Hematopoietic Stem Cell Development, Niches, and Signaling Pathways

Hematopoietic stem cells (HSCs) play a key role in hematopoietic system that functions mainly in homeostasis and immune response. HSCs transplantation has been applied for the treatment of several diseases. However, HSCs persist in the small quantity within the body, mostly in the quiescent state. Understanding the basic knowledge of HSCs is useful for stem cell biology research and therapeutic medicine development. Thus, this paper emphasizes on HSC origin, source, development, the niche, and signaling pathways which support HSC maintenance and balance between self-renewal and proliferation which will be useful for the advancement of HSC expansion and transplantation in the future.


Introduction
Hematopoietic stem cells (HSC) are adult stem cells that contain the potentiality in self-renew and differentiation into specialized blood cells that function in some biological activities: control homeostasis balance, immune function, and response to microorganisms and inflammation. HSCs can also differentiate into other specialized cell or so called plasticity such as adipocytes [1], cardiomyocytes [2], endothelial cells [3], fibroblasts/myofibroblasts [4], liver cells [5,6], osteochondrocytes [7,8], and pancreatic cells [9]. Most HSCs are in quiescent state within the niches that maintain HSC pool and will respond to the signals after the balance of blood cells or HSC pool is disturbed from either intrinsic or extrinsic stimuli.
In addition, HSCs have been studied extensively, especially, for the therapeutic purposes in the treatment of blood diseases, inherited blood disorders, and autoimmune diseases. Nonetheless, advanced development in this field needs knowledge in the biological studies as a background in performing strategy and maintaining of HSCs. Thus, HSC source, origin, niches for HSC pool, and signaling pathways, essential for the regulation of HSCs, will be discussed in this review.

HSCs Origin and Development
In the hematopoietic system, the discovery of HSCs has shed the light on stem cell biology studies including connection to other adult stem cells through the basic concepts of differentiation, multipotentiality, and self-renewal. In the early period of those discoveries, lethally irradiated animals were found to be rescued by spleen cells or marrow cells [17,18]. After mouse bone marrow cells were transplanted into irradiated mice, the clonogenic mixed colony of hematopoietic cells (often composed of granulocyte/megakaryocyte and erythroid precursors) were formed within the spleen, which these colonies were then termed colony-forming unit spleen (CFU-S) [19]. Some colonies of primary CFU-S could reconstitute hematopoietic system in the secondary irradiated mice after receiving transplantation [20]. Initially, CFU-S was first proposed that it may be differentiated from HSC, but subsequently, CFU-S was demonstrated to be originated from more committed progenitor cells [21]. The discovery by Till and McCulloch embarked on a new journey toward many investigations to clarify HSC biology, functional characterization, purify, cultivation, and other stem cells research. Intraembryonic yolk sac is the first site of blood cells observation at around E7.0-E7.5. The de novo hematopoiesis in the placenta and AGM occurs at nearly similar wave of gestation (around E8.5-E10.5) beforeit circulates into fetal liver where there is the large HSC pool during gestation. At around E16.5, the HSCs migrate and reside within the bone marrow which finally becomes the source of HSC in adult life (adapted from [10]).
Hematopoiesis and HSC development are the key role to improve efficient HSC expansion for the transplantations. Embryogenesis study has been performed to identify HSC origin and activity from various anatomical sites of several kinds of animals such as zebrafish, chicken, and mouse including human embryos model have been emerging. Initially, Moore and Metcalf showed that hematopoietic cells in the yolk sac could generate hematopoietic progenitors that restricted to only erythroid and myeloid lineages [22]. Moreover, the Runx1 (transcription factor for the onset of definitive hematopoiesis) was first identified to express at embryonic day 7.5 (E7.5) in the yolk sac, the chorionic mesoderm, and parts of allantoic mesoderm [23]. However, HSCs found in the yolk sac lacked the definitive hematopoietic stem cells which did not show long-term hematopoietic reconstitution activity in mouse embryo prior to E11.5 [24]. On the other hand, long-term repopulating HSCs (LT-HSCs) were shown to increase largely in the aorta-gonad mesonephros (AGMs) region of the mouse embryo including the serially transplantable irradiated mice, suggesting that AGM region is the first site for HSCs detection [24,25]. Vitelline and umbilical arteries were also endowed with hematopoietic potential [26]. The presence of HSC phenotype in the embryo was supported by the evidence that a high number of nonerythroid progenitors with high-proliferative potential was observed from which the liver rudiment has been removed [27]. A dense population of CD34 + cells adhering to the ventral side of the aortic endothelium within the embryonic compartment was shown to display a cell-surface