Nature Immunology
- 7, 1129 - 1130 (2006)
doi:10.1038/ni1106-1129
The unpredictable stem cellGay M Crooks1 & Kenneth Weinberg21 Gay M. Crooks is in the Division of Research Immunology and Bone Marrow Transplantation, Childrens Hospital Los Angeles, University of Southern California, Los Angeles, California 90027, USA. gcrooks@chla.usc.edu 2 Kenneth Weinberg is in the Division of Stem Cell Transplantation, Department of Pediatrics, Lucile Packard Children's Hospital, Stanford University School of Medicine, Stanford, California, 94305, USA. How individual hematopoietic stem cells contribute to blood cell formation throughout a lifetime has remained a subject of debate. A new analysis suggests there is substantial variation in hematopoietic stem cell fate and self-renewal activity.The existence of hematopoietic stem cells (HSCs) was established nearly 60 years ago in seminal experiments demonstrating the existence of clonogenic marrow cells with the ability to regenerate the entire blood-forming system of lethally irradiated mice1. Justifiably, that work has been called transformative. The study of hematopoiesis changed from a descriptive to an experimental science, stem cell biology emerged from embryology and a new clinical discipline of hematopoietic stem cell transplantation was born. The 'Holy Grail' sought in those endeavors was the rare and elusive HSC. Subsequent studies have focused on methods for identifying, purifying, functionally assessing, and experimentally and therapeutically manipulating first mouse and then human HSCs. Three key steps in the process of identifying the human HSC were the recognition that the cell surface glycoprotein CD34 is expressed by immature human hematopoietic progenitors, including HSCs; the general validation of the postulate that HSCs are lineage negative (Lin-) and thus do not express lineage-specific or lineage-associated surface antigens characteristic of differentiated blood cells; and the development of in vivo assays for human stem cell activity in immunodeficient mice2,
3,
4,
5,
6. In this issue of Nature Immunology, McKenzie et al. have brought together the entire range of experimental methods for study of HSCs to elegantly and meticulously address key issues of how human HSCs act in vivo7. In so doing, they raise the issue of whether the HSC is a unique and finite cell type or a compartment of cells comprising a spectrum of interchangeable states.
Over the decades since HSCs were first identified, two general ideas have been proposed to explain the process by which these cells provide a stable yet regulated reservoir for a lifetime of hematopoiesis. The 'clonal succession' model states that small numbers of quiescent stem cells are sequentially recruited into the cell cycle and then progressively divide and differentiate until they are exhausted and ultimately are replaced by the next cohort of activated stem cells (Fig. 1). However, elegant studies in mice8 and larger animals9 have shown that although multiple clones contribute to engraftment early after HSC transplantation, long-term hematopoiesis is sustained by a limited number of self-renewing stem cells. Using lentiviral vectors to track the clonal ancestry of human cells after serial transplantation, McKenzie et al. provide evidence that the latter model of 'clonal maintenance' also seems to exist in humans7. Notably, and in contrast to mouse studies, in addition to stable clones that persisted throughout the 7 months of analysis, other clones were identified that fluctuated in their contribution to hematopoiesis over time and were not always active during early engraftment. Some caution, however, should be used in attributing those data to a difference between human and mouse hematopoietic regulation. The human HSCs in this study were isolated from umbilical cord blood, and there are inherent differences in the cell cycling and generative capacity of human marrow and cord blood stem cells that provide different patterns of engraftment and potentially self-renewal10.
 | | |
The McKenzie et al. study supports the idea that the fate of stem cells to self-renew or differentiate is achieved as a stochastic process affected by chance interactions with the environment or random signals generated in the cell itself. That conclusion is based on the assumption that the immunophenotype used defines a uniform population of stem cells whose intrinsic self-renewal capacity is patterned only after transplantation. In this study, only 25–30% of the CD34+Lin-CD38- cells that repopulated mice with severe combined immunodeficiency showed evidence of self-renewal, and among those, the pattern of activation varied widely and unpredictably. Many other studies have also found functional heterogeneity in stringently defined immunophenotypic populations, but that still raises the issue of whether the heterogeneity is a result of different types of cells with a common immunophenotype or of random events that affect intrinsically identical or very similar cells. Thus, it is not possible to absolutely disprove the alternative 'deterministic' theory, in which stem cell fate is predetermined by the fixed intrinsic biology of each cell11. Perhaps the strongest evidence for the stochastic mechanism in this study is the finding that clonally related daughter cells that migrate to different areas in the skeleton and retain self-renewal capacity nonetheless demonstrate very different patterns of activation.
