English
中文Abstract
Haemopoiesis requires a highly complex series of cellular events in which a small population of stem cells needs to generate continuously large populations of maturing cells. Normally, the diverse proliferative, differentiative, and maturation events required to achieve this occur with precision, which indicates that the regulatory mechanisms involved are complex. Control mechanisms tend to be intrinsic or extrinsic to the stem cells, or a combination of both. The external factors include the cell-cell interactions in the haemopoietic microenvironment, cytokines, haemopoietic growth factors and interleukins. Whilst intrinsic regulation is achieved due to the genetic factors; the expression of several transcription factors has been shown to be essential for haemopoietic cell development from the earliest stages. Determination of the control of haemopoietic stem cell proliferation kinetics is the critical factor for the regulation of haemopoietic cell production. However, information about the control of stem cell renewal versus differentiation, and how this might be manipulated to improve haemopoietic cell regeneration, is still not fully understood. Overall, extrinsic and intrinsic control mechanisms may be considered separately, but a picture is emerging of the integration of extracellular signaling, signal transduction, transcription factors and cell cycle control in the determination of stem cell fate.
Introduction
The process of producing the cellular constituents of the blood is known as haemopoiesis. This process occurs from fetal throughout adult life to replace cells which are removed from the circulation. In the embryo, this blood cell formation originates in the yolk sac, and subsequently takes place in the fetal liver and spleen (the latter to a lesser degree). By the time of birth haemopoiesis develops in the bone marrow, which is the principal haemopoietic organ. Bone marrow is one of the largest organs in the body (in an adult) and is located in the upper ends of the humerus and femur, cranium, clavicles, scapulae, sternum, ribs, vertebrae and pelvis (Hoffbrand et al. 2001). The highly organized structure of bone marrow is composed of haemopoietic cells, fat cells and vascular structures enclosed within bony tissue. An extensive network of vascular sinusoids that communicate with the peripheral circulation traverses haemopoietic tissue. The volume of haemopoietically active bone marrow in an adult is about 1-2 litres, but this can expand several-fold to meet increased needs.
Blood cells are derived from the pluripotent stem cell or haemocytoblast. This cell type can divide rapidly and differentiate into committed stem cells. The committed stem cells are colony-forming in that they are committed to produce large quantities of erythrocytes, granulocytes (neutrophils, eosinophils and basophils), monocytes-macrophages, megacaryocytes-blood platelets, and B- & T-lymphocytes depending upon various growth inducers or cytokines (Das-Gupta and Russell 2004) (Figure 1). Each day a normal adult produces approximately; 2 x 1011 erythrocytes, 1 x 1011 leukocytes and 1 x 1011 platelets, these rates of production can increase more than 10-fold under conditions of increased need (Kaushansky 2006). The pluripotent early stem cell in the bone marrow gives rise to the haemopoietic stem cell (HSC). HSC is defined as a multi-potential cell with extensive self-renewal and proliferative capacity that can differentiate into progenitors of the different blood cell types. It is estimated that HSCs comprise 0.01-0.05% of the total marrow cell population in humans (Lee 2002). A massive amplification of cell numbers is achieved as the progenitor cells (developmentally more restricted) generated by HSC division undergo further proliferation and differentiation into mature cells of the various lineages.
Aim
A highly complex series of cellular events in which a small population of stem cells needs to generate continuously large populations of maturing cells in eight major lineages is required for haemopoiesis. The precision achieved from the diverse proliferative, differentiative, and maturation events; leads to the expectation that the regulatory mechanisms involved would need to be complex (Brown at el. 1990). Also, sophisticated regulatory control would be required for the events where the entry of mature cells into the circulation, their selective localisation in appropriate tissues, and their functional activation occur. Control mechanisms can be intrinsic or extrinsic to the stem cells, or a combination of both. The aim of this paper is to discuss the extrinsic and intrinsic factors/regulators and their function in haemopoiesis.
Extrinsic factors
Responsiveness to demands for increased haemopoietic cell production can be controlled by external factors, such as cytokines or cell-cell interactions in the haemopoietic microenvironment. The haemopoietic microenvironment consists of stromal cells, including fibroblasts, macrophages, reticular cells, adipocytes, endothelial cells, and extracellular matrix that support haemopoietic progenitor cell division and maturation. Physical contact between haemopoietic progenitor cells and stromal cells is required to sustain in vitro long-term bone marrow cultures (Das-Gupta and Russell 2004).
