SEARCH

SEARCH BY CITATION

Keywords:

  • Hematopoietic stem cells;
  • Tissue stem cells

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. The Hematopoietic Family Tree
  5. The Origin of Stem Cells—A Single Episode or Continuous Generation?
  6. The Bone Marrow Origin of Gonadal Stem Cells—A Sterile Aside
  7. Self-Generation by Stem and Other Cells
  8. Blast Colony-Forming Cells
  9. Lineage-Committed Colony- Forming Cells
  10. The Behavior of Supposedly Mature End Cells
  11. Conclusions from the Reculturing Experiments
  12. Can the Pattern of Self-Generation and Commitment Be Manipulated?
  13. General Conclusions
  14. Disclosure of Potential Conflicts of Interest
  15. Acknowledgements
  16. References

The term hematopoietic stem cells has at times been used to include a miscellany of precursor cells ranging from multipotential self-generating cells to lineage-restricted progenitors with little capacity for self-generation. It is probable that the stem cells of other tissues also vary widely in their multipotentiality and proliferative capacity. This review questions several dogmas regarding the self-generative capacity of various hematopoietic cells, the single episodic origin of hematopoietic cells, and the irreversible nature of progressive mature cell formation in individual hematopoietic lineages.

Disclosure of potential conflicts of interest is found at the end of this article.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. The Hematopoietic Family Tree
  5. The Origin of Stem Cells—A Single Episode or Continuous Generation?
  6. The Bone Marrow Origin of Gonadal Stem Cells—A Sterile Aside
  7. Self-Generation by Stem and Other Cells
  8. Blast Colony-Forming Cells
  9. Lineage-Committed Colony- Forming Cells
  10. The Behavior of Supposedly Mature End Cells
  11. Conclusions from the Reculturing Experiments
  12. Can the Pattern of Self-Generation and Commitment Be Manipulated?
  13. General Conclusions
  14. Disclosure of Potential Conflicts of Interest
  15. Acknowledgements
  16. References

It is now slightly more than 45 years since the discovery of spleen colony-forming cells (CFU-S) [1] and the belief that these were the sought for hematopoietic stem cells. Since then, hematopoietic stem cells (HSC) and their progressively committed progeny have become the prototypical examples of what might be expected of candidate stem cells in other tissues. It is useful, therefore, to review somewhat critically what actually has been firmly established about HSC, what is believed or assumed, perhaps incorrectly, and what skeletons are tucked away in the family tree because they do not fit well in an orderly model. This review is the written version of a recent lecture and is deliberately referenced only lightly. The objective is to allow workers with candidate stem cells in other tissues to compare the properties they have established about their stem cells with what is known about hematopoietic stem cells and in this way to assess what information remains in need of discovery about their cells.

The word “stem cell” needs some definition because, for different workers, the term can embrace quite different cells. My view is that a hematopoietic stem cell should be capable of self-generation and be multipotential and, thus, potentially be able to form maturing cells in all eight major hematopoietic lineages. What has become apparent is that cells more or less fitting this definition are likely to be quite heterogeneous. Some—a minority—may also be totipotent and able to form cells of other tissue types [2]. Most stem cells are quite heterogeneous in their proliferative and self-generative capacity. I suspect of particular relevance to workers on tissue stem cells is the fact that some cells with a claim to the term “stem cells” do not possess the complete properties of a stem cell. Some hematopoietic cells may have substantial self-generative capacity but no longer be multipotential. More numerous are lineage-committed hematopoietic cells with considerable clonogenic (proliferative) capacity but no capacity for self-generation. Finally, there are some mature cells with a quite surprising proliferative (self-generative?) capacity. Cells of all the above types are likely to be encountered in studies on tissue stem cells and, because the various cells are likely to be related ancestrally, they need to be fitted into some sort of family tree.

