Following chemotherapy and/or the administration of growth factors, such as granulocyte-colony stimulated factor (G-CSF), hematopoietic stem cells (HSC) mobilize from bone marrow to peripheral blood. This review aims to systematically present the structure of the HSC “niche” and elucidate the mechanisms of their mobilization. However, this field is constantly evolving and new pathways and molecules have been shown to contribute to the mobilization process. Understanding the importance and the possible primary pathophysiologic role of each pathway is rather difficult, since they share various overlapping components. The primary initiating event for the mobilization of HSC is chemotherapy-induced endogenous G-CSF production or exogenous G-CSF administration. G-CSF induces proliferation and expansion of the myelomonocytic series, which leads to proteolytic enzyme activation. These enzymes result in disruption of various receptor-ligand bonds, which leads to the disanchorage of HSC from the bone marrow stroma. In everyday clinical practice, CXC chemokine receptor-4 (CXCR4) antagonists are now being used as mobilization agents in order to improve HSC collection. Furthermore, based on the proposed mechanisms of HSC mobilization, novel mobilizing agents have been developed and are currently evaluated in preclinical and clinical studies.
Autologous hematopoietic stem cell transplantation (ASCT) is a widely used therapeutic strategy in the treatment of multiple myeloma and relapsed/refractory lymphomas. It can offer long-term disease control or even cure in a substantial proportion of patients. The prerequisite of ASCT is a successful and adequate stem cell mobilization and collection. Initial observations regarding the steady state circulation of hematopoietic stem cells (HSC) in the blood led to the study of HSC kinetics after the administration of chemotherapy with or without growth factors. Thus, nowadays, collection of HSC from the bone marrow (BM) has been neglected at least in the autologous transplantation setting and has been largely replaced by peripheral blood stem cell collection through cell separators. Patients who fail to collect ≥2.0 × 106 CD34+ stem cells/kg of body weight cannot undergo ASCT and thus experience its benefits. The deep knowledge and understanding of HSC mobilization will give insight into the mechanisms of “poor mobilization” and moreover may help in developing new mobilizing agents.
The term HSC was introduced for the first time by Alexander Maximov in 1909 [
Anatomically, the niche is located in close proximity to the endosteum and is supported by a variety of cells and molecules referred as “stroma.” The main representatives of the stroma are bone tissue cells (osteoblasts (OB), osteoclasts (OC), osteomacrophages (OMAC), chondrocytes, fibroblasts, and fat cells), reticuloendothelial cells (dendritic cells, lymphocytes, and macrophages), endothelial cells, as well as mesenchymal stem cells (MSC), myocytes, and cells of the autonomous nervous system. Noncellular stromal elements include the extracellular matrix (ECM), collagen, and minerals [
Three types of niches have been recognized: the endosteal (osteoblastic), the reticular, and the vascular (endothelial). The former is located at the endosteum and consists mainly of the “spindle-shaped N-cadherin+CD45− osteoblastic cells” (SNO) [
The stem cell niche is essential for the quiescence of HSC. More than 70% of them are in the G0 phase of the cell cycle, while only 10% of their progenies are quiescent. It has been shown that approximately 30% of the quiescent HSC divide every 145–193 days, while a more active subpopulation does so every 28–36 days [
LT-HSC protect themselves from DNA damage by limiting the number of their cellular divisions. The main DNA-repair mechanism of HSC is the nonhomologous end joining, NHEJ [
It is obvious that the description of the aforementioned three types of niches is undertaken for understanding purposes. They do not represent different anatomical compartments, since bone and vessels are in close proximity to each other [
A concise presentation of the main cells, receptors, and ligands that interact with each other during the process of mobilization is shown in Table
Receptors, ligands, and adhesions molecules involved in the homeostasis of the HSC niche.