and molecular phenotype of primitive hematopoietic progenitors (CD45 + , CD34 + , CD31 + , CD38 − , negative for lineage markers, GATA-2 + , GATA-3 + , c-myb + , SCL/TAL1 + , c-kit + , flk-1/KDR + ) [28,29]. Moreover, the autonomously emergence of myelolymphoid lineage from progenitors was found in splanchnopleural mesoderm and derived aorta within the human embryo proper, while restricted progenitors were generated in the yolk sac [30]. Altogether, AGM region in the embryo is suggested as the source of definitive hematopoiesis as the generation occurs between E10.5 and E12.0 with the enhance activity of HSC after mid-day 11 of gestation [10,[31][32][33]. Even though, the main source of fetal hematopoiesis was considered in AGM including vitelline and umbilical arteries, the question is raised whether the rare population produced in those regions would be enough for the distribution into fetal liver for alternative development of enormous HSCs before the transition of hematopoiesis continues to occur in the fetal thymus and bone marrow in postnatal life. Recently, the placenta, an extraembryonic organ, has been considered as the other hematopoietic organ for de novo hematopoiesis [34,35]. This may be due to the physiology of the placenta containing highly vascularized blood vessels, and cytokines and growth factors rich environment for proper microenvironment of hematopoiesis and development [36]. Additionally, privilege site within the placenta may hide the HSCs from the promoting signal into differentiation stage. However, there is no experimental evidence to support that HSCs are generated de novo in the extraembryonic tissues. Therefore, future works will be needed to elucidate this enigma. Summarization of the source of blood cells during gestation through adult life has been elucidated in Figure 1.
The origin of HSC in the placenta is being questioned. Understanding how theplacenta develops might be useful to define the source and the niches supporting HSC development. Mouse and human placentas are anatomically similar and its genes have analogous identity [37,38]. The placenta is formed from trophectoderm, mesodermal tissues, chorionic mesoderm, and allantois ( Figure 2) [39]. At E8.5 of mouse gestation, the allantois develops and fuses with chorionic mesoderm through its distal part generating the chorioallantoic mesenchyme in the chorionic plate and continuing to form the fetal vascular compartment of the placental labyrinth, while the proximal part becomes the umbilical cord [11].
The umbilical cord (a constitution of the fetal arteries and veins that inserted within chorionic plate of the placenta) is attached to the center of fetal surface for uteroplacenta circulation through maternal blood. Maternal blood passes through the placenta from uterine arteries to spiral arteries in the maternal decidua. Thereafter, the maternal blood percolates through the villous tree in humans (or the showing the chorioallantoic mesenchyme lies cover the placenta labyrinth with fetal vessels lined by fetal endothelium (dark vessels with lumen) and trophoblast lined by maternal blood spaces (grey vessels surrounded by dark trophoblasts). al, allantois; ch, chorion; am, amnion; epc, ectoplacental cone; ys, yolk sac; psp, para-aortic splanchnopleura; da, dorsal aorta; ua, umbilical artery; va, vitelline artery; fl, fetal liver (modified from [11]).
labyrinth in mice) known as chorionic villi which created and lined by fetal trophoblast cells [37,39]. The inner core of the chorionic villi consists of allantoic mesenchyme and vasculature which is continuous with that of the umbilical cord. The chorioallantoic vasculature connects the placenta via the dorsal aorta and fetal liver through the umbilical cord vessels. These regions are localized by an equally dense network of fetal capillaries where the fetomaternal exchange occurs [37]. Because of the mesoderm layer gives rise to all blood cells, the chorionic and allantoic mesoderms are considered as the origin of HSC in the placenta. This can be explained by the observation that hematopoietic potential emerging from both tissues and has been identified with myeloerythroid potential [40]. In addition, hematopoietic cells (CD34 + CD45 + ) collected from placental villi stroma and highly expression of CD45 + cells that appear to be budding from the vasculature have been found from human placenta during midgestation [35]. Moreover, cells harvested from term human placenta vessels and tissues could generate human hematopoietic repopulation of nonobese diabetic (NOD)-SCID mice, which harbored and/or amplified in vascular labyrinth placenta niche [35]. These observations imply that the placenta is the HSC source along with umbilical cord blood. At E10.5, first HSC emerge in the dorsal aorta before the onset of heart beat where the circulation has not been formed. One study showed that in the absence of heat beat in Ncx1 (the sodium and calcium exchange pump1) knockout embryos, the HSC development was verified to initiate in the placental vasculature [41]. Additionally, multilineage hematopoietic potential could be obtained from placentas of Ncx1 knockout embryos. Thus, within the extraembryonic tissues, fetal HSCs were observed in placenta, vitelline, and umbilical arteries.
