interest to declare.