Hematopoiesis in the liver and marrow of human fetuses at 5 to 16 weeks postconception: quantitative assessment of macrophage and neutrophil populations.

During human development, the liver and marrow both function as hematopoietic organs, but little is known about differences in the production of macrophages and neutrophils by these two organs. We used immunohistochemical stains to quantify the ratio of neutrophils to macrophages within the liver and the marrow of 16 fetuses from 5 to 16 wk postconception. At 5 wk the liver had a ratio of one granulocyte [myeloperoxidase (MPO)-positive cell] to every 9 +/- 5 (X +/- SD) macrophages (KP-1-positive cells). Between 5 and 16 wk, the granulocyte to macrophage ratio in the liver was constant, whereas it changed markedly in the marrow. Before 8 wk no MPO-positive or KP-1-positive cells were observed in bones. At 10 wk, bones still had no MPO-positive cells, but KP-1-positive cells were abundant. At 11-12 wk the granulocyte to macrophage ratio was 1 to 1 +/- 1, but by 13-16 wk it had increased to 8 +/- 3 MPO-positive cells to one KP-1-positive cell. We hypothesized that at 13-16 wk the abundance of MPO-positive cells in the marrow and their scarcity in the liver was the result of production of granulocyte colony-stimulating factor (G-CSF) and its receptor (G-CSF-R) in the marrow and their absence in the liver. However, by reverse transcriptase-PCR mRNAs for G-CSF and G-CSF-R were positive in both organs at all gestations, and G-CSF and G-CSF-R proteins (by immunohistochemistry) were also abundant in all liver and marrow specimens. We then hypothesized that progenitors in the fetal liver were intrinsically different from those in the marrow, and were unable to generate clones of neutrophils. However, progenitors from the liver produced neutrophils abundantly in culture. Thus, the explanation is likely related to as yet undescribed environmental differences between the fetal liver and marrow.