The Hematopoietic Family Tree

  1. Top of page
  2. Abstract
  3. Introduction
  4. The Hematopoietic Family Tree
  5. The Origin of Stem Cells—A Single Episode or Continuous Generation?
  6. The Bone Marrow Origin of Gonadal Stem Cells—A Sterile Aside
  7. Self-Generation by Stem and Other Cells
  8. Blast Colony-Forming Cells
  9. Lineage-Committed Colony- Forming Cells
  10. The Behavior of Supposedly Mature End Cells
  11. Conclusions from the Reculturing Experiments
  12. Can the Pattern of Self-Generation and Commitment Be Manipulated?
  13. General Conclusions
  14. Disclosure of Potential Conflicts of Interest
  15. Acknowledgements
  16. References

A typical family tree diagram for hematopoietic cells is shown in Figure 1. The rudiments of this family tree date back at least as far as Pappenheim and Maximov in the early 20th century. At that time it was guesswork based largely on an extrapolation between two more or less secure reference points—at one end, a likely common origin of hematopoietic precursors and endothelial cells and, at the other end, the quite evident orderly morphological progression in various lineages of blast cells through to mature end cells.

thumbnail image

Figure Figure 1.. The conventional view of hematopoiesis in which multipotential stem cells are self-generating and also produce precursor cells with increasing restriction of their lineage and proliferative potential. Within each lineage, population cell numbers rise with increasing maturity. Mast cell category includes basophils. In the interest of simplicity, the diagram has omitted natural killer cells and dendritic cells. The hematopoietic stem cells in the marrow are heterogeneous, and the marrow also contains mesenchymal stem cells and perhaps a few totipotential stem cells. Abbreviations: BFU-E, burst-forming units–erythroid; CFU-S, colony-forming units–spleen; CLP, common lymphoid progenitors; CMP, common myeloid progenitors; Eo-CFC, eosinophil–colony-forming cells; G-CFC, granulocyte–colony-forming cells; GM-CFC, granulocyte macrophage–colony-forming cells; GMP, granulocyte-macrophage progenitors; HSC, hematopoietic stem cell; M-CFC, macrophage–colony-forming cells; Mast-CFC, mast–colony-forming cells; Meg-CFC, megakaryocyte–colony-forming cells; MEP, megakaryocyte-erythroid progenitors.

Download figure to PowerPoint

What the family tree in Figure 1 implies is that continuous cell formation occurs with individual stem cells being both self-generating and generating more mature progeny. The model further proposes that a succession of precursor cells exists with cell-generating capacity but a progressively lesser capacity for self-generation. Finally, a series of irrevocably lineage-committed cells is generated that ultimately loses all proliferative capacity as it progressively matures.

Some major reservations should be raised about the reality of this proposed sequence. Need multipotential stem cells be continually involved in hematopoiesis? These cells are agreed to be nondividing or very slowly dividing. How good is the evidence excluding the possibility that more mature progeny are perfectly capable of sustaining hematopoiesis in normal life with stem cells perhaps only becoming activated in extreme emergency? Furthermore, with much of the current interest in the possible plasticity of commitment, how secure are the dogmas of lineage fidelity and the irreversibility of the progressive restriction in proliferative capacity implied by the model?

It needs to be emphasized that much of the dogma regarding the hematopoietic family tree has arisen from the use of lethally irradiated mice and the properties of cells able to generate maturing cells in such extremely stressed animals. The rarer use of less extreme recipient animals has painted a quite different picture of how hematopoiesis may occur. An important general cautionary note is therefore to beware of drawing broad and exclusive conclusions from the use of extreme damage models. In this regard, work so far on stem cells for tissues such as liver, lung, or muscle has found the need to first induce major organ damage in order to observe repopulation. This approach runs the same risk encountered by hematopoietic workers who have been led to conclusions that may be misleading.

The Origin of Stem Cells—A Single Episode or Continuous Generation?