Transmembrane proteins e.g. SCF and Flt3 ligand are produced by stromal cells. SCF binds to its receptor, c-Kit, expressed by haemopoietic stem cells and is essential for normal blood cell production. Flt3 ligand binds to Flt3 on haemopoietic cells and is important for cell survival and cytokine responsiveness. Cell cycle control and intracellular signaling can be integrated with extracellular signals via the Notch-1-Jagged pathway. Notch-1 is an example of a surface receptor present on haemopoietic stem cells that binds to its ligand, Jagged, on stromal cells. This results in cleavage of the cytoplasmic portion of Notch-1, which can then act as a transcription factor. c-Kit, the receptor for SCF, and receptors for TGF-β and tumour necrosis factor α (TNF-α) may also act in this way (Gordon 2005).
In vitro haemopoietic culture systems, advances, together with rapid progress in protein biochemistry and recombinant DNA technology, have enabled the identification, purification, gene cloning and recombinant synthesis of most of the haemopoietic growth factors. These growth factors are glycoprotein hormones that regulate the proliferation and differentiation of haemopoietic progenitor cells as well as the functions of mature blood cells. Haemopoietic growth factors effects are exerted by binding to specific receptors on target cells. A signal to the cell to proliferate or differentiate is transduced by the receptor to the cell nucleus, upon binding to the receptor. Usually, each haemopoietic growth factor acts on more than one haemopoietic cell lineage and at several stages of differentiation.
A glycoprotein hormone that is produced by the interstitial peritubular cells of the kidney is erythropoietin (EPO), a major humoral regulator of erythropoiesis. It promotes the survival, proliferation and differentiation of late erythroid progenitor cells. Absence of EPO leads to programmed cell death, or apoptosis, of erythroid progenitors. EPO acts on erythroid cells through an EPO receptor that is found on all committed erythroid cells. EPO receptors decline in number as erythroid cells mature and disappear by the stage of the late erythroblast. EPO is produced at a steady rate to maintain a normal level of erythropoiesis, under normal conditions. Hypoxic conditions in the kidney result in increased EPO synthesis. The mechanism by which interstitial peritubular cells respond to hypoxia is unknown, but probably involves a haemoprotein that senses tissue hypoxia. A reduction in erythropoietin production occurs when tissue oxygenation exceeds a certain threshold (Lee 2002).
The primary regulator of megakaryocyte maturation and platelet production is thrombopoietin (TPO). Other growth factors stimulate thrombopoiesis including IL-3, stem cell factor, IL-11 and EPO (which has significant sequence and structural similarity to TPO). TPO receptors are found on megakaryocytes and platelets and it works by promoting the proliferation and differentiation of megakaryocytes and their fragmentation into platelets. TPO is produced constitutively by several organs (such as the liver and kidney), where blood levels vary inversely with the combined megakaryocyte and platelet mass (Lee 2002).
Colony-stimulating factors are glycoproteins that stimulate the proliferation and differentiation of granulocytic and monocytic progenitors. The following colony-stimulating factors: granulocyte colony-stimulating factor (G-CSF), macrophage colony-stimulating factor (M-CSF), granulocyte-macrophage colony-stimulating factor (GM-CSF) and interleukin-3 (IL-3) regulate granulocyte and monocyte formation. Other haemopoietic cytokines with effects on the myeloid lineage include stem cell factor, flt3-ligand, IL-1, IL-5, IL-6, IL-11 and IL-12 (Lee 2002). GM-CSF and IL-3 are multi-lineage growth factors acting on multi-potential haemopoietic progenitors, whereas G-CSF and M-CSF act principally on committed haemopoietic progenitors. In addition, G-CSF, GM-CSF and M-CSF stimulate the functional activation of neutrophils and macrophages. GM-CSF also stimulates the proliferation, mobilisation and differentiation of dendritic cells.
Using ‘gene knockout’ mice the physiological roles of the colony-stimulating factors have been investigated. It was found that inactivation of the gene for G-CSF resulted in severe neutropenia and defective responses to infection. Whereas, M-CSF gene inactivation caused a severe reduction of osteoclast and macrophage numbers in mice, and developed the bone disorder osteopetrosis. Inactivation of the gene for GM-CSF resulted in defective lung macrophages and defective immune responses to infection and the clinical syndrome of alveolar proteinosis.