Although the use of lentivirally transduced cells and immunodeficient mice represents the best method available for studying the clonality of human HSCs, results from two clinical gene therapy trials point out the potential pitfalls of that approach. An important assumption in 'clonal marking' by analysis of viral integration sites is that the generative potential and differentiation of the transduced cells is not biased by the processes of culture or transduction. In a clinical study of retroviral gene therapy for X-linked chronic granulomatous disease, insertional activation of several genes that regulate hematopoiesis caused expansion of the transduced clones and increased myelopoiesis12. Similar results have been obtained for mouse hematopoiesis13. In the paper by McKenzie et al. here7, transduction did not seem to induce bias in the hematopoietic lineages generated from each clone, suggesting that at least differentiation was not affected by lentiviral marking. Another caveat relates to the definition of long-term hematopoiesis. The nonobese diabetic–severe combined immunodeficiency mouse model used here allows analysis of hematopoiesis for only approximately 6 months. Although that is relatively long-term for a mouse, it represents a small fraction of both a human lifespan and the predicted cycling time for an individual HSC in large animals9. In another clinical gene therapy study of retroviral transduction of CD34+ cells from children with adenosine deaminase deficiency, a single clone was found to contribute to hematopoiesis in one recipient for at least 8 years, but other clones had a much shorter duration of activity14. Notably, feline studies have shown that clonal stability is not achieved until 1–4.5 years after transplantation9, much longer than in the study here with human cells7. It is not known if and how the kinetics of human HSCs are altered by the mouse environment, but caution should be taken in concluding that 6 months of HSC activity from individual clones represents long-term stability.
Another fundamental issue relates to the extrapolation from transplantation models to stem cell activity during steady-state hematopoiesis. The use of serial transplantation is an essential tool for measuring the self-renewal capacity of stem cell clones. Nonetheless, in this artificial setting, the stress of serial transplantation could theoretically induce long-term changes in the biological activity of stem cells, providing selection bias toward those with particular self-renewal patterns.
For 60 years, the HSC has been the methodological model for studies of postnatal stem cell populations. The considerable heterogeneity of human HSCs described by McKenzie et al. is probably a general phenomenon of other stem cell populations. Although genetic models have defined necessary pathways for self-renewal, other approaches will be needed to understand the subtle differences that result in the observed variability in stem cell compartments15. As noted by the authors, understanding of the heterogeneity of HSC (or other stem cell) populations is unlikely to be achieved by static analyses of pooled cells. What is required are refinements in the techniques for identifying HSC subsets as well as new tools for dynamic analyses of the activation state of individual stem cells and their intercellular interactions, while preserving functional potential.
REFERENCES
-
Till, J.E.
&
McCulloch, E.A.
Radiat. Res. 18, 96–105 (1963). | PubMed | ISI | ChemPort |
-
Civin, C.I.
et al. J. Immunol. 133, 157–165 (1984). | PubMed | ISI | ChemPort |
-
Spangrude, G.J.
,
Heimfeld, S.
&
Weissman, I.L.
Science 241, 58–62 (1988). | PubMed | ISI | ChemPort |
-
Terstappen, L.W.
,
Huang, S.
,
Safford, M.
,
Lansdorp, P.M.
&
Loken, M.R.
Blood 77, 1218–1227 (1991). | PubMed | ISI | ChemPort |
-
Conneally, E.
,
Cashman, J.
,
Petzer, A.
&
Eaves, C.
Proc. Natl. Acad. Sci. USA 94, 9836–9841 (1997). | Article | PubMed | ChemPort |
-
Gan, O.I.
,
Murdoch, B.
,
Larochelle, A.
&
Dick, J.E.
Blood 90, 641–650 (1997). | PubMed | ChemPort |
-
McKenzie, J.L.
,
Gan, O.I.
,
Doeden, M.
,
Wang, J.C.Y.
&
Dick, J.E.
Nat. Immunol. 7, 1225–1233 (2006). | Article |
-
Jordan, C.T.
&
Lemischka, I.R.
Genes Dev. 4, 220–223 (1990). | PubMed | ISI | ChemPort |
-
Abkowitz, J.L.
et al. Proc. Natl. Acad. Sci. USA 92, 2031–2035 (1995). | Article | PubMed | ChemPort |
-
Hao, Q.L.
,
Shah, A.J.
,
Thiemann, F.T.
,
Smogorzewska, E.M.
&
Crooks, G.M.
Blood 86, 3745–3753 (1995). | PubMed | ISI | ChemPort |
-
Morrison, S.
&
Weissman, I.L.
Immunity 1, 661–673 (1994). | Article | PubMed | ISI | ChemPort |
-
Ott, M.G.
et al. Nat. Med. 12, 401–409 (2006). | Article | PubMed | ChemPort |
-
Kustikova, O.
et al. Science 308, 1171–1174 (2005). | Article | PubMed | ISI | ChemPort |
-
Schmidt, M.
et al. Nat. Med. 9, 463–468 (2003). | Article | PubMed | ISI | ChemPort |
-
Park, I.K.
et al. Nature 423, 302–305 (2003). | Article | PubMed | ISI | ChemPort |
|