  1. Top of page
  2. Abstract
  3. Introduction
  4. The Hematopoietic Family Tree
  5. The Origin of Stem Cells—A Single Episode or Continuous Generation?
  6. The Bone Marrow Origin of Gonadal Stem Cells—A Sterile Aside
  7. Self-Generation by Stem and Other Cells
  8. Blast Colony-Forming Cells
  9. Lineage-Committed Colony- Forming Cells
  10. The Behavior of Supposedly Mature End Cells
  11. Conclusions from the Reculturing Experiments
  12. Can the Pattern of Self-Generation and Commitment Be Manipulated?
  13. General Conclusions
  14. Disclosure of Potential Conflicts of Interest
  15. Acknowledgements
  16. References

Controversy continues about the site of origin of the first genuine hematopoietic stem cells. Initially, the yolk sac was proposed as the site of origin of such cells, which then migrate into the body in parallel with primordial stem cells and melanoblast precursors. Removal of the yolk sac prevented the subsequent development of hematopoietic populations in the fetal liver [3]. Later studies using quail/chicken chimeras disputed this view and appeared to document that the yolk sac-derived populations were transient and that definitive, persisting hematopoiesis arises from cells in the aortic-gonadal region (AGM) around the great vessels (Fig. 2) [4].

thumbnail image

Figure Figure 2.. Hematopoiesis begins in the yolk sac with migrants then seeding in the fetal liver. This population is transient and is replaced by the progeny of hematopoietic cells generated in the aortic-gonadal mesonephros region. Subsequent migration streams populate the developing bone marrow and spleen.

Download figure to PowerPoint

It has become accepted that the initiation of hematopoiesis occurs during a strictly limited period after which no further de novo formation of hematopoietic cells can occur. If the yolk sac is indeed the true initiating source of hematopoietic cells, then the notion of such a finite hematopoietic episode is logical because the yolk sac later disappears. If, however, the AGM is the true source, then there is no real evidence to indicate that the initiation of hematopoiesis is necessarily a process of finite duration. There is now good molecular evidence that local cells of a somewhat undefined source are induced to differentiate into HSC by local signals in the AGM [5]. If some of these future HSC are of intraembryonic origin, the generation de novo of hematopoietic cells might continue throughout life if required by the circumstances. For that matter, the AGM need not continue to be the actual location of such hematogenesis. It is a familiar observation in transplanted lethally irradiated mice that some cells in the regenerating hematopoietic tissue are of host, not donor, origin. Usually, these have been regarded as clonogenic cells that have survived irradiation, but this cannot be distinguished from the alternative that de novo formation of stem cells has occurred. The question needs re-examination whenever suitable tools arise and is a question needing to be posed for all types of tissue stem cells—episodic generation or continuous generation?

Much has been made of the obvious differences between embryonic (nucleated) erythroid cells of yolk sac origin and definitive (non-nucleated) erythroid cells of the AGM/fetal liver origin. However, the studies of Keller et al. on clonally derived embryoid bodies have shown that both types of erythroid cell can have the same ancestral embryonic stem cell [6]. Similarly, transplantation studies of Turpen et al. in frog eggs showed that the erythroid cell phenotype of cells originating in the ventral marginal zone will differ according to whether the cells are grafted to the dorsal lateral plate (definitive erythropoiesis) or the ventral blood island region (primitive erythropoiesis) [7]. These studies indicate that the phenotype and properties of stem cells can be directed by external signals that can differ according to location. The same biological processes could well change the phenotype and properties of stem cells for other tissues.

On the question of the type of HSC present in the yolk sac, the elegant studies of Yoder and Hiatt [8] documented the occurrence of major errors if standard irradiated recipients are used as the assay system. With this latter assay, the yolk sac appears to contain no cells capable of long-term repopulation of the recipients. However, if yolk sac cells are engrafted into milder pretreated neonatal recipients, hematopoietic cells of yolk sac origin develop in the adult animal that are readily able to engraft lethally irradiated recipients. The irradiated recipient assay clearly was too demanding to allow such yolk sac stem cells to reveal themselves.