Expression from HSC | Expression from BM stroma | |
---|---|---|
Tie2 |
|
Ang-1 |
Mpl |
|
TPO |
c-kit (CD117) |
|
SCF |
CXCR4 |
|
SDF-1 (CXCL12) |
TGF- |
|
TGF- |
FGFR1-4 |
|
FGF |
Notch |
|
Jagged-1 |
GRP78 |
|
Cripto |
PTH-R | ||
HIF-1 |
|
Cripto |
N-cadherin |
|
N-cadherin |
BMP-R2A |
|
BMP |
VLA4 |
|
VCAM-1 |
Frizzled |
|
Wnt (ECM) |
FLT-3 (CD135) |
|
FLT3 ligand |
CaR |
|
Ca++ (ECM) |
HCAM (CD44) |
|
Hyaluronan (ECM) |
LFA-1 |
|
ICAM-1 |
VEGF |
|
VEGFR |
Agrin-R |
|
Agrin proteoglycan (ECM) |
S1P1 |
|
S1P |
Ang-1: angiopoietin-1, SCF: stem cell factor, SDF-1: stromal-derived factor-1, FGF: fibroblast growth factor, PTH-R: parathormone receptor, Cripto: teratocarcinoma derived growth factor-1/TDGF-1, BMP: bone morphogenic protein, VLA4: very late antigen 4, VCAM-1: vascular cellular adhesion molecule-1, FLT-3: fms-like tyrosine kinase 3, HCAM: hyaluronan binding-cellular adhesion molecule, LFA-1: lymphocyte function-associated antigen-1, VEGF: vascular endothelial growth factor, S1P: sphingosine-1-phosphate, and ECM: extracellular matrix.
In 1976 Richman et al. described an increase of HSC in the blood of patients who had undergone chemotherapy [
G-CSF promotes proliferation and maturation of the myeloid series, while at the same time it induces substantial changes in the bone marrow stroma, leading to the rise of HSC in the circulation by 60 times compared to baseline [ It directly activates OMAC and macrophages, a fact that downregulates neighboring SNO, Nes+ MSC, and CAR, leading to reduced production of stromal-derived factor-1 (SDF-1) by these latter cells [ The hyperplastic myelomonocytic series (through G-CSF) secrete a large variety of proteases, which induce proteolytic cleavage/clearance of SDF-1, leading to the release of the CXCR4 receptors on HSC and their subsequent liberation from the BM stroma. The activity of the proteases is further assisted by the cleavage of protease inhibitors [ The same proteolytic G-CSF-induced mechanism is responsible for the degradation of VCAM-1, fibronectin, and OPN, leading to reduced cellular adhesion of HSC through their receptor VLA-4 to BM stroma [ G-CSF evokes a shift of HSC to more central locations in BM, close to the vascular endothelium due to increased oxygenation needs of the HSC [ G-CSF reacts as a potent noradrenalin reuptake inhibitor. An additive effect of G-CSF and tricyclic antidepressants in HSC mobilization has been shown [
The two basic G-CSF-induced mechanisms of SDF-1/CXCR4 and VCAM-1/VLA4 disruption have a synergistic effect. Coadministration of G-CSF and CXCR4 inhibitors or G-CSF and anti-VLA4 antibodies results in additive and more potent HSC mobilization compared to G-CSF alone [
SDF-1 (CXCL12) is a CXC chemokine, secreted by various BM stromal cells, such as CAR, Nes+ MSC, osteoblasts, and endothelial cells. The interaction of SDF-1 with its receptor CXCR4 on HSC plays a key role in HSC retention and trafficking. The expression of CXCR-4 on HSC is enhanced through a signaling cascade involving cAMP, phosphatidylinositol-4,5-biphosphate 3 kinase (PI3K), a number of GTPases, and atypical protein kinase C isoform
The recently introduced mobilizing agent CXCR4 inhibitor plerixafor (AMD3100) disrupts the SDF-1/CXCR4 axis in a synergistic way to G-CSF. The following hypotheses have been proposed for its mechanism of action [
Comparing the differences between G-CSF- and plerixafor-induced mobilization, it should be noted that CXCR4 antagonists lead to a more rapid mobilization. Additionally, CXCR4 antagonists do not induce either myeloid hyperplasia or proteolysis. It has been shown from experiments in primates that myeloid hyperplasia is not necessary for mobilization [
The SDF-1/CXCR4 axis seems to be influenced by sphingolipids, with the main representative being sphingosine-1-phosphate (S1P), which is abundantly produced by red blood cells, activated platelets, and endothelial cells [