The true origin of HSC in the intraembryonic hematopoiesis remains controversial. One of the main hypotheses is hemangioblasts or hemogenic endothelial while the alternative model is mesodermal precursors. The blood islands originated in the yolk sac are derived from mesodermal cell aggregates, which contain the ability to differentiate into both hematopoietic and endothelial cells. The common precursor by those lineages is suggested to be so called the hemangioblast [42]. Hematopoietic phenotype originated from hemogenic endothelium has been found in avian and mouse during ontogeny [43,44]. Imaging and cell-tracking study explored that hemogenic endothelial cells could give rise to hematopoietic cells [45]. By time-lapse imaging study in single-cell mouse mesodermal cells demonstrated that it could generate endothelial sheet colonies and some colonies developed the hematopoietic morphology that upregulating the blood-specific proteins CD45, CD41, and CD11b and losing their intact morphology. Recently, this evidence has been supported by the observation on time-lapse confocal imaging from live mouse aorta showing that HSCs (Sca + , c-kit + , CD41 + ) could emerge directly from ventral aortic endothelial cells [46]. Moreover, Oberlin and colleagues proved that the origin of adult bone marrow HSCs which most of them derived were from the vascular endothelialcadherin ancestor [47]. Taken together, these studies pinpoint the evidence that definitive hematopoietic stem and progenitor cells emerge from the hemogenic endothelium at the AGM region.

HSC Niches
Homing of HSC from other definitive hematopoiesis to fetal bone marrow is thought to involve some signaling factors such as stromal derived factor-1 (SDF-1 or CXCL12)/chemokine C-X-C receptor 4 (CXCR4) axis [48,49]. Soluble factors are not only mediated in fetal bone marrow but also in adult bone marrow to maintain HSC in undifferentiated state and regulate HSC in proliferative and differentiated states within the specific microenvironments termed "niche" throughout the life [12]. Stem cell niche was first proposed by Schofield [21], with the later identification in Drosophila melanogester's ovary to confirm the existence of HSC niche [50]. Germline stem cells resided in the Drosophila ovary that is surrounded by differentiated somatic cells have been shown to be essential for maintaining stem cells survival and division [50]. Thus, HSC niche is the special local environments of HSCs that maintains and controls HSCs function by regulating survival, self-renewal ability, and cell fate decision. Such molecules have been identified to be associated with HSC homing to bone marrow, for example, SDF1-α, β1-integrins, metalloproteinases (MMP), and serine-threonine protein phosphatase (PP)2A [51,52]. By using real-time imaging, it is possible to explore the localization of HSCs with their function [53]. HSCs lodge in the endosteal surface, osteoblasts, and blood vessels, particularly in trabecular regions, in the mouse calvaria. On the contrary, more mature cells reside away from the endosteum. Similarly, a study by developed ex vivo realtime imaging in irradiated mice show the homing and lodgment of transplantable HSCs in the endosteal region of the trabecular bone area where they respond to bone marrow damage by rapidly dividing [54].
Recently HSCs niches are suggested to be mediated in two main microenvironments within bone marrow: endosteal niche and vascular niche ( Figure 3). First, endosteal niche: osteoblasts derived from mesenchymal precursors are localized in the endosteal regions which are well vascularized. The activation of osteoblastic differentiation is in part mediated by HSC-derived bone morphogenic protein-2 (BMP-2) and BMP-6 [55]. Osteoblasts are suggested as the niche due to the finding that the number of osteoblasts is increased from parathyroid hormone activation and results in an increase HSCs number in vivo [56]. This signal was found to be activated through Jagged1, a serrate family of Notch ligand, on osteoblasts [57]. Study by Chitteti and colleagues supports this evidence and shows that enhancing hematopoiesis promoted by osteoblast via Notch signaling not only through Jagged1 upregulation, but also Notch2, Jagged2, Delta1 and 4, Hes1 and 5, and Deltex ligands [58]. Soluble factors produced from osteoblasts function in regulating HSC quiescence, HSC pool and fate such as angiopoietin-1 (Ang-1) [59], SDF-1 (CXCL12) [60], and osteopontin [61]. Recently, osteoblasts secreted cysteine protease cathepsin X have been found to catalyze the chemokine CXCL-12, a potent chemoattractive cytokine for HSCs, and ablate the attachment of CD34 + cells with the osteoblasts [62]. This result suggests the role of osteoblasts in regulate HSCs trafficking in the bone marrow.