The Bone Marrow Origin of Gonadal Stem Cells—A Sterile Aside

  1. Top of page
  2. Abstract
  3. Introduction
  4. The Hematopoietic Family Tree
  5. The Origin of Stem Cells—A Single Episode or Continuous Generation?
  6. The Bone Marrow Origin of Gonadal Stem Cells—A Sterile Aside
  7. Self-Generation by Stem and Other Cells
  8. Blast Colony-Forming Cells
  9. Lineage-Committed Colony- Forming Cells
  10. The Behavior of Supposedly Mature End Cells
  11. Conclusions from the Reculturing Experiments
  12. Can the Pattern of Self-Generation and Commitment Be Manipulated?
  13. General Conclusions
  14. Disclosure of Potential Conflicts of Interest
  15. Acknowledgements
  16. References

When initial experiments suggested the possible presence of a miscellany of organ-specific stem cells in marrow, we were encouraged to explore the possibility that marrow populations might contain gonadal stem cells. The seminiferous tubules contain quasi clonal populations generating mature spermatozoa (Fig. 3A). After lethal irradiation (2 × 5.5 Gy) and marrow transplantation, many seminiferous tubules remain empty of spermatozoa, but some (Fig. 3B) display normal spermatogenesis. Could such populations have been produced by the injected marrow cells used to restore hematopoiesis? We had generated in C57BL mice two lines of knockout mice (SOCS-5−/− and Asb-1−/− mice) in which spermatogenesis and fertility were normal but, because the lacZ gene had been knocked in during gene deletion, in both animals spermatogenic cells were able to be stained intensely blue by lacZ staining [9, 10] (Fig. 3C). In the experiments, adult male C57BL mice were lethally irradiated then injected with 2 × 106 marrow cells from either SOCS-5−/− or Asb-1−/− mice. Two months later, the testes of all irradiated recipient mice were examined for lacZ-positive spermatocytes. As shown in Figure 3D, no lacZ-positive cells were observed, from which it could be concluded that the intriguing appearance of regenerating spermatocytes in irradiated mice must be due to surviving gonadal stem cells and that marrow populations do not generate such repopulating spermatogenic cells. Other explanations for the negative results are possible, but we decided to return to hematopoietic populations.

thumbnail image

Figure Figure 3.. Bone marrow cells fail to form spermatocytes. (A): A section of mouse testis with seminiferous tubules filled with maturing spermatozoa. (B): A testis after whole body irradiation and bone marrow transplantation. Note that one tubule contains a spermatogenic population but that other tubules are empty. (C): LacZ-staining of spermatozoa in a seminiferous tubule of a SOCS-5−/− mouse. (D): Absence of any LacZ-positive spermatozoa in the testis of an irradiated mouse whose hematopoietic tissues were repopulated 2 months previously by the injection of SOCS-5−/− bone marrow cells.

Download figure to PowerPoint

Self-Generation by Stem and Other Cells

  1. Top of page
  2. Abstract
  3. Introduction
  4. The Hematopoietic Family Tree
  5. The Origin of Stem Cells—A Single Episode or Continuous Generation?
  6. The Bone Marrow Origin of Gonadal Stem Cells—A Sterile Aside
  7. Self-Generation by Stem and Other Cells
  8. Blast Colony-Forming Cells
  9. Lineage-Committed Colony- Forming Cells
  10. The Behavior of Supposedly Mature End Cells
  11. Conclusions from the Reculturing Experiments
  12. Can the Pattern of Self-Generation and Commitment Be Manipulated?
  13. General Conclusions
  14. Disclosure of Potential Conflicts of Interest
  15. Acknowledgements
  16. References

Stringent tests on the properties of HSC and their progeny really require protocols in which the performance of individual cells can be monitored and, ideally, in which the fate of individual daughter cells can also be monitored clonally. The current gold standard definition of an HSC is that it is a cell capable of generating long-term repopulation by cells in all lineages in a lethally irradiated recipient. Despite the use of limit dilution protocols, this is not a situation allowing individual stem cells or their immediate progeny to be monitored clonally. Extensive fluorescence-activated cell sorting (FACS) can achieve a stem cell population of uniformly small mononuclear cells with little cytoplasmic RNA. Evidence differs on whether such populations are homogeneous or remain a heterogeneous population. Most would say the population remains heterogeneous, that it might possibly even contain some totipotential cells, but is certainly heterogeneous in the proliferative potential of individual cells. Others, however, have claimed that all cells in such preparations are repopulating stem cells [11].