Adhesion molecules are surface antigens that facilitate cell-cell and cell-extracellular matrix (ECM) interactions, through their respective receptors. They play a major role in inflammation, mediating trafficking, endothelial rolling, adhesion, and extravasation of leukocytes and lymphocytes. They are essential for cellular immune responses, normal hematopoiesis, and differentiation, as well as for the organization of cells within tissues during ontogenesis. They are widely distributed on hematopoietic and nonhematopoietic cells [ The VLA-4/VCAM-1 axis: the CD44 (HCAM): CD44 has many isoforms. Among these, CD44s, the smallest one, is the most common one expressed on HSC. The major ligand of CD44 is hyaluronic acid (HA), a component of ECM. The most HA-rich regions of the bone marrow stroma are the endosteum and the sinusoidal endothelium, the sites that are considered the niches of the most primitive HSC. CD44/HA interactions are essential for homing of HSC in the BM. Membranal CD44 cleavage on HSC is associated with mobilization and depends on the equilibrium of a membrane-bound proteolytic enzyme (membrane-type-1 metalloproteinase-MT-1 MMP) and its inhibitor RECK. G-CSF administration leads to increased expression of MT-1 MMP on CD34+ cells, which in turn results in membranal CD44 cleavage and HSC egress from the BM [
Stem cell factor (SCF) is an important chemokine produced by SNO and CAR. Its active form binds to c-kit on the surface of HSC. C-kit (CD117) is a type III tyrosine kinase receptor and is also expressed by endothelial progenitor cells and MSC. The SCF/c-kit axis plays an important role in embryonic hematopoiesis and cellular differentiation. SCF has been shown to augment engraftment in mice through increased CXCR-4 expression on HSC [
It is nowadays widely accepted that bone tissue plays an important role in maintaining and regulating the HSC niche. Bone remodeling is a complex process and its consequences on HSC retention and mobilization are an evolving field [
Receptor activator of NF-
The administration of RANKL in mice resulted in an expected increase of osteoclastic activity, but also a simultaneous HSC mobilization, was observed [
In an effort to elucidate the role of bone tissue in human HSC mobilization, we undertook a study of various biochemical markers of bone remodeling, osteoclast/osteoblast regulators, and angiogenic cytokines in the process of HSC mobilization. We studied 24 patients, 10 with multiple myeloma, 5 with Hodgkin’s Lymphoma, and 9 with non-Hodgkin’s lymphoma who were mobilized with G-CSF +/− chemotherapy [
CD45, known as leukocyte common antigen, is expressed by all leukocytes, including HSC. It is a transmembrane protein tyrosine phosphatase. Through dephosphorylation of Src kinase, it is involved in signal transduction pathways, regulating several cell processes. CD45 has been mainly studied in lymphocyte maturation, proliferation, and activation. Recent data have demonstrated a pivotal role of CD45 in HSC motility and microenvironmental regulation in both mice and human. From CD45 knockout mice experiments and CD45 treated human HSC in xenotransplantation experiments, it has been shown that CD45 is essential for HSC mobilization. G-CSF leads to increased expression of CD45 in bone marrow mononuclear cells, which in turn correlates with HSC release from the bone marrow. HSC from mice lacking CD45 show reduced mobilization, impaired ability to cross extracellular matrix barriers, and a hyperadhesive phenotype. In addition CD45 is required for normal osteoclastic development and function. CD45 negative osteoclasts display abnormal morphology and have a reduced ability to form multinucleated cells. Moreover they secrete lower amounts of MMP-9 and have a reduced bone resorbing activity. CD45 negative osteoclasts display a reduced response to RANKL resulting in poor HSC mobilization. Thus CD45 is a paradigm of the concurrent regulation of hematopoiesis and the microenvironment [
The main regulator of hypoxia in the endosteal niche is HIF-1, consisting of HIF-1
It has been shown that activation of m-TOR and the resultant increase of reactive oxygen species (ROS) in the BM lead to the egress of HSC, whereas the inhibition of mTOR with rapamycin has the opposite effect. It appears that a “critical” level of ROS and HIF-1
