Adult HSCs can be highly

Adult HSCs can be highly purified (Purton and Scadden, 2007; Osawa et al., 1996). This is in contrast to the first HSCs found in the embryo, which have been more difficult to enrich thus far (Taoudi et al., 2005). Many markers currently used to isolate adult HSCs do not apply for embryonic HSCs. Moreover, surface marker expression on HSCs varies during development (Robin et al., 2011; McKinney-Freeman et al., 2009) and also between strain and species. For example, the SLAM marker CD150, which allows for a high enrichment of HSCs in the adult bone marrow (Kiel et al., 2005) and fetal liver (Kim et al., 2006) when used in combination with the classical marker combination LSK (Lin−Sca1+ckit+), is not a marker of AGM HSCs (McKinney-Freeman et al., 2009). The lineage antibody panel (Lin) has been designed and used routinely to deplete mature cells from the erythroid, lymphoid and myeloid lineages (Muller-Sieburg et al., 1986). Such depletion does not apply to purifying embryonic HSCs because Mac-1 (CD11b), a marker classically expressed by macrophages/monocytes in adults, is also expressed by a fraction of AGM HSCs (Sanchez et al., 1996). Moreover, HSCs and endothelial cells share many surface markers, reflecting their close developmental relationship (Garcia-Porrero et al., 1998). The only reliable method used to identify HSCs is to perform a long-term in vivo assay where the multilineage repopulation and self-renewal abilities of the cells are tested after transplantation into primary and secondary adult wild-type irradiated recipients. Using this standard assay, HSCs were first detected in the AGM region starting at E10.5 (Fig. 1A) (Medvinsky and Dzierzak, 1996; Muller et al., 1994). More precisely, HSCs are restricted to the aorta, as shown by the subdissection of the AGM region to separate the rad51 inhibitor from the urogenital ridges before performing transplantation (de Bruijn et al., 2000). HSC activity in the AGM is transient and stops after E12. Interestingly, HSCs are also found at E10.5 in two other major vessels, the vitelline and umbilical arteries (de Bruijn et al., 2000) (Fig. 1A). Slightly later (E11–11.5), HSCs are also detected in other major highly vascularized hematopoietic sites: the YS, placenta and fetal liver (Kumaravelu et al., 2002; Ottersbach and Dzierzak, 2005; Gekas et al., 2005). Only very few HSCs (~4–11 cells) are present in the complete mouse conceptus at E11. By E12, the HSC number multiplies by at least 14 times, mainly in the placenta and fetal liver (Kumaravelu et al., 2002; Gekas et al., 2005). The liver becomes the main HSC reservoir at mid-gestation until the HSCs start to colonize the bone marrow at E17 (Christensen et al., 2004). HSCs behave differently in the embryo (compared to adult) as they transit through several anatomical sites or niches, and are actively self-renewing (Morrison et al., 1995; Bowie et al., 2006). The embryonic microenvironment that composes the successive niches is still poorly described, but it certainly influences the equilibrium between HSC self-renewal and differentiation. In comparison, HSC niches in adult bone marrow are well described (Levesque and Winkler, 2011). Two types of niches, very close spatially, have been reported so far. In the endosteal niche, HSCs are in close contact to the endosteal bone surface where the main supportive cell type, the osteoprogenitor population, maintains HSCs in a quiescent/slow-cycling state (Calvi et al., 2003; Zhang et al., 2003; Raaijmakers et al., 2010; Lo Celso et al., 2009). In the second niche, HSCs are associated with the sinusoidal endothelium (Kiel et al., 2005), but it remains to be determined whether it represents a functional niche or only a transition site (Purton and Scadden, 2008). The use of in vivo and ex vivo time-lapse confocal microscopy has nicely shown that transplanted HSCs preferentially localize close to the endosteal bone surface while committed progenitors localize further away (Lo Celso et al., 2009; Xie et al., 2009). The most potent HSC niches are most likely hypoxic (Winkler et al., 2010). The well-defined medullar microenvironment of the niches maintains most HSCs in an immature and quiescent/slow-cycling state, the quiescence status being the hallmark of their long-term HSC properties (Wilson et al., 2008; Cheshier et al., 1999). A small pool of HSCs will eventually self-renew only a few times during the lifetime (Wilson et al., 2008). However, perturbation of homeostasis promotes their self-renewal more rapidly (Wilson et al., 2008). The mesenchymal stem cells (which have adipogenic, osteogenic and chondrogenic potential), that express the intermediate filament protein nestin, were recently described as a very important player in maintaining the function of the HSC niche (Mendez-Ferrer et al., 2010). Such cells produce SDF-1α (CXCL12) and SCF (Stem Cell Factor) that are important for HSC maintenance. These factors are part of a long and non-exhaustive list of intrinsic and extrinsic factors essential for HSC fate regulation. It includes transcription factors (e.g. SCL, Runx1, Cbfβ, Lmo2, GATA2), cell cycle regulators (e.g. p27kip1, p21cip1/waf1), hematopoietic cytokines (e.g. TPO, Flt3/Flk2 ligand, IL-3) and developmental regulators (e.g. BMP-4, Tie2/Angiopoietin-1, Wnt/β-catenin, TGF-β/p21, VCAM-1, Hedgehog, Notch/Jagged 1) (Zon, 2008). Nevertheless, no studies have yet pinpointed the exact molecular network architecture that distinguishes self-renewing from non-self-renewing hematopoietic cells.