An early progeny cell of HSC is the cell forming large (107 cells) hematopoietic colonies in the spleens of irradiated mice (CFU-S). Such colonies themselves contain CFU-S, indicating a capacity for self-generation on the part of CFU-S, and certainly there is also multilineage potentiality. However, CFU-S are now regarded as not having true long-term repopulating capacity. Furthermore, when colony-to-colony passage of CFU-S was undertaken, self-generation ceased with adult marrow-derived cells after three sequential passages [12]. More recent studies using long-term repopulation as the readout, and therefore monitoring stem cell self-generation, have also documented an inability of repopulating cells from adult marrow to undergo more than three sequential transplantation cycles (see for example [13]). On these grounds, both HSC and CFU-S have self-generative capacity, but their self-generative capacity is quite restricted. No hematopoietic stem cell has the unlimited capacity for self-generation possessed by embryonic stem cells.

FACS separation allows three populations of bi- or possibly multipotential precursor cells to be segregated—common myeloid progenitors (CMP), granulocyte-macrophage progenitors (GMP), and megakaryocyte-erythroid progenitors (MEP)—with some evidence suggesting that CMP can generate the more restricted GMP and MEP precursors [14]. The exact cellular origin of each of these populations remains unclear. Clonal cultures of each population show that only the CMP population contains cells able to form blast colonies [15] (Table 1). Blast colonies are multicentric colonies stimulated to develop by stem cell factor plus granulocyte colony-stimulating factor or interleukin (IL)-3 in this laboratory that are analyzed after 7 days of incubation. They appear to contain a uniform population of blast cells, but many of these cells are in fact lineage-committed progenitor cells.

Table Table 1.. Colony formation by fractionated myeloid progenitor cells
Thumbnail image of

Blast Colony-Forming Cells

  1. Top of page
  2. Abstract
  3. Introduction
  4. The Hematopoietic Family Tree
  5. The Origin of Stem Cells—A Single Episode or Continuous Generation?
  6. The Bone Marrow Origin of Gonadal Stem Cells—A Sterile Aside
  7. Self-Generation by Stem and Other Cells
  8. Blast Colony-Forming Cells
  9. Lineage-Committed Colony- Forming Cells
  10. The Behavior of Supposedly Mature End Cells
  11. Conclusions from the Reculturing Experiments
  12. Can the Pattern of Self-Generation and Commitment Be Manipulated?
  13. General Conclusions
  14. Disclosure of Potential Conflicts of Interest
  15. Acknowledgements
  16. References

Blast colonies not only contain lineage-committed colony-forming cells but also blast colony-forming cells among their progeny. These blast colony-forming cells are therefore the most immature cell type able to be analyzed for self-generation in clonal cultures. Data from an analysis of this question are shown in Table 2. In these experiments, 25 randomly chosen blast colonies grown using IL-3 were analyzed from cultures of (C57BL × SJL) F1 marrow cells. Each colony was resuspended and recultured in secondary cultures again containing IL-3 as the stimulus. Table 2 shows the calculated total number of progeny colonies formed by the 25 recultured blast colony cells. In the line above, colony numbers in primary donor cultures stimulated by IL-3 have been adjusted so that blast colony numbers in the cultures total 25. This manner of expressing the results allows a direct comparison of ability of the clonogenic cells generated in the secondary cultures to replace the clonogenic cells present in the primary cultures. It can be seen that the 25 blast colonies produced 64 daughter blast colony-forming cells plus more than 20,000 lineage-committed colony-forming cells—many more than needed to replace the 144 in the primary cultures. The second panel in the table shows data from a similar set of recultures of 14 blast colonies grown from DBA/2 marrow cells. Here, the self-generating capacity of blast colony-forming cells is even more dramatically demonstrated. The 14 primary blast colonies contained a total of 120 cells able to form blast colonies identical in size to the primary colonies. By contrast, a similar experiment using 10 BALB/c blast colonies failed to detect any daughter blast colony-forming cells.

Table Table 2.. Excess production of lineage committed progenitor cells by blast colony-forming cells
Thumbnail image of

These results are therefore somewhat conflicting but, at least for (C57BL × SJL) F1 and DBA/2 marrow cells, blast colony-forming cells demonstrated a notable capacity for self-generation plus an ability to form massive numbers of committed progenitor cells. For cells in the granulocytic, macrophage, eosinophil, and megakaryocyte lineages, the data imply that the blast colony-forming cells are perfectly capable of sustaining cell formation without cellular input from more ancestral HSC or CFU-S.