As already stated, the main initiating molecule of the HSC mobilization process is exogenous or chemotherapy-induced endogenous G-CSF. For many years, the administration of G-CSF with or without chemotherapy has been the mainstay of mobilization in every day clinical practice. Ifosfamide-containing (IGEV and ICE) or platinum-containing (ESHAP and DHAP) regimens with G-CSF are usually applied in lymphomas [
The endpoint of every mobilization attempt is the collection of adequate numbers of HSC, reflected practically by the absolute number of CD34+ cells collected per kg of body weight. Adequate number is considered the dose of HSC that ensures rapid and sustained long-term hematopoiesis after the administration of myeloablative chemotherapy and HSC infusion. The optimal dose of HSC is 5 × 106/kg, with little clinical benefit with doses 5–8 × 106/kg and no further improvement with grafts containing >10 × 106/kg CD34+ cells [
Still, there are a substantial proportion of patients who fail to mobilize and collect adequate numbers of HSC and consequently cannot proceed to autologous stem cell transplantation. The inability of adequate HSC mobilization is associated with HSC reserve exhaustion and disruption of BM stroma [
Risk factors for poor mobilization include age >65 years, bone marrow cellularity <30%, bone marrow infiltration, multiple previous chemotherapy regimens, especially alkylating agents, fludarabine, >4 cycles of lenalidomide, and previous extended field radiotherapy including the pelvis [
There is not a widely accepted definition of the “poor mobilizer.” For this purpose, the Italian group of apheresis has proposed criteria not only for the definition, but also for the early recognition of the poor mobilizer [
Plerixafor is a reversible CXCR4 antagonist and its mechanisms of action have been already described. It has been approved in both the US and the EU for the mobilization of patients with lymphoma and myeloma. With its use in combination with G-CSF, approximately 70% of poor mobilizers may succeed in the collection of adequate numbers of HSC in order to proceed to ASCT [
The high cost associated with plerixafor requires a stringent protocol with well-defined criteria for its proper use [
Although plerixafor rescues a significant proportion of patients from failing collection of HSC, there is still a 35% possibility of its failure. Towards this aim, new mobilizing agents are being tested either in preclinical or clinical studies.
Agents targeting the SDF1/CXCR4 axis: POL6326 (Polyphor, Allschwil, Switzerland) is a potent selective CXCR4 inhibitor, which has shown considerable efficacy as monotherapy in newly diagnosed multiple myeloma patients [ Proteasome inhibitors: bortezomib is nowadays considered among the leading therapeutic drugs in multiple myeloma. Surprisingly, in animal studies, bortezomib proved to be a potent mobilizing agent, increasing significantly CFU-Cs compared to placebo. In addition, a synergistic effect of bortezomib with G-CSF, plerixafor, and chemotherapy was evident [ Parathormone (PTH): Brunner et al. showed that PTH induces a significant increase of progenitor cells in the peripheral blood (1.5- to 9.8-fold) of mice. The authors postulated that this activity relates to endogenous release of G-CSF [
Other novel agents currently evaluated in preclinical studies are FG-4497 that stabilizes HIF-1 through inhibition of its hydroxylation [
The mechanisms of HSC mobilization are overlapping and not fully elucidated yet. Experimental studies have given light into many aspects of this cascade of events. However, we are still far away from establishing an integrated model of the mechanisms that control the equilibrium of HSC between quiescence and mobilization. Laborious research is mandatory for the development of newer agents that might render HSC mobilization and collection possible for all patients.
The authors declare that there is no conflict of interests regarding the publication of this paper.