The issue whether haemopoietic regulators can initiate or influence maturation events in the progeny of committed cells has been discussed (Metcalf 1998). The G-CSF receptor (a C-terminal region) was shown to be required for granulocyte maturation in cell lines, and the maturation arrest of congenital neutropenia is often associated with mutations in the C-terminus of the G-CSF receptor. A clear example of the mandatory action of a regulator in achieving maturation is provided by TPO (although IL-3 is a better proliferative stimulus for megakaryocytes than TPO) only TPO can achieve full cytoplasmic maturation and platelet formation. The distinct phenotypic difference imposed on the morphology and behaviour of macrophage colony cells when part of the colony is subsequently stimulated by M-CSF and part stimulated by GM-CSF is an example of haemopoietic regulator action. However, it has been questioned whether haemopoietic regulators are necessary to initiate maturation. Actual maturation programs executed were not dependent on the particular haemopoietic regulator used as the proliferative stimulus. When growth factor was withdrawn from cultures of an immortalised cell line, but cell survival was ensured by overexpression of Bcl-2, many cells were able to undergo substantial maturation. In clonal cultures of normal marrow cells (a similar phenomenon was observed), when developing clones were transferred to factor-free cultures, cell division ceased but some maturation to granulocytic or macrophage cells was observed (Metcalf 1998).
Advances made over recent years involve EPO, which was first purified from the urine of anaemic patients. The EPO gene has now been cloned and synthetic recombinant EPO (although expensive) is used clinically for the treatment of anaemia of chronic renal failure/ diseases such as AIDS, cancers and rheumatoid arthritis. With TPO, low concentrations of made it difficult to purify from the plasma of thrombo-cytopenic patients. Sequencing of a proto-oncogene, c-mpl revealed gene sequences characteristic of a haemopoietic growth factor receptor, and inhibition of c-mpl expression in haemopoietic cells using antisense oligonucleotides resulted in the inhibition of megakaryopoiesis, indicating that c-mpl was the TPO receptor. This was confirmed by the subsequent purification of TPO (or c-mpl ligand) and cloning of the TPO gene (Lee 2002).
Over thirty years, the semi-solid colony assay system was developed and led to the purification and cloning of the colony-stimulating factors. The assay uses soft agar or methylcellulose, supplemented with liquid culture media that had been pre-conditioned by the growth of other cell types to grow clonal colonies of granulocytes and macrophages. Diffusible peptides in the conditioned media were found to preferentially stimulate either granulocyte or macrophage colony- forming activity, or both. Subsequently, murine, and then human colony-stimulating factors were purified from conditioned media and their genes cloned.
Intrinsic factors
Transcription factors, GATA-2 and SCL have been shown to be essential for haemopoietic cell development from the earliest stages. They are required for the development of haemopoiesis in the yolk sac; candidate genes that are targeted by these transcription factors include c-Kit, globin genes and myeloperoxidase (Gordon 2005). Studies (Keller et al. 1993; Weiss and Orkin 1996) from different inbred mouse strains indicated that haemopoietic progenitor cells vary widely in number and proliferative activity. The observations indicated that genetically determined constitutional variation in human haemopoiesis is also likely to exist. This was supported by the fact that parameters such as clonogenic cell numbers and frequency, proliferation ability and capacities for mobilisation and expansion vary widely among individuals in the general human population. Associations have been reported between genetic markers and the activity of stem cells and frequency in mouse strains.
A failure of the appropriate haemopoietic progenitor to differentiate into functional osteoclasts, resulted in osteopetrosis in c-fos-null mice (Shivdasani and Orkin 1996). The defect was intrinsic to the haemopoietic cell precursor, and was reversible by restoration of c-fos expression within the mutant cells. By virtue of directing a tissue-specific program of gene expression, c-fos was required for the differentiation of a specific subset of precursors. This was similar with the related bZip transcription factor p45 NF-E2 (lineage-restricted). Each plays an essential role in terminal differentiation, but not in lineage commitment, of a specific subset of haemopoietic progenitors.
It was concluded (De Haan et al. 2002) that the expression levels of a large number of genes may be responsible for controlling stem cell behaviour. These collections of genes may be analogous to those responsible for the inter-individual behaviour of human haemopoietic stem cells. Certain specific genes and gene products have been demonstrated to influence haemopoietic cell kinetics, such as the cyclindependent kinase inhibitors (CKIs). They have been shown to enhance proliferation and repopulating efficiency of bone marrow cells in gene knockout/ knockin and transfer models. Loss of CKIs was found to increase clonal expansion by haemopoietic progenitor cells and the size of the stem cell pool; the Fas and Fas ligand genes, which generally are associated with the process of cell death by apoptosis, also influence haemopoiesis as part of a mechanism suppressing progenitor cell proliferation (De Haan et al. 2002).
