10
2016 ). Macrophages are another type of cell that does not associate with the vascu-
lature but it regulates HSC trafficking (in this case indirectly) by targeting perivas-
cular stromal cells. Many macrophages are found adjacent to the bone surface and
are defined as “osteomacs”. Macrophage ablation induces a reduction of CXCL
production by perivascular niche cells thus triggering HSC mobilization from the
bone marrow to the periphery (Chow et al. 2011 ; Christopher et al. 2011 ; Winkler
et al. 2010 ). Another cell that indirectly regulates HSC trafficking are aged neutro-
phils that return from the periphery to the bone marrow where they are phagocy-
tosed by macrophages activating them and ultimately triggering CXCL
downregulation in perivascular niche cells and HSC release to the circulation
(Casanova-Acebes et al. 2013 ). An open question is that while the crosstalk between
macrophages and perivascular niche cells is well established the molecules and
receptors involved in this crosstalk are not known.
Other candidate niche cell: In addition to the ones described above other cell
types in the bone marrow have been proposed to regulate HSC function in some
studies. However, other reports have found little to no effect of these cells on HSC
and additional studies are needed to clarify their function. Bones (vertebrae)
enriched in adipocytes contain fewer HSC and “fatless” mice, which have fewer
BM adipocytes, showed faster HSC recovery after lethal irradiation and transplanta-
tion in long bones suggesting that adipocytes negatively regulate HSC during regen-
eration (Naveiras et al. 2009 ). In contrast, drug-induced adipocyte expansion in vivo
had no effect on HSC suggesting that adipocytes do not form part of the niche dur-
ing homeostasis (Spindler et al. 2014 ). A third, recent, report suggests that adipo-
cytes are a proregenerative component of the niche: The authors found that BM
adipocytes also produce SCF and that Adipoq-Cre/ER mice can be used to lineage-
trace adipocytes and a subset of LepR+ cells that contain the adipocyte progenitors.
They also found that SCF derived from Adipoq-Cre/ER+ lineage-traced cells had no
effect in HSC numbers or function in the steady-state but it was required for HSC
engraftment after transplantation in long bones (Zhou et al. 2017 ). In agreement
“fatless” mice, which have a twofold reduction in BM adipocytes, had normal HSC
numbers in the steady-state but impaired HSC recovery after BM transplantation in
long-bones (Zhou et al. 2017 ). In contrast, the vertebrae of the fatless mice had
increased HSC numbers and increased sinusoidal vessels in the steady-state sug-
gesting that the vertebrae-specific effects are mediated via adipocyte-induced
remodeling of the sinusoids in these bones (Zhou et al. 2017 ). More analyses are
needed to dissect direct and indirect effects of adipocytes in HSC and niche cell
regulation in different bones. It is also not clear what parts of the observed pheno-
types are mediated by mature BM adipocytes or more immature Adipoq-Cre+LepR+
progenitors.
Another non-hematopoietic cell that has been proposed to be a component of the
HSC niche is the osteoblast. These are cells located in the endosteal surface of the
bone and that are actively producing new bone matrix. They were shown to support
HSC maintenance in vitro (Taichman and Emerson 1994 ; Taichman et al. 1996 ).
This together with the fact that the central marrow contains fewer HSC than areas
closer to the bone (Nilsson et al. 2001 ); and that mice with constitutive expansion of
D. Lucas