A group of de Borros supports this hypothesis by showing that the 3D spheroid of noninduced and one week osteo-induce bone marrow stromal cell (active osteoblasts) formed an informative microenvironment that control migration, lodgment, and proliferation of HSCs [63]. Bone marrow endosteal cells, particularly, osteoblast-enriched ALCAM + Sca-1 − cells promoted LT-reconstitution activity of HSCs via the upregulation of genes related in homing and cell adhesion [64]. In addition, HSCs were found to adhere with spindle-shaped N-cadherin + osteoblastic (SNO) cells which are a subpopulation of osteoblasts [65]. BMP receptor type IA mutant mice have been shown to increase in the number of SNO cells that correlated to an increase in HSC number [65]. Consistently, green fluorescent proteinpositive (GFP + ) HSCs derived from Col2.3-GFP + transgenic mouse were found to attach to SNO cells but not all GFP + HSCs were in contact with SNO cells showing that Ncadherin − component might be the other niche for HSCs [54]. Cumulatively, osteoblasts and SNO cells are suggested as the niche for hematopoietic stem and progenitor cells where this microenvironment termed "Endosteal niche." Some observations have suggested that another niche, vascular niche, might involve in HSC maintenance within the bone marrow. Studies in osteoblast depletion demonstrated that there was a loss of B lymphopoiesis but not immediately loss of HSC number [66,67] and few bone-marrow HSCs (CD150 + CD48 − CD41 − lineage − ) were localized to the endosteum [68]. Mice model defected in osteoblast function conferred no changes in LT-reconstitution function of HSCs [69]. Additionally, the loss of N-cadherin did not any effect on HSC maintenance and hematopoiesis [70]. Most HSCs in the bone marrow have been observed to reside in the sinusoid, where fenestrated endothelium persists and allows blood flow for an exchange of blood cells and small molecules. Taken together, the vascular niche is suggested as the other niche for HSC maintenance [68]. Bone-marrow endothelial cells have been proposed to play a role in HSC controlling within vascular niche. Primary CD31 + microvascular endothelial cells can restore hematopoiesis in mice when they receive bone-marrow lethal doses of irradiation [71]. Study by a group of Salter shows a consistent observation that endothelial progenitor cells injected in total body irradiated mice can stimulate HSC reconstitution and hematologic recovery [72]. Furthermore, selective activation of Akt in endothelial cells produced angiocrine factors mediated in the reconstitution, expansion, and maintenance of HSCs [73]. Nonetheless, constitutively activation of Akt, a binding ligand of phosphoinositide 3 in the phosphoinositide 3-kinase pathway, impaired engraftment ability and preferable generated leukemia in mice [74]. Sinusoidal endothelial cells are essential for engraftment of hematopoietic stem and progenitor cells (HSPCs) and restoration of hematopoiesis after myeloablation [75]. Angiocrine factors, such as Notch ligands, released by endothelial cells in vivo contributed to the replenishment of the LT-HSC pool and resulted in reconstitution of hematopoiesis [76]. Altogether, vascular niche containing endothelial cells is suggested as the major HSC pool and maintenance conferring proliferation and differentiation selection.
Additionally, Sugiyama and colleagues demonstrated that reticular cells located around the sinusoid endothelium could produce stromal cell-derived factor 1 (SDF-1, aka CXCL12) mediated in HSC niche [77]. These cells have been named CXCL12 abundant reticular cells (CAR cells). This study showed that almost all HSCs were found in contact with CAR cells and all HSCs allocated at endosteum were also found to be in contact with CAR cells, suggesting that these cells play HSCs are in contact with SNO cells, bone-lining osteoblasts, within endosteal niche. Osteblasts produce several signal molecules such as Notch ligands, angiopoietin-1 (Ang-1), CXCL12, and cathepsin X mediated in control HSC pool and maintenance. Most HSCs are found in sinusoids, particularly adherence to CAR cells that surround sinusoidal endothelial cells (reticular niche). Similarly, CAR cells produce CXCL12 in association with CXCR4 signaling essentially for HSC maintenance (modified from [12]). a crucial role in HSC niches by homing HSCs in both vascular and endosteal niches [77]. CXCL12/CXCR4 signaling is essential in maintaining the HSC pool, development of B cells and plasmacytoid dendritic cells [78][79][80][81]. Moreover, short-term ablation of CAR cells resulted in the impairment of adipogenic and osteogenic differentiation. Thus, CAR cells are suggested as the adipo-osteogenic progenitors [82]. Study in CAR cell-depleted mice demonstrated that HSCs were reduced in number and cell size, which were more quiescent and highly expressed early myeloid selector genes [82]. CAR cells were suggested to coincide with CD146 + stromal progenitors that express CXCL12 and Ang-1. CD146 + cells could generate osteoblast that form bone and could function as skeletal progenitor cells [83]. Taken together, CAR cells provide or generate the hematopoietic microenvironment that link to the hematopoietic regulation in both vascular and endosteal niches.