From studies of disease pathogenesis (Tischkowitz and Hodgson 2003), many cell cycle control genes and genes promoting cell death by apoptosis are tumour-suppressor genes that have been found to be deleted or mutated in leukaemia and other cancers. Fanconi’s anaemia is an autosomal recessive bone marrow failure syndrome associated with an increased tendency for spontaneous chromosome breaks. The disease can be caused by mutations in at least seven different genes. Some of the genes have been cloned, and the corresponding proteins played important roles in DNA repair. In dyskeratosis (abnormal keratinisation) congenita mutations have been identified in the DKC1 gene, which encodes dyskerin. Dyskerin is a component of small nucleolar ribonuclear protein particles and the telomerase complex, indicating that the disease is due to defective telomerase.
Differentiation is fundamentally a process of ordered gene regulation that culminates in the expression of a unique complement of specific and widely expressed genes in each cell type. Studies (Thomis et al. 1995) of critical lineage-specific factors have provided an invaluable window into the molecular basis of cell differentiation. Within haematopoietic cells, an essential role has been shown for several signal transduction molecules, including erythropoietin, the growth factors G-CSF, GM-CSF, thrombopoietin, the receptor tyrosine kinase Flk-1 and the Janus kinase Jak3. Haemopoietic defects have been noted in diverse knockout models, including the C-myc, platelet-derived growth factor, RelB and Retinoblastoma genes (Shivdasani and Orkin 1996).
Selection between alternate lineages likely involves both activation and silencing of distinct subsets of genes; the competing activities of various transcriptional regulators may lie at the heart of this process. An important goal for future research is to define the mechanisms by which transcription factors operate within regulatory networks to orchestrate lineage selection and haemopoietic development (Shivdasani and Orkin 1996). Numerous influences modulate transcription factor function, including signals delivered by growth factors and cytokines, the stage of the cell cycle and cell-cell interactions. Although the impact of some of these influences on transcriptional regulators is known, their effect on lineage-specific factors has largely been unexplored and highlights a future area of study. The initiation of key transcription factors in progenitor cells is unclear. Possibilities include an unregulated process or a finely tuned response to external signals. Further study of the regulation of some key transcription factors will allow this question to be addressed.
Due to the simple studies used such as phenotypic analysis of knockout mice, questions remain largely unanswered e.g. whether the mechanisms by which loss of critical target genes interferes with cell differentiation. The real hierarchy among regulatory factors is in fact speculative, and it is likely that multiple interactions among the listed proteins and among other lineage-restricted or ubiquitous factors occur throughout the developmental cascade; these are not shown by knockout phenotypes (only the earliest block in a pathway is shown). Combinations of transcription factors rather than single proteins should be used to determine the genetic program of a cell. The benefits of knockout experiments have been the establishment of candidate proteins as essential regulators of differentiation (Verfaille 1998). Gene knockouts have thus set the stage for future identification of crucial target genes and for the study of how transcriptional regulation of individual genes impacts on cell differentiation. Due to molecular cloning and expression in suitable vector systems of the genes for many growth factors, in vivo experiments are readily available (Dexter 1987).
At present, the arguments for and against particular models for haemopoiesis are based on circumstantial evidence. The availability of a large number of antigenic markers for haemopoietic cells at various stages of maturation allows the possibility to define and purify haemopoietic stem cells. Computational models, to predict stem cell systems and discern underlying regulatory mechanisms governing stem cell fate decisions, have been investigated (Viswanathan and Zandstra 2003). HSC were the system of choice to investigate mechanisms that regulate stem cell fate decisions, and have thus served as an experimental and computational model for other stem cells.
Overall, intrinsic and extrinsic control mechanisms may be considered separately, but a picture is emerging of the integration of extracellular signalling, signal transduction, transcription factors and cell cycle control in the determination of stem cell fate. It seems incorrect to regard the multiplicity of haematopoietic regulators as representing a highly redundant control system of dubious value. There is mounting evidence of the necessity to use regulator combinations to achieve certain types of cell production and of the higher efficiency of multiple regulatory factors, even when their actions appear to exhibit considerable overlap (Metcalf 1993). These advantages also include the ability to achieve more subtle localisation of cells produced and to achieve the complex mixtures of the cells required in certain situations.
