- Red blood cells: RBC production (and reticulocyte maturation), structure and function.
- Neutrophils: WBC production, structure and function.
- Platelets: Platelet production, structure and function.
Hematopoiesis is the production of blood cells (myeloid cells, erythroid cells, platelets and lymphocytes). In the fetus, hematopoiesis occurs in three distinct waves. The first, or primitive wave, occurs in the yolk sac. This initial wave produces mostly erythroid precursors, diploid platelet precursors, and macrophages from a common precursor cell known as a haemangioblast (this also produces endothelial cells). During the second wave, a more differentiated common erythroid-myeloid progenitor (EMP) emerges. The EMPs migrate to the fetal liver and ultimately to the bone marrow. Homing to the bone marrow is largely driven by a cytokine, stromal derived factor-1α (SDF-1α) or CXCL12, and its receptor, CXCR4. SDF-1α is secreted in high concentrations by stromal cells around the vascular sinuses in marrow, thus helping to recruit CXCR4-expressing stem cells (Blau 2014). In the third wave, hematopoietic stem cells (HSC) arise from the haemogenic endothelium within specific locations distributed throughout the vasculature and again seed the fetal liver and bone marrow. The exact time when the second wave turns over to the third wave in the fetus is not known.
In most adult vertebrates, hematopoiesis occurs within the bone marrow. When hematopoiesis occurs outside the marrow, it is termed extramedullary hematopoiesis. Extramedullary hematopoiesis can occur in many sites, but the most common sites, where we expect to see it under physiologic conditions, is the spleen (red pulp) and liver (portal areas). Localized hypoxia has been postulated to be the driving force behind extramedullary hematopoiesis, whereby stromal cells secrete high concentrations of SDF-1α, causing homing of stem cells similar to that which occurs normally in the bone marrow (Johns and Christopher 2012). The continued production of blood cells depends on the maintenance of the hematopoietic niches in the bone marrow and each lineage has unique means of regulating the day to day production of cells (adult or steady-state hematopoiesis) and mechanisms to respond to increased demands in the context of inflammatory or infectious disease (stress hematopoiesis).
Within the bone marrow, there are different niches in which hematopoiesis occurs. The area closest to the endosteal surface of the trabecular bone is called the "osteoblastic" or "endosteal" niche. It is the furthest away from the vascular sinuses that feed the marrow (with oxygen and nutrients) and is considered to be relatively hypoxic. This is also the area where the earliest stem cells are maintained in a quiescent state, which is thought to be largely mediated by the hypoxic environment in this location. The area closest to the vascular sinuses is called the "vascular" or "endothelial" niche and this is where megakaryocytes and erythroid progenitors largely reside. Throughout the marrow, along with hematopoietic cells, is also an array of bone marrow or mesenchymal stromal cells (MSCs) that make up the marrow microenvironment. The MSCs include osteoblasts, chondrocytes, macrophages, adipocytes, endothelial cells and pericytes. These cells secrete cytokines (such as stem cell factor) and interact directly with hematopoietic cells via adhesive ligands, both of which drive differentiation of hematopoietic stem cells (Blau 2014). Macrophages have a prominent role in marrow with supporting hematopoiesis, particularly erythropoiesis (erythroid progenitors frequently surround marrow macrophages in so-called "erythroblastic islands"). In contrast, the role of osteoprogenitor cells (osteoblasts, osteoclasts) supporting hematopoiesis is not fully elucidated, but a number of factors that affect bone metabolism and turnover can also affect hematopoiesis. Neurons are also an important component of the bone marrow stroma. The innervation of the bone marrow is partly responsible for some of the changes we can see in an animal’s hemogram when they are excited or stressed and is mediated through norepinephrine and epinephrine. The adipocytes in the bone marrow are also an active part of the stromal compartment. Marrow adipose tissue, unlike fat beds in soft tissues, is confined by the margins of the skeleton. Therefore, if the adipose tissue expands, there is less room for hematopoiesis; this is thought to contribute to the decreased hematopoiesis noted in obese patients. Based on studies in people, marrow adipose tissue also expands in patients with diabetes mellitus, anorexia, and with aging and glucocorticoid therapy. In inflammatory diseases, the MSCs in marrow develop more readily into adipocytes. Cell culture experiments have shown that marrow adipocytes are capable of secreting various cytokines (including IL-6, G-CSF and GM-CSF), which influence hematopoiesis. Clearly, adipocytes in marrow should not be viewed as simple space-fillers.
Self-renewing, multipotent HSCs are the source of all blood cells. Long-term HSC are thought to divide once or twice per year, whereas more short term HSC divide on average once per month. The differentiation of these HSCs into the various lineages (erythrocytes, myeloid cells, platelets and lymphocytes) has been extensively investigated and the proposed models are ever evolving. Central to all models is the progressive loss of self-renewal with concurrent differentiation towards a more-defined cell type. The HSCs develop into multipotent stem cells, then oligopotent stem cells and ultimately into unipotent progenitor cells (these are committed not just to a lineage, but to one cell type). The early models of hematopoiesis were depicted by a tree-like branching system that relied on what was termed binary cell fate decisions at each branching point (no going back). In these models, the HSC split off early into separate lymphoid and myeloid compartments. Although the latter concept helps with understanding different leukemia (myeloid versus lymphoid), from mice models of hematopoiesis where lineage fate can be tracked, it does not appear that this is the case. These mouse studies have led to newer models of hematopoiesis, such as the cloud model (Velten et al 2017), which propose more cross-lineage and continuous reprogramming. In addition, some of these newer models show that lymphoid and myeloid progenitors can have more commonality in differentiation for longer than previously considered.
The balance between self-renewal and differentiation of HSC is regulated by metabolic signals, adhesion molecules, hormones and cytokines. Aerobic respiration produces a high yield of ATP (36 molecules from each glucose fully oxidized). While anaerobic glycolysis yields far less ATP, it generates more NADPH which serves as a building block in the synthesis of macromolecules such as amino acids and nucleotides that are required for cell division. Thus, it is postulated that the hypoxic niche in the bone marrow helps regulate their “stemness.” HSC cultured in low oxygen environments have been shown to be more efficient at repopulating the bone marrow than those cultured under normoxemia. The oxygen tension in the bone marrow is lowest near the endosteal surface where cells tend to remain in a quiescent state. Moving towards the central area of the bone marrow, there is not only an increase in oxygen tension but also differentiation of the cells. The low oxygen levels in the stem niche are also thought to limit production of reactive oxygen species and damage to the DNA of these precious cells. The signals that drive asymmetric division of stem cells, where one cell differentiates into a daughter cell and the other retains stem cell properties, are not fully characterized. However, fatty acid oxidation has been shown to be a key regulatory switch in this pathway and serves largely to retain the stem cell nature of the HSC. The switch from fatty acid oxidation to oxidative phosphorylation is concurrent with differentiation of the HSC and a decreased potential for self-renewal.