Lineage-Committed Colony- Forming Cells

  1. Top of page
  2. Abstract
  3. Introduction
  4. The Hematopoietic Family Tree
  5. The Origin of Stem Cells—A Single Episode or Continuous Generation?
  6. The Bone Marrow Origin of Gonadal Stem Cells—A Sterile Aside
  7. Self-Generation by Stem and Other Cells
  8. Blast Colony-Forming Cells
  9. Lineage-Committed Colony- Forming Cells
  10. The Behavior of Supposedly Mature End Cells
  11. Conclusions from the Reculturing Experiments
  12. Can the Pattern of Self-Generation and Commitment Be Manipulated?
  13. General Conclusions
  14. Disclosure of Potential Conflicts of Interest
  15. Acknowledgements
  16. References

When this type of recloning experiment was repeated with lineage-committed granulocytic, granulocyte-macrophage, or macrophage colonies grown using granulocyte macrophage–colony-stimulating factor (GM-CSF) and with GM-CSF, macrophage colony-stimulating factor, or stem cell factor (SCF)+IL-3+erythropoietin (EPO) used in the secondary cultures, a very different situation was observed. Using C57BL marrow cultures, recloning of 115 granulocytic colonies failed to detect a single granulocytic or granulocyte-macrophage colony in the readout secondary cultures. What did develop with 10% of the recultured colonies were small numbers of macrophage colonies barely large enough to be classified as colonies. A similar outcome was observed with 71 recultured granulocyte-macrophage colonies. Again, there were no secondary granulocytic colonies, but 55% of the colonies contained cells able to form small macrophage clones. Not surprisingly, of 111 macrophage colonies recultured, no secondary granulocytic colonies were observed, but 70%–94% of the secondary cultures contained small macrophage colonies.

In all these sets of experiments, the secondary macrophage colonies were of small size and, on further reculture, failed to form clonal progeny of large enough size to be scorable as colonies. The finding of cells able to form small macrophage clones in recultured granulocytic colonies was surprising because such primary colonies appeared to contain only granulocytic cells. However, the secondary culture of areas of colony-free agar medium adjacent to granulocytic colonies generated equivalent numbers of small macrophage colony/clones, and it is apparent that the result with the recultured granulocytic colonies was based on contaminating clonogenic macrophage precursors present, possibly merely as single cells, in the medium being occupied by the expanding granulocytic clones.

In these recloning experiments, occasional eosinophil colonies (that have the same general size and morphology as neutrophil colonies) were accidentally recultured. In this case, the secondary cultures invariably contained small numbers of eosinophil colonies of a size equal to that of the primary colonies.

These recloning experiments using colonies formed by lineage-committed progenitors allowed several conclusions. Granulocyte-committed progenitors have no capacity for self-generation, whereas eosinophil-committed progenitors do. Macrophage-committed progenitors have some capacity to form clonogenic progeny, but this capacity is restricted and is lost on secondary reculture.

The Behavior of Supposedly Mature End Cells

  1. Top of page
  2. Abstract
  3. Introduction
  4. The Hematopoietic Family Tree
  5. The Origin of Stem Cells—A Single Episode or Continuous Generation?
  6. The Bone Marrow Origin of Gonadal Stem Cells—A Sterile Aside
  7. Self-Generation by Stem and Other Cells
  8. Blast Colony-Forming Cells
  9. Lineage-Committed Colony- Forming Cells
  10. The Behavior of Supposedly Mature End Cells
  11. Conclusions from the Reculturing Experiments
  12. Can the Pattern of Self-Generation and Commitment Be Manipulated?
  13. General Conclusions
  14. Disclosure of Potential Conflicts of Interest
  15. Acknowledgements
  16. References

The above conclusions appear to complete the story for the lineage-committed lines in the hematopoietic family tree, but with three lineages a much less clear situation exists. T and B lymphocytes both have a capacity to generate memory cells, which have a significant capacity for generating progeny populations when restimulated. Can such cells be regarded as quasi stem cells—the only qualification being that presumably they are not multipotential? They do meet part of the definition of stem cells—the more so if multipotentiality is not required, as may be the case for many tissue stem cells.

