12
et al. 2014 ; Nakamura-Ishizu et al. 2014 , 2015 ; Yamazaki et al. 2011 ; Zhao et al.
2014 ); or retention (e.g. macrophages (Chow et al. 2011 ; Christopher et al. 2011 ;
Winkler et al. 2010 ). These raise the possibility that different spatial locations in the
bone marrow regulate different HSC pools. In this section we discuss the evidence
for, and against, spatially distinct niches.
Arteriolar and sinusoidal niches: A plethora of imaging and functional analy-
ses have conclusively demonstrated that most HSC reside close to, and are regulated
by, perivascular niches (Acar et al. 2015 ; Bruns et al. 2014 ; Chen et al. 2016 ; Kiel
et al. 2005 ; Kunisaki et al. 2013 ; Mendez-Ferrer et al. 2010 ; Nombela-Arrieta et al.
2013 ). In addition several studies have proposed the existence of distinct sinusoidal
and periarteriolar niches. Kunisaki et al., showed that ~30–35% of all
Lin−CD48−CD41−CD150+ HSC were located within 20 μm of CD31+Sca1+ BM
arterioles (Kunisaki et al. 2013 ). These authors also showed that BM arterioles were
ensheathed by Ng2+/Nestin-GFPbright cells and that these structures were enriched in
quiescent HSC when compared to sinusoids. In agreement with these results deple-
tion of Ng2+ perivascular cells led to reduced BM HSC numbers, relocation of HSC
away from arterioles and loss of HSC quiescence (Kunisaki et al. 2013 ). In a follow
up manuscript the same group showed that conditional Cxcl12 deletion in Ng2+
perivascular cells using Ng2-creER or Myh11-CreERT2 mice caused loss of BM HSC
and relocation of HSC away from arterioles. In the same study Cxcl12 deletion in
LepR+ cells had no effect on BM HSC numbers (Asada et al. 2017a). In contrast Scf
deletion in Ng2+ cells had no effect on BM HSC numbers but Scf deletion in LepR+
cells causes a dramatic HSC loss. These results suggest that Ng2+ periarteriolar and
LepR+ (which are mainly perisinusoidal albeit some LepR+ are close to arterioles
(Ding and Morrison 2013 ; Ding et al. 2012 ; Kunisaki et al. 2013 )) maintain HSC
via the production of different cytokines (Asada et al. 2017a). It is important to note,
however, that these results contrast with a previous study that showed that LepR+
cells maintain HSC via CXCL12 (Ding and Morrison 2013 ). Megakaryocytes pro-
mote HSC quiescence ((Bruns et al. 2014 ; Nakamura-Ishizu et al. 2014 , 2015 ; Zhao
et al. 2014 ) and are located exclusively in the sinusoids. Megakaryocyte ablation
induces HSC proliferation and HSC relocation away from sinusoids but does not
disrupt HSC interaction with arterioles suggesting that sinusoidal and arteriolar
HSC niches are functionally independent (Bruns et al. 2014 ). The fact that non-
myelinating Schwann cells, that are associated with the sympathetic nerves in arte-
rioles (Kunisaki et al. 2013 ), restrict HSC proliferation (Yamazaki et al. 2011 )
further supports the concept of a periarteriolar niche that promotes HSC quiescence.
Additional data supporting the existence of functionally distinct arteriolar and sinu-
soidal niches was presented by Itkin et al. These authors quantified the percentage
of Lin−CD48−CD150+ hematopoietic stem and progenitor cells (HSPC) that stained
positive for reactive oxygen species (ROS) in the bone marrow. They found that all
HSPC adjacent to arteries were ROS− whereas ~35% of all HSPC adjacent to sinu-
soids where ROS+. Increases in vascular leakiness using Cdh5-creERT2:Cxcr4lox/lox or
Cdh5-creERT2:Fgfr1/2lox/lox mice revealed increased numbers of ROS+ HSPC close to
sinusoids indicating that vascular permeability controls HSPC metabolic state (Itkin
et al. 2016 ). The same study showed that BM sinusoids are the exclusive trafficking
site and that increased vascular permeability promotes BM HSPC release into the
D. Lucas