Hierarchy of Human Hematopoiesis
Based on the study of molecular marker expression by flow cytometry analysis has led the identification of each blood cell subpopulations in terms of their biology and potential when combined with other functional assays. As a result, schematic demonstration of hematopoietic hierarchy has been proposed ( Figure 4) [13]. The origin of all blood cell in hematopoietic system is believed to be derived from HSCs that contain self-renewal capacity and give rise to multipotent progenitors (MPPs) which lose self-renewal potential but remain fully differentiate into all multilineages. MPPs further give rise to oligopotent progenitors which are common lymphoid and myeloid progenitors (CLPs and CMPs, resp.). All these oligopotent progenitors differentiate into their restricted lineage commitment: (1) CMPs advance to megakaryocyte/erythrocyte progenitors (MEPs), granulocyte/macrophage progenitors (GMPs), and dendritic cell (DC) progenitors, (2) CLPs give rise to T cell progenitors, B cell progenitors, NK cell progenitors and DC progenitors. Notably, DC progenitors (CD8α + DC, CD8α − DC, and plasmacytoid DC) could be derived from both CMPs and CLPs [84][85][86]. Among the isolation and characterization of HSCs and progenitors, CD34 molecule is the first widely chosen for the study by several researchers. CD34 is comprised in the CD34 family of cell-surface transmembrane proteins together with podocalyxin and endoglycan [87][88][89]. CD34 expression on blood cells is about 0.1-4.9% in human cord blood, bone marrow, and peripheral blood [90][91][92]. The first candidate human HSCs was a population of cells  expressing CD34 + CD90 + (Thy-1)Lin − which could give rise to T and B lymphocytes and myeloerythroid activities in both in vitro and in vivo human fetal thymus transplanted into SCID mice while some subset of CD34 − , CD90 − , Lin − lacked of multipotent progenitors [93]. Further isolation of HSCs was based on the expression of CD38 [94,95] and CD45RA [96]. This data could be concluded that Lin − CD34 + CD38 − CD90 + CD45RA − population enriches for human HSCs and the candidate human MPP fraction of multipotency with an incomplete self-renewal capacity is enriched in Lin − CD34 + CD38 − CD90 − CD45RA − population [97]. However, recently observation using HSC xenograft assay in NOD-SCID-IL2Rgc −/− (NSG) mice has shown that both Lin − CD34 + CD38 − CD90 − CD45RA − and Lin − CD34 + CD38 − CD90 + CD45RA − contain LT repopulating activity in secondary recipients with different frequency [98]. In addition, CD49f (integrin α6) marker has been shown as a specific HSC marker within Lin − CD34 + CD38 − CD45RA − population which as single-sorted HSC is highly efficient in generating long-term multilineage grafts while the loss of CD49f expression results in the absence of long-term grafts [98]. Furthermore, Rhodamine-123 marker (efflux of the mitochondrial dye) is added to enrich for HSCs where high Rho efflux (Rho lo )Lin − CD34 + CD38 − CD90 + CD45RA − can also repopulate all blood lineages in secondary recipients [98]. Taken together, these results demonstrate that human HSCs are enriched in the Lin − CD34 + CD38 − CD90 +/− CD45RA − Rho lo population of hematopoietic cells (Figure 4).

Signaling Pathways in Self-Renewal and Maintenance of HSCs
The balance that controls between self-renewal and differentiation (or cell fate decision) of HSCs in the bone marrow is mediated by several factors. There are a number of animal models promoting the concept that the niches inside bone marrow provide the maintenance and regulation of HSCs by some microenvironmental-dependent signals. Most HSCs are in quiescent state (i.e., in G0/G1 phase of the cell cycle) [99], however, when the hematopoietic cells disturbance occurs, hematopoiesis system will respond by shutting down or turning on the regulators mediated in the 7 regulations. Several pathways have been studied in relation to that circumstance which are SDF-1 (CXCL12)/CXCR4 signaling, BMP signaling, Mpl/Thrombopoietin (TPO) signaling, Tie2/Ang-1 signaling, hedgehog and Notch signaling, as well as Wingless (Wnt) signaling.

SDF-1 (CXCL12 )/CXCR4 Signaling Pathway.
