Nestin⁺ Pericytes In The Fetal Liver Are Necessary To Maintain HSCs

Although most hematopoietic stem cells (HSCs) are quiescent under homeostasis in the adult bone marrow, they are actively proliferating during development. Definitive HSCs, marked by the ability to repopulate a lethally irradiated adult mouse, are first detectable in the aorta-gonad-mesonephros region around E10.5, and then colonize the fetal liver (FL) to expand in this organ until E15 when hematopoietic activity shifts to the fetal bone marrow. The role of the microenvironment, or niche, in the regulation of HSCs in the FL, a site of physiological expansion, is unclear. Fetal liver sinusoidal cells as well as hepatic progenitors have been proposed as sources of supportive signals that drive the exponential expansion of HSCs, however, exact cell type(s) supporting HSC expansion have not been defined. Since stromal cells marked by Nestin-GFP form niches for HSCs in the BM, we have hypothesized that similar cells exist in the FL. Indeed, a rare population of Nestin⁺ cells (CD31^-CD45^-Ter119^-) comprising 0.02 ± 0.01% of nucleated FL cells was isolated. FL Nestin⁺ cells express surface markers similar to their BM counterparts, including PDGFRα, CD51, and endoglin. Since BM Nestin⁺ cells are enriched in mesenchymal stem and progenitor cell (MSPC) activity, we next tested this activity in FL Nestin⁺ cells. We found that the entire CFU-F forming capacity of the FL was contained within FL Nestin⁺ cells, further suggesting similarities of these cells between these two hematopoietic organs. In addition, FL Nestin⁺ cells were enriched for the HSC maintenance genes SCF and CXCL12, suggesting that they may also serve as the HSC niche in the FL. FL Nestin⁺ cells were localized on the abluminal side of large-bore arteries and expressed the pericyte markers, α-smooth muscle actin and NG2. To investigate the function of the FL Nestin⁺NG2⁺ cells in vivo, we generated a genetic depletion mouse model of NG2⁺ cells (NG2-cre / inducible diphtheria toxin-A (iDTA) transgenic mice). We found that E14.5 NG2-Cre;iDTA mouse embryos developed normally compared to fl-DTA control littermates. The frequency and absolute numbers of HSCs per FL, however, were reduced by about 40% in NG2-Cre;iDTA embryos (control / depleted; 0.01146 ± 0.0008% / 0.00660 ± 0.0008% of nucleated cells, p=0.0015; 606 ± 76 / 377 ± 48 HSCs/FL, p=0.03, , N=6-7 per group) suggesting that Nestin⁺ cells are required to maintain HSCs/progenitors in vivo. To interrogate further the relationship between FL Nestin⁺ cells and HSCs, we adopted the re-aggregate organ culture assay (Sheridan, Genesis 2009). Sorted lineage^- FL hematopoietic progenitor cells and FL parenchymal cells either with or without sorted Nestin⁺ cells were re-aggregated and cultured for 7 days in serum-free, cytokine-free, media. After 7 days of culture, we found that CD150⁺ CD48⁻ CD41⁻ Lineage⁻ HSCs were maintained in re-aggregates containing Nestin⁺ cells, but not when Nestin⁺ cells were absent. To confirm that FL Nestin⁺ cells were essential to maintain functional HSCs, re-aggregated cells cultured for 7 days were transplanted together with competitor bone marrow into lethally irradiated mice. We found that the contribution to the peripheral blood at 8 weeks post-transplant was only observed in the re-aggregated group containing Nestin⁺ cells. These preliminary data indicate that factors derived from FL Nestin⁺ cells are required to maintain HSCs in re-aggregate cultures. These findings suggest that HSCs inhabit similar microenvironments in temporally and spatially distinct hematopoietic organs. Further studies on differences between Nestin⁺ cells in these tissues may shed light on the mechanisms that determine the finely tuned quiescence, self-renewal and differentiation of HSCs.