An important population needing to be considered is the mast cell population. Mast cell progenitors do exist in the bone marrow and can readily be grown in agar cultures using a combination of SCF+IL-3, with or without EPO. These mast cell colonies can readily be recloned and produce secondary colonies identical to the primary colonies. From such colonies it is very easy to derive continuous cloned cell lines using SCF+IL-3 as the stimulus, and in our laboratory such cloned cell lines have been maintained for at least 18 months with no loss of proliferative potential. At all times, the cultures contain mast colony-forming cells. Mast cells have been a difficult population to interpret in vivo—precursor cells exist in the marrow, but peripheral mast cells in regions such as the skin or fatty tissue can appear largely independent of the clonogenic precursors in the marrow [16]. Are the cloned mast cell lines comparable with tissue stem cells, and how do they relate to such cells?

Conclusions from the Reculturing Experiments

  1. Top of page
  2. Abstract
  3. Introduction
  4. The Hematopoietic Family Tree
  5. The Origin of Stem Cells—A Single Episode or Continuous Generation?
  6. The Bone Marrow Origin of Gonadal Stem Cells—A Sterile Aside
  7. Self-Generation by Stem and Other Cells
  8. Blast Colony-Forming Cells
  9. Lineage-Committed Colony- Forming Cells
  10. The Behavior of Supposedly Mature End Cells
  11. Conclusions from the Reculturing Experiments
  12. Can the Pattern of Self-Generation and Commitment Be Manipulated?
  13. General Conclusions
  14. Disclosure of Potential Conflicts of Interest
  15. Acknowledgements
  16. References

With the notable exceptions of T and B lymphocytes, mast cells, and eosinophil progenitors, any capacity for self-generation appears to be lost at the committed progenitor cell stage. The family tree in Figure 1 with the above qualifications therefore appears to describe hematopoiesis. However, the capacity for substantial self-generation by CFU-S and by blast colony-forming cells would argue that the figure may well not properly represent hematopoiesis in normal life if in fact self-maintenance can be sustained by cells at the CFU-S/blast colony-forming cell level. Only in extreme emergency situations such as following whole body irradiation or intensive chemotherapy might genuine multipotential stem cells need to be activated to generate progeny continuously, as is implied in Figure 1. There is, therefore, a need for caution in basing hematopoietic models on cell behavior in lethally irradiated mice, and alternative models considering also the role of cell cycle status and using data from engrafted normal recipients need to be considered [17].

Can the Pattern of Self-Generation and Commitment Be Manipulated?

  1. Top of page
  2. Abstract
  3. Introduction
  4. The Hematopoietic Family Tree
  5. The Origin of Stem Cells—A Single Episode or Continuous Generation?
  6. The Bone Marrow Origin of Gonadal Stem Cells—A Sterile Aside
  7. Self-Generation by Stem and Other Cells
  8. Blast Colony-Forming Cells
  9. Lineage-Committed Colony- Forming Cells
  10. The Behavior of Supposedly Mature End Cells
  11. Conclusions from the Reculturing Experiments
  12. Can the Pattern of Self-Generation and Commitment Be Manipulated?
  13. General Conclusions
  14. Disclosure of Potential Conflicts of Interest
  15. Acknowledgements
  16. References

Dogma holds that the pattern of proliferation and lineage commitment in hematopoietic cells is fixed and irreversible. However, much experimental literature shows clearly that this is not so.

For example, if marrow cells are transfected and overexpress certain homeobox genes such as HOXB6 or HOXB4, cells that appear to be committed progenitor cells can acquire a capacity for extended self-generation, if not immortalization [18, 19]. Many similar experiments allow the same conclusion. These are nonphysiological experiments, but they do show that the pattern of progressive reduction of self-generation can be reversed. If this is so with nonphysiological concentrations of these homeobox gene products, it can be argued that similar events might be possible, if infrequently, with physiological levels of these gene products if appropriate emergency situations arise.