Stromal cell-derived factor 1 (SDF-1) is constitutively expressed in several organs including lung, liver, skin, and bone marrow [100]. SDF-1 belongs to α-chemokines that functions as chemoattractant for both committed and primitive hematopoietic progenitors and regulates embryonic development including organ homeostasis [100]. There are two main splicing forms that have been identified, SDF-1α and SDF-1β, which ubiquitously expressed with highest levels detected in liver, pancreas, and spleen [101]. Additionally, another variant form, SDF-γ, has been characterized in the nervous system [102]. Subsequently, SDF-1δ, SDF-1ε, and SDF-φ have been identified with highly expression in pancreases and lower levels detection in heart, kidney, liver, and spleen [103]. SDF-1 counteracts with its cognate receptor, CXCR4 that expresses widely in numerous tissues including hematopoietic and endothelial cells to stimulate the physiological processes. SDF-1/CXCR4 signaling plays a critical role during embryonic development by regulating B-cell lymphopoiesis, myelopoiesis in bone marrow and heart ventricular septum formation [104][105][106]. In addition, SDF-1 has been shown to be mediated in the recruitment of endothelial progenitor cells (EPCs) from the bone marrow through a CXCR4 dependent mechanism suggesting the functional role in vasculogenesis in which EPCs could form blood vessels [107]. A number of observations demonstrated that there was an increase in SDF-1 expression the ischemic sites [108,109]. More evidence demonstrated that locally injection of SDF-1 augmented vasculogenesis and subsequently contributed to ischemic neovascularization in vivo by promoting EPC recruitment in ischemic tissues [110]. Recently, Liu and colleagues have shown that signal of SDF-1/CXCR4 together with CXCR7 can increase the mobilization and paracrine actions of mesenchymal stem cells (MSCs) ischemic kidneys under hypoxia condition [111]. Moreover, SDF-1/CXCR4 not only plays a role in HSC maintenance but also regulates HSC attachment within the niche. The mechanism mediated in this regulation was found to be activated through matrix metalloproteinase-9 that mediated in the releasing of soluble Kit-ligand [112]. Inactivation or deletion of CXCR4 in mice resulted in HSC pool reduction and hyperproliferation responsive to HSC defections [77,78]. Tzeng and colleagues also confirmed the role of SDF-1 in HSC maintenance by demonstrating that a conditional SDF-1-deficient mice conferred an impairment in HSC quiescence and endosteal niche localization [113].

BMP Signaling Pathway.
Bone morphogenic proteins (BMPs) are a group of growth factors that belongs to a TGF-β family member [114]. BMPs are mainly produced by osteoclasts in HSC niche [115]. During embryogenesis, BMP-4 regulates hematopoietic lineage commitment from mesodermal cell, while HSC number and function within bone marrow niche is controlled by Bmp-4 during adult life [116,117]. Knowledge of BMP signaling and receptor related adult HSC within bone marrow has been studied in a small number and is elusive. BMP signaling impairment displayed an increase in the niche size, leading to the enhancement in the number of HSCs [65]. Differential response of HSC to soluble BMPs observed by a group of Bhatia showed that higher concentrations of BMP-2, BMP-4 and BMP-7 maintained human CB HSCs in vitro while at lower concentrations of BMP-4-induced proliferation and differentiation of HSCs [118].

c-Mpl/TPO
Signaling Pathway. c-Mpl and its ligand, thrombopoietin (TPO), are known to regulate megakaryopoiesis [119]. c-Mpl receptor is expressed mainly on HSCs, with a lesser extent on megakaryocytic progenitors, megakaryocytes and platelets [120]. Various tissues expressing c-Mpl are mediated in hematopoiesis, including bone marrow, spleen, and fetal liver [14]. Based on the crystallographic EPO receptor study and its analogy to the TPO receptor have led to the postulation that TPO initiates the signal transduction by binding to the c-Mpl at the distal part, which in turn a homodimer of c-Mpl becomes active [15]. Consequently, Janus kinase 2 (JAK2) can phosphorylate tyrosine residues within the receptor itself which at least two tyrosine residues, Tyr625 and Tyr630, are phosphorylated on c-Mpl [15,122], thereby stimulating the downstream cascade STATs, PI3K, the mitogen-activated protein kinases (MAPKs), and extracellular signal regulated kinases-1 and -2 ( Figure 5) [123,124]. c-Mpl/TPO signaling involved in postnatal steady-state HSC maintenance and cell-cycle progression at the endosteal surface [125,126]. Mpl-expressed LT-HSCs were found in correlation to cell cycle quiescence and that was closely associated with TPOproducing osteoblastic cells in the bone marrow [125]. Additionally, the inhibitory adaptor protein Lnk was suggested as a negative regulator of JAK2 in HSCs following TPO stimulation, in which HSC quiescence and selfrenewal controls were predominantly through Mpl [127]. Therefore, TPO/Mpl/JAK2/Lnk pathway can be concluded as a gatekeeper for HSC quiescence. Recently, TPO knockin RAG2 −/− γc −/− mice has been shown to improve human engraftment in the bone marrow and maintenance of HSPCs pool by serial transplantation [128]. Taken together, TPO has an important function in maintenance and self-renewal of HSCs.

Hedgehog
Signaling Pathway. Hedgehog (Hh) is proposed as a negative regulator of the HSC quiescence [130]. Hh ligand binds to the transmembrane receptor Patched (Ptc) and subsequently allows the signaling function of a second transmembrane protein, Smoothened (Smo), essentially for the Hh signal to be active. Trowbridge and colleagues demonstrated that constitutive activation of the Hh signaling pathway in Ptc heterozygous (Ptc-1 +/− ) mice resulted in induction of cell cycling and expansion of primitive bone marrow hematopoietic cells [130]. To support this hypothesis, deletion of Smo in the in utero of transgenic mice was performed and the result demonstrated that there was an impaired stem cell self-renewal and the inhibition in engraftment activity of HSCs [131]. Furthermore, the common downstream positive effector of Hg signaling, Gli1, has been shown to play a critical role in normal and stress hematopoiesis [132]. Nonetheless, the discrepancies on Hg role in hematopoiesis were shown in some studies claiming that the conditional loss of Smo within adult HSCs is dispensable for hematopoiesis [133,134]. These conflicts might be due to the difference of the mice model and conditional system used to impair Hg signaling. 5.6. Notch Signaling Pathway. Notch signaling plays a key role in several fundamental functions including proliferation, differentiation and cell fate decision [135,136]. Four notch receptors (Notch 1-4) and five ligands (Jagged1-2 and Delta-like 1, 3, and 4) have been identified in mammals [137]. Cells expressing Notch ligands or engineered immobilized Notch ligands could maintain or enhance HSC self-renewal in the culture [138,139]. Some investigations demonstrated that there were an impaired HSCs differentiation both in vitro [140,141] and in vivo [139,141] studies following interaction of Notch receptors and Notch ligands. Transcription factor act upstream of the Notch signaling cascade, Hes2, was shown to be essential in HSCs formation in zebrafish embryos when hes2 expression was knockdown, whereas HSC formation could be rescued by the activation of Notch signal [142]. One study showed that an increase in in vitro maintenance of hematopoietic functions and repopulating potential on osteoblasts and Lineage − Sca-1 + CD117 + (LSK) cells coculture was mediated by the up-regulation of Notch signal (Notch2, Jagged1 and 2, Delta1 and 4, Hes1 and 5, and Deltex) [58]. Taken together, these studies support the role of Notch signaling mediated in HSC hematopoiesis and maintenance. In the contrary, some investigations proposed that Notch signaling was not important for HSC self-renewal and maintenance [143,144]. Inactivation of Notch1 and Jagged1 in bone marrow progenitors and bone marrow stroma, respectively, did not impair HSC maintenance and reconstitution [144]. The inhibition of Notch1-4 signaling via a developed dominantnegative Mastermind-like1 construct was transfected into LSK and demonstrated similar result of LT reconstitution in bone marrow compared to LSK control, except for Tcells [143]. Nevertheless, the study by Kim and colleague explored the important of Notch in normal hematopoiesis [145]. Mind bomb (Mib)-1, which regulates the endocytosis of Notch ligands and activation, was inactivated in mice leading to myeloproliferative disease (MPD). Surprisingly, when transplanted with wild-type bone marrow cells into the Mib1-null microenvironment, it results in a de novo MPD. The MPD progression was suppressed by transplantable Notch activating cells, suggesting that MPD develops from the nonhematopoietic microenvironmental cells with defective Notch signaling. Therefore, Notch signaling is indeed required for normal hematopoiesis. Santaguida and colleague developed JunB-decient mice which resulted in impairment of Notch and transforming growth factor-β (TGF-β) signaling, in part via the transcriptional regulation of Hes1 [146]. This study showed an increase in LT-HSCs proliferation and differentiation without impairing their selfrenewal in vivo, suggesting that LT-HSC proliferation rate is not exclusively compelling to self-renewal activity and maintenance of HSC in the BM niches.

Wnt Signaling Pathway.