Similar comments can be made regarding lineage commitment. Again, by way of a representative example, overexpression of GATA-1 in common lymphoid progenitors leads to the development of megakaryocyte-erythroid precursors, as does overexpression of GATA-1 in GMP populations [20]. Here, lineage commitment has been switched in a dramatic way, albeit again by using nonphysiological concentrations of a normal cellular transcription factor. Again, do these experiments tell us not only that lineage switching is possible but also that similar switches could sometimes occur under physiological conditions?

General Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. The Hematopoietic Family Tree
  5. The Origin of Stem Cells—A Single Episode or Continuous Generation?
  6. The Bone Marrow Origin of Gonadal Stem Cells—A Sterile Aside
  7. Self-Generation by Stem and Other Cells
  8. Blast Colony-Forming Cells
  9. Lineage-Committed Colony- Forming Cells
  10. The Behavior of Supposedly Mature End Cells
  11. Conclusions from the Reculturing Experiments
  12. Can the Pattern of Self-Generation and Commitment Be Manipulated?
  13. General Conclusions
  14. Disclosure of Potential Conflicts of Interest
  15. Acknowledgements
  16. References

The family tree of Figure 1 has served hematologists well for more than a century and experimentally for approximately half a century. More recent experiments that raise doubts regarding its exact form and inflexibility do not really destroy its broad concepts. These recent experiments do, however, demand a preparedness to accommodate a much more flexible behavior pattern in this remarkable population.

What lessons does the hematopoietic population have for those working on the corresponding hierarchical structure of stem cells in other tissues? In hematopoietic populations, multipotential stem cells do exist but have a restricted capacity for self-generation and may not participate in everyday cell formation. Lineage-committed cells exist with limited multipotentiality that again have some self-generative capacity but again of a restricted nature. Clonogenic cells exist with no capacity for self-generation and, finally, some apparently mature cells can have substantial self-generative capacity. All of these stem cell or clonogenic cell types can be predicted to occur in other organ systems, and this requires a careful analysis of the properties of any cell type with some claim to being a tissue stem cell.

Finally, plasticity is now recognized to be a cellular variable both for proliferative and lineage-commitment parameters. Are these properties capable of being manipulated under physiological or disease conditions, or are they extreme experimental artifacts? There is, indeed, no shortage of intriguing questions to be resolved.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. The Hematopoietic Family Tree
  5. The Origin of Stem Cells—A Single Episode or Continuous Generation?
  6. The Bone Marrow Origin of Gonadal Stem Cells—A Sterile Aside
  7. Self-Generation by Stem and Other Cells
  8. Blast Colony-Forming Cells
  9. Lineage-Committed Colony- Forming Cells
  10. The Behavior of Supposedly Mature End Cells
  11. Conclusions from the Reculturing Experiments
  12. Can the Pattern of Self-Generation and Commitment Be Manipulated?
  13. General Conclusions
  14. Disclosure of Potential Conflicts of Interest
  15. Acknowledgements
  16. References

The work from the author's laboratory referred to in this review was supported by grants from the Cancer Council Victoria, the National Health and Medical Research Council, Canberra, and the National Institutes of Health, Bethesda, Grant number CA22556. This work is based on the paper “Stem Cells and Hematopoietic Stem Cells” delivered on June 20, 2007, at the Fifth Annual Meeting of the International Society for Stem Cell Research held in Cairns, Australia.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. The Hematopoietic Family Tree
  5. The Origin of Stem Cells—A Single Episode or Continuous Generation?
  6. The Bone Marrow Origin of Gonadal Stem Cells—A Sterile Aside
  7. Self-Generation by Stem and Other Cells
  8. Blast Colony-Forming Cells
  9. Lineage-Committed Colony- Forming Cells
  10. The Behavior of Supposedly Mature End Cells
  11. Conclusions from the Reculturing Experiments
  12. Can the Pattern of Self-Generation and Commitment Be Manipulated?
  13. General Conclusions
  14. Disclosure of Potential Conflicts of Interest
  15. Acknowledgements
  16. References