Notch signaling is involved in the cross-talk with other pathways particularly Wnt signaling not only in hematopoiesis [147] but also in other cellular development [136,[148][149][150][151]. In addition, Wnt signaling pathway is mediated in the regulation of stem cell fate and maintenance of mouse ESCs and human ESCs in undifferentiated state [152,153]. There are at least two independent pathway comprised in Wnt signaling: canonical Wnt and noncanonical Wnt signaling pathways. The canonical Wnt signal interacts with Frizzled (Fz) receptors and single-pass co-receptors LDL-receptor-related proteins 5 and 6 (LRP 5 and 6). The Fz protein contains a conserved motif, a cysteine rich domain (CRD) located on the extracellular domain that binds to multiple Wnts with a high affinity ( Figure 6) [154]. Specifically, Dishevelled (DVL) is phosphorylated by casein kinase Iε (CKIε), which then binds typically to FRAT and confers the assembly between Fz to DVL (Fz-DVL complex) and LRP5/6 to AXIN and FRAT (LRP5/6-AXIN-FRAT complex) [155,156]. After that, β-catenin is stabilized and released from phosphorylation by CKIα and GSK3β. Then, β-catenin forms the complex with T-cell factor/lymphoid enhancer binding factor (TCF/LEF) family transcription factors and also with Legless family docking proteins (BCL9 and BCL9L) associated with PYGO family coactivators for stabilization and nuclear accumulation [157][158][159]. Typically, the downstream effectors for transcriptional activation target genes are FGF20, DKK1, WISP1, MYC, and CCND1 [160][161][162][163]. In the absence of Wnt, β-catenin is destabilized by phosphorylation of CKIα and glycogen synthase kinase 3β (GSK 3β), which then resulted in a formation of a destruction complex facilitating by Axin (β-catenin-APC-AXIN) that is polyubiquitinated by βTRCP1 or βTRCP2 complex for the degradation by proteasome [16,159].
In the second pathway, "noncanonical Wnt signal" exerts the independent β-catenin signaling. The Wnt subfamily members, for example, Wnt5a binds to the Frizzled receptor, stimulates downstream intracellular signaling, resulting in an increase in intracellular Ca2 + , and then activates protein kinase C and calmodulin-dependent kinase [159]. The cross-talk between Notch and Wnt signaling pathways was found in the stabilizing β-catenin on bone-marrow stroma cells that promoted maintaining and self-renewal of HSCs [164]. Moreover, the induction of Jagged1 and delta-like 1 was observed in Wnt/β-catenin-activated bonemarrow stroma or in bone-marrow stroma cultured with Wnt3a-conditioned medium [164]. Mice lacking Wnt3a resulted in prenatal death [165]. Moreover, Wnt3a deficiency reduced the number of HSCs in fetal liver and impaired the repopulating activity in vivo [165]. However, the exact role of Wnt signaling pathway in regulation of HSCs remains a controversy. Some studies demonstrated that constitutive activation of Wnt/β-catenin in transgenic mice resulting in the multilineage differentiation block and loss of repopulating stem cell activity due to the induction of quiescent stem cells entering into cell cycle and arresting their differentiation [166,167]. In contrast to previous works, the administration of an inhibitor of GSK-3β in vivo displayed the enhancement in the recovery of hematopoietic cells for neutrophil and megakaryocytic lineages as well as primitive LSK cell population together with the upregulation of Wnt, Notch, and Hedgehog genes [168].
Furthermore, inhibition of Wnt signaling in HSCs by overexpression of the paninhibitor of canonical Wnt signaling, Dickkopf1 (Dkk1), resulted in the induction of cell cycling and reduction in repopulating ability in transplanted induction mice [169]. When the inhibitor of GSK-3β, 6bromoindirubin 3 -oxime was used to treat CB-CD34 + cells, cell cycle progression was delayed including promoted engraftment of ex vivo-expanded HSCs [170]. Cumulatively, these studies suggest the positive regulatory role of Wnt/βcatenin signal on the proliferative or repopulating activity of HSCs. 12/15-lipoxygenase-mediated unsaturated fatty acid metabolism has been implicated in canonical Wnt-related signaling in the maintenance of LT-HSC quiescence and number [171]. Taken together, the canonical Wnt signal is mediated in the regulation of HSC function by maintaining quiescence and balance in proliferation.

Concluding Remarks
HSCs have been studied extensively for HSC source, hematopoiesis, biological functions, and signaling pathways related to the maintenance and regulation of HSCs. Advance researches such as imaging system clearly provide useful information on tracking the HSC origin, pool, and transplantation outcome in the mouse models. The observation In the absence of Wnt, β-catenin-forming complex with AXIN and APC is phosphorylates, leading to be a target for ubiquitination and degradation by proteasome (adapted from [16]).
of molecular mechanisms downstream the signaling cascade of self-renewing and proliferation of HSCs will also provide the knowledge through the new discovery in the treatment of diseases including the development in the performing a large scale preparation of HSCs for clinical transplantation. In addition, the signaling pathways will also provide understanding insight into the cancer stem cells which are now challenging scientists to explore their possible strategy for the treatments.